University of Southern California Dissertations and Theses

To what extent can wicked problems of health and place be addressed through clumsy solutions? Three experiments in reconstituting networks of spatial knowledge production

(USC Thesis Other)

To what extent can wicked problems of health and place be addressed through clumsy solutions? Three experiments in reconstituting networks of spatial knowledge production

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content To what extent can wicked problems of health and place be addressed through clumsy solutions?
Three experiments in reconstituting networks ofspatial knowledge production.
by
Leo Lerner
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(POPULATION, HEALTH AND PLACE)
May 2024
Copyright © 2024 Leo Lerner

ii
Acknowledgements
I am extremely grateful to the participants who contributed their time and amity. Their
willingness to share stories and talk about their lives made this project possible. I would also like
to thank the community partners in El Sereno, Lincoln Heights, and Boyle Heights, and the
directors who took the time to listen to my plans and put me on the straight and narrow.
My development as a doctoral candidate was made possible by the support of my cochairs Drs. Lourdes Baezconde-Garbanati and John Wilson. Thank you both for your patience
and consistent attention over the past five years. I am grateful especially for the exposure to so
many amazing research teams during this time, including Stay Connected Los Angeles and Smart
and Connected Community Food Systems, which informed my thinking no end.
I would also like to thank my outstanding committee members. Thanks Dr. Robert Vos
for your joint supervision as committee member, service as Director of Graduate Studies,
brainstorming the structure and format of my dissertation, and including me in Community
Health Equity Solutions. Thanks Dr. Edgar Rivera Colón for helping me convert my English
theory into the American context, introducing me to the LA Tenant’s Union, and meeting up on
Zoom and in Mariachi Plaza when I needed to be steered back on course. Thanks to Dr. Juan De
Lara for your direction in Latinx geographies, and for your challenging questions during my
qualifying exam.
Thank you to the USC staff, without whom large parts of this project would not have
been able to go ahead. In particular, thanks to Stephanie Tran in the Spatial Sciences Institute
and Rosa Barahona in the Center for Health Equity in the Americas.

iii
I am grateful to my tutor at Lancaster, Dr. Duncan Whyatt, for teaching me GIS, and
teaching it well. The practices you taught me are in everything that followed. I am also grateful
to my tutors at Oxford, especially Professors Derek McCormack, Gillian Rose, Tim Schwanen,
and Catharina Landström. What I learnt from you is now my way of life.
Last but by no means least, thank you to the others who completed or are completing
PhDs at the same time as me. Drs. Avery Everhart and Bita Minaravesh were willing to answer
any question at any time of day, for which I am eternally indebted. My Population, Health and
Place cohort showed great resilience, staying strong during the strange times. Good luck to the
six of us. Thanks finally to my compatriots and best friends Alex, Emilie, and William, whose
kind spirits make anything possible.

iv
Table of Contents
Acknowledgements......................................................................................................................... ii
List of Tables ................................................................................................................................. ix
List of Figures................................................................................................................................. x
Abbreviations................................................................................................................................ xii
Abstract........................................................................................................................................ xiii
Chapter 1 Introduction .................................................................................................................... 1
1.1. Situated knowledges ...........................................................................................................1
1.2. Place....................................................................................................................................3
1.3. Health and place..................................................................................................................5
1.4. Expertise and participation .................................................................................................7
1.5. Knowledge controversies, wicked problems, crisis............................................................9
1.6. Participatory GI Science ...................................................................................................11
1.6.1. Participatory GIS for food .......................................................................................12
1.6.2. Public participatory GIS for mobilities....................................................................13
1.6.3. Volunteered geographic information for housing....................................................14
1.7. Significance of the dissertation.........................................................................................14
1.8. Structure............................................................................................................................17
Vignette 1 Ave. 26 Night Market ................................................................................................. 18
Chapter 2 Structurally competent food systems science in the city: Reflections on a goldstandard food environment audit in five neighborhoods on the Eastside of Los Angeles.... 23
Abstract....................................................................................................................................23
2.1. Introduction.......................................................................................................................24
2.1.1. Food deserts.............................................................................................................27
2.1.2. The role of data ........................................................................................................30

v
2.1.3. Mixing methods in food research ............................................................................32
2.1.4. The NEMS-S............................................................................................................34
2.2. Methods.............................................................................................................................36
2.2.1. Study Area ...............................................................................................................36
2.2.2. Identifying outlets....................................................................................................39
2.2.3. Reduced item food audit ..........................................................................................40
2.2.4. Additions to NEMS-S..............................................................................................42
2.3. Results...............................................................................................................................46
2.3.1. Outlet type................................................................................................................46
2.3.2. Reduced item food audit ..........................................................................................47
2.3.3. Benefits....................................................................................................................48
2.3.4. Sample Prices...........................................................................................................49
2.3.5. LILA Analysis .........................................................................................................49
2.3.6. Kernel density analysis............................................................................................53
2.3.7. Service Areas...........................................................................................................56
2.4. Discussion.........................................................................................................................60
2.4.1. Efficiency of the audit tool and comparison with existing research........................60
2.4.2. Adapting the NEMS-S and implications for future research...................................61
2.4.3. Alternative food networks........................................................................................62
2.4.4. Food assistance programs........................................................................................63
2.4.5. Temporal resolution.................................................................................................63
2.4.6. Culturally appropriate food......................................................................................64
2.4.7. 15-minute cities........................................................................................................65
2.4.8. Relational and affective approaches to food environment studies...........................66
2.4.9. Food environment-health behavior nexus................................................................68

vi
2.5. Conclusion ........................................................................................................................69
Vignette 2 Buses and neighbors.................................................................................................... 74
Chapter 3 Affective geographic information science: Using carto-photo elicitation interviews
to examine the connections between mobility and place amongst the carless in Los
Angeles.................................................................................................................................. 80
Abstract....................................................................................................................................80
3.1. Introduction.......................................................................................................................81
3.1.1. Visual methods.........................................................................................................82
3.1.2. Place.........................................................................................................................83
3.1.3. Spatio-visual methods..............................................................................................84
3.1.4. Contributions ...........................................................................................................87
3.2. Methodology.....................................................................................................................88
3.2.1. Study area.................................................................................................................88
3.2.2. Study population ......................................................................................................91
3.2.3. Practicalities and the initial meeting........................................................................93
3.2.4. Generating materials................................................................................................95
3.2.5. Carto Photo Elicitation Interview (CPEI)................................................................98
3.2.6. Analyzing the carto photo materials......................................................................101
3.3. Results and discussion ....................................................................................................103
3.3.1. Photo and carto elicitation .....................................................................................104
3.3.2. Interviews...............................................................................................................109
3.3.3. Mobility and place .................................................................................................111
3.4. Concluding remarks........................................................................................................151
Vignette 3 Verdemour Avenue ................................................................................................... 155
Chapter 4 Countermaps as resistance: Using volunteered geographic information (VGI) to
record experiences of housing insecurity on the Eastside of Los Angeles.......................... 163

vii
Abstract..................................................................................................................................163
4.1. Introduction.....................................................................................................................164
4.1.1. Housing, Place and Health.....................................................................................164
4.1.2. Countermapping.....................................................................................................167
4.1.3. Goals and objectives..............................................................................................172
4.2. Methods...........................................................................................................................173
4.2.1. Study area...............................................................................................................174
4.2.2. Study population and survey..................................................................................176
4.2.3. Deidentification......................................................................................................181
4.2.4. Data analysis..........................................................................................................183
4.2.5. User dashboard.......................................................................................................184
4.3. Results.............................................................................................................................184
4.3.1. Sample population .................................................................................................184
4.3.2. Housing history......................................................................................................186
4.3.3. Housing insecurity .................................................................................................187
4.3.4. Countermap dashboard ..........................................................................................196
4.4. Discussion.......................................................................................................................199
4.4.1. Privacy concerns....................................................................................................199
4.4.2. Moderating the countermap ...................................................................................201
4.4.3. Geographic context and spatial data quality ..........................................................204
4.4.4. Ownership..............................................................................................................206
4.4.5. Multidimensionality and the role of the locale ......................................................209
4.4.6. Who produces knowledge......................................................................................211
4.5. Conclusion ......................................................................................................................213
Chapter 5 Conclusion.................................................................................................................. 217

viii
5.1. Key findings....................................................................................................................218
5.2. Limitations and future work............................................................................................220
5.2.1. Small data for interdisciplinary learning................................................................220
5.2.2. Geographies of despair and new publics ...............................................................221
5.3. Summative reflection......................................................................................................223
5.4. Dissertation takeaway .....................................................................................................225
References................................................................................................................................... 226

ix
List of Tables
Table 2.1 – The 269 outlets on the Eastside which sell retail food ..............................................46
Table 2.2 – Percent of stores selling specific foods on the Eastside.............................................48
Table 2.3 – Price ranges of sampled food products on the Eastside.............................................49
Table 2.4 – Walking time service areas for four different types of outlets ..................................59
Table 3.1 – Demographic characteristics of study participants....................................................92
Table 3.2 – Section of Google Location History ...................................................................... 106
Table 4.1 – Demographic characteristics of the sample ............................................................ 185
Table 4.2 – Reasons why participants had moved home .......................................................... 187
Table 4.3 – Physical dimensions of inferior housing................................................................. 188
Table 4.4 – Landlord actions associated with housing insecurity ............................................. 189

x
List of Figures
Figure 2.1 – Study area, 1,737 identified food outlets and 364 selected outlets ..........................41
Figure 2.2 - 269 audited retail food outlets classified by LILA status .........................................51
Figure 2.3 - 36 audited supermarkets/large grocery stores classified by LILA status..................52
Figure 2.4 - Kernel density showing non-supermarket/large grocery store..................................54
Figure 2.5 - Kernel density showing supermarket/large grocery store outlets.............................55
Figure 2.6 - Service area map of walking time from all outlets ...................................................57
Figure 2.7 - Service area map of walking time for supermarkets/large grocery stores................58
Figure 3.1 – Participant image at a bus stop after a spiritual healing appointment ................... 100
Figure 3.2 - Google Location History of all participants........................................................... 107
Figure 3.3 - Google Location History of all participants, centered on Los Angeles................. 108
Figure 3.4 - Spatial clustering of Google Location History data............................................... 109
Figure 3.5 - Tree diagram for mobility themes.......................................................................... 110
Figure 3.6 - Tree diagram for place attachment themes ............................................................ 111
Figure 3.7 – CPEI extracts discussing financial reasons for carlessness................................... 113
Figure 3.8 – CPEI excerpts discussing participant attitudes to unhoused people...................... 126
Figure 3.9 – Participant image taken after administering Narcan ............................................. 127
Figure 3.10 – CPEI excerpts discussing smell on the bus ......................................................... 134
Figure 3.11 – CPEI excerpts on navigating the city after dark.................................................. 140
Figure 3.12 – Participant image showing the contrast in lighted spaces................................... 141
Figure 3.13 – CPEI excerpts on the relationships between safety and cleanliness ................... 143
Figure 3.14 – Participant image of abandoned buildings near a school .................................... 145
Figure 3.15 – CPEI excerpts on the story-telling properties of the materials............................ 149

xi
Figure 4.1 – Four conventional maps of Los Angeles............................................................... 168
Figure 4.2 – Smartphone preview of VGI survey...................................................................... 178
Figure 4.3 - The displacement method used for deidentification .............................................. 182
Figure 4.4 - How safe participants felt at each recorded location of housing insecurity........... 191
Figure 4.5 – Excerpts outlining monthly financial decisions .................................................... 192
Figure 4.6 – Excerpts outlining the stress and anxiety from housing insecurity ....................... 193
Figure 4.7 – Excerpts describing coping mechanisms for housing insecurity........................... 194
Figure 4.8 – Excerpts outlining a collective solidarity .............................................................. 195
Figure 4.9 – A prototype of an ArcGIS Dashboard for housing insecurity............................... 197
Figure 4.10 – Dashboard pop-up of an individual response ...................................................... 198

xii
Abbreviations
AFN Alternative Food Network
ANT Actor-Network Theory
CPEI Carto-Photo Elicitation Interview
DDHS Department of Health and Human Services
DMV Department of Motor Vehicles
DWG Data Working Group
EBT Electronic Benefit Transfer
GIS Geographic Information System
GLH Google Location History
LILA Low-Income Low-Access
MCU Measured Contextual Unit
NDSC Neighborhood Data for Social Change
NEMS-S Nutrition Environment Measures Survey in Stores
PGIS Participatory Geographic Information System
PPGIS Public Participatory Geographic Information System
SNAP Supplemental Nutrition Assistance Program
TCU True Contextual Unit
USC University of Southern California
USDA United States Department of Agriculture
VGI Volunteered Geographic Information

xiii
Abstract
Where scientific disputes become heated, generating wider social interest, knowledge
controversies emerge. This dissertation argues that these cannot be solved through the singleperspective knowledge politics dominant in conventional technoscience. This is especially the
case in the context of “wicked” problems, which are unique, persistent, insoluble, and
characterized by contradictory certitudes. The challenge for social research is to deliver
meaningful solutions which rekindle public confidence via different forms of engagement.
“Clumsy” solutions build on emergent and creative dialogue amongst an extended peer
community to incorporate many relevant worldviews, providing the kinds of noisy and
discordant evidence for decision-making which represents the multiple situated sites and forms
of expertise extant in the social world. The three empirical chapters each propose their own
clumsy solution to a specific context within health and place. Chapter 2 uses participatory
geographic information systems (PGIS) to challenge the capitalist production of food systems.
Chapter 3 uses public participatory GIS (PPGIS) to explore mobility practices in the
dispossessive infrastructures created by urban renewal. Chapter 4 employs volunteered
geographic information (VGI) to countermap housing insecurity, adopting the manners of
representation of the political elite to resist their historical use for enclosure and conquest.
Throughout, the role of expertise is examined for its capacity to generate new objects and sites of
disagreement which may lead to new political subjectivities and collectives. Overall,
participatory spatial science is commended as a powerful way to attach meaning to spatial
relations, as long as it is preceded by sophisticated methodological thought, from both the
participatory and the GIS sides of the elision. The findings suggest that clumsy solutions are a
viable response to wicked problems of knowledge, and that more theoretical and empirical
experimentation is needed to reconfigure expertise in health and social research.

1
Chapter 1 Introduction
There are inscriptions of inequality throughout Los Angeles, manifest in the socioenvironmental
exclusionary dynamics embedded into its everyday geographies. This introduction lays out a
conceptual framework for a dissertation which investigated the experience and materiality of
food, mobility, and housing in the uneven city. It explains the rationale behind the reimagination
of the knowledge politics attached to a set of conventional platforms and practices, deployed to
subvert the representational devices of contemporary health and social research.
1.1. Situated knowledges
At its core, this is a study into the role of situated knowledges in science. For a definition,
this dissertation uses Haraway’s (1988), rooted in the critique of the methodological pursuit of
universal truths. It emphasizes the role of geographic methods for deconstructing the truth status
of claims made from hostile technoscience which signify unmarked positions of masculine and
western through the attachment of the “objective” reference. The usual assumptions of scientific
method are tied to a historiography of capital, conquest, supremacy, and exploitation, distancing
subjects and subjectivity from the results to maintain an illusory power dynamic. By using
interpretive theories of culture (Geertz, 1973; Habermas, 1983), I aim to recast science and
method as tools for reconfiguring power, using the contradictions, connections, and unexpected
openings made possible through situated-ness to begin to understand the social world.
One such opening is made possible through the situation of this entire epistemological
enterprise. It takes place in the communities surrounding the University of Southern California
(USC) Health Sciences Campus on the Eastside of Los Angeles. Imagining a compass with the

2
campus at its center, four neighborhoods radiate out: El Sereno to the northeast; City Terrace to
the southeast; Boyle Heights to the southwest; and Lincoln Heights to the northwest. The
combined area of these four neighborhoods, referred hereon to as the collective “Eastside”, is the
highest proportion Hispanic and Latinx area in the country, home to the largest overall Mexicandescendent Chicanx population outside of Mexico (US Census Bureau, 2023). It lags in almost
every indicator of health and social equality in comparison to the remainder of the city,
particularly its geographic opposite, the majority white Westside, comprising neighborhoods like
Brentwood, Bel-Air, Beverly Hills, and Westwood (ibid, U.S. Census Bureau, 2022). The Health
Sciences Campus generates annual revenue in the billions of dollars (USC, 2019), bordering
census tracts with some of lowest life expectancies in the country (U.S. Census Bureau, 2022).
The adjacent poles of wealth and deprivation work on a prima facie relationship: the enduring
success and profit of the health-industrial complex rely on a ready availability of ill and broken
bodies, perpetuating the association between science and abandonment (Hyra et al., 2019;
Kaufman et al., 2019).
This is where I found myself for nearly five years, observing what Gilmore (2008) calls
the “organized abandonment” of neoliberalism. Where the Eastside used to be a site of vibrant
multiculturalism and radical political radicalism, with abundant public services (Sanchez, 2004),
the political elite guarded the systematic withdrawal of its communal resources, facilitating the
percolation of circulating capital into almost every daily action and interaction within the postFordist urban. As money and exchange value became the key focus of the capitalist city,
translation, convertibility, and universality – what Haraway (1988) calls “reductionism” –
became the main implication in scientific enquiry. I argue in the main chapters that this

3
reduction, and its according elimination of the particular and the local, is a gross disservice to the
populations on which it is inflicted.
1.2. Place
In particular, hierarchical positivism (here realist empiricism rather than logical) has
ordered what counts as knowledge. This is troubling in the social sciences, especially in light of
an ongoing aim to be considered the equal of the natural, physical, or medical sciences. Foucault
(1980) asks what we might fail to see in the elevation of reductionism and positivistic research
over an interpretivist worldview. The dominance of a reductive objectivism is what Foucault and
Haraway rail most strongly against. As one of the many discourses which constitutes knowledge
production, its attained dominance has drowned out competing positions (Cheek, 2007).
Amongst other outcomes, this has created epistemic gaps in research, producing different power
relations in different disciplines.
In human geography, concerned with the spatial relations which make up societal
meaning, the resolution to this has been the turn to place. This began during the cultural turn in
the late c. 20th, with the spatialization of concepts from philosophy and social theory. In a placebased worldview, humans are understood to depend on, identify with, and become attached to
the world around them (Brown et al., 2015). These connections, whether emotional, social,
physical, or phenomenological, constitute attachment to a material or imaginary locale, called
place. If social science is a study of meaningful events, geography’s contribution is to account
for the agency of the places where those events are happening. Just as the natural sciences has a

4
set of techniques to investigate the world, geography is methodologically the same – it uses
systematic, evidence-based interpretation to build understanding (Ragin, 2009). Even though the
things it studies are radically different, geography as a social science is as rational, and therefore
as credible, as natural science.
The key difference is geography achieves this. Naturalists have the well-defined space of
a lab, where the performance of science is developed and trusted. Social science is positioned in
society, which is the subject of continuous contestation, not only in and of itself, but also with
regards to our capacity to understand it. This is a Marxist take on ideology, distinguishing
between the “precision of natural science”, and the intrinsic non-precision of the legal, political,
religious, aesthetic, or philosophic (Marx, 1859). Once this distinction entered consciousness, it
created an unavoidable tension, resolved in some disciplines by a formal removal from the social
world. Observations were recorded by an infinitely mobile vision, seeing everything from
nowhere, a “god-trick” (Haraway, 1991). This positionless gaze definitionally removes place as a
unit of analysis, rejecting the researcher’s own attachment to space as a factor in their production
of knowledge.
This is a complete fallacy. When scientists enter the lab, the hermetic seal from the
outside world may prevent dust, germs, or other contaminants from affecting an experiment, but
it cannot bar the social attachments of the scientists themselves. As social beings, they bear the
weight of their backgrounds, relationships, places, memories, emotions, and so on. Rather than
putting place in science, which geography has done, science is put into its place. For a long time,
assertions that knowledge was geographically produced were widely interpreted as suggesting

5
that the knowledge in question was not authentically true at all (Shapin, 1998). My response to
the god-trick is to argue that scientific knowledges are in fact “views from somewhere”, the
result of particular studies conducted at particular times and in particular places and are therefore
socially constructed.
1.3. Health and place
Geographers like me are a determined bunch. Because of our proclivity for
interdisciplinarity, place has slowly but surely taken hold as a concept theorized seriously
enough for other fields to take notice. In the health sciences, space (i.e., physical location,
locatable by coordinates on a map) and its determinants are well-understood, thanks to the
spatialization of enquiry into phenomena of global health like infectious diseases and
environmental hazards. Not so well-established is the acknowledgement of the infinite variations
in the construction and experience of space. Health needs new methods which render this, place,
intelligible. Geography is well-positioned to inform these methods, three of which I set forth in
the empirical parts of this dissertation.
Like many things though, “health” and “place” are themselves fundamentally contested
definitions (Kemp, 2011). This creates an uneasy reality for geographers, who assume that other
scientists should also view the association between health and place as intuitive, despite the
inherent notion that the pathways linking them are difficult to prove and convey. So far there
have been two responses to the question of what it would mean to provide a place-based
explanation of health phenomenon.

6
The first has been to explain away the health sciences as a social phenomenon of place,
common from interdisciplinary approaches which lean more heavily on the geographic side, like
the work in exposomic geographies (Prior et al., 2019; Jacquez et al., 2015). This is risky though,
as in explaining it away, place-based health scholars are undermining their own credibility. In the
effort to elevate the perceptions of objectivity and authority of place, level with, say, the health
sciences, they have succumbed to what Latour (2000) calls “physics envy”. Scholars of place
have mistaken it for something like the objects which health scientists have fully mastered and
dominated, searching for situations in the social world which resemble as much as possible the
mythical posture of the god trick, becoming “disinterested scientists gazing over objective
entities that they could master at will and they could explain by strictly causal chains” (ibid.).
The second is to ignore it altogether, and only explain social functions of the
phenomenon itself, like bias, power relations, and legitimacy, common in health justice (e.g.,
Bravemen et al., 2011) and, as Macintyre et al. (2002) explained, in health behavior. In avoiding
making a place-based explanation of health, scholars have undermined a different possibility,
neglecting to explore the dimensions of health phenomena, such as genetic, which transcend
purely social factors, failing to address crucial aspects of those phenomena. Consequently, its
analysis will almost certainly lack a conventionally empirical depth, leading to partial
understanding and reduced transferable legitimacy.
This dichotomy has to stop, and I call for a serious review of the remit of place in all
interdisciplinary endeavors. The solution in this dissertation is to reject an assumption from the
outset that the social can be known, apart from its composition through associations. This

7
assertion is the foundation of actor-network theory (ANT), which poses society as an assemblage
of agents and things (Callon and Latour, 1981; Law, 1993). ANT proposes a different type of
object-ivity. Rather than matching the legitimacy of natural and health sciences through the
maintenance of a detached perspective, or by increasing distance to a subject matter, it is about
how the agency and resistance of objects themselves – here social phenomena, identities, and
perhaps most importantly, places – resist and challenge the definitions and interpretations
imposed on them by researchers.
For the three papers of the dissertation, place and attachment to place are granted primacy
in the understanding of the social world, but not as a hidden source of causality which points to
the existence and stability of some unknown action or behavior. As a description, “place-based”
does not code a substance, nor a specifically identifiable domain of reality, like “natural”,
“technical”, or “economic” do (Latour, 2000). Rather, it denotes a process or method of
connecting diverse social elements across space, translating different types of entities into one
another. This is evident in the multiple allusions to another ANT concept, “circulating
reference”, throughout the dissertation.
1.4. Expertise and participation
Central to health-place relations are questions of expertise: who or what counts as
expert[ise]? How is expertise formed, authorized, circulated and disputed? And where does
expertise take place? These are politically generative ideas, questioning how the distinction
between experts and non-experts is negotiated, becoming critical as science and politics move
towards evidence-based policy, technocratic governance, and intellectual fundamentalism.

8
Expertise and politics are supposed to be separate domains, with the latter consulting the former,
which has led to some troubling distinctions in research, in particular the proliferation of a
“public deficit” view amongst scientists (Braun, Whatmore and Stengers, 2010).
The lay public, those outside of the academy, are designated non-expert, deprived of
rights to participate in decision-making due to their supposed ignorance and lack of knowledge.
Yet this belies the academy’s own definition of expertise, grounded in siloed, disciplinary
knowledge, and increasing specialty, which means most “experts” are non-experts outside their
relatively narrow knowledge domains. This is proven time and again by the challenging of expert
claims by local, first-hand knowledges (Whatmore, 2009; Beresford, 2000; Wynne, 1996).
Clearly, expertise is not a fixed set of skills and agents, but a social and cultural process which is
neither guaranteed nor easy to maintain (Collins and Evans, 2007). What counts as expertise
depends on the question of who, how, or where assumes the competence to produce knowledge,
which then becomes authorized to inform decision making.
This dissertation approaches this question using a core tenet, that everyone is expert in
their own lives. It asks what lives, which bodies, does geographic enquiry impact, and how the
experiences of those lives can be incorporated. Its three empirical experiments serve as examples
of reconfigurations of the science, attempting to rekindle public confidence via different
permutations of public engagement. It builds on the participative methods within the social
sciences of the 1980s, the deliberative methods of the 1990s, and the collaborative methods seen
from the 2000s onwards, to engage first-hand experience of three different facets of social life in
uneven Los Angeles.

9
The medium used to translate this expertise is place. Spatial attachments to the Eastside
are taken as contingent and dynamic outcomes, rather than as external, homogeneous, or inert
spheres for simple consultation. Participation within geography and the spatial sciences is wellestablished, following the recognition that most information used in research and policy contains,
or can be related to, a geospatial component. Historically, participatory methods in this field
have been driven by decision makers’ desire to increase citizen satisfaction, create realistic
expectations, and produce better outcomes based on local knowledge and skills (Maquil et al.,
2015). When these are carried out using the platforms and tools of geographic information
systems (GIS), these are referred to as public participation GIS (PPGIS). I will later subdivide
this into three separate categories.
So far, a majority of participatory spatial science has been informed by rationalist drivers;
normative and assumed needs to educate and persuade different publics to build social consensus
(Landström et al., 2011). This rests on an unsound and fundamentally unjust theoretical base,
with the objective of one side (science) to engender trust through the management and control of
the other’s (public) response. This creates challenges for participatory researchers wanting to use
the platforms for new forms of knowledge generation.
1.5. Knowledge controversies, wicked problems, crisis
Instead, I undertook three experiments in “civic” engagement, using participatory
methods to critique the assumptions underpinning conventional expertise and decision-making
(Whatmore and Landström, 2011). I chose three sites where claims and counter-claims by

10
different actors has resulted in scientific dispute: food; mobility; and housing. By definition,
scientific knowledge in these realms is provisional, contested, and amended over time. These
disputes have become heated without a prevailing consensus, generating wider social interest
which has led to a loss of public trust. The response is a cultural attribution of “crisis” as a status:
“the obesity crisis”; “the transit crisis”; “the housing crisis” are commonly voiced and tabled in
Los Angeles science and politics. In these cases, science itself is seen as the arbiter of
controversy. Conventionally, these crises are solved through reviews of the science in question,
plus attempts by “experts” and the political elite to educate the public about the methods and
technologies involved.
ANT redefines these disputes as controversies of knowledge. Where uncertainty prevails
and enters the realm of public scrutiny, the disputes become public knowledge controversies, or
“matters of concern” (Whatmore, 2009). Food and eating, mobility and movement, homes and
housing, are naturally under constant public scrutiny. Everyone needs to eat, move, and live
somewhere. Knowledge controversies complicate the division between experts, counter-experts
and the lay public, requiring the kinds of experimentation and empirical research which I lay out
in the three main chapters.
These challenges for science and politics are now considered under the “wicked problems”
framework codified by Rayner (2006) and Rittel and Webber (1973). They are unique issues
with no standardized solutions. They are persistent and insoluble, and there are no definitive
problem formulations, which as a notion becomes itself part of the problems. They are
symptomatic of other, deeper problems. As place-based health issues, there are no clear set of

11
crisp alternative solutions, and are characterized by contradictory certitudes – there are different,
incompatible answers of what is good or bad, depending on one’s worldview. Levin et al. (2012),
in their description of climate change as a specific type of wicked problem, add the following
qualifiers for their “super wicked” designation: time is running out; those seeking to end the
problems are also causing them; and there is no clear central authority. There are clear
implications for entrenched interests in the wicked and super wicked problem conceptual
framework, usually related to the circulation and accumulation of capital. Here are three brief
examples of entrenchment taken from each empirical setting of the dissertation:
i. The centrality of supermarkets in the current capitalist iteration of food systems –
where they are sited is the result of cost-benefit analysis, rather than of altruism by
shareholders.
ii. Urban renewal and lobbying by the auto industry created a city dependent on cars and
freeway construction – to navigate Los Angeles without a car is to do so as an
infrastructurally subaltern body.
iii. Real estate speculation and the rewarding of absentee landlordism – low income
Angelenos might now be expected to spend a higher proportion of their income on
rent than ever before, effectively eliminating any chance of ever being able to own
property.
1.6. Participatory GI Science
From this, I turned to three tools from PPGIS, developing what Ney and Verweij (2015)
termed “clumsy solutions” to address the knowledge practices producing and maintaining

12
wicked problems. The clumsiness emerges from the possibility for compromise, negotiation, and
coalition-building. Rather than privileging singular perspective knowledge, endemic in
institutional technocracy, the dissertation aims to conceive polyrational solutions which
incorporate noise, discord, and contradiction. The clumsy approach mixes different worldviews
in a flexible, emergent manner, producing a set of plural results. This aligns with ANT, where
every worldview is partial, needing others to be effective (McCormack and Schwanen, 2011).
The three tools are outlined as follows:
1.6.1. Participatory GIS for food
The first study in the dissertation, entitled Structurally competent food systems science in
the city: Reflections on a gold-standard food environment audit in five neighborhoods on the
Eastside of Los Angeles, builds on the assertion earlier that research needs to take more formal
account of situated knowledge. Food science, with its historical focus on calorie intake and the
biomechanics of diet and digestion, lacks convincing accounts of place (Patel, 2009; Fonte,
2008). In this study, I spend a summer walking down every street on the Eastside, recording each
retail food outlet. Rather than a passive observer, I was an active participant in the generation of
spatial data, using my body to get a sense of the food environment.
I gained primary experience and insight into the condition of one aspect of food,
generating a rich and contextually grounded set of results. Once again this is an embrace of
principles from ANT, adopting a perspective which acknowledges the active agency of both
human and non-human actors in the networks process of research. As I engaged with the subject,
I became entangled in the interactions of the food system, where humans, material artefacts, and
socio-technical systems were playing influential roles in shaping knowledge construction. My

13
definition of participatory GIS (PGIS) here revolves around the participation of myself as a
researcher, with blurred boundaries between observing and observed. This was a direct
perspective into the relational dynamics at play within food-places, pointing towards the
emergence of new understandings and insights.
1.6.2. Public participatory GIS for mobilities
For the second study, Affective geographic information science: Using carto photo
elicitation interviews to examine the connections between mobility and place amongst the carless
in Los Angeles, I recruited participants with strong place attachment to the Eastside. None of
them owned a car, and I was interested to see how this played out in terms of their movements
and perception of the city. This was explored through the geography of affect, prioritizing the
body for making sense of the world.
Knowledge here was produced by participants, translated using a newly developed
method, carto-photo elicitation. By situating the participatory knowledge claims in the spaces
where they were developed and attaching social meaning from visual materials produced in those
spaces, place became a central discursive arbiter. Rather than treating maps and spatial
information as a static backdrop, common in PPGIS applications in planning and environmental
consultancy (Landström et al., 2011), place in this PPGIS was a dynamic arena where social
relations and inequalities were enacted. In particular, this study proved the potential for the
spatial sciences to go beyond knowledge production as a representative exercise, using ANT
principles to explore non-representational place attachment in the subjective, experiential, and
embodied aspects of the geography of affect.

14
1.6.3. Volunteered geographic information for housing
The final study, entitled Countermaps as resistance: Using volunteered geographic
information (VGI) to record experiences of housing insecurity on the Eastside of Los Angeles,
harnessed the potential of citizen science as a resistance to exploitation. Over 100 participants
contributed place-based accounts of inadequate home-places, from which I produced a prototype
of a platform for sharing these experiences amongst a locale.
This study included a description of the dynamic interactions between individuals,
technologies, and environments. The VGI was intrinsic to the socio-spatial context within which
it was produced, with participants able to contribute and engage with data about their domestic
surroundings. The platform was the result of individuals sharing geographic information
connected to their own perspectives and experiences, reflecting the diverse realities of the
housing situations in the four Eastside neighborhoods. Lastly, it recognized the distributed nature
of agency within a socio-technical network of the postmodern city, highlighting the knowledge
politics of a collaborative effort which empowered individuals to co-create spatial meaning.
Through their participation, community members became active agents in the production of
knowledge, influencing the ways in which spatial information was collected, shared, and
interpreted.
1.7. Significance of the dissertation
The significance of this dissertation lies in its three-part investigation of participatory
spatial science methods, examining to what extent they could be applied to knowledge
controversies like the so-called wicked problems of food insecurity, mobility insecurity, and

15
housing insecurity. Through the three studies, it builds place-based understanding in three ways,
all centered on the Eastside of Los Angeles. The study areas have subtly different boundaries and
extents but use a guiding principal that “the Eastside” lies to the east of the Los Angeles River.
This stands in contrast to the recent trend to term places like Los Feliz, Silver Lake, or Echo park
as “Eastside”. This contested terminology is wrapped in a rhetoric of gentrification, and I argue
should be rejected on two bases. Firstly, in the words of one-time mayor Eric Garcetti, “true east
is east of Downtown” (Los Angeles Times, 2014). The river is east of downtown, and the
Eastside east of the river. Beyond the truth of cardinal directions, this makes sense
geographically because of the cultural divide that the river represents (chapter 3 describes in
detail how the systematic withdrawal of public transit reduced porosity between the Eastside and
the rest of the city to such an extent that the channelized stream still acts as a substantial barrier
today). There are also substantial demographic and historical differences between the four
neighborhoods on the Eastside and the others west of the river: they are not as predominantly
Hispanic/Latinx, so do not share the stories of Chicanx political activism. I have increasingly
encountered compensatory monikers like “The New Eastside”, or “The West Eastside”, which
fail to capture what I now appreciate about the place.
The constitution of a region or a neighborhood is understandably contested, and in many
cases politically charged, as I encountered during my participation in a National Science
Foundation (NSF) research project investigating many of the same areas. What emerges are three
socially-constructed ideations of “the Eastside” as a place, reflective of the understanding
brought forward by the agents participating in each case.

16
The first empirical chapter is a tool for developing my own understanding of place, the
second is a method for my understanding of participants’ attachments, and the third a platform
where the public can understand their community’s attachments. This is a first-, second-, and
third-person knowledge framework, working with techniques and practices which have not yet
been applied for this purpose.
As a matter of course, the dissertation explores the multiple factors of oppression which
have created precarity in Los Angeles. These principally result from neoliberalism and the
privatization of public goods, abandonment, and a state-led corporate political economy. It is my
hope that the methods and platforms which I developed during my doctoral studies will one day
be adopted by geographers and spatial scientists, or if not the schemes themselves, the kinds of
methodological thinking which I used to challenge the conventional hierarchical structures of
scientific knowledge claims.
I think about Latour’s trip to the Amazon when he was writing Pandora’s Hope (1999).
As a pivotal figure in the theorizing of ANT, he might have faced criticism for indulging in
endless contemplation from a Parisian armchair. Instead, he travelled halfway around the world
to a remote part of the Brazilian rainforest to join a soil science project in the Amazon. In the
distant and unfamiliar locale, he undertook an ethnography of the science at hand, participating
in the labor of pedology, and studying empirically the epistemological question of scientific
reference. This is what I have tried to emulate in my doctoral studies, travelling across the
Atlantic and continental U.S. to a different place, to eat, move, and live in the laboratory of the
Eastside for nearly half a decade. As Latour’s site for ethnography was the intractable jungle,

17
mine is the intractable urban metabolism of the city. Where his dialogues were with a
geomorphologist and a botanist, mine were with neighbors and local practitioners. This
dissertation is my expanded field notes, a formal record of my engagement with theory through
situated practices across space.
1.8. Structure
The three articles which compose this dissertation are stand-alone manuscripts. Each one
contains its own abstract, introduction and literature review, methods, results and discussion
sections. They are connected by some common threads; participation in three senses of the word,
ANT, action research, and study area. Each one validates different arguments made from certain
analytical standpoints, leading to a conclusion which pulls the threads tight, provides the key
findings and restates the connections across the three. Also in the conclusion are discussions on
my vision for future work at the nexus of health and place, limitations, a summative reflexive
statement, and the implications for science and politics.
Each empirical chapter is prefaced by a vignette. As this work hopefully shows, there are
no clear demarcations between theory, data, and method. The short stories serve to set the scene
for what follows, making clear some part of my own situated knowledge for each context. They
reproduce experiences which I had eating, moving, and living in different places on the Eastside.
They are full of meaning for me, which may or may not be obvious to anyone else. As field
experiences they are central to my performance of science, not necessarily intended to invoke
any specific interpretation, records of my own attachments to this extraordinary place.

18
Vignette 1 Ave. 26 Night Market
I can’t stand supermarkets. I think they’re cathedrals to agribusiness, each aisle an altar to a
different pillar of the global capitalist food system, so I’m always trying to find ways to avoid
giving them too much of my money. My master’s dissertation was on the capacity for alternative
food networks to feed a city, and since before then I’ve always been interested in other ways of
eating and solving problems of food and nutrition. In Los Angeles I’ve found some great
farmers’ markets in Alhambra, Pasadena, and South Pasadena, which I love for their short
lengths of chain between producer and consumer. Even at the employee owned Winco in
Lakewood, the only chain supermarket I regularly patronize, you can’t meet the farmer who
grew the apples, or put money directly into the hands of the women who baked the vegan
brownies. In the autumn of 2020, I built a chicken run in my back garden and picked up three
chickens from a farm in Lompoc. They laid an egg a day each without fuss, and livened the place
up a great deal when there were so many terrible things going on. I also had space and means to
grow lots of my own food. Anything that didn’t make it into a salad made it into a sandwich or a
soup, and anything left went to the chickens. I mapped out the fruit trees which overhung the
pavement and made friends with the owners of ones that didn’t. When they were ripe, I ate as
many loquats as I could, making ice cream and jam out of the ones I couldn’t finish. The final
agent in my multispecies food network was a huge sourdough starter, which I kept on top of my
washing machine, making bread once a week and feeding it flour and water in between.
That is all to say, when everyone’s food networks changed during the pandemic, I had
built my own resiliencies to stay fed and sane. I thought about how lucky I was, that I had a
garden, and enough time, money and wherewithal to organize that system. I of course knew that

19
others in Los Angeles were not so able to prepare. This was the overwhelming conclusion of
several of the papers which I helped work on as a research assistant from 2021-2023. It was also
evident from just being around food in the city during that time. People were redressing what
was important to their families dietetically, changing how they ate and even conceived of food.
The market latched onto that, seeking new opportunities for profit and exploitation. I remember
neighbors making genuine cash offers for a carton of my hens’ eggs, unable to justify the double
or triple expense at the height of egg-flation. I was quite happy to give them away, not only
because three-a-day was too many for me on my own, but because I knew that they would return
with a box of nopales or a carrier bag full of pomelos in turn.
Whilst the Kroger supermarkets and multinational fast-food restaurants were thriving,
making more money than ever in fact, small, independent retailers were under the cosh. Between
stay-at-home orders and mandatory restrictions on indoor dining, lots of places on the Eastside
shuttered their doors. Those that didn’t or couldn’t afford for the owner’s sake not to stay open,
turned to different techniques. One such was a short bike ride away from me on Artesian Street
in Lincoln Heights. I had heard about a regular night market opening there from my neighbors,
carrying on a tradition that had been put on pause because of city and state ordinances. Local
businesses had started serving food on fold-up tables, out of truck beds, or from hand-pulled
carts fueled by portable butane tanks. Even early on, the number and diversity of options were
extraordinary. There were the taco stands, coolers of aguas frescas, and shopping carts full of
tamales which are pretty much de rigeur for the Eastside, but also all manner of other amazing
things. Pizzas paddled in and out of bicycle-towed wood-fired ovens, baguettes being sliced open
on car bonnets for banh mi, mobile bars serving beer from all over the world. There were also

20
amazing offerings from the businesses in nearby Chinatown – my favorite was an unmarked
vegetable chow mein, which seemed to be sold at a different price for each customer.
The market exploded in popularity by midsummer, largely due to the vendors’ children –
all being schooled at home whilst LAUSD was still in its remote-learning phase – posting on
TikTok and Instagram. The posts were attracting serious views, which drummed up customers
for the food stalls. I ended up going at least once a week that summer, amazed at the scale of the
thing. It stretched for several hundred yards parallel to the gold line tracks, and after about seven
at night, was a complete throng. COVID-19 regulations didn’t exist down there, although people
were mostly respectful of personal space and masking. Even though it wasn’t on the same road,
it became known as the Avenue 26 Night Market, probably because of the eponymous taco stand
which started just around the corner.
Of course, the whole thing was completely illegal. None of the businesses had paid to be
there, which meant lost income for the city, nor were there any food safety or licensing
certificates to be seen. The market ran every night of the week, up until around half-two in the
morning on the weekends. And it was packed. There was no parking at all for streets around –
hence my cycling there – and definitely no public bathrooms. The local residents were
resourceful however, charging for toilet access in their homes, or $5 a head to sit and eat on
plastic chairs in their front gardens. All of it fell outside the reach of either the conventional food
network, or the state. And so, after several weeks of communal joy, LAPD came down one
afternoon and blocked off the entire street with an eight-foot-tall chain-link fence. The system
couldn’t deal with this scale of resistance, and blamed “crime” for its forced closure, rather than

21
its own lament for the lack of access to the undeclared cash or money-transfer revenue. From one
night to the next, the night market was over.
I saw two great injustices in this. First was the proximity to Dodger Stadium, around a
mile away in Elysian Park. By this point in the pandemic, the games were opened to fans again,
and every home game tens of thousands of people flooded into the stadium. The city managed to
overlook the perceived and actual spreading of the virus in this case, presumably because it stood
to make a great deal of money. When those same fans came down from the ballpark and walked
over the 110 to the night market, they suddenly became an outbreak risk. Yet the behavior hadn’t
changed. Rather than paying fifteen dollars for a branded “Dodger Dog” on the stadium-side of
the turnstiles, they were paying five for one wrapped in a Costco paper napkin. And this one
came with onions and peppers and wrapped in bacon. Instead of buying a Heineken in a plastic
cup for ten dollars, people were getting four Modelos in glass bottles with two dollars change.
The second was related to the meatpacking plants which were kept open by the president,
who signed an executive order to that effect in the first month of the pandemic. The plants
already had some of the highest rates of occupational hazard in the country and went on to have
some of the highest infections and death rates in the world in the first throes of the disease. One
plant called Farmer John, a few miles away in Vernon, just south of Boyle Heights, was the
principal suppliers of Dodger Dogs, and almost entirely staffed by Mexican and Central
American migrant workers. The perpetuation of racial violence was legalized to feed baseball
fans. I stood in the middle of the night market with a taquito, looking up to the stadium
floodlights, and remembered Achile Mbembé, who I was reading that summer, and his new

22
spatial relations of necropolitics: COVID-19 had legitimized the assignation by those in power of
different rights to differing categories of people for different purposes within the same space.
Keeping Farmer John open was killing in the name of necessity, at least as far as the market was
concerned, and shutting the market down was a simple exercise of discipline, reproducing statecorporate violence and maintaining criminogenics on the Eastside.
After Artesian was boarded off, I helped to set up the El Sereno Night Market, which still
takes place on Wednesday nights. It is a fraction of the size but has pleasant echoes of the
creativity and resilience that the Eastside showed during the intervening crises of the early
2020s. I’ve since made new friends, with whom I swap eggs from my current flock for limes,
passion fruit, and new strains of sourdough, influenced by the hyperlocal atmosphere and the
hands that fed it. In the second empirical paper, I wanted to show some of the frustrations I was
harboring for conventional food systems and the research done on them. The methods used to
monitor and predict diet and nutrition map very poorly onto the social realities of food and eating
in the city, but starting with its understanding as a knowledge problem, I am always thinking of
ways to resist the formal subsumption of food and digestion under capital.

23
Chapter 2 Structurally competent food systems science in the city:
Reflections on a gold-standard food environment audit in five
neighborhoods on the Eastside of Los Angeles.
Abstract
Background: Food environments are key actants within food systems but are frequently
misunderstood and misrepresented in research. This article explores the consequences of
adopting a structural competency approach when investigating contemporary food
environments. It challenges the limitations of the conventional models, which focus on
proximity to large supermarkets, and argues for a more nuanced and multi-dimensional
understanding of the environment.
Methods: The study introduces a tailored audit tool, adapted from the Nutritional
Environment Measures Survey in Stores (NEMS-S) to ground-truth retail food outlets in the
specific context of five neighborhoods on the Eastside of Los Angeles, California. The audit
collected geographic information so that spatial analysis techniques could be applied to the
results, namely kernel density and service area geoprocessing.
Results: The results show the disparate access to outlets selling healthy foods as a result of
the profit-maximizing incentive of hegemonic food systems agents. The discussion
emphasizes the importance of considering the relational aspects of food environments in
research, calling for a shift from individual-focused approaches toward a framework which
holistically appraises socioeconomic, and historical factors.
Conclusion: The article concludes by highlighting the need for practitioners to adopt
interdisciplinary, mixed methods approaches to deepen knowledge of the complex
relationships between food environments, systems and place.
Keywords: Food environments, mixed methods, retail food audit, spatial analysis, structural
competency

24
2.1. Introduction
Over the last two decades, there has been a growing interdisciplinary focus on the nutrition
and dietetics of residents living in “poor” or “inadequate” food environments, including work
on food access and availability (Walker et al. 2010), food security (Barrett 2010; Jones et al.
2013), and food deserts (Beaulac et al. 2009; Cummins and Macintyre 2002). Despite the
broad-ranging enquiry, the food “environment” itself remains a concept with multiple and
contested applications, likely in part due to its draw to a wide range of disciplines. For
example, in different fields, food environment studies may examine the physical or built
features of a study area (Pradeilles et al. 2021); the placing and arrangement of different
foods within stores (Winkler et al. 2019); the display of different options at restaurants
(Allman-Farinelli et al. 2019); or the sustainability or affordability of different diets over time
(Downs et al. 2020; Kaiser et al. 2019).
For this reason, it becomes important for food scholars to clearly define their version
of the food environment from the outset. So, in this study, the food environment is the set of
points where capital accumulation is most intense, namely retail outlets selling food for
profit. This is somewhat reductive, distilling the production of a system by an elite white
bourgeoisie used primarily to exploit labor and profit for maintaining structurally racist
inequalities in the post Fordist American metropolis to a set of coordinates and attributes. The
rationale for this reduction is that these points have a disproportionate impact on the lives of
people in poor food environments, especially where they are the only source of nutrition.
This is informed by the “structural competence” framework of Hansen and Metzl
(2019), in part to discern how the clinical issues of symptoms and disease represent the
“downstream implications of upstream decisions” (ibid. p.128). The study particularly draws

25
on the first “competency”, which calls for clinicians and researchers to incorporate an
appreciation of structural barriers in their day-to-day practices.
The specific focus of this structural competency analysis of the food environment is in
critique of a symptom of the academic and public-policy development of the urban “crisis,”
and the resultant application of interdisciplinary expertise to the management of food systems
in American cities. The dependencies of the hegemonic global order on slavery, violence,
imperialism, and genocide (Kelley 2017), are especially clear in agribusiness and the
nutritional networks of capitalism, which have merged to become the dominant global
schema. The brazen safeguarding of structural inequality by property owning elites has come
under increasing scrutiny by health and social science researchers since the sudden
widespread recognition in 2020 of systemic racism in the Unites States (U.S.) and has forced
new perspectives on routinized and interlinked inequality in all parts of the American social
fabric (Bowen et al. 2021; Motta 2021; Smith 2022; Vonthron et al. 2020).
Compare on one hand the incredible life-producing technologies made possible during
the third agricultural revolution of the late 20th century with the seemingly incompatible
notion of malnutrition and deficiency in the 21st century on the other. Hunger as a condition
is not solely confined to the global economic hinterland: rather, the Current Population
Survey Food Security Supplement (CPS-FSS) estimates tens of millions of U.S. households
being hungry because of too little money to buy food (Coleman-Jensen et al. 2022), and
whilst 38 million people in the U.S. claimed food assistance in 2019 (USDA Economic
Research Service 2022b), which is an incredible figure in itself, the Food and Nutrition
service of the U.S. Department of Agriculture (USDA) estimated in a 2021 report that around
one-in-six eligible claimants did not participate in the program (Foster et al. 2021).

26
This claim is proof of the narrative of improvement which has come to justify
population-level advances delivered at the expense of chosen individuals. Urban
policymakers and the remainder of the governing elite pursued state-led corporate capitalism,
absentee landlordism, and monocropping as the way forward in feeding American cities. The
terms on which food is, or is not, made available to people has been almost entirely placed
into the open hands of the market, with precious little consideration of the social control of
the food system: “food security” as a term makes no mention of this. As Patel (2008) points
out, it is entirely possible to be “food secure” whilst in prison. Access to food does not come
with a guarantee of any power at all in an agential food system.
The sacrificial in the American instance was the poorest tranche of society,
unsurprisingly including a multitude of Black, Latinx, indigenous, and minority ethnic
households (Reese, 2019). Alienating agricultural production is part-and-parcel of the
productivist paradigm, lengthening the chain of transactions between producer and consumer.
As links in the chain, additives and preservatives are added to food in quantities inversely
proportional to the shelf-price (Baker et al. 2020), increasing the energy density but reducing
the nutritional quality. Supplementing the diets of the American poor satisfies the politicians
and electorate concerned with dispensing food into the hungriest mouths (Patel 2008), as well
as the agribusinessmen, for whom the costs of production have never been lower, and hence
the margins of profit never higher (Joyner et al. 2022; Zhang and Ghosh 2016).
This article asks how the structures of American society, as shaped by historical
contingency and socioeconomic path dependency, have influenced the place-making effects
of food environments. It uses conventional methods of representation to draft responses to

27
this question but goes on to ask how such practices in food systems science have resulted in
the predominance of discipline-specific worldviews and problem formation. In this way it
aims to demonstrate the importance of developing innovative and interdisciplinary
approaches to wicked problems such as inequitable access to food.
2.1.1. Food deserts
An important part of this narrative is the historical delineation of certain areas as
“food deserts” by the USDA – itself a federal department devoted to expanding American
productivism. Considering its primary function at its inception, preserving the practices and
standards of an elite white property-owning class (Grim 2012; Sloan 2022), it is no wonder
why the USDA has such a strong focus on increasing production: more output results in
greater profit for the protected landowners. The USDA definition today has jettisoned the
“desert” label, opting instead for “Low-income low-access” (LILA) areas. This is a
population estimate of an indicator of spatial access to healthy and affordable foods, using
demographic data from the U.S. Census as well as estimates from the American Community
Survey (ACS) and food availability data from a national business directory (USDA 2022b).
As a dimension of the neighborhood-level food environment – the delineations are
determined at a census-tract scale – LILA status has become widely used by academics,
government officials and charitable organizations, as well as in reports for all manner of food
policy decision making. As concepts, LILA and the erstwhile food desert label have become
part of the public consciousness, as easy-to-grasp metrics with perceivably household-level
ramifications; how far you have to travel to be able to buy healthy food.
This is an interesting sociocultural development. Ostensibly the causation is clear
here: diet has strong, well-studied, associations with health (National Institute of Health

28
2017), so there should be similarly clear associations between life in a LILA neighborhood
and the reduced health of that population, as the commonly held perception is that the only
available food comes from fast food restaurants, convenience stores, and other limited
markets (Robbins 2020). Crucially though, the entire principle of LILA in the U.S. rests on
the fundamental principle that this is a commodity relationship. The USDA classification
model only considers supermarkets, supercenters (e.g., Target, Walmart), and large grocery
stores, each of which must have annual sales of over $2 million (USDA 2012) as spatial
modifiers for determining food access. Grocery stores are businesses, so are therefore
profoundly invested in profit and its maximization. Retail food outlets are nodes in the
capitalist food network, a network that the USDA was established in the early American
agrarian economy to support. Reflect on the ramification from the preceding paragraph. The
embedding of this iteration of access into public consciousness is not a deliberation of food
and eating, it is a dissolution of a survival function into a transaction – the real subsumption
of diet under capital. It is necessary to consider what this means. Rather than predicting
spatial conditions for scarcity and hunger, LILA models simply map areas which supermarket
owners have deemed profitable.
These ideas are supported by much of Patel’s (2008) work, like the “stuffed and
starved” concept. Firstly, it describes the characteristics of “food desert food” as highly
processed and nutrient poor, but also contends that obesity and hunger are two points on a
continuum of poverty, not food supply. Food outlets, especially the large chain retailers
included in the USDA LILA model, are not designed with the intention of feeding
populations. Instead, they have a driving aim to seek as much profit as possible from each
location. This business incentive shapes the formulation of food environment, evident in the
long-held associations of the increased prevalence of unhealthy options in the census tracts

29
which the USDA deems LILA. It is unsurprising that whilst LILA areas exist in many
different communities across the U.S., Walker et al.’s (2010) review found that they are
densely concentrated in communities of color, specifically amongst areas of high Black and
Latinx populations.
This finding portends the continuation of racialized biowarfare through the disparate
prevalence of ill-health and disease, for example non-white Americans are nearly twice as
likely to suffer from type II diabetes (Robbins 2020). So that is where many researchers stop.
The face-value of the USDA classification stands for a catch-all evidence base when
designing food policy, which all too often resolves in protractions of the productivist
paradigm: recommending the construction of a new supermarket in one neighborhood, or the
siting of a large grocery store in another. Such recommendations are endemic in science and
politics today. From the data alone, the USDA model is ill-equipped to answer questions of
how race and capitalism have influenced the relationships between food environment, diet,
and health. It is incredibly problematic therefore when researchers and decisionmakers draw
inference from it without acknowledging its shortcomings. This is not to say that it should be
dismissed altogether without question. Distance to retail locations does play a significant role
in food consumption and has been long recognized in doing so (Bitler and Haider 2011; Shaw
2006; Wrigley 2002), but it is just one part of the decision matrix resulting in either healthy
or unhealthy outcomes. This article outlines the requirement for scientists to pay a deep
attention to racial capitalism and other oppressive histories when building their matrix,
whether it includes food desert ideology or not, and to track the entangled production of
health and social knowledge throughout the entire enquiry.

30
Notably too, the classification is an outsider label. LILA, or “desert”, are not
descriptions used by impacted communities, and in many instances clashes with
neighborhood and community efforts to address inequitable food. Hill (2017) described this
juxtaposition in Detroit, where the persistence of the term served to sustain the unequal
power relations between local advocates and the predominant media and political institutions.
Kolb (2021) discusses this inequality at length, before adding that outsiders who are
compelled to intervene in LILA areas do so because of the impression of scarcity that the
label engenders. Outsiders are usually ill-equipped to create useful or lasting solutions and
holding scarcity perceptions, tend to deliver interventions straight out of the productivist
playbook. This provokes another of Metzl and Hansen’s (2014) structural competencies, the
requirement to develop a language of structure. Food desert as a label is simply no longer
appropriate for describing the geographic injustices created by capitalist food networks,
because of its adoption by media and politics as a metaphor which discredits those living in
the areas (Hill 2017) at the same time as obscuring their deliberate creation by the forces of
free market neoliberalism. Although appealing for natural and health scientists studying food,
the disciplinary reaches for universality and convertibility have in this instance resulted in the
adoption of a singular language as the standard for all such conversion. Haraway (1988) calls
this reductionism, the language positivism, and notes that its authors tend to hold the
positions of a white, male, governing elite.
2.1.2. The role of data
Examining the role of reductionism in research is important, especially when
investigating something as complex and nebulous as the food environment. Although largely
dismissed as a specific critical framework (Adams St. Pierre 2016), the inertia of logical
positivism continues to reverberate in the practice of food science, where the kinds of “brute”
data which form the inputs to food models perform the role of the “solid bedrock” of thought

31
and knowledge (ibid.). There is only one way of reading them (the data), with little chance
for ambiguity or uncertainty. “Data” should be recognized for the philosophical baggage it
brings to the research process; not as neutral essential elements abstracted from the world
through objective enquiry, but as constructions created within complex assemblages which
themselves actively shape their constitution (Kitchin 2014). In other words, making sense of
the world should always be framed by how data are generated and examined, and then
through which theoretical lenses they are interpreted. Failing to name this in research, or
indeed to reflect on it at all, does not mean that this process is absent, merely that a researcher
is using a theoretical framework which they have not acknowledged, and that their work is
being fundamentally guided by a process which they have not credited. This is a failure
commonplace in many studies making use of the LILA concept and should be considered
problematic if they claim to understand the food environment.
All data are partial. No single source can serve as a complete and accurate
representation of social reality. So too are the methods used to generate them. Any method
has its specific strengths and weaknesses which render the materials it produces unable to
address all dimensions of the part of the world it set out to examine. Researchers have tackled
this problem by combining methods, in “mixed methods” approaches. The traditional
definition of such an approach is using different methods and data to answer the same
research question, and the employment of mixing is so common in health and social research
that it has become relatively unremarkable as a facet of research design (Bryman 2006). The
work for the wider food systems project encompassing this study is truly mixed methods,
with researchers from geography and the spatial sciences, public health, psychology and
behavioral science, public policy, and an assembly of non-academics comprising local
residents and practitioners who each bring their own methodological experience, ideas, and

32
interpretations. The study is thus able to serve a split purpose: firstly, to critique and contest
LILA and food desert labels which have been placed on large parts of the study area; and
secondly to improve the richness and quality of the claims that the wider project will make by
its conclusion. The position of this work as part of a broader research remit means that it can
act as one pole of knowledge production – it is a research tool which is being used to build a
matrix of understanding of the food environment.
2.1.3. Mixing methods in food research
It is hoped that this study will initially provide the grounds for the now conventional
rationales of mixing methods: complementarity; development; and triangulation (Greene et
al. 1989). The idea here is that if different methods are biased in their own direction – or ask
different questions altogether (Shotland and Mark 1987) – yet still come up with the same
results, the biases should have been offset or counteracted in the investigation of the shared
phenomenon. These rationales are effective, even a little prosaic. The expectation is the
production of strong, rigorous evidence which is palatable for the publication and other
conventional means of disseminating results. This is common in the food scarcity literature,
which uses the LILA model alongside other common quantitative and non-situated methods
to produce more narratives of the symptoms of capitalist food networks. These other methods
typically include anthropometrics like BMI (e.g., Testa and Jackson 2019), dietary recalls
(e.g., Dubowitz et al. 2015), or mobility and other spatial data (e.g., Shannon 2015). These
first three motives do have clear traces of a positivist perspective, in that there is an
expectation of a positively defined reality (Law 2004). Taken to their logical extreme, they
suggest that if enough methods are added, eventually the research should converge on the
objective truth. This doesn’t really make sense in a food systems context, as food is
subjective, personal, place-based, socially constructed, and constantly evolving across space
and time geographies.

33
More relevant to food systems research with this context is the proposition that
methodology scholars like Rossman and Wilson (1985) call “initiation”. Rather than seeking
to produce confirmatory evidence, initiation rests on the idea of finding paradoxes and
contradictions between results produced through different approaches, which can then be
used to identify the gaps and omissions produced by any one specific method. This is relevant
for research on the food environment, as part of food systems science, as the kinds of
provocative knowledge which is required to develop new interpretations of the social and
human experiences of food and diet is based on subjectivity and uncertainty, so requires a
large creative leap. Areas of nonconverging evidence, such as the expected tensions
surrounding LILA classification, can now initiate new interpretations, suggest areas for
further analysis, and recast the framing of individual research questions as well as the broader
areas of enquiry for the collective project. This finally allows the study to move beyond the
positivist perspective, within which the principal concern would be the technical “nuts and
bolts” of science. Instead of rejecting any emerging incongruence, and redeveloping methods
to reduce the perceived inaccuracy or bias, this post-positivist interpretation cherishes the
unanticipated outcomes, leading to the subjectivity which Haraway (1988; 1991) calls for in
the rejection of reductionism. Instead of working as hard as possible to reconcile the
dissimilar results from the mixed methods, the divergence will instead create opportunities to
enrich the evolving explanation. At the very least it means that the multiple and contested
definitions and applications of the “food environment” concept will be granted multiple,
contesting methodologies and theoretical frameworks to enrich the debate.
Rather than a solid bedrock, the methodologies upon which science and its
observations are formed are interpretations piled on interpretations (Adams St. Pierre 2013).

34
Meaning in this worldview is not centered on the object-subject divide, rather they are
contingent, plural, uncertain, and emergent. As long as this is upheld, whereby the enquiry of
science and scientists themselves are at stake and open to debate, the real-world food
environment could never be converted into a stable and static representation. However, this is
seemingly at odds with the stated framework of this study – structural competency – which
spotlights the impact of the social structures of race, gender, and class for understanding and
addressing health disparities. This is where the novel insights of this study become clear:
questioning the assumed stability and objectivity of the methodological structures which have
come to shape contemporary food environment discourse. By undertaking a deep-dive
investigation into one aspect of the impossibly complex reality of the food environment, this
chapter seeks to expose the ways in which power relations and discourses have shaped and
reinforced the disparities which structural competence emphasizes, and how these disparities
are produced and reproduced through scientific study. To the authors’ knowledge, there is no
existing work which adopts a post-structural structural competency approach for any
application, so this represents an exciting chance to deconstruct the discourses and power
relations of the types of methodology which have characterized the understanding of lived
experiences in a postmodern metropolis. It also crucially aims to highlight the importance of
reflexivity and critical subjectivity amongst practitioners and researchers in this field, in order
to avoid perpetuating or reinforcing the harmful practices of colonial and racial capitalist
technoscience.
2.1.4. The NEMS-S
The method used to add to the understanding of the food environment in this project
is the Nutrition Environment Measures Survey in Stores (NEMS-S, Glanz et al. 2007), which
was developed to assess food availability, price, and quality in retail food outlets, referred to
as the “gold standard” of food audits by Partington et al. (2015). The NEMS-S, a directly

35
observational audit of individual food establishments using detailed assessment of 11
different types of foods, has considerable benefits. Inter-rater and test-retest reliability is
exceptionally high (Glanz et al. 2007), and the tool is fairly complete, in that it provides a
useful and verifiable account of the food for sale in a study area. The tool is widely used – the
original paper has been cited nearly 1,000 times, with hundreds of empirical publications
using NEMS-S to answer a variety of research questions from around the world.
This method has been deployed as part of mixed-methods approaches previously,
usually to evaluate commercial listings – very high-level categorizations of food outlets (e.g.,
full-service restaurant, fast food restaurant, supermarket/grocery store, other retail store, etc.),
with no finer detail about the composition of food available for sale. Authors use NEMS-S
audits on sampled outlets to gauge how accurate the list databases are and use that
information to characterize their own versions of the food environments they are studying
(e.g., Caspi and Friebur 2016; Franco et al. 2008; Saelens et al. 2007). This is an example of
the development and complementarity rationale to mix methods. Commercial listings are
useful for business purposes, rather than for understanding the range or nutritional quality of
foods available. To use these data alone to define a food environment, for example based on
the number, ratio, or spatial proximity, to outlets in each category, can only provide limited
insight into the actual food a population has access to, which matters from a public health and
food equity perspective. Mixing data from commercial listings with the results of a NEMS-S
audit, however, allows for an identification of outlets which can then be analyzed for the
affordability, quality, and cultural appropriateness of food on sale.
There is also a chance for some basic initiation of methods in those cases. The
disciplinary differences between business and dietetics creates different observations of the

36
same features, the tension between which can be used to examine the predetermined
perceptions held by researchers in each field. Collective deliberation of these differences,
however small, still represents a mixing of epistemological frameworks and a chance to
enrich the outcomes of research. This type of methodological enrichment is what is sought
after in the undertaking of this study, specifically by the comparison of conventional LILA
methods with this iteration of a NEMS-S audit, and a reflection on what this will mean for
wider food environment research.
2.2. Methods
2.2.1. Study Area
The study was conducted immediately east and northeast of Downtown Los Angeles,
California, USA. The boundary was drawn as part of a National Science Foundation (NSF)
project researching community food systems in Los Angeles. This included the four Eastside
neighborhoods: El Sereno; City Terrace; Boyle Heights; and Lincoln Heights. It extended
contiguously to a further section of the unincorporated city of East Los Angeles stretching
south to the 60-freeway, called Wellington Heights. The neighborhood boundaries followed
those in the Los Angeles Times’ (2010) Mapping L.A. project, but with edges adjusted
according to input from a small advisory council comprising resident local experts and
researchers familiar with these neighborhoods. In particular, it was expanded to include the
William Mead public housing development in Chinatown just west of the LA River, and part
of Montecito Heights directly north of Lincoln Heights.
The area is of interest to food scholars for two main reasons. Firstly, by all measures
of conventional classification, there are numerous census tracts which are LILA (USDA
Economic Research Service 2022a) and beneath the federal poverty line (U.S. Census Bureau
2021). Secondly, four of the five neighborhoods border the USC Health Sciences Campus, a

37
multi-billion-dollar site for medical teaching, practice, and research, with Wellington Heights
only a few miles away. This jarring spatial proximity results from a local history of
segregationist and other exclusionary measures including redlining, decommissioning public
transport, and many other imposed acts under the umbrella of “urban renewal” (Gibbons
2023). These acts of governance and population management culminated in the 1961 decision
to site what became the world’s busiest freeway interchange a few miles west in Boyle
Heights. The 135-acre junction, which at some points is 27 lanes wide (Estrada 2005), has
come to serve as a physical and metaphorical border between the Eastside and the remainder
of Los Angeles – a membrane for the auto-dependent city across which non-automotive
populations are effectively prohibited from passing. The area is now almost 100%
Hispanic/Latinx, the highest proportion census-designated place in the U.S. (U.S. Census
Bureau 2022)1
.
These structural factors support the view amongst researchers and policymakers that
the food environment in these neighborhoods is “poor”, and that there is a need to develop
strategies for intervention. Meeting this need requires devising reliable and valid
measurement tools that are tailored specifically to the people of the place and the food
environment in question. Without this, interventions in the Eastside will be based on the
kinds of non-specific, distanced metrics which have stunted progression in food research and
policy over the past two decades.
And yet the outsider attribution of LILA betrays the foodways of resistance formed by
the residents living in them. The Eastside is widely recognized, beyond the consideration of
1 Census Designated Places are used by the U.S. Census Bureau in the American Community Survey
to represent singly named, contiguous communities with no legally defined boundary.

38
the academy, as an area with some of the best and most extensive Mexican cuisine in the U.S.
In particular, there are outlets which have catered to the Chicanx community, serving a wide
range of regional food to the Mexican-descendent population. As well as restaurants and
cafes, an alternative and informal food network has been built out of food trucks, carts,
stands, markets, and home kitchens. It is not uncommon to observe people queuing along
residential streets on different nights of the week, waiting to order through kitchen windows
or over propane burners set up in front yards. Few of these are registered businesses, and
while some develop a reputation and inevitably become so, the clandestine, word-of-mouth
nature of eating there serves to enhance the sense of community and authenticity – people
cooking for and dining with their neighbors on folding tables in driveways, broadly removed
from many aspects of the formality of capitalist food environments.
Another example was the Avenue 26 Night Market in Lincoln Heights, which at one
point was touted as the largest in the country (Los Angeles Times 2021). The short-lived
community venture, open only for a few weeks in the summer of 2021, was a vibrant
collective of hundreds of local vendors who were experiencing hardship during COVID
lockdowns. Its popularity was fueled by posts on social media showing an incredible range of
food, both prepared and for at-home consumption, which celebrated the perseverance of the
historically migrant community in the face of the intervening failures of conventional
globalized food networks. Consider that just a few miles away, just south of the study area,
an executive order signed by the president had kept industrial meatpacking plants in Vernon
open and running in the face of some of the highest rates of coronavirus infection and death
in the country, in service of continued profits for the plants’ owners (Los Angeles Times
2020) and maintenance of the status quo. There is a double cruelty here, in that one of the
plants supplies Dodger Stadium’s “Dodger Dogs,” the construction of which in the 1950s had

39
displaced hundreds of Mexican American families from the Chavez Ravine, and whose
floodlights dominate the skyline over Avenue 26. The closure of the market by the Los
Angeles Police Department and the Mayors’ office, out of respective concerns of crime and
COVID infection, are in direct contrast to the acts of violence enforced by the same structures
which traded lives for meatpacking and homes for a baseball parking lot. The LILA label in
Lincoln Heights is therefore a violent and political imposition, imposing a presumed
endogenous “nothingness” onto a fabrication of targeted housing, financial, and
governmental techniques to structurally discipline the neighborhood. The minority ethnic and
immigrant enclaves which materialize in similar areas of poor health are not an outcome of
coincidence, but rather of a constructed political tactic.
2.2.2. Identifying outlets
This study focused solely on the characterization of the retail food environment, and
specifically outlets selling groceries (i.e., foods that are typically prepared and consumed at
home). Whilst an inventory of available restaurant food is generally straightforward – many
outlets publish copies of their menus online, which can be accurately analyzed and audited
without a ground-truth – retail food data are not so readily available. At the same time,
although there is research on scraping data on foods sold at large grocery stores using
barcodes, like FoodSwitch (Dunford and Neal 2017) or online information, like Open Food
Facts (Open Food Facts 2019) which links to the type, nutritional quality, and price, many
smaller or independent outlets do not list their food online, or even have barcodes on the
products. This means that a comprehensive audit of a neighborhood’s food environment
cannot rely on these data sources when trying to capture the food available in the small
markets and corner stores and that are a key feature of the environment on the Eastside.

40
Similar to other NEMS-S studies, retail food outlets were identified using multiple
data sources. County retail food licenses were matched with location data in a geographic
information system (GIS) and verified using Yelp and Google Maps data alongside previous
experiences of the field researcher. For the purposes of this study, only outlets in which retail
food is accessible to the public were included, which in these neighborhoods meant the
exclusion of several wholesale and distribution centers, and any churches, schools, libraries
or similar institutions where food was sold. Outlets which offered service-only food sales,
such as full and fast-service restaurants or cafés were excluded, but specialty stores such as
bakeries and butchers were included. This left mostly outlets which sold food “requiring
additional preparation” a category explained in Lucan et al. (2020), with some outlets
offering “ready-to-consume” foods requiring no additional effort than perhaps opening a
container. The result was a hybrid of mostly “grocery” and a few “grazing” (ibid.) outlets.
Also excluded were any outlets which were listed as permanently closed according to most
recent data. In total from all datasets there were 1,737 food outlets in the study area.
Excluding the non-retail, limited service, closed, and duplicate data, the final list includes 364
food outlets, shown in Figure 2.1.
2.2.3. Reduced item food audit
The development of the audit tool followed a review of existing literature and tools
(i.e., Cohen et al. 2020; Glanz et al. 2007; Lucan et al. 2020; Ohri-Vachaspati and Leviton
2010). Researchers doing cognate work were also consulted, and what resulted was an
amended version of the NEMS-S. The granular and repeated sections 1-11 were curtailed to
reduce the burden and complexity of repeated assessments of retail food without
compromising measurement quality (Carins et al. 2019; Partington et al. 2015). The reduced
item audit focused on availability, acceptability, and price, removing sections on reference

41
brands, specific nutritional information like percent fat of meat or grams of sugar in cereals,
alternative products, and packet sizes. Also excluded was the on-site calculation of the
aggregate NEMS score. This was because the study focused on the use of a tool which could
be used for targeted interventions in the relatively small study area, perhaps on an individual
outlet level, so general scores were less important.
Figure 2.1. Map showing study area and the 1,737 identified food outlets within the boundary
and the 364 outlets selected for audit (Sources: Data Axle 2021; County of Los Angeles
Public Health 2022)

42
2.2.4. Additions to NEMS-S
There were several additions to the tool as follows.
2.2.4.1. Outlet storefront
A photograph of the outlet storefront was taken to be used for cross-referencing to the
existing datasets in the instance that the store’s name had changed or was listed differently in
the business list.
2.2.4.2. Benefits
The acceptance of government food assistance benefits was also recorded, as this
makes up a crucial and often overlooked facet of food access which increases financial
accessibility to food for paying consumers. In California, the most widely used programs are
CalFresh (known nationally as the Supplemental Nutrition Assistance Program, previously
“food stamps”) and WIC, the Special Supplemental Nutrition Program for Women, Infants,
and Children. These programs provide recipients with electronic benefit transfer (EBT) cards
which are accepted as payment for food by the outlet. Records were made from either the
presence of a sign that displayed EBT, Supplemental Nutrition Assistance Program (SNAP),
CalFresh, or Food Stamps, or the “Quest” sign which is attached to point-of-sale equipment
telling recipients that their cards can be used at that location. The presence of an ATM which
allows cash withdrawals using an EBT card was also taken as an outlet’s acceptance of
benefits.
2.2.4.3. Outlet type
The primary business purpose of the outlet was also collected, whether principally a
food store or not, like a liquor store, gas station, or drugstore. This was based on researcher
discretion, determining the primary purpose of the business based on what products were
most prominently sold, or indicated as “other” with a short explanation if the business

43
retailed food but the type of outlet could not be classified. Additional information collected
on each outlet included how much of the outlet space was dedicated to retail food; whether
the outlet had refrigerators or freezers for food; and whether there were food preparation or
re-heating facilities which customers could use. The last feature is important to consider in all
areas, but especially on the Eastside, where poverty and housing insecurity have combined to
leave many residents without their own space or facilities to prepare and cook meals.
2.2.4.4. Culturally appropriate foods
Another observation noted was the measurement of the availability of products
associated with one or more ethnicities or cultures, in order to assess the range of culturally
appropriate food in the study area.
2.2.4.5. Spatial information
Our version of the NEMS-S also collected spatial coordinates alongside the food
attribute data. This was achieved by completing the audit using ArcGIS Survey123, a digital
data collection tool which records the location of the user device at the time of the
observation using GPS. This meant that the data could be easily and quickly visualized
spatially on a map. During collection, this allowed for continual monitoring of which areas in
the study area had been audited, and subsequently facilitated spatial analysis for any or all of
the recorded data in a GIS. It also meant that the ground-truthed, “gold standard” data could
be compiled and examined alongside the pre-existing datasets and commercial listings. In the
future this could be used as a verification tool for a study area, or to show what parts of the
food system are incorrectly, or not at all, represented in conventional sources. It was
important to record precise location data so that the outlets would be mapped distinctly (i.e.,
not on top of each other). ArcGIS Survey 123 allows for sub-meter accuracy on most devices,
including the Android tablets that were used for this study.

44
2.2.4.6. Sample price data
Guided by the literature and consultation with the public advisory council, ten
products were chosen as likely to be available at a majority of outlets and purchased by many
individuals at least once a week. These were apples, bananas, roma tomatoes, whole milk,
wholewheat bread, tinned beans, chicken breast, eggs, frozen strawberries, and frozen
lasagna. Prices were taken either from shelf labels or stickers on individual items, and
normalized for weight (e.g., per lb.) or size (e.g., 64 oz carton) to allow for comparison across
the study area. Where no prices were displayed, the records were left blank.
2.2.4.7. Supermarket/Grocery Store classification
Using a classification system developed from a synoptic literature review (Bernsdorf
et al. 2022; Canalia et al. 2020; Han et al 2012; Ohri-Vachaspati et al. 2011; Powell et al.
2011), and consultation with the public advisory board, I decided on a criterion for defining
outlets as supermarkets or large grocery stores. For the classification, an outlet had to be
primarily a retail food outlet, plus have at least two types of the audited fresh fruit and
vegetable items, and at least one type of grain available.
2.2.4.8. Pilot test
The audit tool was finalized following a short pre-test in a small sample (10 outlets)
of the whole study area. This pilot test highlighted unclear and redundant questions, as well
as audit elements which were unreasonable to collect without extended time in each outlet. In
particular, it was still important to be conscious of social distancing and perceptions of health
researchers on the Eastside, where many low income and high minority ethnic neighborhoods
had suffered disproportionately from COVID-19. The audit was therefore designed to be able
to be completed without having to speak to clerks or customers. Where appropriate and

45
convenient, shopkeepers or clerks were greeted, and the study was introduced and explained.
The audit was designed to take around 10 minutes to complete in each outlet.
2.2.4.9. Data analysis
Spatial and statistical data analyses were completed in ArcGIS Pro (version 2.9).
Records from the audit were compared with records from the identified outlet dataset, using
Google Maps data as a secondary reference. The records were projected as a point feature
class, with individual points for each audit completed. These records were then compared
with LILA data from the USDA Economic Research Atlas, using both the spatial information
and the non-spatial food attribute data.
Kernel density estimates were generated using 5000 m2 hexagon-shaped cells with
densities as the output values to difuse the spatial impact of outlets across the cell radius and
create a raster surface showing the varying weights across space. The search radius was set to
the long diagonal of the hexagon (87.74m) to produce a highly detailed raster. The planar
method was used on account of the small study area size.
Service area analysis was undertaken using the network analyst extension of ArcGIS
Pro. The outlets were classified according to different attributes (e.g., accepts EBT, sells
produce) and saved as separate facilities feature classes. Cut-offs were set at 5-, 10-, and 15-
minute walk-times. The Esri road and pedestrian network was used to calculate polygons,
which were drawn with high precision and dissolved to merge the service areas of any outlets
with the same cut-off values into a singular polygon. Data from the LA County Office of the
Assessor Parcel Map Service (2022) were used to calculate the number of households within
each service area using zonal statistics.

46
2.3. Results
2.3.1. Outlet type
A total of 269 retail outlets were audited as part of this study. This represents 74.9%
of the 364 outlets identified from the pre-existing datasets and commercial listings. The
remaining 95 outlets had either shut down, changed business type, or showed no evidence of
ever having been a retail food outlet.
The audited outlets were a mix of retail types that included conventionally defined
supermarkets as well as many other stores whose primary business was not selling food
products. The types of stores are summarized in Table 2.1.
Table 2.1. The 269 outlets on the Eastside which sell retail food
Category No. of stores
(percent)
Supermarket/large grocery store 36 (13.4)
Other stores that sell food products 129 (48.0)
Non-specialty stores (e.g., mini-marts, convenience stores) 81 (30.1)
Bakeries 25 (9.3)
Butchers 17 (6.3)
Fishmongers 4 (1.5)
Health food stores 1 (0.4)
Confectionary/candy stores 1 (0.4)
Stores whose primary business is not grocery sales 104 (38.7)
Gas stations 30 (11.2)
Liquor stores 23 (8.6)
Dollar / discount stores 15 (5.6)
Donut stores 17 (6.3)
Drugstores/pharmacies 9 (3.4)
Other/not known (e.g., money transfer stores, water stores) 10 (3.7)
There were 36 supermarkets or large grocery stores (13%) according to the criteria
used and 129 stores selling more limited foods (48%), with a majority of these being “nonspecialty” stores like mini-marts and convenience stores. A smaller number (48) were
specialty stores like bakeries, butchers, and fishmongers. The remaining 104 outlets (39%)
were other types of stores whose primary business was not selling food but did have some

47
food products available. These included gas stations, liquor stores, dollar stores, and
drugstores. Around a quarter of the outlets (66, 24.5%) were chain stores.
A majority of the outlets allocated only a small amount of their total retail space to food.
Almost half (127) of the stores had less than 25% of space for food. Only one in four stores
(66) had more than 50% of their space dedicated to food.
2.3.2. Reduced item food audit
The majority of the stores sold multiple unhealthy food items, whilst options like fruit
and grains were less common. Almost all stores (n = 259, 96.28%) sold candy/confectionery
and sugar-sweetened beverages like soda. Healthy grocery staples like fresh produce and
milk alternatives were sold at all supermarkets and most grocery stores, but these outlets were
also equally likely to sell unhealthy foods too. Just over a quarter of all stores (76, 28.25%)
sold any produce at all. Of those, 75 (27.88%) sold fruit only, and 60 (22.30%) sold
vegetables only.
Fifty-nine stores (22%) predominantly sold food associated with one or more
ethnicities or cultures, almost entirely Latinx- or Mexican-associated foods. Overall, 53% of
supermarkets sold some kinds of culturally relevant foods, compared to 17% of the other
stores.
Table 2.2 below shows the number and percentages of stores on the Eastside selling
specific foods.

48
Table 2.2. Percent of stores selling specific foods out of all, supermarket, and nonsupermarket outlets on the Eastside
Category
All stores
that sell retail
food items n=
269 (percent)
Supermarkets/large
grocery stores (n=36)
(percent)
Stores that are not
supermarkets/large
grocery stores
(n=233) (percent)
Alcohol 125 (46.5) 23(63.9) 102(43.8)
Baked goods/pastries 219 (81.4) 31(86.1) 188(80.7)
Beans 170 (63.2) 36(100.0) 134(57.5)
Bread 86 (32.0) 27(75.0) 59(25.3)
Candy/confectionery 258 (95.9) 34(94.4) 224(96.1)
Chips 236 (87.7) 35(97.2) 201(86.3)
Cookies 218 (81.0) 30(83.3) 188(80.7)
Fresh fruits 43 (16.0) 36(100.0) 7(3.0)
Fresh meat 65 (24.2) 25(69.4) 40(17.2)
Fresh vegetables 60 (22.2) 36(100.0) 24(10.3)
Frozen meals 28 (10.4) 15(41.7) 13(5.6)
Frozen meat 15 (5.6) 13(36.1) 2(0.9)
Ice Cream 202 (75.1) 31(86.1) 171(73.4)
Meat alternatives 14 (5.2) 12(33.3) 2(0.9)
Milk 179 (66.5) 31(86.1) 148(63.5)
Milk alternatives 39 (14.5) 19(52.8) 20(8.6)
Pasta 124 (46.1) 31(86.1) 93(39.9)
Rice 139 (51.7) 35(97.2) 104(44.6)
Sugar-sweetened drinks 258 (95.9) 36(100.0) 222(95.3)
Vegetarian frozen meals 20 (7.44) 13(36.1) 7(3.0)
2.3.3. Benefits
One hundred and seventy-one of the outlets (64%) accepted electronic benefit transfer
as payment. All of these accepted CalFresh (the federal SNAP program), and 17 accepted
some other type of benefit like Special Supplemental Nutrition Program for Women, Infants,
and Children (WIC) vouchers and cards. There was no significant difference in the
availability of healthy and unhealthy food between SNAP accepting and non-accepting
outlets. Overall, I was surprised at how many outlets accepted some kind of benefit and
would be interested to compare this proportion with different areas in Los Angeles and other
cities across the U.S.

49
2.3.4. Sample Prices
Food products had a wide price range across all stores, with the range for the 10
sample products summarized in Table 2.3.
Table 2.3. Price ranges of sampled food products across the retail food outlets on the Eastside
Category Price range ($US)
Apples 0.60-2.99/lb.
Bananas 0.55-1.69/lb.
Chicken breast, 2lb 3.49-13.98
Eggs, 1 dozen 1.00-5.59
Frozen Lasagna 3.54-9.16/lb.
Frozen strawberries 1.49-5.73/lb.
Roma tomatoes 0.75-1.99/lb.
Tin black beans, each 0.89-3.49
Whole milk, 64 oz 1.29-4.99
Whole wheat bread, loaf 0.99-5.49
In general prices tended to be higher in smaller outlets like convenience stores and gas
stations. This was most evident when observing the prices of fresh fruit and vegetables – the
highest prices were in non-supermarket outlets and were accompanied by a slightly smaller
proportion of “acceptable” quality produce. Gosliner et al. (2018) previously found that
outlets with wider varieties of range and availability of produce (e.g., a large display of many
apples of different kinds), were associated with lower prices. This corresponds to the data
from this audit, where the small outlets with only very limited produce selection (e.g., a small
display of only one type of apple) advertised the highest prices. Outside of the distinction
between supermarkets and grocery stores (of which all 36 accepted SNAP), I did not observe
any significant association between price and benefit acceptance.
2.3.5. LILA Analysis
Of the 67 census tracts within or intersecting the study area, 33 are classified by the
USDA as LILA (49.2%) (USDA Economic Research Service 2022a). This classification was
the smallest radius of access measured by the USDA, at 0.5 miles in urban tracts and 10 miles
in rural tracts. Comparing this with Los Angeles County, where 472 of 2,495 tracts are

50
classified as LILA (18.9%) and the U.S. overall, which has 17,392 of 84,112 tracts classified
as LILA (20.7%) (ibid.), the census tracts in the study area were classified proportionally
more as LILA compared to the remainder of the county and the country.
Of the 269 outlets audited, 119 were located within the designated LILA census tracts
(44.24%). 255 outlets were within half a mile of one or more LILA tracts, leaving only 14
outlets (5.2%) further than that distance from a LILA area. All outlets are shown in Figure
2.2, displayed by location either within or outside of a USDA-classified LILA census tract.
There is clearly a strong network of existing food infrastructure in the study area, even within
some of the census tracts designated as only low access (i.e., not low income) by the USDA
model. Twenty-seven tracts classified by the USDA as LILA contained at least one outlet,
with one containing 10 and another 16. Eleven tracts overall, including five not classified as
LILA, contained no outlets within their boundary. Within the study area there were
differences in outlet concentration between the five neighborhoods, with more outlets located
in Boyle Heights and Lincoln Heights in the southwest and northeast of the study area,
respectively.
Using the established classification of supermarkets/large grocery stores, 14 were
located within LILA census tracts (38.9%), and 3 supermarkets/large grocery stores were
further than 0.5 miles from one or more LILA tracts (8.3%). These outlets are shown in
Figure 2.3. Four LILA census tracts had two supermarkets/large grocery stores within them,
six others had one each. The other 23 LILA tracts did not have a supermarket/grocery store,
with 17 containing at least one kind of another outlet.

51
Figure 2.2. Map showing 269 audited retail food outlets, classified by location within or
outside of USDA-classified LILA census tracts (Source: USDA Economic Research Service
2022a)

52
Figure 2.3 – Map showing 36 audited supermarket/large grocery stores, classified by location
within or outside of USDA-classified LILA census tracts (Source: USDA Economic Research
Service 2022a)

53
2.3.6. Kernel density analysis
The resulting raster surface from the kernel density analysis represents a greater
spatial resolution indicator of access compared to the diffusion offered by census-tract level
measures. Further, generating a kernel density allows for visualization of the commonly held
assumption in food research that the influence of an outlet is strongest in the closest areas,
with the influence diminishing with distance (Hallum et al. 2020). The resulting maps show
how the relative densities of non-supermarket/grocery stores and supermarket/grocery stores
vary over the study area.
First, there are clear concentrations of non-supermarket outlets visible in Figure 2.4,
which according to my audit carry disproportionately lower nutritional value foods (as shown
in Table 2.2). Comparing this with Figure 2.5, there are areas where lower nutritional quality
food outlets are much more concentrated than outlets where proportionally greater nutritional
value food is available. These two maps have been normalized so that cells with “0” values
represent the sparsest areas and cells with “1” values represent the densest, in order to
account for the different number of outlets in each category.
Note that these kernel density results (Figures 2.4 and 2.5) can be used as a proxy for
access, but still rely on Euclidean distance as in the LILA model. Zonal statistics can be
calculated for each census tract to revert to a standard areal unit, but this does reduce the
spatial resolution, as well as introduce edge effects, the ecological fallacy, and modifiable
areal unit problems to the analysis. For these reasons, the kernel density analyses are
maintained as unaggregated surfaces, apart from a resample to a hexagonal grid, reducing the
sampling bias generated by the higher perimeter-to-area ratio of square-shaped cells.

54
Figure 2.4. Kernel Density (hexagonal cell size = 5000m2
) analysis showing spatial
concentration of non-supermarket/large grocery store outlets

55
Figure 2.5. Kernel Density (hexagonal cell size = 5000m2
) analysis showing spatial
concentration of supermarket/large grocery store outlets

56
2.3.7. Service Areas
Figures 2.6 and 2.7 show areas with the same walking time to the nearest outlet (i.e.,
service areas) for supermarkets/large grocery stores and other outlets. Using the finding from
the audit that supermarkets/large grocery stores have proportionately healthier food available, I
thus identified areas where individuals have to walk much further to reach a healthier outlet
compared to an unhealthy one, supporting spatial assumptions about anticipated food decisions.
Overall, there is a much larger area which is within a five-minute walk of the nearest
unhealthier outlet compared to the area within a five-minute walk of the nearest healthier
outlet, and several areas where the nearest supermarket is well beyond a 15-minute walking
time, for example.
I also calculated these service areas as a proportion of the study area, to determine how
much of the area falls within each cut-off travel time, including service areas of stores with
specific attributes, such as EBT acceptance or availability of produce, shown in Table 2.4.
Nearly all (95.9%) of the study area was within a 15-minute walk of a non-supermarket/grocery
store outlet, compared to 72.6% within 15 minutes of a supermarket/grocery store. Over half of
the area (57.1%) was within a five-minute walk from any outlet accepting EBT, but under half
(40.2%) was within five minutes of an outlet selling any type of produce. A much greater
proportion of the area was a further distance away on foot from any store that sold produce – a
common indicator of a “healthy” food outlet (Lucan et al. 2020) – than any store which did not.
This kind of analysis extends beyond straight-line distance by examining travel across a
traversable network, important in this study area; an outlet may be a short distance away as the
crow flies, but the multiple transecting freeways pose impassable barriers when walking that
may result in a much longer travel time. These outputs can be combined with county parcel
data to show the number of households in each service area, also shown in Table 2.4.

57
Figure 2.6. Service areas for non-supermarkets/large grocery stores (n = 233), showing 5, 10
and 15-minute walking distance from each outlet

58
Figure 2.7. Service areas for supermarkets/large grocery stores (n = 36), showing 5, 10 and
15-minute walking distance from each outlet

59
Table 2.4 – Walking time service areas with 5, 10 and 15-minute cut offs for four different types of outlets
(Study area = 17.23mi2)

60
2.4. Discussion
2.4.1. Efficiency of the audit tool and comparison with existing research
Overall, this study found that using a tailored tool to audit retail food outlets in a
neighborhood was feasible and provided new and useful information about food access. On
average the researcher for this study was able to complete a maximum of around 20 records a
day, which should give some idea of how long this would take in a different study area
The conventional, long-from NEMS-S is accepted as generally providing reliable and
robust information for research, but at substantial cost: a median time of nearly 45 minutes
(Carins et al. 2019; Glanz et al. 2007). This is not feasible for many research applications; hence,
the common tendency to reduce the audit to only the elements pertinent to the applicable
research questions. The version for this study reduced the median time taken for the audit to
under 10 minutes for all types of outlets. This was likely because of the uncommonly broad
definition of “retail food” adopted here. The limited range of food at gas stations and liquor
stores was to be recorded on the same instrument as well-stocked large chain grocery stores. This
required a degree of compromise, in order that relevant and useful data be collected across this
large range of outlet type on the same platform.
Whilst Partington et al. (2015) found comparable results when reducing the length of the
tool, Carins et al. (2019) used the same reduction but could not identify a similarly acceptable
equivalence. This non-equivalence in the two previous studies – one where the long- and shortform NEMS-S scores were comparable, and one where they were not – means that assertions

61
based on the observed dimensions of price, quality, and placement of different food in stores
must come with the kind of reflexive acknowledgement of methodology given in this study.
Knowledge claims related to the food environment are limited by the understanding of any of its
features and social processes. That is to say, the validity of any blanket claim is limited by the
lowest common denominator, or the least well-understood aspect. This hurdle is important to
consider in systems-based (here food systems-based) approaches, which rely on, by definition,
relational and joined-up thinking. Such an epistemological barrier often presents problems when
developing meaningful conclusions in health and social research. Troubling assumptions become
magnified, and when parameters for a model are compiled, these assumptions leave a trace
throughout entire research frameworks.
2.4.2. Adapting the NEMS-S and implications for future research
The amendment of the NEMS-S, through reduction in some areas and expansion and
addition in others, affected the results and conclusions of this study. Considering the reductions
first, it was originally determined that the same instrument would be used to record observations
in all the different types of outlets. This meant that the same audit was completed in gas stations
as in health food stores as in chain supermarkets. Of course, these are distinct retail and
nutritional environments in and of themselves, so aiming the same measurements at them all has
resulted in lower resolution results. If this study had only audited convenience stores, then in the
time available a much finer and more specific detail could have been captured with the same cost
in time and resources. In this case, “bespoke” meant capturing high-level summary indicators for
food availability across many types of outlets, rather than a granular account of the contents and
composition of food and how that changed from one store type to another.

62
2.4.3. Alternative food networks
Certainly, the study design necessitated that some critical food outlets were excluded
altogether. Notably farmers markets were not recorded, and neither were any of the night
markets which these neighborhoods are renowned for. Also left out were the various sites for
food assistance, such as food banks and pantries. Equally, community resources like gardens and
fridges were not recorded. All of these outlets contribute to the food environment, in what have
been termed alternative food networks (AFNs), and all take place at a discrete location which
could have been visited during the study. This was a choice in the research design, on the
exclusion criteria for food outlets to audit, and represents a limitation on the potential to
understand the food environment in the study area. Future research should seek to include AFNs
in their analysis, despite the difficulty involved in auditing them in the same way as conventional
food networks, as they make up a critical and frequently overlooked source of food.
AFNs are, to varying degrees, removed from the harbingers of the dominant capitalist
agri-food system: profit maximization; extended chains between producer and consumer; and the
productivist paradigm to name just three. Whilst many would agree that holistic descriptions of a
food environment are vital for understanding and recommending interventions for health policy,
this gap poses a perhaps insurmountable challenge for this methodological iteration. It then
becomes necessary to ask what tool could allow a researcher to measure and record every node
in a food system. In the structural competence worldview, the singular tool would lead to the
development of a singular language to describe the heterogeneity. Instead, I propose that
researchers account for the relational food environment, and sacrifice the commitment to
universality for a richer, more situated discourse. As it stands, the NEMS-S is suitable for

63
auditing stores – for which it was designed – but unsuitable for AFNs. The detailed and varied
results presented earlier show that the audit tool can be successfully deployed to describe the
conventional sectors of the food environment, but I argue that it should be one part of a
multimodal approach, rather than the only part of a universal approach.
2.4.4. Food assistance programs
The expansions of the NEMS-S in other areas did lead to valuable and insightful results.
Of particular interest in these neighborhoods was the high acceptance of EBT, or other food
assistance. This was higher than expected, yet whilst this should aid financial access to groceries
for qualifying low-income residents, no assumption can be made that this environmental effect
has led to greater food security. Tiehan and Frazão (2016) estimate that 84% of SNAP
redemptions were at the types of outlets included in their conventional model, which according
to the audit, includes only 36 outlets in the study area (13%). Notably there were no
“supercenters” (e.g., Walmart, Target etc.) or warehouse/discount clubs (e.g., Costco, Sam’s
Club) in the study area, which are effective outlets to maximize the value of payment by EBT:
the benefits are administered monthly, and many recipients spend them quicky at these large
retailers (Widener and Shannon. 2014) to take advantage of lower cost-per-unit prices. Over the
course of the month though, many individuals may come to rely on “top ups” from convenience
stores and the other types of outlets in the study area. More research is needed on these
interesting time geographies, again likely as part of a mixed methods approach.
2.4.5. Temporal resolution
One useful example of methodological triangulation in this study was related to the crosssectional design of both my audit and the LILA model and the importance of timely data. The

64
USDA data are only updated every four years, with approximately a two-year delay in making
the data publicly available (USDA Economic Research Service 2012). This delay in modelling
creates a temporal gap in research, which is concerning during rapidly developing crises like the
COVID-19 pandemic. Specifically, regarding the food environment, research and news reports
from the early stages of the pandemic showed changes in food access, such as reduced opening
hours of grocery stores (Jingnan 2020); sold-out food staples (Markowitz 2012); and grocery
store closures (Juarez 2021; Tyko 2020). Food insecurity increased, quite drastically, in many
regions of the U.S. (de la Haye et al. 2021; Dubowitz et al. 2021; Fang et al. 2022; Mui et al.
2022). Yet research on how LILA areas changed throughout the pandemic, and the experiences
of populations living in them, was limited because of the lack of a responsive, regularly updated
model of the food environment. While the USDA model was updated in mid-2021, this update
incorporated data from the 2014-2018 ACS and the 2019 Trade Dimensions TDLinx – thus did
not reflect local changes in employment and income or any stores that closed, temporarily or
permanently, during the pandemic. My audit, whilst only over a small study area, did serve as a
relatively quick review of the state of retail food – one that could be repeated more quickly and
frequently than the less specific LILA model.
2.4.6. Culturally appropriate food
One initial area of interest for the research team was the availability of culturally
appropriate food. This harks to the preliminary definition of food security offered at the 1996
World Food Summit, which stipulated that alongside sufficient, safe and nutritious food, it
should also “meet their… food preferences” (Food and Agricultural Organization 1996). In a
study area with a predominantly Hispanic/Latinx population, it was determined important to
undertake some observation of culturally appropriate food in the retail environment. Early on in

65
the audit process, however, it became apparent that the conditions for any one type of food to be
especially “appropriate” to any one culture vary greatly, and it was challenging for a singleperspective audit to account for this. Certainly, this is an area that the original NEMS-S is not
designed to address. It should be a priority for food scholars moving forward to develop a multipurpose tool which can account for the multiple and diverse interpretations of “culturally
appropriate” food and access thereof.
2.4.7. 15-minute cities
The spatial distribution of retail outlets and the resultant assumptions of access,
particularly the service area analysis, are similar to “15-minute city” approaches as in Allam et
al. (2022). In this study I calculated service areas around each outlet, but in conventional 15-
minute cities approaches the analysis is reversed; the quantity and range of amenities are
measured in a travel radius around home addresses. In a strict sense this approach does still rely
on the amenity-proximity paradigm of conventional food and nutritional security models, future
work that combines this kind of analysis with the detailed store attribute data will allow
researchers to draw stronger inferences from the spatial information. This was the most
important finding in the study in this regard, that 95.9% of the study area and 51,984 households
(99.8%) were within a 15-minute walk from an outlet with disproportionately unhealthy food,
and that substantially less of the area (72.6%) was within the same radius from an outlet with
proportionally healthier food. If access research continues to focus on distance to nearest healthy
outlet, which has been shown to be a significant determinant of diet, then it should make use of
service area and 15-minute cities approaches. The “low access” tracts identified by LILA models
obscure important environmental features and create misleading ecological fallacies. Advanced
methods from the spatial sciences, like kernel density estimates and other interpolation

66
techniques, are powerful and useful in food research, but until very recently have been relatively
under-applied (Hallum et al. 2020).
2.4.8. Relational and affective approaches to food environment studies
For the spatial sciences to play a role in structural competency, its practitioners must
acknowledge the legacies of imperialism and capital accumulation which can be traced
throughout the discipline. In the food context, there has been a tendency to shift the locus of
control onto the individual or the household – an “individualization” of hunger, which shifts the
analytic focus away from the structural inequalities of production and consumption (Sonnino et
al. 2016). Instead of using geospatial analysis to map the acquisition of capital, as it has done so
often before, it should be used to visualize the political and socioeconomic intersecting
constraints that produce and maintain poverty across space and time. Rather than corporation-led
efforts to map the globalization of food as a commodity, the spatial sciences should present the
concomitantly global and local systems which Metzl and Hansen (2014) describe as either
“central arteries” in the health of some areas and “sclerotic” in others. Attention to critical
geographies is paramount here, particularly when dealing with the diverse trans-local networks
which are interchanging food and knowledge between communities. This is another rationale to
incorporate AFNs into future work, as they frequently cut across scale and place as diverse social
movements which create relational networks of resistance (Sonnino et al. 2016). Fundamentally,
food security research should be undertaken across and between scales, rather than focusing on
fixed units like household, census tract, or food policy council jurisdictions. Such relationalism is
typical in human geography, and it transcends the false dichotomy of global and local to
foreground the mutual constitution of sometimes distant places (Massey 2004). This may
provide an answer to the question of why the nutritional needs of Eastside residents are

67
continually unmet despite the surprising plenty of healthy food that I found, and how to modify
public policy to match the specific demands of differently situated neighborhoods. This
challenge is not specific to the study area: two centuries of productivism has meant the world is
in its largest ever food surplus, and yet billions remain food insecure and subject to the stuffedand-starved conditions of diet under capitalism.
An interest in the study of the changing processes that human bodies undergo as we
experience, encounter, and perform within food environments is growing with a resurgence in
interest in psychological and emotional geographies. Affect, which prioritizes the body as a
means for making sense of the world, is apt here, to address and examine the states that combine
sense and perception for rendering the world intelligible. The record produced in this study is
appreciated as participatory; several weeks spent walking, seeing, touching, eating, consuming,
and getting a sense of the food environment. This is difficult to quantify by any standard means.
The detailed picture of the retail food environment that resulted is invaluable, lending
authenticity to the claim of any adequate collective understanding of the situation on the
Eastside. As well as practical, non-quantitative, insights from the experience, it creates the
ferment for an animated supplement to existing approaches to food systems knowledge
production, here and elsewhere tending to overemphasize the representation of food
environments as a collection of undifferentiated entries in a secondarily sourced dataset. Affect
should be important in mixed approaches, as a concept that transcends disciplinary boundaries
(Gregg and Seigworth 2010).

68
2.4.9. Food environment-health behavior nexus
It is important to discuss the implications of ground-truthing the food environment for the
preparation of suggestions for such intervention. Absolutely, the study of neighborhood-level
physical environments is important when attempting to understand the ecologic framework of
food systems (Cannuscio et al. 2013). The diversity and proximity of retail food outlets and the
different products and their prices is widely accepted as a factor in the neighborhood health
(Walker et al. 2010). Emphatically though, the food environment alone does not explain the
health of individuals, families, and communities. Consider the contradictory findings of some
studies which document the reduced health of populations who live far away from supermarkets
(Hendrickson et al. 2006) or close to fast-food outlets (Mackenbach at al. 2019), with others that
identify no spatial relation at all (Ghosh-Dastidar et al. 2017). Now picture the multi-milliondollar homes in the winding hills of Los Angeles’ Westside, just a few miles west of the study
area, many of which lie far outside of the strict 1-mile radius of the nearest chain supermarket
specified in the USDA LILA classification. Should I classify these households as “low access”?
Likely not, based on a holistic appreciation of spatial and non-spatial features of that particular
neighborhood’s food system. An adequate understanding of the food environment, likely built on
the kinds of methods employed by this study, must be paired with a deep appreciation of the
geographic features and character of different places. Universal-level rules and classifications
will never be able to account for the social and historical contingencies which have formed
contemporary environments like the ones in Los Angeles, influenced by so much more than
Euclidean distance from nearest supermarket. Equally, the consideration of actions within the
simple retailer-consumer framework, must be expanded upstream to account for the powerful
forces of productivism under Western agriscience and profit-maximizing corporate interest.

69
Within the simplified framework, the only solutions which can emerge are confined to efforts
that preserve the existing system, like the fairly recent arrival of the logically inconsistent
“sustainable intensification” paradigm (Garnett et al. 2013). Interventions which are confined to
agriculture, production, and trade – the conventional areas of study for many food systems
scientists – serve only to expand the productivity gains of industrialized food into new markets.
At a minimum, this article demonstrated the use of an audit to confirm and clarify the actual state
of the food environment on the Eastside. For researchers undertaking similar research in the
future in other neighborhoods, they should diligently address the paradigmatic perception of the
lack of access in neighborhoods classified as “food deserts”.
2.5. Conclusion
Despite mixed evidence, the influential literature and rhetoric surrounding LILA areas
and food deserts has inspired a mainstream politic blithely aimed at decreasing the Euclidean
distance between home address and supermarket. This focus on amenity proximity lacks nuance,
focusses on a metaphor of scarcity which this study has proven inaccurate, and ignores the
social, cultural, and behavioral factors informing nutrition and diet. Assuming that the physical
food environment is the sole or most important determinant in health outcomes is a fallacy.
There is no disadvantage to ground truthing the food environment if a nuanced, accurate, and
detailed understanding is sought, but the research design must appreciate the generated data as
one pole in a study area requiring simultaneous plurality of knowledge. The power of one kind of
conventional method was shown in the results section of this chapter, demonstrating the kinds of
insight that can be produced by advanced techniques from the spatial sciences. Through the

70
discussion though, it became clear that the methodological power was compounded when used as
part of multimodal approaches.
Undertaking the study in the wake of the pandemic, alongside a renewed academic
interest in health equity, there is one pertinent observation. Whilst COVID-19 did not start the
structural processes of nutritional inequality, it absolutely helped to reveal it. Neoliberalism was
proven to have only one charge: the extraction of value and return of investment to shareholders,
even amongst mass death. The low-income, minority ethnic populations on the Eastside of Los
Angeles represented such value: providing the lowest cost, lowest quality food, in exchange for
maximal labor-power. The intervening failures of the global food network were present on the
streets and on the shelves of the five study area neighborhoods, where it felt especially clear that
conventional systems were unable to meet the changing demands of the population. The surging
demand in food banks and other food assistance programs is evidence of this.
This article outlines some of the lessons gleaned from completing a “gold standard” audit
of the retail food environment in five neighborhoods (El Sereno, City Terrace, Boyle Heights,
Lincoln Heights, and Wellington Heights) on the Eastside of Los Angeles. It discussed the
consideration of why and how to adapt existing research tools for use in specific contexts, here a
variation of the NEMS-S (Glanz et al. 2007). In place-based health and social research,
maintaining attentiveness to the place itself should be paramount, and any adaptations should be
justified and evaluated from the outset. For this tool, it was suggested that reduction in one area,
even if there is expansion in another, would serve to limit the understanding of the food
environment by reducing the lowest common base of understanding. Here, the inclusion of

71
acceptance of supplemental nutrition assistance program benefits (e.g., CalFresh) and the
collection of spatial information alongside the store audits allowed for novel insights into the
food environment and laid a foundation for further powerful enquiry using a GIS. For the many
scholars seeking to examine the food environment-health outcome nexus, a ground-truth audit is
a powerful tool, but cannot itself serve as a key for absolute understanding. Food security as an
outcome of governance is contingent on the convergence of different interventions and policy
decisions, all of which should be based on evidence as accurate and sensitive as possible. This
study represents one way of measuring the food environment using an adapted set of existing
methods and techniques, providing a useful start to building this base of knowledge.
The aim of this article was not to impose particular views on which methods or suites of
methods are “better” for place-based research. Instead, it attempted to show that doing research
can never be separate from worldviews – epistemologies – or from the methods which are cocreated. All methods are situated in social relations, hinging on Haraway’s (1991) notion of
partial objectivity. The case for engaging in mixed methods work for understanding place-based
health like food environments and systems is clear, perhaps imperative. The experiences of life
in a low-quality nutritional environment have filtered down from race and imperialism as the
grounds upon which capitalism and its system of subjugation in its entirety have been built.
Engaging with food in this way inherently demands grappling with such logical telos, hence the
emphasis on structural competency. Defining, approaching, and ultimately solving food
problems all depend on different, often incompatible, responses, which contend themselves on
the context and worldview of the affected agents. Considering these notions, seemingly innocent
questions like “how do you feed a city?” or “what does food security look like?” quickly become

72
inconceivably complicated. Disentangling the socioeconomic factors from the geographic – for
example, to propose interventions – cannot justifiably be performed from the platform of
granular quantitative data. Decision making and relationships with food and nutrition are
complex, individual, place-based, gendered, racialized, socially and culturally produced,
dynamic, and temporally and spatially heterogeneous. So why not model food systems like this?
Conventional, singular-method approaches survey the food environment without critical
reflection, leading to the promotion of inadequate solutions. Without structural competency,
rather than steps on the convoluted path towards structural transformation, the solutions are
perceived as endpoints (Montenegro de Wit 2020). It becomes easy to mistake progress with a
little less abuse (ibid.). Remember the meatpacking plants in 2020, kept open by executive order.
The delivery of masks and other personal protective equipment designed to reduce death was
celebrated as a success; tools to keep the productivist food system running. Nobody involved in
the decision-making was asking if the system even made sense. The same is true for food
security studies. The only treatment that can exist for LILA according to its own classification is
to place another node in the capitalist food network in the census tract, which provides great
benefit to large scale landowners who continue to gain from the commodified system of food as
in any other. The audit in this study proves that this is not the only, or even most appropriate,
possible intervention. Instead of questioning the precision and manners of technique, a
structurally competent food scholar should probe the structural components that have led to the
current condition. Not the food-supply questions like “where are the food deserts?” or “where is
the nearest supermarket to a home address?” but questions which reflect the knowledge and
methods problems like “how do conventional food systems exploit class-, gender-, and race-

73
based relations to architecture networks of exploitation and capital accumulation?” “What are
the crucial components that transform relations in food systems?” and “What is the role of place
as a site for the performance of food systems.” Reese’s (2019) Black Food Geographies provides
an exemplary model of this kind of research process, albeit absent of the advanced spatial
methods which I deployed in the measurement of the retail food environment.
In sum, this study has made significant contributions to an understanding of the food
environment and its potential implications for health outcomes on the Eastside of Los Angeles.
By conducting a comprehensive retail food audit and employing a mixed-methods approach, I
have shed light on the complex dynamics of food access and highlighted the limitations of
conventional approaches such as the LILA model. The findings have shown that access to
healthier food options is limited, with disproportionate concentrations of non-supermarket outlets
offering lower nutritional value foods. This study calls for a more nuanced understanding of the
food environment, one that considers the social, cultural, and historical factors that shape it.
Future research should aim to incorporate alternative food networks, such as farmers markets and
community resources, into the analysis to capture a more comprehensive picture of food access.
Additionally, further investigations should explore the intersecting constraints of poverty, race,
and imperialism which contribute to nutritional inequality, and develop interventions that
challenge the underlying structures of capitalist food systems. By adopting a structural
competency approach and fully embracing interdisciplinary perspectives, transformative change
to address the complex challenges of food security and health equity in diverse communities
becomes possible.

74
Vignette 2 Buses and neighbors
As a Londoner in Los Angeles, I often think about my attachment to this place. Almost as soon
as I moved here, people were asking me “do you think you’ll stay in America or go home after
you’re finished?” For me, the answer revolves around the concept of situated cognition, which as
a geographer, is fundamentally a proposition of space and place. Of course, there are corporeal
and more-than-physical aspects of each city, Los Angeles and London, which have contributed
to the formation of my identity at different points in time, just as there are for any transplant.
Different people attach to different themes, but I posit now some large catch-all categories:
relationships; consumption; sensing; and movement. These are entangled and mutually
implicating, bound up in the power-laden and recursive production of identity across space.
Consider movement. In London, I only had a car for a few months, only driving a white
1992 Peugeot 205 which had belonged to my great grandmother after I was in a position to be
able to afford to keep it running. Moving through the city in the car completely changed my
relationship to the place I had lived in my entire life. Instead of coming and going according to
the rhythms of the bus, train, and tube, I was suddenly able to set my own schedule, pick up my
friends – all of whom were carless – and go to a new set of places made available by my new
mode. In the end though, keeping the car in London resulted in a huge inconvenience. Its huge
two-liter diesel engine was thirsty, I had to pay the twelve-pound congestion charge every time I
wanted to drive into the middle of town, I kept forgetting to add a daily parking permit on the
council website which meant huge fines, and there was always so much damn traffic. After a
while the novelty wore off, and I drove the car back to my granny’s field in Suffolk, relieved of
the burden of car ownership. I had commuted for seven years on the London Underground and

75
knew my part of the network like the back of my hand. When the chassis of the trains “looked a
bit rusty,” it wasn’t me who had to pay for a patch-job, and to be honest, the seats were probably
comfier anyway. My father listed the car on eBay, including photos of the oxidized axles, and
sold it for three hundred pounds the summer I moved to California.
I made a three-month attempt to live in Los Angeles without a car. Contrary to what had
been reported to me, I thought that the light rail was actually pretty great, especially for the $1.75
flat fare. There were some clear drawbacks to the network, notably that it didn’t have anything
near full coverage, with large quarters of the city cut off from rail access. The buses filled some
of these gaps and were great when you were on them.
The trouble was their unpredictable schedule, seemingly ungoverned by any regularity,
and that the “estimated arrival” times on the transit apps were unmatched by reality. I ended up
using a bike for a lot of trips, and enjoyed the ease with which I could roll it onto the train or
hook it onto the bus-front racks, neither of which are possible in London. By October I relented
and bought an electric Fiat 500 from one of the expansive dealerships in West Covina. I
remember driving back into LA down the 10, excited at the new mobilities which I had
“unlocked”, but slightly trepidatious having joined the fray as a driver in the most famous trafficcapital of the world.
I noticed two things straight away. The first was that I was already used to the roads and
routes around the university campus and downtown. I thanked my time on the bicycle, for which
I still think there is no quicker way to learn a city’s geography. The second was that I was

76
already disagreeing with the wayfinding proposed by Google Maps. It had a preference for
freeways which didn’t do well for the car’s modest battery life. I quickly adjusted my journeys to
prioritize shortest driving distance rather than quickest time, dictated by range and charge
conservation. My movements through Los Angeles were a product of my situated mobilities. I
felt like an active traveler, rather than a passive passenger subjected to the bus route or the
Google Maps driving directions.
Mobilities in Los Angeles are peculiar. Like many other phenomena in the city, they are
highly moderated by social differentiation: youth; race; refugee status; and sexuality; to name
four kinds. I almost never see white children walking on the street, thanks perhaps to their
nannies ferrying them to school in private cars. When I do get the bus now, which I do most
often to avoid traffic and parking on my way to sports matches, the abstraction of American
culture is very quickly brushed aside by the olfactory and audio-visual encounters. I think of
Marisela Norte’s poems, triggered by her acceptance of the city through open eyes, ears, mind
and heart, written by observing – or as she says, eavesdropping – las vidas de ellas. The lives of
working women, destined to move through the city by public transport, animate her writing, all
done from the vantage of different seats on the bus. When she reaches her stop, the writing is
complete. In the words of George Lipsitz, Norte’s buses became sites of containment and
connection, incarceration and affiliation, solitude and sociality. Every time I leave my car behind
to move through Los Angeles, I come across so many people who have been themselves left
behind by the remainder of society, drifting from bus to bench to train to kerb. I remember them
when I’m writing, or being welcomed into institutions where they are prohibited from entering. I
try to let these encounters inform the continuing shaping and reshaping of my own identity.

77
One such tangential encounter was with one of my neighbors in Montecito Heights on the
Eastside. I had been out for a walk one evening and got talking to a dogwalker, Tom, about a
chicken coop that he was building on the weekends. I said I’d gladly take a look and weigh in,
having built several iterations for myself and others, and followed him down his garden path. I’d
walked past his front gate hundreds of times, but didn’t realize how vast his lot was, or quite how
big his house was, both hidden from the street. He said that he had bought the land in the 1980s
for $25,000, built the house with some local contractors, and reckoned that it was now worth
about $3 million. He walked me around his property, showing me where the chickens would go,
to which I voiced my approval. By the time we’d gone round the entire garden, it was pretty
dark, and he told me to follow him to the front gate, where it would be easier to get back to my
house. He said that he’d better go inside and tell his wife, just so she wasn’t spooked by a
stranger walking past the tall windows on the back of the house.
Tom popped his head out to invite me in, as his wife wanted to meet me. Tom was a tall,
white Texan, and his wife, Haina, was a tiny Latina, both in their late 60s. I introduced myself to
Haina as a neighbor, and pointed out the big tree in my back garden which you could see across
the hillside. They moved here when they were very young, after marrying in Mexico. She told
me that she had never been taught how to drive, and spent almost every day up on the hill,
sending her husband down to get groceries or pick up prescriptions. We then had a quite
extraordinary conversation, standing around their kitchen island, which I will try to recollect.
“What do you do?”
“I’m doing a PhD at USC.”

78
“At USC? So, they are brainwashing you?”
“I don’t know about that…”
“They’re sending you to protests for Hamas? And they do the critical race theory?”
This went on for a little while, with a string of other questions about gender politics and
immigration. I didn’t want to act out the culture war there and then, so deflected a lot of the
questions and tried to edge towards the back door. Tom remained silent in the corner.
“Where are you from? Your accent is British?”
“London!”
“Oh I love London. We’ve been there twice.”
“Amazing, did you like it?”
“I loved it the first time, in the nineties. Everyone was so polite, and beautiful. Pale,
Viking looks. And the manners! It was so wonderful.”
“Oh really?”
“But then we went back before covid. It was so bad. Too many immigrants. There were
Muslims in the street confiscating alcohol, because of sharia law. It was disgusting, I feel sorry
for you.”
“I don’t know about any of that.”
“It’s true! What do your parents think?”
“My dad loves multiculturalism, he loves-”
“That’s not multiculturalism. That’s colonization. You have been colonized.”
I left with an odd feeling. I had seen that there was some rolling news channel on the TV
in the kitchen where Haina had been preparing their supper. But the stuff she had wasn’t from

79
normal broadcast news. I wondered where she was getting this ersatz information from, about
brainwashing on college campuses and a new era of faith-driven prohibition on the streets of
London. I charted the encounter up to experience, reflecting on the paradox that Haina had
clearly travelled to Europe – she told me about her impressions of Paris (“full of North
Africans”) and Rome (“dirty” and “criminal”) – but because of her carlessness rarely ventured
down from Montecito Heights into the rest of Los Angeles. I wondered what she talked about
with her husband, or whether they talked much at all. If I was the first person who had been over
in a while – Tom did take special care to warn her that I was about to walk past the windows – or
whether she had similar things to share with every visitor.
As anecdata, it’s proven to me the great range of experiences which different people can
have in the same place, a micro ethnographic cross-section. I haven’t been back to see Tom and
Haina’s chicken coop, which is probably finished by now, but I hope they’re doing well. I think
about her when I’m watching the news, wondering what version of events might be recounted
around the kitchen island, looking out of the windows over the city. This place is full of people,
an assemblage of 18 million identities, rooted in complex and usually contradictory experiences.
How all these people are moving, at what speeds, and where, make a huge difference in the
representations and meanings of the city at the individual and societal levels. I was incredibly
excited to be given a platform to empirically examine the geographies of mobility on the
Eastside, and to build my own vision of an affective geographic practice.

80
Chapter 3 Affective geographic information science: Using carto-photo
elicitation interviews to examine the connections between mobility and place
amongst the carless in Los Angeles.
Abstract
Background: This study presents a novel approach, carto-photo elicitation interviews (CPEI),
combining participatory spatial and visual methods to investigate the dynamic intersections of
mobility and place in the contemporary metropolis. As action research, it recasts the network of
knowledge production away from the conventional power relations of social research, outlining
how subject-elicited photographs and location traces could creatively depict new configurations
beyond their configuration as representational or illustrational.
Methods: Participants (n = 7) were drawn from the Eastside of Los Angeles and did not have a
personal vehicle at the time of the study. Using smartphones, they collected geo-tagged images
and location data (using Google Location History) over a period of 1-3 months. After data
collection, participants were interviewed using the resultant images and maps as prompts for
enquiry into the affective, psychosocial, and corporeal experience of life without a car in the
auto-dependent urban, including issues of mobility, housing, mental health, and class and
belonging. The materials; spatial, visual, and textual, were analyzed separately using
conventional techniques of exploratory geospatial, content, and framework analysis respectively,
before holistic interpretation of their interrelations.
Results: The results of the study were a set of eight thick descriptions which circulate reference
between the dissimilar ontological varieties of experience and account, underscoring the
significance of situated and place-based knowledges in studies of place and mobility.

81
Conclusion: The outcomes of the study demarcate the boundaries of effectiveness for CPEI,
offering insight into the nature of mobility and place attachment in an uneven city, and highlight
the method’s potential for generating affective knowledge.
Keywords: Los Angeles, carto photo elicitation interview, mobility, place attachment, affect
3.1. Introduction
In the wake of the cultural turn, the focus of human geographers progressively shifted towards a
spatial exploration of behavior and emotion. This extended to the contingent social processes
which produce geographies of affect: identity; change; and conflict (Connell and Hilton, 2015).
Hall (1997) suggested that in scrutinizing the central object of the turn – culture – our primary
focus should be on the “production and exchange of meaning between members of society.” This
entails interpreting what is around us, developing ways of “making sense” of it, and interpreting
those ways in turn. Through this conceptual framework, investigating affect and its cognate
geographies transforms into an epistemological endeavor, emphasizing the processual dynamism
of knowledge production and the exchange of meaning between people (Anderson and Harrison,
2006). This study explores these ideas in the context of a participatory research project,
evaluating a new method and its capacity to render intelligible the experiences, encounters, and
performance of movement and sensing in the city through the generation of thick descriptions. It
prioritizes daily existence in space as a first order phenomenon, asking how place attachment is
informed by mobility, and how it can be examined and made sense of for problems of social
research.

82
3.1.1. Visual methods
Meanings are frequently made through representation, and it is these representations
which structure the expressions and articulations of people’s behavior and lives (Rose, 2012).
Many of the meanings which constitute the discourses of consciousness are produced and
mediated by visual imagery, all of which can be understood as constructed through socially
formed codes of recognition (Bal and Bryson, 1991). Investigation of images and the practices
and knowledges involved in their production and construction, through collectively termed
“visual methods”, have been performed in many different applications, from child development
studies (i.e., McCloy et al., 2010) to accounting and management organizational research (i.e.,
Davison and Warren, 2017).
Another effect of the cultural turn was a rising concern with the power differential
between researcher and participant (McDowell, 2016; Rose, 1997; Veroff and DiStefano, 2002).
Geographers have been attentive to this in designing new methods for social enquiry, testing
different ways of flattening the hierarchical relationship between “expert” scientist and “lay”
public. Practitioners of visual methods are at the forefront of this movement, using photo
elicitation interviews to generate materials. Based on the simple idea of inserting a photograph
into a research interview (Harper, 2002), it has become a widely used method across a range of
disciplines (Rose, 2012). Where their perspectives may have been previously underreported as a
matter of course (Sen, 2000), photo elicitation interviews bring participant perspectives to the
forefront literally. They enable participants to decide themselves on which parts of their life are
important to capture and make visible, then determine in partnership how those representations
are analyzed.

83
3.1.2. Place
This study takes the concept of photo elicitation interviews and combines them with
another established practice, in this case from the spatial sciences. As well as photographs, it also
elicits location from participants, by tracing the GPS pings of their mobile phones during the
study period. As a study investigating questions of social difference and the power relations that
sustain them, this multimodal approach is unique, triangulating new knowledges of place
attachment that go beyond conventional static categories of analysis. The mix of visual and
spatial methods is apposite, based on Rose’s (2012) theory that the actions of creating and seeing
images cannot be separated from cultural practices. What this study offers is an analysis of these
cultural practices, rooted in a discourse of place and place attachment. It draws on
Swyngedouw’s (2004) insights into geography’s role in delineating the material and experiential
characteristics of urban life. Adopting a place-based framework, seemingly incidental events,
such as the recording of GPS signals over 4 weeks, gain significance in conveying essential
aspects of culture, community, history, neighborhood change, and human relationships (Fratini
and Jensen, 2017). The central goal for this kind of enquiry is to explore how and why
individuals within the same city encounter markedly divergent experiences, despite sharing many
of the same infrastructures and activities.
The recent interest in place as a critical mediator in human well-being is by now a wellunderstood pathway (Kemp, 2011). Unlike the biological markers typically analyzed by natural
and medical scientists, scholars of place attempt to account for the infinite variation in the
construction of conscious and subconscious experience of the social world. Conventional
techniques tend to meet the fluidity and unpredictability of reality with a set of concrete and

84
predetermined collections and manipulations of data. This is the most commonly understood
definition of method in science (Law, 2016). Using a single method to explore something as
elusive as humans’ connection to place has been shown to be limiting. In fact, conclusions from
single-method enquiries may be so narrow in design that they have worked to obscure rather than
disclose the link (Zohrabi, 2013). This is not to say that the methods themselves are wrong, on
the contrary; hundreds of years of science has resulted in a powerful set of tools through which
parts of the world can become very accurately known. Instead, normative attachments to method
within the study of the definite and the regular have combined with a tendency to focus on the
minute ‘nuts and bolts’ of technique (Kwan and Schwanen, 2009), resulting in stagnating
practices bearing increasingly diminished resemblance to the realities which they attempt to
uncover (ibid., Law and Urry, 2004).
3.1.3. Spatio-visual methods
The economies, rules, disciplines, and histories of place exert considerable influence on
how a particular image is apprehended (Rose, 2012), and different disciplines have scrutinized
these specificities through their respective lenses. The methods of spatial scientists and visual
researchers have not overlapped in this regard, with neither formally nor substantively
incorporating the affectively productive aspects of the other. Yet, every image is intrinsically
bound to a set of tangible sites: the location where it was captured; the location of what is being
photographed; and the site where it is viewed.
Consider a picture of the Hollywood sign proudly displayed in a tourist’s home on a
different continent, thousands of miles away. Whilst the subject of the image is instantly
recognizable to many, the significance of where it was captured, is not. Many people’s first

85
sighting, and often first photograph, is from the 105-freeway leading away from Los Angeles
International Airport, nearly 25 miles due south. The freeway conjures a very different set of
memories and emotions, completely apart from the popular imaginary of the tall white hillside
letters. As well as the spatial coordinates of their sites, there are an accompanying set of
intangible phenomena attached to every image. The connection between people and place is
vague, diffuse, unspecific and unpredictable, and its study has not always led to singular
answers. Researchers of place must acknowledge this complexity, learning new ways to know
the realities of the social world which are unusual to, and often unknown in, conventional modes
of research (Law, 2004).
This is one reason why spatio-visual methods are so well-suited to understanding urban
life in Los Angeles: the visual culture of film and television is integral to the city’s landscape.
Benton’s (1995) critique of three films first described the differences between Los Angeles’
“real” and “reel” representation, in a narration of the effects of urban instability and chaos on
representations of the city. The fabrication of the city as a set of images – more people have
“visited” Los Angeles through imagery than any other city in the world (Soja, 1989) – belie an
enigmatic subjectiveness, with its representations often embodying fragmentary geographies of
ambiguity and change, concealing as much as they reveal. So, visions of Los Angeles are
inherently paradoxical, provocative intertwinements of utopia and dystopia, brilliant sunshine
and noir decadence, opportunity and danger, optimism and despair (Soja, 2014). So, although the
reel city, emblemized by so many photographs of the Hollywood sign, is a sociocultural myth,
the representation is critical to understanding the reality. Through a concerted effort to weave the
notion place into this discourse, this study harnesses the visual symbolism of the locale to

86
elucidate its capitalist urbanism, firmly rooted in the places and first-hand lived experiences of
its inhabitants.
In this study, I outline the development of a new method, the carto-photo elicitation
interview (CPEI). For their part, images and GPS traces provide excellent representations. The
information stored in the pixels which make up photographs and maps is rich, and open to all
manner of interpretations. As they stand, the only possible interpretation of the materials is
representational; made intellectually manageable for purposes of compatibility, standardization,
and universality (Latour, 1999). This stands in contrast to the facets of place already outlined:
locality; particularity; and materiality. By centering the phenomenology and corporeality of
imagery and location, I really tried to inspect the perceptual, experiential, and sensory
understandings of meaning in my encounters with participants. Thrift (2008) describes this as
nonrepresentational theory, inspired by the language and methods of Deleuze’s (1988) attention
to the spatial and visual structures of cinema. Moving past the insights made possible by
analyzing pixels is interpretively demanding, and perhaps aimless. The ultimate goal of CPEI is
not to perfectly interpret what the materials mean, rather to find a richness and vitality in them
that helps to relay experiences that are impossible to account in text, or even in language itself.
Marks (2000) describes the materials as connective tissue between things. CPEI ventures beyond
representation, seeking to unveil the originary events, departing from symbolic representation
and converging on a shared, transcorporeal affective reality. She goes further, suggesting that
there is no need to interpret, only to unfold, “increasing the surface area of experience” (ibid.).

87
3.1.4. Contributions
I believe that the development of CPEI and its testing in Los Angeles has broader
implications for the way social scientists think and practice place-based research and thinking.
The study ultimately makes three contributions, as follows.
Firstly, through an explanation of the development of CPEI, I highlight the kind of
methodological thinking which I believe is necessary to develop convincing accounts of place in
contemporary social science. I argue that the generation and subsequent analysis of data perform
active roles in the research process and are not neutral and objective abstractions. This would be
clearly impossible for place-based research, which is axiomatically and hermeneutically
dependent on the materiality of representations.
Second, while it might seem obvious that distinct mobilities would affect place
attachment, I lay out this relationship through accounts of participants with nonstandard
experiences navigating the post-Fordist American metropolis. The unfolded ecology makes
visible a range of attachments, objects, places, and cultures, forcing themselves onto our
attention and thinking.
Finally, knowledges produced by CPEI are shown to be neither fully representational nor
nonrepresentational, rather I argue that they are more-than-representational. The
transcorporeality which flows within the researcher-participant-audience trialogue is Latour’s
(1999) circulating reference, generating dense affective intensity with immediate effects. I

88
conclude with an evaluation of the likelihood of success for methods like CPEI, promising
optimism for its future.
3.2. Methodology
The materials for this research were generated over the autumn and winter of 2023-2024.
I developed relationships with the neighborhoods and networks of the individuals involved in the
study over three years of formal and informal observation during my doctorate. I had been a
regular presence in the community, working in different research and service roles in projects
and organizations situated on the Eastside of Los Angeles, starting in the summer of 2020. This
project was conceived in large part through interactions with my neighbors and conversations
with community practitioners during that time. For this reason, I did not consider myself a
complete stranger to the places and people under consideration: I shared and share many of the
physical spaces and amenities of the participants. I stop short at suggesting that we possess the
same exact experiences; I am a car owner, do not have children, and do not live in a multigeneration family household for example, but echoes and close reflections of mutual feelings and
affective knowledges did emerge during the collective analysis of materials.
3.2.1. Study area
The study was conducted in the area immediately east and northeast of Downtown Los
Angeles, California, USA, colloquially referred to as the “Eastside”. The area comprises El
Sereno; City Terrace; Boyle Heights; and Lincoln Heights, and in the early 20th century was
considered a success story of multicultural and multiracial urbanism (Sanchez, 2004), with a
booming economy and an outstanding record of democratic and communal decision-making and
political activism (Pardo, 1990; Scott and Soja, 1996; Viesca, 2014). The area should be of great

89
historical interest to mobilities scholars and urban geographers. It was brought into the economic
core of the city by a remarkably dense network of public transport; at one point five separate
companies operated routes from the Eastside into downtown (Post, 1987), facilitating the growth
of relatively wealthy communities living in attractive housing with well-developed and plentiful
community spaces like parks and meeting houses.
By 1961, all the lines were decommissioned in a wave of policies enforced under the
strident “urban renewal” carried out by state-led capital accumulation after the Second World
War (Gibbons, 2023). Suburbanization, in large part driven by the ethnic assimilation of Jews2
into the white Protestant elite (Sanchez, 2004) led to the development of the San Fernando
Valley, racially restrictive covenants and other redlining practices, and increased zoning on the
Eastside for locally undesirable land uses (Cutter, 1995), resulting in a deliberate separation of
Angelenos by race and class3. After almost the entire area was redlined by the Home Owners
Loan Corporation (HOLC), federal programs were seen as the best possible course of action.
Under the narrative of improvement, city planners, underwritten by government departments and
emboldened by the auto lobby, seized entire tracts for demolition and new construction. The
success of the Arroyo Seco Parkway in reducing journey times between Pasadena and Los
Angeles was convincing enough for the California Department of Public Works (now Caltrans)
to draw up a plan for a regional freeway network hitherto unseen. The relics of this plan remain
2 At the height of their population, there were nearly over 35,000 Jews on the Eastside, in some
neighborhoods eclipsing the Mexican-descendent population (Elliott, 1996).
3 The policies were all motivated by racial capitalism, but some were more brazenly race-based
than others; the relocation of Japanese Americans to internment camps during the war removed
thousands of residents, of whom many left the Eastside for the first time in their lives, returning,
in some cases as late as 1946, to a completely different community (Hayashi, 2019).

90
standing today in the form of the six freeways that traverse the Eastside. The areas which city
planners deemed “blight” neighborhoods are bound on all four sides by interstates or state routes,
leaving wealthier, whiter, neighborhoods on the westside freeway free. The concrete megaliths
serve as physical and symbolic demarcations, spatially explicit reminders of the extent of
privilege, and in this case race; the Eastside is now almost 100% Hispanic/Latinx, the highest
both proportionally and absolutely in the U.S. (U.S. Census Bureau, 2022).
The elimination of streetcars and electrified light rail literally paved the way for the shift
to private transit, and slapdash plans to run trains along the central median of the new freeways
were never fulfilled. This was the start of the legitimization in the global imaginary of Los
Angeles as a traffic-scape, as the city almost wholly pivoted to private car ownership in the
1960s. Commuters were funneled into the existing economic centers from the expanding San
Fernando Valley in the west, and the San Gabriel Valley in the east – through the Eastside. New
modes of capital and consumption were only made possible by cars and roads, a transition just as
ideological as material. Dozens of 19th century craftsman cottages were bulldozed in Boyle
Heights to make way for an offramp from the 10 freeway directly into the car park of a flagship
Sears department store (Estrada, 2005), almost exclusively accessible by driving. The urban
structure of the Eastside was redefined by these at-best semi-permeable linear structures, which
produced an uneven geography of exclusion for the carless population. They were forced to
negotiate the inability to benefit from, or even to navigate, the freeway, with the costs of aural
and respiratory impositions from the noise and particulate matter spewing over its shoulder. And
these neighborhoods have some of the highest rates of households which do not possess a
working vehicle. The constituent areas which make up the Eastside have an aggregated 14.47%

91
carless population, compared to a county average of 8.66%, and a national average of 8.30%
(U.S. Census Bureau, 2023).
3.2.2. Study population
One of my goals in this project was to limit the extent to which an inequality between
researcher and participant would define the terms of the participation itself. Like Packard (2008),
I proposed an almost entirely exploratory strategy to my institutional review board, which was
unpopular. The only criterion that I had in mind for participation was place attachment to the
Eastside, which is difficult to spell out in a conventional research protocol. It is impossible to
embark in the social sciences with tabula rasa (ibid.), but I did try to start out without an
unnecessarily strict predetermination, which began with participant recruitment. This started with
purposeful sampling by reaching out to five local practitioners, and then snowball recruiting, as
those five referred individuals in turn. The local practitioners were directors and administrators
of community organizations with whom I had cultivated relationships through previous
community-based research and volunteering. To the review board, place attachment was
described as “including, but not limited to, primary residency,” which six of the seven
participants fulfilled. Other attachments were proximal family history, work and education, and
practically any other reason for spending significant time in or around the fuzzily bounded
Eastside. The other participant, H, lived in Vernon, adjacent to the south of the conventionally
defined Eastside. In a city so associated with cars and driving, I wanted to explore how place
attachment was affected by mobility outside of the standard mode of navigation. Carlessness –
lack of access to their own personal vehicle – was another condition of participation.

92
Finally, because my Spanish is just conversational, a skill which I am refining following
a realization that the Castellano I learnt in Europe is a very different language to the Chicano
spoken in Los Angeles, fluent English was also a requirement to participate in this research.
Before the cut off time, 10 participants initially came forward, of which 7 eventually agreed to
take part in the full activity of the study, listed in Table 3.1.
Table 3.1. Demographic characteristics of study participants
Participant Race/ethnicity Age Gender Education Neighborhood Other notes
AJ Latinx 30-34 Female High
school
El Sereno
H Latinx 30-34 Male Associate
degree
Vernon
L Latinx/White 25-29 Female Some high
school
Lincoln Heights L is married to E, with
whom she has one
child. She has another
child with a different
father. Her mother
bought a car a few
weeks before the
study, which she
sometimes is allowed
to drive. She lives
with E.
E Latinx 25-29 Male High
school
Lincoln Heights E works as a customer
service agent for the
Los Angeles Metro.
He has two children,
one with L.
V Latinx 25-29 Female Some high
school
El Sereno V has been in and out
of prison in her adult
life, most recently
released eight months
before this study.
She has one child.
K White 45-49 Male Bachelor’s
degree
El Sereno
J Latinx 20-24 Female Some high
school
Lincoln Heights J has two children.

93
3.2.3. Practicalities and the initial meeting
Once they had agreed to take part, I arranged an initial in-person meeting to go over the
terms of participation, mostly in the participants homes, but in one case a Starbucks and another
a McDonalds, both at the participant’s request. The aim of this first encounter was to begin to
develop some initial trust, for me to explain the overall aims and context of the research
including what is expected from the participant, and for them to sign the consent form. I chose
not to provide participants with cameras, rather choosing to use their smartphones as the primary
instruments to generate the materials. It is increasingly difficult to move through the city in the
21st century without a phone, meaning that almost everyone has one, and carries it close to their
bodies at almost all times. This is convenient for visual researchers, as practically all phones
have high-definition cameras, but also for spatial scientists, as they also have surprisingly
accurate GPS receivers. As smartphone users are commonly already used to taking photos and
recording their location with their phones, it means that they do not require training or instruction
on how to collect, store, and share the materials, which has been a barrier to participation in
previous projects (Rose, 2012).
During the initial meeting, I discussed with participants how many photos to take, and
what kinds of images to capture. The frequency and volume were left very open-ended, with no
upper or lower limit. I suggested that they try to take at least one photograph a day, and not to
feel pressured to point their phone camera at something which they normally would not. On the
information sheet and the consent form, the following prompt was suggested, but not prescribed:
“Please capture images which you feel represent your places and how you move around the city.
This may include taking a photograph every time you make a journey, every time you travel

94
somewhere on foot, or to photograph every place you visit during the day. You may also capture
any other images which you think are interesting in relation to this prompt.” It was important to
me not to ask the participants to take photographs which they thought that I would find
interesting, but to leave that designation up to them. This notion is discussed in Frith and
Harcourt (2007), whereby participants show concern that their images might be boring, not
meaningful to anyone but themselves, or not “artistic” enough to share. These concerns were
raised by my participants, but I assured them that every image had some meaning worth
discussing, even sharing some of the mundane images which fill up the camera roll on my phone
to those who needed particular convincing, explaining the numerous photos of my flat bike tire,
or my new elastic shoelaces, and what each meant within the context of my own spatial
experiences. Once they realized that any photo taken would provide an insight into their place
attachment and mobility, they were all confident in their capacity to generate an array of visual
materials that would represent their experiences.
Participants were also shown how to enable and disable location services on their phone.
The mode for recording location in this study was Google Location History (GLH). A large, but
unknown, proportion of smartphone users in the U.S. have Google Maps installed on their phone,
and most of those users have not opted out of the passive location recording service, with most
people’s “timeline” showing a complete record since the day they first signed in to Google
(Ruktanonchai et al., 2018). GLH uses a combination of the phone’s antenna and local Wi-Fi to
generate the record, collecting “place visits” and “activity paths”. It matches location to its stored
database of “places” and “activity types”, giving a confidence score for each one. For example, a
place visit may show 88% confidence of being at a coffee shop or park. If the user has input their

95
home and work locations as places, records there usually have 100% visit confidence. Activity
types include different transit types, like walking, driving, bus or train. GLH collects waypoints
along the journey, providing information like bus route, walking speed, or wait time in between
transfers. It is increasingly common to use GLH for activity space research, categorizing the
place visits into different behaviors for understanding different public health exposures like air
pollution, walkability, and green space (i.e., Hystad et al., 2022; Paraday et al., 2022). In this
project, the GLH acted as a mediator, tethering the data and analysis firmly to the geographic. It
worked to connect intricate and often abstract experiences to specific places as part of the
investigation into affect and its spatial configuration. The locations of each photograph were also
recorded, using the internal GPS of the participants’ smartphones.
I also allowed participants to choose the duration of their participation, between 1 and 3
months. The study period included Thanksgiving and Christmas, and some participants were
keen to either include these holidays, or to stop data collection before. Lastly, I explained how
we would transfer the materials, asked whether or not they would like me to send text reminders
to keep taking photographs, and penciled in the time and place where we would meet next.
3.2.4. Generating materials
Participants then went away to generate visual materials over their chosen duration.
There were no requirements for the interval between images, so they took photographs as and
when they felt comfortable. During this time, they were encouraged to interpret the prompt on
the information sheet as they saw fit, and to check in with any questions or ideas that they had.
One participant, L, texted several times asking about “the assignment,” worried that she would
be graded on some aspect of the photos. She also texted to say that she was consistently having

96
second thoughts on whether the photos were “good enough”, going back through her camera roll
and deleting the photos that she had taken that day. I reassured her that she was in control over
all the images produced, that there were no grades available for content or composition, and that
she may want to preserve all the images over the study period, rather than being overly selective.
I’m sure that she did find this assuring, as she said later that she stopped deleting the photographs
after a few days. For Holland (2000), this speaks to the role of photographs outside the scope of
research, “part of the complex network of memories and meanings with which they make sense
of their everyday lives.” Empowering participants with the autonomy to generate materials as
outlined in this study emerged as a crucial gesture, providing a means for the individuals to
assert control during a time when other aspects of their lives might have been characterized by
uncertainty. This is a common point of emphasis in other visual methods (Frith and Harcourt,
2007), and participatory action research in general (MacDonald, 2012).
Once each participant’s generation period had elapsed, I met up with them a second time
to transfer the photos and to download their GLH. I color printed the photographs in large
format, numbered them all, and before detailed inspection, placed them into an envelope. I also
visualized the GLH at different zoom levels and on different basemaps, using a “firefly”
symbology which intuitively shows which routes were travelled the most using thickness and
brightness. I printed these in color in large format too and placed them in the envelope with the
images. The envelopes were sealed and returned to the participants for two purposes. Firstly, the
person who took the photographs may not have wanted to keep them all as units of analysis,
choosing to remove embarrassing or illicit images for example (Rose, 2012), and secondly, I
encouraged participants to look through the photographs and imagine a caption for each one. I

97
also asked them to choose 12 photos which we would focus on during the final meeting, even
though all the photos would be considered in the joint analysis.
Asking the participants to curate their submissions served two purposes. Firstly, it
kickstarted the processes of reflection and introspection, in a manner that should have enriched
the final meeting (Blinn and Harrist, 1991). Secondly, I anticipated the potential inundation of
submissions, given the ease with which a great many digital photographs can be captured very
quickly on a smartphone. Selecting 12 images prioritizes those deemed most captivating by
participants, igniting further primary reflection, and mitigating the impacts of excess. Many
practices of visual methods like this were developed before the smartphone era, centered around
disposable cameras as a mode for image collection. The natural constraint of 36 exposures
demanded participants engage in thoughtful pre-selection. However, the deliberation and slower
pace of film cameras occasionally disrupted the immediate realization of the intended image
conceived in their minds, and Packard (2008) found that despite their mechanical simplicity,
many people just did not know how to operate them.
Today, smartphones are much more woven into the visual and behavioral milieu of urban
existence, but it is essential to account for their material and technological mediations in research
design. I attempted to do this in this study’s methods, but there were still some unexpected
complexities in the generation of materials. Having shown participants how to switch off
location sharing for their GLH collection – in the same way that there may have been images
which they did not want to share, there may have been places which they did not want to show
up on their location record – several of them either forgot to turn it back on, or somehow

98
adjusted another setting in their phone which resulted in incomplete data capture for part of the
study period. This is one of the downsides of using GLH as opposed to a conventional noninteractive GPS transponder (Ruktanonchai et al., 2018), but my intent was to afford participants
the ability to observe, mediate, and even withhold, their location data in real-time using their
GLH timeline, aligning with my efforts for autonomy in the image generation.
3.2.5. Carto Photo Elicitation Interview (CPEI)
After the materials were returned to the participant, we arranged an interview at a place
and time of their choosing. Several participants were interested in seeing the physical spaces
associated with health research on the Eastside, and requested to meet at the USC Health
Sciences Campus, in between Boyle Heights and Lincoln Heights. This stage is vital in clarifying
the meaning of the photographs and the maps; the materials alone are unable to evoke fully the
social conditions within which they were produced. In this way the interviews became more than
another point of data collection, transformed into a shared platform for analysis, led by the
participant and supported by the researcher. The translated expertise, here in the language and
conversation of experience, delivered non-representational knowledge alongside the very
representations of the affect itself. In other words, to have examined the images and maps alone
would have severely limited the understanding of how the participants experienced life in space
(O’Grady, 2018). The CPEIs served as active sites for knowledge generation that were inherently
participatory: to make sense of the situations that these people had found themselves in, it was
necessary to not only attend to their knowledges, but also to include them first-hand in this part
of the sense-making process.

99
In each CPEI, I asked participants to first lay out all their produced materials, and then
choose something to start with. From there, we followed a semi-structured protocol laid out in
Rose (2012), which begins with broad questions like “what does this show?”, “where were you
when you took this picture?”, or “what happens in this area?”. From there, I asked about
whatever topics emerged, pursuing and developing the themes that the participants provided.
This dialogue produced a very rich account of the participants’ experiences, the whole time using
the materials as a shared reference. My own familiarity and shared experience of navigating
many of the same spaces was invaluable here, allowing for joint and reinforcing recollection of
different geographies, reduced the time that we needed to spend elucidating the unimportant
trivial aspects of the situation. This, in turn, provided more room to delve into relationships and
social meanings, transcending the depiction itself (Deleuze, 1988). I believe that this practice has
ramifications for geographic thinking, detailed later on in this chapter.
Several authors have argued that CPEI-style methods have limitations if the researchers
are the ones leading the interview (Fawns, 2020; Meo, 2010; Emmison and Smith, 2000). I
attempted to mediate this by allowing the participant to select the photos, and starting with the
broad questions, letting the participant describe the content and themes which they want to
emphasize. This allowed information to come out that I would definitely otherwise have never
been able to find or interpret. For example, in AJ’s interview, she chose to speak about a
photograph (Figure 1) of two storefronts in a strip mall, one called “Angel Health Center,” and
the other a nail salon sharing a name with her mother. The image held little interest for me at first
glance, but it led to questions about the changes in her relationship with her mother, and how
long journeys on the bus to an energy healer had allowed her to cope with ongoing family

100
trauma. These opened up an entire conversation about intergenerational households and moving
through the city as a spiritual person. This underscored the productivity of CPEIs, supporting
Hodgetts et al.’s (2007) perspective that such methods can prompt participants to produce
articulate and insightful narratives about profoundly personal dimensions of their lives.
Figure 3.1. An image captured by AJ whilst waiting at a bus stop after a spiritual healing
appointment
More than this, Croghan et al. (2008) purport that these become sites for identity work, or
“forms of self-accounting.” In the interviews, participants clarify their presentation of self,
repairing any misrepresentation in the images, working with me as the interviewer to build a
picture of the consequences of this [re]presentation in the broader social context (ibid.). Hodgetts
et al. (2007) and Rose (2012) affirm the role of the photographs themselves in this selfaccounting, pointing out that participants may have wanted to portray themselves in research as
intentionally different than they choose to in their regular lives. Considering this – what photos
were included, were taken but not included, and were not taken at all – is an important part of
assessing the social context of the materials. This is the same for the GLH. Participants may have
altered their regular navigation of the city, consciously or not, knowing that they were being

101
recorded by their phones at all times. This is a key difference between the two types of materials:
the phones, and therefore the GPS receivers, must have been present for each photograph taken,
but were not necessarily on the participants’ person for the entire duration of the study. This is
important to acknowledge too, that the GLH only shows where the phones went when they were
recording location, neither places that they went when they weren’t, nor places that participants
went to without their phones. Discussing what photographs were not taken, as well as which
places were not recorded, is thus an important part of the CPEI.
The interviews were recorded. Each time a photograph was being discussed, I made sure
to reference the number I had given it before returning them in the envelope.
3.2.6. Analyzing the carto photo materials
The interviews were all transcribed manually using ExpressScribe. Just as in the analysis
of the other materials, the conversion of conversation to text here was an exercise in discourse
analysis. Though it allowed me to understand, interpret, and relate the actual interaction, the
creation of the representation in the writing of the speech event, transfiguring it into units of
analysis, was a stage of reduction (Du Bois, 1992). It is important to acknowledge this in CPEI
practice, being aware of what is lost or gained in each successive mediation from matter to form
(Latour, 1999). This reduction did allow the incredibly rich encounters to be analyzed using
existing techniques however and having taken an overview of all the data collected, the three
modes were first examined separately: the spatial; the visual; and the written.
This started with exploratory spatial data analysis of the GLH, calculating attributes such
as furthest distance travelled from home and route density. For the visual, this was a content

102
analysis: photos were divided by what they depicted, into categories like means of transportation,
private places, special persons, and consumption. The transcripts were input into NVivo for
framework analysis, as laid out in Ritchie et al. (2003). This method uses a hierarchical thematic
framework to classify and organize data according to key themes, organizing and describing the
patterns and relationships of emergent but comprehensive categories. Using these techniques, I
began to be able to describe and interpret the data intuitively. As a participatory project, I
abstained from a priori hypotheses. Instead, I hoped that there would be an emergence of
meaning through the unfolding of uncertainty and unanticipated outcomes (Latour, 2005;
Lincoln and Guba, 1985). This preliminary analysis as a composite undertaking proved
invaluable in delving into the contextual intricacies, and I started to generate themes from open
coding of all the data. This was useful, as it allowed the specific roles of the materials, and then
the relation between them, to be considered more directly.
I think that it is also important to mention here that relatively few of the images are
presented in this study, and the maps do not detail many specific places visited. This is common
in elicitation research and Rose, (2012), for example,, who suggests that the photographs are
more valuable for their generation of rich and complex dialogue than they are for their content. I
agree somewhat, but whilst many studies choose not to reproduce any of the photography at all, I
have incorporated some images for illustrative purposes, deploying them as exemplary instances
of their active contribution to the analytical and interpretive insight within the framework of the
research. More than Rose, I think that the images captured by the participants are imbued with
particular contexts of display and discourse, operating as mediators of meaning. Reproducing
them en masse in a context outside the CPEI, such as in the pages of an academic article, is

103
analogous to the transformations discussed earlier; it would alter their meaning and ultimately
divert their relevance from the core arguments I wanted to make in this project.
3.3. Results and discussion
The results of this study are a curated series of eight thick descriptions. These are accounts of
behavior which describe social context alongside the behavior itself (Geertz, 1973). In this study,
part of the “thickness” is given by a persistent attention to place, rejecting a notion that behavior
is discrete from geography. This triangulation of meaning is a strength of the mixed method of
CPEI, which works to uncover and unfold more relevant connotations of the processes and
events which contribute to affective knowledge production. This section is organized as follows.
First, I summarize the nature and extent of the generation of materials, including a short content
analysis of the images and some exploratory spatial data analysis of the GLH traces. In the
following sections, I relate participants’ experiences through seven thick descriptions:
i. Reasons for carlessness
ii. Freedom and independence
iii. Free time in public space
iv. Family car and family place
v. Public transit
vi. Senses
vii. Safety
Focusing on the links between mobilities and attachment to place, I discuss in each context
the potential of CPEI to reconceptualize spatial and visual methods as embodied, rather than

104
solely technical modes of attending. Finally, moving onto the participants’ experience of CPEI, I
present theirs and my reflections on the method, evaluating how far it allows participants to
articulate their matters of interest and concern. The thick descriptions, plus the discussions and
evaluations which they provoke, are key to the three study aims: (1) they necessitate an
examination of the situated knowledges in the materials – acknowledging who made, saw, and
analyzed them is an intrinsic part of their role; (2) they pertinently conjure the embodied and
quotidian patterns and rhythms of life in Los Angeles without a car; and (3) they re-enliven the
nonrepresentational turn by rejecting the role of photographs and maps as illustrations for
modelling, communication, or impact, rather emphasizing them as a research method in their
own right (Langdridge et al., 2019).
3.3.1. Photo and carto elicitation
In total, the 7 participants captured 512 images. This ranged from a minimum of 9, to a
maximum of 313, with a mean of 75 and a median of 42. Of these, 336 were photographs
containing no discernible humans (64.3%). 87 were photographs with people as their primary
focus (16.7%), and the remaining 99 were photographs which included people not as their
primary focus (19.0%). 177 images were captured at the participants’ home addresses (33.9%),
174 were photographs taken whilst travelling (33.3%), and 171 were captured in other places
(32.8%). A majority (n = 463, 88.7%) of the images were in the “standard” vertical orientation,
which makes sense considering that all the images were captured using smartphones. There were
no notable differences in orientation according to age or gender, though two female participants
did note how the portrait design of most mobile photo sharing apps affected their decision to
capture portrait images. This aligns with the findings of other contemporary visual methods
practitioners (i.e., Thomson and Uddin, 2023), but finer content analysis, such as camera to

105
subject distance, blurriness or lack of focus, and placement context, was outside of the scope of
this study. Generally, the images were “good quality”, in that they all represented something that
the participants had chosen to capture. I suspect this is due to the self-selection process, both in
their decision on which photos to submit in the first place, and which to choose and caption as
their top twelve for discussion. No participants reported any painstaking difficulty when making
these selections, and all seemed content with the images that they had chosen.
For the GLH data, the most commonly recorded “place” was participants’ home
addresses. The GLH is unreliable when recording long periods with no movement (Ruktanonchai
et al., 2018), resulting in large gaps in the data. Where there are large (6-10 hour) gaps, common
practice is to assume that people are at home sleeping (Elevelt et al., 2021; Ashbrook and
Starner, 2003). These gaps are tedious to identify but do seem accurate when triangulating a
participant’s home address coordinates with their oral account. Table 3.2 shows an accurate
excerpt of V’s timeline over a few days. The place names are taken from Google Maps user
submissions, so not everywhere that a participant visited is identified by name. The place visits
are generally very high (above 90% confidence), but the durations are not always reliable based
on the timestamps provided by the GLH.
Based on these assumptions, participants spent an average of 62.0% of the time at their
home address during their elected study period duration, ranging from 51.5 to 68.8%. Of all the
places visited, an average of 23.5% were within a 2-mile radius of the participants’ home
addresses, 45.8% were within 5 miles, and 61.1% were within 10 miles. The furthest distance
travelled away from home was 115.03 miles, when H was driven to visit friends in Tijuana.

106
Table 3.2. Section of V’s Google Location History data over several days during the study period
Latitude Longitude Place name Place Visit
Confidence (%)
Duration
(hh:mm)
34.087341 -118.174896 Huntington / Pueblo 93 00:04
34.0935961 -118.1614241 Sohi Ithread 95 00:38
Home 95 09:14
34.104135 -118.152999 Starbucks 96 00:09
Home 93 11:52
34.0477728 -118.3369717 Unknown 97 00:03
34.0329 -118.334493 Chevron 88 00:13
34.0700548 -118.0134449 Cottonwood Manor Apartments 94 06:11
33.8183092 -117.2294596 Discount Mini Mall 97 01:03
33.8181481 -117.2111597 Unknown 95 07:32
33.802656 -117.228651 Plaza de Perris 95 00:54
33.8000749 -117.2275431 Perris Towne Square 94 00:04
Home 92 09:43
34.0898303 -118.148657 Costco Wholesale 96 00:54
Home 92 05:44
34.1290978 -118.0025328 Southern California Service Office
of Narcotics Anonymous
95 01:13
34.0505256 -118.0855655 Walmart Supercenter 95 00:29
34.0778937 -118.1420023 Yoshinoya - Alhambra 95 00:17
34.0805603 -118.1521464 Starbucks 94 00:08
Home 84 01:31
34.0212365 -118.1298075 Los Amigos Group 96 00:56
Home 95 12:19
34.0946618 -118.1404935 Unknown 91 00:03
34.0956428 -118.1513172 Walgreens 89 00:15
34.0798829 -118.1530505 Kohl's 94 00:21
For insertion into the CPEI, maps were created showing the entirety of each participant’s
GLH. The aggregated routes are shown in Figures 3.2 and 3.3. When presenting these maps to
the participants, the most common response was surprise at how far they had travelled. Many
participants immediately started tracing the routes with their fingers, intuitively recounting “this
is when I went down to my meeting,” “you can see where I went for my Christmas dinner,” and
“this one here is walking up to the creek with my dogs.” The cartographic part of the elicitation
took on an interesting dynamic; participants had created the geographic information by moving

107
around with their smartphones but did not have the experience or technical skills to display all of
the GLH at once. Several of them said that they had been looking through the daily timeline in
Google Maps for the first time during the study, but seemed impressed, sometimes taken aback,
that their movements had taken them so widely.
Figure 3.2. The aggregated Google Location History of all participants

108
Figure 3.3. Aggregated Google Location History, centered on the Eastside
Six of the participants lived very close to each other, with H living in Vernon. A kernel
density surface showing the spatial clustering of the GPS records is shown in Figure 3.4,
alongside the participants’ home addresses. Two areas of the highest density – the westernmost
yellow areas in the center of the map – are on the southern and north-eastern corners of the “East
Los Angeles Interchange”, the 135-acre intersection of freeways and interstates. The other
hotspot, the easternmost yellow area, is in southeast El Sereno, close to California State
University, Los Angeles, where several participants transferred buses at the campus
transportation center. The other densely visited areas were Downtown Los Angeles, again where
several public transit routes converge, and along the arterial transit routes crossing the Eastside.

109
Figure 3.4. Spatial clustering of Google Location History data
3.3.2. Interviews
The interviews focused on participants’ attachment to place, and their movements
between and within different places. Because the participants did not own cars, large parts of the
dialogue revolved around their experiences on Los Angeles’ public transit system, namely trains
and buses. All sorts of anecdotes and descriptions followed, grounded in the shared reference of
the photographs and maps. Figures 3.5 and 3.6 are hierarchy charts which show the framework
coding strategy I used when going through the discussions of mobility and place attachment,
respectively. The size of each rectangle represents the number of items coded, a proxy for

110
frequency across all participants. Every participant talked about the experience of carlessness,
public transit, encounters with homeless individuals, safety, and trips during discussions about
mobility. When talking about place attachment, all seven spoke about community, gentrification,
and their neighborhood. At the end of the CPEI, when I would start a conversation about the
experience of the research process, all participants spoke about how the materials related to their
embodied experiences, and whether or not they thought that their stories were “good” on
balance.
Figure 3.5. Tree diagram showing open coding strategy for mobility themes

111
Figure 3.6. Tree diagram showing open coding strategy for place attachment themes
3.3.3. Mobility and place
The cultural turn affected mobilities studies too, welcoming transportation geography
into the social sciences. This had numerous beneficial effects, notably the widespread acceptance
that mobility is endemic to life, society, and space, and a renewed focus on how mobility relates
to the core concept of place (Kwan and Schwanen, 2016). Miller and Ponto (2016) assert the
social, cultural, and political reification of mobility, is recursively bound up in the production of
place. This is vividly encapsulated in Los Angeles, where movement and navigation are
inherently power-laden. Health and well-being are central to the mobility-place nexus; access to
and use of health care facilities are inextricably tied to human movement at various scales
(Kwan, 2012), and as people live out their daily routines, they are under the geographic

112
influences of many different places, coming into contact with different people and
neighborhoods. People’s exposures to those different social and environmental influences evolve
across space and time in incredibly complex ways (Kwan and Schwanen, 2016), some of which I
have tried to divulge here, focusing on the idea of how carlessness impacts things like time and
everyday life. Over the span of 120 years, a brazen interplay of marketing and state-led corporate
capitalism have rendered personal vehicles nearly indispensable to the American way of life. The
participants in this project demonstrated an acute awareness, expressed in the vernacular, of the
intersection between race and class with automobility. Just so, there were numerous unfolding
narratives depicting states of immobility, wherein livelihood or identity had trapped them in
specific places. The thick descriptions which fill out this section are imbued with multiple, often
profoundly ambivalent meanings and affective resonances, illuminating the spatial and social
contingencies inherent in mobility and place-making.
3.3.3.1. Reasons for carlessness
The main reason why participants didn’t have a car was money. When asked why they
didn’t have a car, five participants pointed principally to finances, especially not being able to
afford a large down payment. Some of these responses are shown in Figure 3.7. Low credit
scores usually mean unfavorable terms for car-buyers, necessitating a large cash payment
followed by long-term, high-interest monthly payments.
This opened up conversations about the other costs of car ownership, with differing
perspectives on cost and value. E, AJ, and H were convinced that they were saving money by not
having to buy fuel or pay monthly insurance premiums. Others, like H and V, were certain that
using public transport was worked out to be more expensive, and several participants referenced

113
the high price of Uber rideshares to explain this, describing how when public options let them
down, they had to pay $20 or more to hire a private car for a bus journey that would have
otherwise cost $1.75. These ongoing cost-benefit analyses are challenging internal calculations,
ultimately premised on the centrality of car production in the American economy at a time when
its cities were expanding (Schwanen, 2016).
Cars are the first consideration for urbanism and city planners; owning and paying for their
use is fundamentally wrapped up with capital accumulation and being an active agent in the postFordist political economy. In many ways, the success of government in the U.S. depends on the
cultural values of cars. Independence, flexibility, freedom, social status and other affectivities
which connote car ownership have funded the popular strategies of road expansion and
construction, advanced by the coalition of private and public sectors for stimulating growth
“At least not right now. Not until my money situation is more steady. So I know that I’ll be able to
afford it and feel comfortable.”
AJ
“Why wouldn’t I go and get one tomorrow? I just feel like now that I’m riding it, [the metro] it’s easier
for me to want to save more money, to continue just keep adding up, so if I do get one, I’m still able to
have extra money.”
E
“Just when I know that I’m actually financially stable… I haven’t even got my paycheck yet. I don’t
even know what my paycheck’s going to look like.”
V
“I think it was just because we couldn’t afford one. I’ve been working since I was 15, but I just couldn’t
afford one at the time. My mum couldn’t just afford to give me a car, like most of my friends.”
L
“I’m saving money. I’m saving money on gas and I’m saving money on insurance. I try to see the pros
more than the cons.”
H
Figure 3.7. CPEI extracts discussing financial reasons for carlessness

114
(ibid.). This urban environment is almost solely receptive to automobility, intensely and mutually
concerned with it much more deeply than other types of mobility. And yet the precarity and
instability of capital accumulation leaves little room at the margin to subsume low-income nondrivers, like some of my participants.
a. Other reasons
It is perhaps no surprise then that the only voluntarily carless participant was also the only
white participant, K. He described how he relinquished his car after moving with his wife to the
Eastside from Koreatown, in a more built-up area of the city. Rather than driving, he primarily
travels by bicycle, using an all-purpose commuter/road hybrid bike to get around. This is unusual
for white men in America, the highest proportion demographic of car owners (Blumenberg et al.,
2018). It points to a need to trade sweeping generalizations about automobility in Los Angeles
for localities and particularities. Unlike the other participants, K rarely used public transport
options, preferring instead to walk, cycle, or ride in friends’ cars when convenient. This is an
interesting fray in the dominant narrative of racial capitalism, and perhaps a sign that certain
white residents in Los Angeles are poised for post-automobility.
Several participants were parents, and J cited her first pregnancy as a reason for never having
learnt to drive: “honestly when my friends started driving, I got pregnant the first time, and they
just drove me around, or my family would drive me around like to get food or to go to the store
or whatever.” Having two children relatively young and looking after them during the day meant
that she had never found time to learn to drive. She also cited a lack of time to study for the
written test for a driver’s license, or even to go to the Department of Motor Vehicles (DMV) to
wait in line. V also talked about the written test, saying that even though she knew how to drive a

115
car, she was almost paralyzingly scared of the process, having heard anecdotes from friends she
had made in prison. I suggested that she sounded well-prepared after she told me that she had
downloaded several study guides and three different mock-test apps to her phone, to which she
replied that if she did fail from nerves, or not being able to remember the material, that she
wouldn’t be able to afford the retest fee payable after the third failed attempt.
3.3.3.2. Freedom and independence
Another interesting topic of discussion with many participants was the idea of freedom.
Where the literature points to the freedom-granting norms of car ownership (Schwanen, 2016), it
seems to have discounted the inverse. Four participants spoke about the different freedoms that
came from not owning a car, ranging from H’s description of relief at never experiencing the
stress of getting behind the wheel himself, to AJ’s talk about how the lack of trepidation she
faces when choosing whether to have one more drink, knowing that her friends are worrying
about having to come and collect their cars in the morning after a night out. K enjoys the
newfound freedom of decelerated travel through the city, slowing down to appreciate different
features of the neighborhood like mature trees and old homes that he would otherwise have
missed when driving. These are not part of the typical advertising campaign for cars, which
focus on the constant pursuit of the optimization of incumbent technology.
Even so, the techno-economy that has turned travel time into dead time, folding speed
and efficiency into the means of production, removed many freedoms afforded to the automobile
population. Everyone talked at length about the longer journey times without cars, and how the
time on the bus or train was frustrating, seeing people speed past when waiting at a stop, or when
held up by service delays and planned maintenance. For most, this meant time spent on their

116
phone; H said that he had tried to study during this “dead” time, but often felt too distracted on
the bus, V said that trying to read whilst moving made her feel carsick. Reading and studying are
obviously not possible when driving either, but most participants felt as if they were missing out
doubly; taking many times longer to get to where they were going, and not being able to feel
productive or occupied on the way there. The continued alienation of the poorest groups in the
global North will come as a result of the endurance of urban automobility, giving many people in
the same situations as the participants a lessened capacity to endure (Newell and Mulvaney,
2013). Carlessness has become another motif for path dependency, limiting the possibility for the
redistribution of life chances and wellbeing across the full range of urban populations
(Schwanen, 2016).
This meeting point of production and mobility was especially evident when talking to the
participants about their commutes. V had brought a photograph of her nearest bus stop, taken
whilst she was waiting for the bus to go to her new job. This started a discussion about her
morning routine, which began before 6 a.m., three hours before she normally clocked in. After
waking up she would reheat a small frozen breakfast sandwich in the microwave before heading
out the door to walk a short distance to the bus stop. If the bus came promptly, it was only a short
ride east 6 stops, then a change of buses, and another short ride 3 stops to the strip mall where the
restaurant she worked at was. In the few weeks that she had been going on that route, the bus
never did come promptly. V always arrived at the same time, anxious that she would miss an
ahead-of-schedule service. Her anxiety was blended with boredom and hunger. The photograph
was taken on a long, straight section of Huntington Drive, a major east-west route through El
Sereno, on which she pointed out that she could see the bus coming from around half a mile

117
away, where there was a stop at another crossroad. V explained how if she couldn’t see a bus
when she arrived at her stop, she would have at least enough time to go into the doughnut store
on Huntington to buy a Monster energy drink, a high sugar, high caffeine, second course to her
light reheated breakfast. This became a daily practice, filling some time and some appetite.
Carlessness had channeled nutrition and consumption into the social process of commuting,
creating a daily mobility practice that was defining and confining V’s courses of action (Harvey,
1989).
3.3.3.3. “Free” time in public space
When looking at the maps, I asked V about one of her further trips, to Inglewood in
southwest Los Angeles. She told me that she had been told about a program that gives pre-loaded
TAP cards, the system of contactless payment for LA Metro, to low-income ex-prisoners,
working out to around 20 free rides per month. This left her in an interesting situation, whereby
she was not paying out of pocket to ride the bus, but because of her morning routine and the
unreliable service, she had a different set of daily expenses. When the bus did come, she had
built in a buffer so that she knew she was unlikely to arrive late to work, where she aimed to be
at least 30 minutes before her shift started. In that time, she went almost every day to the
Starbucks next to the restaurant, purchasing a large drink and sometimes another breakfast wrap
if she was still hungry, usually around $12, saying “that’s time that I get back, free time, before I
have to go to work,” when she could sit down and use her phone before an 8-hour shift. Except
of course her use of that place wasn’t free; she was only legitimized as a user through a
transaction that would equate to just under an hour of her work at the restaurant next door. Users
of coffee shops have to demonstrate intent as a customer.

118
J recounted a similar experience when talking about a photo of a coffee shop in Lincoln
Heights, where she had stepped inside with her child in a stroller to escape the rain: “I felt bad
for being there without buying anything for so long. I think I bought like four drinks because I
was there for hours.” In these instances, legitimacy is granted to anyone that the amenity was
trying to attract, so for businesses that is paying customers (Sanusi and Palen, 2008). Austerity
and rent-seeking urban policy have resulted in a systematic under provision of true public space.
Now, free public parks and the public libraries may be the only places where judgements of
legitimacy are not solely made on the basis of a commercial transaction (ibid.), and even so,
those places are not open to all at all times. It is overly simplistic to compare public places with
private vehicles, but in the specific cases of V and J, having a car would have provided a
comfortable place to wait in before work, or a dry place to wait in while it was raining.
Several participants also spoke about one specific store, Target, describing it as a place
they go to when bored, to socialize, or as a pastime of itself. L said that it was somewhere she
had recently spent a lot of time with her mother, who only a few weeks before the study started
had bought a car. Before then, they had not been spending much time together after the birth of
her children. For L, her mother’s new car had opened up Target as a destination to refresh their
relationship, attaching meaning to another otherwise meaningless space of consumption.
3.3.3.4. Family car and family place
Apart from H, all the participants had close family members who owned a car. AJ, L, E,
V, and K all lived in the same house as car owners, whilst J lived only a short distance away. V,
J, and H did not have driver’s licenses, but the others were able to borrow and use the car when it
was not being used by its principal owner. That being said, none of them had actually driven the

119
cars that much at all, in AJ’s case not once in the last 5 years. L and E used L’s mother’s car the
most, around once a week or so, but only after the mother was back from work, and only then
after L’s sister was not using it – she was given priority under the auspices of her enrolment in a
fairly distant California State University. K’s wife owned a car, which again he very infrequently
used. One of his pictures showed a rain-filled bike lane, taken from the passenger seat of his
wife’s car, which he used as an example of a time when she would drop him off at work, rather
than cycle along the wet roads.
Those who couldn’t drive had even less reliance on the family car. In V’s case, she knew that
she would never be able to drive it, even if she got her license; the insurance company had added
her as an exception on the policy because of her felonies and the increase in the premium would
have been too much for her parents to ever be able to afford. J’s mother had also just recently
bought a car, in her case to make it easier caring for her grandparents, whom her mother lived
with. It hadn’t really changed her quotidian mobility however, still walking most often. There are
of course gendered ramifications to this. L, J, and V all cited one of the main uses of their
mothers’ cars as making elder care more manageable, and then at other times in the CPEI talking
about the difficulty of caring for young children as carless mothers themselves.
a. Carless mothers
Gender theory was not central to this study, but there are clearly areas where it structures
mobility intensely. Uteng and Cresswell (2008) described the entanglement within mobility
studies of actual and potential physical movement, inciting analysis of the meaning and
embodied practices of motherhood as a carless individual on the Eastside. L, when talking about
a photograph of her mother’s car, described how much more comfortable her children were on

120
the school run when it was very hot, as she said it had been in the summer of 2023, within the
air-conditioned space. Yet she had only been able to take them a few days of that summer, when
her mother had called in sick to work. While L and J were not employed full-time, V had decided
to “work whenever they want me to work. As long as I’m getting paid for it.” She spoke about
how, since getting back from prison, she thought that not having a car meant that she wasn’t able
to redevelop her relationship with her child, leaving before they woke up for school, arriving
back home often after bedtime. J, pointing to a line on the map between her apartment and the
county hospital, talked about how she was in constant fear of her children getting sick after she
had faced no other option than to walk to the emergency room with her eldest child. Worried
about the eventual cost of an ambulance, she didn’t have enough money in her bank account to
call an Uber, and her mother was miles away with her car. Carlessness raises difficult questions
for gender roles, and although normative ideas that a mother’s care is best for a child’s
development are largely outdated (Van Wel and Knijn, 2007), it is evident from the mothers here
that the material constraints imposed by carlessness in a city structured around auto-dependency
are profoundly impacting the relationships and emotional bonds of maternity.
b. Family on the block
Five participants lived with or within walking distance of family members. AJ lived in her
grandparents’ house, L and E lived together in the back unit of L’s grandparents’ house where
her mother also lived, V lived with her parents, and J and her boyfriend lived in an apartment
owned by her uncle around a 7-minute walk from her grandparents’ house, where her mother and
siblings also lived. H had previously lived with his parents in Minnesota, moving to California
aged 22. K moved out of his parents’ house aged 18, moving to California from out-of-state.
Several participants suggested that living with large families was a core Latinx experience,

121
recounting the good and bad sides of intergenerational cohabitation. E, V and AJ had grown up
in places where their extended families lived in adjacent houses or apartments, creating a hyperlocal tight knit familial place. V and AJ still lived in those places, but in both instances the
families had coalesced into one unit, with siblings or cousins moving out, like E had done after
he married L. In all cases except K, the only white participant, and H, housing costs were
defrayed by living with family. J told me how her uncle had made a unit available to her so that
he could go and live at her grandparents’ house to help with the elderly care, moving into her old
room. L’s grandparents owned their home, as did AJ’s. V’s parents had been renting the same
house since she was born from her mother’s friend, who had raised the price very little in over 20
years.
Overall, without speculating on real estate trends, it seemed that the participants’ incomes
were not high enough to be able to afford to rent or buy homes through the conventional market.
I thought of the house I had been renting during my doctorate, for which I was paying over five
times as much as V’s parents. I knew that the owners, a Latinx family, had lived in my house for
over 50 years. Once they decided that the opportunity to generate passive income through leasing
it out was great enough, they moved into a bigger house further east, affording the mortgage by
way of my monthly rent. H, showing me an image of his living situation – a room in a shared
house – told me that he would never have been able to afford to rent a house in California
without “the benefits of the Latino community.” His room was cheap, the owners hadn’t asked
for a credit check, and hadn’t even asked him to sign a lease. The tenancy worked on an
agreement that as long as an appropriate amount of money came in roughly once a month, he
could stay. H didn’t have family with housing in California but relied on a migrant network to

122
find a place to live. These family and community networks are difficult to measure and quantify
but are clearly working at providing housing.
Participants were mixed in their opinions over whether family spaces were restraining or
liberating, but it was clear that living amongst family had a material impact on their connections
to place. AJ showed me photographs of a crepe myrtle tree which she remembered planting as a
child with her since passed paternal grandfather in the house opposite her maternal
grandparents’, where he used to live. Since he died, AJ had seen an uncle move in and default on
the mortgage shortly after, meaning the house was foreclosed by the bank. Six years later, a
young white family moved in. They remodeled the front garden, but left AJ and her
grandfather’s tree. Every time she walked out of the front door of her house for several months
after the new family moved in, she would look at the tree over the road and cry at the memory of
her grandfather. She also cried at this point in the interview, talking about the photograph of the
tree. Family memories can evoke intense emotional responses, turning banal objects into
landmarks of meaning (Rose, 2004). The images of these family places and the record of
participants’ movements through them initiated talk of powerful emotional geographies,
performing as traces of bodies. The materials were sites of togetherness, both where participants
produced images of actual congregated family members, but also of scattered-ness. It was only
through CPEI that the embodiment and subjectivity could be interpreted from representations of
family places, conjuring a formidable and corporealized loving – of family both proximal and
distant – that although was probably beyond words, was evidently taking place somewhere. I
argue that the different “somewheres” were an embodiment of the systems and structures of
cultural meaning, simultaneously sites of superficially ordinary, everyday practice, as well as of

123
deeply personal memory and affect. The situation and movement of family across physical space
had a profound impact on the way that the participants spoke and reflected on their mobilities
and place attachment, reproducing its own shifting ground of space and encounter.
3.3.3.5. Public Transit
Five participants, all apart from J and K, brought images captured while on public transit.
For many, the bus in Los Angeles is thought of as the transportation mode of last resort (Lipsitz,
2004), a busy, dirty, loud and unpredictable assemblage of people and vehicles moving through
the city, largely unseen. There is a considerable reason in most people’s decision to use the bus;
they can’t drive. E and AJ talked about the embarrassment they felt as bus users, with AJ
describing the feeling of sharing the bus with children on their way to high school every
morning: “Slightly embarrassing because I would hope by now that I would have a car, but I
don’t. But also connected in a way where I mean that was me too.” Five participants, including
AJ, had at one point been in the place of those children, and shared mostly fond memories of
riding packed buses early in the morning with their friends. For all of them, this was among their
first experiences of independence, trusted by their parents with the bus fare to dutifully get to
school by themselves, or with siblings. For some, this was their first experience with fare
evasion, preferring not to pay the driver, but to save the money for a luxurious breakfast at a fastfood restaurant along the route (L), or saving the change for longer trips on the weekend (V). For
Los Angeles’ carless, the bus is not the nightmare from Weird Al Yankovic’s “Another one
Rides the Bus”, but a very real necessity (ibid.).
It is frustrating then that the worst services are imposed on the people who rely on them
most. I’ve already recounted some of the frustrations the participants faced, such as long wait

124
times at bus stops, unsure if one would arrive in time for them to get to work or to pick up their
children from school. But many participants also talked about the environment on the buses
themselves. Several of them had unwittingly sit down on what they believed were urine-soaked
seats, grabbed handrails with unknowable stickiness, sweated in July when the air conditioning
had failed, or been refused entry late at night by an anxious driver. E, a subcontracted customer
service representative for LA Metro, had spent the most time on buses, likely travelling the
furthest by way of his job. He had taken several photographs of graffiti-tagged route boards and
other user-damaged features. Despite facing this array of damage and uncleanliness daily, E was
incredibly passionate about his role in improving people’s experiences of buses and trains, and
generally expressed great satisfaction in his work. Even so, for most, public transit was
uncomfortable and inconvenient, becoming a series of spaces which participants wanted to spend
as little time in as possible. This was an unfortunate paradox, as of course they ended up
spending many hours in these places over the course of the study. H spoke about his commute
which he knew only took 15 minutes to drive but stretched to an arduous hour and a half on two
buses every day, longer if the transfer didn’t line up neatly. He described the “dread” of this
experience as deeply embodied, something that he felt from the first moment he woke up every
workday. I felt a large contrast with the dread that I feel when I have to commute, not so much
embodied as a mild apprehension for being in my comfortable and familiar car a few minutes
longer in the Los Angeles traffic. The pervasiveness of the dread, accumulated over days and
weeks, cultivated a palpable aversion to transit among the participants, markedly removed from
the memories that they had shared of relatively carefree bus rides to school, skipping the fare and
chatting with friends.

125
a. Encountering unhoused people
There is a significant group of people whom the setting of public transit provides more than
transportation. For many people experiencing homelessness, buses and trains are places to sleep,
shelter from extreme cold or heat, escape surveillance and discipline, or use free high-speed
wireless internet (Ding et al., 2021; Los Angeles Metro, 2017). Ultimately this is due to a lack of
public funding and resources for unhoused people in the US, but it has meant that public transit
spaces have become coincidental places which provide amenities and services. As a pressing and
much politicized matter of concern, all the participants had taken photographs and brought
experiences of encounters with unhoused people to the CPEI, with mixed and contrasting
accounts, shown in Figure 3.8.
Encountering unhoused people is a routine experience in Los Angeles, but for many people
the encounters are from behind a windscreen, while exiting a freeway, or driving past an
encampment. For the participants though, who were coming face to face with unhoused people,
some nearly every time they stepped onto a bus or train, they were navigating the cartographies
of homelessness with no barriers. Their movements and experiences transcended the political
and social spatialities of homelessness, and the materials that they produced could come to stand
for the deep failings of the welfare state, as well as the myth of the vitality and wealth of postindustrial urban heartlands (Cloke et al., 2008).

126
One of E’s images (Figure 3.9) showed an emergency medical team and some firefighters
wheeling a person along a train platform. E had said that while he was on his work route,
performing customer service duties, he spotted a man, who he believed to be unhoused, having a
drug overdose. He went into great detail about how only a few days earlier his team had been
given Naloxone (Narcan) kits, which effectively reverse opioid overdoses (Gould, 2019), and a
morning of training in how to administer the medication:
What I did first was make sure to open my kit, and I gave one dose in one nose, waited
two minutes, give another dose to the other side of the nose, and the patron was not
responsive at the time, so I made sure before I started doing that, that my partner called
911, to get the full report going, and then ended up using my partners Narcan kit. I took
her kit, and I used 2 more. I also did chest compressions that day. It was a pretty scary,
interesting day. He looked pretty dead to me, he wasn’t responding with 4 kits, 4 Narcan
kits under his nose. I did chest compressions for about 15-20 more minutes, and when
they had just arrived, he took a breath.
“The transients walk a lot up and down, and like I said, they’re unstable. If I give them money they ask
for more money. It’s difficult to know how to respond in those situations.”
AJ
“… being in LA with what type of environments there is, and homelessness and whatnot, I can’t trust
people, I can’t”
H
“I have nothing against them, I’ve always helped out the homeless in ways that I could. I just think the
poop, the trash, the leftover needles and syringes, stuff like that I don’t like.”
L
“I feel pretty sad honestly, because at the end of the day, they’re human, that’s how I see them. I don’t
see them as bad people, or people that lost their mind, they’re human, just like everyone else. It’s sad
that they went that route, even if they didn’t want to.”
E
“They’re not harmlessly there, they create an abundance of waste and trash.”
K
Figure 3.8. Interview excerpts discussing participant attitudes to unhoused people

127
Figure 3.9. An image taken by E after he had administered Narcan and chest compressions to a
person having a drug overdose
This led to a discussion on the difference between the actual and expected responsibilities of
his job in customer service. I asked him whether he thought delivering emergency medical care
was in the description for the role, to which he replied “Naturally it shouldn’t be. Our
description was customer service. And then they added the Narcan kit. I never thought I would
be doing it.” This was a shift in the locus of responsibility from government, in this case directly
onto customer service workers like E. We both agreed that it was amazing that he had been in the
position to save the patron, but that it shouldn’t have come down to him as a last resort. The

128
enaction of harm reduction practice in Los Angeles has meant putting Narcan into the hands of
people who are close, like E, and so they become a resource. It leaves unsettled the “housing
first” debate and transfigures spaces like the train station on E’s route into treatment rooms.
Giving Narcan kits to customer service workers is neither a long-term structural solution, nor a
sustainable critical intervention, to the structural violence imposed on unhoused Americans
through policies like redlining and criminalization. I was impressed when E said that after saving
the person’s life, he went about the remainder of his workday practically as normal, continuing
along his route, and clocking in the next day almost as if nothing extraordinary had happened. He
said that he would always remember the paleness of the person’s skin when he would go through
the station after the event.
b. Routes and digital wayfinding
One thing that many participants shared was how they had built up a mental route map after
several years of using buses and trains in Los Angeles. For AJ, L, E, V, and J, who all still lived
in very close proximity to the houses where they grew up, this started from a young age, when
they had been taught how to use the network by their friends or family. L explained how when
she got a new friend, her mental map expanded to include the routes from school to their house,
and from their house back home. Her map grew until she dropped out of high school during her
second year, after which she said that she hardly visited any new places until her mother bought
a car, when her route knowledge could grow once again. For V, a lot of her routes were based on
what she called “bad stuff,” and she showed me on the maps where different activities would
happen and how she used to get between them – usually by bus or in a somehow-obtained car
with her old friends. Since coming back from prison, she has been very careful to avoid the bad

129
stuff places, taking special care not to walk with her child through those areas. Relearning the
network in this way enables her to get around without seeing the same associated landmarks and
people, or even as she would say, “types of people”.
For some, the routes they knew and took were invisible boundaries on their quotidian
mobilities. Several participants were anxious when going outside of their regular activity spaces.
The predominant apprehension shared amongst participants revolved around the prospect of
prolonged time in unfamiliar places, particularly while waiting for unreliable services. V and AJ
explained on the one hand their ability to navigate through the discomforts of grime, unsettling
fellow passengers, and delays in places that they knew, such as on their daily commutes, but on
the other becoming hyper-vigilant and increasingly anxious when confronting these same
features in distant, unfamiliar places. When we were looking together at a photograph of a
graffiti-covered bus stop, V said that she would feel immediate relief as soon as she returned to
the parts of her mental map where she recognized the tags. This is corporeal place attachment,
where emotional states are invoked by what bodies are sensing and perceiving. The relations
between people, their mental route maps, and the temporary submission to the mobile collective
of the bus, are diverse, differentiated by affective relations. Through the “energetic outcome of
encounters between bodies in particular places” (Conradson and Latham, 2007), the interplay of
frustrations, anxieties, and familiarities embedded within the mental maps of these participants
flickers into view, lending a new, place-based, discursive register to articulate their experiences
with heightened clarity.

130
Consider the inverse too. H’s description of his mental map was more one of opportunity. He
brought a picture he had taken of an LA Metro map and pointed out sections that he wanted to
explore and memorize, starting with the connections between the Eastside and West Hollywood.
He spoke about the perceptions of the area amongst gay men in Los Angeles, and that he had
previously felt excluded from the place as a gay man who did not like clubbing. It was only after
he visited the area during the day, he had a doctor’s appointment during the study period, that he
began to appreciate the significance of the place in the construction of his own gay identity. The
place had become formative in his own life over the next few weeks, showing his best friend the
new routes he had found from their neighborhoods, visiting and spending more time there and
building his mental map. In this case, place attributes had been melded with the personal
qualities of a gay man, who told me that one of the reasons he had chosen to move to California
from his family home in Minnesota was to “figure out who I am,” including his sexual identity.
Attachment to real and imagined facets of place like this transcends a mere outlining sites of
routine activities on a map, encompassing a further-reaching concept of sites of meaning that
human geographers have long been focused on (i.e., Tuan, 1977). The concept points to an
atmosphere of affect which McCormack (2008) described as distinctly spatial, yet difficult to
locate. The mental maps inhere feelings that arise and dissipate diffusely, embodiments sensed
through movement, and create a ubiquitous backdrop to everyday life on the move (Bissell,
2010).
H’s mental map was also embellished through his interaction with the Pokémon Go app. He
brought a photograph of someone who he had met on his work commute, telling me how he
connected with him after seeing him bring up the game on his phone. Pokémon Go is an

131
augmented reality, location-based game, where users move through physical space as they tag,
collect, trade, and battle for digital collectibles and achievements (Hjorth and Richardson, 2017).
Through the app, a highly abstracted digital overlay across the actual environment displays
objects and checkpoints, georeferenced to real-world landmarks. H told me about some of the
different landmarks, there were several that he knew he could reach from inside the bus on his
routes. He also explained how the collectibles refresh every 24-hours, playing a part in his
decision to take a different route to work in the morning than he takes going home in the
evening, intending to “hit as many pokéstops in a day as possible.” Through the augmented
layering, banal and familiar spots like intersections and bus stops have become significant game
loci for H. He also showed me on the map a trip he had made to Santa Monica, to attend a
Pokémon Go “community day”; paid events when specific types of collectibles are made
available within a geofence not usually found in the game. Unlike the LA Metro map, the
Pokémon map is defined by play, and the community days and other social aspects of the game
highlight the creative dimension of place making as enacted through play (Lammes and Wilmott,
2018). The game prescribes certain actions (e.g., chasing a Pokémon avatar across a park) which
invite interactions with the game, as well as other players, humans, and things (ibid.), creating
new cartographies as players move through physical space. H told me that his sibling was
coming to visit from Minnesota for the next community day, held at the Rose Bowl in Pasadena,
and had already started planning the journey they would take there from Vernon, looking along
different bus routes to see how many pokéstops they could get on the way.
Aside from the game, there were two other apps which participants were using as spatial
interfaces. The first was the Transit app, which provides bus and train schedules updated in real

132
time. AJ, H, L, and E all used the app daily, timing actions such as leaving the house or paying a
restaurant check with the next nearest transit arrival. The other was Google Maps, which also
provides service updates and transit trackers. AJ showed me how she uses the app, explaining
how she checks the “on time” as opposed to the “scheduled” arrivals. V, another Google Maps
user, wasn’t aware of this feature, and we had a long conversation about all the times when she
had been looking at her phone showing a “live” bus going past her location on the map, when in
reality no service had materialized. Like Pokémon Go, these interface apps were doing
performative work, becoming “sites of contestation” (Hookway, 2014) which play a crucial part
in informing meaning. I would consider the spatial realms within the relatively small window of
the smartphones as representational, essentially ocularcentric metaphors for physical and virtual
specificities. Once more, the affective only becomes perceptible by moving past the
representational, when the mind-smartphone hybrid is understood to have produced new and
distributed spatial relations. The app interfaces cannot themselves produce meaning. Take the
nearest park to J, L, and V’s houses, which appears on the phone screen as a green triangle. This
relies on a singular semiotic relationship between a real-world feature and its representation
(November, Camacho-Hübner, and Latour, 2010): the park is likely to contain green objects. In
their mental maps though, the parks contain a chain of production for participants (ibid.): taking
their respective children to the playground; touching, feeling, seeing, hearing a whole array of
things which cannot be inferred from the green pixels.
Because the participants were not used to thinking like spatial scientists, I stopped short of
asking outright whether they were sensitive to their own roles in the shifting alliances of the
interfaces – seeing the map pan and rotate around their representative “blue dot” centered on the

133
smartphone screen – but there was plenty of evidence for the “unfolding mapping practices”
(Kitchin et al., 2013) of digital wayfinding. The embodied tactile engagement with the apps – H
tracing his routes in Pokémon Go with his finger, E uploading photos of damaged Metro property
to his work’s hazard reporting service – engendered a perpetual spatial becoming, amalgamating
mental maps with visual representations, plus participant’s physical movements across space. I
thought these amalgams were most interesting when the hybridity included the networked,
personal, and quotidian aspects of experience, and would be really interested to see how the
kinds of knowledges made possible through CPEI can contribute to the reframing of maps as part
of a new cartographical epistemology, moving beyond their relatively basic and mimetic
representations of space. In particular, I anticipate a future examination of the possibility of
virtual reality and gamified ethnography to reconsider spatial relations in place attachment.
3.3.3.6. Senses
Rose (2012) describes the rejection by the affective emphasis on the distinction between
vision and visual. Like Clough (2008), she emphasizes the immanence of the dynamism between
body matter and matter more generally. To explore this, I’m going to articulate two other sensory
encounters which the participants had with specific materials.
a. Smelling
For all its talk of atmospheres, affective geography pays a great deal less attention to the
olfactory than to the other senses (Tan, 2012). Certainly though, the photographs and location
traces that participants brough to the CPEIs conjured up some vivid accounts of smell and
smelling. Smell is a marker of space, delineating self-other boundaries (Lefebvre, 1991; Low,
2008). These boundaries are fused in confined spaces like the bus, especially as I learnt from

134
participants, when sharing rides with unhoused people, who may not have had recent access to
hygiene facilities. E and V recalled this in dialogue about some of their images in Figure 3.10.
Figure 3.10. Excerpts from E and V discussing smell on the bus
Sensory politics are deeply implicated in the affectual relations between urban residents.
The immediate sensual responsiveness to someone walking onto a bus blend with cultural
responses like repulsion, fear, and shock, and in some ways become politically validated if they
align with the dominant moralized narrative. Attitudes to unhoused Angelenos are justified by
discourse like “clean up the streets,” or “they stink of shit,” spatialized and corporealized by
olfaction. The linguistic and political conflation of “odorous” and “odious” is built on the
embodiment of certain bodies being sensorially out of place, perpetuating stigmatization and
othering (Tan, 2012).
J, talking about a photograph of her children playing in Lincoln Park, spoke about how
she sympathized with the rough sleepers, “I definitely feel bad for them, like it’s not their fault
they’ve got nowhere else,” but added “it’s just frustrating that it has to be in the same place as
where people take their kids.” We talked about how she liked to take her children outside at least
once a day, and because she didn’t have a back garden used the park as an outdoor space.
“The bus system isn’t all that pretty. I mean here and there you would catch a good bus, here and there
you would catch some smelly people, some unhoused people. I try my best not to be too much where
they’re at, if not then honestly, I would wait for the next bus.”
E
“I get up and I move away. I don’t care if it’s mean to them, because I would throw up. I have a very
uneasy stomach; it doesn’t take a lot for me to want to throw up. Sometimes there’s one person, they
walk in and just the whole place stinks, and I’ve got to go open one of the little windows.”
V

135
Underpinning her frustration was what she would smell on her daily trips: “They just smell. Even
in the park, you can smell where they’ve been because of all their trash and their stuff. They
obviously can’t shower, so they smell bad themselves too.” Visiting Lincoln Park myself on a
Sunday morning, I saw a mobile shower stall provided by a collective of non-profit organizations
which tours the city on a schedule (Shower of Hope, 2023). Austerity and failures in governance
have once again moved the burden of public services onto different amenities, meaning that J’s
children have to share a space for play with unhoused people’s space for sleeping and washing.
Duarte (2017) suggests that this is only a relatively recent sensory phenomenon, and that the
olfactory “blanding” of urban smell-scapes through the removal of untreated wastewater,
slaughterhouses and factories, and particularly potent vehicle emissions has meant that
unpleasant smells (the definition of which is contentious) are now indelibly connected to poverty
and even specific ethnicities. The concentration of these people in specific areas, like unhoused
people in parks because of a systematic failure to provide housing, serves to strengthen the
spatial and social correlations.
The contention of what smelt bad surfaced when K showed me an image of someone who
was starting a composting business on an empty lot that he owned next to his house (in total K
owned 5 lots, 2 of which had standing buildings on them). He said that the wind usually carried
the smell away from his house, but when he did smell it, he thought it smelt good, “like,
pleasant.” He went into great detail about where the organic material came from, and how
recently the composter had been given “mounds and mounds of green onions” by a nearby
restaurant. Many would find the hypoxic decay of organic material to be nauseating, but K spoke
about times when he specifically walked up his garden path to smell the pile, “it smelled like they

136
were simmering in a skillet, it smelt really good.” Smell is collectively experienced but yields
involuntarily different memories and attachments. The volatile and endless assemblage of
molecules, bodies, odorants, associations, reflexes, and emotions should be more woven into
discussions of space and place, rethinking the material, embodied, nature of smell, and its links
with the discursive and cultural.
b. Hearing and headphones
Like vision, the phenomenology of hearing is well-codified in urban geographies (Tuan,
1974; Simpson, 2009; Pocock, 1989). During the cultural turn too, there was burgeoning interest
in the spatiality of aural cultures (Krims, 2007). In a car, drivers have almost complete control
over what they are hearing and listening to, but for the carless, there are competing auditory
signals from a diverse array of emitters. Five participants spoke with me about how they regain
agency over what they are hearing, facilitated by wearing wireless headphones.
H and V spoke at length about how they use their headphones. H said “I don’t think there’s
ever been a time when I’ve left the house without my headphones. So it’s kind of rare when I’m
listening to the outside noise around me.” His music of choice was techno pop, which he listens
to every morning on his commute. In one photo showing his every day carried items, I asked
why there were two sets of headphones, one large over-ear pair, and a smaller set of in-ear buds.
He explained how with the big pair, all external noise is completely eliminated, “all of it, you
don’t hear nothing,” but wearing them for an hour or longer induced migraines, when he would
switch to the lighter pair which allowed some ambient sounds to filter through. We talked about
the different experiences that this led to whilst moving through the city; he preferred to blast
techno in the mornings to create his own sound space, preparing for the noisy interactions at

137
work by “just being in my own space, thinking a million things in my own head.” V had a pair of
in-ear buds too, which she said had very good noise cancelling technology, still allowing “talk
through” so she could hear things like driver announcements. She first used them in prison,
where she had been sent them before the bus ride from Folsom to Los Angeles, the furthest she
had ever travelled at the time after she was released. She said that unlike the buses around her
neighborhood, she didn’t recognize the route and had been scared of missing her stop, constantly
pulling the earbuds out to hear the stops being called out over the loudspeaker. It was only when
she got home that her sister showed her the talk through feature, which she felt much more
comfortable using.
Unlike the production of bland urban smellscapes, cities remain heterophonous places. Its
different sound signals connote neighborhood, custom, even transport (Revill, 2013). Sounds and
hearing express multitude, especially in shared spaces of mobility. Several participants spoke
about their dismay at getting onto a bus and hearing videos played out loud through phones, or
waiting at a traffic light where an idling car was broadcasting from its stereo. E moderated the
unruliness by wearing “one in and one out,” keeping one ear open to listen for outside noise. He
also spoke about how in his customer service role, he thought that people would find it offputting if he had both earbuds in at once. For all, the material production of sound in the most
personal space – in between the ears – afforded a sense of control and embodied belonging,
demarking the boundaries of what they could hear, and others couldn’t, in spaces where clear
auditory agency is reduced.

138
For AJ and J, there were clear times when they preferred not to have headphones on. AJ had
made it a habit not to wear ear headphones when she was walking around at night; “I have to
make sure that I could hear anybody nearby, if I don’t know them. I mean people have been
attacked, even in my neighborhood.” For her, hearing was clearly related to awareness and
safety, connecting unexpected sounds, even in familiar places, to corporealities of danger and
surprise. J, for whom almost all her journeys was accompanied by either one or both of her
children, carried her headphones with her, almost never wearing them: “I can’t. I have to be
concentrating on them. I guess I also talk to them a lot, and I couldn’t do that if I’m wearing
headphones.” Hearing is of course part of speech and listening, further evidence of the coproduction of affect through sensing and culture.
CPEI in part relies on ocularcentrism, in that affective talk is initially prompted by the
insertion of images and maps. I think that this becomes a strength, rather than a limitation, if
practitioners rely on the evoking of other senses by visual and spatial, using CPEI as a platform
on which to build solid accounts of all kinds of sensorimotion. I thought that it was really
successful in charting overlapping meaning, showing how olfactory and aural sensing and
culture all constituted, and are in turn constituted by, place and place attachment. In their
descriptions of photographs and maps, participants were superimposing layers of meaning and
sensing, the analysis of which superimposes another layer of meaning in turn, once more
showing the power and promise of CPEI to unfold experience.
3.3.3.7. Safety
To understand the subjective feeling of safety and the related notion of comfort CPEI is
useful for its consideration of the emotional and experiential dimensions of place. Places are felt

139
as more comfortable because of their affective importance, independent of actual risks
(Zavattaro, 2019), not clearly inscribed in routine urban cartographies (Cloke et al., 2008). All 7
participants spoke about safety, usually in the context of feeling unsafe.
a. Night-time mobilities
Darkness changed the way that the participants perceived space, as well as the meanings
which they attributed to it. Night spaces were interesting in the context of safety and moving
through the city. For most, moving around at night was a risky practice (Figure 3.11), full of
dangers and threats. The affective implications are clear: losing touch with daylight sensibilities
and being made open to unexpected encounters with others (Duarte, 2017). For J, night had
become a self-inflicted curfew. She had shown me an image earlier in the CPEI of a place that
she would go and hang out at night with her cousins, showing me on the map the route she would
take from her grandmother’s house where she lived at the time. Some time since then though, she
had become increasingly reluctant to go out after sunset, citing “the world is so crazy now” as
her reasoning, and that “there are so many crazy people in the street now, and they definitely
come out more at night.” Her change in attitude seemed to have roughly coincided with her
having children, which L also explained, “the world is too crazy to be going out at night with
kids that young.” Both of them spoke about how hard it was to be seen by cars at night, and how
worried they were about their children getting run over in the low light.

140
This is another example of gender demarcations in space, in this case the contrast between
light and dark has created new territories depending on the time of day. Valentine (1989)
describes this territorization, whereby at night women are forced to avoid the same spaces which
they frequent worriless during the day. AJ showed me a photograph of downtown Los Angeles
which she had taken to express the injustice of the dark streets underneath the brightly lit highrise buildings (Figure 3.11). She pointed to the bus stop, one that she uses regularly when
coming home from friends’ houses and spoke about the discomfort she felt waiting there at night,
the relief she felt seeing the headlights round the corner, the comfort of finally boarding the
illuminated bus, and of finally being able to relax back in the lit space. The nightly burden of the
fear of male violence has an acute impact on women’s use of space, like the production of the
“taken for granted” behavior of route choice (ibid.). When looking over L’s location record, I
asked her why she didn’t take the shortest path from the main road to her house through a park
and over a bridge, instead electing to walk a much longer way round. Unsurprisingly, she said “I
“Once I get off of the bus there’s no safety net of the light and needing to be seen in case of anything
happening.”
AJ
“If you see around it, it’s mostly dark, and it gives you that impression that somebody could be hiding in
the back. Even though you do still need to go through steps. I don’t know why I think like that, but I’m a
human so…”
H
“At night, I feel like there are people being sketchy or anything. They come at night. Instead of being
out there at night, they’re all over the train system if they can, so I don’t want to be there with them.”
E
“I get scared of walking around at night. Even here, my mum dropped me off so close to my front door,
but I still made my boyfriend come out and meet me.”
J
Figure 3.11. Interview excerpts on navigating the city after dark

141
would walk all the way around because there was weirdos at the park. Weird guys, weird men at
the park, so I would have no other choice but to go around.”
Defensive tactics and spatial coping strategies are necessitated by the construction of a
geography of fear, building up place associations in the participants’ mental maps. These
associations are entirely situated. Consider V’s spatial association, about which she had earlier
spoken of the “bad stuff” places. I asked her about what kinds of things made her feel unsafe in
her neighborhood at night, to which she replied, “Oh no, I’m not worried about that.” She said
that “because I was one of those people that fucking was up to no good late at night. Now I’m
not, and it just doesn’t faze me. I’ve always had that ‘nothing can touch me’ complex.”
Figure 3.12. A photograph from AJ showing the contrast in lighted downtown spaces

142
It is also worth pausing to think about the urban night itself. Duarte (2017) proposes that
complete darkness has been completely eliminated from cities, thanks to electricity and public
illumination. When those lights go out though, the darkness where there was once an expectation
of light leads to new perceptions, forcing people to call on other senses to identify and move
through portions of space. K spoke about his bicycle commute home along a largely industrial
road, very slightly uphill from his work to his home, just enough to require extra pressure on the
pedals. He highlighted a portion of the route on the map, where he said that he had seen people
stripping copper wire from the streetlamps, plunging a mile or so of the road into a more intense
dark. He described what it was like to feel cars coming up from behind him at high speed – there
are no traffic lights or stop signs, so he said that cars often reached high speeds along the stretch
– sensing the headlights and hearing the approaching engine, not knowing how wide they were
going to pass until the last second. His front bike light only provided a very narrow beam, and it
was mounted to his handlebar so would only illuminate straight ahead, rather than where he was
looking. This cast disconcerting shadows, K described now knowing whether they were people
or bushes, and how he sometimes felt like someone was going to rush out of the dark and knock
him off his bike. Cycling is of course a very different night experience than driving, with no
glass or metal barrier between the rider and their surroundings. Rather than pressing the
accelerator to speed away from perceived threats, a cyclist has to exert their muscles, literally
moving their body out of the space in concert with the machine.
The temporal and gendered accounts of experiences of night mobilities were informative.
There were many variations in the perception and use of dark spaces even amongst the 7
participants in this study, suggesting a continued need to examine how individuals and groups

143
cope with moving through cities in the not-quite-total-darkness, plus the interactions of
environments, society, bodies, and senses.
a. Safety, cleanliness, and aspiration
Part of the discussions about safety that I had with the participants was comparing their
neighborhood to other places that they had moved through, forcing thought on what was
producing perceptions of safety and comfort. For many, “cleaner” places were considered safer.
For some, like E, V, and AJ, this was primarily related to people sleeping on the street,
raising contrasting language of clean, rich realms and dirty streets of low-income areas. Apart
“It would be a dream to live there one day, just because it’s such a nice place to live. It has a lot of trees,
a lot of green. And it’s wide-open streets, and it’s not too noisy. It just has everything there, it’s nice, it’s
clean. It makes me feel, it sounds bad, but higher class.”
AJ
“I would just want to move somewhere where it’s quiet, clean, where I could raise my kids. And safe,
so they could play outside, and I won’t have to worry about them getting ran over.”
L
“Even out there, or honestly, we’ve gone towards like Glendale, or when you go to Pasadena, Arcadia,
or Alhambra. There’s a lot of beautiful neighborhoods over there. I definitely want to live in those
beautiful neighborhoods where they have the big palm trees, or big trees.”
E
“I really like South Pasadena. It’s so green there, and the streets are so nice and clean. Plus all the
houses have their own yard and there’s nice stuff there like parks and stores.”
J
“Where you don’t have to lock your car. You can sometimes leave your door open, and nothing crazy is
going to happen. So I like that. I feel like moving forward, when I’m older, maybe in a couple more
years, I want to be in the suburbs or somewhere like that.”
H
Figure 3.13. Participants discussing the relationships between safety and cleanliness across space

144
from K, who said that he already lived in his dream house, all the participants had aspirations to
move to different parts of Los Angeles. These were mostly tied up with ideas of safety,
cleanliness or the physical environment (Figure 3.13).
L showed me four photographs she had taken of abandoned buildings around her eldest
child’s elementary school, one shown in Figure 3.14. She told me that she could remember when
she was at the same school that the homes were full of families, and the houses were always
well-maintained. Since they had fallen into disrepair, she was worried that they would attract
squatters and illicit activity. Although “broken windows” policing strategies have largely been
shown to be ineffective and structurally violent (Modly, 2009; Harcourt and Ludwig, 2006),
there is still an inexorable social link between visual signs of disorder and perceptions of higher
rates of criminality. The attribution of crime to visual disorder by police and the sanctity of
private property within broken windows theory have legitimized fears that participants like L
project onto different places. In the affluent neighborhoods which they identified as aspirational,
visual order of the physical environment was largely upheld by excessive policing. So it was
jarring to hear accounts of police violence which some participants had been subjected to, like E,
who had been stopped and searched by gang units on his way home from school, or V, who had
been physically abused during her arrests, only for them to praise the work of the Los Angeles or
South Pasadena Police Departments in “cleaning up” different neighborhoods. It revealed a lot
about the assumptions people may hold concerning place, fear, and police, and why broken
windows policies and perceptions may have persisted so long in the dearth of empirical evidence
(Modly, 2009).

145
Figure 3.14. One of L’s photographs of abandoned buildings near her child’s school
The association between cleanliness, safety, and dream places was entangled with class
and belonging. Both L and J worked part time as house cleaners and domestic staff in expensive
houses in nearby affluent neighborhoods and had brought photographs of the inside or exteriors
of their client’s properties. They spoke about the large, “I’ve never worked in a house that big”
(L), lavishly decorated, “I thought it was a store or something” (J) houses, and how they
imagined the owners to live in luxury and tranquility, free of having to worry about crime or
danger. This out-of-placeness was tough for them to deal with, both becoming tearful when I
asked what it would feel like to live in the kinds of places which they were cleaning now, saying

146
it was hard for them to imagine. E was worried that if he did move to a more affluent area that a
majority white population might take offence to him being a “messy Mexican,” and end up
calling the police on him, which had never happened to him before in his majority Latinx
neighborhoods. Lastly, V showed me one journey on the map that she made every week with her
mother, to an upmarket grocery store in Pasadena. She told me that her family had developed a
negative impression of their nearest supermarket where lots of the products were locked up,
ostensibly to prevent theft, the aisles were always untidy, and the shoppers were “too ghetto.”
The market they went to was “where the nice people live, the right people,” and her and her
mother were “the most ghetto people in the store.” The class-based place attachments were
mediated by the visual, racialized, and resulted in material shifts in spatial behavior and activity.
This again links back to the flawed basis of broken windows theories, where communities were
wrought apart by “untended” behavior (Kelling and Wilson, 1982).
Although I was surprised, it is perhaps easy to begin to understand why all the
participants generally spoke favorably about gentrification. This was either in explicit terms – all
participants knew the term and were familiar with the concept – or indirectly, when they would
speak about “nicer” (E), “richer” (AJ), “whiter” (L) newcomers to their home places. As broken
windows theory posits a stable neighborhood being deterministically dilapidated through visual
disorder, the participants implored the logical counter; that incoming families would exclude
unwanted vagrants and intruders through discipline and gaze, actively reshaping the perceived
inhospitable urban crimescape into safe, clean neighborhoods. These interactions highlighted the
pervasive influence of neoliberalism in the co-constitution of social norms and urban landscapes.
The cultural prioritization of private property was reinforcing what felt like to me very outdated,

147
yet still very corporeal and material perceptions of crime, cleanliness, and safety. The
participants’ talk underscored the intricate interplay between society, imperatives of capital and
property, and spatial urban arrangements. The appeal of the logic of neoliberalism perpetuates
power structures, legitimizes policing, and deeply influences the lived experiences of American
residents. CPEI had uncovered incredibly detailed relationships between fear, race, and poverty,
and I feel like there is a lot more to be done to unfold the problematic contingencies which
reproduce institutions of power and class-based privilege.
3.3.3.8. Summary
The seven thick descriptions of carlessness on the Eastside interrogate the unequal
exercises of mobility and place making. Mobilities, particularly non-automotive, were shown to
be situated and subjective, interconnected with affect and space. The role of culture is prevalent
throughout, surfacing in the historical contingency and path dependency of uneven urban
movements in Los Angeles. The contexts shown here are by no means exhaustive, participants
brought many other images of routine and extraordinary encounters that led to fascinating
discussions and co-analyses. Above all, their movements and embodied experiences challenge
the idea of space as a container for social processes (Sheller and Urry, 2016), as shown by the
analysis of networks, relations, and flows, not fixed places. The connection between physical
movement and place is not a neutral black box, rather a set of iterations of permissions of social
and spatial meaning (Sheller and Urry, 2006). E and H’s wide-ranging travel across the city are
negotiations with several sociotechnical systems, demanding interaction with assemblages of
capital, people, images, information, markets, and emotions. And so is J’s comparatively
miniscule radius of mobility, determined by a set of gendered, racialized, and class-based factors.
CPEI showed that participants’ attachments to place are tied to networks of connection that

148
stretch beyond any singular place (ibid.), layering meaning into their mental maps and their
corporeal praxes.
Toward the end of each CPEI, the discussion turned to the process of elicitation itself. I
asked participants “How well do you think these materials captured your experiences of the last
few weeks?” which started a reflexive dialogue, excerpts of which are shown in Figure 3.15.
Unanimously they thought that the maps and images were useful and accurate in “telling their
story,” and I agree, the materials provided an important insight into their lives, without which the
depths might have been impossible to garner. The materials themselves were invaluable, for the
information held in their content and composition, but also for the meaning that they were
connected to. By performing as sites of memory and affect, they became a means for making
sense of the participants’ worlds. The invoked states of what their bodies had sensed and
perceived were reinvigorated in the CPEI, expressing their states of affairs spatially, and across
people and things.
Some were very keyed into their personal role in the reinvigoration. V, reflecting at the
end of her interview, said “there’s no-one else that will give you this stuff. It might be similar, or
pertaining to their area, but it’s not mine.” To me, that is powerful testament to the centrality
and importance of expertise in social science. While the materials do possess a degree of
discursive capacity, they fall short in being able to express entire affectual experiences. To
effectively circulate reference between the world and these words, active participation in the
translation of knowledge was imperative.

149
If everyone is expert in their own lives, to engage that knowledge should involve some
type of participation. That idea is fairly key to Hall’s (1997) idea of cultural studies; that culture
depends on its participants interpreting meaningfully what is around them, and also semiotically,
with participants having connected and drawn on accounts of meaning in social settings (Rose,
2012). Where conventional participatory methods uphold the power relations between expert and
non-expert, I thought that CPEI did work to dismantle some of the structural barriers which
proscribe co-production. This was achieved principally by enabling participants to set the terms
for what was at stake. It is a fallacy however to suggest that this method inherently creates coproduction (Packard, 2008). Even by nature of my situated presence as a researcher, a white male
funded by a huge research university and so on, it was impossible to completely inhibit the
occurrence of inequitable interactions. At the very least though, CPEI demanded some
“I think they do. I mean they’re capturing moments in my life, so I think you start putting the moments
of your life together, that tells your story.”
K
“This is pretty much all I do, like you’re seeing it all. This is just what my life looks like, and obviously
you can see where I’ve been on the trace, so it’s where my life is too. There’s nothing else.”
J
“I think honestly, they are actually very accurate. These are places that I move from and go to.… we are
in all these places that the map is actually showing.”
H
“Well really well. Because I feel like I’m a real socialite, and I have so much with friends, and with
moments that I really cherish. There was a lot of funny stuff going on, like a lot of family stuff, a lot of
night life, some silent moments. That’s all in front of us here.”
AJ
“Seeing them gives me flashbacks and memories of my childhood. That picture gives me how I helped
save a life, it’s pretty interesting seeing my life, places I go through on a daily basis.”
E
Figure 3.15. Participants responding to questions on the story-telling properties of the materials

150
introspection of my role in situated knowledge production, and I have been pleased to play a part
in its life in visual and spatial ethnography through this exploratory study.
Lastly, I asked the participants whether they thought that they had told a “good” story. I
wanted to determine what kind of process CPEI had been for the participants, after much of the
talk centered on the inequality that carlessness had produced, touching on several difficult times
in their lives and intense feelings of inferiority and out-of-place-ness. Nevertheless, they all
agreed that going through their stories made them feel good, which I would argue is a placebased response. The retrospections touched on mobility, “it just makes me smile. It’s amazing
how far one human can move around,” (H); memory, “I think they are happy. This reminds me of
my childhood, this reminds me of my favorite place to eat that I can never go to often, the gun
range I’ve never been to,” (L); routine, “It’s mainly just the small things in my life which I do
over and over… yes, it’s a good story,” (J); and relationships, “It’s a good chapter of my story,
looking back at my interaction with the spaces I’m in and the people in those spaces,” (K). It was
a great privilege to provide the grounds for these good feelings.
The participants were all grateful to have been given a space to tell their stories, and to
divulge certain experiences which they had never been previously able to share and discuss, in
many cases literally invisible. Thus underscores the empowering potential inherent in CPEI,
positioning it as a transformative research approach worthy of recognition within the domain of
participatory action research. By providing a space for co-production and collaborative sensemaking, CPEI transcends the conventional techniques characterized by extractive social science
and perpetuation of hierarchical dichotomies between experts and non-experts, facts and
opinions, and powerful and powerless. As a methodology, it fosters inclusive dialogue which

151
empowers participants as active agents in research, ultimately serving to greatly enhance the
collective understanding and appreciation of complex social phenomena.
3.4. Concluding remarks
CPEI is an analytical endeavor into the transmediation of subject matter. I used it as a
method to create a performative reimagining of movement and meaning, studying how mobilities
inform place. As a platform for coanalysis between participant and researcher, it triggered an
unfolding of meaning, increasing the interpretive surface area of experience (Marks, 2000),
emphasizing the role of geography in delineating experience across space (Swyngedouw, 2004).
The materials elicited from participants were embedded with meaning from within the social
formation (Bal and Bryson, 1991), and I worked with participants to triangulate affective
meaning and knowledge through talk and reflection. As place-based enquiry, it offered so-far
unparalleled insight into the locality of nonautomotive mobility and place attachment, elevating
seven thick descriptions of narrative and unexpected outcomes to a broad expanse of application
and engagement.
And CPEI was definitionally based in the local. This stands in contrast to many
conventional methods in the mobilities, which use big data to cut through the “bias” of human
geography (Kwan and Schwanen, 2016). This is not a limitation. Rather, it can be used to
complement and enrich the analysis performed in those methods, initiating new knowledge and
meaning. By forging new connections between CPEI and other approaches of mobility
geography, affect can transition from a theoretical concept into a tangible reality, refining the
framework which has guided social research since the cultural turn. I believe that this would be

152
useful, as it expands the social realm beyond the sole preserve of what is taken to be “human”
(Bissell, 2010). An affective lens forced thought on how the participants’ experiences came into
being through socially constructed assemblages of bodies, objects, technology and movement,
transmitting agency between and within their networks of relations. This was evident when
describing the different experiences of people in the same physical spaces, which I posit was
engendered through situated place attachment.
My arguments for the relevance of CPEI for generating affective knowledges are built on
interviews with seven people on the Eastside. I visited all the participants’ homes over the course
of the study, which is also to say that they all did have homes – there is an established sub-field
which carries out visual methods, including photo elicitation, with unhoused urban residents
(Packard, 2008). I recruited them all by “snowballing” through community organizations. I
recorded the CPEIs, then worked to unfold social meaning through content, exploratory spatial,
and qualitative analysis, using many of the techniques outlined by Rose (2012). My argument
here is therefore based on a series of in-depth conversations with community-engaged, ablebodied, housed, all Latinx but one, individuals with strong attachment to the Eastside of Los
Angeles. As Rose (2004) explained in her study recruiting middle-class mothers, comparative
work is needed before the salience of the details for other people in different situations – in other,
less autodependent cities for example – can be assessed. That outstanding though, I believe that
the method has proved itself as able to generate registers far exceeding the discursive, and the
embodied, affective knowledges which that engenders are vital for developing meaningful
conclusions in social research.

153
The relatively simple research design required substantial methodological thought to
develop convincing accounts of place. It created sites which accounted for the partiality of data,
using gaps and subjectivities as actants in participants’ stories, cherishing the materials as
connective tissue with which to establish meaning. It also created sites for reflecting and
redrawing ethical orientation which I will punctuate as follows: I do think that CPEI addresses
some issues of power and trust in social research, but I also recognize that asking participants to
play the role of co-collaborator was problematic. They did not possess the full capacity, not to
mention the time or energy, to fulfil what I initially imagined. Although, like Packard (2008),
valuable insights were made, it was done in a way which will ultimately reinforce the divide
between science and public. In other words, the sense-making of participant-generated data was
still performed by me, a researcher, acting in isolation from the field. Like many participatory
methods, there is still work to be done to map the limits of CPEI, particularly when dealing with
marginal groups.
As exploratory work, it did make visible the range of phenomena which connect mobility
and place. In particular, it used the spatial causalities of being carless in Los Angeles to examine
how people move through the prototypical auto-organized metropolis as marginal bodies, asking
what that marginality might mean for place and attachment. CPEI worked well in unfolding the
paradoxes of Los Angeles, taking its simultaneous historical and mythical status seriously. As
Soja (2014) and Sanchez (2004) suggest, urban regeneration and racial capitalism have imprinted
themselves onto the real and imagined cartographies of the city, which requires a new
ontological foundation for understanding space here. Mobility scholars must acknowledge the
contingent ways that movement and navigation are impacted by complex networks of

154
knowledge, discourse, power, and bodies, moving beyond singular understandings and shallow
engagements with theory.
Finally, the knowledges derived from CPEI eschewed the representationalnonrepresentational binary, taking on a more-than-representational nature. Their dynamic
transcorporeality, gaining meaning through transposition, echoed Latour’s (1999) circulating
reference and points toward the abandonment of purified categories of phenomenology. By
focusing on what happens when participating bodies were moving through the city, the study
sought to diversify understandings of how affect-driven methods can help to produce new spatial
knowledge and reconstitute networks of scientific production. The immediate consequences of
the affective knowledges are promising, and I am filled with optimism for the future prospects of
CPEI and other place-based methods like it.

155
Vignette 3 Verdemour Avenue
During my first semester at USC, I subleased a room from an undergrad in the business school
who was studying abroad. It was in a high-rise “resort-style” apartment complex near the
University Park campus just north of Historic South Central. The complex had four swimming
pools, a climbing gym, a cinema, and a library amongst other amenities. It offered shuttles to
campus every 15 minutes, which was convenient in the period before I bought a car but was
otherwise fantastically horrible. It felt like a cheap airport hotel, with its long windowless
corridors and cheaply furnished units. It was also sinisterly fortress-like, manufacturing a sharp
divide between inside and out, policed around the clock by a private, armed, security force. It
was extremely verboten to sublet the flats, and my lessor warned me never to speak to anyone at
the front desk lest they question why my name and face didn’t match up to the details in their
electronic system. This also meant that I couldn’t do things like make maintenance requests or
open the parking garage. It was only when I worked out how to wedge open the side gate from
the inside that I could really have people round – you otherwise had to check them in with a
security guard in the main lobby.
The convenience wore off quickly. For what it was, it worked out to be rather expensive,
and I decided that I would rather have much more control over my living situation. I had also
found out that the owner was a strong opponent of mandatory affordable housing requirements
for Los Angeles developments, had sued the city several times, and was a big-time donor to
some pretty awful political action committees. I didn’t sleep well at all knowing that I was
directly contributing to any of that, and after my last coursework assignment was handed in,
packed up my little car and headed to my friend Scott’s house.

156
Scott was part of my migrant network. He had sat next to my mother’s cousin at public
school in England, became firm friends, and later ran odd jobs around London’s West End in the
late 1980s with my father. He made friends with some filmmakers, taught himself how to cut and
edit, moved to California in 1996, the year I was born, at the same age as I had moved twentythree years later in 2019. We had only met once before, but through my parents had extended a
generous offer to look out for me if I ever needed it. He lived in Highland Park, where he had
bought a two-bedroom house for very little money, knocked it down and built a new one which
was now worth nearly $2 million. In the few days before Christmas, I moved my stuff into his
shed and moved into his spare room-cum-office, house-sitting and looking after his dog Scruffy
whilst he was back in England for the holidays. I felt so much more relaxed in that house, even
though it wasn’t mine, not having to tiptoe around security guards, pay dues to an evil billionaire
tycoon. I slept much better away from the faux lux-scape and was grateful to Scott for the place
to stay for a few weeks whilst I looked for my own house.
At the same time, a girl I was going out with had just moved to a huge new house with a
couple of musicians in El Sereno, just over the 110 Highway from Highland Park. I found myself
spending a lot of time there, quickly becoming enamored with the quiet hills and friendly
neighbors. It felt a world away, but in reality, was only 7 miles from my old flat. I really loved it,
and what I loved most was Ascot Hills Park – 93 acres of steep unadorned hills, covered in brush
and native grasses. From the top of the first hill, there was a clear view past the skyscrapers in
downtown to the Pacific, sweeping from the Vincent Thomas Bridge in Long Beach to the Santa
Monica mountains. The space was very wild, especially the northern edge. It wasn’t really a

157
proper park, more of a blank green space, bleeding uphill into the unfenced land around a radio
relay mast and down into the valley where there was an LADWP reservoir. I was quite taken by
it as a place, especially in the evenings, when it became territory of the coyotes. I found the sight
of these wild things, stalking against the silhouetted skyline, very affecting, grateful for their
wild and natural resistance against the mythic, oppressive urbanism which I was slowly getting
used to.
It was happy chance then, when one January evening I was walking Scruffy back to my
car, I saw a tiny handwritten sign for an open house on one of the side streets leading down from
the park’s north gate – the intriguingly French-sounding Verdemour Avenue. It was just about
the last house on the hill, and I came back the next day not knowing quite what to expect. The
whole time, I had been trawling Craigslist, rental sites and online estate agency pages, met
usually with grotty, overpriced apartments, dodgy landlords, or just outright scams. I appreciated
the fortuity of walking past the sign. The property wasn’t listed anywhere on the internet, and
already I was hugely relieved to potentially be free of the disheartening digital chore of house
hunting. I was the only one turned up to the viewing, which was being run by a woman who
didn’t speak great English. She said her name was Sabine, and explained that she was the
owner’s sister-in-law, who was a Korean artist who lived full-time in Seoul. He bought the house
very cheaply in 2009 when he was visiting the US, renting it out to medical students at USC ever
since.
The house itself was charmingly ramshackle, perched on the hillside, looking out towards
Alhambra and the San Gabriel Valley, with outstanding views up to Mounts Wilson and Baldy.

158
You walked down ten concrete steps from the road to the front door, which opened onto the
open-plan living room and kitchen diner. The biggest bedroom was on this floor too, plus a
bathroom. There were wooden stairs down to two adjoining rooms, the second of which had a
back door which let out onto a rickety deck bounded by three enormous blue agave plants. On
the side of the deck was a gate which led round to more concrete steps, leading back up the side
of the house to the front door. The whole interior had been refitted after the last tenants moved
out, so I would benefit from a brand-new bathroom and kitchen, redone wooden floorboards, and
a whizzbang air conditioning unit on the ground floor. I asked Sabine how much it would cost
and was shocked at how little they were asking for – less than I was paying for the sublease the
semester before, but still a sizeable portion of my monthly stipend. I told Sabine I was definitely
interested and could move in straight away. We walked out together, and she locked the door
behind, presumably thankful that someone had finally agreed to take it. She said that she would
speak to the owner and get back to me in a few days.
I waited for her reply, excited at the thought of finally having my own place in my new
city. I fantasized about walking over the road to the park, cycling up and down the steep hills,
having people over freely, and living alone for the first time in my life. It was lots of money,
more than I knew other grad students at USC were paying, but I reasoned that living in a nice
place, in a seemingly fine house with a garden and amazing views, was worth it to me. I was
thrilled when I got back a garbled email a few days later saying that the owner had approved of
me, and that I could move in immediately. I was less thrilled that they needed a social security
number from me – something I did not have at the time – to run a credit check. This is a
routinised part of renting property in America of course, which I thought many other people must

159
have run into trouble with. I asked if there was any other arrangement we could make. They
asked if I could pay the entire first year’s rent up front, as a sort of guarantee. I thought that was
ridiculous, but how about six months? Fortunately, I had a good wad of cash from working in
London in the year before I moved to Los Angeles, some of which I had used to buy my car.
They agreed, and I headed to the bank to get a cheque for the biggest amount I had ever written.
I met Sabine the next day at the house. She didn’t have a proper lease for me to sign. I
thought this was odd, so quickly looked one up on my phone and copied it out by hand on some
paper torn out from my notebook. She signed it, took a picture to send to the owner, readily
snatched my cheque, handed over the keys, and sped off in her white BMW. It was a great
moment, knowing that I finally had my own place. Scott lent me his Toyota Tacoma to go and
buy a mattress and a few other large things, and I moved the remainer of my stuff from his shed
that night.
Straight away there were all sorts of sounds coming from under the floorboards. I
couldn’t think what it could have been, and when I asked my neighbors down the road the next
day, they said to burn some sage to expel any unwanted presence. I did this two nights in a row,
but the noises kept and kept emanating, especially from the crawlspace around dusk. I was sitting
on the deck one evening and finally realized that it must be rats and mice after I saw a couple of
long tails disappear into the agave leaves. I went down to Home Depot to get a live trap and set it
the next day under the house. Within five minutes I heard it go off, and nervously pulled it out to
find a huge fat rat enjoying the crunchy peanut butter I had used as bait. I took a picture and

160
walked the captive rodent up to the radio mast, giving it a new life far enough from my home
that I thought it wouldn’t come back.
This became a daily practice. Sabine couldn’t care less, even after I sent her over 50
photos of different rats and mice. I bought more and more traps and blocked up more and more
holes under the house. The owner didn’t care either, not at least until I rung up the Los Angeles
Public Health Department and got them to come and do a spot-inspection. They cited the
landlord and gave 48 hours to sort it out. That put a fright in them, and Sabine sent round a
buffooning “handyman” to set some bait stations and line the crawlspace with expanding foam.
It definitely slowed the infestation, making it slightly more difficult for the rodents to find their
way in. In the end, it just became something I lived with. My friends called it “rat house”, and
jokingly brought me snap-traps every time they came round.
There were several other problems with the house too. The downstairs cupboards were
filled with asbestos panels. Every so often a swarm of bees would come and live in the roof for a
day or two. The fridge gave out electric shocks if you didn’t close it with by handle. The water
heater exploded one morning in April. A tree fell down and blocked the front door from opening.
I learnt a lot in the two-and-a-half years that I lived there, all throughout COVID-19, particularly
about how awful absentee landlords can be. I’m sure I ended up being a nightmare too: a white,
English-speaking, twenty-something year old with a Marxist identity and a firm grasp of tenant
law. I became completely intolerant of Sabine turning up unannounced, and clearly, she was
losing patience too. She once shouted at me from the top of the steps, lamenting her charity for
her brother-in-law, how she wasn’t paid for any of this, and that I was making life very difficult

161
for all of them. I largely ignored this, knowing that the tens of thousands of dollars I had paid in
cumulative rent was reward in itself.
In the end, the back wall of the house fell down. It sounds it, but it wasn’t all that
dramatic, and it was still just about livable. The bigger problem was that it had ruptured the main
sewer line, which was flooding the downhill neighbor’s terrace every time I had a shower or did
the washing up. Sabine was distraught when I told her. Perhaps she thought that I had pushed the
wall over myself. She told me that the owner had no money to fix it, and that if I wanted to do
anything about it to pay for it myself. I spent that month’s rent on some emergency plumbers,
who effected a temporary repair but warned me that a more permanent fix would be tens of
thousands of dollars – the pipe was completely collapsed, and they would have to dig a 50-footlong trench down the entire hillside to replace it.
USC has free drop-in legal counselling a couple of times a week, where a lawyer told me
to withhold rent immediately, but to look for somewhere else to live pretty sharpish. A month or
two after that, Sabine sent me another email saying that the owner was completely out of money,
that he had no funds to make any repairs whatsoever. He was putting the house on the market,
and I was to leave immediately. I was surprised at the landlord’s own precarity – that my lack of
rental dollars for two months had sent him into a downward financial spiral. I asked what they
were selling the house for, and if I could buy it for the same as he had in 2009. They never
replied.

162
I moved out at the end of April 2022. They ended up selling it for over six times what
they paid. I found out that the reason why it was so cheap in the first place was that it had been
the site of a domestic violence incident where someone had shot and killed everyone inside the
house, going to ground shortly after. The bank foreclosed the house, auctioning it off to the sole
bidder. I look back on my time in Rat House as a slightly ironic chapter. For sure there were
some egregious oversights by the owner and his property manager relative – most of which are
classified by some measure of housing insecurity – but I was fortunate enough to have been able
to deal with every blow. I had enough money, time, and patience to get over things like the rats
and what in eventuality turned out to be a no-fault eviction, which I absolutely acknowledge that
others on the Eastside do not. Ultimately, I left with a set of entertaining stories, and enough
savings from not paying rent for a few months to move to a much nicer place a couple of hills
over. The worst that would have happened to me would have been moving back into Scott’s in
Highland Park. I loved and still love El Sereno, and the house, while it was standing, was just
about fine. It did make me realize the state of housing in the city though, and the challenges
faced by millions of renters with poorly behaved landlords. I have great sympathy for all renters,
and hope to one day liberate as many people as possible from the crushing exploitation of the
illegitimate profiteering and rent-seeking villainy of landlords.

163
Chapter 4 Countermaps as resistance: Using volunteered geographic
information (VGI) to record experiences of housing insecurity on the Eastside
of Los Angeles.
Abstract
Background: Housing has significant impact on health, often serving as a focal point for
individuals’ place attachment. Housing insecurity is an evolving multidimensional concept
which measures inadequacy in physical conditions, as well as the interrelated social costs of
different living arrangements. This study uses volunteered geographic information (VGI) to
construct a countermap which aims to dismantle the cartographic and cultural associations
between profit maximization and home-places on the Eastside of Los Angeles.
Methods: Participants (n = 113) had all experienced some form of housing insecurity in the
study area. They completed a web survey which attached spatial coordinates to their accounts of
life in precarious housing. The data was mapped and used to produce an interactive platform for
communities to upload, store, manage, analyze, and publish their own data.
Results: Extreme rental burden was shown to be a foundational causal mechanism for housing
insecurity. Other important factors included low-quality rental housing stock and slumlord
behavior from property owners. Participants expressed a strong desire to upend the systems of
oppression likely caused by the historical and social processes typical in post-urban-renewal
cities across the United States.
Conclusion: The countermapping platform highlighted a set of emergent challenges for
researchers and communities related to privacy, moderating, geographic context, ownership, and
multidimensionality for place-based knowledge production. Future projects should prioritize
local expertise when designing the data inputs, and strive to record aspirational, rather than

164
despairing, geographic information. The actor-network of participatory mapping requires
prolonged methodological reflection to ensure meaningful outcomes.
Keywords: Housing insecurity, Los Angeles, countermapping, place attachment, VGI
4.1. Introduction
Housing researchers focus primarily on how populations engage with their surrounding social,
economic, political, and physical landscapes, especially within the context of their residential
spaces (Easthope, 2004). In geography, the connexion between housing and the concept of
“home” has considerable significance, intertwining various facets of shelter, architectural
structure, relationships, economy, and ideology into an intricate milieu of social meaning (Blunt
and Dowling, 2022). This article investigates this connection by exploring housing inequality in
four neighborhoods on the Eastside, using volunteered geographic information (VGI) to produce
a collective platform for mapping place-based housing issues under the broad umbrella of
“housing insecurity”.
4.1.1. Housing, Place and Health
Place is a useful concept for addressing human-environment interactions (Tuan, 1979).
Harvey (1996) notes that culture and affect are rooted in the socially constructed processes of
place, which is crucial to acknowledge in housing. Place-based enquiry enables insights into
identity, well-being, conflict, and political economy (Easthope, 2004), particularly when dealing
with significant, meaning-laden places like home. In Massey’s (1995) description of the
“increasingly unstable and uncertain world”, people have clung strongly to notions of secure and
stable places, the interpretations of which are based not only on their built physical materiality,

165
but also on feelings, perceptions and imaginations (Soja, 1996a). The association between
psychological attachment to places and well-being is increasingly well-understood in population
and public health sciences, with proven links to physical health (Stahre et al., 2015; Bhat et al.,
2022), child development (Bess et al., 2023), access to primary care (Martin et al., 2019), diet
(Parekh et al., 2022), and several other outcomes (Bratt, 2002). There is an ontological security
in home, related to constancy and familiarity in daily routine. The introduction of precarity, or
displacement or removal of home-places entirely, therefore, will predictably provoke a state of
distress (Giuliani, 1991).
Precarity is measured by housing “insecurity” in the literature, a multidimensional
framework which translates a lack of adequate housing to stakeholders including researchers,
housing analysts, and policymakers (Cox et al., 2016). The United Nations (2014) describes
housing security as consisting of the following criteria:
i. Tenure security that guarantees legal protection against forced eviction
ii. Availability of materials and infrastructure such as safe drinking water, sanitation, energy
for cooking, heating, lighting, food storage, and refuse disposal
iii. Affordability such that housing payments do not compromise other human rights
iv. Habitability that provides protection against the cold, damp, heat, rain, wind, and other
threats to health
v. Location that is not polluted or dangerous
vi. Accessibility that meets the needs to marginalised groups and does not compromise the
expression of cultural identity

166
The U.S. Department of Health and Human Services (DDHS) elaborated on these core rights,
adding indicators on costs relative to income, inferior housing quality, neighborhood instability,
overcrowding, and homelessness (Office of the Assistant Secretary for Planning and Evaluation,
1969). The last indicator is clearly the most extreme form of housing insecurity, and unhoused
people have received a great deal of attention from policymakers, researchers, journalists, and
the broader public. Similar to many contemporary social issues, its definition has become a
moving target (Cox et al., 2016), with the framing becoming part of the problem itself. As more
explicit attention was paid to differences in housing – variations within living situations along
lines of race/ethnicity, class, age and so on – the framing is now studied more systematically.
This followed growing recognition that the experiences of housing and attachments to homeplaces of, for example, middle-class white single men in their early 20s are too different from
multi-generational Latinx families surviving on social security and unemployment benefits to be
agglomerated into a single category (ibid., Bess et al., 2023).
These juxtapositions occur frequently in Los Angeles, an archetypal multicultural urban core,
and especially on the Eastside of Los Angeles, with its history of cultural adaptation and role as
harbinger of the social complexities that other metropolises will have to contend with in the
future (Monahan, 2002). The situatedness of the Eastside, plus its adjacency to the so-called
“homeless capital of the United States” (Wolch, 2005) in Skid Row, demands a reflexive
inspection of why different human experiences emerge from the same physical spaces. Soja
(1996b) argues that researchers must appreciate the complexity and significance of historical and
spatial factors when investigating cities like Los Angeles, cautioning against the binding of

167
disciplines into solely material or symbolic enquiry. Geographers are proven navigators of coconstruction, so integrating a public health and place attachment framework here is
unproblematic. This article does argue for a deeper engagement with social theory to address this
interdisciplinary research design, moving rhetorically from “housing and space” to “home and
place”.
4.1.2. Countermapping
Housing and place are inextricably linked to economic power and flows of capital.
Harvey (1996) outlines the nature of attachment to home-places as evident in spatial breakdowns
of economic rationality, wherein individuals’ decisions are based on a number of factors,
including their place attachment. Globalization and deindustrialization created incredible
instability for the working class in a restructured Los Angeles (Scott and Soja, 1996), who began
to worry about the meaning of place in the turbulence of the changing global patterns of capital
accumulation (Harvey, 1996). Whilst the conditions were hardly optimum for the promotion of
homeownership, millions of Angelenos, facing a plethora of perceived threats by the volatile
forces of globalization (Easthope, 2004), were desperate to invest in a construction that gave
them a sense-of-place, “rooting” themselves and their families equally in a physical
infrastructure as a social identity (Fullilove, 2020). In the global economic order, the “winners”
in Los Angeles, as in other U.S. cities, were property owners, who capitalized on the city’s
desperation through rent.
These processes are spatially inscribed into the maps and other manners of representation
with which the political elite represents the city, four examples of which are shown in Figures
4.1.1, 4.1.2, 4.1.3 and 4.1.4.

168
Figure 4.1.1. Redlining by the Home Owners Loan Corporation (1939)
Figure 4.1.2. Los Angeles County Metropolitan Transportation Authority’s proposed aerial
tramway to Dodger Stadium (2023)

169
Figure 4.1.3. Los Angeles County Office of the Assessor’s property assessment information
system (2024)
4.1.4. Rental properties on the Eastside listed on Zillow.com (2024)

170
These maps ignore the devastating impacts which urban neoliberal politics and profitmaximizing real estate have had and continue to have on the everyday lives of Los Angeles
residents. They assert the status quo of accumulation by dispossession, charting the continued
enclosure of public goods and undermining of society by technocapital (Maharawal, 2014).
These are familiar critiques of maps and geographic information systems (GIS) more widely,
first outlined by Pickles (1995), later expanded by Wright, Goodchild and Proctor (1997). Their
argument, that GIS’ “unusual exposure to general view”, portends key social implications: what
message does it send, whom is it empowering, and what responsibility its developers have
(ibid.). Creating housing maps is an unavoidable exercise in power, explicitly defining the
concept, privileging the knowledge practices and politics which have shaped its understanding,
and deciding how housing will be managed.
I have come to recognize this dynamic even in my own mapping pursuits, navigating the
field of GIS whilst keenly attuned to the most incisive critiques put forth by human geographers
and social theorists. Most simply, I believe that data-driven cartography is presently unable to
represent the diverse and multiple human experiences within the confines of its structure and
ontological framework. At the same time, the power of mapping is undisputed, as is the success
of its adoption as a technique by the political elite. With that in mind, this study adopts the same
techniques, using the power of spatial data and cartography to tell individual and collective
stories of housing insecurity. Rather than becoming mired in the academic debate about the use
and power of GIS, or on its status as science or tool (Wright, Goodchild and Proctor, 1997), it
endeavors to return power to the citizenry, using the manner of representation which has
historically been employed to discipline and subjugate, as a means for expression and liberation.

171
This is the essence of countermapping, an inherently political act rooted in critical
cartographic and feminist data visualization practices (Maharawal and McElroy, 2018). It aims to
unveil and make spatially explicit the elision of capitalist, colonial, and neoliberal frameworks
into urban social life (Kwan, 2002; Wood and Krygier, 2009). Countermapping addresses
concerns surrounding the production of maps – questions of how, why, and for whom they are
created. As Maharawal and McElroy (2018) plead, it is frequently accompanied by the political
action of fostering solidarity and collectivity amongst participants. In contrast to conventional
maps which highlight investment opportunities or bolster eminent domain claims (e.g., in fig.
4.1), the countermap in this study takes a different approach, focusing on mapping struggle, loss,
and marginalization. Drawing inspiration from the Anti Eviction Mapping Project in San
Francisco (Maharawal and McElroy, 2018), it diverges from housing maps which produce and
maintain a fragmented urbanism of capital accumulation. Rather, it uses an intersectional lens to
theorize the complexities of risk and insecurity on the Eastside through direct collaboration with
individuals experiencing insecurity related to their home-places. This approach yields a new
geography of resistance, encouraging alternative modes of analysis to understand the quotidian
experiences of insecurity imposed by corporate-led state capitalism. This is an arming of the
proletariat, repurposing tools of knowledge production to empower those involved by recording
the complex social world of housing and place in America. In this way, GIS transcends the
science-tool dichotomy, serving as a means to express the insurgency and counter insurgency of
capital and people (Pavlovskaya, 2009).

172
4.1.3. Goals and objectives
The goals of the study are multifaceted: (i) design and test a dynamic participatory
platform for recording housing security using VGI; (ii) produce spatial artefacts conducive to
tenant advocacy and political mobilization; and (iii) critically evaluate the countermapping actornetwork in a community setting. The countermap which it generated serves as an antidote to the
interpretation of the maps and materials which helped to initiate the alienation that absentee
landlordism inflicted on the fabric of American urban life. It also serves as an empirical attempt
to recatalyse Los Angeles studies, emphasizing the importance of geography in the shaping of
postmodern cultural studies (Davis, 1990). In City of Quartz, Davis rejects Soja’s teleologic
mapping of the city, from which I will repurpose a quote meant to distance himself from the
predilection of Los Angeles as paradigmatic: “The city [Los Angeles] is a place where
everything is possible, nothing is safe and durable enough to believe in, where constant
synchronicity prevails, and automatic ingenuity of capital ceaselessly throws up new forms and
spectacles…” (1990, p. 86). In this instance, though social inequality, racial segregation, and
class conflict have been surfaced by capital, resistance – facilitated by me as a researcher, fueled
by the power of participant contributions – has emerged in the hope of producing a convincing
and enduring alternative. I was inspired and compelled to act after reviewing the frank and
affective VGI, and see no reason why other geographers, public health practitioners, or housing
scholars alike couldn’t have the same reaction.
The housing insecurity countermap aligns with Soja’s (1996a) emphasis on spatial
practice, documenting the ways in which Angelenos inhabit and interact with the city. It
acknowledges both the complexities highlighted by Soja (1996a) and the inequalities emphasized

173
by Davis (1990), accounting for the diversity of experience which influences urban society.
Although it stops short at developing solutions to housing insecurity itself, the countermap offers
a platform for dialogue, collaboration, and collective action amongst a diverse public. In
summary, this study aims to advance scholarly knowledge at the nexus of housing, health and
place, and also to catalyze meaningful change and empowerment within the community it is
situated through reflexively applied VGI.
4.2. Methods
This study was undertaken in the winter of 2023-2024. My relationship with the
community developed over the course of several years of involvement in community service as
well as formal observation through other research projects and outreach. I have lived in the study
area for nearly five years, and although my situation was very different to many of my neighbors
and participants in this study, did experience facets of housing security during that half decade.
This study was inspired by the conversations and shared experiences I had with Eastside
residents and community practitioners over the course of my doctorate. In other words, when I
formally entered the field as a researcher, I did not do so as a complete stranger. Prior to meeting
at community centers and outreach events, I was a white transplant, tracing the stories from
Rodriguez’ (2002) The Republic of East LA, trying to understand the history and meaning of the
place I was in. This section describes the study area, outlines the steps taken to design and
implement a VGI survey, to analyze and get a sense of the results, and to create a user dashboard
as a platform for viewing and gaining insight to the data.

174
4.2.1. Study area
The Eastside of Los Angeles is an area just east of the Los Angeles river. Generally
considered as comprising four neighborhoods: Boyle Heights; Lincoln Heights; El Sereno; and
City Terrace, the area has a history of political radicalism. This stems from its previous life in the
mid 20th century as a coincidental site of multiculturalism following redlining practices in the
remainder of the city: Jewish workers arrived with their tradition of enthusiastic trade unionism
(Sanchez, 2004), meeting dedicated Japanese Americans cultural organizers (Elliott, 1996),
Black leftists (Pulido, 2002), and perhaps most significantly, the hub of the Chicanx power
movement (ibid.). All four groups had suffered egregious immigration, employment, and
property exclusions, but on the Eastside, housing was fair, fulsome, and crucially not restricted to
those of the “Caucasian race” (Sanchez, 2004). Housing demands were codified in many of the
activist groups, like the Black Panther Party’ Ten-Point Program which demanded “decent
housing” and people’s “community control” (Newton, 1996). The units, mostly single-family
residences, though not expansive, luxurious, or modern, were generally viewed as places of
security, and the area even became the site of some of the first public housing communities for
low-income families in California (Los Angeles Conservancy, 2023).
After the Second World War, there was a two-decade period when the multicultural
structures were dismantled through the social engineering practices of applied racial and
population science, fiscal policy, and direct intervention by the state and federal governments.
The reshaping of the four communities on the Eastside started with local housing and
transportation policies like land-use zoning and freeway construction and culminated in a
wholescale redefinition of the terms of racialization (Sanchez, 2004). Firstly, the new color line

175
placed Jews into the “white race”, at the same time as the rise of Jewish-owned real estate in the
San Fernando Valley. Jews moved west, leaving the racially unregulated Eastside behind. After
Japanese internment ended, many returners could not afford to move back into their own homes,
especially as mortgage assistance had been funneled away to the wartime coffers. In the
aftermath, racism had time to evolve into the “model minority” construct (Pulido, 2002), when
many Japanese Americans sought to assimilate as quietly as possible, drawing minimal attention
to themselves (Kashima, 1980). Self-employment, even in highly racialized occupations, was
paying more than the old working-class employment (Light and Roach, 1996), and they too were
moved over the river. As for Black Eastside residents, Kim (1999) and Pulido (2002) argue that
their position as the universal racial subordinate enhanced the position of Chicanx Angelenos,
especially after the Watts Riots and other primarily Black events deemed “problematic” by
white, universally dominant, city officials. They moved out of the area, suffering the material
effects of the structural division produced by the new racial ideology.
After the smaller ethnic groups – Italian, Russian, Armenian – were revised into
Caucasian-ness, the Eastside became almost completely racially and socially homogeneous. The
neighborhoods are now almost entirely Hispanic/Latinx, with most census tracts in Boyle
Heights and City Terrace over 90%, and in El Sereno and Lincoln Heights over 80% (U.S.
Census Bureau, 2023a). The firm Chicanx attachment to industrial labor, resulting from
employment exclusions in other sectors, created an enduring working class, deepened following
the new international division of labor and deindustrialization. This was nonconductive to
homeownership, and resultingly the Eastside has some of the lowest proportions of housing units
occupied by the owner, an average of 29.5% across the neighborhoods compared to 46.2% in

176
Los Angeles County, with some census tracts as low as 6.5% (U.S. Census Bureau, 2023b). This
can be traced to the selling off of homes by the departing racial groups, leaving a renting class of
mostly low-income, multi-generational households.
4.2.2. Study population and survey
Participants were sampled using an exploratory strategy, using purposive sampling to
reach out to community practitioners with whom I had volunteered and worked with before, then
snowball sampling as they referred community members to the research. Practically, this entailed
attending community events at after school clubs, community centers, clinics, and pop-up events,
and setting up a table with touchscreen tablets and a scannable code linking to the web GIS
designed on ArcGIS Survey123. I encouraged most participants to complete the survey on the
tablets with my help in case there were problems or questions, and asked participants if they
would rather fill out the survey themselves or for me to read out the answers and dictate their
responses. The latter was consistently more popular and ended up forming a large majority of the
responses. This is something that similar projects should be aware of, making sure that the data
collection is accessible to diverse groups with varying amounts of time.
Participants were first presented with a short description of housing insecurity, which
they either read or I read aloud, and then a screening question, which asked “Have you ever
experienced housing insecurity anywhere in Eastside Los Angeles?” The screening was kept
intentionally broad, focusing on the feeling of insecurity in someone’s own home rather than on
fulfilment of one or more of the UN criteria, in order to sample a more diverse experience. After
screening, participants were presented with an electronic informed consent form, detailing in

177
particular the nature of the de-identification of spatial data, and the use of their responses on the
prototype of the shared platform.
The survey was offered in English and Spanish, which I translated into English at a later
date. For the researcher-assisted entries, I read out the questions and multiple-choice answers in
Spanish, asked the open-ended questions in Spanish, but recorded the responses in English. I
tried to facilitate it so that there were no unnecessarily strict requirements for participation. No
participants asked if other languages were available, but this is something that would have to be
adapted based on locality. The survey took around 10 minutes to fill out, extended if participants
had experiences of housing insecurity in more than one place.
The survey itself started with demographic information recording age, race/ethnicity, and
education attainment. It then collected a history of housing, asking whether participants had a
fixed address, how many times they had moved in the last two years, what the reasons for the
moves were, and whether they had ever lacked a safe and adequate place to sleep. These are
fairly standard questions in housing research (Cox et al., 2016), and allow the countermap to be
compared with conventional materials.
Then participants are directed to record the places where they experienced housing
insecurity. This starts with recording spatial data, which presents participants with three options:
use the device’s recorded location; enter an address; or pan and zoom on the map to draw a
point. Most participants found it easiest to enter an address, especially when completing the
survey with a researcher assistant, using the point on the map to visually confirm the location. A

178
representation of this step is shown in Figure 4.2. This section then asks when the experience
took place, their annual income, and whether the participant was the owner of the property. This
was designed to countermap homeownership in the neighborhoods, with a guiding assumption
that owner-occupiers experience greater security in their living situations (DeLuca and Rosen,
2022), are more financially stable (Aaronson, 2000), and personally invested in their
neighborhoods (Ranganath, 2024).
Figure 4.2. Preview of geographic information section of the survey, as it would appear in a
smartphone browser.

179
If participants did not own the property, the survey asked if they paid rent and if so, how
much each month. Combined with the annual income data, this can be used to calculate rental
burden. For this study rent-only burden was considered the most apposite measure as opposed to
housing cost burden, which includes utilities (Gawrys and Carswell, 2020). Even though utilities
and other costs are often overlooked in their contribution to the severity of an overall housing
burden, they lie in aggregate of more direct expenditures, and are usually subject to policy
directives other than housing, such as energy, sanitation, or telecommunications. Households can
become rent burdened due to low income, high rent prices, or as is common on the Eastside, both
(California Housing Partnership, 2023). This is a crucial measure, I would argue one of the most
important in the survey, after a majority of rent burdened Angelenos were shown to be cutting
back on their consumption of basic needs like food, clothing, and energy in order to afford rent,
or going into debt to keep up with the monthly payment (Kredell, 2020).
It then collected information on number of bedrooms and number of inhabitants. A
household is considered to be overcrowded if there are more people than rooms, which leads to
higher rates of physical and mental health problems (Solari et al., 2012). It is frequently used as
another gauge of the relationship between income and housing affordability, and Los Angeles
has some of the highest rates of overcrowded housing in the U.S. (US Census Bureau, 2022).
Finally on this section, they survey collected experiences related to the UN definitions of
housing insecurity and a checklist of actions that a landlord had ever performed. The first part is
another conventional checklist related to the physical conditions of the unit, and along with the
latter part, were taken from an amalgam of the California Civil Rights Department’s (2024)

180
definitions of housing discrimination, the California Department of Real Estate’s (2022) guide
for residential tenants, my conversations with neighbors and community members, and my
experience as a tenant on the Eastside. Attached to each incidence of insecurity, it also asked
how safe participants felt based on these experiences when they were living in that housing.
Participants had the option to add further experiences in other places in the study area, which
followed the same format.
The last part of the countermapping survey is a set open ended questions which expand
and unfold the place making effects of housing and insecurity. The responses to these are
intended to be presented alongside the relatively abstract quantitative data, lending vibrancy and
relatability to the map. Gilmore (2002) wrote about the “violence” of abstraction; that a removal
of place from enquiry serves to maintain normative views of how people fit into the world. These
questions emphasize embodied knowledges to address that violence, asking how participants’
monthly finances are affected by rent, how housing insecurity affects their lives, what coping
mechanisms they have for dealing with the problems, and what experiences they would like to
map that were not otherwise captured within the confines of the short survey. I also included an
opportunity within one question to describe how their emotions and management strategies for
housing insecurity changed during and after the COVID-19 pandemic given the wholescale
impact that the pandemic had on many Angelenos lives, particularly as related to rent and
eviction outlined in Blasi’s (2020) early report on the crisis, and Manville et al.’s (2023) review
three years later.

181
4.2.3. Deidentification
For the purpose of data presentation in this study, the spatial data were deidentified using
a context-sensitive displacement. There are technical and ethical considerations to make when
using VGI for science that create problematic outcomes for so-called “open data”. There is a
clear conflict here; a desire to present locations which relate accurately to the places where
housing insecurity was experienced without providing materials for retaliation by a landlord or
some other agent. Participants who were concerned that the owner or property manager of their
unit could see the data were reassured that the exact location would be hidden, and that the
deidentification would greatly reduce the chance that a landlord would be able to identify
specific tenants. This contrasts with many popular countermaps which chart evictions, like the
Anti-Eviction Mapping Project’s (2020) map of Ellis Act evictions in San Francisco, which
mostly draw from already publicly available data recording unit addresses and eviction notice
details, rather than from VGI like in this study. In this case, steps need to be taken to reduce the
possibility of singling out, linking, or inferring identity from the points.
Two steps were taken. The first was an adapted version of Hansanzadeh et al.’s (2020)
deidentification technique, which uses a bimodal Gaussian function coupled with donut
anonymization to displace points at random angles within a minimum and maximum radius.
Because local context was key, plus the situation in a relatively compact study area, a spatial
mask was applied on the donut to ensure that the displaced points fell within the enclosing
geometry of census blocks groups – a stable and contiguous geographic unit containing between
600 and 3,000 people – shown in Figure 4.3. The minimum value was set at 50 m and the
maximum to half of the furthest straight line across the enclosing block group. This approach

182
aligns with other work in VGI and public participatory geographic information science (PPGIS),
as well as conventional GIS (e.g., Burgert et al., 2013), which avoids unnecessarily large
displacements in dense areas where identification is reasonably unlikely, like the Eastside. The
context map was also limited to remove building outlines so that deidentified points could not be
corresponded to any specific unit.
Figure 4.3. The displacement method used to deidentify responses based on location
The second was a presentation of some data as aggregated, reducing the likelihood of
attribute or spatial identification. For attributes, this included age, annual income and monthly
rent. For presentation in this study, some of the data is aggregated to the census block group
level.

183
As this study was largely a proof-of-concept exercise, the sample size is relatively small.
Had the web GIS remained open for submissions, or should other countermapping projects
receive many more responses, aggregating data spatially would prove prudent. It requires less
technical skill from the GIS practitioner or data manager and allows for more intuitive spatial
reasoning like comparisons between neighborhoods.
4.2.4. Data analysis
Preliminary data analysis was undertaken to demonstrate the kinds of insights that the
countermap could produce. Summary statistics of demographic and housing data were compiled
in Microsoft Excel. The open response questions were coded using a ground-up inductive
approach.
Given the small sample size, it is difficult to make representative statements. Instead, the
VGI is prized for its locality. The countermap offers a transformative potential to mediate
between the abstract and the concrete, leveraging situated knowledges to enrich the uniformities
and conventions inherent in traditional spatially independent housing research. Given the
profound ontological fractures surrounding housing within the broader frameworks of health and
social research (Easthope, 2004), the participatory approach holds promise, but engenders
difficult analytical prospects. The imperative for countermapping researchers is to remain open
to interrogating, repurposing, and reimagining advancements in interdisciplinary data and
methods, fostering a dynamic engagement with the demands of evolving housing and homeplace landscapes. The steps outlined above are typical in the geographic methodological canon,
relatively straightforward, and I suggest are important groundwork for the development of
building a robust basis for rigorous and meaningful data analysis for housing insecurity.

184
4.2.5. User dashboard
In the spatial sciences, a dashboard is an application used to manage and represent data.
This project used Esri’s ArcGIS Dashboard program to host the housing insecurity platform.
This consisted of a database management system on ArcGIS Online, with the dashboard
designed to provide interactive data visualization. To create a holistic impression, the dashboard
was designed to include a map, a serial chart, embedded content, and several live indicators.
These all presented the data on a real time basis, updated when new VGI data was moderated and
added to the central database.
Like many software-as-a-service (SaaS) models, the Esri programs facilitate collaborative
data collection, management, analysis, and visualization. They also provide on-demand cloud
storage, allowing data to be accessed from any internet browser. The dashboard was formatted
for mobile devices too, in the event that users might not have access to desktop computers.
4.3. Results
4.3.1. Sample population
113 individuals filled out the survey. 26 individuals recorded two places where they had
experienced housing insecurity, and 4 individuals recorded three places, for a total of 143
records. Descriptive demographic statistics are shown in Table 4.1.

185
Table 4.1. Demographic characteristics of the sample
n %
Gender
Female 60 53.1
Male 53 46.9
Age
18-24 6 5.3
25-34 40 35.4
35-44 35 31.0
45-54 21 18.6
55-64 7 6.2
Over 65 4 3.5
Race/Ethnicity
Hispanic or Latinx 99 87.6
White (Non-Hispanic or Latinx) 7 6.2
Black or African American 1 0.9
Asian 3 2.7
Two or More 3 2.7
Highest educational attainment
Some high school 39 34.5
High school 49 43.4
Trade/Vocational Degree 11 9.7
Bachelor's degree 8 7.1
Master's degree 5 4.4
Doctoral degree or higher 1 0.9
The median income across all participants was $26,000, with a mean of $27,395.70.
Excluding participants who reported their income as zero, the mean was $28,195.58. The nonzero range was $90,000, with a minimum of $10,000 and a maximum of $100,000.
The demographic and income data are slightly different from the American Community
Survey data for these neighborhoods. This was to be expected given the requirement to
participate of experiencing housing insecurity in the study area, either currently or in the past.
This is supported by other demographic studies of housing insecurity, which predict lower

186
income, a greater proportion of minority ethnic population, and lower educational attainment
amongst the housing insecure (e.g., Gawrys and Carswell, 2020; Cox et al., 2016).
4.3.2. Housing history
66 participants (58.4%) had not moved within the last two years. 26 (23%) had moved
once, 9 (8%) had moved twice, 4 (3.5%) three times, and 8 (7.1%) four or more times. Across all
participants, 111 (98.2%) were presently housed, and 2 (1.8%) were unhoused at the time of the
data collection. 16 (14.2%) had, in the last two years, lacked their own safe place to regularly
sleep. This might not necessarily indicate homelessness but is increasingly present as a
normative measure of housing insecurity in similar studies (Shelton et al., 2009, Cox et al.,
2016).
Homelessness is the most extreme form of housing insecurity (Henry et al., 2015), and as
such has its own dedicated infrastructure and oversight to address it. Although it is beyond the
scope of this study to determine the appropriateness of the definition of homelessness, or the
success of the resources allocated to the issue, it is prudent here to point out that it is no longer
considered part of the increasingly unified construct of housing insecurity, or at least not on the
same continuum set forth by the most commonly adopted metrics (Cox et al., 2016).
Participants who had moved more than once could enter more than one reason why they
had done so, which produced a total of 209 responses. These are show in Table 4.2.

187
Table 4.2. Reasons why participants had moved home within the last two years (there may have
been more than one reason for each move)
Reasons why participants moved n %
At fault eviction (e.g., non-payment of rent) 12 5.7
No fault eviction (e.g., landlord sells unit) 28 13.4
Contract/lease ended 64 30.6
Financial uncertainty (e.g., couldn't afford housing) 45 21.5
Landlord problem (e.g., failure to make repairs) 37 17.7
Moved from a waiting list into housing 3 1.4
Other 20 9.6
Amongst “other” responses, the reasons ranged from moving out of a family or parents’
house (n = 11), moving to a larger unit (n = 2), moving in with a partner (n = 2), moving for job
or educational reasons (n = 2), buying a home, seeking a safer neighborhood, or being removed
from city housing (all n = 1).
4.3.3. Housing insecurity
Of the 143 locations where housing insecurity had been recorded, 86 (60.1%) were at a
current address, and 57 (39.9%) were at a former address. 46 experiences were happening at the
time of the survey (40.7%), 43 were within the last two years (38.1%), 27 from two to four years
prior (23.9%), 16 from five to nine years prior (14.2%), and 11 over ten years prior (9.7%).
Of reported physical conditions, participants reported 294 separate incidences of inferior
housing quality, shown in Table 4.3.

188
Table 4.3. Physical dimensions of inferior housing
Physical housing inferiority n %
Unit was uncomfortably cold for 24+ hours 36 12.2
Heating broke down 27 9.2
Air Conditioning broke down 17 5.8
Water leak during rain 38 12.9
Holes in the floor 13 4.4
Large cracks in walls 21 7.1
Evidence of rodents or other vermin 41 13.9
Plumbing broke down for more than 6 hours 31 10.5
Windows broken 6 2.0
No working elevator in multi-unit building of four or more storeys 1 0.3
Any other similar experiences not listed 63 21.4
Amongst the other similar experiences, issues with appliances were the most common (n
= 13), followed by excessive heat in the unit and other broken features like fences or roofs (both
n = 9), excessive noise, utilities issues, or broken doors or gate locks (all n = 5), the unit being
extremely dirty or unclean when the tenancy started (n = 4), issue with neighbors or overly
crowded shared units or lots (both n = 3), problems with mold or fire alarms (both n = 2), and
finally flood damage, evidence of roaches and ants, and issues with parking (all n = 1).
Participants recorded whether they pay rent or not for 140 of the locations. Of these, 19
were places where participants did not pay rent (13.6%), and 121 were places where monthly
rent was paid (86.4%). Of those 86, the median monthly rent was $1,300 and the mean was
$1,405.13. This ranged from a minimum of $100 to a maximum of $4,147. The modal value was
$900, with nine locations charging that amount each month.

189
Those who paid rent were also asked about the actions of the landlord or property
manager. Participants recorded 299 landlord actions associated with housing insecurity at 130
places, summarized in Table 4.4.
Table 4.4. Landlord actions associated with housing insecurity
Landlord action n %
Turn up at the unit without reasonable (24 hours) warning 36 12.0
Show the property to prospective tenants or buyers without warning 11 3.7
Enter the property without warning or otherwise abuse their right of
access 23 7.7
Fail to make timely repairs 96 32.1
Threaten tenant by word or gesture 19 6.4
Directly interfere with a right to quiet enjoyment of housing 50 16.7
Discriminate based on actual or perceived characteristics 14 4.7
Refuse to accept or acknowledge receipt of rent 6 2.0
Request information that violates a right to privacy (e.g., residence or
citizenship status) 12 4.0
Any other similar experiences not listed 32 10.7
Amongst other responses, landlords had failed to communicate according to the terms set
out in a lease or as required by state law (n = 8), broken a verbal or otherwise non-official
agreement (n = 8), pressured tenants to move out or threatened eviction (n = 5), failed to clean or
maintain public or common spaces on their lots or complexes (n = 5), made racist remarks (n =
2), raised rent illegally (n = 2), stored property in the rental unit (n = 1), or described simply as
“slumlord” (n = 1). Slumlords are absentee private owners, generally of low-income housing,
who maximize rental income by cutting back on maintenance costs of their properties, often in
violation of housing law. This is a market failure, where the neglect of the property reduces its

190
value in such a way that even though short run operating income is increased, the total
profitability is lower than following a non-slumlord strategy (Lind and Blomé, 2011).
The inefficient allocation of resources is sustainable in Los Angeles, and many other
cities in the US, due to rapidly rising property prices. Although slumlord behavior was only
mentioned by name once, there are many landlord behaviors reported in this study which
correspond with such a strategy. Note that while this is markedly different from the “slum”
designation of many urban areas in developing countries, there is a large American literature
from the 1950s and 1960s which first coined the term (e.g., Stokes, 1962), an overview of which
is provided in Quigley (2000).
Participants were also asked how safe they felt at each home-place where they had
experienced housing insecurity on a 5-point Likert scale. The results of this question are shown
in Figure 4.4, which follows a fairly normal distribution. Safety and its perception are complex,
socially constructed phenomena. Reflected by crime, community integrity, having basic needs
met, and the “sense” of security, safety links to psychological and mental health (Soria et al.,
2022; Wright et al., 2020). Continuously feeling not at all safe is a key component of insecurity.
It increases chronic stress and decreases cognitive function (Soria et al., 2022), which has
disastrous implications for Angelenos and their families who are already facing structural
barriers to education, amongst other challenges.

191
Finally, participants were given the opportunity to respond to four open ended questions.
The first, which asked about how monthly finances were affected by rent, elicited many accounts
of sacrifice and hard choices based on the fundamental opportunity cost of the more that is spent
on housing means less disposable income to spend elsewhere. A selection of these accounts is
given in Figure 4.5.
There was also a common theme surrounding frustration at not being able to build up
savings to be able to afford the down payment for a mortgage, being forced to rent in perpetuity.
Another common theme was the anticipation of not being able to afford emergency expenses
(e.g., medical expenses, auto repair, and unexpected child-related costs).
0
10
20
30
40
50
Not at all safe A little bit safe Somewhat safe Very safe Extremely safe
Figure 4.4. Responses to a Likert scale question on how safe participants felt at each
recorded location of housing insecurity

192
The second question asked how housing insecurity affected the participants’ lives,
particularly whether it was something which was causing stress. Few participants reported “no
stress” or “not really”, with some referencing rent control or strong tenant protection in
California as relief. A majority of responses did recount some kind of stress or anxiety however,
with some typical excerpts shown in Figure 4.6.
There were also some specific accounts where inadequate or inferior housing had resulted
in acute effects. In one case the landlord had failed to replace a faulty freezer, and when it finally
failed, the participant had to throw out several weeks’ worth of frozen food. In another, the house
was so cold that the participants’ infant child was showing signs of hypothermia, so the family
had to relocate to the participant’s partner’s parents’ house to warm up. Even from these short
“Makes it hard to have a financial backup for other monthly expenses.”
“Literally can't afford anywhere better than this. Struggle to have groceries by the
end of the month.”
“I have no money to invest in the future, its all sucked up by rent.”
“Paying rent often leaves me with little to no savings and makes it difficult to afford
other necessities.”
“Takes a significant proportion of income, have to choose between rent or other
essentials like groceries or gas.”
“The rent means I can't afford good food for my family, and sometimes I have to skip
medical appointments. Last month I skipped a dentist appointment because I wouldnt
have afforded the copay after the rent. We also have no family outings like meals out
or vacations.”
Figure 4.5. Excerpts from participants outlining the financial decisions they have to
make each month

193
insights, it’s clear that housing insecurity does impact people’s lives deeply and in diverse,
frequently unpredictable ways.
The third question asked about coping mechanisms, and how management of housing
insecurity had changed during the COVID-19 pandemic. Many participants pointed toward
family and other support networks to cope with the challenges, and some detailed the relief
payments and tax credits which were made available during the COVID-19 stay-at-home orders.
California enacted a fairly robust set of tenant protections, which were lapsing around the time
that I collected data for this study, which few participants seemed aware of. Some typical
responses are recorded in Figure 4.7.
“The unstableness makes mental health challenges, not so much for the rest of my
family as they don't know the full story but they don't exactly think its peaches and
cream.”
“…hard to focus on being happy if you don't know whether you'll have a roof over
your head…”
“An added layer of stress to my family's everyday life. I'm probably more stressed
just because I'm supposed to be the stable parent in charge of everything, so it's hard
when the owner is late to fix things or doesn't pay the water bill.”
“Yes, moved so much in my life that never had a same house for more than five or
six years. That makes it hard to settle down. I wish I could stay somewhere for more
than a year or two.”
“It's like a heavy weight that never lifts off my shoulders. It's not just about having a
stable place … it's also fear and stress. I'm always worried that I'll get an eviction
notice if I don't pay straight away. It messes [sic] with everything, my job, my
relationships. It eats at me day in day out.”
Figure 4.6. Excerpts from participants outlining the stress and anxiety they felt from
housing insecurity-related challenges

194
Some participants named specific communities which they relied on, including church
congregations, local community organizations, ethnic communities and networks in certain
neighborhoods, and co-workers in essential sectors like healthcare and food service. Several
participants were dismayed at their landlords’ insistence that the cost of owning property had
increased, leading to inflated rent, whilst their salaries had either increased less than the rise in
cost of living, stayed the same, or fallen during the pandemic as they were furloughed or laid off.
That of course increases rental burden, creates more difficult decisions about budgeting and
monthly expenditure, and increases precarity.
To close, participants were offered the chance to record any other experiences related to
housing insecurity which they wanted to share as part of the countermap. I was unsurprised by
the apathy felt by many in the current provision and quality of housing: “… need a new system”;
“Became harder to access resources [during COVID-19] because more people were
competing for the funds.”
“I rely on community like neighbors and church to find resources. The stimulus
checks were helpful during Covid but they went straight to backrent.”
“Stay hopeful and call friends for help when things break in my apartment that the
landlord won't do. Pandemic made it hard because people couldn't come round for a
long time.”
“Indefinitely lean on my wife and she leans on me. It was tough in covid when she
was pregnant as she couldn't work full time, so we had no extra money.”
“…budget very carefully. During the pandemic my job slowed down as a restaurant
chef, so I had to postpone moving out of my family house.”
Figure 4.7. Excerpts from participants describing their coping mechanisms for housing
insecurity, and how they had changed over the course of the COVID-19 pandemic

195
“The city doesn’t care about tenants”; or “… we need real change, not just quick fixes.” More
striking though was a pervasive demand to collectivize and share these accounts and feelings, as
well as the sense of solidarity amongst the renting class. Some of these expressions are shown in
Figure 4.8.
There were also instances where participants used this open-ended section to directly ask
for help or for advice where to find different resources. If I was helping those individuals with
the survey, I attempted to provide that as best I could. It was clear though that some people
didn’t know about the resources and assistance available to them in their specific situation, or
even that there were other people with a similar set of problems.
“I appreciate the opportunity to share my experiences and hope it leads to positive
change.”
“Everyone should have affordable housing, irregardless of how much money they
earn. Rich people should not b taking homes from people who cannot afford to live
elsewhere.”
“It's tough but we keep pushing forward. I want to help the situation by doing this
project.”
“I know it's not just me who's struggling, and it isn't fair that people have to worry
so much about what should be something provided by the state.”
“I believe there should be more funding for programs to meet the demand for
everyone who needs it, especially in East LA.”
“I wish there was more resources for changing the situation. For us it's turning out
ok but I know other people who got evicted or had to relocate.”
Figure 4.8. Excerpts from participants demonstrating a shared solidarity, and a desire to
voice housing insecurity concerns as a collective

196
4.3.4. Countermap dashboard
The eventual aim of a countermapping project is to compellingly represent and make
public the VGI. Given the kinds of sentiment expressed in Figure 4.8, as well as the rationale for
Maharawal and McElroy’s (2018) anti-eviction map, it seems as though a user platform showing
located experiences would serve a useful purpose in addressing housing insecurity in urban
communities. To this end, I designed a prototype for an accessible and easy to navigate resource
showing the salient information on one interactive webpage. A screenshot of the result, an
ArcGIS Dashboard, is shown in Figure 4.9.
The dashboard contains several key elements. The main window shows a map of the
study area, which each record of housing insecurity identified by a point, deidentified using the
displacement method detailed earlier. Users have the ability to change the symbology to
represent different demographic and socioeconomic characteristics. In Figure 4.8, the size of
each point is scaled to represent income, with larger points symbolizing higher income. Also
shown in the main data frame are the census block groups which enclose each point. The righthand ribbon is a legend for the main data frame. Any feature on the main map can be selected
and queried, giving a complete, deidentified, summary of the account of housing insecurity at
that place, an example of which is shown in Figure 4.10.

197
Figure 4.9. A prototype of an ArcGIS Dashboard as a user platform for housing insecurity on the Eastside

198
Figure 4.10. Individual points can be selected, producing a pop-up of the more detailed
individual response
The lower ribbon has five boxes. The leftmost box is an embedded video which shows
how to check if a unit is protected by Los Angeles’ rent stabilization ordinance. The remaining
boxes show key data insights from the countermap, including age, number of times displaced,
rental burden across all responses, and average monthly rent. These boxes are configured to be
“live”, whereby filtering and selecting the data in the main data frame will change the results
shown in each indicator. The left-hand ribbon prompts visitors to the site to fill in the survey
themselves, which would populate new results to the map and update the indicators
automatically to reflect the new data. In an edition of the countermap as true community
resource, rather than one moderated by an institutional review board, the data would be available
for download (reflected in an updated consent at the start of new data collection), so community
members would be able to use, analyze, present, and share the data on their own terms.

199
4.4. Discussion
This discussion section focusses on the implication of the results for countermappers. It
discusses the findings and outcomes of my own approach to creating a housing insecurity
platform, addressing the strengths and weaknesses of the employed methods. It begins by
outlining the privacy concerns, then moves onto a deliberation of moderation and maintaining a
countermap platform. The next sections evaluate: the importance of context and spatial data
quality; ownership of data; and multidimensionality as part of the situated knowledges of local
geographic enquiry. The discussion closes with a consideration of the role of a community itself
in monitoring housing insecurity, and a determination on who is given the ability to speak in
multi-stakeholder conversations like these.
4.4.1. Privacy concerns
The principal risk of being identified through the countermap is retaliation by a landlord.
A property owner might seek out their rental units on the map to see if a tenant is complaining,
reporting harassment, contract or legal infractions, or otherwise resisting the power dynamic
inherent in renting a home. With this information in hand, they might raise the monthly rent,
pressure the tenant to move out, begin eviction proceedings unfairly, or perform some other kind
of landlord abuse. This is not a problem with maps like the anti-eviction map (Maharawal and
McElroy, 2018), as the tenants have already been evicted. In this countermap, individuals may be
living in the place where they are currently experiencing housing insecurity, as 60.1% of
participants were. With that in mind, careful steps need to be taken to maintain privacy, like the
displacement technique outlined in the methodology.

200
This is a unique challenge for PPGIS data, as other survey-based methods have not
typically collected spatial coordinates (Brown and Kyttä, 2014). Home address by definition
singles out records which identify an individual in a dataset. In the presented form of the
countermap, I altered the accuracy of the data using randomized donut displacement to weaken
the links between the data and individuals. This made sense based on the relatively small sample
size, but should the countermap be expanded substantially, generalization is likely a better
approach. In that case, records would be spatially generalized into larger spatial regions – census
block groups for example – which contain at least K – 1 other users (Ghinita et al., 2010).
Practically this means aggregating points to enclosing units, grouping them with other
individuals in what is now referred to as “K-anonymity” (Cassa et al., 2006). This would match
other data platforms geared for social change, like the Neighborhood Data for Social Change
(NDSC) or the Opportunity Atlas platforms.
There are two challenges for using K-anonymization here. The first is recruiting
sufficient participants to reduce the likelihood of individual identification, likely many responses
per census block group. The second is that spatial aggregation like this reduces the relatability of
the countermap. One of the great successes of Maharawal and McElroy’s (2018) map is the
ability to inspect individual stories of eviction. This platform grounds knowledge production in
the stories and experiences of neighbors and community, rather than generalized polygons of
speculative real estate return. However, countermaps like this end up being adopted, used, and
shared, so I would strongly advocate for the maintenance of a point-based data structure to
render visible the sites of resistance, as long as there is a meaningful disclaimer that proper steps
have been taken to reduce the chances of retaliation.

201
4.4.2. Moderating the countermap
A key consideration for countermappers then is who will make that disclaimer. Ultimately
some party will need to take responsibility for maintaining the platform. This includes recruiting
new participants, checking the data for accuracy, deidentifying the records in whichever way is
deemed appropriate, and uploading responses to the countermap. As yet, there are no studies
which discuss this issue in the academic literature. I propose two options here: the social
approach and the crowd-sourced approach, with each having distinct benefits and drawbacks.
i. Social approach
Goodchild and Li (2012) outline the social approach thusly: “a hierarchy of trusted
individuals act as voluntary gatekeepers”. Whilst the common application is slightly different,
taken from the Open Street Map (OSM) platform where users who make a large number of VGI
contributions with a low number of subsequent edits are given a high score and assigned new
roles in the Data Working Group (DWG). The DWG takes on the responsibility for dealing with
vandalism, copyright violation, disputes and so on, and have the final say on what is accepted to
the platform. In effect, the social approach emulates the hierarchy of conventional authoritative
mapping, just without the job title or salary.
There are positive and negative aspects of this, some of which could relate to the housing
insecurity countermap. Firstly, this approach usually results in high quality outcomes (Goodchild
and Li, 2012). Despite their constitution from a wide variety of subjective knowledge praxes,
maps are commonly perceived as objective, geographic “truths”. This is one of the subversive
strengths of countermaps, which repurpose the perception of legitimacy to resist the spatial

202
knowledge claims of professionally produced cartographies. Nevertheless, good quality spatial
science is essential for this countermap, to reinforce its validity and shield it from criticism by
logical positivists who might discredit it based on technical deficiency alone.
There are considerable drawbacks to the social approach in this context, however. Firstly,
there is no clear way to establish the hierarchy based on the design of the platform. Participants
submit one response, rather than a large number of contributions common on platforms like
OSM or Wikimapia. Secondly, the role as moderator does demand a considerable time
commitment, almost always uncompensated. It will require initial training by a researcher, and
an expectation to train subsequent moderators. The position also requires a strong degree of
sensitivity. Many of the volunteered contributions contained deeply personal accounts of housing
insecurity, some of which are personally identifying, inflammatory toward specific landlords,
and often both. Selecting which information to deidentify, obfuscate, or even exclude, would be
a key responsibility of a moderator. Lastly, this approach does maintain some of the hierarchical
knowledge networks which produced and maintain the original geographies of inequality.
Goodchild and Li (2012) explicitly use the term “gatekeep” to describe the moderators’ core
onus, which is not ideally aligned with the spirit of countermapping. I don’t doubt that there are
affected individuals in communities across the Eastside who would step up and bear the
responsibility willingly, even with enjoyment, but ultimately, I think that the imposition of
unpaid labor on a demographic already stretched thin by housing insecurity is unfair.

203
ii. The crowd sourcing approach
With regards to PPGIS, crowd sourcing approaches solve problems by referral to a large
number of people, without respect to their qualification or status (Goodchild and Li, 2012). It
relies on Linus’s law, an assertion from software development positing that with a sufficiently
extensive user base, nearly every issue is swiftly characterized, and the solution elucidated. Most
research on crowdsourcing in VGI is related to OSM-style platforms (e.g., Haklay, 2010, Haklay
et al., 2010, Mooney, 2011), where the relationship between number of users and data quality is
clearer than for a countermap.
Crowdsourcing the moderation of the housing insecurity map carries two distinct advantages.
Primarily, it emphasizes the countermap as a product by and for the community in which it was
made. This is a core democratizing value of countermapping and would ensure that the resource
remains a community asset and means for advocacy, rather than the preserve of a few users.
Secondarily, it distributes the workload between many individuals. Once again though, I am
weary of adding to the capacities of community members, especially ones who may want to
submit their response and never have to think about the platform again.
In truth, I anticipate that a community willing to countermap housing insecurity would be
best advised to take on some kind of hybrid between the social and crowdsourced approach.
Distributing the agency and ownership amongst a collective is an epistemologically appealing
notion but should likely be reinforced by a “super user” or similar who matches technical
expertise with sensitivity to their community. Ensuring longevity beyond the research timeline is
often overlooked by countermappers in the academy, who should consider seriously the

204
ramifications of to whom and how they hand over ownership of their work. That is to say, I
strongly believe that participatory work like countermaps should be returned to the communities
who provided the raw materials for their creation, but there needs to be more theoretical and
empirical development on tailoring moderation of conventional VGI to platforms like the one in
this study, based on experiential, embodied data, rather than can purely be described using the
language of accuracy.
4.4.3. Geographic context and spatial data quality
As well as the other requirements of moderation, spatial data remains an important
quality concern. To generate reliable place-based knowledge, there needs to be a meaningful
attachment to the geography of home and home-places. As it stands, home address is a useful
envoy for this, especially in the context of housing insecurity, which is usually tied to the
physical structure as a starting framework. This is convenient for data collection, as people are
familiar with their address, as well as for visualization and translating knowledge, as users can
pan and zoom the map to view their neighbors’ experiences. Home address is a familiar context
for many health and social researchers too, so it should facilitate comparison with other
conventional data.
The measured contextual unit (MCU) of home address does not always relate to the true
contextual unit (TCU) of human experience (Robertson and Feick, 2018), even when dealing
with the home-place geographies of housing insecurity. The results of this study clearly
demonstrate that housing insecurity is an assemblage of diverse spatiotemporal social behaviors.
If a participant was at the grocery store checkout, anxious about whether or not they would have
sufficient money left in their bank account to afford their rent after buying enough to feed their

205
family, or walking back and forth from their parents’ house to their apartment with a backpack
full of clothes because the machines in their laundry room were broken again, home address
starts to make less sense as a solitary site to record insecurity. This is a common assumption
across the spatial sciences, where the conditions of the MCU are taken to be causally related to
the outcomes of individuals in that space (ibid.). Behaviorally, home address is a poor MCU,
failing to account for mobility and environmental heterogeneity. This is not a newly
acknowledged problem, even within housing insecurity literature. The DDHS, in their 1969
description, referred to neighborhood instabilities like crime, unemployment, litter, and noise as
key tenets which might induce feelings of insecurity (Office of the Assistant Secretary for
Planning and Evaluation, 1969), and there are several other spatially tangential attachments
which play into the intersectionality of place-based experiences of home.
I do think that such notions should form part of the critical reflection of countermapping,
but they should not impede the overarching aim to effect social change. With that in mind, home
address should for now serve as a satisficing compromise. It is still a signifier of where
individuals spend time and form attachment, which Jacquez et al. (2015) define as an exposome
centroid.
With regards to spatial data quality, the main concern should not be hyper accuracy –
after all the home addresses are displaced – but rather contextual pertinence. Locals and
community members have a well-developed knowledge of neighborhoods, so should know best
whether the VGI makes sense in the location where it is represented. I foresee spatial error from
two sources related to this, either in mistyping the address which will then be geocoded

206
incorrectly or misplacing a point on a map manually. Addressing this will likely form part of the
moderation, whichever approach or hybrid is taken, using context from the remainder of the
response to appropriately situate the experience. Overall, this countermap approach is unlikely to
be able to provide accurate or reliable results from detailed spatial analysis – there are too many
geographic analysis problems related to inference with spatial data which arise from the use of
conventional data structures to describe the contextual housing insecurity variables. For this
reason, it would be difficult to decipher causation based on the spatial data stored in the
countermap alone in terms of phenomena like environment-individual processes (Robertson and
Feick, 2018). Untangling the spatial associations of multiple and conflicting interpretations has a
stronger chance of success if it is tackled as an intervening and interdisciplinary problem.
4.4.4. Ownership
Another three of the core theses of countermapping and decolonized cartography are
transparency, ownership, and reproducibility. Taking ownership of the VGI is an important step
in restoring the power of the authoritative resource of maps to the participants and their
communities. It is difficult to obtain complete ownership when the data is collected, managed,
analyzed, stored, and presented on proprietary platforms. The tools used for this study are part of
the Esri ecosystem of software, which often require a license and a paid subscription. The license
tends to restrict the following four freedoms, outlined in Steiniger and Hunter (2013) as GIS
freedoms: (i) to run the software for any purpose, (ii) to study and adapt the software for the
users’ own needs, (iii) to freely redistribute the software, and (iv) to improve the software and
make these improvements available to other users. I am under no illusion that every
countermapping project is run by software developers eager to access and adapt the source code,
but I do think that Steiniger and Hunter’s freedoms are nonetheless important. Communities will

207
have different intentions for and demands of their GIS, which may or may not be possible using
the industry- and research-standard Esri systems.
As they stand, ArcGIS Survey123 and ArcGIS Dashboard worked well in the context of
this small-scale survey. Esri has been working to challenge the conventional connotations of socalled “closed source” providers. Firstly, instead of making the data available in proprietary file
formats incompatible with other software, Esri platforms promote interoperability. Many types
of external data, such as the tabularized survey responses in this study, can be integrated,
analyzed, and converted using ArcGIS. Using database connection tools, the data can be rehosted for processing and sharing in real-time, streamlining the flow of information between
users. Esri also offers numerous capabilities for free to communities intending to use GIS for
equity and social justice. These “exceptions”, where users are given access to the tools for no
cost, are made widely in the U.S. and globally for organizations leveraging the technology for
environmental justice, indigenous health services, and education access amongst many other
applications (Esri, 2024).
Alternatively, countermapping initiatives may elect to operate their platforms using opensource technologies. The benefits are manifold, headlined by the lack of a license fee. This grants
a different sense of ownership to the spatial information, reducing reliance on large a large
corporation to make or maintain the types of case-by-case exceptions which Esri makes.
Opensource platforms like QGIS, GeoDa, and GRASS GIS are the products of collaboration
amongst many users, resulting from a mixture of divergent perspectives with no commercial
pressure. This paradigm has historically been more aligned with the goals of countermapping,

208
especially with regards to issues such as housing insecurity. Countermappers will not be
“evicted” from their own platforms, and the spatial knowledge which is rendered visible is not
bound to any institutional continuity. The source code is perennially published for universal
access, which fosters an accountability in the development process, and facilitates continuous
improvement through community feedback and contributions.
Like closed source, these platforms have their own drawbacks too. The lack of
institutional oversight and maintenance could compromise integrity and security, particularly
when it comes to storing data online. Secondly, whilst most open-source platforms have an
enthusiastic user base who solve problems and respond to questions in collaborative digital
spaces like Stack Exchange (the “QGIS” tag has over 10,000 pages of answered questions), there
are no full-time support teams, which may lead to problems for users who consistently encounter
technical issues. There are also growing concerns over data privacy and unauthorized access
resulting in data breaches (Yang et al., 2020), which should be acknowledged by users collecting
sensitive data like housing experiences. Lastly, many open-source GIS rely on third-party
libraries and dependencies. If these are not regularly maintained or updated, they can lead to
disruptions for end-users. Addressing these and other related concerns requires proactive
measures such as regular audits, robust access controls, and community-driven security best
practices. Countermappers should balance the advantages and disadvantages of closed and opensource platforms alike when deciding on a digital site for the performance of their resistance. The
choice should follow a period of reflection on the demands of the project and the needs of the
community in question.

209
4.4.5. Multidimensionality and the role of the locale
There is no convincing unified measure of housing insecurity (Cox et al., 2016). Instead,
researchers, housing analysts, and policymakers use a multidimensional approach to capture
aspects of security and stability, as I did in my VGI survey, using different components and
statistical combinations to indicate different outcomes. The framing of housing insecurity in the
associated disciplines is uncertain and there are multiple definitions of housing, home, place,
insecurity. The countermap which I facilitated provides a wide base from which to interpret the
materials. I wanted to balance this breadth of figures and measures collected with convenience
for participants, not taking up unnecessary amounts of their time. The results collected are useful
for indicating the nature and extent of the problem on the Eastside and serve as a useful proof of
concept for the feasibility of such a platform.
Should this type of countermap be adopted by other communities, there is clear room for
expansion in a number of directions. The route where I see most promise is to develop a set of
specific metrics for the locale. This could be a measurement of response to a local ordinance,
such as Maharawal and McElroy’s (2018) mapping of Ellis Act evictions in San Francisco, a
direct engagement with residents who live along a gentrification corridor like Doucet et al.’s
(2022) demolition and renovation map, or something like Boudjedra’s (1988) literary
countermap of the temporal landscapes of Algerian migration in the Parisian metro. These use
the same guiding philosophy of countermapping – to create new spatial imaginaries which
question cartographic conventions – to upset specific power relations in the local places which
were originally mapped. A community version of the countermap in this project, freed from the
constraints of review by an institution which upholds many of the same power relations, could

210
subscribe to a common ethic – a collectively agreed upon moral code – to map the features
necessary for its own resistance, like naming individual landlords or property management firms
who are breaking the law or withholding their tenants’ rights to housing in some other ways. A
community could decide to subscribe to some nationally used metric, like rental burden, or use
hyper local expert knowledge to map a specific issue. Whichever path is taken, or some
combination of both, the principles of countermapping and VGI are flexible enough to take
account.
In the context of housing there is a conventional emphasis on quantifying, mapping,
surveilling, and otherwise managing bodies through their home-places. Countermaps are
uniquely positioned to collect embodied and experiential information not portrayed in other
maps. Even in the short open-ended responses at the end of the survey in this study, there was a
wealth of place-based knowledge and experience which adds meaning to the countermap
transcending the Cartesian plane. The brief accounts are evidence of corporeal and psychic
experience which demand further inspection and understanding within the context of housing.
There is also eminent possibility within contemporary mapping platforms to attach multimedia
like videos and audio recordings, which would add depth to the sense-making and accounts of
place attachment.
Amidst these discussions is a common critique that the focus on place limits the ability to
“scale up” knowledge or apply the findings from one study area to another. This makes sense
based on Latour’s (1999) notion of circulating reference. Similar to Latour’s argument regarding
the challenges of articulating the bodily experience of sensing a place through written journal

211
entries, I contend that conveying the realities of 113 participants’ housing insecurities through
143 points on the digital map generated here is inherently unattainable. Instead, the countermap
serves as a connective tissue of practice between the ontology of the local, particular, and
material on one side, and the standardized, universal, and textual on the other. This resists the
standardizing thrusts integral to disciplines like global health (Lock and Nguyen, 2010) or the
quantitative social sciences (Neely and Nading, 2017), and dissolution of the world into a
geographically coherent, homogenous globe. The notion, that bodies are interchangeable,
standardized units, fits well in such disciplines, but a normative goal of objectivity has worked to
obscure the power relations which produced inequality in the first place. By prioritizing
experience as rooted in place, the knowledges produced by the countermap are poised to directly
challenge the historical and spatial contingencies producing phenomena like housing insecurity
at the scale of the local. Whilst I think that it will be an enduring challenge to resist the global
forces of capital accumulation and hegemony, a developed network of countermaps designed and
used by communities and collectives themselves to map the geographies of suppression seems
like a strong place to start.
4.4.6. Who produces knowledge
As already discussed, one of the drawbacks of countermapping is in the division of labor.
The data collection is often performed by individuals who live in the places most at-risk of
insecurity. In the daily struggle of life as a marginal body in Los Angeles, filling out forms and
doing the other work of research may be an impossible additional task. What is different about
countermapping endeavors like this one is ownership and agency. Rather than the results being
syphoned to an elite institution or government department, the materials to enact change are held
by the community itself.

212
Maharawal and McElroy (2018) outline some of the differences which this produces.
When the materials are extracted, by journalists or researchers for example, the narratives are
often reduced to stories of victimhood and loss. Tenants are portrayed as docile and passive
actants in the housing network, characterized by inability rather than anything else. In the
“shared authority” engendered by communal and publicly maintained resources like
countermaps, these narratives are countered with stories of effective resistance by a commune
enacting intentional contestation (Frisch, 1990). The countermap serves to share that analytic
authority beyond the time and place of the experience, circulating reference from one space of
insecurity to another.
Even though I have lived on the Eastside for several years, I am still a relative outsider,
representative of a predominantly white research institution. In its current iteration, the platform
records the complexities of social and political dimensions of housing reductively. Should it be
adapted by the community, I suspect that the platform could unfold individual accounts to
incorporate the stories of happiness, aspiration, and contradiction which define home-places as
centers of human attachment. As it stands, the stories produced are City of Quartz-esque (Davis,
2006), punctuated by geographies of despair rather than of hope. I think that countermaps such
as this one hold a promise of being transformed into a participatory ethnographic format,
highlighting stories of individuals navigating the politics of place, forging resistance within the
fabrics of intimate, everyday life centered around the concept of home. I view my role as a
PPGIS researcher as a facilitator and supporter of the production of place-based knowledge,
helping to create the best possible conditions for participation. Place and its attachments are vital

213
components of social justice, with its associated knowledge able to foment collective political
analyses whilst building resistance (Kerr, 2008). The countermap in this study produced very
nuanced data, whose meaning is perhaps best interpreted by the actors who produced it.
Researchers should not stand in the way of this interpretation but should appraise the uncertainty
and messiness of the politics and history of their study areas, helping the communities which
they pledged to serve by providing tools and materials for resistance.
4.5. Conclusion
This countermapping study was a political project of community and place attachment. In
constructing the platform, my guiding enquiry revolved around the possibility for digital
storytelling through narrative data and cartography to authentically represent embodied and
experiential knowledge of the social world, and whether such representation could ever catalyze
meaningful societal transformation. I believe that the countermap holds plenty of answers to
grand questions related to the urban housing “crisis” in Los Angeles and other post-Fordist cities.
Its capacities for archiving, mapping, storytelling, and fomenting political organizing represent a
formidable collection of tactics and strategy for resisting the sinister practices of speculative real
estate, slumlording, and other acts of capital accumulation tied to the human right to housing.
The first of three goals for this study was to design and test a dynamic dashboard of
housing security using VGI. This was successfully achieved but pointed towards a new set of
challenges with regards to moderation, privacy concerns, context and spatial data quality, data
and platform ownership, and multidimensionality as part of a consideration of the locale. It also
raised important questions over who is able to speak in these discussions, and how mapping and

214
countermapping privilege and silence certain worldviews. Ultimately, I advocated for the
handing over by the researcher of as many tools and concepts as possible for the enabling of
social change, but to remain attentive to the requirements and specificities of the community
where the work is being done. Lastly, I acknowledged the substantial labor required to design,
build, and maintain a countermapping platform like this one, and to avoid adding to the burdens
of communities by unloading that research work nonreflexively. In the case of this project, I am
working to find local stakeholders who may be interested in learning more about housing
countermapping insecurity and would seek to share the data and platform with them as
constructively as possible.
The second goal was to produce spatial artefacts conducive to tenant advocacy and
political mobilization. The VGI platform collected important place-based data related to
conventional definitions of housing insecurity (e.g., United Nations, 2014, Office of the
Assistant Secretary for Planning and Evaluation, 1969). Amongst this, the most striking figure
was the average rental burden of 67% across all participants, meaning over two-thirds of
monthly income was dedicated to rent alone, far exceeding the national and county averages.
There are many developed empirical practices for measuring all manner of housing outcomes,
which I argued should be tailored and adapted to match the geographic context of the locale, to
showcase the marginalization of entire communities. In this eventuality, all of the metrics were
designed and collected by me, with my limited, but growing, understanding of the four
neighborhoods and their situations within the history and space of the city. As is common for
urban geographers and housing researchers, many of the narratives which resulted were bleak,
drawing on Davis’ (2006) dystopian and apocalyptic accounts of Los Angeles. These have been

215
understandably inspiring for scholars, but the countermap which I produced relies too heavily on
a cartography of suffering, despite the numerous contributions from participants who expressed
home and willingness to address and change their and their neighbors’ social situations.
The third goal was to critically evaluate the countermapping actor-network in a
community setting. This was a running thread throughout the work, demonstrating the kind of
reflection that I believe is critical for the successful and enduring application of VGI to
challenges of housing, health and place. Although it exists solely as a digital resource, I think
that the materiality of the finished platform is important: how it looks, how people are able to
interact with it, and how it is able to circulate reference as a dashboard. The arguments presented
have broader implications for the way geographers and housing scholars think and practice
embodied research, and challenge common assumptions about the technologically mediated
discipline of GIS and the spatial sciences.
Lastly, I would like to use this work as a case study to remind researchers that Los
Angeles, as Scott and Soja (1996, p. 460) have written, “as always, is worth watching). The
mythic of one of the world’s largest cities by population, size, and economy offers a great
diversity of gifts, challenges, and opportunities. Soja (1996b) explained himself how population
growth and sprawling development created some of the cruelest repercussions of urban renewal,
which are still seen across the country and around the world nearly 30 years later. The jobshousing balance combined with powerful forces of race and class create sharp topographies of
inequality, but also geographies of resistance. The Eastside, with its histories of economic
deprivation and political radicalism, is a fascinating site of American urbanism. Exploring

216
housing insecurity and learning about so many people’s experience and attachment to their
home-places have been incredibly insightful and inspiring, and I hope to inspire researchers and
community members to practice countermapping and other defiant techniques in the face of
state-led corporate capitalism, and in doing so begin to reclaim the means of knowledge
production.

217
Chapter 5 Conclusion
There can be no single solution to addressing the inequalities inscribed into the urban
fabric of Los Angeles. This dissertation raised a multitude of further questions, highlighting
profound issues of what kinds of knowledge are possible and desirable, and how they might be
achieved. Instead of proposing a singular answer, it asked what might even begin to constitute an
answer: who or what is able to answer back, what is a solution, what is truth, what is credible,
and what is expertise and how is it made. This formulation echoes Rose (2003), who asserted the
non-linear process of research design, and Latour (2005), in his tying together of heterogeneous
bundles to translate some type of knowledge into another.
To revisit one of the statements in the introduction, this was fundamentally a project in
situated knowledges, engaging with subjects not as fixed, but as a process in which they were
changing and transforming across space. The other fundamental assertion, that everyone is expert
in their own lives, was addressed by centering the effect of the place. Identity and knowledge are
produced in part by the effect of the places in which they are situated, so challenging or
destabilizing those places becomes a political act. The contexts of the three experiments, in food,
mobility, and housing, are continuously subject to the destabilizing forces of capital, and the
subsumption, dispossession, or abandonment which follows. I attempted to identify the issues at
stake for actors who are routinely excluded from the conventional boundaries of knowledge
controversies, namely a “lay” public, and create alternative forums to reconfigure their expertise
into political agency.

218
5.1. Key findings
The first experiment was a test of my own capacity as a researcher to participate in the
production of knowledge beyond the site of a desk or office. I went into the Eastside, my study
area, and walked down every street looking for evidence. By moving my body, using my senses
to get a sense, and interacting with the human and more-than-human features of the assemblage,
I gained a totally different understanding of the place. I spent weeks immersed in the food
environment, literally tasting the products of global supply chain. This was a different experience
than was afforded to the other members of the National Science Foundation-funded team, which
was investigating food systems in the same place, who used a different set of methods to build
their understandings. What resulted was truly greater than the sum of its parts, a triangulated and
“initiated” (Rossman and Wilson, 1985) base of evidence. The key empirical finding was that the
“low-access, low income” model used by the USDA belies the true provision of the food system,
depending on the definition of “food environment” used in the chapter. Still, the food
environment on the Eastside stands as a material testament to the distortions of long-standing
structural conflict that is yet to be resolved, or even clearly articulated.
The second experiment was a blend of spatial and visual methods, out of which I built a
new technique, the carto-photo elicitation interview. I used it to explore the corporeal and
psychological outcomes of navigating Los Angeles without a car, which because of urban
renewal and decades of automobile path dependency, is to do so in a hostile physical
infrastructure as a marginalized body. The PPGIS created a set of deep thick descriptions,
simultaneously outlining different behaviors and the social conditions in which they take place.
If this method were to become stabilized within the spatial sciences as a useable and useful

219
participatory technique, the uptake of co-produced knowledge would widen the notion of
expertise and change the power relations intrinsic to so many conventional tools of inquiry. If the
breadth of expertise gathered by CPEI was anything to go by, this should be an ambition across
disciplines, changing the ways in which scientists and the public engage with each other in
relation to shared matters of concern like movement and mobility. The key finding was that
carlessness produces a wide variety of psychic experiences of difference, and mobilities scholars
should account for the co-constitution of movement and navigation with networks of knowledge,
power, and bodies.
The last experiment was an outcome of participative practice through volunteered
geographic information, using collective experiences of exploitation within home-places to resist
the disordering conditions of the Los Angeles housing market. The end-product, the interactive
user platform, is an artifact of the generative event in the knowledge controversy. As Whatmore
(2009) explains, this should create opportunities which arouse a different awareness of the
problems and situations around us that mobilize a public, reconfiguring the scientific division of
labor and reclaiming accountability and innovation from the political elite who created these
conditions in the first place. The key finding was that experiencing housing insecurity permeates
into all manner of everyday geographies, especially in places like the Eastside where the average
rental burden amongst participants was 67%. Rather than public deficit (Braun, Whatmore and
Stengers, 2010), which produces an expert-lay boundary to deprive people of rights to participate
in decision-making, countermapping like this authorizes a completely different population to
produce and take ownership of knowledge, taking advantage of the generative politics which
result to challenge the status quo and work to improve the social situation.

220
These approaches, as first-, second-, and third-person accounts of expertise, produced
unpredictable accounts of experience on the Eastside, with interesting findings that have
important consequences for the situation both within and beyond their contexts. Each one
validated arguments made from certain theoretical and analytical standpoints, undermining
conventional wisdom and proving their importance in the science and politics at the nexus of
health and place. Where it is not clear who has the expertise or “epistemic authority” (Cheek,
2007) to inform decision-making, or how to carry out open dialogue with stakeholders, hopefully
there is something from these methods which will improve the likelihood of success. I think that
clumsy solutions are not just relevant for addressing wicked problems, where it is not clear what
the controversy is about, what types of knowledge need to be produced, or where the limits of
debate ought to be, but imperative. The spatial sciences are well positioned to inform the
development of clumsy solutions, as just as everyone is expert in their own lives, everything
happens somewhere. Basing health research in an investigation of place yields results which are
meaningful and translatable, far more so than if investigated from a spatially abstract, “view
from nowhere” position.
5.2. Limitations and future work
5.2.1. Small data for interdisciplinary learning
One of the core features of this work was that it focused on small sample sizes in specific
locales. Whilst this allowed for deep understandings of the particular and the material, there is an
understandable concern that this limits the possibilities for standardization and relative
universality. It is no coincidence that these are the two poles at either end of Latour’s (1999)
reversible chain of transformations making up circulating reference. At each step between them,

221
some properties are lost, and others are gained, rendering phenomena at the meeting point
between things and the human mind compatible with already-established centers of calculation.
An important area for future research is thus how traditional “small data”, including the
qualitative and critical work here, can be used together with “big” data for overcoming the
limitations of the latter. There are plenty of datasets on food, mobilities, and housing, all of
which serve valuable purposes from different interpretive standpoints.
Whilst I view the multi-sited ethnography as a strength, paying close attention to the
details of both the social world and the practices used to record it, others might see this as a
blinkered view of reality. Fortunately, the techniques and methods I employed have an intrinsic
tendency for interdisciplinary collaboration, due in large part to the shared language and culture
with big data scientists within the spatial sciences. Because actor-network theory is in some way
an ethnographic approach in itself, for understanding the social as a set of associations, there is
no reason why it can’t fold in perspectives from outside the worldviews of human geography or
science and technology studies. In particular, I would encourage collective learning and
empirical experimentation with macro data scientists and health behavior researchers.
5.2.2. Geographies of despair and new publics
There was a tendency in this work, because of its ANT focus on tensions, conflicts and
power dynamics, to present a pessimistic or gloomy view of science and politics. The
deconstruction of power structures at the core of its three chosen knowledge controversies
focused on the hidden dynamics and inequalities which impact the daily lives of marginal bodies
in Los Angeles. In many cases, marginality was highly racialized, as the entrenched power
dynamics of post-Fordist America usually are. On the Eastside this meant a focus on the

222
Hispanic and Latinx population, thanks to its proximity and close ties to Mexico and Central
America.
Whilst I questioned the reliability of many institutions, practices, and knowledge systems,
particularly those maintained and operated by a predominantly white political elite, I do think
that many of the “bottom up”, “grassroots” organizations I worked with over the course of these
studies offer promise for empowering individuals and communities to take collective action and
shape their own futures. The community centers, street clinics, rehabilitation programs,
communal gardens, and countless social initiatives have built community resilience through
addressing local challenges including the social inequalities I researched myself. As
collaborative efforts, their success highlights the potential for collective action to effect
meaningful change, and I would love to see more scholars of health and place collaborate with
these kinds of groups.
Building networks like this, between the academy and the grassroots, harnesses diverse
knowledges, and gives rise to a new “public” for political agency. In this way, the alternative
forums for reconfiguring expertise can work to challenge the association of wicked problems
with geographies of despair, catalyzing long-term systemic change by mobilizing people around
a shared value system. There are well-defined arenas for this kind of work, like environmental
injustices embedded within the low-income communities of color found all over the City of
Quartz, but the analytical frameworks used here should be able to deepen understanding of the
heterogeneous networks of actors involved in many different urban settings.

223
5.3. Summative reflection
Something in this place forced me to think. Through my fundamental encounters with
people and things, I moved through this work and through these places conscious of how I could
hope to formalize the experience. In the end, I think that the affective tones I grasped were
appropriate, even unavoidable. Every interaction I had over the last five years was a result of
experimentation; moving to a different place to study new meaning in ways that I never had
before. I found it to be interesting, important, and intriguing. The Eastside is unlike anywhere I
had been before, materialized from the politics, creative practices and cultural production its
Chicanx and other minority ethnic residents. I was constantly reminded of my place: walking
down Whittier as a white Englishman; talking to neighbors about how much rent I was paying on
the doorsteps of houses their parents had built; driving past hundreds of people sleeping on the
pavement; asking in broken Spanish whether the chef could make me some vegan tacos. My
experience was a product of the places in which it happened, and the relationships I built there,
all of which is important to restate as a scholar of situated knowledges.
At the end of the dissertation, I am uncertain of the degree to which the research
questions did guide my methodological choices. I am very aware of the view of how the
relationship between research questions and research methods should operate, thanks in no small
part to the grants I have worked on as a research assistant during my doctorate. I don’t think this
was the way it operated in practice. I was not constantly thinking in terms of Rayner’s (2006) or
Rittel and Webber’s (1973) coherent perspectives on world policy, though perhaps this is one of
the defining characteristics of wicked problems after all: nobody really knows what they are,
how to define them, or who the agents who bring about the solution will be. In terms of scientific

224
norms then, perhaps this does convey a sense of rationality to the craft of research, and the entire
thing has been one clumsy solution in itself, balancing my own ontology, epistemology,
methodology, and normative expectations to enact practice.
As a dissertation which was built on fieldwork, it produced some observations of the field
itself, which is often taken for granted in geography as a place separated from the theory, where
a researcher can be expected to have authentic empirical encounters with the world. Ultimately, I
do not think that the field was a distinct location for me, a terra incognita to brood and speculate
within and then to be removed from to write a paper. There was no great separating out of neat
domains, partially because most of it took place within a shared physical space. Clearly though,
even this was ontologically unstable, as the notion of the study area itself was constantly brought
into question: what is the Eastside of Los Angeles? Demanding oneself to inspect such seemingly
sturdy categories is a formative part of work like this, not only for geographers who are prying
apart the assumptions connecting space and place, but for the growing group of public health
researchers undertaking place-based research. Rather than creating one huge ontological gap
between the world and the words I used to describe it, my fieldwork created many tiny gaps,
across which I tried to build bridges and connections using methodology. This is a somewhat
complex spatial metaphor, closely related to circulating reference, and one which I will continue
to think about beyond my pages and words. I definitely think that paying close attention to the
details of practice was invaluable, particularly from as close up as I was. Becoming an active
participant in the production of knowledge required critical reflexion, particularly on the radical
indeterminacy of the materials which were generated. The most exciting outcomes were when
the field “spoke back”.

225
5.4. Dissertation takeaway
To conclude, I have argued that upstream processes of capital accumulation and a stateled corporate political economy have resulted in geographies of inequality, felt in places like the
Eastside through the impacts on people’s everyday lives. As outlined in the introduction,
expertise in the knowledge controversies of wicked problems is a social and cultural process
which creates politically generative events in its formation, authorization, circulation, and
dispute. Where expertise takes place is a critical consideration as it constructs situated
knowledges and sites for participation.
Chapters 2, 3, and 4 outlined PGIS, PPGIS, and VGI approaches to knowledge
controversies respectively, with detailed accounts of each technique’s possibilities, as well as the
set of deliberations which must accompany their practice. From the findings, I suggest that
clumsy solutions are viable procedures for engaging an extended peer community innovatively
and flexibly, derived from cultural theory. Research is as much an affective as an intellectual
process, and I advocate for scientists to take greater account of their own and others’ first-order
experiences as part of the distributed assemblage of geographic enquiry.

226
References
Aaronson, D. (2000). A Note on the Benefits of Homeownership. Journal of Urban Economics,
47(3), 356-69.
Adams St. Pierre, E. (2016). The long reach of logical positivism/logical empiricism. In N.
Denzin & M. Giardina (Eds.) Qualitative Inquiry through a critical lens (pp. 27-38).
Routledge: Abingdon.
Allam, Z., Bibri, S. E., Chabaud, D., & Moreno, C. (2022). The ‘15-Minute City’ concept can
shape a net-zero urban future. Humanities and Social Sciences Communications, 9(1), 1-5.
Allman-Farinelli, M., Rahman, H., Nour, M., Wellard-Cole, L., & Watson, W. L. (2019). The
role of supportive food environments to enable healthier choices when eating meals prepared
outside the home: Findings from focus groups of 18 to 30-year-olds. Nutrients, 11(9), 2217.
Anderson, B., & Harrison, P. (2006). Questioning affect and emotion. Area, 38(3), 333-335.
Anti-Eviction Mapping Project (2020). All San Francisco Eviction Notices, 1997-2020.
Retrieved from: https://antievictionmap.com/sf-evictions (Accessed 3/6/24).
Ashbrook, D., & Starner, T. (2003). Using GPS to learn significant locations and predict
movement across multiple users. Personal and Ubiquitous Computing, 7, 275-286.
Baker, P., Machado, P., Santos, T., Sievert, K., Backholer, K., Hadjikakou, M., Russell, C.,
Huse, O., Bell, C., Scrinis, G., & Worsley, A. (2020). Ultra‐processed foods and the nutrition
transition: Global, regional and national trends, food systems transformations and political
economy drivers. Obesity Reviews, 21(12), e13126.
Bal, M., & Bryson, N. (1991). Semiotics and art history. The Art Bulletin, 73(2), 174-208.
Barrett, C. B. (2010). Measuring food insecurity. Science, 327(5967), 825-828.
Beaulac, J., Kristjansson, E., & Cummins, S. (2009). Peer reviewed: A systematic review of food
deserts, 1966-2007. Preventing Chronic Disease, 6(3), A105.
Benton, L. M. (1995). Will the real/reel Los Angeles please stand up? Urban Geography, 16(2),
144-164.
Bernsdorf, K. A., Bøggild, H., Aadahl, M., & Toft, U. (2022). Validation of retail food outlet
data from a Danish government inspection database. Nutrition Journal, 21(1), 60.
Bess, K. D., Miller, A. L., & Mehdipanah, R. (2023). The effects of housing insecurity on
children’s health: a scoping review. Health Promotion International, 38(3), daac006.

227
Bhat, A. C., Almeida, D. M., Fenelon, A., & Santos-Lozada, A. R. (2022). A longitudinal
analysis of the relationship between housing insecurity and physical health among midlife and
aging adults in the United States. SSM-Population Health, 18, 101128.
Bissell, D. (2010). Passenger mobilities: affective atmospheres and the sociality of public
transport. Environment and Planning D: Society and Space, 28(2), 270-289.
Bitler, M., & Haider, S. J. (2011). An economic view of food deserts in the United States.
Journal of Policy Analysis and Management, 30(1), 153-176.
Blasi, Gary (2020). UD Day: Impending Evictions and Homelessness in Los Angeles. UCLA
Luskin Institute on Inequality and Democracy: Los Angeles, CA.
Blinn, L. and Harrist, A.W. (1991) Combining native instant photography and photo- elicitation.
Visual Anthropology 4: 175-92.
Blumenberg, E., Brown, A., & Schouten, A. (2020). Car-deficit households: determinants and
implications for household travel in the US. Transportation, 47, 1103-1125.
Blunt, A., & Dowling, R. (2022). Home (2nd ed.). Routledge: Abingdon.
Boudjedra, R. (1988). Topographie idéale pour une agression caractérisée (extraits). Diversité,
72(1), 45-46.
Bowen, S., Elliott, S., & Hardison‐Moody, A. (2021). The structural roots of food insecurity:
How racism is a fundamental cause of food insecurity. Sociology Compass, 15(7), e12846.
Bratt, R. G. (2002). Housing and family well-being. Housing studies, 17(1), 13-26.
Brown, G., & Kyttä, M. (2018). Key issues and priorities in participatory mapping: Toward
integration or increased specialization? Applied Geography, 95, 1–8.
Bryman, A. (2006). Integrating quantitative and qualitative research: how is it done? Qualitative
Research, 6(1), 97-113.
Burgert, C. R., Colston, J., Roy, T., & Zachary, B. (2013). Geographic displacement procedure
and georeferenced data release policy for the Demographic and Health Surveys. DHS Spatial
Analysis Reports No. 7. ICF International: Calverton, MD.
California Civil Rights Department (2024). Housing Discrimination. Retrieved from:
https://calcivilrights.ca.gov/housing/ (Accessed 3/6/24).
California Department of Real Estate (2022). California tenants: A guide to residential tenants’
and landlords’ rights and responsibilities. Official Publication of the California Department
of Real Estate: Sacramento, CA.

228
California Housing Partnership (2023). Los Angeles County 2023 Affordable Housing Needs
Report. Retreived from: https://chpc.net/wp-content/uploads/2023/05/Los-AngelesCounty_Housing-Report_2023.pdf (Accessed 3/6/24).
Canalia, C., Pinho, M. G. M., Lakerveld, J., & Mackenbach, J. D. (2020). Field validation of
commercially available food retailer data in the Netherlands. International Journal of
Environmental Research and Public Health, 17(6), 1946.
Cannuscio, C. C., Tappe, K., Hillier, A., Buttenheim, A., Karpyn, A., & Glanz, K. (2013). Urban
food environments and residents’ shopping behaviors. American Journal of Preventive
Medicine, 45(5), 606-614.
Carins, J. E., Rundle-Thiele, S., & Storr, R. J. (2019). Appraisal of short and long versions of the
Nutrition Environment Measures Survey (NEMS-S and NEMS-R) in Australia. Public Health
Nutrition, 22(3), 564-570.
Caspi, C. E., & Friebur, R. (2016). Modified ground-truthing: An accurate and cost-effective
food environment validation method for town and rural areas. International Journal of
Behavioral Nutrition and Physical Activity, 13(1), 1-8.
Cassa, C. A., Grannis, S. J., Overhage, J. M., & Mandl, K. D. (2006). A context-sensitive
approach to anonymizing spatial surveillance data: Impact on outbreak detection. Journal of
the American Medical Informatics Association, 13(2), 160–165.
Cloke, P., May, J., & Johnsen, S. (2008). Performativity and affect in the homeless
city. Environment and Planning D: Society and Space, 26(2), 241-263.
Clough, P.T. (2008) ‘The affective turn: Political economy, biomedia and bodies’, Theory,
Culture and Society. 25: 1-22.
Cohen, N., Chrobok, M., & Caruso, O. (2020). Google-truthing to assess hot spots of food retail
change: A repeat cross-sectional Street View of food environments in the Bronx, New York.
Health & Place, 62, 102291.
Coleman-Jensen, A., Nord, M., Andrews, M., & Carlson, S. (2012). Household food security in
the United States. Economic Research Report ERR-141. USDA ERS: Washington, DC.
Connell, K., & Hilton, M. (2015). The working practices of Birmingham's Centre for
Contemporary Cultural Studies. Social History, 40(3), 287-311.
Conradson, D., & Latham, A. (2007). The affective possibilities of London: Antipodean
transnationals and the overseas experience. Mobilities, 2(2), 231-254.
County of Los Angeles Public Health. (2022). Food Businesses. Retrieved from
http://publichealth.lacounty.gov/eh/business/food-businesses.htm#foodBusiness (Accessed
2/2/23)

229
Cox, R., Henwood, B., Wenzel, S., & Rice, E. (2016). Roadmap to a Unified Measure of
Housing Insecurity. University of Southern California Center for Economic and Social
Research. CESR-Schaeffer Working Paper Series, Paper No: 2016-013.
Croghan, R., Griffin, C., Hunter, J., & Phoenix, A. (2008). Young people's constructions of self:
Notes on the use and analysis of the photo‐elicitation methods. International Journal of Social
Research Methodology, 11(4), 345-356.
Cummins, S., & Macintyre, S. (2002). “Food deserts”—Evidence and assumption in health
policy making. BMJ, 325(7361), 436-438.
Cutter, S. L. (1995). Race, class and environmental justice. Progress in Human
Geography, 19(1), 111-122.
Data Axle. (2021). Business Data. Retrieved from https://www.data-axle.com/our-data/businessdata/ (Accessed 2/2/23)
Davis, M. (1990) City of Quartz: Excavating the Future in Los Angeles. Vintage Books: New
York, NY.
Davison, J., & Warren, S. (2017). Visual methodologies for accounting and accountability. The
Routledge companion to qualitative accounting research methods, 112-128. Routledge:
London.
De la Haye, K., Wilson, J., Bruine de Bruin, W., Li, K., Livings, M. S., Xu, M., Miller, S.,
Solanky, M., Weber, K., Babboni, M., Wald, F., & Frazzini, A. (2021). Enough to Eat: The
Impact of COVID-19 on Food Insecurity and the Food Environment in L.A. County April
2020-September 2021. Retrieved from https://publicexchange.usc.edu/wpcontent/uploads/2021/10/Enough-to-Eat.pdf (Accessed 1/28/23).
Deleuze, G. (1988). Spinoza: practical philosophy (1st English ed.). City Lights Books: San
Francisco, CA.
DeLuca, S., & Rosen, E. (2022). Housing insecurity among the poor today. Annual Review of
Sociology, 48, 343-371.
Ding, H., Loukaitou-Sideris, A., & Wasserman, J. L. (2022). Homelessness on public transit: A
review of problems and responses. Transport Reviews, 42(2), 134-156.
Doucet, B., Diwan, F., Turman, W., Mehta, N., & Petrovic, A. (2022) Mapping displacement
through lived experiences: Countermapping transit-induced gentrification in KitchenerWaterloo, Ontario. Radical Housing Journal, 4(2), 67-87.
Downs, S. M., Ahmed, S., Fanzo, J., & Herforth, A. (2020). Food environment typology:
Advancing an expanded definition, framework, and methodological approach for improved

230
characterization of wild, cultivated, and built food environments toward sustainable diets.
Foods, 9(4), 532.
Du Bois, J. W. (1992). Discourse transcription. Santa Barbara Papers in Linguistics, 4, 1-225.
Duarte, F. (2017). Space, place and territory: A critical review on spatialities. Taylor and
Francis: New York, NY.
Dubowitz, T., Dastidar, M. G., Troxel, W. M., Beckman, R., Nugroho, A., Siddiqi, S., Cantor, J.,
Baird, M., Richardson, A., Hunter, G., Mendoza-Graf, A. & Collins (2015). Healthy food
access for urban food desert residents: Examination of the food environment, food purchasing
practices, diet and BMI. Public Health Nutrition, 18(12), 2220-2230.
Dubowitz, T., Dastidar, M. G., Troxel, W. M., Beckman, R., Nugroho, A., Siddiqi, S., Cantor, J.,
Baird, M., Richardson, A., Hunter, G., Mendoza-Graf, A. & Collins, R. L. (2021). Food
insecurity in a low-income, predominantly African American cohort following the COVID-19
pandemic. American Journal of Public Health, 111(3), 494-497.
Dunford, E. K., & Neal, B. (2017). FoodSwitch and use of crowdsourcing to inform nutrient
databases. Journal of Food Composition and Analysis, 64, 13–17.
Easthope, H. (2004). A place called home. Housing, Theory, and Society, 21(3), 128–138.
Elevelt, A., Bernasco, W., Lugtig, P., Ruiter, S., & Toepoel, V. (2021). Where you at? Using
GPS locations in an electronic time use diary study to derive functional locations. Social
Science Computer Review, 39(4), 509-526.
Elliott, W. (1996). The Jews of Boyle Heights, 1900-1950: The Melting Pot of Los
Angeles. Southern California Quarterly, 78(1), 1-10.
Emmison, M., & Smith, P. D. (2000). Researching the Visual: Images, Objects, Contexts and
Interactions in Social and Cultural Inquiry (1st ed.). Sage: London.
Esri (2024). Equity and Social Justice Community. Retrieved from: https://www.esri.com/enus/industries/corporate-equity/community (Accessed 4/14/24).
Estrada, G. (2005). If you build it, they will move: The Los Angeles freeway system and the
displacement of Mexican East Los Angeles, 1944-1972. Southern California
Quarterly, 87(3), 287-315.
Fang, D., Thomsen, M. R., Nayga, R. M., & Yang, W. (2022). Food insecurity during the
COVID-19 pandemic: Evidence from a survey of low-income Americans. Food Security, 14,
165-183.
Fawns, T. (2020) The photo-elicitation interview as a multimodal site for reflexivity. In P.
Reavey (Ed.) Visual Methods in Psychology (pp. 487–501). Routledge: London.

231
Food and Agriculture Organization. (1996). Rome Declaration on World Food Security. 13–17th
November, Rome.
Foster, T. B., Knop, B., & Bhaskar, R. (2021). Access and Eligibility for Supplemental Nutrition
Assistance Program Varies County by County. United States Census Bureau.
Franco, M., Roux, A. V. D., Glass, T. A., Caballero, B., & Brancati, F. L. (2008). Neighborhood
characteristics and availability of healthy foods in Baltimore. American Journal of Preventive
Medicine, 35(6), 561-567.
Fratini, C. F., & Jensen, J. S. (2017). The Role of Place-specific Dynamics in the Destabilization
of the Danish Water Regime: An Actor–Network View on Urban Sustainability Transitions. In:
Frantzeskaki, N., Broto, V. C., Coenen, L., & Loorbach, D. (Eds.). Urban Sustainability
Transitions (pp. 86-105). Routledge: Oxford.
Frisch, M. (1990). Shared authority: Essays on the craft and meaning of oral and public history.
State University of New York Press: Albany, NY.
Frith, H., & Harcourt, D. (2007). Using photographs to capture women's experiences of
chemotherapy: Reflecting on the method. Qualitative Health Research, 17(10), 1340-1350.
Fullilove, M. T. (2020). From ‘root shock’ to ‘Main Street’: For a biopsychosocial urban
psychology. Journal of Urban Regeneration & Renewal, 13(3), 251-256.
Garnett, T., Appleby, M. C., Balmford, A., Bateman, I. J., Benton, T. G., Bloomer, P.,
Burlingame, B., Dawkins, M., Dolan, L., Fraser, D., Herrero, M., Hoffman, I., Smith, P.,
Thornton, P., Toulmin, C., Vermeulen, S. & Godfray, H. (2013). Sustainable intensification in
agriculture: Premises and policies. Science, 341(6141), 33-34.
Gawrys, M. R., & Carswell, A. T. (2020). Exploring the cost burden of rural rental
housing. Journal of Rural Studies, 80, 372-379.
Geertz, C. (1973). The interpretation of cultures. Routledge: New York, NY.
Ghinita, G., Zhao, K., Papadias, D., & Kalnis, P. (2010). A reciprocal framework for spatial Kanonymity. Information Systems, 35(3), 299–314.
Ghosh-Dastidar, M., Hunter, G., Collins, R. L., Zenk, S. N., Cummins, S., Beckman, R.,
Nugroho, A., Sloan, J., Wagner, L. & Dubowitz, T. (2017). Does opening a supermarket in a
food desert change the food environment? Health & Place, 46, 249-256.
Gibbons, J. (2023). Examining the long-term influence of New Deal era redlining on
contemporary gentrification. Urban Studies, 60(14) 2816-2834.

232
Gilmore, R.W. (2002). Fatal couplings of power and difference: Notes on racism and geography.
The Professional Geographer, 54(1), 15–24.
Giuliani, M. (1991) “Towards an analysis of mental representations of attachment to the home”,
The Journal of Architectural and Planning Research 8(2), 133–146.
Glanz, K., Sallis, J. F., Saelens, B. E., & Frank, L. D. (2007). Nutrition Environment Measures
Survey in stores (NEMS-S): Development and evaluation. American Journal of Preventive
Medicine, 32(4), 282-289.
Gosliner, W., Brown, D. M., Sun, B. C., Woodward-Lopez, G., & Crawford, P. B. (2018).
Availability, quality and price of produce in low-income neighborhood food stores in
California raise equity issues. Public Health Nutrition, 21(9), 1639-1648.
Gould, K. A. (2019). Got Narcan? Dimensions of Critical Care Nursing, 38(1), 1-4.
Greene, J. C., Caracelli, V. J., & Graham, W. F. (1989). Toward a conceptual framework for
mixed-method evaluation designs. Educational Evaluation and Policy Analysis, 11(3), 255-
274.
Gregg, M., & Seigworth, G. J. (2010). The affect theory reader. Duke University Press, Durham,
NC.
Grim, V. (2012). The Black Farmers Class Action Lawsuit Against the United States Department
of Agriculture, 1997-2010. In D. Reid & E. Bennett (Eds.), Beyond Forty Acres and a Mule:
African American Farm Families after Freedom (pp. 271-296). University Press of Florida:
Greensville, FL.
Habermas, J. (1983). Interpretive social science vs. hermeneuticism. In N. Haan, R. Bellah, P.
Rabinow & W. Sullivan Social science as moral inquiry (pp. 251-270). Columbia University
Press: New York, NY.
Haklay, M. (2010). How Good is volunteered geographical information? A comparative study of
OpenStreetMap and ordnance survey datasets. Environment and Planning B: Planning and
Design 37(4), 682–703.
Haklay, M., Basiouka, S., Antoniou, V., & Ather, A. (2010). How many volunteers does it take
to map an area well? The validity of Linus’s Law to volunteered geographic information.
Cartographic Journal 47(4), 315–322.
Hall, S. (1997). Culture and power. Radical Philosophy, 86(27), 24-41.
Hallum, S. H., Hughey, S. M., Wende, M. E., Stowe, E. W., & Kaczynski, A. T. (2020). Healthy
and unhealthy food environments are linked with neighborhood socio-economic disadvantage:
An innovative geospatial approach to understanding food access inequities. Public Health
Nutrition, 23(17), 3190-3196.

233
Han, E., Powell, L. M., Zenk, S. N., Rimkus, L., Ohri-Vachaspati, P., & Chaloupka, F. J. (2012).
Classification bias in commercial business lists for retail food stores in the US. International
Journal of Behavioral Nutrition and Physical Activity, 9(1), 1-9.
Hansen, H., & Metzl, J. M. (2019). Structural competency in mental health and medicine: A
case-based approach to treating the social determinants of health. Springer: New York, NY.
Haraway, D. (1988). Situated knowledges: The science question in feminism and the privilege of
partial perspective. Feminist Studies, 14(3), 575-599.
Haraway, D. (1991). Simians, Cyborgs and Women. Routledge: New York, NY.
Harcourt, B. E., & Ludwig, J. (2006). Broken windows: New evidence from New York City and
a five-city social experiment. University of Chicago Literature Review, 73, 271.
Harper, D. (2002). Talking about pictures: A case for photo elicitation. Visual Studies, 17(1), 13-
26.
Harvey, D. (1989). From managerialism to entrepreneurialism: the transformation in urban
governance in late capitalism. Geografiska Annaler: Series B, Human Geography, 71(1), 3-
17.
Harvey, D. (1996) Justice, Nature and the Geography of Difference. Blackwell Publishers:
Malden, MA.
Hayashi, K. T. (2019). Making Home Again: Japanese American Resettlement in Post-World
War II Los Angeles, 1945-1955. University of California, Riverside.
Hendrickson, D., Smith, C., & Eikenberry, N. (2006). Fruit and vegetable access in four lowincome food deserts communities in Minnesota. Agriculture and Human Values, 23, 371-383.
Hill, A. B. (2017). Critical inquiry into Detroit's “food desert” metaphor. Food and Foodways,
25(3), 228-246.
Hjorth, L., & Richardson, I. (2017). Pokémon GO: Mobile media play, place-making, and the
digital wayfarer. Mobile Media & Communication, 5(1), 3-14.
Hodgetts, D., Chamberlain, K., & Radley, A. (2007). Considering photographs never taken
during photo-production projects. Qualitative Research in Psychology, 4(4), 263-280.
Holland, J. H. (2000). Emergence: From chaos to order. Oxford University Press: Oxford.
Home Owners Loan Corporation (1939). Residential Security Maps. Creative Commons.
Hookway, B. (2014). Interface. MIT Press: Cambridge, MA.

234
Hystad, P., Amram, O., Oje, F., Larkin, A., Boakye, K., Avery, A., Gebremedhin, A., & Duncan,
G. (2022). Bring Your Own Location Data: Use of Google Smartphone Location History Data
for Environmental Health Research. Environmental Health Perspectives, 130(11), 117005–
117005.
Jacquez, G. M., Sabel, C. E., & Shi, C. (2015). Genetic GIScience: Toward a place-based
synthesis of the genome, expo- some, and behavome. Annals of the Association of American
Geographers, 105(3), 454–472.
Jingnan, H. (2020). Grocery stores start to cut hours as Coronavirus prompts surge in panicbuying. NPR, March 14. Retrieved from:
https://www.npr.org/2020/03/14/815911461/grocery-stores-start-to-cut-hours-as-coronavirusprompts-surge-in-panic-buying (Accessed 7/25/23).
Jones, A. D., Ngure, F. M., Pelto, G., & Young, S. L. (2013). What are we assessing when we
measure food security? A compendium and review of current metrics. Advances in Nutrition,
4(5), 481-505.
Joyner, L., Yagüe, B., Cachelin, A., & Rose, J. (2022). Farms and gardens everywhere but not a
bite to eat? A critical geographic approach to food apartheid in Salt Lake City. Journal of
Agriculture, Food Systems, and Community Development, 11(2), 67-88.
Juarez, L. (2021). 3 Kroger grocery stores in Los Angeles to permanently close over “hero pay”
Saturday. ABC7 Los Angeles, May 15th.
Kaiser, M. L., Carr, J. K., & Fontanella, S. (2019). A tale of two food environments: Differences
in food availability and food shopping behaviors between food insecure and food secure
households. Journal of Hunger and Environmental Nutrition, 14(3), 297-317.
Kashima T (1980) Japanese-American internees return, 1945–55: Readjustment and social
amnesia. Phylon, 41, 107–115.
Kelley, R. D. G. (2017). “What did Cedric Robinson Mean by Racial Capitalism?” Boston
Review, January 12th.
Kelling, G. L., & Wilson, J. Q. (1982). Broken windows. Atlantic Monthly, 249(3), 29-38.
Kemp, S. P. (2010). Place, history, memory: Thinking time within place. In Burton, L. M.,
Kemp, S. P., Leung, M., Matthews, S. A., & Takeuchi, D. T. (Eds.). Communities,
Neighborhoods, and Health: Expanding the Boundaries of Place (pp. 3-19). Springer: New
York, NY.
Kerr, D. (2008). Countering corporate narratives from the street: The Cleveland Homeless oral
history project. In P. Hamilton and L. Shopes (Eds.), Oral history and public memories
(pp.231–51). Temple University Press, Philadelphia, PA.

235
Kitchin, R. (2014). Big Data, new epistemologies and paradigm shifts. Big data and society,
1(1).
Kitchin, R., Gleeson, J., & Dodge, M. (2013). Unfolding mapping practices: a new epistemology
for cartography. Transactions of the Institute of British Geographers, 38(3), 480-496.
Kolb, K. H. (2021). Retail inequality: Reframing the food desert debate. University of California
Press: Berkeley, CA.
Kredell, M. (2020, December 14). Many L.A. renters cut back on basic needs to make rent USC
Price Center Study shows. USC Sol Price School of Public Policy. Retrieved from:
https://priceschool.usc.edu/news/many-l-a-renters-cut-back-on-basic-needs-to-make-rent-uscprice-center-study-shows/ (Accessed 3/6/24).
Krims, A. (2007). Music and Urban Geography (1st ed.). Taylor and Francis: Abingdon.
Kwan, M.P. (2002). Feminist visualization: Re-envisioning GIS as a method in feminist
geographic research. Annals of the Association of American Geographers, 92(4), 645–61.
Kwan, M.P., & Schwanen, T. (2009). Quantitative revolution 2: The critical (re) turn. The
Professional Geographer, 61(3), 283-291.
Kwan, M.P., & Schwanen, T. (2016). Geographies of mobility. Annals of the American
Association of Geographers, 106(2), 243-256.
Kwan, M.P. (2012). The uncertain geographic context problem. Annals of the Association of
American Geographers, 102 (5): 958–68.
Lammes, S., & Wilmott, C. (2018). The map as playground: Location-based games as
cartographical practices. Convergence, 24(6), 648-665.
Langdridge, D., Gabb, J., & Lawson, J. (2019). Art as a pathway to impact: Understanding the
affective experience of public engagement with film. The Sociological Review, 67(3), 585- 601.
Latour, B. (1999). Pandora’s Hope: Essays on the Reality of Science Studies. Harvard University
Press: Cambridge, MA.
Latour, B. (2005). Reassembling the Social: An Introduction to Actor-Network-Theory. Oxford
University Press: Oxford.
Law, J. (2004). After Method: mess in social science research. Routledge: London.
Law, J. (2016) STS as Method in Felt, U., Fouché, R., Miller, C. A., & Smith-Doerr, L. (Eds.). The
handbook of science and technology studies. MIT Press: Cambridge.

236
Law, J., & Urry, J. (2004). Enacting the social. Economy and society, 33(3), 390-410.
Lefebvre, H. (1991). The production of space. (Trans. Nicholson-Smith, D). Oxford: Blackwell.
Light, I. and Roach, E. (1996) Self-employment. In P. Waldinger and M. Bozorgmehr (Eds.)
Ethnic Los Angeles (pp. 193–213). Russell Sage Foundation: New York, NY.
Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. Sage: New York, NY.
Lind, H., & Blomé, G. (2011, June 15th-19th). The return of the Swedish slumlord: Analysis of a
recent case. European Real Estate Society Annual Conference, Eindhoven, Netherlands.
Lipsitz, G. (2004). Learning from Los Angeles: Another one rides the bus. American
Quarterly, 56(3), 511-529.
Lock, M., & Nguyen, V.K. (2010). An Anthropology of Biomedicine. Wiley-Blackwell, Malden,
MA.
Los Angeles Conservancy (2023). William Mead Homes. Retrieved from:
https://www.laconservancy.org/learn/historic-places/william-mead-homes/ (Accessed 3/5/24).
Los Angeles County Office of the Assessor. (2022). LA County Parcel Map Service. Retrieved
from: https://hub.arcgis.com/datasets/lacounty::la-county-parcel-map-service (Accessed
8/14/23).
Los Angeles County Office of the Assessor (2024). Property Assessment Information System.
Retrieved from: https://maps.assessor.lacounty.gov/m/ (Accessed 3/4/2024).
Los Angeles County Metropolitan Transportation Authority (2023). Final environmental impact
report for the Los Angeles aerial rapid transit project, Los Angeles, California. Prepared by
Kimley-Horn.
Los Angeles Metro (2017). Pilot program begins to put WiFi on 150 Metro buses. Retrieved
from: thesource.metro.net/2017/06/15 (Accessed 2/10/24).
Los Angeles Times (2014). East-sigh story: Just where is the Eastside of Los Angeles? Retreived
from: https://www.latimes.com/local/la-me-eastside-battle-20140218-m-story.html (Accessed
5/6/24).
Los Angeles Times (2020). Coronavirus impacts worsen conditions for California workers.
Retrieved from: https://www.latimes.com/politics/story/2020-09-04/meatpacking-wasalready-a-dangerous-job-for-california-workers-the-pandemic-made-it-worse (Accessed
5/2/23).
Los Angeles Times (2021). An outdoor night market blossoms in Lincoln Heights as L.A.
recovers from COVID-19’s devastation. Retrieved from:

237
https://www.latimes.com/california/story/2021-05-27/tacos-tiktok-help-la-avenue-26-nightmarket-shine (Accessed 8/9/23).
Los Angeles Times (2010). About Mapping L.A. Retrieved from:
https://maps.latimes.com/about/#:~:text=In%20June%202010%2C%20the%20map,Hughes%
20landed%20on%20the%20map (Accessed 8/9/23).
Low, K. (2009). Scents and scent-sibilities: smell and everyday life experiences. Cambridge
Scholars Publishing: Cambridge.
Lucan, S. C., Maroko, A. R., Patel, A. N., Gjonbalaj, I., Elbel, B., & Schechter, C. B. (2020).
Healthful and less-healthful foods and drinks from storefront and non-storefront businesses:
Implications for ‘food deserts’,‘food swamps’ and food-source disparities. Public health
nutrition, 23(8), 1428-1439.
MacDonald, C. (2012). Understanding participatory action research: A qualitative research
methodology option. The Canadian Journal of Action Research, 13(2), 34-50.
Mackenbach, J.D., Charreire, H., Glonti, K., Bárdos, H., Rutter, H., Compernolle, S., De
Bourdeaudhuij, I., Nijpels, G., Brug, J., Oppert, J.M. & Lakerveld, J. (2019). Exploring the
relation of spatial access to fast food outlets with body weight: a mediation analysis.
Environment and Behavior, 51(4), 401-430.
Maharawal, M. (2014). Protest of gentrification and eviction technologies in San Francisco.
Progressive Planning, 199, 20–24.
Maharawal, M. M., & McElroy, E. (2018). The Anti-Eviction Mapping Project: Counter
Mapping and Oral History Toward Bay Area Housing Justice. Annals of the American
Association of Geographers, 108(2), 380-389.
Manville, M., Monkkonen, P., Lens, M. C., & Green, R. (2023). Renter nonpayment and
landlord response: Evidence from COVID-19. Housing Policy Debate, 33(6), 1333-1367.
Markowitz, L. (2012). Highland Haute Cuisine: The Transformation of Alpaca Meat. In E.
Finnis (Ed.). Re-Imagining Marginalized Foods (pp. 34–48). University of Arizona Press:
Tucson, AZ.
Marks, L. U. (2000). The skin of the film: Intercultural cinema, embodiment, and the senses.
Duke University Press: Durham, NC.
Martin, P., Liaw, W., Bazemore, A., Jetty, A., Petterson, S., & Kushel, M. (2019). Adults with
housing insecurity have worse access to primary and preventive care. The Journal of the
American Board of Family Medicine, 32(4), 521-530.
Massey, D. (1995) “The conceptualization of place”, in D. Massey and P. Jess (Eds.) A Place in
the World? Places, Cultures and Globalization (pp. 45–85). Oxford University Press: Oxford.

238
Massey, D. (2004). Geographies of responsibility. Geografiska Annaler: Series B, Human
Geography, 86(1), 5-18.
McCloy, L., White, S., Lee Bunting, K., & Forwell, S. (2016). Photo-elicitation interviewing to
capture children's perspectives on family routines. Journal of Occupational Science, 23(1), 82-95.
McCormack, D. P. (2008). Engineering affective atmospheres on the moving geographies of the
1897 Andrée expedition. Cultural Geographies, 15(4), 413-430.
McDowell, L. (2016). ‘Doing Gender: Feminism, Feminists and Research Methods in Human
Geography'. In L. McDowell, & J. Sharp (Eds.) Space, Gender, Knowledge: Feminist
Readings (pp. 105-114). Routledge: Abingdon.
Meo, A. I. (2010). Picturing students' habitus: The advantages and limitations of photoelicitation interviewing in a qualitative study in the city of Buenos Aires. International
Journal of Qualitative Methods, 9(2), 149-171.
Metzl, J. M., & Hansen, H. (2014). Structural competency: theorizing a new medical engagement
with stigma and inequality. Social Science and Medicine, 103, 126-133.
Miller, B., & Ponto, J. (2016). Mobility among the spatialities. Annals of the American
Association of Geographers, 106(2), 266-273.
Modly, A. R. (2009). Geography of fear: Understanding students’ sense of
place. Chrestomathy, 8, 115-146.
Monahan, T. (2002). Los Angeles studies: The emergence of a specialty field. City &
Society, 14(2), 155-184.
Montenegro de Wit, M. (2021). What grows from a pandemic? Toward an abolitionist
agroecology. The Journal of Peasant Studies, 48(1), 99-136.
Mooney, P. (2011, September 7-8th). The evolution and spatial volatility of VGI in
OpenStreetMap. Hengstberger Symposium Towards Digital Earth: 3D Spatial Data
Infrastructures, Heidelberg, Germany.
Mui, Y., Headrick, G., Raja, S., Palmer, A., Ehsani, J., & Porter, K.P. (2022). Acquisition,
mobility and food insecurity: Integrated food systems opportunities across urbanicity levels
highlighted by COVID-19. Public Health Nutrition, 25(1), 114-118.
National Institute of Health. (2017). How dietary factors influence disease risk. Retrieved from:
https://www.nih.gov/news-events/nih-research-matters/how-dietary-factors-influence-diseaserisk (Accessed 5/5/23).

239
Neely, A. H., & Nading, A. M. (2017). Global health from the outside: The promise of placebased research. Health & Place, 45, 55-63.
Newell, P., & Mulvaney, D. (2013). The political economy of the ‘just transition’. The
Geographical Journal, 179(2), 132-140.
Newton, H.P. (1996) The War against the Panthers. Harlem River Press: New York, NY.
November, V., Camacho-Hübner, E., & Latour, B. (2010). Entering a risky territory: Space in the
age of digital navigation. Environment and Planning D: Society and Space, 28(4), 581-599.
O’Grady, N. (2018). Geographies of Affect. Oxford Bibliographies. Accessed 2/3/24. Retrieved
from: https://www.oxfordbibliographies.com/display/document/obo-9780199874002/obo9780199874002-0186.xml
Office of the Assistant Secretary for Planning and Evaluation. (1969). Ancillary services to
support welfare to work: Housing instability. U.S. Department of Health & Human Services:
Washington, DC.
Ohri-Vachaspati, P., & Leviton, L. C. (2010). Measuring food environments: a guide to available
instruments. American Journal of Health Promotion, 24(6), 410-426.
Ohri-Vachaspati, P., Martinez, D., Yedidia, M. J., & Petlick, N. (2011). Improving data accuracy
of commercial food outlet databases. American Journal of Health Promotion, 26(2), 116-122.
Open Food Facts. (2022). Methodology. Retrieved from: https://world.openfoodfacts.org
(Accessed 5/2/23).
Packard, J. (2008). ‘I'm gonna show you what it's really like out here’: the power and limitation
of participatory visual methods. Visual Studies, 23(1), 63-77.
Parady, G., Suzuki, K., Oyama, Y., & Chikaraishi, M. (2022). The effectiveness of using Google
Maps Location History data to detect joint activities in social networks. Travel Behaviour and
Society, 30, 344-357.
Pardo, M. (1990). Mexican American Women Grassroots Community Activists: "Mothers of
East Los Angeles". Frontiers: A Journal of Women Studies, 11(1), 1-7.
Parekh, T., Xue, H., Cheskin, L. J., & Cuellar, A. E. (2022). Food insecurity and housing
instability as determinants of cardiovascular health outcomes: A systematic review. Nutrition,
Metabolism and Cardiovascular Diseases, 32(7), 1590-1608.
Partington, S. N., Menzies, T. J., Colburn, T. A., Saelens, B. E., & Glanz, K. (2015). Reduceditem food audits based on the nutrition environment measures surveys. American Journal of
Preventive Medicine, 49(4), e23-e33.

240
Patel, R. (2008). Stuffed and starved: The hidden battle for the world food system. Melville
House Pubublishing: New York, NY.
Pavlovskaya, M. (2009). Non-quantitative GIS. In M. Cope, & S. Elwood, (Eds.) Qualitative
GIS: A mixed methods approach (pp. 13–37). Sage: Los Angeles, CA.
Pocock, D. (1989). Sound and the Geographer. Geography, 74(3), 193–200.
Post, R. C. (1987). America's electric railway beginnings: trollers and daft dummies in Los
Angeles. Southern California Quarterly, 69(3), 203-221.
Powell, L.M., Han, E., Zenk, S.N., Khan, T., Quinn, C.M., Gibbs, K.P., Pugach, O., Barker,
D.C., Resnick, E.A., Myllyluoma, J. & Chaloupka, F.J. (2011). Field validation of secondary
commercial data sources on the retail food outlet environment in the US. Health & Place,
17(5), 1122-1131.
Pradeilles, R., Irache, A., Wanjohi, M. N., Holdsworth, M., Laar, A., Zotor, F., Tandoh, A.,
Kmlomegah, S., Graham, F., Muthuri, S., Kimano-Murage, E., Colemna, N., Green, M., Osei,
Kwasi, H.A., Bohr, M., Rousham, E., Asiki, G., Akparibo, R., Mensah, K., Aryeetey, R.,
Brixas, N., & Griffiths, P. (2021). Urban physical food environments drive dietary behaviors
in Ghana and Kenya: A photovoice study. Health & Place, 71, 102647.
Pulido, L. (2002). Race, class, and political activism: Black, Chicana/o, and Japanese‐American
leftists in Southern California, 1968–1978. Antipode, 34(4), 762-788.
Quigley J. (2000) A Decent Home: Housing Policy in Perspective. UC Berkeley: Berkeley
Program on Housing and Urban Policy. Retrieved from
https://escholarship.org/uc/item/8f57x42q (Accessed 3/14/24).
Ranganath, P. (2024). Homeownership. Neighborhood Data for Social Change. Retrieved from:
https://la.myneighborhooddata.org/2021/06/homeownership/ (Accessed 3/6/24).
Reese, A. M. (2019). Black food geographies: Race, self-reliance, and food access in
Washington, DC. UNC Press Books: Raleigh, NC.
Revill, G. (2016). How is space made in sound? Spatial mediation, critical phenomenology and
the political agency of sound. Progress in Human Geography, 40(2), 240-256.
Ritchie, J., Spencer, L., & O’Connor, W. (2003). Carrying out qualitative analysis. In J. Ritchie,
J. Lewis, C. McNaughton Nicholls, R. Ormston (Eds.) Qualitative research practice: A guide
for social science students and researchers (pp. 219-262). Sage: London.
Robbins, O. (2020). From food deserts to Food Oases: Addressing access to healthy food.
Retrieved from: https://foodrevolution.org/blog/food-deserts-food-oasis-healthy-food-access/
(Accessed 5/5/23).

241
Robertson, C., & Feick, R. (2018). Inference and analysis across spatial supports in the big data
era: Uncertain point observations and geographic contexts. Transactions in GIS, 22(2), 455-
476.
Rodriguez, L.J. (2002). The Republic of East L.A.: Stories (1st ed.). Rayo: New York.
Rose, G. (1997). Situating knowledges: positionality, reflexivities and other tactics. Progress in
Human Geography, 21(3), 305-320.
Rose, G (2003) ‘Asking Questions’ in M. Pryke, G. Rose, S. Whatmore (Eds.) Using Social
Theory: Thinking Through Research (pp. 9-11). Sage: London.
Rose, G. (2004). ‘Everyone's cuddled up and it just looks really nice’: an emotional geography of
some mums and their family photos. Social & Cultural Geography, 5(4), 549-564.
Rose, G. (2012). Visual methodologies: an introduction to researching with visual materials (3rd
ed.). Sage: London.
Rossman, G. B., & Wilson, B. L. (1985). Numbers and words: Combining quantitative and
qualitative methods in a single large-scale evaluation study. Evaluation Review, 9, 627-643.
Ruktanonchai, N. W., Ruktanonchai, C. W., Floyd, J. R., & Tatem, A. J. (2018). Using Google
Location History data to quantify fine-scale human mobility. International Journal of Health
Geographics, 17, 1-13.
Saelens, B. E., Glanz, K., Sallis, J. F., & Frank, L. D. (2007). Nutrition Environment Measures
Study in restaurants (NEMS-R): development and evaluation. American Journal of Preventive
Medicine, 32(4), 273-281.
Sánchez, G. J. (2004). "What's Good for Boyle Heights Is Good for the Jews": Creating
Multiracialism on the Eastside during the 1950s. American Quarterly, 56(3), 633-661.
Sanusi, A., & Palen, L. (2008). Of coffee shops and parking lots: Considering matters of space
and place in the use of public Wi-Fi. Computer Supported Cooperative Work, 17, 257-273.
Schwanen, T. (2016). Rethinking resilience as capacity to endure: Automobility and the
city. City, 20(1), 152-160.
Scott, A. J. and Soja, E. W. (1996) Introduction to Los Angeles: City and Region. In A.J. Scott,
and E.W. Soja (Eds.), The City: Los Angeles and Urban Theory at the End of the Twentieth
Century (pp.1-21). University of California Press: Berkeley, CA.
Sen, A., (2000). Social exclusion: concept, application, and scrutiny. Asian Development Bank:
Manila.

242
Shannon, J. (2015). Rethinking food deserts using mixed-methods GIS. Cityscape, 17(1), 85-96.
Shaw, H. J. (2006). Food deserts: Towards the development of a classification. Geografiska
Annaler: Series B, Human Geography, 88(2), 231-247.
Sheller, M., & Urry, J. (2006). The new mobilities paradigm. Environment and Planning
A, 38(2), 207-226.
Sheller, M., & Urry, J. (2016). Mobilizing the new mobilities paradigm. Applied Mobilities, 1(1),
10–25.
Shelton, K., Taylor, P., Bonner, A., & van den Bree, M. (2009). Risk factors for homelessness:
Evidence from a population-based study. Psychiatric Services, 60, 465–472.
Shotland, R. L., & Mark, M. M. (1987). Improving inferences from multiple methods. In M. M.
Mark & R. L. Shotland (Eds.), Multiple methods in program evaluation: New Directions for
Program Evaluation, pp.77-94. Jossey-Bass: San Francisco, CA.
Shower of Hope (2023). Locations. Retrieved from: https://theshowerofhope.org/resources/
(Accessed 2/15/24).
Simpson, P. (2009). ‘Failing on deaf ears’: A postphenomenology of sonorous
presence. Environment and Planning A, 41(11), 2556-2575.
Sloan, A. (2022). Pigford v. Glickman and the Remnants of Racism. Oil and Gas, Natural
Resources, and Energy Journal, 8(1), 19.
Smith, B. J. (2022). Food justice, intersectional agriculture, and the triple food movement.
Agriculture and Human Values, 36, 825-835.
Soja, E.W. (1989). Postmodern geographies. Verso: London.
Soja, E. W. (1996a) Thirdspace: Journeys to Los Angeles and Other Real and-Imagined Places.
Blackwell: Cambridge.
Soja, E. W. (1996b) Los Angeles, 1965-1992: From crisis-generated restructuring to
restructuring-generated crisis. In: A.J. Scott and E.W. Soja (Eds.) The City: Los Angeles and
Urban Theory at the End of the Twentieth Century (pp.426-462). University of California
Press: Berkeley, CA.
Soja, E. W. (2014). My Los Angeles: From urban restructuring to regional urbanization.
University of California Press: Berkeley, CA.
Solari, C. D., & Mare, R. D. (2012). Housing crowding effects on children’s wellbeing. Social
Science Research, 41(2), 464-476.

243
Sonnino, R., Marsden, T., & Moragues‐Faus, A. (2016). Relationalities and convergences in
food security narratives: towards a place‐based approach. Transactions of the Institute of
British Geographers, 41(4), 477-489.
Soria, K. M., Horgos, B., & Shenouda, J. D. (2023). Disparities in college students’ financial
hardships during the COVID-19 pandemic. Journal of Student Affairs Research and Practice,
60(1), 31-48.
Stahre, M., VanEenwyk, J., Siegel, P., & Njai, R. (2015). Housing insecurity and the association
with health outcomes and unhealthy behaviors, Washington State, 2011. Preventing Chronic
Disease, 12(7), E109.
Stokes C J (1962) A Theory of Slums. Land Economics, 38(3), 187-197.
Swyngedouw, E. (2004). Social power and the urbanization of water: flows of power. Oxford
University Press: Oxford.
Tan, Q. H. (2012). Towards an affective smoking geography. Geography Compass, 6(9), 533-545.
Testa, A., & Jackson, D. B. (2019). Food insecurity, food deserts, and waist-to-height ratio:
variation by sex and race/ethnicity. Journal of Community Health, 44, 444-450.
Thomson, T. J., & Uddin, S. (2023). Contemporary ways of seeing: exploring how smartphone
cameras shape visual culture and literacy. Journal of Visual Literacy, 42(4), 269-286.
Thrift, N. (2008). Non-representational theory: Space, Politics, Affect. Routledge: London.
Tiehen, L. & Frazão, E. (2016). Where Do WIC Participants Redeem Their Food Benefits? An
Analysis of WIC Food Dollar Redemption Patterns by Store Type, USDA ERS, EIB-152.
Tuan, Y-F. (1974) Topophilia: a study of environmental perception, attitudes, and values.
Englewood Cliffs: London.
Tuan, Y-F. (1979). Space and Place: Humanistic Perspective. In S. Gale, & G. Olsson (Eds.)
Philosophy in Geography. Theory and Decision Library (pp.387-427). Springer, Dordrecht.
Tyko, K. (2020). Stores cutting hours or closing because of coronavirus: Walmart, Apple, Nike,
Albertsons, Trader Joe’s, more. USA Today, March 14th.
U.S. Census Bureau. (2021). ACS Poverty Data Tables. Retrieved from:
https://www.census.gov/topics/income-poverty/poverty/data/tables/acs.html (Accessed 8/9/23).
U.S. Census Bureau. (2022). American Community Survey 5-year estimates, Table B25014,
Tenure by Occupants per Room. Retrieved from
https://data.census.gov/table/ACSDT5Y2021.B25014?q=B25014&g=1400000US481576722
02 (Accessed 3/5/24).

244
U.S. Census Bureau. (2022). Areas Published. Retrieved from:
https://www.census.gov/programs-surveys/acs/geography-acs/areas-published.html (Accessed
8/9/23).
U.S. Census Bureau. (2023). American Community Survey 5-year estimates, Table B25044. Retrieved
from: https://www.census.gov/data/developers/data-sets/acs-5year.html (Accessed 2/15/24).
U.S. Census Bureau. (2023a). American Community Survey 5-year estimates, Table B03002,
Hispanic or Latino Origin by Race. Retrieved from
https://www.census.gov/data/developers/data-sets/acs-5year.html (Accessed 2/2/24).
U.S. Census Bureau. (2023b). American Community Survey 5-year estimates, Table B25003,
Tenure. Retrieved from https://www.census.gov/data/developers/data-sets/acs-5year.html.
(Accessed 3/5/24).
United Nations. (2014). The right to adequate housing. (Fact Sheet No. 21, Rev. 1). Retrieved
from: http://www.ohchr.org/Documents/Publications/FS21_rev_1_Housing_en.pdf (Accessed
3/4/2024).
USDA Economic Research Service. (2012). Access to Affordable and Nutritious Food: Updated
Estimates of Distance to Supermarkets Using 2010 Data. USDA ERS Report Number 143.
USDA Economic Research Service. (2022a). Key Statistics and Graphs. Retrieved from:
https://www.ers.usda.gov/topics/food-nutrition-assistance/food-security-in-the-u-s/keystatisticsgraphics/#:~:text=In%202021%3A,with%20adults%2C%20were%20food%20insecure
(Accessed 4/25/23).
USDA Economic Research Service. (2022b). Food Access Research Atlas Documentation.
Retrieved from: https://www.ers.usda.gov/data-products/food-access-researchatlas/documentation/ (Accessed 1/9/23).
Uteng, T. P., & Cresswell, T. (2008). Gendered mobilities. In T. Cresswell, T. Uteng (Eds.)
Gendered Mobilities (pp. 1-12). Routledge: Abingdon.
Van Wel, F., & Knijn, T. (2007). Single mothers’ motivation to work and their participation in
the labour market in the Netherlands. International Journal of Sociology of the Family, 33(1),
183–197.
Veroff, J., & DiStefano, A. (2002). Special issue: Researching across difference. American
Behavioral Scientist, 45(8), 1185-1307.
Viesca, V. H. (2014). the battle of Los Angeles: the Cultural Politics of Chicana/o Music in the
Greater eastside. In Understanding Deviance (pp. 496-503). Routledge: Abingdon.

245
Vonthron, S., Perrin, C., & Soulard, C. T. (2020). Foodscape: A scoping review and a research
agenda for food security-related studies. PLoS One, 15(5), e0233218.
Walker, R. E., Keane, C. R., & Burke, J. G. (2010). Disparities and access to healthy food in the
United States: A review of food deserts literature. Health & Place, 16(5), 876-884.
Widener, M., & Shannon, J. (2014). When are food deserts? Integrating time into research on
food accessibility. Health & Place, 30, 1–3.
Winkler, M. R., Lenk, K. M., Caspi, C. E., Erickson, D. J., Harnack, L., & Laska, M. N. (2019).
Variation in the food environment of small and non-traditional stores across racial segregation
and corporate status. Public Health Nutrition, 22(9), 1624-1634.
Wolch, J. (2005). From Global to Local: The Rise of Homelessness in Los Angeles during the
1980s. in A.J. Scott, and E.W. Soja (Eds.) The City: Los Angeles and Urban Theory at the
End of the Twentieth Century (pp. 200–210). Routledge: New York, NY.
Wood, D., & Krygier, J. (2009). Critical cartography. In R. Kitchin & N. Thrift, N. (Eds.), The
International Encyclopedia of Human Geography (pp. 340–344). Elsevier: New York, NY.
Wright, D. J., Goodchild, M. F., & Proctor, J. D. (1997). GIS: Tool or science? Demystifying the
persistent ambiguity of GIS as" tool" versus “science”. Annals of the Association of American
Geographers, 87(2), 346-362.
Wright, E., Haskett, M. E., & Anderson, J. (2020). When your students are hungry and homeless:
The crucial role of faculty. Communication Education, 69(2), 260-267.
Wrigley, N. (2002). 'Food deserts' in British cities: policy context and research priorities. Urban
Studies, 39(11), 2029-2040.
Yang, P., Xiong, N., & Ren, J. (2020). Data security and privacy protection for cloud storage: A
survey. IEEE Access, 8, 131723-131740.
Zavattaro, S. (2019). Using feminist geography to understand feelings of safety and
neighborhood image. Journal of Health and Human Services Administration, 42(2), 167-205.
Zhang, M., & Ghosh, D. (2016). Spatial supermarket redlining and neighborhood vulnerability:
A case study of Hartford, Connecticut. Transactions in GIS, 20(1), 79–100.
Zillow.com (2024). “All rental listings”. Retrieved from: https://www.zillow.com/homes/for_rent
(Accessed 3/4/2024).
Zohrabi, M. (2013). Mixed method research: Instruments, validity, reliability and reporting
findings. Theory and Practice in Language Studies, 3(2), 254.

Abstract (if available)

Abstract Where scientific disputes become heated, generating wider social interest, knowledge controversies emerge. This dissertation argues that these cannot be solved through the single-perspective knowledge politics dominant in conventional technoscience. This is especially the case in the context of “wicked” problems, which are unique, persistent, insoluble, and characterized by contradictory certitudes. The challenge for social research is to deliver meaningful solutions which rekindle public confidence via different forms of engagement. “Clumsy” solutions build on emergent and creative dialogue amongst an extended peer community to incorporate many relevant worldviews, providing the kinds of noisy and discordant evidence for decision-making which represents the multiple situated sites and forms of expertise extant in the social world. The three empirical chapters each propose their own clumsy solution to a specific context within health and place. Chapter 2 uses participatory geographic information systems (PGIS) to challenge the capitalist production of food systems. Chapter 3 uses public participatory GIS (PPGIS) to explore mobility practices in the dispossessive infrastructures created by urban renewal. Chapter 4 employs volunteered geographic information (VGI) to countermap housing insecurity, adopting the manners of representation of the political elite to resist their historical use for enclosure and conquest. Throughout, the role of expertise is examined for its capacity to generate new objects and sites of disagreement which may lead to new political subjectivities and collectives. Overall, participatory spatial science is commended as a powerful way to attach meaning to spatial relations, as long as it is preceded by sophisticated methodological thought, from both the participatory and the GIS sides of the elision. The findings suggest that clumsy solutions are a viable response to wicked problems of knowledge, and that more theoretical and empirical experimentation is needed to reconfigure expertise in health and social research.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Incomplete data & insufficient methods: transgender population health research in the US

PDF

Implementation of peer providers in integrated health care settings

PDF

The role of dyadic and triadic factors on psychosocial wellbeing and healthcare interactions among childhood cancer survivors, parents, and medical providers

PDF

Characterizing self-reported spatial and provider barriers to maternal health care utilization in Malawi

PDF

Examining the built and online environments to understand cannabis and e-cigarette availability, marketing, and product use in Southern California

PDF

Smoke-free housing policies and secondhand smoke exposure in low income multiunit housing in Los Angeles County

PDF

Energy drink consumption, substance use and attention-deficit/hyperactivity disorder among adolescents

PDF

Examining tobacco regulation opinions and policy acceptance among key opinion leaders and tobacco retailers in low socioeconomic status African American, Hispanic, and non-Hispanic White communities

PDF

Addressing unmet needs and harnessing social support to improve diabetes self-care among low-income, urban emergency department patients with diabetes

PDF

Smoking in pregnancy: from effects to solutions

PDF

Integrating spatial visualization to improve public health understanding and communication

PDF

Transforming unstructured historical and geographic data into spatio-temporal knowledge graphs

PDF

Homeless encampments and access to Water, Sanitation, and Hygiene (WaSH) services in Los Angeles, CA

PDF

Anxiety symptoms and nicotine use among adolescents and young adults

PDF

The effect of an educational fotonovela on the prevention of secondhand smoke

PDF

Defining neighborhood for health research in Arizona

PDF

Health care utilization and spending of the U.S. aging population

PDF

Intimate partner violence in the Black community: empowering Black families through education and engagement to increase survivor safety via a strengthened social support network

PDF

Climate change, air pollution and health: how policy can induce adaptation

PDF

Vulnerable station: perceptions of suicide in America's Last Best Place

Asset Metadata

Creator Lerner, Leo (author)

Core Title To what extent can wicked problems of health and place be addressed through clumsy solutions? Three experiments in reconstituting networks of spatial knowledge production

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Health and Place,Population

Degree Conferral Date 2024-05

Publication Date 05/21/2024

Defense Date 04/30/2024

Publisher Los Angeles, California (original), University of Southern California (original), University of Southern California. Libraries (digital)

Tag actor-network theory,clumsy solutions,GIS,OAI-PMH Harvest,public participatory geographic information science,volunteered geographic information,wicked problems

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor (Baezconde-Garbanati, Lourdes), Wilson, John P. (committee chair), De Lara, Juan (committee member), Rivera-Colón, Edgar (committee member), Vos, Robert (committee member)

Creator Email leolerne@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113950052

Unique identifier UC113950052

Identifier etd-LernerLeo-12966.pdf (filename)

Legacy Identifier etd-LernerLeo-12966

Document Type Dissertation

Format theses (aat)

Rights Lerner, Leo

Internet Media Type application/pdf

Type texts

Source 20240521-usctheses-batch-1156 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu