University of Southern California Dissertations and Theses

In/visible constituents: the representation of undocumented immigrants

(USC Thesis Other)

In/visible constituents: the representation of undocumented immigrants

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content IN/VISIBLE CONSTITUENTS: THE REPRESENTATION OF UNDOCUMENTED IMMIGRANTS by Matthew Synclair Mendez A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirement for the Degree DOCTOR OF PHILOSOPHY POLITICAL SCIENCE AND INTERNATIONAL RELATIONS August 2015 i DEDICATION To Juanita Alvarado, my mother, for nurturing my intellectual curiosity and for teaching me the importance of kindness. Your love has guided me through the tough times. To Elaine Pettis, my grandmother, thank you for being the rock of our family. You have taught me the importance of hard work and living up to my potential. To Robert Garcia, my partner, I cherish you and the support you have given me throughout this journey. I am in awe of you. To Veasey Tucker, my late great-grandmother, and Abigail Mendez, my late grandmother, for devoting your lives to your families. I miss you both and will love you always. ii Acknowledgements All journeys must come to an end and I would like to take this opportunity to thank the many people who have helped me get here today. I must start by thanking my excellent committee for their enthusiastic support for this project-Ange-Marie Hancock, Christian Grose, and Veronica Terriquez. I was fortunate enough to find three people who were committed to this project as much as I was and helped mentor me along the way. To my chair, advisor, and mentor, Ange-Marie Hancock, I owe you an immense debt of gratitude for your unwavering support of me throughout my entire time at USC. You are an exemplary role model in how to not only study transformative politics, but also how to live it. You taught me how to be a scholar and an educator and I aspire to one day touch lives the way you have touched mine. I still remember when you told me you would help me find a way to use the tools of political science to answer the questions I wanted to ask, and I am so grateful that you did just that. To Christian Grose, thank you for also being a mentor to me and for showing me exactly what political science has to offer. To Veronia Terriquez, thank you for going above and beyond as my outside member;. There are many others I must thank for their kindness. I must thanks Drs. Jefferey M. Sellers, Carol Wise, Diana O’Brien, and Sarah Gualtieri for their support. A special thanks goes out to Cathy Ballard for helping to make life easier not only for myself but all the other graduate students. Whenever I needed help, Cathy was always there to listen and help find solutions. I would have been lost without her. To Veridiana Chavarin, Aurora Ramirez, and Jody Battles, thank you for your help in the office and for making the Political Science office such a friendly place. Graduate school can be a very lonely place, but I was fortunate to find a community among the graduate students at USC that always gave me support and encouragement. Thank you Eunice Kang Kim, Kym MacNeal, Jordan Peterson, Elsa Menouno, Thora Giallouri, Sarah iii Sadhwani, Tyler Curley, Mark Paradis, Keira Stearns, Juvenal Cortez, Scott Wilbur, Denise Gonzalez, Tom Jamieson, Simon Radford, Denise Diaz Payan, Christian Paiz, and Stefan A. Alexander, Jiun Bang, and In Young Min. To my dear friend Laura Yen, I could not have done this without you. You were there for the hard times and you have been a constant source of affirmation and aid since. Thank you for all the love you have given me over the past few years. I would also like to thank Joshua Reed, Dennis Dean, David Allen, Cindy Allen, Isaac Romero, Devin Cotter, and Sharon Weissman for their technical expertise. To my interviewees, thank you for taking time the time to sit with me and share your experiences. Your insights helped me to write the dissertation I always intended to write and taught me the importance of activism. I must thank my wonderful family. To my mother, Juanita, and my grandmother, Elaine Pettis, thank you for loving me throughout the years. I could not have done this without both of you sacrificing so much for me. To my brothers, Paul and George Mendez, both of you were instrumental in helping me to get my BA degree and I am so grateful to both of you. To my step- grandfather, Ray Pettis, thank you for accepting us as your family and loving us as if we were your own. To my grandfather, Joseph Alvarado, thank you for supporting me throughout the years. To my nephew, Christopher Mendez, thank you for helping show me what really matters in life. To my Aunt Adriann, thank you for your love and for teaching me the importance of keeping our heritage alive. Lastly, I must thank my partner, Robert Garcia. I could not have done this without you. You have been my rock throughout this five-year journey and no matter how hard it difficult it became I could always count on you to be there for me. I feel like we did this together and I am iv so grateful to have you in my life. Thank you for believing in me and teaching me to believe in myself. v TABLE OF CONTENTS Dedication…………………………………………………………………………………………i Acknowledgements……………………………………………………………………………....ii Table of Contents………………………………………………………………………………...v List of Tables……………………………………………………………………………………vii Abstract………………………………………………………………………………………...viii Chapter 1: Introduction..…………………………………………………………..……………1 1. Introduction………………………………………………………………………………1 2. Toward a Study of the Representation of Undocumented Latinos……..…………….3 3. Outline of Empirical Chapters…..……………………………………………………...6 4. Conclusion………………………………………………………………………………10 Chapter 2: Descriptive Representation and Historically Disadvantaged Groups.................11 1. Introduction……………………………………………………………………………..11 2. Concepts of Representation……………………………………………………………12 3. Legislative Representation……………………………………………………………..15 4. Descriptive Representation and People of Color……………………………………..18 5. Intrinsic vs. Extrinsic Motivations…………………………………………………….23 6. Descriptive Representation and its Limits…………………………………………….26 7. Theorizing the Representation of Undocumented Immigrants……………………...34 8. The Issue of Accountability…………………………………………………………….36 9. Determinants of Representation for Undocumented Immigrants…………………...38 10. Conclusion…………………………………………………………………………...….45 Chapter 3: Who Represents Undocumented Latinos? A Field Experiment of State Legislators……………………………………………………………………………………….47 1. Introduction……………………………………………………………………………..47 2. Service Responsiveness as Representation……………………………………………48 3. Why Field Experiments?.................................................................................................50 4. Representation and the Undocumented Immigrant………………………………….51 5. Operationalization and Identity Influence……………………………………………53 6. Research Design: A Field Experiment of U.S. State Legislators…………………….54 7. Deception and Ethical Considerations………………………………………………...58 8. Methods………………………………………………………………………………….59 9. Results…………………………………………………………………………………...60 10. The Conditional Effect of Legislator Ethnicity……………………………………….62 11. Discussion and Implications…………………………………………………………....67 12. Conclusion………………………………………………………………………………71 Chapter 4: Disentangling Ethnicity and National Status…………………………………….73 1. Introduction……………………………………………………………………………..73 2. Building on the Previous Experiment…………………………………………………73 3. Rationale for An Eastern European Alias…………………………………………….75 4. Hypothesis……………………………………………………………………………….76 5. Research Design………………………………………………………………………...76 6. Results…………………………………………………………………………………...79 7. Discussion……………………………………………………………………………….82 8. The Conditional Effect of Legislator Ethnicity……………………………………….85 vi 9. Results: Do Latino Legislators Respond Differently?………………………………..85 10. Discussion……………………………………………………………………………….89 11. Conclusion………………………………………………………………………………92 Chapter 5: A Case Study of Immigrant Rights Advocacy in California……………………93 1. Introduction………………………………………………………………......................93 2. Representation as a Deliberative Process……………………………………………..95 3. Case Study: California…………………………………………………………………98 4. Research Design: Grounded Theory Method and an Interpretive Framework…………………………………………………………………………….100 5. Paradigm Intersectionality Approach………………………………………………..102 6. Advocacy on Behalf of Undocumented Students……………………………………104 6.1 Inegalitarian Traditions Meet in the Wake of the Civil Rights Movement…………104 6.2 Experiences with Descriptive Representation in the 1980s and Early 1990s……………………………………………………………………………………….....110 6.3 Post-AB 540 and Activism within the University………………………………….……113 7. Immigrant Advocacy in Politically Competitive Regions…………………………..118 8. Grassroots Organizations and Gaining Access to Legislators……………………...130 9. Discussion and Implications…………………………………………………………..138 10. Conclusion……………………………………………………………………………. 142 Appendix A: Letters.....…...…………………………………………………………………..143 Appendix B1: Robustness Checks for Field Experiments on National Status…………….144 Appendix B2: Robustness Checks with State Dummy Variables…………………………..144 Appendix B3: Robustness Checks: Regression Models with Interactive and Conditional Variables…………………………………………………………………...…….145 Appendix C1: Balance Check Across Co-Variates for Experiment Disentangling Ethnicity and National Status………………………………………………………………..146 Appendix C2: Robustness Checks……………………………………………………………146 Appendix C3:Robustness Checks with State Dummy Variables…………………………..147 Appendix C4: Robustness Checks: Regression Models with Interactive/Conditional Models Based on Latino Ethnicity…………………………………………………………...148 Appendix D: Questions for the Semi-Structured Interview………………………………..149 References……………………………………………………………………………………...151 vii List of Tables Table 3.1: Constituent Treatments…………………………………………………………….55 Table 3.2: Balance Checks Across Co-Variates………………………………………………58 Table 3.3: Legislator Responsiveness to Undocumented & Citizen Constituents………….61 Table 3.4: Non-Latino Legislators are Less Likely to Respond to Undocumented Constituents……………………………………………………………………………………..62 Table 3.5: Latino Legislators are as Likely to Respond to Undocumented Constituent as Other Constituents……………………………………………………………………..........64 Table 3.6: Difference-in-Difference between Response Rates to the Undocumented vs. Citizen Constituent for Latino Legislators and non-Latino Legislators……………………65 Table 3.7: Robustness Checks: Regression Models with Interactive/Conditional Variables………………………………………………………………………………...............66 Table 3.8: Robustness Checks: Regression Models with Interactive/Conditional Variables………………………………………………………………………………………...67 Table 4.1: Constituent Treatments…………………………………………………………….77 Table 4.2: Legislator Responsiveness to Constituent Emails as Compared to the Undocumented Latina…………..……………………………………………………………...80 Table 4.3: Legislator Responsiveness to Undocumented and Citizen Emails as Compared to the Eastern European Undocumented Immigrant………………………...82 Table 4.4: Latino legislator Responsiveness to the Undocumented and Citizen Emails as Compared to the Undocumented Latina…………………………….85 Table 4.5: Non-Latino Legislators Responsiveness to Undocumented and Citizen Constituents as Compared to the Undocumented Latina……………………………………87 Table 4.6: Non-Latino Legislators Responsiveness to Undocumented and Citizen Constituents as Compared to the Undocumented Eastern European………………………88 Table 5: A Selection of Pro-Immigrant Bills Passed Since 2000 in California……...............99 viii ABSTRACT Traditional mechanisms of legislative accountability have largely stressed the importance of sanction-based models of representation. Such models are largely premised on the assumption that legislators are responsive to constituents in order to ensure their reelection. This dissertation asks the question of what kinds of representation are necessary for the representation of undocumented immigrants, a group that is structurally barred from the franchise. An overwhelming majority of undocumented immigrants are Latinos and over the last decade they have mobilized forcefully to advocate for their rights. As such, this dissertation specifically asks what mechanism of accountability serves as the link between descriptive representation and substantive representation in regards to undocumented Latino immigrants. The dissertation substantively engages with the question of how to represent individuals with multiple identities and subsequently looks to intersectionality theory for possible answers. This dissertation utilizes a multi-method approach to investigate these research questions. In Chapter 2, I review the literature on representation and offer a theory on how the insights of intersectionality theory can help scholars understand the representation of undocumented immigrants. Field experiments are utilized in Chapters 3 and 4 to assess whether state legislators provide representation for undocumented immigrants. My results show that legislators do provide some representation for undocumented immigrants, but are generally more responsive to citizens. In Chapter 3, specifically, there is evidence to suggest that Latino legislators do provide equitable representation for undocumented Latino immigrants. Chapter 5 utilizes semi- structured, in-depth interviews with immigrant rights advocates to illustrate how these advocates built relationships with state legislators to act affirmatively on behalf of undocumented ix immigrants. The results show that descriptive representation along the lines of ethnicity is a possible route to representation, but that it can be mediated by electoral factors. 1 Chapter 1: Introduction 1. Introduction Undocumented Latinos represent nearly 3% of the total U.S. Population and account for 17% of the Latino population (Lilley 2012). Yet, for decades they have found themselves the targets of restrictive laws and public policy initiatives. That said, their story should not be misconstrued as one of victimhood, as there have been many instances in which undocumented Latinos have engaged the political system to advocate for themselves and their rights. They have created social movements; formed organizations at the local, state, and federal level; built coalitions across racial and sectarian lines to build awareness for the plight of the undocumented; and lobbied elected officials to legislate on their behalf. That last activity has been especially understudied, but presents important questions about the political incorporation of undocumented Latino immigrants. The primary research question of this dissertation is, what kinds of representation are needed for undocumented Latino immigrants? Other important questions include the following: Do legislators discern between identities, or do they view the undocumented Latino as an intersectional identity made up of national status and ethnic identity? What role do immigrant rights advocates play in ensuring that legislators listen to undocumented Latinos? This dissertation aims to provide answers to those questions by focusing on the relationship between undocumented Latino immigrants and those who could and do represent them in the political sphere. Specifically, this dissertation primarily examines the representation of undocumented Latinos by state legislators and seeks to answer what kind of representation is needed for this group. Constituents have traditionally been theorized to hold legislators accountable through the threat of voting them out of office. Undocumented immigrants do not 2 have this ability and instead this dissertation will offer possible avenues for representation. The term undocumented immigrant 1 refers to people who are in the United States without formal authorization and who, at the moment, currently have very limited avenues to attain legal status, yet are able to exercise some rights under U.S. law (Bosniak 2006). While the undocumented immigrant has largely been constructed as a Latino body (Chavez 2008; Johnson 1996; Ngai 2003) in popular consciousness, this dissertation will not conflate membership in the pan-ethnic Latino identity with the identity of the undocumented immigrant. Instead, this dissertation seeks to address the privileges associated with citizenship as well as the disadvantages associated with the undocumented status. This approach to the category of “national status” will feature prominently in this dissertation as it not only focuses on Latino political insiders, such as elected officials and those who work for advocacy organizations, but also on how multiple identities impacts and structures the relationship between undocumented Latinos and the political sphere. This dissertation will bring the representation literature into dialogue with intersectionality as it studies a population with two distinct, but deeply intertwined identities. Since its scholarly articulation, intersectionality (e.g. Crenshaw 1989, 1991; Collins 1990) has become a highly utilized theory to analyze the constitutive nature of socially constructed multiple identities and their political consequences. Intersectionality explicitly calls for an exploration of how the intersection of disadvantaged identities situates certain social groups in a sociopolitical location that then shapes their experiences. Crenshaw's original articulation was concerned with how race and gender together, with their corresponding histories of oppression, 1 Undocumented immigrants are not a monolithic group, but a heterogeneous population that includes people from different ethnic, racial, and national groups. The conditions by which people have acquired this people from different ethnic, racial, and national groups. The conditions by which people have acquired this status also varies greatly by circumstance. 3 work to marginalize Black women in discussions regarding anti-discrimination policy and domestic violence. The focus on race and gender, instead of the standard question of race or gender, has opened the door for others to ask how public policy and the political discourse silence those who straddle multiple oppressions (e.g. Hancock 2004). Various scholars have recently begun to call for an expansion of the categories that are normally interrogated by studies of intersectionality, including age, legal status, and disability among others (e.g. Romero 2008; Hancock 2009, 2011, 2012; Dhamoon 2010). For this dissertation, an acknowledgement of intersectionality seems more than appropriate when applied to a large, heterogeneous pan-ethnic group like Latinos. The steady stream of immigration from Latin America, an increasingly repressive immigration regime, and the continuing reality of structural racism has given scholars many points of entry in understanding the barriers and opportunities for Latinos in the political sphere. Undocumented Latinos have two deeply intertwined identities that leaves them without formal access to representation, but also allows for the potential of representation. 2. Toward a Study of the Representation of Undocumented Latinos What is the state of representation for undocumented Latinos and what are the implications for greater political science? Not much is known about the representation of undocumented immigrants in general, or of undocumented Latinos specifically. President Obama’s recent executive action granting relief from deportation to certain segments of the undocumented population 2 and movement at the state level to establish in-state tuition for undocumented youth might lead some to believe that undocumented immigrants are overrepresented relative to their position. Whether they are overrepresented in terms of policy is 2 These executive orders established the Deferred Action for Childhood Arrivals (DACA) and the Deferred Action for Parents of American or Lawful Permanent Residents programs. 4 unknown at this time as no study has compared the subjective interests of undocumented immigrants with the voting records, or ideologies of their legislators. This dissertation cannot answer that question, but it does provide evidence that undocumented immigrants receive less representation in regards to constituency service than citizens. The qualitative evidence from Chapter 5 helps detail the efforts, and struggles, of activists to get legislators to act affirmatively on behalf of undocumented immigrants. The primary goal of this dissertation is to help unpack the relationship that exists between the Latino undocumented immigrant and those who are in positions to represent them. The secondary goal is to specifically explore the ways that immigrant rights advocates represent the needs of undocumented Latinos to state legislators. Chapter 2 will detail the literature on representation and identify the mechanisms that explain the behavior of legislators. It will also argue that descriptive representation that occurs upon one single axis is not enough to represent groups that experience advanced marginalization as a result of being at the intersection of two disadvantaged identities (e.g., Cohen 1999). I use Mansbridge’s (2003) formulations of representation to help me argue what type of representation would be necessary for such groups. I then argue that intersectional linked fate on the part of the descriptive representative is necessary for her to act affirmatively on behalf of the intersectionally marginalized constituent who shares at least one aspect of the identity. Chapters 3 and 4 of the dissertation will consist of questions regarding the identity, or identities of, the legislator and that of the undocumented immigrant and how the salience of particular identities determines the level of representation given to the undocumented immigrant through constituency service. Chapter 5 will examine the political dynamics that shape the representation of Latino undocumented immigrants by immigrant rights advocates, and how 5 these advocates build relationships with state legislators. Both parts are united by a broad concept of representation and power that both speak to formal political institutions (e.g. Pitkin 1967), and social movement politics. While putting these together might not seem intuitive, this dissertation argues that the political representation of undocumented immigrants requires an understanding of not only the formal political apparatus and whether its actors take into account the demands and needs of the undocumented Latinos, but also on the people who have been selected at the community level to bring forth the political needs of this group of people. While this dissertation does not, and cannot, demonstrate a direct relationship between political institutions, immigrant rights advocates, and representational outcomes, it is my intention to imply that social movements, as symbolized by immigrant rights advocates, are integral to helping voice the demands of undocumented Latinos to elected officials. Both sections will utilize different methods that are best suited to identifying the conditions under which these different groups of “representatives” could and do represent undocumented Latinos within their respective contexts. Quantitative methods are most suitable for legislators as political science has identified areas of substantive representation they are expected to fulfill, such as constituency work, bill sponsorship, and policy advocacy (e.g. Poole and Rosenthal 1997; Grose 2011). As for the immigrant rights advocates, qualitative methods will be the method of choice as this dissertation seeks to illuminate the representational process that occurs between state legislators and the advocates. The specific qualitative method used for this dissertation will be semi-structured, in-depth interviews with a purposive sample of advocates. 6 3. Outline of Empirical Chapters This dissertation will utilize different approaches to the study of multiple identities. Field experiments will be utilized to help gauge the relationship between the undocumented immigrant and the state legislator. Field experiments are a novel approach that help to identify the private motivations, whether strategic or personal, of legislators. Previous studies have shown that state legislators hold racial biases (Butler and Broockman 2011; Broockman 2014) against African American constituents even when party and electoral considerations are taken into consideration. These particular field experiments are illustrative in that they have allowed the researchers to vary the rewards and costs for the state legislators (e.g., Grose 2014; Bergen 2009; Butler and Nickerson 2011). Experiments allow researchers to assess causality in a more direct fashion as it gives the researcher more control over what treatment is applied to the subjects (e.g., de Rooij, Green, and Gerber 2009; Morton and Williams 2010). Since the direction of causality is easier to assess, at least compared to observational studies, control variables becomes less important. In the case of legislators, this research design can help reveal the relationship between them and their constituents by directly testing different messages with fictional aliases (e.g. Butler and Broockman 2012). The experiments utilized in this dissertation will seek to test the legislator's relationship to her constituents based on the multiple identities of the legislator and the constituent. The legislator is largely seen as a rational actor (Mayhew 1974) interested first and foremost in reelection. The duty of the representative is to “stand in for” the constituency (Pitkin 1967) and adopt behaviors that are in accordance with the best interests of their constituents and her own reelection. The legislator is also a person, so they are bound to also have personal preferences 7 based on life experiences (Burden 2007) and their ascriptive characteristics (e.g. Whitby 1998). Membership in groups based on race, ethnicity, and gender have produced numerous studies identifying that these identities do influence the behavior of elected officials (e.g. Swers 2002; Bratton 2005; Tate 2003; Grose 2011; Rouse 2013). The two experiments that will be run for this dissertation dovetail with the two main approaches to empirical intersectionality found in the social sciences(Hancock 2013). The “intersectionality-as-testable explanation” approach involves a first order question that presupposes that one aspect of identity has a causal effect on a certain outcome. 3 This particular strategy pays mere lip service to intersectionality, and as a result, scholars have come up with a multiple approach that includes multiple identity variables and an interaction effect related to two identities. The first experiment follows this approach, and will be explained in more detail below, as it tests to see whether Latino legislators are more likely to respond to an undocumented Latina requesting help with finding financial aid for college than non-Latino legislators. The rather small sample size of Latino legislators means this group cannot be split more than three ways, meaning that only the identity associated with national status can be examined. The first experiment in this dissertation seeks to build upon the theory of intersectionally linked fate (Strolovitch 2007)which calls for advantaged members of a particular group to feel as if their fate is linked to that of the more disadvantaged members of their group. There is a direct link between responsibility for the disadvantaged subgroup on the part of the advantaged subgroup, and the expectation that those whose represent the group in an official capacity should represent the disadvantage subgroup. This idea carries special resonance for those who are tasked with representing groups that have experienced marginalization, or occupy a subordinate position 3 Hancock (2013) refers to this as a Unitary Quantitative Formulation, and specifies a qualitative formulation of this model as well as addresses Multiple Quantitative Approach, but for brevity's sake, I will not delve too deeply into the latter. 8 in U.S. society (ibid, 63). This theory of intersectionally linked fate and the experimental method offer a novel approach to testing how multiple identities can affect the consciousness of legislators. As formulated by Strolovich, intersectionally linked fate might take into account the intersectional identities of the constituent populations serviced by the interest group, but it operates on a single identity dimension for the interest group, its officers, and theoretically, other agents as well. 4 Since the Latino legislator is the one acting for the larger group, these two experiments will theorize a specific theory of intersectionally linked fate that operates on a single-axis of identity. The experiment in Chapter 3 will operationalize an “identity salience model of representation” that uses an undocumented Latina alias to test whether Latino representatives will respond to them at the same rate as a Latina citizen. I further theorize that intersectional linked fate operates specifically for Latino elected representatives within a larger theory of intersectional linked fate. This form of representation remarks upon the diverse nature of the Latino pan-ethnicity while acknowledging that there is a desire on the part of Latino elected officials to represent this large and diverse group. Latino legislators have explained their legislative styles as being adaptable to the concerns of certain Latino constituencies (e.g., Casellas 2011; Geron 2005). Regardless of whether it is referred to as an identity salience model of representation or a unitary identity formulation of the intersectionality-as-testable-explanation approach, it allows for an investigation into whether Latino representatives experience a sense of linked fate with co-ethnics who are undocumented. Chapter 4 will build upon Chapter 3 by explicitly incorporating an analysis of ethnicity into the model. An Eastern European alias is incorporated into the experiment in order to test 4 Strolovich explains how and why social justice organizations frame linked fate this way as it relates to creating issue frames, and “justify claims and issues in particular ways” (59). 9 whether ethnicity and national status, together or separately, determine the responsiveness of legislators to undocumented immigrants. Latino legislators are also evaluated as to whether they will respond more to an undocumented immigrant as a co-ethnic than one who does not share the same ethnicity. It too is an example of the intersectionality-as-testable-explanation approach but the larger sample of legislators allows for actual consideration of categorical multiplicity, as both ethnicity and national status can be examined together. As stated previously, Chapter 5 of the dissertation will utilize qualitative methods to ascertain how immigrant rights advocates build relationships with state legislators. The primary method used here will be in-depth, semi-structured interviews with immigrant rights advocates who have experience with lobbying state legislators. This chapter uses grounded theory and the Paradigm Intersectionality Approach (Hancock 2011, 2013) to understand how the advocates were able (or in some cases unable) to build relationships with state legislators for a population that sat at the intersection of oppressions related to race and national status. The advocates also reveal potential motivations for why legislators do (or do not) build relationships with them. Since the focus is on the experience of undocumented immigrants, this study will employ the approach of purposive sampling which “bases inclusion in the sample on the shared common experience or shared group status” (Cuadraz and Uttal 1999). This will not be a comparative study, but will instead seek to understand the lobbying experiences of immigrant rights activists through the respondent’s own subjective lens. The purposive sample of 20 immigrant rights advocates features mostly Latina participants, but also includes Latino males, and Whites and Asian Americans of both genders. It is a localized look at lobbying in the state of California. 10 4. Conclusion This dissertation can be summed up as substantively being focused on the political representation of undocumented Latino immigrants; theoretically concerned with the role of multiple identities and intersectionality in questions of representation; and how experimental and qualitative methods can be used to analyze the relationship between undocumented Latinas and the people who could, and do, represent them. It seeks to combine elements from the fields of legislative politics, intersectionality theory, field experiments, and Latino politics, in order to situate undocumented Latinos in the political sphere of the United States. While it will not provide an exhaustive account of all the facets of representation, it will hopefully shed light on how constituency service, probably the most interactive and democratic aspect of a legislator’s careers can be used to aid social justice. 11 Chapter 2: Descriptive Representation and Historically Disadvantaged Groups 1. Introduction The political representation of historically disadvantaged racial and ethnic groups has been one of the defining themes of the American political science literature over the past three decades. Both normative theorists and empirical political scientists have sought to unlock the puzzle of how best to increase not only the presence of disadvantaged groups within formal political institutions but also enhance their representation through substantive means and outcomes. While some scholars have focused on the structural design of electoral systems (i.e., Young 1990; Guinier 1994), others have focused on changing the electoral incentives to make legislators more responsive to the needs of racial and ethnic minorities (Canon 1999; Grose 2011). Still others focus on the nature of the personal preferences of the legislator themselves, whether it is based on individual characteristics (i.e., Burden 2007); preexisting knowledge (i.e., Butler 2014); or personal insight gained through membership in a historically underrepresented group (Swers and Rouse 2011; Broockman 2013). The following chapter will explore the last two sets of literature as they speak to the current debate about whether changed electoral incentives, or the personal knowledge gained through membership in a historically underrepresented racial or ethnic group is the best way to achieve representation of these groups. The current literature largely operates under the assumption that racial and ethnic groups have monolithic policy interests, and therefore ignores the fact that within these groups exist disadvantaged subgroups with needs that are unique from the rest of the group (i.e., Cohen 1999). This chapter will also address the contemporary dilemma of how to understand the role that diversity within these groups plays in representation. This particular issue will be addressed with the analytical tools of intersectionality theory, which 12 originated from one instances of disciplinary intersections: Legal Studies and Sociology (Crenshaw 1989, 1991). 5 Intersectionality theory addresses the fact that people are made up of constitutive ascriptive identities that are associated with perceptions of power relations that help drive their sociopolitical standing. (Crenshaw 1991; Hancock 2011). This particular chapter will accomplish the previously defined goal of examining the broader literature on the formal political representation of historically underrepresented minorities, and then move to examine how recent innovations in our understanding of descriptive representation can help us evaluate the representation of undocumented immigrants. The following chapters will empirically investigate whether descriptive representation is sufficient to provide representation of undocumented Latino immigrants. Chapters 3 and 4 both use field experiments on constituency service to assess the motivations of Latinos legislators in representing this particular group. Chapter 3 uses a Latina alias and treatments that vary the national status of the that alias to assess whether state legislators in general will respond less to a constituent request from an undocumented immigrant than a citizen. Chapter 4 incorporates a second constituent alias, that of an Eastern European, in an effort to disentangle the effects of national status and ethnicity. Chapter 5 features an interpretive analysis of the experiences of immigrant rights advocates who lobby, and in some cases, work with state legislators in California. 2. Concepts of Representation Most empirical U.S. politics scholars view political representation through the conceptual classifications theorized by Hanna Pitkin (1972): formal representation, substantive representation, descriptive representation, and symbolic representation. In her treatment of 5 This refers to the naming of the term in academia. The basic tenets of intersectionality theory have been present in U.S. Black women's political thought for over a century (i.e., Maria Stewart [1987], Sojourner Truth 1851; Julia Ann Cooper 1892). 13 representation, formal representation is concerned with representatives within political institutions, and deals with authorization and accountability; substantive representation is concerned with the way the representative acts on her constituents’ preferences; descriptive representation is the way in which the representative “mirrors” her constituency; and symbolic representation, which is concerned with the meaning the representative has for her constituency. What ties these four types of representation together is the idea that someone has been elected to participate in government by some constituency. These four classifications have proven to not be mutually exclusive, and empirical scholars have used them to investigate the ways that legislators provide representation to their constituents. Eulau and Karps (1977) define the actions taken by representatives as responsiveness. Responsiveness takes four distinct forms: policy responsiveness, where the representative takes policy positions similar to that of her constituents; service responsiveness, where the representative provides particular benefits to an individual constituent, or a specific group of constituents; symbolic responsiveness, where the representative makes “public gestures of a sort that create a sense of trust and support” between her and her constituents; and allocation responsiveness, where the representative seeks to obtain projects, and funds, for her district (ibid, 241). Pitkin’s conceptualization of representation and Eulau and Karp’s specification of responsiveness have both influenced how political scientists evaluate representation. For example, Grose (2011), in his examination of racial representation, collapses policy responsiveness, allocation responsiveness, and service responsiveness under the banner of substantive representation. Some empirical scholars have argued that Eulau and Karp’s conceptualization of symbolic responsiveness can be linked to descriptive representation. Tate 14 (2003), for example, argues that the presence of descriptive representatives is symbolically important to African Americans because it “represents their inclusion in the polity…and their political power in the U.S. system (ibid, 15).” Political theorist Jane Mansbridge (2003) has argued for the consideration of three additional forms of representation. In addition to the traditional accountability model of promissory representation, a sanctions model of representation that threatens the representative with being voted out if she violates promises made in the authorizing election. Advancements made in the empirical literature on representation (e.g., Kingdon 1989; Arnold 1990) helped Mansbridge formulate anticipatory representation, which acknowledges the dynamic nature of both the representative and her constituency. This is not a sanctions model of representation in the traditional sense as the representative is thought to engage in educational activities to gauge changing opinions of the constituency, as well as potentially influence them, in order to be rewarded with reelection. The issue of accountability manifests itself quite simply in promissory representation as the sanction acts as punishment for failing to deliver on promises made to the constituency. It is a bit trickier for anticipatory representation, which does not as much punish the representation for delivering on promises, as it incentivizes her to look at ways to please the constituency in the future election. In anticipatory representation, the representative is not so much held accountable through the vote, but instead evaluated by the nature of deliberation that exists between her and her constituency (ibid, 520). The last two of Mansbridge’s new forms of representation also do no fit traditional forms of accountability. Gyroscopic representation describes the ability of a representative to provide meaningful representation for her constituency by looking within to internal values. The 15 electorate is able to hold the elected representative accountable not through the power of the sanction, but by initially selecting someone with the same values whose behavior in office will largely reflect the “constellation of values” held by her constituents (ibid, 521). The elector is able to trust the representative to act in her best interest and therefore the representative’s behavior cannot be explained by the threat of sanction. Gyroscopic representation is a selection- based model of representation that gives the elector greater influence over the larger polity by electing someone who will act for them regardless of external incentives. Surrogate representation is where a representative responds to the needs of those that live outside of her district. This form of representation presupposes that no electoral relationship exists between the representative and those whom she represents outside of her district. Accountability can only be maintained through financial contributions or support from broader interest groups. Mansbridge’s rethinking of representation leads to new questions of motivations for the behavior of elected representatives, as she argues that electoral incentives are not always necessary, or sufficient, explanations. As a result, she offers some new ways to think of causality in terms of representation and accountability, which will be discussed below. 3. Legislative Representation Miller and Stokes’ (1963) classic article sought to determine how constituency control influenced the opinions of Members of Congress (MCs). Their study, while methodologically flawed (Clausen 1977; Erickson 1978), found that MCs do find themselves responding to constituency pressures in some cases, but not in others. The policy congruence model is most traditionally aligned with the sanction based models of representation as it uses statistical measures to determine how far, or close, the representative’s policy preferences, determined either by elite surveys or roll call votes, is from the aggregate preferences of her constituents, as 16 determined by opinion surveys. The expectation from this model is that congruence between the representative and her constituency means that the former is exercising control over the latter as the representative responds to electoral pressures fueled by mass opinion (e.g., Downs 1957). Earlier studies sought to determine whether policy congruence merely existed, with some finding little to no congruence (Brand 1969; Jones 1973), and others offering a qualified yes (Fiorina 1974; Page, Shapiro, Gronke, and Rosenberg 1984). However, more recent studies find that such incongruence can have negative electoral outcomes in cases of ideological extremism, party loyalty, or controversial roll call votes (Canes, Wrone, Brady, and Cogan 2002; Ansolabehere and Jones 2010; Nyhan, McGhee, Sides, Masket, and Greene 2012). These latter examples, which were quantitative studies, find that voting shares for representatives decreased in these circumstances because of the perception that the representatives had moved away from the wishes of their constituency. These are examples of promissory representation as the voters signaled their disapproval of their representatives’ actions by voting them out of office. The focus on Congressional elections has primarily used a rational choice model to explain the behavior of legislators. David Mayhew (1974) offered a theory that Members of Congress were rational actors concerned first and foremost with reelection. The desire for reelection was offered as the sole cause for the behavior of MCs, and explained why they valued constituency service and allocation over policy representation. Mayhew’s rational choice theory adheres to the tenets of promissory representation because of their fear of losing reelection. MCs behave in ways to avoid being sanctioned by their constituencies. Fenno’s (1978) seminal book on representation also offered reelection as the primary explanation for the behavior of MCs, but only because it was the condition that must be met by MCs in order to accomplish their other goals. 17 The difference between Fenno and Mayhew is that the former also articulated other incentives for the legislator’s behavior, such as the desire to make good policy, and sought to theorize about Congressional behavior by engaging in participant-observation. Fenno referred to his research methods, as the “soak and poke” method, and involved accompanying MCs back to their districts to see how they represented themselves to their constituents. While the MCs articulated the importance of trust and mutual education, Fenno also finds that they engaged in strategic behavior, such as engaging different segments of their constituencies with different messages, and believing that every roll call vote is potential negative campaign attack to be used against them. Fenno is still largely concerned with sanction-based models of representation, after all the representative is always concerned about losing reelection, but engages in behaviors designed to build trust with the electorate and learn about what matters to his constituents. Fenno was not just concerned with promissory representation, although the MCs were certainly aware that their roll call votes could come to haunt them, but also identified the antecedents of anticipatory representation: mutual education and the potential to influence voters preferences. Arnold (1990) would develop a theory about the behavior of MCs that directly includes the potential preferences of voters, but still primarily views legislators as acting through the threat of sanctions. As Mansbridge (2003, 518) notes, Arnold acknowledges that voter preferences, and interests can change over time, but that their primary function is still to register approval, or disapproval, of their MC’s actions. Representation is still defined as the voting electorate exercising power over the representative by Arnold, instead of the representative working through deliberative means, both dyadic and systemic, to identify, and influence, the interests of voters (ibid, 520). 18 4. Descriptive Representation and People of Color The United States might have been founded on egalitarian ideals, but in practice has excluded many different groups from political power. The remedies to these discriminatory practices would require changing the U.S. Constitution, and adding new rights, including the 15 th Amendment, which “specifically prohibits denying the right to vote on the basis of ‘race, color, or previous condition of servitude” (Whitby 1997, 19), and the 19 th Amendment, which granted the franchise to women in 1920. The 15 th Amendment was passed in 1870 and was the last of the post-Civil War era amendments 6 by Congress designed to fully extend citizenship to African Americans following the Civil War. The post-Civil War Congress was also very active in protecting the voting rights of African Americans, as Section 2 of the 15 th Amendment granted the Congress the power to pass laws designed to enforce the amendment. These laws included the Enforcement Act of 1870, which “prohibited state election officials from applying local election laws in a racially biased manner” (Whitby 1997, 20). The Congress passed additional acts, like the Enforcement Act of 1871, which “placed federal elections in the south under direct federal supervision,” and the Ku Klux Act (1871) which authorized the use of federal troops in the South to protect Black voters from violent White terrorist groups (ibid, 20). Congress’ efforts to protect the Black franchise would be undone by the Supreme Court, which would interpret the 15 th Amendment in a manner favorable to the southern states. In the United States v. Reese (1876), the Supreme Court ruled that the 15 th Amendment only prohibited discrimination in voting based on race, color or previous servitude and did not guarantee the right to vote. This allows the southern states to begin to devise tactics to prevent Blacks from 6 The other amendments included the 13 th Amendment, which officially ended chattel slavery and the 14 th Amendment, which grants equal protection under the law for all citizens, among other things. 19 voting as long as they do “not blatantly violate the voting rights of Blacks” (Whitby 1997, 20). The Supreme Court would strike down the enforcement acts by requiring that federal prosecutors essentially prove racially discriminatory intent on the part of White perpetrators against Black voters (ibid, 20). The 15 th Amendment was rendered powerless leaving Black voters to the mercy of local registrars and state officials. Effectively barred the right to vote, African Americans would be subjected to many discriminatory laws, including poll taxes, literacy tests, and the so-called White primary, which only allowed White voters to participate in Democratic primaries (Kousser 1984). African Americans also faced violence at the hands of White vigilantes if they tried to vote, and this further drove participation rates down for both groups in the South. Racial gerrymandering also ensured that African Americans were always the minority within any Congressional district across the south (Parker 1990); however, as Tate (2003, 54) notes, this practice was also common in the North when Black voters constituted a majority in urban areas. These discriminatory laws and practices would make it impossible for African American to exercise any kind of accountability over their representatives. They were effectively without representation in the South as they were structurally barred from voting, so they could not sanction their representatives through the vote. Civil Rights organizations would effectively agitate against these structural barriers, including a successful challenge to the White primary in Smith v. Allright (1944) and passage of the 24 th Amendment, which abolished the poll tax in 1964 (Whitby 1997, 21). The biggest advancement for voting rights would come with the Voting Rights Act of 1965, which would specifically eliminate all barriers to voting in the South, including suspension of the literacy test, and put the region, as well as other states and municipalities with histories of racial discrimination under federal review (ibid, 24). 20 The Voting Rights Act of 1965 opened the door to increased political participation by racial and ethnic groups who had been excluded through both formal and informal barriers. The VRA would later ensure that historically underrepresented groups like African Americans and Latinos would have the opportunity to elect candidates of their own choice, and this would be done, at times controversially, by mandating majority-minority districts (Groffman, Handley, and Niemi 1992). 7 Black or Latino population within a district came to be seen as the mechanism by which Blacks and Latinos can hold their representatives accountable (Bullock 1981; Combs, Hibbing, and Welch 1984; Lublin 1997; Grose 2005). This new political reality led to questions about the functionality of descriptive representation. Originally theorized as a stand-alone form of representation, scholars began to ask the question of whether descriptive representation and substantive representation are linked. In this chapter, the question becomes, does descriptive representation lead to substantive representation for racial and ethnic minorities because of promissory representation, or are other mechanisms of representation involved? Swain (1995) argues that there is no link between descriptive representation and substantive representation, and posits that White Democrats can represent African Americans as well as Black Members of Congress because MCs are ultimately accountable to their electorate. Furthermore, she and Cameron, Epstein, and O’Halloran (1996) argue that drawing districts designed to elect African American MCs actually harms their chances of getting substantive legislation passed. Later research has challenged this view with convincing data using roll call votes and bill sponsorship, including Whitby (1997), Lublin (1997), and Canon (1999). These 7 In 1975, the VRA was amended to include protections for single-language minorities, including Spanish speakers, Asian Americans, Native Americans, and Alaskan natives. It also expanded the scope of states and territories that were now under federal review, including the Arizona, Texas, Alaska, and “jurisdictions in Colorado, South Dakota” and parts of Florida and North Carolina (Grofman, Handley, and Niemi 1992, 21). 21 scholars argue that African American Members of Congress provide a greater degree of policy responsiveness to African Americans than White members of either party. In an exhaustive study of the “captured” status of African American voters, Frymer (1999) argues convincingly that the policy preferences of African Americans remain apart, due to a number of reasons having to do with a history of slavery and legal discrimination, from the White majority, making it costly for either party to fully advocate for their needs (ibid, 8-9). This means that despite the loyalty of African American voters, party leaders find it in their best interest to marginalize them. Frymer also examines the Congressional Black Caucus, using them as proxies for Black interests 8 , as he describes them as “the most visible and vocal organization attempting to represent Black interests” (ibid, 149). He finds that the CBC’s agenda includes a mixed level of success, as they oftentimes were at odds with both the larger Democratic Party and the Republican Party. His interpretation of the data suggests that the race of the MC does matter in terms of providing substantive representation for African American voters. With strong evidence that the race of the Congress Member does impact policy responsiveness, scholars began to go beyond roll call votes. Tate (2003) finds that African American Members of Congress place more emphasis on symbolic representation and district casework, which resonates strongly with their co-ethnics. Tate argues that the race of the legislator does matter and while her work does not specifically address concerns with reelection, an argument can be made that she does not believe that such concerns wholly drive their decision-making process. Tate believes that African American legislators are different, but the exact mechanism remains unclear because her work is geared more towards the importance of African Americans in the legislative branch. As she puts it, “Black members of Congress defend 8 He extensively argues that despite the internal conflicts in the Congressional Black Caucus (Frymer, 1999, 148), and the more liberal attitudes of this group, “they offer as a good a proxy as we are likely to find” for African American interests (ibid, 149). 22 and act on a set of interests that most other members of Congress don’t take up” (ibid, 160) making it possible that she believes that African American MCs are acting as surrogate representatives for the national African American community. Grose (2011) presents a unified theory of African American representation that examines resource allocation, casework, and roll call votes on civil rights legislation, and finds that African American Members of Congress do provide better representation to African Americans. He uses district level data to run statistical analyses on the various areas of responsiveness. However, Grose argues that the behavior of Black MCs is driven by their concern for reelection. For him, African Americans emphasize different areas of substantive representation depending on the racial makeup of their district, and whether they need to appeal to White voters to win. He appears to endorse the view that African American MCs are using the traditional sanction-based models of accountability, such as promissory representation, and posits that Black MCs have the same motivations as other MCs. The story of Latino representation is often told through quantitative studies of roll call votes. The effect of Latino ethnicity is not as clear in these works as Democrats in general are found to provide policy representation for Latinos, with Latino population often serving as the most prominent factor (Welch and Hibbing 1984; Hero and Tolbert 1995; Santos and Huertas 2001). The key variable to these findings seem to be Latino population, which makes it possible that substantive representation for Latinos is driven by electoral concerns. Both Griffin and Newman (2007) and Wallace (2014) find that Latinos provide better representation for Latinos that non-Latinos, but also find that Latinos can still expect a high-degree of representation from Democrats. These findings possibly indicate that Democrats in general want to represent their Latino constituents in order to win their votes. 23 5. Intrinsic vs. Extrinsic Motivations The wealth of evidence produced on the subject of whether substantive and descriptive representation are linked begs the question as to why this occurs. As noted above, the political science literature on representation has largely privileged strategic concerns, such as reelection (i.e., Mayhew 1974; Fiorina 1974; Fenno 1978). The concern with reelection has been referred to as an extrinsic motivation, with the focus on a tangible reward (Broockman 2013); therefore, all actions of the representative, and her office, are geared toward winning reelection (Fiorina 1989; Jacobson 2004). The literature on descriptive representation and race and ethnicity are largely based on assumption that the size of the minority population ,as well with reelection concerns influencing the representation of Latinos (i.e., Griffin and Newman 2007), and the representational styles of African American MCs (i.e., Canon 1999; Grose 2003). Canon argues that a “politics of commonality” influences African American MCs to engage in a balancing act that sees them emphasizing issues that impact all of their constituents. He believes that creating majority- influence districts will lead to better representation for all as it incentivizes MCs to reach out to voters regardless of race. Grose (2011) largely reaches the same conclusions, especially with his results in the determining factors of how African American MCs substantively represent Black constituents. He finds that the percentage of the Black population is related to the level of policy responsiveness, allocation responsiveness, and constituency responsiveness provided, with Black MCs providing more policy responsiveness to their Black constituents in Black-majority districts, while Black MCs in white-majority districts provide more allocation responsiveness. He theorizes that this occurs because of electoral concerns in white-majority districts, such as the 24 need to appease White voters on roll call votes, while still providing robust representation to African American voters. An alternative to the theory that electoral concerns, and related variables, are the main explanation for the link between descriptive representation and substantive representation is that the group consciousness might explain the behavior of representatives of color. Whitby (1997) proposes that African Americans have a group consciousness that serves as an internal basis to act on issues that advance the group. His thoughts on group consciousness are built upon Dawson’s theory of linked fate (1994), which posits that individual African Americans believe that their interests are bounded up with that of their racial group due to a shared history of oppression. In this sense, it serves as form of gyroscopic representation as African American MCs look within to their group consciousness to help them make decisions about how to represent their group, but it also potentially serves as a motivation for reelection (Whitby, 1997, 85). Whitby’s analysis does not allow for him to draw conclusions about the actual effects of the MC’s race, and by extension, whether she is driven by something other than reelection concerns (ibid. 97). While there is qualitative evidence that Latino legislators also possess a group consciousness based around ethnicity (Geron 2005; Casellas 2011), this claim has never been specifically tested with anything other than measures of legislative choice or roll call votes (Bratton 2005; Knoll 2009). Broockman (2013, 3) shies away from specific discussions of linked fate based solely on race or ethnicity, and instead focuses on a legislator’s personal preference for helping those with shared personal characteristics. There are examples in the literature of legislators providing representation those for with shared personal characteristics, such as women (Bratton 2002, 2005; Thomas 1991; Carroll 1994; Reingold 2000; Swers 1998, 2002; 25 Thomas 1991); the working class (i.e., Carnes 2012, but see Butler 2014, 49-50); and even smokers (i.e., Burden 2007). Broockman (2013) instead posits that there are intrinsic motivations for legislators to represent certain groups. He defines intrinsic motivations as something that legislators do solely for the benefit of others that leads to a psychic benefit for the legislator (Broockman, 3). His experiment on constituency service and whether African American state legislators are more likely to respond to a request from a fellow co-ethnic who lives outside of the legislator’s district, and hence no electoral reward, found that African Americans do respond to a non- resident co-ethnic. His findings identified a causal mechanism distinct from rational-choice based arguments about electoral success that comes from within the legislator. Broockman’s (2013) findings are an example of Mansbridge’s (2003) gyroscopic representation, where the “representative looks within, for guidance in taking action, to a contextually derived understanding of interests, interpretive schemes, (“common sense”), conscience, and principles” (Mansbridge, 2003, 521). Voters exercise power in this regard not by “inducing preferences”, but by placing in the system someone who has “a constellation of values” that are similar to the constituent’s own, thus ensuring that even without the threat of some external sanction, the representative will act in accordance with the wishes of her constituents. These values can be signaled to the constituent through “descriptive characteristics, party identification, faithfulness to a single issue, or some personal characteristics (ibid, 521). In Broockman’s case, his results are driven by the personal desire of the African American legislator to advance the interests of his co-ethnics, and not by reelection concerns. If we follow the logic of gyroscopic representation, African American voters put someone in office who 26 shares their group consciousness and desire to help the group, and he is now acting in their interest by responding to a fellow African American even though there is no electoral award. 6. Descriptive Representation and its Limits There is a wealth of evidence to suggest that descriptive representation does lead to substantive representation for racial and ethnic minorities. Previous studies have acknowledged the problems of assuming that racial and ethnic groups have monolithic policy views on all issue areas (Griffin and Newman 2007), while still demonstrating that on certain issues, racial and ethnic groups have distinct policy preferences that are supported by overwhelming majorities of the racial or ethnic group. This is usually done by identifying objective interests and subjective interests, with the former being issues that disproportionately impact a particular group, and the latter being issues that members of a particular group self-report as being important (i.e., Swain 1995; Canon 1999; Haynie 2001; Rouse 2013, 21). While diversity of public opinion within groups is acknowledged, there is little research into how diversity within groups affects political representation. Intersectionality theory teaches us that individuals are made up of constitutive social identities, such as race, gender, class, sexual orientation, or national status, that influence their sociopolitical locations, and oftentimes, can place individuals under overlapping forms of oppression (i.e., Collins 1990; Hancock 2011; Simien 2006), that often takes the form of public policies that either ignores or harms intersectional populations (i.e., Crenshaw 1989, 1994; Cohen 1999; Hancock 2004, 2011). If increased political power for groups like African Americans and Latinos is supposed to lead to more substantive outcomes, then the question becomes, better for whom within these groups? Intersectionality theory has its roots among Black and Latina feminists who found themselves frustrated with a mainstream U.S. feminism that largely privileged the experiences 27 and goals of White, middle class, heterosexual women, and U.S. civil rights movements that privileged Black and Latino men, and largely ignored the contributions of Black women and Latinas (i.e., Collins 1990; hooks 1981; Davis 1981; Anzaldua and Moraga 1987). As a theory, intersectionality does not privilege one particular axis of domination, be it patriarchy, racism, classism, or heterosexism, as the primary source of oppression, but instead seeks to understand the interlocking nature of these forms of domination and how they structurally disadvantage certain groups. 9 Originally concerned with the intersection of race, gender, and class oppression, intersectionality theory has broadened in scope to include other relevant categories, including sexual orientation, national status, disability, and age (i.e., Fogg-Davis 2008; Romero 2008; Dhamoon 2010; Hancock 2011; Hancock 2012), and even examines the way that privileged identities and disadvantaged identities can co-exist within oppressed groups (i.e., Hancock 2009; Strolovitch 2007). Intersectionality theory is useful in discussions of descriptive representation because it directly challenges the problems of essentialism, which has been seen as one of the biggest problems with the concept. Mansbridge (1999) highlights the concern that some scholars (see, Kymlicka 1993; Young 1997) had with the notion of descriptive representation, namely that it leads to an essentialism that could potentially limit the political power of disadvantaged groups (ibid, 637, 638), and subsume all sub-identities under the dominant group (ibid, 638). As Hancock (2011) notes: “Intersectionality scholars analyze…categories as social constructions that retain political influence far beyond any actual meaning of the biological, phenotypical, and chromosomal differences among us. Many scholars recognize this claim as a constructivist one –based on the conviction that humans cognitively construct the world around them in order to best navigate a complex society. While intersecionality starts with this constructivist premise, it recognizes the material reality that these social constructions impose on us (ibid, 34). 9 In her seminal book Black Feminist Thought, Patricia Hill Collins refers to this as the Matrix of Domination (1990). 28 Intersectionality does not take any category as a given, and rejects the notion of essentialism as defined by some inherent trait specific to some certain group. Instead, it produces “historically, politically, and socioeconomically accurate information” about historically disadvantaged or stigmatized groups (Hancock 39-41). These benefits offered by intersectionality theory fits with Mansbridge’s own prescription of how to solve the problem of essentialism. Mansbridge (1999) states, “descriptive representation should stress nonessentialist and contingent reasons” that “highlights the historical contexts in which descriptive representation is likely to advance the substantive interests of the group” (ibid, 638). These historical contexts include a “previous history of exclusion from political power by the dominant group, and a history of strong prejudice toward the disadvantaged group” (ibid, 639). She also states that descriptive representation can help groups with “uncrystalized interests” more quickly ascertain their interests in an institutional context, and thus act in a way that the voter would on the basis of having a shared identity, and shared experiences (ibid, 644). Dovi (2002) would also endorse descriptive representation, but on the condition that “descriptive representatives possess strong mutual relationships with dispossessed subgroups of historically disadvantaged groups” (ibid, 729). She highlights the need for preferable descriptive representatives to engage with these subgroups and engage in acts of mutual recognition with one another (ibid, 735). Dovi is arguing for particular criterion for the selection of descriptive representatives so that the group in question, and all of its members, will receive proper representation. This means having a more interactive relationship that involves the representative “acting in concert with the disadvantaged subgroup”, and with the representative specifically identifying with the disadvantaged subgroup (ibid, 736), while still leaving room for political disagreements (ibid, 737). Dovi is advocating for a more deliberative model of representation 29 between the descriptive representative and her constituents, especially the dispossessed subgroup. It appears that she is advocating for a deliberative model of representation akin to Mansbridge’s anticipatory representation (2003), which calls for mutual education between the representative and her constituency among others. This allows the descriptive representative to learn about “latent interests,” and ensures that the she is providing representation for the subgroup. These arguments for the importance of understanding the historical determinants of disadvantage, and the need to recognize that subgroups exist within historically disadvantaged groups leaves the door wide open for an intersectional analysis of representation. The empirical studies of representation mentioned earlier in this chapter largely operated under a single-axis of identity, race or ethnicity, but more recent literature has begun paying attention to the way intersectional identities impact representation. 10 Scholarly work in this vein in political science has largely focused on women of color legislators in the U.S. states and the Congress (Hawkesworth 2003; Bratton, Haynie, and Reingold 2007; Fraga et al. 2005; Smooth 2006; Reingold and Smith 2012; Bejerano 2013; Brown 2014), or the way that stereotypes and cultural images impact intersectional groups in public policy debates (i.e., Hancock 2004). In studies of descriptive representation, intersectionality theory is most often represented in the form of political intersectionality 11 , which describes “how the representational failure created by using an elite or privileged group member as a prototype for policy remedies prevents the comprehensive representation, and by extension, remedy of the obstacles created by the 10 Earlier studies examined the electoral opportunities, and challenges, that presented themselves to women of color candidates (i.e., Prestage 1977; Darcy and Hadley 1988; Moncrief, Thompson, and Schuhmann 1991). 11 The other two domains of intersectionality are “structural intersectionality,” which “highlights the contextual factors that produced an inability to obtain legal remedies that are presumed to be available to legal subjects; and “representational intersectionality,” which addresses the ways that people who straddle multiple social locations are culturally constructed” (Hancock 2014, 45). 30 drivers of racial, gender, class, and sexuality disparities” (Hancock 2014, 43). Cohen coins the term “advanced marginalization” to describe groups, such as Black lesbian, gay, bisexual, and transgender (LGBT) people, that experience a secondary marginalization within their own group (1999, 63). The failure of Black political elites to represent the interests of Black LGBT who suffer from AIDS was indicative not only of their refusal to expend capital, but also of their adherence to “a set of middle-class racial uplift norms regarding Black sexual behavior” (Hancock 2014, 43). Hancock also details how representation along a single-axis of identity can fail disadvantaged subgroups as her analysis on the Congressional debate on welfare reform shows that White women MCs and Black men MCs did not voice support for the poor women on welfare, who were often discussed in racialized and gendered terms (Hancock 2004). 12 Cohen’s insights, and use of qualitative and quantitative data, would empirically demonstrate that not only are group interests not monolithic, but that they can actually serve to marginalize further certain subgroups. Strolovitch (2007) would later use Cohen’s idea of advanced marginalized and apply it to the study of social justice interest groups. She finds that these groups suffer from an “intersectional stigma” that makes representing them costly to the organizations. Both empirically demonstrate the dangers of relying on a single-axis of identity to fully represent all the members of any particular group, but also highlight the need for scholars to investigate the representation of advanced marginalized subgroups apart from the larger group. Aside from Cohen and Hancock (2004), the majority of work on political representation and intersectional groups in the American Politics literature remains focused on women of color legislators. The vast majority of them used the intersectionality embodied approach, which places those who hold intersectional identities at the center of the analysis, following the 12 Hancock’s analysis largely focuses on how the social cognition of the welfare queen as a poor Black woman based on crude racist and sexist stereotypes went largely unchecked, and was often reinforced by the very MCs who might have objected to characterizations based on such stereotypes (Hancock 2004). 31 “general spirit of Crenshaw’s original articulation” to “put women of color at the center of analysis” (Hancock 2014, 44-45). These studies utilize qualitative approaches like in-depth interviews and participant observation (Hawkesworth 2003; Brown 2014), but more generally, they use quantitative approaches (Bratton and Haynie 1999; Bratton et al 2006; Fraga et al 2007; Reingold and Smith 2012). The quantitative approaches are a very pragmatic way to study the influence of intersectional identities as the focus turns to which identity has the most influence on the outcome variable (Hancock 2013). For example, Bratton (1999) and Bratton et al (2006) study the legislative behavior of Black women, but largely follow a multiplicative model, that seeks to understand whether Black women act more like Black male legislators or (White) women legislators in terms of bill sponsorship, or instead act differently, possibly due to the need to represent both interests independently. This approach is indicative of what Hancock (2013, 269) calls the Multiple Quantitative Formulation, which treats social categories like race and gender as mutually exclusive, and takes an additive approach to determine which social identity, or combination of identities, has an effect on the dependent variable. Both Fraga et al. (2005) and Bejerano’s (2013) work use what Hancock refers to as the Unitary Quantitative Formulation, which asks the question of which social category, race or gender, is more important in determining the behavior or electoral chances of Latinas. 13 Fraga et al. (2005) advance a theory of “strategic intersectionality” and claim that Latinas have an advantage in that they can emphasize their gender identity over their ethnic identity in order to build support from non-Latino lawmakers, especially other women. The authors deem that Latinas have a gender-inclusive advantage because of their greater likelihood of supporting the 13 Hancock refers to the Unitary Quantitative Formulation and the Multiple Quantitative Formulation as examples of the “intersectionality-as-testable-hypothesis” approach (2013, 268-270). 32 women’s caucus than the Latino caucus, but do not analyze how gender or race relations might factor into the decision-making process of the Latina lawmakers and their interactions with Latino male legislators and non-Latina women legislators. Bejerano (2013) argues a complimentary theory to “strategic intersectionality” that posits that their gender identity gives Latinas an advantage with non-Latino voters. She does not properly theorize why the intersection of power relations related to gender and race would work in the Latinas’ favor, but as its implied, works against Latino men. Both studies do not properly attend to the power relations associated with race and gender and instead simply assume that gender, or being female, gives Latinas an advantage, putting them in direct opposition with much of the women and politics literature (e.g., Lawless and Fox 2010; Anzia and Berry 2011). While intersectionality theory has begun to explore categories related to agency and privilege (Fogg- Davis 2008; Hancock 2011), the onus is on both Fraga et al and Bejerano to explain why gender provides these benefits to women, and requires extensive analysis of both race and gender together. While more recent quantitative studies have sought to incorporate the power relations associated with their categories of analysis into their theoretical frameworks (Reingold and Smith 2012; Reingold and Haynie 2014), they are still primarily concerned with the identity influence of the legislator. This is certainly important, but for this chapter, there is one big question that has now been addressed: what type of representation can hold representatives accountable to intersectionally stigmatized groups? Hawkesworth’s (2003) qualitative account of the experiences of women of color MCs suggests that that the experiences of marginalization experienced by these women on account of their race and gender led them to forcefully advocate 33 for poor women on welfare in the debates on welfare reform, thus alluding to Mansbridge’s gyroscopic representation. While her study is not directly about accountability, Hawkesworth offers examples of women of color engaging in actions akin to anticipatory representation, where they held forums and town halls in order to educate and influence their constituents, and press conferences with welfare recipients in order to change the debate (2003, 545-546). Hawkesworth’s aims in her article, and the fact that is “suggestive, not definitive” (ibid, 546) leaves the mechanism of accountability unclear, but at different times, alludes to gyroscopic representation, anticipatory representation, and possibly surrogate representation, as the Congresswomen of color attempted to represent all welfare women in the debate on this bill, not just those in their districts. Brown would offer the most direct assessment of how intersectional groups, but not necessarily intersectionally stigmatized groups, can achieve substantive representation. In a qualitative study of the experiences of African American female state legislators in Maryland, Brown (2014) employs an intersectional framework to understand how their race-gender influences their legislative behavior. Her innovation of representational identity theory borrows from Mansbridge’s gyroscopic representation in that Black women legislators look within to their understandings of how power relations related to race and gender, among other categories, shapes their approach to representing Black women, African Americans, and representation generally (ibid, 6, 11-13). Brown’s analysis of the legislative actions of Black women legislators uses the embodied intersectionality approach to highlight the individual attributes of each one and analyzes how their “lived experiences” work in tandem, or in tension, with their ascriptive identities (ibid, 10, 12). 34 The literature review above demonstrates that descriptive representation is generally linked to substantive representation, but leaves questions about how to ensure that all members of the racial or ethnic group are properly represented. Most of the studies of representation focus on a single axis of identity, and do not extend their analysis to intersectionally stigmatized members of the group. The studies that do exist mostly use the embodied intersectionality approach, and focus on the legislative behavior of the legislators with intersectional identities. There is little in the way of examining the relationship between the intersectionally stigmatized constituent and the legislator. The current literature has shown that descriptive representation is not by itself sufficient for the representation of intersectionally stigmatized groups, but offers potential clues for what mechanisms can help facilitate their representation. Mansbridge’s typology is most helpful here, especially her formulations of gyroscopic representation, anticipatory representation, and surrogate representation. These formulations all move away from sanction-based models of representation and provide the foundation for plausible explanations as to why legislators would represent intersectionally stigmatized groups. The next chapter will move forward with the discussion of how legislators, primarily Latino legislators, come to represent the interests of undocumented Latinos. 7. Theorizing the Representation of Undocumented Immigrants As a population, undocumented immigrants face a host of disadvantages unique to their status in the United States. These include facing the constant threat of deportation; the inability to work legally in most states; lack of access to federal programs, including financial aid for education and health insurance; being prohibited from obtaining a driver’s license in most states; and the absence of political rights. The lack of legal status has forced over 11 million people in 35 the United States to live, and work, in the shadows of our society in the early part of the 21 st century. This changed drastically in the 2006 when over a million undocumented immigrants and their supporters took to the streets across the country to protest H.R. 4437, formally known as the Border Protection, Anti-Terrorism, and Illegal Immigration Control Act. 14 The restrictive House bill sought to make it a crime to be in the United States without legal documents, and it led to extensive mobilizing by undocumented immigrant communities, especially among Latino immigrant groups (i.e., Ramirez 2013; Zepeda-Millan 2014). Since then, much of the empirical research in political science has been focused on the effects of the political marches on political attitudes and behavior (Zepeda-Millan and Wallace 2013; Ramirez 2013; Wallace, Zepeda- Millan, and Jones-Correa 2014). The emergence of primarily Latino undocumented immigrants on the political scene advocating for their rights also demonstrated that even though they could not vote, they could still exercise political power. Political theorist Cristina Beltran (2010) argues in her landmark theoretical treatment of Latino political identity that the political nature of the mass immigration protests have been overlooked in favor of political arguments related to the issues political issues surrounding the marches (ibid, 131). By doing so, scholars and the mass public ignore how undocumented immigrants were “actualizing a power they did not have yet” to formally enter the public space and contest openly for their rights (ibid, 132). The marchers also created their own discourse, using what Beltran describes as “festive anger,” to both challenge the injustices they faced, but also to reaffirm their commitment to the promise of the traditional U.S. immigrant narrative (ibid, 143). I believe that the main implication of Beltran’s treatment of the immigrant 14 Undocumented immigrants and their allies had been active at the local and regional level prior to the 2006 protests (see Hondagneu-Sotelo 2008), but I emphasize the significance of these protests for their national character. 36 marches for empirical political scientists is that undocumented immigrants must use participatory forms of democratic action in order to achieve recognition as political actors from both the mass public and from the government. Exactly how this representation is achieved through recognition is not made apparent by the literature in political science. While previous studies have examined the determinants of passage certain types of immigration bills (i.e., Wallace 2014), and others have assessed the link of descriptive representation to the passage of bills (e.g., Rouse 2013), none have ventured too deeply into theorizing why legislators would represent the interests of undocumented immigrants (see Rocha and Matsubayashi 2012). Chavez et al (2014), for example, use qualitative methods to explain the political behavior of immigrant youth like the DREAMers and identify what enables, or hinders, their activism, such as state factors, as well as what motivates them to act, but it does not focus so much on representation. This chapter will seek to answer this question by offering an explanation for why legislators should not necessarily represent the interests of undocumented immigrants, and then offering a new direction for why legislators might represent them. 8. The Issue of Accountability Despite the many disadvantages that undocumented immigrants face in U.S. society, there is one that is particularly influential for them in terms of political representation: their inability to directly punish or reward legislators through the exercise of voting. This lack of formal political power has no doubt hampered the ability of political science to study the representation of undocumented immigrants because electoral rewards, and sanctions, are thought to drive much of legislative behavior (i.e., Mayhew 1974). If this is the case, then legislators should not be concerned with representing undocumented immigrants at all since they 37 theoretically cannot help legislators achieve their primary goal of reelection. The question for political science then becomes, why would legislators represent a marginalized group that cannot vote, and might never have the opportunity to vote? Undocumented immigrants belong to a larger group of people that also experience degrees of political marginalization: the larger category of immigrants. The recent wave of immigration has largely brought in populations from Asia and Latin America, with a declining number coming from Europe. The increasing diversity of immigrants adds a racial dimension to the discussion of immigrants and voting. As pan-ethnic groups with large immigrant populations, Asian Americans and Latinos vote at much lower rates than African Americans and Latinos, even when accounting for citizenship status (i.e., Lien 2000; Ramakrishnan 2005). Scholarly research has shown that community-based organizations and labor unions can step in and help mobilize immigrant voters (i.e., Wong 2006; Garcia Bedolla and Michelson 2012), but their limited resources makes the possibility of a mass mobilization of immigrant voters unlikely (i.e., Wong 2006). 15 These studies offer a critique of pluralist theory (i.e., Truman 1951; Dahl 1961; Wilson [1974] 1995) by acknowledging that there are barriers to the participation of immigrants. However, the underlying theme of these studies is that immigrants and can have some kind of influence if they naturalize and are mobilized to vote (i.e., Pantoja, Ramirez, and Segura 2001; Ramakrishnan and Espenshade 2001; Ramakrishnan 2005). These studies, for the most part, do not entertain the notion of what happens if immigrants, like the undocumented, are completely barred from voting. While there are discussions that today’s immigrants are tomorrow’s voters, 15 Wong argues that political parties due to the financial constraints of the modern campaign are not mobilizing immigrants. 38 this idea does not hold much promise for undocumented immigrants whose legal status, and future in the United States, remains uncertain. Since they lack the ability to sanction, or reward, legislators directly through the exercise of the vote, undocumented immigrants are excluded from traditional forms of incentivizing legislators to represent them. In order to understand the dilemma of the undocumented immigrant and the polity, it is necessary to first understand how they became a part, albeit an invisible part, of the political body in the first place. The next section will briefly examine how the legal structure of the United States set up the necessary condition for undocumented immigrants to receive representation. 9. Determinants of Representation for Undocumented Immigrants As Beltran (2010) writes, undocumented immigrants in the United States cannot be thought of in the same light as the stateless, and therefore, “rightsless” people that Hannah Arendt writes about in The Origins of Totalitarianism (ibid, 138-139). While undocumented immigrants “live under conditions of acute vulnerability” but are still granted personhood that “does in fact offer a form of civic standing, limited though it may be” (ibid, 139). The United States grants undocumented immigrants a certain set of rights, including the right for children to attend public education and due process in criminal rights proceedings. Undocumented immigrants might not have citizenship status, but they still are able to enjoy certain attributes of substantive citizenship. The fact that they have rights means that the undocumented cannot only demonstrate for new rights, but it means that they are in fact entitled to certain services at the different levels of government (Bosniak 2006). The availability of rights to the undocumented also extends to another area: the right to be counted. In the 1960s the Supreme Court ruled in a series of cases that legislative districts 39 must be made of equal population. These cases did not make any stipulation for what type of person is to be counted, meaning that those who could not vote, like children, and those who are not citizens must be counted and included in legislative districts. 16 The Court required that “precise mathematical equality” be the standard for drawing districts and rejected the many arguments that sought to justify any population disparity (McConnell 2000). Undocumented immigrants, thus, became part of what Fenno described as the “geographic constituency” (Fenno 1978). While there is very little research on elected officials and their responsiveness to noncitizens, Butler et al. (2012) found that both Members of Congress and state legislators were responsive to constituent letters concerning how to help noncitizens begin the citizenship process. I believe that the availability of some rights to the undocumented immigrants, and the fact that they are counted as part of all district populations satisfies the most basic requirement of representation: presence. They are physically present, and to some extent, they are legally present as well. What might prove to be especially important is that large populations of Latino undocumented immigrants are often drawn into Latino-majority districts, making their presence especially of consequence to officials elected from those areas (Schmidt et al 2010). These conditions lead to the kind of political empowerment that Beltran writes about, and I believe, helped lay the foundation for representation. The lack of voting rights means that undocumented immigrants cannot use the threat of sanction to get legislators to respond to their interests, but I posit two alternative suppositions for what kinds of representation could guarantee some form of representation from legislators for 16 These cases are Baker v. Carr (1963), Gray v. Sanders (1963), and Reynold v. Sims (1964). 40 undocumented Latino immigrants. Both suppositions highlight the importance of the larger Latino category to the representation of undocumented immigrants. Anticipatory representation (Mansbridge 2003) might serve as a mechanism for undocumented immigrants to gain representation as legislators learn about the issues affecting undocumented immigrants, and how it influences constituent opinion. The immigrant marches were certainly a moment of recognition for the political power of undocumented immigrants, but it is far from the only way for politicians to learn about undocumented immigrants. At the local, state, and national level, there are interest groups and activists engaged in lobbying and educating legislators about the needs of undocumented immigrants (e.g., Nicholls 2013; Seif 2004), and that extends to other representatives educating their colleagues about these issues (e.g., Seif 2004; Rincon 2010). From these lobbying efforts, legislators might learn their voting constituents also care about these issues and thus seek to address these issues in some fashion in order to please them in future elections. Descriptive representation is another avenue in which undocumented immigrants might expect representation. While undocumented immigrants can be of any race or ethnicity, the fact is that they have multiple identities that might be salient to legislators. Since undocumented immigrants cannot run for office, they cannot count on undocumented immigrants to provide them direct representation, they might be able to seek representation from legislators who share the same ethnicity. I posit that the since undocumented Latinos lack the ability to sanction or reward legislators with reelection, they must count on a different kind of representation: gyroscopic representation (Mansbridge 2003). This form of representation would require a representative to look within to her inner values to make decisions. Her decision to represent undocumented 41 immigrants would also reflect the desire of her constituents to act on issues to be responsive to undocumented immigrants. A normative function of descriptive representation is to provide a voice in government for marginalized groups. Latino elected officials have been found to believe it is important for them to represent the interests of Latinos, but not at the expense of other constituent groups (Casellas 2011, 125; Geron 2005, 197-200. While immigration does not rank highly among the concerns of Latino elected officials or state legislators, many Latino elected officials generally express personal support for immigrant rights (Geron 2005, 198; Rouse 2013). Therefore, group consciousness would need to be present among the Latino legislators, and possibly their constituents, in order for the former to represent undocumented Latinos. Group consciousness “involves identification with a group and a political awareness or ideology regarding the group’s relative position in society along with a commitment to collective action aimed at realizing the group’s interests” (Miller, Gurin, Gurin, and Malanchuck 1981). It is the concept that forms the basis for linked fate, which in Dawson’s (1994) formulation, is a heuristic used by individuals where they use “the social standing of the group as a proxy for the well- being of the individual” (McClain et al. 2009). McClain et al. (2009) argue that it is a particular measure of “racial group consciousness” that relies on race as a cue to activate the heuristic. Greenwood (2008) explicitly identifies that group consciousness can occur on both a singular level, meaning along one axis of identity and an intersectional level, which accounts for the need to recognize multiple identities. In her study, she found that women activists with high degrees of intersectional consciousness did not view diversity within a social change organization as negatively as women with a singular consciousness. The concept of “intersectional political consciousness” is very close in character to what Strolovitch describes as “intersectional linked fate” (2007, 11). Like Greenwood, Strolovitch originated this term in 42 connection with social justice organizations, and used it to describe the responsibility of social justice organizations to advance the interests of advanced marginalized subgroups. This form of linked fate means that not only do members of a particular group believe that their political interests are bounded to one another, but that the interests of the disadvantaged are important to the political well-being of the privileged members of the group (2007, 63). Both Greenwood and Strolovitch would publish their respective works within a year of one another, but would largely come to the same conclusion: political activism that reflects only one axis of identity is not sufficient to deal with diversity that exists within groups, like women. There are important differences in their approaches, however, as Greenwood situates her study in the political psychology literature, and Strolovitch writes in the context of interest groups and advocacy. Greenwood’s analysis thus exists at the mico-level as she conducts lab experiments with female activists (2008) to test their individual consciousness, and Strolovtich’s work exists at the macro-level as she surveys social justice organizations on their commitment to disadvantaged subgroups (2007). As such, Greenwood’s analysis does not necessarily deal with how such consciousness works in the context of actions related to social change, and Strolovitch does not necessarily utilize how intersectional linked fate works in the context of political psychology. However, I believe that the two works do speak to each other, and when brought together can help fill in the blanks of how intersectional political consciousness can play out in the political arena. Strolovtich does not specifically address group consciousness in detail, but does write that she draws from Dawson’s theory of linked fate in developing her concept of “intersectional linked fate.” As McClain et al. (2009) have stated, linked fate is seen as a particular application of group consciousness, one the emphasizes the well-being of the individual group members with 43 that of the larger group. Linked fate had been originally thought to explain African American political behavior, including voting, (Dawson 1994), but others (e.g., Chong and Rogers 2005) argue that it has its most robust effects on African American policy attitudes. Therefore, linked fate would not exist without group consciousness, and this means that Strolovitch’s theory of intersectional linked fate would not exist without it as well. This requires a rethinking of what kind of group consciousness underlies intersectional linked fate. Dawson’s theory of linked fate has been used to study the political behavior of all racial groups in the U.S., (e.g., Sanchez and Masuoka 2008; Masuoka and Junn 2013), and it has been done so only on one axis of identity: race. A singular orientation of group consciousness would not suffice for intersectional linked fate as it requires the social justice officer, or any representative, to be able to account for how multiple identities, or domains of oppression, serve to disasdvantage particular subgroups that her group represents. Greenwood’s particular insight about intersectional political consciousness is most useful here as I theorize that in order for a representative to have a sense of intersectional linked fate, they must possess an intersectional political consciousness. In this sense, intersectional linked fate is a particular application of intersectional political consciousness as it relates to representatives. I will use the term intersectional linked fate to describe the sense of responsibility that a representative from an advantaged subgroup, like Latino political representatives, has to represent members of a disadvantaged subgroup, like undocumented Latinos. The rationale here is that intersectional linked fate is an application of intersectional political consciousness to the context of representation. This is done both to acknowledge the spirit of Strolovtich’s original articulation of the term, as well as to not unintentionally limit the scope of Greenwood’s innovation of intersectional political consciousness. It is also an acknowledgment of the fact that 44 intersectional linked fate is a specific measure designed as an alternative to rational choice theories of legislative behavior. Intersectional linked fate is a measure designed to measure whether a representative that belongs to a marginalized group (like Latinos) is able to provide representation for an advanced marginalized member of her group (like an undocumented Latino). This theory of intersectional linked fate will be thought of in the same terms as Broockman’s (2013) intrinsic motivations, which he used in his study of African American legislators and their representation of co-ethnics who live outside their districts. According to Broockman (3), intrinsic motivations are done solely for the psychic benefit of the legislator (also, see Citrin and Green, 1990), as opposed to extrinsic ones, which are done for some outside reward, such as ensuring reelection. Intersectional linked fate 17 works as a more specific intrinsic motivation because it stems from the recognition of the Latino legislator that in order to be a representative of all Latinos, she must represent the interests of those that are disadvantaged, like the undocumented. This would mean that she is not only cognizant of the issues that undocumented Latinos face, but must be active in her official capacity as a legislator to help undocumented Latinos. As a very specific intrinsic motivation, intersectional linked fate fits into Mansbridge’s theory of gyroscopic representation because it means that the Latina representative is helping the undocumented Latino immigrant because she feels it is a part of her duty as a legislator, a part from her reelection concerns. It is also possible that her constituents, or at least some of them, 17 Broockman (2013) shies away from the term “linked fate” because it is a rational choice explanation for why a racial group, like African Americans, tends to have a largely unified policy agenda. I use the term “intersectional linked fate” because I interpret Strolovitch’s definition of the term denotes the need to move beyond self-interest (2007, 63). She states that social justice organizations believe in justice, and that Melissa William’s (1998) argument in favor descriptive representation is predicated on the idea that such representation for historically disadvantaged groups is done in favor of the entire group. 45 elected her because they care about undocumented immigrants and they believe that she will act accordingly even without the threat of sanction. Latino legislators have embraced the concept of the pan-ethnic identity of Latino and the responsibility it engenders to represent all the national groups under its umbrellas, even those different from their own national-origin group 18 (Geron 2005, 199). To embrace the pan-ethnic group means that the legislator must be more malleable to changes within her own sub-ethnic group and be more willing to act on issues that does not affect her, or even the majority of her sub-ethnic group directly. There exists some empirical evidence that this might be the case as Geron (2005), Casellas (2011), and Rouse (2013) all find that Latino elected officials, including state legislators, are supportive of immigrant rights. Among the greater Latino public, research has found that U.S. born Latinos have connected to the immigrant rights marches to their own feelings of racialization (Zepeda-Millan and Wallance 2013), and expressed support for undocumented immigrants (Wallace, Zepeda-Millan, and Jones-Correa 2013). These empirical studies demonstrate that the conditions might exist for Latino elected officials, such as state legislators, to engage in acts of gyroscopic representation. 10. Conclusion This chapter identified that one of the current debates about descriptive representation involves a discussion of causal mechanisms, more specifically, the influence of extrinsic motivations, such as the desire for reelection, versus intrinsic motivations, those done for the psychic benefits of others. The discussion of such specific mechanisms for the purpose of 18 For example, Geron (2005, 199) shares the stories of a Latino legislator of Puerto Rican ancestry who became especially active on issues related to the burgeoning Mexican immigrant population. This is important because Latinos are a pan-ethnic group that encompasses many different national groups with different histories of migration and discrimination, and vary across important identity markers, such as national status, generation, and race. 46 descriptive representation fits with the idea that such representation is done for normative purposes of justice and fairness for historically disadvantaged groups. Intersectionality theory helps us to understand that racial and ethnic groups are not monolithic, but instead are made up of subpopulations that are advantaged or disadvantaged by their intersectional identities and that descriptive representation theorized along a single axis of identity is not be enough to represent them. This chapter ended with a discussion of how an advanced marginalized group like undocumented immigrants might be able to count on intersectional linked fate among those who share the same ethnicity to represent them. The next chapter will operationalize intersectional linked fate among Latino legislators as it relates to undocumented immigrants. 47 Chapter 3: Who Represents Undocumented Latinos? A Field Experiment of State Legislators 1. Introduction The previous chapter demonstrated how descriptive representation on a single-axis of identity is not sufficient in and of itself to address the needs of intersectionally stigmatized subgroups. This is especially true when one considers that most studies of representation are implicitly based upon sanction-based models of representation that finds the legislator prioritizing her reelection above all else. This would seem to only bode ill for undocumented Latino immigrants who are both structurally barred from voting, but are also often the targets of highly emotional and sometimes racist rhetoric (e.g., Chavez 2011). In this chapter, I offer a scenario in which undocumented immigrants might be able to not only obtain representation, but also equitable representation. I theorize that most legislators have few incentives to be responsive to undocumented residents in their districts. I further theorize that Latino legislators are responsive to undocumented constituents living in their districts because of intersectional linked fate (Strolovich 2007); while non-Latino legislators are less likely to engage in responsiveness to undocumented constituents. This theoretical argument is an important contribution to our understanding of status, representation, and the extent to which elected officials may respond to those where the electoral connection is tenuous or even nonexistent. In order to investigate these questions, I have developed a novel field experiment that tests whether legislators have an intrinsic concern for representing undocumented Latinos. The experiment consisted of constituent contacts that varied the citizenship status of a Latina constituent by specifically identifying her as an undocumented immigrant or as a citizen to state legislators in 14 U.S. states with significant Latino populations. In addition, a control group of 48 legislators received the same message but without identifying the constituent as either a citizen or undocumented. This article is the first to ever examine whether legislators represent undocumented Latinos, and it is also the first to examine whether citizenship status influences a legislator’s decision to represent a constituent. 2. Service Responsiveness as Representation Constituency service, also known as casework, has long been thought of as one of the central duties of representatives (Eulau and Karps 1978; Fiorina 1989; Rosenthal 1998). It is often thought that such work is done to help the representative secure the votes of individuals for whom they might not be providing policy responsiveness (Cane, Ferejohn, and Fiorina 1987). While the literature is mixed as to the actual effects of the so-called “personal vote” (Johannes 1984; King 1991; Serra and Cover 1992; Serra and Moon 1994) it is apparent that representatives believe it provides them with electoral benefits (i.e., Ellickson and Whistler 2001). Many of these studies have been focused on self-reporting of time spent on casework, but other work has survey data of the mass public and their opinions of their legislators constituency work (Serra and Cover 1992; Serra and Cover 1995). These works have largely evaluated how constituents and legislators perceive casework, and in some cases, sought to use them to explain the incumbency advantage. Casework is also important beyond its possible connection to the legislators’ electoral goals, as it is one form of responsiveness that is solely dependent on the legislator (Butler 2014). Other forms of responsiveness, such as roll call votes and bill sponsorship are measures of legislator choice, but are also dependent on the complex institutional decision-making process. Constituency service, on the other hand, is entirely left up to the discretion of the legislator and 49 her office, and other research demonstrates that it can have different purposes for both legislators and constituents. Using elite interviews with legislators and their staffs, Swain (1995), Cannon (1999), and Gross (2011) find that gender and race hold different meaning for legislators and their staffs. Swain (2005) and Canon (1999) argue that having staff members of diverse background can make members of minority communities feel that the MC cares about their interests, which helps make the MC and her office seem more welcoming. Grose (2011) finds through his interviews with staffers of Black MCs in the South that Black constituents feel more comfortable reaching out to Black MCs. Thomas (1991) finds that female MCs personally prefer constituency service to legislating. The insights gained through these studies and methods are important but they come with important limitations. Surveys of legislators, for example, do tell us in what order they rank their duties, but not how they treat all of their constituents. The qualitative treatments, which I believe do allow for analysis of how race and trust can impact outreach to minority constituents, are susceptible to criticisms that they are too limited in scope. Qualitative methods, such as elite interviews and participant observation, are also vulnerable to selection bias. Both methods are also subject to charges of bias to social desirability. For scholars of descriptive representation, or who are generally concerned with representation and equality, these studies cannot tell us whether legislators respond to all constituents equally. Service responsiveness is an important aspect of representation because it is the one aspect of substantive representation that involves a direct a relationship between an individual constituent and her legislator. As such, it is also an area of study that has come under increasing scrutiny as scholars debate whether the perceived benefits of descriptive representation affects 50 this particular domain of representation. Previous research indicates the race of the constituent does impact service responsiveness (Butler and Broockman 2011; Broockman 2013; Butler 2014), and argues against the dominant rational choice model that prioritized reelection concerns over other mechanisms of representation. 3. Why Field Experiments? Field experiments offer political scientists a novel way to evaluate the behavior of individual representatives. They have been used extensively in studies of political behavior, most prominently in those studies concerned with voter turnout and campaigns (Green and Gerber 2000; Green and Gerber 2008; Panagoplous 2011). When it comes to the study of race and ethnicity, field experiments have been especially helpful in determining the effect of being contacted by someone of the same race (Michelson 2003; Garcia Bedolla and Michelson 2012). Following the lead of political behaviorists, institutional scholars have begun to use field experiments to determine how representatives at the local, state, and federal level respond to inquiries from constituents. Such studies have been used to uncover legislator bias either in favor of, or against certain groups (i.e., Butler and Broockman 2011). The ability to examine legislative bias is made possible due to the ability to directly randomize the assignment of subjects to treatment groups, therefore automatically controlling for confounding factors (e.g., McDermott 2002; Morton and Williams 2010; Grose 2014). Researchers are able to directly compare response rates across treatment groups to determine whether disparities exist by holding the message constant. With the message similar across all treatment groups, the only variation occurs along the independent variable, which is usually indicated by the text or the name of the constituent. This independent variable could be the 51 randomization of electoral rewards (i.e., Broockman 2013; Butler 2014, 100-110), or the ethnicity of the sender (i.e., Butler Butler and Broockman 2011; Mendez and Grose 2014). 4. Representation and the Undocumented Immigrant The discussion in Chapter 2 laid out justifications for why we might see differential responsiveness to constituency requests from undocumented immigrants. If legislators are motivated almost exclusively by electoral goals and undocumented immigrants are marginalized constituents, then the direct electoral goals hypothesis, a null hypothesis, would suggest the following: Legislators will be unlikely to be responsive to requests for assistance from undocumented constituents. There are, however, reasons to anticipate that legislators might provide responsiveness to an undocumented immigrant. It is possible that a legislator may perceive some indirect electoral benefit from responding to undocumented constituents. Perhaps legislators representing districts or states with high Latino and/or foreign-born populations may think an undocumented immigrant may know and communicate about the legislator with potential voters. This anticipation of potential benefits to or consequences for the legislator factors into her logic (Arnold 1990) of whether to respond to the undocumented Latino immigrant. This calculation could be based on the importance of the Latino vote to the legislator’s reelection. It is also possible that the debate regarding issues surrounding undocumented immigrants had reached the legislator, possibly through direct contact with immigrant rights activists. Failing to respond could make the legislator think she will be targeted by immigrant rights activists for demonstrations, but not be electorally threatened. Even if this is the case, however, it is likely that legislators will prioritize responsiveness to constituents who are citizens, and then make time for undocumented constituents only after 52 citizen requests are fulfilled. It is also possible that some legislators will intrinsically be motivated to be responsive to undocumented constituents, but that citizens will still be prioritized for electoral purposes. For this reason, I posit the differential citizen-undocumented responsiveness hypothesis: Most legislators will respond less to a constituent request from an undocumented Latino than they will to a Latino citizen. Latino legislators have previously stated t they believe they should represent Latino interests, and have been found to be active on pro-immigration measures according to indicators of legislative choice (i.e., Casellas 2011; Rouse 2013; Wallace 2014). It should be remembered that Latino legislators still have the same goals as other legislators with reelection being the top concern. If Latino legislators are primarily engaged in vote-seeking behavior, then we should also expect them to behave in a manner consistent with the hypotheses above. If Latinos do, however, have an intrinsic motivation to represent all Latinos, regardless of their national status, then I theorize that intersectional linked fate could explain the higher propensity of Latinos to respond, since a desire to represent Latinos should include representing undocumented Latinos. For these reasons I posit the conditional Latino legislator hypothesis: Latino legislators should be as likely to respond to a constituent request from an undocumented Latino as to a constituent request from a Latino citizen. Thus, I expect that a sub-sample of Latino legislators will respond similarly to constituent requests from an undocumented Latino and a Latino citizen. Non-Latino legislators are expected to be less responsive to undocumented Latinos relative to citizen Latinos because there will be little intrinsic motivation to represent a group that lacks the ability to vote and does not share an intersectional linked fate with the non-Latino legislators. 53 5. Operationalization and Identity Influence This chapter will utilize a field experiment to assess the motivations of legislators to provide representation for undocumented immigrants. The first field experiment will test whether the national status of a fictitious Latina determines the degree to which legislators respond to a constituent email. The primary emphasis of this field experiment will be whether Latino legislators provide service responsiveness to an undocumented Latina immigrant at the same rate as a Latina citizen. In this chapter, I only vary national status given my theory is about legislator responsiveness due to national status. In addition, given that my theory predicts that Latino legislators in particular will treat undocumented and citizen Latinos similarly, creating any additional treatment groups would reduce statistical power too much for this theoretically important subgroup of legislators (there are 199 Latino legislators in the sample). Thus, this study incorporates one particular aspect of intersectionality, that of recognizing the differing levels of privilege and power within a racial group as it relates to certain categories of difference. In the case of this field experiment, the subjects of the analysis, state legislators, are only examined along a single-axis of identity, their ethnic or racial identity. This particular experiment requires that in order to determine the effect of national status, it must be disaggregated from ethnicity both theoretically and empirically, which stands in opposition to the normative intersectionality theory (Hancock 2013, 280). This research objective and the lack of categorical multiplicity among the subjects, the legislators, make this experiment a standard social science experiment on identity influence. This experiment is able to maintain its link, albeit tenuously, to intersectionality theory because of its acknowledgment that diversity does exist within the Latino community, and it could lead to different effects. 54 This experiment uses the a variant of the “intersectionality-as-testable hypothesis” known as the Unitary Quantitative Formulation which only focuses on a single category of difference and asks whether it leads to certain effects. While certainly a pragmatic way to answer questions related to concerns of identity influence, it does not satisfy the theoretical or empirical requirements of an intersectional analysis (Hancock, 2013). This analysis brings undocumented Latino immigrants into the discussion on representation, and challenges their invisibility on this topic, albeit in a limited fashion. This limited inclusion does not place them at the center as much comment on the ability of institutions to recognize them. Part of this limitation stems from the advantage of field experiments: the ability of carefully designed field experiments to better capture causal effects (Grose 2014). This limitation would require the researcher to assemble a sample large enough to accommodate multiple interaction effects, and including variables that speak to the intersectional subjects’ socio- historical context (Rouhani 2014). As stated previously, the subjects of this analysis are the subjects and the main research question concerns whether national status influences their ability to respond. For this reason, I consider this to be more of a study of identity salience and whether the Latino ethnicity will prove more salient to the Latino legislators than the lack of electoral reward that undocumented status symbolizes. 6. Research Design: A Field Experiment of U.S. State Legislators To test the hypotheses, a field experiment was conducted with state legislators as subjects. Legislators were randomly assigned into three groups and each of these groups received a specific constituent message requesting assistance. The first group of legislators is the undocumented constituent treatment group, the second group of legislators is the citizen 55 constituent treatment group, and the third group of legislators is the control group who received a message from a constituent who did not identify her citizenship status. Table 3.1: Constituent Treatments Treatment 1 Treatment 2 Treatment 3 Undocumented Latina Latina Citizen Latina Control For all three legislator groups, a Latina alias was created for the purpose of sending emails to state legislators. 19 The Latina constituent wrote the legislator to simply ask for help finding information regarding federal and state aid. The messages to legislators from constituents did not require much in the way of legislator (or staff) time and resources, and carried no political risk for the legislators. The issue of student financial aid was chosen because it was realistic. Financial aid for students is an issue that falls under the domain of state policy, and it is a timely issue for undocumented immigrants, as the repeated failure of the U.S. Congress to pass a federal DREAM Act provided an incentive for undocumented immigrant youth activists to push for state level DREAM Acts (Nicholls 2013). Concurrent with the study’s time period, states such as Maryland and California passed legislation granting in-state tuition to undocumented students (Marimow 2012; Medina 2011). It is also an issue that would concern a constituent who was a student. The name used for the constituent was Maria Fernanda Garcia. Garcia was used as it is the most common Spanish surname in the United States (Roberts 2007), and Maria Fernanda was 19 A Latina alias was chosen because the focus of the study is on the immigration status of the constituent, and not ethnicity. Thus, the ethnicity of the constituents is held constant across all treatment groups. However, I anticipate that Latino legislators will be more responsive to undocumented Latino immigrants than will non-Latino legislators. I sought and was granted exemption by my university’s Institutional Review Board. The human subjects are public elected officials, a category officially classified as exempt. 56 used because it is one of the most common contemporary children’s names in Mexico (Top Baby Names 2010). A female alias was chosen because the majority of Latino college students are female (Branch-Brioso 2008). The text of the emails sent to the three groups of legislators was exactly the same except the citizenship status of the constituent was varied. In the first treatment group, the message to the legislator explicitly states that Maria Fernanda Garcia is an undocumented immigrant, and that her family is undocumented as well. The second treatment email message explicitly states that Maria Fernanda Garcia is a citizen of the United States, as designated by her status as a third-generation Mexican-American. The inclusion of her generational status is intentional, as it indicates to the legislator that not only was the constituent born here, but that her family has a historical presence in the country making her a citizen of the United States. The third email message (control group) does not state whether Maria Fernanda Garcia is a citizen or an undocumented immigrant. The text, in full, is below. The sample of legislators is from U.S. states with significant Latino populations. In order to maximize the possibility that the state legislators will be from districts that feature undocumented Latino constituents, I included states with Latino populations of over 10 percent (U.S. Census 2010). This yielded a sample of state legislators from 28 legislative chambers in 14 states: Arizona, California, Colorado, Connecticut, Florida, Illinois, Kansas, Nevada, New Dear Assemblymember/Representative/Senator, My name is Maria Fernanda Garcia and I need help finding financial aid for college. I would be the first person in my family to attend college, which is very meaningful since we are [national status]. As a resident of your district, I was hoping you could help me find information on state, federal, or community grants. Thank you for your time. Sincerely, Maria Fernanda Garcia 57 Jersey, New Mexico, New York, Texas, Utah, and Washington. 20 Nine of these states (California, Connecticut, Illinois, Kansas, New Mexico, New York, Texas, Utah, and Washington) have in-state tuition policies for undocumented youth with four (California, Ilinois, New Mexico, and Texas) providing direct access to financial aid. Only two of the states in the sample, Arizona and Colorado, had policies explicitly banning in-state tuition for undocumented immigrants. These states yielded a total sample of 1,871 state legislators with email addresses. The benchmark of 10% Latino population was chosen in hopes of producing a sample that would yield a substantial number of Latino legislators, and also by extension, states with sizable populations of undocumented immigrants. I also believed that using these states would yield similar contexts as 2012 saw activism on issues related to in-state tuition for undocumented immigrants. These states serve the purpose of my study well as the activism occurring around this time helped heighten the potential rewards, or risks, for legislators as failure to respond could lead to additional activism by immigrant advocates, but not necessarily an electoral sanction. These states were also chosen as a precaution as I was originally concerned legislators might not believe that the letter from an undocumented immigrant was real (this was a concern that would later prove to be unfounded, see Chapter 4). The emails were sent over a two-week period in June 2012 (emails within the same legislature were sent on the same day). Legislators’ email addresses were collected from the official websites of state legislatures. Table 3.2 below presents a balance check across a number of covariates, and it shows that the two treatment groups and the control group features little variation of Democrats, Latinos, and women. These groups were checked because they are coded 20 This study also originally included the states of Idaho and Oregon, but these two states were excluded due to evidence of spillover/contamination between legislators (based on post-treatment debriefing and inquiries). 58 1 for the following co-variates: party, Latino, and gender. Differences across the two treatment and control groups were found to be statistically significant, suggesting that the random assignment was implemented properly. Table 3.2: Balance Checks Across Covariates Undocumented Treatment Citizen Treatment Control Group Democratic legislators (%) 48.8% 44.5% 46.6% Female legislators (%) 27.0% 26.5% 27.5% Latino legislators (%) 11.6% 8.7% 8.8% 7. Deception And Ethical Considerations The goal of this dissertation was not to expose a condition of animus against undocumented immigrants, but rather determine whether they are considered less of a priority by the legislator, and whether intrinsic motivations related to ethnic identity can override this calculation. This is a rational calculation on the part of the legislator, and a worthwhile question as the previous chapter has argued. However, studying this important question raises a number of important ethical questions, including social desirability bias, as legislators might respond to any questions with answers they believe are socially acceptable, and also will not harm their reputations. More importantly, the questions asked by this study could not be done without putting individuals, in this case, undocumented immigrants, at risk. To recruit them in such a role could potentially put them in danger of being caught up in the immigration dragnet, as well as possibly cause them psychic harm. For these reasons, deception is used as tool to help this study accomplish its goal. Deception is a quite common, and useful, tool to social scientists interested in uncovering discrimination against certain groups. It has been used commonly in audit studies examining 59 racial bias in the labor market labor market (Riach and Rich 2002; Bertrand and Mullainathan 2003). Public officials are also officially counted as an exempt category by institutional review boards (IRB), and it is common practice for all scholars to obtain approval from their immediate IRB in order to move forward with their field experiment. This study follows the precedent (i.e., Putnam 1993; Butler and Broockman 2011) that the request should come at a minimal cost to the legislator in terms of time and effort. The requests in both experiments are simple and ask for information about state programs regarding differing levels of education. The confidentiality of the legislator is of the utmost importance to me, and that is why this data will be altered to protect the identities of the legislators if it is ever shared for replication purposes. 8. Methods The results were analyzed for significance through one-tailed difference of means tests between the various mean point estimates. The first analysis focuses on all legislators and the secondary analysis divided legislators into two groups: Latinos and non-Latinos, and each group had their mean point estimates analyzed for significance through the one-tailed difference of means tests. Critical intervals at the 95% level serve as a robustness check to evaluate the significance of the mean point estimates as overlapping values should indicate there is no difference between the response rates. Ordinary least squares (OLS) regressions were run as robustness checks and included control variables related to political party, Latino voting age population, the number of legislative staff per state, and the Squire Index. Because Latino legislators are more frequently Democrats and a large number are elected from majority-Latino districts, it may be that any conditional effect of ethnicity is due to a legislator’s party or district demographics (Grose 60 2005). 21 The conditional effect of Latino population is represented as an independent continuous variable representing the percentage of a district's Latino voting age population (see Table 3.7). The separate variables for legislative staff and legislative professionalism are included because it is possible that a state's professionalism and the availability of staffers might influence the ability of legislators to respond to constituents (see Table 3.8). The conditional effect of legislative professionalism is represented as an independent variable based on a state's 2003 score in the Squire Index of Legislative Professionalism (Squire 2007). Finally, legislative staff is represented as an independent variable based on the number of total legislative staff divided by the number of total legislators per state. 9. Results Of the 1,871 legislators in the sample, 558 legislators sent email responses to constituents (a 29.8% total response rate across all three legislator groups). The response rate for the two treatment groups and control group are as follows: 25.2% for the undocumented constituent letter; 31.1% for the citizen constituent letter; and 33.2% for the control group. Table 3.3, which shows these mean response values as well as confidence intervals, does not take legislator ethnicity into account. The 8 percentage point difference between the control group and the undocumented constituent treatment offers support for the differential citizen-undocumented response hypothesis, as does the 5.9 percentage point difference between the undocumented constituent and the citizen constituent treatments (see Table 3.3; both of these differences in means are statistically significant at p < 0.01). Given that the undocumented treatment group was responded 21 In the dataset used here, however, this may not be a problem. The correlation between Latino legislator (coded 1 if Latino; and 0 otherwise) and party (coded 1 if Democrat and 0 otherwise) is only 0.26 and the correlation between Latino legislator and legislators representing districts 50 percent Latino or higher (1=yes; 0=no) is only 0.64. 61 to less frequently than the citizen treatment group and the control group, the results suggest that legislators generally have less of an incentive to reply to an undocumented Latina. When the constituent was identified as an undocumented immigrant, only 25.2% of legislators replied. As a whole, the legislators were more likely to reply when the constituent was not identified as an undocumented immigrant. 22 Table 3.3: Legislator Responsiveness to Undocumented and Citizen Constituents Undocumented Treatment Citizen Treatment Control Group (national status not mentioned) % of legislators responding to each constituent group (n=1871) 25.2% [22.3, 28.0] 31.1% [28.0, 34.2] 33.2% [30.1, 36.3] Difference in means between undocumented treat group N/A +5.9* +8.0* *p<0.01, one-tailed difference of means test. Confidence intervals for each group displayed in brackets. However, the results for the undocumented letter are surprisingly higher than one might assume given the lack of a direct electoral reward. One-quarter of all legislators responded to a person who cannot vote for them, which is fairly remarkable. Legislators are responding to constituents even when there is little electoral motivation, suggesting that reelection is not the only factor affecting legislator representational decisions. What makes this even more surprising is this was an election year when 13 of the 14 states featured state legislative elections. I read and coded the emails to make sure the responses were helpful and not hostile. Only two responses of the 558 total responses were not helpful. Thus, we can interpret the choice by the legislator to respond to the constituent as helpful. The nature of the legislative responses were 22 In Appendix B, I estimated an OLS model with legislator response as the dependent variable (results are similar to those presented as crosstabs in Table 1. The model shows that the citizen group and the control group were responded to at higher rates than the undocumented group. These results hold when dummy variables for state fixed effects are included. 62 also generally benign, as they legislators either offered links to financial aid websites, encouraged Maria to contact them with more information, or recommended she speak to someone at a local college for help. 10. The Conditional Effect of Legislator Ethnicity As for the hypothesis that Latino legislators will be as responsive to the undocumented constituent as other constituents; while non-Latino legislators will be less responsive to the undocumented constituent relative to other constituents, I find support (see Tables 3.4 and 3.5). Table 3.4 shows the response rates to the different constituents for non-Latino legislators. Among those legislators who are not Latino, the response rate was 24.9% to the undocumented constituent; 32.3% for the citizen letter; and 33.0% for the control group. This difference in response between the undocumented treatment and the control group is statistically significant and substantively quite large. Legislators who are not Latino were over 7 percentage points less likely to respond to the undocumented constituent compared to the citizen constituent. Table 3.4: Non-Latino Legislators Are Less Likely to Respond to Undocumented Constituents Undocumented Treatment Citizen Treatment Control Group (national status not mentioned) % of non-Latino legislators responding to each constituent group (n=1871) 24.9% [21.8, 27.9] 32.3% [29.0, 35.6] 33.0% [29.8, 36.3] Difference in means between undocumented treatment group N/A +7.4* +8.1* *p<0.01, one-tailed difference of means test. Confidence intervals for each group displayed in brackets Table 3.5 shows the results for Latino legislators only. Statistically, Latino legislators are as likely to respond to the undocumented constituent as they are to the citizen constituent and the control group. The response rate for the Latino representatives is 27.5% for the undocumented 63 constituent treatment; 20.0% for the citizen treatment letter; and 33.8% for the control group. Even though not statistically significant, Latino representatives responded to the undocumented constituent at a higher rate than to the citizen constituent, though at a lower rate than to the control group. The evidence in Table 3.5 suggests that Latino legislators were as likely to respond to an undocumented person as another constituent. Additional support for the conditional role of ethnicity can be found by taking the difference-in-difference between non-Latino and Latino legislators for each group, which can be found in Table 4. As reported in the table below, non- Latino legislators responded to the citizen treatment relative to the undocumented treatment at a rate of +7.4 percentage points; non-Latino legislators responded to the citizen treatment relative to the undocumented treatment at a rate of -7.5 percentage points. As shown in the bottom right cell of Table 4, the difference-in-difference is thus 14.9 percentage points, which is a statistically significant and substantively large effect. This provides support for the hypothesis of greater response to citizens conditional on legislator ethnicity. The difference-in-difference, however, between Latino and non-Latino legislators’ response rates to the control group (not shown in Table 3.6) and the undocumented group was not statistically significant), suggesting that only when the Latina constituent noted that she was a citizen did the non-Latino legislators respond in greater numbers. In general, the results in Tables 3.3, 3.4, and 3.5 suggest that Latino legislators were more likely than non-Latino legislators to respond equally to both undocumented and citizen constituent requests. To provide additional evidence of this, I also estimated a regression model. 23 The dependent variable was 1 if the legislator responded and 0 if not. The independent 23 These models were also estimated as logit models, and the effects were largely similar. 64 variables are the undocumented treatment (coded 1 if in this treatment group; 0 otherwise); and the citizen treatment (coded 1 if in this group; 0 otherwise); an independent variable indicating if the legislator is Latino (1 for Latino; 0 if not); an interaction variable undocumented treatment x Latino legislator; and an interaction variable citizen treatment x Latino legislator. This OLS model is displayed in Table 3.7, and provides support consistent with Tables 3.3 and 3.4. Table 3.5: Latino Legislators Are as Likely to Respond to Undocumented Constituents as other Constituents Undocumented Treatment Citizen Treatment Control Group (national status not mentioned) % of non-Latino legislators responding to each constituent group (n=1871) 27.5% [19.1, 35.8] 20.0% [11.2, 28.7] 33.8% [23.5, 44.2] Difference in means between undocumented treatment group N/A -7.5 +6.4 Note: Difference of means are not statistically significant, implying no difference across groups. Confidence intervals for each group displayed in brackets. In general, the results in Tables 3.3, 3.4, and 3.5 suggest that Latino legislators were more likely than non-Latino legislators to respond equally to both undocumented and citizen constituent requests. To provide additional evidence of this, I also estimated a regression model. 24 The dependent variable was 1 if the legislator responded and 0 if not. The independent variables are the undocumented treatment (coded 1 if in this treatment group; 0 otherwise); and the citizen treatment (coded 1 if in this group; 0 otherwise); an independent variable indicating if the legislator is Latino (1 for Latino; 0 if not); an interaction variable undocumented treatment x 24 These models were also estimated as logit models, and the effects were largely similar. 65 Latino legislator; and an interaction variable citizen treatment x Latino legislator. This OLS model is displayed in Table 3.7, and provides support consistent with Tables 3.3 and 3.4. Table 3.6: Difference-in-difference between response rates to the undocumented vs. citizen constituents for Latino legislators and non-Latino legislators Undocumented Treatment Citizen Treatment Difference of Undoc. Treatment minus Citizen Treatment % of Latino legislators responding to each constituent group (n=199) 27.5% 20.0% 7.5% % of non-Latino legislators responding to each constituent group (n=1,678) 24.9% 32.3% -7.4%* Difference +2.6 -12.3 14.9* (diff.-in-diff.) P<0.05,one-tailed difference-in-difference test. The results presented in the robustness checks estimations in Table 3.7 largely confirm my findings. Latino population is not found to be a significant factor in determining whether legislators respond to the undocumented immigrant. There is a conditional effect related to Democratic legislators. Democrats are more likely to respond to the citizen treatment than the undocumented treatment, but this positive result is not found when comparing the control group variable and the undocumented variable for Democrats. 66 Table 3.7: Robustness Checks: Regression Models with Interactive/Conditional Variables Independent Variables Coef. (s.e.) Coef. (s.e.) Coef. (s.e.) Undocumented Treatment -0.083 (0.027)** -0.130(0.032)*** -0.083 (0.027)** Citizen Treatment -0.007 (0.027) -0.061 (0.035) -0.008 (0.027) Undoc. treat. x Latino legislator 0.028 (0.082) -0.015 (0.084) 0.028 (0.082) Citizen treat. x Latino legislator -0.131 (0.088) -0.181 (0.090)* -0.128 (0.088) Latino legislator 0.007 (0.062) 0.057 (0.064) 0.028 (0.068) Undoc. treat. X Dem. leg. -------------------- 0.104 (0.037)** -------------------- Citizen treat. x Dem. leg. -------------------- 0.120 (0.053)* -------------------- Democratic Legislator -------------------- -0.102 (0.037)** -------------------- Latino Voting Age Population (VAP) -------------------- --------------------- -0.001 (0.001) Constant 0.330 (0.019)*** 0.376 (0.025)*** 0.339 (0.022) Note: Reference category is undocumented treatment group. N=1871 ***p<0.001 **p<0.01; *p<0.05 Non-Latino legislators are still more likely to respond to the citizen treatment than the undocumented treatment in the additional models (evidenced by the negative and statistically significant coefficients on the undocumented treatment variable in Table 3.7, as the control group is the reference category); and Latino legislators do not differentiate between undocumented constituents and the other constituent groups in terms of overall response rates even when also conditioning on the number of legislative staffers and legislative professionalism. Legislative professionalism as measured by the Squire Index was found to be statistically insignificant at the 0.05 level. Legislative staff was also found to be statistically insignificant. 67 Table 3.8: Robustness Checks: Regression Models with Interactive/Conditional Models Independent Variables Coef. (s.e.) Coef. (s.e.) Coef. (s.e.) Undocumented Treatment -0.127 (0.032)*** -0.083(0.027)** -0.082 (0.273)** Citizen Treatment -0.060 (0.035) -.007 (0.027) -0.008 (0.027) Undoc. Treat. x Latino legislator -0.013 (0.084) 0.028 (0.082) 0.028 (0.082) Citizen Treat. x Latino legislator -0.179 (0.090)* -0.130 (0.082) -0.131 (0.087) Latino legislator 0.061 (0.069) 0.006 (0.062) 0.001 (0.062) Undoc. Treatment x Dem. Leg 0.102 (0.038)** ------------------ ----------------------- Citizen treatment x Dem. Leg 0.119 (0.053)* ------------------ ----------------------- Democratic Legislator -0.099 (0.037)** ------------------ ----------------------- Latino VAP -0.003 (0.001) ------------------ ----------------------- Squire Index (2007) --------------------- 0.114 (0.071) ----------------------- Legislative Staffer --------------------- ------------------ 0.003 (0.002) Constant 0.380 (0.027)*** 0.303 (0.025)*** 0.302 (0.026) Note: Reference category is undocumented treatment group. N=1871 ***p<0.001 **p<0.01; *p<0.05 11. Discussion and Implications There are significant theoretical implications of these findings and important avenues for future research. The fact that Latino legislators responded at statistically similar rates to all constituents suggests that Latino legislators do indeed have a different relationship than non- Latino legislators to undocumented Latinos – and that this relationship could be mediated by a sense of intersectional linked fate. These results suggest that at least among the Latino legislators in these states, undocumented Latinos can expect to be represented at a level similar to Latino citizens. The lack of marginalization of undocumented constituents among Latino legislators is only underscored by the evidence of the lower response to undocumented constituents by non- 68 Latino legislators. Non-Latino legislators were less responsive to the undocumented Latina than to the citizen Latina, suggesting that undocumented Latinos are marginalized by their lack of legal status among most legislators. The intersectional location of an undocumented Latina results in heterogeneous treatment (compared to a citizen Latina) by most elected legislators, as the large majority of legislators are not Latino. While I theorized that the primary explanation for the lack of response differential to undocumented and citizen constituents among Latino legislators is due to an intersectional linked fate, this cannot be definitively proven, as there may also be indirect incentives, electoral and otherwise. One of the reasons for this might be that the sample of states with large Latino populations have large foreign-born populations. This might make legislators of all backgrounds (and particularly Latino legislators from majority-Latino districts) aware of indirect electoral consequences for their actions regarding undocumented Latinos. Legislators might not want to risk coming off as insensitive to the needs of an undocumented Latina as it might influence how potential voters, possibly friends and family members who are citizens, would view the incumbent. Not wanting to alienate anyone, legislators might have an incentive to respond. On the other hand, the request to the legislator is an individual email, and it is unlikely the response would be shared widely with friends and family who are also voters (Wong and Tseng 2008). It is also possible that constituency work offers its own intrinsic rewards for Latino legislators, a possibility that fits logically with other work on minority populations and constituency case work (e.g., Broockman 2013; Grose 2011; Tate 2003). Surprisingly, the results also suggest that undocumented Latinos might not be as politically disadvantaged as we might have previously assumed because of their inability to vote 69 (Ramakrishnan 2005; Ramakrishnan and Espenshade 2001; Wong 2006). The fact that about one-quarter of legislators responded to an undocumented immigrant is fascinating evidence that those in the shadows of legality may be able to achieve some level of political representation. The surprisingly high response to the undocumented constituents suggests that some legislators are representing the undocumented Latino population. It could be a matter of context for this specific sample of states, as issues regarding undocumented Latinos, and undocumented immigrants in general, are frequently debated and voted on in these states. This creates a sense of familiarity with the issues of this undocumented population and might make lawmakers more sensitive to their needs and concerns. It may also be due to the fact that the representational decision was essentially a response to a constituency service request, and not a decision on public policy. There is also the question of what role an election year contest played when immigration reform was a hot button issue and saw the two political parties take divergent positions, with President Obama in support of undocumented immigrants, and the Republican field often endorsing conservative positions. This was the same election that saw heavy activism by the undocumented youth activists, the DREAMers, to both pressure President Obama and Members of Congress. While this is beyond the scope of this dissertation, and its methods, it is still something worth considering. In addition, states with histories of Latino incorporation and political empowerment might be able to rally immigrant interest groups, and/or Latino interest groups, to put pressure on state legislators to support proposals in favor of undocumented Latinos. This activism on behalf of undocumented immigrants has been heightened at the state level since the U.S. Congress has failed to pass comprehensive immigration reform, a federal DREAM Act, forcing immigrant rights activists to take the fight for immigrant rights to the states. These states have also 70 experienced a dramatic increase in immigration bills since there has no been action on the federal level. This could trickle down to constituency service from legislators as well. These examples suggest that perhaps undocumented Latino immigrants can count on anticipatory representation to provide them service responsiveness, as legislators might learn about their issues in the manner described above. On the other hand, the states in the sample examined include those with recent histories of exclusionary policies toward undocumented immigrants (e.g., Arizona). My findings were tested using an undocumented Latina constituent alias. It is unknown whether these findings would extend to undocumented immigrants of all ethnic backgrounds because this article only focused on undocumented Latino immigrants. It would be advised that a future research project examine similar questions focused on undocumented immigrants of non- Latino ancestry, such as those of Asian, European, or Middle Eastern ancestry, to further understand how undocumented immigrants are represented. Further research in this vein should be able to examine whether the Latina constituent would have received more assistance from legislators than a Latino male due to the former being seen as less threatening, or more sympathetic (see Bejerano 2013). Scholars often claim that political marginalization is caused by the inability for groups who lack voting power or rights to influence elected officials and their behavior. Some legislators did, however, in fact respond to an undocumented immigrant as much, in the case of Latino legislators, as the control group, this raises interesting theoretical implications about representation. This suggests that the controversial space in which this group inhabits may not necessarily inhibit representation, which leads to a need to reformulate exactly how scholars think about this group’s marginalization in the political sphere. It also suggests that scholars of representation should more broadly consider how and whether those that are barred from voting 71 are represented by elected officials. This study also points to the need to continue to examine gyroscopic representation (Mansbridge 2003), and how private motivations can influence legislators to act when there are few rewards. The theory and findings have presented new questions to scholars of representation, racial and ethnic politics, and legislative politics regarding who is perceived as a constituent, and what determines political representation for some groups. When going forward, it is important to ask what factors can make legislators consider the needs of groups that are considered politically marginalized. The study also demonstrated the limits of quantitative studies and their incompatibility with intersectionality, and offers support for Hancock’s proposition that we should think of such studies as an analysis of “multiple identities”. There are still questions that remain, such as, are undocumented Latinos really responded to less only because they are undocumented, or because they are undocumented Latinos? The next chapter will further explore this question. 12. Conclusion The evidence consistently shows a differential in response by non-Latino legislators and no differential in response by Latino legislators. Latino legislators appear to respond to Latina constituents regardless of national status identification, while non-Latino legislators respond more frequently to citizen Latinas than to undocumented Latinas. However, a surprisingly large percentage of legislators responded to the undocumented request, also suggesting that a populace with no electoral power and few legal rights, nevertheless receives some representational responsiveness. In sum, I conducted a field experiment on state legislators in 14 states with significant Latino populations to test whether Latino legislators are more likely to represent undocumented 72 Latinos than non-Latino legislators. This was the first field experiment ever conducted to determine whether legislators will respond to undocumented immigrants, and it is one of the few field experiments ever conducted on elected officials. It is also one of the only studies to ever examine the representation of undocumented immigrants. The main findings are that Latino legislators will respond at a similar rate statistically to a co-ethnic constituent regardless of national status, but that constituent national status does affect whether a non-Latino legislator will respond to a Latino constituent. 73 Chapter 4: Disentangling Ethnicity and National Status 1. Introduction The previous chapter included an experiment that found evidence for the concept of intersectional linked fate among Latino legislators. It also found that despite the lack of a direct electoral reward, non-Latino legislators still provided responsiveness to undocumented immigrants albeit at a lower rate than to a Latina citizen. These findings help fill in the gap about representation concerning undocumented immigrants, but also had some limitations related to sample size and the influence of particular identities. This chapter will feature an experiment that is a replication of the study from the last chapter. This replication is concerned with the questions of whether Latino legislators would provide equitable responsiveness to undocumented Latinos, possibly out of a sense of intersectional linked fate. It is also looked to examine whether legislators would provide any responsiveness to undocumented Latino immigrants. The previous experiment both provided some answers to those questions, but also had some important limitations. First, the sample, which consisted of 14 states with Latino populations over 10%, was too limited to account for multiple interaction effects. Second, the previous experiment only allowed for the testing of the influence of one social identity, national status, whereas this experiment will allow for the examination of ethnicity and national status together. 2. Building on the Previous Experiment The previous experiment showed that legislators do respond to an undocumented Latina immigrant, but this experiment did not allow me to determine the role of ethnicity either in tandem with national status, or apart, due to limitations related to sample size. This experiment will allow me to attend to Categorical Multiplicity, and ask, what categories are relevant to my 74 research question, which is, what might determine whether or not a legislator responds to an undocumented immigrant (Hancock 2013). Using Hancock’s (2012) criterion for category selection, I ask, what are the inegalitarian traditions that lead to my research question? Ethnicity, and more specifically, the process of racialization that accompanies it, is important because of the fact that undocumented immigration is a highly racialized political issue that often features anti-Latino rhetoric (Johnson 1997; Garcia Bedolla 2005; Martinez 2008; Chavez 2011). Latinos make up the largest group of undocumented immigrants, and are also the most frequent targets of punitive immigration policies. The issue is so regularly thought of in racial terms that undocumented immigrants of other racial groups, such as Asians and Eastern Europeans, have been found to think that the issue does not affect them (Zepeda-Millan 2014). Thus, it is important to consider the role that ethnicity and national status has together in determining legislator responsiveness to an undocumented immigrant. The second experiment will build on the first by incorporating an Eastern European alias in order to test whether the intersectional identity influences legislator responsiveness. This experiment follows the Multiple Quantitative Formulation and asks whether the constituent’s ethnicity, national status, or ethnicity and national status has the strongest effect on the outcome variable, which in this case is the response rate (Hancock 2013, 269). This experiment is still an example of the “intersectionality-as-testable-hypothesis” since it examines identity influence, but it will help to clarify a few lingering questions from the previous experiments. 25 The last experiment left open the question of whether Latino legislators are just more responsive to undocumented immigrants regardless of ethnicity. This experiment will finally 25 Some of the limitations of this approach in regards to intersectionality theory include its assumption of the static nature of the categories of the identities involved, and that these categories can be disaggregated. This approach also “assumes that there is a single causal combination of factors” for a particular phenomenon and that this does not allow for proper attention to the different domains of intersectionality (Hancock 2013, 277). 75 engage with categorical multiplicity in regards to the constituents as ethnicity and national status will both be analyzed as they are both relevant. There is also the lingering question of the actual effect of ethnicity. While an undocumented immigrant of any ethnicity offers little in the way of a direct electoral reward, it is still important to test a Latina alias against an Eastern European alias in order to assess responsiveness to the intersectional identities. There is also the possible time element to account for, as the first experiment occurred in the summer 2012 of a Presidential election year where issues surrounding undocumented immigrants, especially that of undocumented youth, who were especially active in targeting President Obama and Republican candidates. For the previous experiment, the summer might have also served as downtime for legislators, allowing them to answer emails. This experiment occurred mere weeks before Election Day in an ultimately low turnout but highly contested midterm Congressional election. Could this dynamic also shape response rates? 3. Rationale for an Eastern European Alias An Eastern European alias was chosen for the second alias because of the need to pick an alias that was ostensibly White, and could realistically pass for an undocumented immigrant. It is important to have them be racially White as Whiteness comes with privileges that are not available to people of color in the United States (McIntosh 1989). Eastern Europeans currently account for 44% of all European immigrants (Russell and Batalova 2012), and are largely believed to make up the majority of European undocumented immigrants (DuMontier 2014), although scholarship on undocumented European immigrants is quite limited. Most undocumented Eastern Europeans are believed to have arrived here on tourist visas but then stayed and worked without authorization in the United States (Kole 2003; DuMontier 2014). 76 Due to the diversity of cultures and languages of Eastern Europe, a specific nationality needed to be chosen, and in this case, the Russian nationality was chosen because they are one of the largest groups of Eastern European immigrants (Russell and Batalova 2012). The selection of a Russian nationality was done to ensure consistency among surnames for the Eastern European alias. Russians also have one of the younger populations of the Eastern European immigrant (EEI) groups in the United States, with more young adults among the first-generation than most other EEI group (Rumbaut 2004; Kasinitz, Mollenkopf, and Waters 2010). The relative youth of the Latino population, both immigrants and native-born, makes it necessary to have a comparison group that features a, relatively, substantial number of young people. 26 4. Hypotheses The hypotheses for this experiment are roughly the same as for the first experiment, and so are the expectations. There is the direct electoral goals hypothesis, which posits that legislators are likely to not respond at all, and the differential citizen-undocumented responsiveness hypothesis, which holds that legislators will respond to the undocumented immigrant, but they will do so less than to the citizen. Legislators are assumed to be engaging in vote-seeking behavior, and as such, will seek to prioritize the citizens over the undocumented immigrants regardless of ethnicity. 5. Research Design This replication study follows the same basic tenets of the first experiment, in that legislators are the subjects of analysis, and then legislators were then block randomized into the different treatment groups. There are two distinct constituent aliases, a Latina alias and an 26 Eastern European immigrants are on average older than Latino immigrants (Robila 2013) and also exhibit more variation among surnames. The Russian language is part of a larger family of historically related, but distinct, Slavic languages located largely, but not exclusively, in Eastern Europe (i.e., Sussex and Cubberley 2006). For this reason, I feel justified in having a Russian name for the Eastern European alias. 77 Eastern European alias, 27 and three different messages used for each alias, meaning that there are six treatment groups in total. Specifically, there is a treatment that explicitly designates the constituent as an undocumented immigrant; another treatment that specifically designates the constituent as a citizen; and a control group that does not mention national status, and instead uses the term resident. Ethnicity will be signaled by the name of the constituent. The content of the message is the same with the only differences being the legal status and ethnicity of the constituent. Table 4.1: Constituent Treatments Treatment 1 Treatment 2 Treatment 3 Treatment 4 Treatment 5 Treatment 6 Undocumented Latina Undocumented Eastern European Citizen Latina Citizen Eastern European Control Group Latina Control Group Eastern European For all six treatment groups, the constituent wrote the legislator with a simple request asking only for information about early childhood education programs on behalf of her sister (see more below). The content of the messages was the same except for the variation of the ethnicity and national status of the constituent. Like the previous experiment, the messages to legislators from constituents did not require much in the way of legislator (or staff) time and resources, and carried no political risk for the legislators. Early childhood education programs were made the subject of the requests because undocumented immigrants are eligible for enrollment in public education programs. This is a legacy of the Plyer v. Doe Supreme Court ruling, which ruled in favor of undocumented 27 This experiment also differs from the first for it uses multiple aliases for both the Latina constituent and the Eastern European. The Latino constituent aliases include Veronica Hernandez, Juana Martinez, and Margarita Lopez. These names were all selected from among the top ten Spanish surnames in the United States according to the U.S. census. The first names were selected from a list of Spanish first names for girls. The Russian constituent aliases include Anastasia Popov, Anna Smirnov, and Daria Ivanov. The names were randomly selected from a list of Russian surnames and a list of Russian girls names. Butler (2014, 93-94) used multiple aliases, in addition to other randomization schemes, to lessen the likelihood of contamination. 78 immigrant children being able to attend public schools (Bosniak 2008). Even the Head Start program, an early education program funded by the federal government, does not permit employees to ask about the national status of the children attending or their parents, even though undocumented immigrants are technically barred from receiving any benefits from the federal government (Matthews and Ewen 2010). Furthermore, 46 U.S. states in total have publicly funded early childhood education programs (Pre-K funding 2014). The need to accommodate the additional alias and its three treatment groups meant that the sample population had to be expanded to include state legislators from all states with publicly available e-mail addresses. This means that the experiment will have an N of 6,082 state legislators from 44 states and 85 state legislative chambers. 28 Each of these legislators received a request from a constituent asking about information on early childhood education on behalf of her sister. The exact text of these treatments can be found below. The emails were sent out a couple of weeks before Election Day November 2014. The timing is different from the first experiment, as it occurs during a timeframe when electoral incentives are at there most salient for the legislators who are most likely facing reelection. It is 28 Only six states are missing from this data set due to lack of publicly available email addresses. They are Idaho, Kentucky, Ohio, Oregon, South Carolina, and South Dakota. There are 89 chambers because Nebraska is has unicameral government, and only one chamber had email addresses publicly available for Michigan (House) and Pennsylvania (Senate). Dear Rep/Senator/Assembly member _______, My name is [constituent name] and I asking on behalf of my sister for information on state programs for early childhood education. As citizens in your area we are hoping you can help us. Thank you for any help you are able to provide. Sincerely, [constituent name] 79 also likely that they are monitoring the reelection efforts of their fellow elected officials at the state and national levels. Like the first experiment, the results will be analyzed for significance through one-tailed difference of means tests. In this section, the analysis will concern all legislators. Critical intervals at the 95% level serve as a robustness check to evaluate the significance of the mean point estimates as overlapping values should indicate that there is no meaningful difference between the response rates. OLS regressions were run as an additional robustness check, and featured models that just included the treatments, as well as other models that incorporated variables for political party and state dummy variables. 6. Results Of the 6,082 state legislators in the sample, 1,309 sent email responses to constituents, meaning there was a 21.5% response rate to all six legislator groups. 29 The response rate for the treatment groups were as follows: 16.2% responded to the undocumented Latina; 15.2% responded to the undocumented Eastern European; 27.3% responded to the Latina U.S. citizen; 22.2% responded to the Eastern European U.S. citizen; 19.05% responded to the Latina control group; and 29.1% responded to the Eastern European control group. Table 4.2 below shows the mean point estimates with the critical values. Table 4.2 also shows that legislators responded more to the Latina citizen than to the undocumented Latina immigrant by a difference of 11.1 percentage points, and this gap is statistically significant at the 0.001 level. The difference between the undocumented Latina immigrant and the Latina control group is a much smaller 2.85 percentage points, a result that is found significant at the 0.05 level. More legislators responded to the Latina citizen than to the 29 I received emails from six individual legislators from the states of Connecticut, Delaware, Indiana, Missouri, New Hampshire, and Washington indicating that spillover might have occurred. The states are still included in this analysis but might possibly be removed in some future iteration of this study. 80 Latina control by 8.25 percentage points, a finding that was significant at the 0.01 level. These results are partially consistent with the first experiment and offers support for the differential citizen-undocumented responsiveness hypothesis as they responded more to the citizen than to the undocumented immigrant by a large gap. They diverge in terms of the citizen treatment receiving greater responsiveness than the control group at a statistically significant level. For the Eastern European constituent, the legislators responded more to the Eastern European citizen than to the undocumented Eastern European immigrant by 7 percentage points. This finding was statistically significant at the 0.001 level. The difference in response rates between the Eastern European undocumented immigrant and the control group is 13.9 percentage points, which was also found significant at the 0.01 level. These findings are also mostly consistent with the first experiment and the differential citizen-undocumented responsiveness hypothesis. Table 4.3 below shows the mean estimates with the critical values for the Eastern European immigrant. Table 4.2: Legislator Responsiveness to Constituent Emails as Compared to the Undocumented Latina Undoc. Latina Undoc. East Euro. Latina Citizen East Euro. Citizen Latina Control East Euro. Control % of legislators responding (n=6,082) 16.2% [14.3,18.2) 15.2% [13.4,17.2] 27.3% [25.0,29.7] 22.2% [20.1,24.4] 19.05% [17.0,21.2] 29.1% [26.7,31.5] Difference in means between undoc. Latina N/A -1.0 +11.1*** +6.0** +2.85* +12.9*** Note: ***p<0.001, one-tailed difference of means test. Confidence intervals for each group displayed in brackets. **p<0.01 *p<0.05 81 The results in Table 4.3 demonstrate that national status is indeed salient as all legislators responded less to the undocumented immigrant constituent of both ethnic groups; however, at least 15% of legislators responded to the undocumented immigrant constituents, allowing me to reject the direct electoral goals hypothesis. The results show that legislators do respond to the citizen treatments more than they do to the undocumented treatments, offering much support for the differential citizen-undocumented responsiveness hypothesis. One of the control groups was responded to more at statistically significant level than the undocumented treatments and it is not clear as to why this occurred. The robustness checks largely confirm these findings with only the difference between the undocumented Latina constituent and the Latina control group being called into question. While the one-tailed difference of means test found the differential to be significant, the OLS regression found it to be insignificant. Additionally, the critical intervals overlapped, suggesting that it is not significant. Now that I have demonstrated the salience of national identity, the question now turns to what role ethnicity plays in service responsiveness. There is no statistical difference between the undocumented Latina and the undocumented Eastern European immigrant, but a difference does manifest in the other treatment pairings. Legislators are more like to respond to the Latina citizen than to the Eastern European citizen by 5.1 percentage points, a finding that is statistically significant at the 0.01 level. However, they are much more likely to respond to the Eastern European control than to the Latino control by 10.85 percentage points, a finding that is statistically significant at the 0.001 level. What this means remains unclear, but it could have to do with how legislators interpret the interaction of the ethnicity and the terms used to describe the constituent’s residency. It is possible that when left unaddressed, the respondents are making an association between the Eastern European name and having citizenship, an association they 82 may not be making for the Latina whose national status goes unaddressed. It does appear that, on balance, ethnicity of the constituent does not appear to wholly offer an advantage or disadvantage to either group, making it appear that national status is the most salient category of difference. 30 The legislators might be engaging in a different thought process when it comes to how they perceive the association of each ethnicity with the terms used to denote national status, or lack thereof. 31 Table 4.3: Legislator Responsiveness to Undocumented and Citizen Email as Compared to Eastern European Undocumented Immigrant Undoc. East. Euro. Undoc. Latina East. Euro. Citizen Latina Citizen East. Euro. Control Latina Control % of legislators responding (n=6,082) 15.2% [13.4, 17.2] 16.2% [14.3, 18.2] 22.2% [20.1, 24.4] 27.3% [25.2, 29.4] 29.1% [26.9, 31.2] 19.05% [17.0,21.2] Difference in means between undoc. East. Euro. N/A +1.0 +7.0* +11.1** +13.9** +3.85* Note: **p<0.001, one-tailed difference of means test. Confidence intervals for each group displayed in brackets. *p<0.05, one-tailed difference of means test. 7. Discussion Overall, the experiment shows that legislators do respond less to the undocumented immigrant, regardless of ethnicity. This experiment took place in October 2014, a few weeks before a national midterm election, that also saw all of the states in the sample hold legislative 30 An OLS regression in Appendix C2 largely confirms these findings. In Appendix C3, there is also an OLS regression that features the treatments with state dummy variables; they did not have an independent effect on the results. 31 A robustness check was run with a party variable (Democrat=1) and the following party interaction variables: party_x_undocumented_Latina; party_x_Eastern_European_citizen; party_x_Latina_citizen; party_x_Eastern_European_citizen; party_x_Latina_control; party_x_Eastern_European_control. The party variables do not have an independent effect on the results. 83 elections for at least some seats in their chambers. It is reasonable to assume that many legislators were focused on electoral efforts and had themselves and their staff (if they lived in state that provided staff for state legislators) devoted to such efforts. The salience of the upcoming election might have made responding to all constituent emails a lower priority. Another finding worth discussing is the differential in responsiveness to the Latina and Eastern European constituent aliases across the different treatment groups. While the undocumented Latina and undocumented Eastern European were responded to at about the same rate by all legislators, ethnicity seemed to play a role in how legislators responded to the citizen and control groups. The Latina citizen was responded to more by all legislators by a statistically significant margin. The Eastern European control was responded to more by all legislators by a statistically significant margin than the Latina control. These results show that the ethnicity by itself does not necessarily advantage one group over another, but points to the potentially racialized meanings of pairing these terms with ethnic signifiers. These findings lead to interesting questions about the possible associations of the signifiers of national status and residence (the control referred to the constituent as a “resident” and the citizen treatment referred to the constituent explicitly as the “citizen”) with ethnicity. The association of the word “citizen” with the Eastern European alias might have made the legislator unsure of whether she actually lived in the district; the exact reason for this is unknown but it could have something to do with the comparatively small percentage of people in the United States who self-report Eastern European ancestry in the United States (7%). 32 Perhaps it was because an Eastern European surname by itself does not make a legislator think about who 32 This figure was calculated by adding together the raw numbers of the nationality groups that constitute Eastern Europe as reported by the 2009 American Community Survey. See, U.S. Census Bureau, 2009 American Community Survey, B04006, “People Reporting Ancestry,” http://factfinder.census.gov, January 2011. This might be an imperfect measure, but it matches the same percentage of self-reported Eastern European ancestry as reported by the 2000 U.S. Census. 84 exactly is writing her, but the “citizenship” marker might have given them pause for an unspecified reason. Conversely, the Latina citizen might have received more responses because of the relatively large size of the national Latino population (17.03%), and the fact that Latinos are an established and/or growing presence in many individual states. It is also possible that receiving an email from a Latina who only describes herself as a “resident” during such a critical time led legislators to question whether she was a citizen. If so, then some legislators might not have replied because they could not assumer her national status, and therefore her ability to vote, and decided to expend their resources on constituents they assumed could vote. If this is the case, then it potentially leads to questions about the racialization of Latinos and whether it influences how legislators view them. The opposite happened in regards to the control groups, where the Eastern European control was responded to more than the Latina control. This result was surprising, but the association with an Eastern European alias with the word “resident” might have signaled to the legislator that she was a voter. It is unclear as to what exactly occurred with the Latina control group, which was also responded to less at a statistically significant level less than the Latina citizen. These results are puzzling in that the Latina control group was statistically responded to at the same rate in the first experiment. One possible explanation is that the expanded sample of legislators included those who might have associated the Eastern European name and the word “resident” with someone who can vote. Perhaps it was because an Eastern European surname by itself does not make a legislator think about who exactly is writing her, but the “citizenship” marker might have given them pause for an unspecified reason. 85 8. The Conditional Effect of Legislator Ethnicity The previous experiment (see Ch. 3) found support for the conditional Latino hypothesis, which posits that Latinos will respond equally to a co-ethnic regardless of national status; however, this hypothesis is modified to fit this experiment by predicting that Latino legislators will respond at a lower rate to the Eastern European alias regardless of national status. This modified hypothesis is partially based on the recent findings by Butler (2014) that Latino legislators responded less to a White constituent alias than a Latino constituent alias. It is also consistent with the theory as Latino legislators should not feel a sense of intersectional linked fate with an Eastern European undocumented immigrant who does not share an ethnic identity with the Latino legislators. Non-Latino legislators are expected to respond less to the undocumented immigrant regardless of ethnicity than to the citizen. 9. Results: Do Latino Legislators Respond Differently? Table 4.4 shows that 14.5% of Latino legislators responded to the undocumented Latina; 28.5% of Latino legislators responded to the Latina citizen; and 10.8% responded to the Latino control group. The same table shows that Latino legislators respond to the undocumented Eastern European at a rate of 16.2%; to an Eastern European citizen at a rate of 16.3%; and to the Eastern European control group at a rate of 16.3%. There were 260 Latino legislators in the sample, and their overall response rate was 23.8%. As seen in Table 4.4, Latino legislators responded less to the undocumented Latina by 14 percentage points than the Latina citizen. This large differential was found to be statistically significant at the 0.05 level by the one-tailed difference of means test. The differential of 17 percentage points between the Latino citizen and the Latino control was found to be statistically significant at the 0.01 level. This is in direct contrast to the first experiment, which did not find 86 any statistically significant differences between the Latino legislator response rates to the Latina constituent in any of the statistical analyses. Table 4.4: Latino Legislator Responses to the Undocumented and Citizen Constituents as Compared to the Undocumented Latina Undoc. Latina Undoc. East. Euro. Latina Citizen East. Euro. Citizen Latina Control Group East. Euro. Control Group % of Latino legislators responding (n=260) 14.5% [7.0, 25.6] 16.2% [7.8, 28.4] 28.5% [18.8, 40.1] 16.3% [8.3, 27.5] 10.8% [3.7, 23.0] 29.6% [15.6, 47.1] Difference in means between undoc. Latina N/A +1.7 +14.0* +1.8 -3.7 +15.1 *p<0.05 level, one-tailed difference of means test. Confidence intervals included in brackets. The robustness checks, however, told a different story as the OLS regression only found the difference between the undocumented Latina and the Latina citizen to be significant the 0.05 level. The critical intervals for all three Latino legislator groups also overlap. The conflicting results mean that a definitive conclusion cannot be reached about the conditional Latino hypothesis, but the evidence does suggest that they did respond less to the undocumented Latina than the Latina citizen. Table 4.3 shows that the Latino legislators responded to the undocumented Eastern European and the Eastern European citizen at virtually the same rate. The Eastern European control group received the highest percentage of legislator responses, but its large differentials with the Eastern European undocumented immigrant (13.4 percentage points) and the Eastern European citizen (13.3 percentage points) were mostly found not to be statistically significant by any measurement. There was also no difference within the national status treatment groups based on ethnicity according to any of the measures. Nothing can be definitively concluded from these results about the effects of national status or ethnicity, either separately or together 87 There were 5,822 non-Latino legislators in the sample, and their overall response rate was 21.6%. The results for the non-Latino legislators largely mirrored the general trend found among all legislators. Non-Latino legislators responded to the undocumented Latina at a rate of 16.2%; to the Latina citizen at a rate of 28.5%; to the Latino control group at a rate of 19.3%; to the Easter European undocumented immigrant at a rate of 15.2%; to the Eastern European citizen at a rate of 22.5%; and to the Eastern European control at 29.1%. (Table 4.4 below has the mean point estimates and critical intervals). Non-Latino legislators responded to the Latina citizen more than the undocumented Latina by a differential of 11 percentage points, and this is statistically significant a the 0.0001 level. The differential between the undocumented Latina and the Latina control group is 3.1 percentage points, a finding which is found to be statistically significant at the 0.05 level according to a one-tailed difference of means test. The robustness checks largely confirm these results (See Appendix). Table 4.5: Non-Latino Legislators’ Responses to Undocumented and Citizen Constituents as Compared to the Undocumented Latina Immigrant Note: ***p<0.0001 level, one-tailed difference of means test. Confidence intervals in brackets. **p<0.001 *p<0.01 †p<0.05 Non-Latino legislators responded less to the Eastern European undocumented immigrant than to the Eastern European citizen by 7.3 percentage points, which was found significant by Undoc. Latina Undoc. East. Euro. Latina Citizen East. Euro. Citizen Latina Control Group East. Euro. Control Group % of non- Latino legislators responding (n=5, 822) 16.2% (14.3,18.3) 15.2% [13.3, 17.2] 27.2% [24.9, 29.6] 22.5% [20.3, 24.8] 19.3%% [17.2, 21.5] 29.1% [26.9, 31.2] Difference in means between undoc. Latina N/A -1.0 +11.0* +6.3** +3.1† +12.9*** 88 the one-tailed diff. of means test (p< 0.01). They responded more to the Eastern European control than to the undocumented Eastern European by 13.9 percentage points, a large differential that was statistically significant at the 0.001 level. The differential of 6.6 percentage points between the Eastern European citizen and the Eastern European control is also statistically significant at the 0.001. The robustness checks largely confirm these findings. These findings support the differential citizen-undocumented immigrant hypothesis as the legislators responded to the undocumented messages less regardless of the ethnicity of the constituent demonstrating the saliency of national status. Table 4.6: Non-Latino Legislator Responsiveness to Undocumented and Citizen Constituents as Compared to the Undocumented Eastern European Undoc. East. Euro. Undoc. Latina East. Euro. Citizen Latina Citizen East. Euro. Control Latina Control % of non- Latino Latino legislators responding (n=5,822) 15.2% [13.3, 17.2] 16.2% [14.3, 18.3] 22.5% [20.3, 24.8] 27.2% [24.9, 29.6] 29.1% [26.9, 31.2] 19.3% [17.2, 21.5] Difference in means between undoc. East. Euro. N/A +1.0 +7.3* +12.0* +13.9* +4.1** Note: *p<0.001, one-tailed difference of means test. Confidence intervals in brackets. **p<0.01 To examine the saliency of ethnicity the different national treatments will be examined in pairs. There is no statistical difference between the two undocumented treatments. For the two citizen treatments, non-Latino legislators responded more to the Latina citizen than to the Eastern European citizen by 4.7 percentage points, a finding that is statistically significant at the 0.01 level. Non-Latino legislators responded to the Eastern European control group more than the Latina control group by 9.8 percentage points, a magnitude that was found statistically significant at the 0.001 level by a one-tailed diff. of means test. Ethnicity does matter within the 89 pairs, but on the whole, it does not advantage one ethnic group over another. Additional analyses were run to determine whether the difference-in-difference between the Latino and non-Latino response rates to the treatment groups were statistically significant. They were not significant, which indicates that ethnicity was not more salient to the Latino legislators than to the Non- Latino legislators. 10. Discussion This experiment and its results demonstrate that national status is a much more salient category for legislators than ethnicity when it comes to responding to constituents. This also appears hold true for Latino legislators, who do not show the same consistency in responsiveness across the statistical analyses that they did in the previous experiment. While the one-tailed diff. of means test and the OLS regression robustness check bear this out, the critical values do not, but even with that I feel justified in interpreting the difference to be significant. One possible explanation why the critical values do not register the difference in responsiveness between the undocumented Latina and the Latina citizen is that the sample size might have been too small to consistently capture the effects of the national status. The fact that the Latino legislator response rates seem to mirror the general legislator response rates at least in the case of the Latina legislator groups, also lends weight to the argument that Latinos are in fact responding more to the Latina citizen than the Latina undocumented, and even the Latina control. There were not any statistically significant differences between the responsiveness of Latino legislators to the different ethnicities, although the gap of 12 percentage points between the Eastern European citizen and the Latina citizen might point to a bias that Latino legislators 90 have towards a co-ethnic who is a citizen. 33 However, this result is possibly negated by the fact that Latino legislators seemed to respond more to the Eastern European control than to the Latina control, mirroring the results found for all legislators. Overall, the experiment shows that legislators do respond less to the undocumented immigrant, regardless of ethnicity. This experiment took place in October 2014, a few weeks before a national midterm election, that also saw all of the states in the sample hold legislative elections for at least some seats in their chambers. It is reasonable to assume that many legislators were focused on electoral efforts and had themselves and their staff (if they lived in state that provided staff for state legislators) devoted to such efforts. The salience of the upcoming election might have made responding to all constituent emails a lower priority. Another finding worth discussing is the differential in responsiveness to the Latina and Eastern European constituent aliases across the different treatment groups. While the undocumented Latina and undocumented Eastern European were responded to at about the same rate by all legislators, ethnicity seemed to play a role in how legislators responded to the citizen and control groups. The Latina citizen was responded to more by all legislators by a statistically significant margin. The Eastern European control was responded to more by all legislators by a statistically significant margin than the Latina control. These results show that the ethnicity by itself does not necessarily advantage one group over another, but points to the potentially racialized meanings of pairing these terms with ethnic signifiers. These findings lead to interesting questions about the possible associations of the signifiers of national status and residence (the control referred to the constituent as a “resident” and the citizen treatment referred to the constituent explicitly as the “citizen”) with ethnicity. The 33 If so, then this would be consistent with Butler’s (2014) finding that Latino legislators have an in-group bias toward Latino constituents. 91 association of the word “citizen” with the Eastern European alias might have made the legislator unsure of whether she actually lived in the district; the exact reason for this is unknown but it could have something to do with the comparatively small percentage of people in the United States who self-report Eastern European ancestry in the United States (7%). 34 Perhaps it was because an Eastern European surname by itself does not make a legislator think about who exactly is writing her, but the “citizenship” marker might have given them pause for an unspecified reason. Conversely, the Latina citizen might have received more responses because of the relatively large size of the national Latino population (17.03%), and the fact that Latinos are an established and/or growing presence in many individual states. It is also possible that receiving an email from a Latina who only describes herself as a “resident” during such a critical time led legislators to question whether she was a citizen. If so, then some legislators might not have replied because they could not assumer her national status, and therefore her ability to vote, and decided to expend their resources on constituents they assumed could vote. If this is the case, then it potentially leads to questions about the racialization of Latinos and whether it influences how legislators view them. The opposite happened in regards to the control groups, where the Eastern European control was responded to more than the Latina control. This result was surprising, but the association with an Eastern European alias with the word “resident” might have signaled to the legislator that she was a voter. It is unclear as to what exactly occurred with the Latina control group, which was also responded to less at a statistically significant level less than the Latina citizen. These results are puzzling in that the Latina control group was statistically responded to 34 This figure was calculated by adding together the raw numbers of the nationality groups that constitute Eastern Europe as reported by the 2009 American Community Survey. See, U.S. Census Bureau, 2009 American Community Survey, B04006, “People Reporting Ancestry,” http://factfinder.census.gov, January 2011. This might be an imperfect measure, but it matches the same percentage of self-reported Eastern European ancestry as reported by the 2000 U.S. Census. 92 at the same rate in the first experiment. One possible explanation is that the expanded sample of legislators included those who might have associated the Eastern European name and the word “resident” with someone who can vote. Perhaps it was because an Eastern European surname by itself does not make a legislator think about who exactly is writing her, but the “citizenship” marker might have given them pause for an unspecified reason. 11. Conclusion This experiment shows that legislators respond less to an undocumented immigrant than to a citizen, regardless of ethnicity. The saliency of national status is understandable given the incentive for legislators maximize electoral rewards by prioritizing citizens. What is surprising here are the actions of the Latino legislators, who might have made the same rational calculation as non-Latino legislators in choosing to respond less to the undocumented Latina than to the Latina citizen. The results might be inconclusive, but they still raise interesting questions about intersectional linked fate. One such questions is whether intersectional linked fate should be thought of in terms of absolutes, or whether it could be thought of as a phenomenon that is mediated by electoral factors. 93 Chapter 5: A Case Study of Immigrant Rights Advocacy in California 1. Introduction How do undocumented immigrants obtain representation from legislators? Undocumented immigrants do not possess the right to vote, and thus lack the traditional mechanism needed to incentivize legislators to act on their behalf. This chapter will further this dissertation’s attempt to answer that question by focusing on the role of the deliberative process between legislators and immigrant rights advocates. The answer to the question of how undocumented immigrants obtain representation must include the understanding that representation begins with movements and community organizations that are specifically oriented to their needs. This chapter presents an in-depth case study of those involved in advocating for immigrant rights in the state of California, and deals with how they obtained access to state legislators; how they determined the interests of the communities of undocumented immigrants they represent; and their experiences lobbying state legislators. Data was gathered through semi- structured interviews with a purposive sample of individuals who are either currently involved in advocating on behalf of undocumented immigrants or have done so in the past. This interpretive approach focuses on the personal experiences and observations these advocates have had with state legislators because these interactions serve as evidence for why legislators come to represent a population like the undocumented. This approach also adds to our understanding of immigrant rights advocacy, the importance of diversity of the advocates themselves, and how this might affect what issues come to the attention of legislators. As the interviews make clear, the experiences of the advocates point to different reasons as to why legislators might represent undocumented immigrants. For this reason, the primary 94 analysis is supplemented with a secondary analysis that uses the Paradigm Intersectionality Approach (Hancock 2013). This approach is also applicable to the goals of this chapter because of Hancock’s definition of it as a “justice-oriented analytical framework for examining persistent sociopolitical problems that emerge from…sociopolitical fissures as interlocking, process-driven categories of difference” (ibid, 282). The interviewees make clear that their advocacy is not just concerned with national status, but also with race, age, class, and gender, and the data suggests these intersections influenced their activism. The present study is not so much concerned with identity influence, and whether one is more salient to the other, but meaningfully engaging with complex causality-“the reality that multiple causal paths can lead to the same result” (ibid, 282). This formulation of empirical intersectionality moves past the static conceptions of identity as fixed variables and does not assume that they have an independent relationship to the outcome, that is legislative representation. Instead, it focuses on the inegalitarian traditions associated with these categories and asks what influence they have on the representational process. This chapter not only presents information as to how the activists built relationships with state legislators and identifies the conditions necessary for the representation of undocumented immigrants. Since many of my interviewees are Latino the importance of ethnicity came up, but for many of them, it served as the basis for an intersectional group consciousness (see Greenwood 2008)-the recognition that members of your in-group are diverse and subject to multiple oppressions that you might not be subject to- that led them to advocate for undocumented Latino immigrants. Much of political science attributes political participation to SES factors (e.g., Verba, Brady, and Schlozman 1995), such as education and income, and while tertiary education was certainly important to many advocates, so was this sense of intersectional 95 group consciousness that was borne out of personal experience. This is important because it suggests that ethnicity as a determinant of civic engagement is contingent on a number of distinct or overlapping factors. In the case of the activists, it is contingent on experiences of discrimination; experiences with unauthorized immigrants; or coming from a family with a recent immigrant history. While this chapter cannot exhaustively demonstrate causality for any of these factors, it does present evidence to suggest that lived experience can be a contributing factor to civic engagement. Another important factor for the ability to create relationships with state legislators was not only the election of Latino state legislators, but also that these legislators came from electorally safe districts. The evidence from the interviews suggests that the ability of the legislator to act on behalf of undocumented Latinos might have been predicated on whether they had constituencies that would allow them to engage in such behavior. The present analysis does not allow for the identification of a precise mechanism for accountability, but does suggest that representation of undocumented immigrants is determined largely by desire on the part of the Latino legislator to represent them and electoral security. 2. Representation as a Deliberative Process The previous chapters have demonstrated that traditional sanction-based models of representation cannot be relied on to bring equitable representation to undocumented immigrants. That, and the fact legislative descriptive representation along the axis of being undocumented is practically impossible, different approaches to representation are needed to provide representation to undocumented immigrants. One such mechanism for undocumented immigrants to gain voice is through means of mutual education with legislators. While Mansbridge’s theory of anticipatory representation is largely oriented toward satisfying the 96 anticipated interests of voters, it is also geared toward mutual education between legislators and constituents (2003). The motivations for the legislator’s behavior in regard to this style of representation are not driven by virtuous concerns, but by prudent ones. This kind of representation does create an opening for undocumented immigrants if advocates can gain access to legislators and demonstrate this is an issue that other segments of the constituency either care about or could come to are about in the future. Dovi (2002) also calls for mutual relationships between disadvantaged subgroups of historically disadvantaged racial and ethnic groups in her discussion of how descriptive representatives should be chosen, and her criteria for how a good elected descriptive representative should conduct her affairs, and one of those includes mutual recognition by the historically marginalized group and the representative. For Dovi, the burden for representation is placed on both the constituents and the representatives to be active in the process of mutual recognition so they “can coordinate on political activities” (ibid, 735). 35 In the case of the undocumented immigrants, the deliberative process of mutual dialogue is crucial for their representation as they are structurally barred from formal participation in politics. Education of the legislator is theorized to be important by normative theorists for reasons having to do with constituents being able to keep their legislators accountable (Mansbridge 2003), and with norms of justice involving historically disadvantaged groups (Dovi 2002). While Dovi was concerned primarily with criteria for the selection of proper descriptive representatives that fit the normative standards used to justify the selection of representatives based on membership in historically disadvantaged groups (ibid, 732-735), her call for mutual 35 Good descriptive representatives should be judged on whether “they reach out to (or distance themselves from) historically disadvantaged groups” and whether they have shared aims for the group, that “is to possess a similar vision for the future direction of politics-one whose goal is the improvement of the social, economic, and political status of particular historically disadvantaged groups” (Dovi, 2002, 737). 97 recognition has distinct value for questions of intersectionality and representation. Position taking and roll call voting are not enough to guarantee the representation of intersectionally disadvantaged groups, but mutual recognition is crucial to the representation of such groups. Groups outside the political structure have historically turned to protest politics (Piven and Cloward 1979; Tate 1998; Tarrow 1994) in order to have their voices heard by political elites. Undocumented immigrants have also mobilized extensively in the 21 st century, most notably in 2006 when millions took to the streets around the United States to protest H.R. 4437, an immigration restrictionist measure passed by the House of Representatives. Much of the social movement literature on undocumented immigrants has emphasized the protest tradition (e.g., Bloemraad and Trost 2008; Baretto et al. 2009; Nicholls 2013) or coalitions among activists (e.g., Cordero Guzman et al 2008; Hondagneu-Sotelo 2010). These studies demonstrate that the interests of undocumented immigrants are communicated to political elites, like legislators, through activist means. Seif (2004) comes closest to discussing the link between activists and legislators, and she gives a broad view of how activism by Latino educators helped lay the groundwork for the passage of in-state tuition in that state, and the relationship between them and activists. Her work, however, attributes the actions of the Latino legislators largely to their own background as working-class Latinos, and while this might be the case for some in her sample, it might not be the only explanation for their behavior. Her study only interviews the Latino legislators and the youth activists, and does not necessarily focus on how the two interacted with one another. What’s missing from these works is the representational process that describes how the legislator and the advocate worked together on these issues, and whether there were any conflicts between them. The previous chapters of this dissertation demonstrated that Latino legislators might not 98 always represent their undocumented constituents, and that there might be factors that influence them to do the opposite. The general case to be made is that immigrant advocacy leads to representation by state legislators, but what are the underlying motivations for legislators to act affirmatively on that advocacy? What role does legislator ethnicity play? 3. Case Study: California This study is set in the state of California for a number of reasons having to do with the unique history the state has with undocumented immigration. California is believed to have the largest population of undocumented immigrants in the country with an estimated population of 2.6 million; they make up 7% of the state’s population (Hill and Hayes 2013). California was also one of the first states in the U.S. to tackle the thorny issue of undocumented immigrants on its own through the infamous and controversial voter-approved referendum Propisition 187. Prop. 187 is considered the catalyst for the mobilization of Latino voters, a growing voting bloc that has only become more important over time, and their preference for the Democratic party (Pantoja, Ramirez, and Segura 2001; Fraga and Ramirez 2003). California’s turn from being a state that pioneered restrictive immigrant legislation to one that has become the most immigrant-friendly makes it an ideal site for study. Currently the state with the most liberal immigration laws (Marquez and Schraufnagel 2013), and perhaps, a favorable political climate for undocumented immigrants and their allies to organize themselves, partially due to the “density and size of the Latino (and undocumented) population” (e.g., Chavez et al 2014, 113). The rise, and strength of Latino legislators, over the past two decades also makes it important to this study, as one of the goals of this chapter is to investigate the role of descriptive representation along the lines of ethnicity in representing undocumented immigrants. 99 Since then, the state has passed many bills that are explicitly geared toward helping to integrate undocumented immigrants. Table 5 lists a selection of the pro-immigration bills the California legislature has passed, most notably those related to making higher education more accessible to students who are undocumented. Notably, Latino legislators have been active in sponsoring successful bills in the interests of undocumented immigrants. Table 5: A Selection of Pro-Immigrant Bills Passed Since 2000 in California 36 Bill Description Primary Sponsor AB 540 (2001) Grants in-state tuition to undocumented immigrant residents who have attended high school in California. Asm. Marco A. Firebaugh (D) AB 130 (2011) Grants colleges & universities discretion to award scholarships to undocumented immigrants who are eligible for AB 540. Sen. Gil Cedillo (D) AB 131 (2011) Allows undocumented immigrants who are eligible for AB 540 to apply for state-based financial aid. Sen. Gil Cedillo (D) AB 60 (2013) Allows an undocumented immigrant to apply for driver’s license that enables them to drive in the state of California. Asm. Luis Alejo (D) AB 4 (2013) Limits the ability of local jails to hold undocumented immigrants convicted of low-level crime over a certain amount of time in order to deport them. Asm. Thomas Ammiano (D) SB 1159 (2014) Requires professional licensing boards to accept tax identification numbers in lieu of Social Security numbers, making qualified undocumented immigrants eligible to apply. Sen. Ricardo Lara (D) 36 For the sake of space, I chose to highlight bills that focused solely on efforts to integrate the undocumented population, and protect them from deportation. 100 California also has an extensive recent history of activism on behalf of undocumented immigrants that not only includes the transformational activism of the immigrant youth movement, including the DREAMers, (Gonzalez 2008; Rincon 2010; Nicholls 2013), but also earlier activism by Latino educators and their allies on behalf of undocumented immigrants. This facet of activism has been mentioned briefly, or alluded to, by previous work (e.g., Seif 2004; Rincon 2010), but these authors did not conduct interviews with these activists as their work was usually focused on a particular historical moment, usually the passage of a bill. Interviews with activists from this period can help highlight the barriers that existed in this era to the representation of undocumented immigrants, and how they overcame them. 4. Research Design: Grounded Theory Method and an Interpretive Framework The present study uses methods associated with qualitative methods, such as grounded theory and interviews to understand how advocates for the undocumented built relationships with state legislators. The primary method used for data collection was semi-structured interviews with 20 individuals who had experience, either past or current, with advocating for undocumented immigrants. The sample of advocates is not representative of the universe of advocates in California, but is instead purposive as all participants were chosen because of their role in the lobbying process. All advocates were chosen through snowball sampling. I conducted research on various organizations that are concerned with the rights of the undocumented, and after contacting and interviewing some participants, I then asked them to refer others to me. As stated previously, the interviews are semi-structured in that all participants were asked the same questions, but there was flexibility in the interview process to allow for the participant to elaborate on certain answers as well as for me to ask follow up questions regarding the information given. The interviews lasted anywhere from 45 minutes to 90 minutes depending on 101 the availability of the participant and their willingness to elaborate on certain questions. The interviews were conducted over the phone, over Skype, or in person, depending on the availability and preference of the interviewee. Most of the interviews were collected between late November 2014 and early January 2015, with one interview being done in March 2015. The purposive sample of advocates is made up mostly of Latino participants, with Latinas (11) making up the majority and Latino men making up a smaller portion of the sample (3). Of the 13 Latino participants, five have a history of being undocumented. The remaining advocates consist of three White women, one White man, one Asian American woman, and one Asian American man. The ages of the participants ranged from 24 to 69. The present study follows the grounded theory method pioneered by Levi and Strauss (1967). This method was invented as a reaction to the dominance of deductive theory testing in the researchers’ disciplinary home of sociology. Grounded theory calls for researchers to collect data that is situated in a particular context and then undergo a process of analysis until a theory emerges. This inductive method is concerned with theory generation regarding a general research question and largely rejects a priori assumptions surmised through logical deductions (ibid, 3). Such a method, and the data gathered under it, interviews in the case of this study, lends itself to an interpretive framework. An interpretive framework often involves how participants in a particular context make meaning of their actions and social reality (Yanow 2003a). In such a study, the actor’s subjectivity, including their personal knowledge and experience, serves as a guide to the very process that the researcher is interested in (Yanow 2000; also, see Schneider and Ingram 1997; Yanow 2000). This particular study follows this route as it utilizes the insights gained from the activism of the interviewees in order to generate specific ideas about how they have built 102 relationships with state legislators in the state of California. I make no claims of absolute causality and do not claim that their words surmise the totality of activism on the issues facing those categorized as undocumented. Rather, these interviews help shed light on the barriers that these activists faced, as well as the avenues of activism that were opened for them. I believe that their insights help illuminate (but not definitively so) how the relationship between advocates and state legislators evolved over time to allow for the latter to provide political representation for undocumented immigrants. 5. Paradigm Intersectionality Approach The Paradigm Intersectionality Approach is used in conjunction with grounded theory and an interpretive framework as a way to organize the data. Hancock describes the paradigm intersectionality approach as a “justice-oriented analytical framework” that “sets empirical standards of research for structural, political, and representational intersectionality” (Hancock 2013, 282). The benefits of this particular operationalization of intersectionality is that it first of all, primarily adheres to Cresnshaw’s original formulation of intersectionality as an approach (Crenshaw 1994; Hancock 2011, 2013). The other benefit is that it acknowledges causal complexity and moves the researcher away from questions of identity influence, and toward an actual investigation of how inegalitarian traditions linked to identity impacts a particular population. In the case of this study, the question is, how do immigrant rights advocates build relationships with state legislators despite the structural barriers that undocumented immigrants face? This is a multi-faceted question that contains a multitude of possible answers. The five linked dimensions of paradigm intersectionality that Hancock formulates (2011, 2013) provides a useful way to analyze the data. There are three dimensions that directly examine the complex 103 relationships between and within categories of difference, or inegalitarian traditions: Categorical Multiplicity, which acknowledges the need to assess which identity categories, or inegalitarian traditions, need to be included in the analysis; Categorical Intersections, which pays attention to where these categories, or traditions, intersect; Diversity Within, which highlights “systematic variation, whether attributable to individual orientations…or groups orientations” (Hancock 2013, 284). The other two dimensions includes Time Dynamics, which acknowledges the importance of historical context and sociopolitical changes over time, and Individual- Institutional Interactions, which acknowledges that categories of difference, or inegalitarian traditions, are shaped and reinforced by interactions between individuals and institutions. The intersectional paradigm sets forth three criterion of analysis. The first being, what are the inegalitarian traditions that present barriers to the representation of undocumented immigrants by state legislators? This criterion will be addressed more in-depth below, but at a minimum, the interviews show that many of the participants believe that restrictionist or negligent policies toward the undocumented and a tradition of structural anti-Latino racism have been primarily responsible for the lack of representation for the undocumented. The data also allows for an analysis of gender-related issues in conjunction with national status and ethnicity, as it was brought up by three of the women who had been, or, are undocumented. The second criterion to be addressed is, what is the substantive issue of social justice? The lack of representation at the state level has meant that state government has historically been slow to act on issues affecting the undocumented, affecting their livelihoods, limiting their opportunities, and in some cases, making them susceptible to deportation. The dimension of time dynamics proves to be especially important for this criterion as the interviewees includes a group of Latinas that were active in the 1970s and 1980s. The situation they faced was dramatically 104 different than the one that the mostly younger activists faced after the year 2000. For this particular criterion, age and class both become important categories as much of the activism described in this chapter concerns education for college students and the issues of accessibility, such as the absence of financial aid. The third criterion is, the scope and target of critique, which in this case, largely turns out to be state representatives, but not always. While parsimony requires state representatives remain the primary targets of critique, some of the interviewees broaden the scope to include representatives at the local and federal levels. These critiques are mentioned only in the context of what they can tell us about the nature of representation and how it relates to undocumented immigrants. 6. Advocacy on Behalf of Undocumented Students The issue of access to higher education for undocumented immigrants emerged as one of the major themes of this project. Most of the interviewees would have some connection to the efforts to make California public universities more accessible to undocumented students while some of the Latinas explicitly used their positions as college administrators and counselors to advocate for undocumented students. This particular section will focus on experiences of the Latina higher education workers in Los Angeles County and roughly begins in the late 1960s and continues into the 2000s. 6.1 Inegalitarian Traditions Meet in the Wake of the Civil Rights Movement The analysis of the interview data will begin with the interviews with the Latina women who had been active on the issue of immigrant rights since before the 1990s. This is an appropriate place to start, as it will help us to understand the Time Dynamics involved not only with legislative politics, but also their perspective on the issues facing undocumented immigrants 105 allowing for a preliminary analysis along the lines of Diversity Within and Categorical Intersections. Their position as advocates also demonstrates their privilege as not only citizens of the United States, but also as individuals who were empowered to seek change within the political system. Marisol is a first generation immigrant born in Durango, Mexico who came to the United States at five years old with her mother and siblings after her father, who was an authorized immigrant, obtained the necessary paperwork for them to come. This experience would remain with her as she told me that the issue of justice for undocumented immigrants was dear to her heart (Personal Interview 01/06/2015). In college, she became active with an organization known as the United Mexican American Students (UMAS), which would later become the organization known as the student rights organization Movimiento Estudiantil Chican@ de Atzlan (MECHA). Their activism largely focused on student issues affecting Chicano students 37 at the time, including issues of accessibility and retention. Her comments below explain how she became aware of the issues facing Latinos, and how they intersected with immigration issues: I became aware of the issues just through the participation in the organizations and also in college, you know, you have all these different rallies that they had at that time. During the ’60s was when we had a lot of the walkouts in the high school level, high school levels that they have to walk out of the school. The students were saying, “Our needs are not being met,” you know, we are gonna walk out and that… We are not getting the right education. Our teachers are not Latinos. Immigration issues were also a factor. People knew that if you didn’t have your green card, you couldn’t go to college (Personal Interview 01/06/2015). Her comments indicate that she was aware of the racial inequality that impacted Latinos in the education system and how that pushed many high school students to participate in the storied walkouts that occurred as part of the Chicano Civil Rights Movement in the United States 37 In the interview, Marisol used the term “Chicano” in reference to both the organization’s primary constituency, and the time period. She would later use, and prefer, the term “Latino” because she considered it “more inclusive.” 106 (Acuna 1972; Mariscol 2005). She also identifies here the intersection between ethnicity and national status as she understood then that those without citizenship, or a green card, were effectively barred from attending college. Marisol, and UMAS, understood that lacking authorization kept many Latino students from going to college and became active in trying to bring attention to their plight. As a member of UMAS, for example, her group would gain access to state legislators by reaching out first to individuals that were associated with government commissions, who would put them in contact with state legislators who she described as “amenable” to discussions about the need of undocumented immigrants. The legislators, who were primarily from Los Angeles County, would meet with them, and while they were supportive, she also believed they were hesitant, “They would say, ‘It’s a hard issue. We are going to have to wait it out’…it was a very volatile issue” (Personal Interview 01/06/2015). As a politically engaged citizen and activist, Marisol and her fellow Latino activists were able to leverage their networks to get in touch with state legislators, but as she notes, the legislators were hesitant to deal with the issue due to its controversial nature. She recalled that her collective experiences led her to advocate for undocumented Latinos at the time, primarily through “raising awareness and also talking to legislators” (Personal Interview 01/06/2015). She describes her experience with legislators, as thus, “at that time, we didn’t have too many Latino legislators so, there were a few non Latino legislators that were sympathetic, but they would say, ‘Yeah, we are gonna try.’” Her thinking here is interesting in that she frames her interactions with legislators in terms of their ethnicity, making it seem as though she thought Latino legislators might have been the most logical actors within the political system for her to meet with regarding undocumented Latinos. She describes some of the non- 107 Latino, and presumably White legislators, as sympathetic, but also states that they viewed the issue as a “hot potato.” “Legislators,” she stated, “were just worried about their election, whether they would be elected with this issue [immigration] or not.” Her activism also involved trying to raise awareness in the local Catholic Church in the late 1960s, as she was involved with a Latino Catholic group called “Catolicos por la Raza” (Spanish for “ Catholics for the People”). This group was hoping the Los Angeles Catholic Church would take up the undocumented immigration issue since Latinos made up a substantial portion of the Church’s congregation. She was disappointed when the very conservative Catholic cardinal at the time, Cardinal McIntyre, did not take up the issue, leading to the following confrontation at a Christmas Eve mass: We told him we were going to attend church and pray for the immigration issues…to pray for the Latinos, and he called the cops on us. When we all showed up, he locked the doors of the church. They couldn’t get in; so we had our own mass, our own rally, and of course the cops came...a lot of people were arrested. That is how he [the Cardinal] responded to our request for help. That was a slap in the face (Personal Interview 01/06/15). This particular story was shared to demonstrate the lack of institutional support that Latino activists had in getting this particular issue to be addressed in the late 1960s. While the data does not allow me to probe deeper into the relationship between the Catholic Church and the local Latino population, it does indicate that these Latino advocates did not have much influence over the agenda of their local primary religious institution, and this lack of access could be seen as further marginalizing the undocumented. Marisol would later become active both professionally and as an advocate for the undocumented in the world of higher education, when she joined a local college in 1972 in student services (Personal Interview, 01/16/15). It was during the 1970s that another interviewee, Caridad, became aware of the barriers that undocumented immigrants, primarily Latinos, faced. Caridad was a 2.5-generation Latina 108 born to a Mexican immigrant father, and a U.S. born Mexican-American mother whose family had been within the boundaries of the United States since before the U.S. annexation of the Southwest. She told me that her father had come to the United States as an undocumented immigrant and that his story had always resonated with her. Like Marisol, she went to high school at a time of institutional inequality for Latino students, as she told me that even though she was one of the most successful students in her school, the counselors still did not prepare her for college. The structural racism she encountered influenced her to become involved in a career of helping make college accessible to Latino students (Personal Interview 01/02/15). Caridad first encountered undocumented Latino students as she worked as an outreach counselor at a local college to enroll Latino youth. She describes what the experience generally felt like in the following passage: We would ask them “Are you a resident or a citizen?’ and to the students who said resident, we would have to follow up by saying, ‘We need your residency number.’ and so students would get stopped with that because they knew that they didn’t have that and so then they would tell us that they were here "undocumented" although, that was not the word that was used (Personal Interview 01/02/15). As an outreach officer she often met with Spanish-speaking students, and at that time, she was not aware of any proper term for the phenomenon that she was encountering. The students, she said, would often use the Spanish word mojado 38 to describe their situation. Caridad came into her career expecting to help Latino students overcome certain practices of institutional racism, but found that there were Latinos who were barred altogether from higher education because they did not have authorization to be in the United States. This discovery would make her realize that doing outreach and recruitment would not be enough to help all Latino students, as the lack of residency status would prevent even talented Latinos from getting into college. 38 As Caridad describes it, mojado was often a derogatory term that roughly translated to wetback in English. She said that at that time, this was the only word the undocumented students she met with knew to use to describe themselves. 109 Her experience advising undocumented Latino students led her to see the connection between the history of structural racism that affected Latinos and exclusion based on national status and how it negatively impacted access to education. She says, “In addition to not having access to education, there’s the issue of residency status you know, so that—again, just another barrier besides not receiving equitable education.” She also recounted one of the particular experiences of advising a young undocumented Latina in the late 1970s that put their predicament in stark relief for her. The young woman had come to Caridad asking for help, and as Caridad advised her that lacking citizenship meant she was not eligible for financial aid, the young woman asked her what she would do. Caridad informed her that she would hypothetically just put down that she was a citizen because at that time, no one would ask for verification, but that it was risky because knowingly lying on a government form was illegal. Caridad ultimately recommended the young woman not lie on the form. The young woman then told her, “You know every morning when I get up and I walk out that door, I take a risk, I take a risk of being caught, and I take a risk of being deported. So what’s one more risk if I’m going to gain an education?” The early histories shared by these two Latinos show how they originally understood the issue of undocumented immigration, as one that one was inseparable from ethnicity. This categorical intersection of race (specifically Latino ethnicity) and national status will remain consistent throughout the rest of the analysis. Class markers, such as poverty and the need for financial aid came up, and while the Latinas acknowledged its importance, I would like to note that both were adamant in acknowledging that it could not be separated from either race or national status. From the interviews, I can explain the process of marginalization as thus: racism assigned Latinos to not only inferior schools, but also prevented them from learning about 110 college; class prevented Latino students from learning about college from their parents, because their parents had often not attended college, and college was also unaffordable without financial aid; and being undocumented made it virtually impossible for these Latino students to even apply to college. Marisol and Caridad would eventually engage in activities to influence the state government in helping undocumented students. 39 6.2 Experiences with Descriptive Representation in the 1980s and Early-1990s In the 1980s, Caridad became involved with the Leticia A. Network, an activist group of Latinos that sought to coordinate efforts to help undocumented students navigate the residency requirements required by colleges and keep them enrolled after a landmark court ruling that granted undocumented immigrants in California residency status for purposes of higher education. 40 Her experience with the network brought her into contact with two Latino Assembly members, one of whom she would work with professionally on a statewide advisory board concerning higher education outreach to ethnic and underserved communities. As a member of the Leticia A. Network, Caridad remembers that Assembly Member Richard Polanco was involved with her group, as well as the larger fight to establish residency for undocumented immigrants. Polanco sent his staffer, Marco Firebaugh, to work directly with the organizers. I asked Caridad directly as to why she thought Polanco would become involved and she answered, “that was his community. He was from LA…areas that were predominantly Latino and also had lots of undocumented students.” She saw his support as being aligned with that of his constituents, even this advanced marginalized group, but that was not all. She also mentioned that the Leticia A. Network’s liaison was in Sacramento lobbying state legislators, 39 It is also important to note that their experiences are primarily with undocumented youth and students, and thus, the analysis of their words do not deal with other segments of the undocumented immigrant population. 40 Guillen (2003) has a chronological accounting of the group’s specific activism. 111 one of whom was Polanco. According to her, he seemed the most responsive, as he became active in working with the group (Personal Interview 01/02/2015). Using her experience as a reference, Polanco seemed to be acting substantively as a descriptive representative of Latinos, but I do not believe his ethnicity is the only relevant factor here. While I cannot fully assess his motivations, his actions demonstrate he was willing to work with these advocates in a manner that appears consistent with Dovi’s concept of mutual recognition (2002). The activists, who were all citizens, are acting in the interest of their undocumented students, and are in essence, serving as mediators for the issues they face. By sending his staffer, Polanco demonstrated his investment in the issue as a staffer represents a key resource for a legislator. It is also important to note that Polanco was working on this issue in the mid-1980s when the issue of in-state residency was being pursued through the courts. I believe this to be significant because he was involved with the activists even though it was not being pursued through the legislative domain, where he arguably had the most influence. In his case, descriptive representation occurs partially because of ethnicity, but also because of his willingness to work with a group like the Leticia A. Network. Norma, who worked on issues related to undocumented students, would eventually become a member of the Assembly in the early 1990s and propose bills to grant in-state tuition to undocumented immigrants. Her motivations for doing so can partially be attributed to emerging from this group of Latino educators who advocated for the undocumented. This was not an empty gesture of support as she introduced these bills in the early 1990s, at the “apex of anti-immigrant legislation,” and thus, received death threats (Personal Interview 01/02/2015). Like Polanco, I can only speculate about Norma’s motivations, and while ethnicity is important, I would also like to highlight the importance of her professional background in higher 112 education and her activism on behalf of undocumented immigrants. This background gave her the requisite knowledge to be an effective legislative advocate on their behalf, knowledge gave her the ability to voice their concerns once in public office. She possessed a relationship of mutual recognition that not only stemmed from her relationship with education activists, but as Caridad noted, also included a history of direct interaction with undocumentedstudents (Personal Interview 01/02/15). Her personal experience with the issues facing undocumented students and subsequent legislative activism on their behalf highlights one of the theorized benefits of descriptive representation: that such a representative can come into a political body and speak authoritatively on issues related to marginalized groups (Mansbridge 1999). This section highlighted the experiences of representation in the 1980s and early 1990s as primarily told by Caridad, but also verified by some of the other interviewees. It is important to call attention to a few trends here, one being that the background of the activists dealt solely with undocumented students, who are but one segment of the undocumented population, and therefore dealt with age. The majority of the people they were seeking remedies for were young Latino adults who were around the age of the typical college student. They appear to see all the categories being linked together as obtaining in-state tuition and financial aid would help young undocumented Latinos access institutions of higher education. I do not mean to imply that my interviewees, or the legislators, did not care about other segments of the Latino undocumented population, but I wanted to highlight that this particular segment was made the focus of this particular policy remedy because their age usually meant that they had attended public schools in California and possibly demonstrated an aptitude for college success, something that older immigrants by virtue of coming later, cannot do. I also believe the theme of intersectional political consciousness is emerging here as the interviewees and the legislators were very much 113 committed to helping college-age undocumented Latino immigrants obtain a legal right to public institutions. 6.3 Post-AB 540 and Activism in the University This section examines two respondents who were active at a local university after the passage of AB 540 in 2001, the California law that allowed “certain types of undocumented students to pay in-state tuition rates at the California State University and California Community Colleges” but excludes “the University of California system” (Chavez et al. 2014, 31). This would signal a big shift in the way the organizers dealt with the issues of undocumented students. The two Latinas in this section worked in an official capacity to represent undocumented students, with one also doing so in an unofficial capacity. Sara is a second-generation Latina citizen who came of age during the Prop. 187 protests. As low-income Latina, Sara grew up in a neighborhood where, “everyone’s parents were undocumented, or had their papers through Ronald Reagan’s law, the Simpson [Mazzoli] Act. I had a lot of family that was undocumented” (Personal Interview 12/29/2014). While she was U.S. born, Sara grew up interacting with undocumented Latinos and she did not see them as different from her. Sara would eventually become active as the director of a student advocacy organization within the California public university system As the executive director of this organization, she would help train the students to lobby state legislators, and her tenure began in 2002, a year after the passage of AB 540, which was championed by then Assembly Member Marco Firebaugh (D). This is the same Marco Firebaugh who worked for Assembly Member Polanco and also organized with the Leticia A. Network in the 1980s. According to Sara, her student organization worked closely with Firebaugh on issues related to students who were undocumented. 114 Sara partially credits her ability to build a relationship with Marco’s office through her previous association with a former Latino colleague from college, who was Firebaugh’s intern at the time. She knew the intern from their time as undergraduates when they both protested Proposition 209 at their respective campuses in 1996. They were both involved in student government at their respective campuses, and Sara believes this relationship she built in college helped her with her advocacy on behalf of students (Personal Interview 12/29/14). “He would organize one-on-one meetings with the student leaders to talk about the bills that were of interest to them,” she said (Personal Interview 12/29/14). Ethnicity certainly proved salient for her, but that was not all as she explained to me that having a college network of educated Latinos proved the most useful, even if it was small. Sara intimates here that ethnicity was not enough, but also education, and possibly class, was important to her success as a director. It is important that ethnicity and education, as a marker of class, be brought together because these two things were necessary in her case, much like the Latina advocates from above, for her to act on issues related to undocumented students. When asked why she thought Assembly Member Firebaugh was so willing to build relationships with her and the student representatives, she answered it was because he was once a student leader. When it came to the issue of undocumented students: “I think it was because he was young and from the hood like the rest of us, like he could relate to the struggle of students to make change through the legislature.” She thought that Firebaugh’s history as the child of Mexican immigrants and as a student activist made him willing to not only work with student leaders, but willing to fight for undocumented students in the legislature. Her statements also correspond with Seif’s (2004) that being a part of this generation of student Latino activists influenced Firebaugh, and Sara’s story shows that he, and his office, were interested in building 115 relationships with the new generation of Latino student leaders. When asked about her student leaders and some of their interactions with other state legislators, she said that they would often try to steer the conversation away from undocumented students and toward college fees (Personal Interview 12/29/14). Sara’s professional, paid position was primarily concerned with helping to advocate for students as dictated by the student leaders. She said that the historical moment she found herself in, which included having Latino student leaders in favor of advocating for undocumented students after the success of AB 540, made it possible for her to advocate for undocumented students. It is unknown whether the Latino student leaders were aware of these issues before the passage of AB 540, but by recognizing the undocumented students as part of their campus constituencies, they were acting affirmatively. Shared ethnicity, however, was not always sufficient for her student leaders to advocate for undocumented Latina students as she remembered having a very contentious relationship with a Latino student leader from a conservative campus who did not support rights for undocumented students. Her position as a professional meant the agenda she advocated for was largely dictated by the student leadership, and when the student leaders became more conservative, and less diverse, the focus shifted away from issues facing undocumented students (Personal Interview 12/29/14). Another interviewee who worked in the university system in the post AB 540 world was Linda, a Latina administrator at a public university in Los Angeles County. She originally learned about the issue facing undocumented students in the early 1970s while working in outreach. Born in Texas in the 1940s, Linda spent part of her childhood raised in a highly racist environment where the local Mexican American community lived with vivid memories of anti- Mexican violence. She also told me that she grew up in an environment where Mexicans were 116 deported, demonstrating that even as a child she understood the connection between ethnicity and national status. Eventually she moved to California as an older youth and encountered what she described as a milder racism. Her experiences and resilience in the face of racism led her to develop empathy for the undocumented students she encountered and see a part of herself in them (Personal interview, 03/21/15). I believe that this helped establish a sense of intersectional political consciousness for her as she recognized that that while racism impacted her, these students faced barriers related to race and national status, yet they did not let that stop them, much like she did not let racism stop her. Linda chose to talk mostly about her experience advocating for undocumented immigrants in the period both leading up to AB 540 and its aftermath. She presented me with a story of her dual roles in lobbying: one as the official campus lobbyist, and the other as a volunteer activist who sought to educate public officials and university communities about ways to help undocumented immigrants. She described the atmosphere in the post-AB540 period at her university as mixed. For example, it was not uncommon for her to encounter university officials who supported undocumented students, but then meet others who were openly hostile. She remembers that other administrators would quietly refer that undocumented students, most of whom were Latino, to her. The lack of information about AB 540 among faculty and administrators led her to found a workshop specifically geared to helping faculty and administrators across the state address the needs of undocumented students in the 2000s (Linda, Personal Interview 03/21/15). She also helped found a support group just for undocumented immigrants on her campus; this is important because it was not specifically tied to any one ethnicity or cultural group, but was geared towards providing a supportive space for all undocumented students. 117 As the campus lobbyist, she was constrained by the university’s agenda when meeting with state legislators. When I asked her if she was able to talk about the needs of undocumented students, she said, “I was only supposed to do it in the context of whether the university supported such bills” but also admitted that at times, she would bring up the issues any way when she could. While she did not go into great detail about how she brought the issues up, she did tell me that learning about the legislators personal likes and building good relationships with their staffs was beneficial to getting the legislator to listen to her. Linda was also a member of a volunteer organization of Latino educators, and had been since the 1980s, that would often set up workshops dedicated to issues facing Latino students, of which undocumented students were a major concern. She says that she and others would build relationships with Latino state legislators to educate them on the issues affecting undocumented Latino students. One of those legislators, she recalled, was Marco Firebaugh. “He was one of our students,” she told me. This comment was told in the context of how the Latino educators, counselors, or administrators often served as mentors of Latino students, and that as members of the relatively small, but growing, members of Latino professionals, they often kept in touch. The above examples show the importance of having descriptive representations within advocacy organizations. These two used their positions within their respective organizations to voice, and their own personal knowledge, to help advocate for undocumented Latino immigrants-even if it was clandestinely, as was the case sometimes for Linda. Sara, especially, was an example of how descriptive representatives in an advocacy organization could help steer the advocacy efforts toward intersectionally stigmatized groups (Strolovitch 224), like undocumented Latino students. It was her connection to a Latino staffer that helped her student 118 leaders build a relationship with a Latino Assembly Member, and she was not afraid to confront the conservative Latino student leader over his opposition either. 7. Immigrant Advocacy in Politically Competitive Regions With in-state tuition granted to students in California in 2001, the struggle for immigrant rights would change in scope and mission among the interviewees as a younger group of activists would seize on the changing political climate to advocate for further ways to integrate those who were undocumented. The interviewees in this section, in particular, would face the challenges of advocating in regions that were more conservative, such as Orange County and the Inland Empire. These conservative regions had smaller infrastructures to support immigrant rights advocacy and also featured more conservative electorates with a more recent history of Latino political empowerment. In these cases, the advocates would confront the limits of descriptive representation as it pertains to intersectionally stigmatized groups, but also become involved with the empowerment of undocumented Latino immigrants themselves. Most of the advocates in this section would advocate for the need to hold elected officials accountable through electoral means in order to provide representation for undocumented immigrants. Robert first came to the United States at the age of 14 and even though he himself was not undocumented, he felt a sense of linked-fate with the other Latino immigrants as they faced the same struggles adjusting to life in the United States as Spanish-dominant immigrants. He says that he first became involved in the immigrant rights movement in 2009 after seeing how being an undocumented student caused his friend great stress. Much like the earlier Latina activists, Robert, and obviously his friend, noticed that AB 540, while an important step, was not enough to ensure equal access to the higher education. In this sense, class, race, and national status served to disadvantage Robert’s friend since he was struggling to pay for his education and 119 was barred from applying for financial aid, state or federal, altogether. Upon seeing his friend stressed out by the lack of financial aid, Robert asked to be brought into the immigrant youth movement to help his friend. He subsequently became a member of a immigrant rights group focused on undocumented students and youth (Personal Interview 01/07/15). As a volunteer and organizer, one of his duties involved advocating elected officials at the local, state, and federal levels, a task that would prove frustrating in a more conservative place like Orange County. During our interview, most of the elected officials he told me about were Latino Democrats, but his experiences would reveal how intersectional stigma can complicate the representation of a group like undocumented Latina immigrants in electorally competitive districts. His experiences will also present an interesting question to scholars: does the larger political structure make the policy representation of undocumented immigrants dependent on Latino representatives? He recounts the following experience with a Latina Congresswoman from Orange County who he says, did not sponsor the federal DREAM ACT until “the last minute”: The reasoning behind this was because she was always afraid of losing her seat and so then nationally people were telling us that people in DC…were always saying, “Well, how do you want me to vote for the Dream Act when your own people don’t even cosponsor it?” We felt that if we have a Congresswoman who is Latina, and is proud to be a woman and a Latina, then she should be a champion of immigrant rights. We took it upon ourselves to do activism…to push her to sponsor the DREAM Act (Personal Interview, 01/07/15). This passage demonstrates Robert’s understanding of elected officials and their reelection concerns, but also about the expectation that he and the DREAMers but also reveals that descriptive representatives along the lines of ethnicity can be expected to act for intersectional co-ethnics by both those co-ethnics and the larger political establishment. Her Latina ethnicity, and pride in her heritage, made her appear as an ideal candidate to support the rights of the 120 DREAMers, who are mostly Latino. Her failure to do so had repercussions for the DREAMers beyond their feelings of disappointment, as he points out that the message from some non-Latino Members of Congress to the national DREAMers seemed to be that this was an ethnic issue and that Latino legislators needed to be the one to bring it up. If ethnicity and national status are being conflated by other political actors, then that presents a conundrum for Latino legislators who find themselves anointed as champions of undocumented Latinos whether they want it or not. The ramifications for this are troubling as it is obvious from this example that electoral incentives can play a large role in determining whether a Latino representative provides representation to undocumented immigrants, even if they do share the same ethnicity. The Latina Congresswoman’s lack of initial action led the DREAMers to pressure her with tactics such as appearing at all her public events, and even calling her out at a county political event. Robert later told me that the Latina MC would eventually come to sponsor the federal DREAM Act while in Congress 41 . The story illustrated above was included because of its direct relevance to the issue of representation for undocumented immigrants. The Latina MC seemed to prioritize her electoral concerns over that of being proactive on issues, or even openly supportive of issues, related to undocumented immigrants. Robert attributed, but did not justify, her actions to the competitive nature of her district, which is situated in conservative Orange Country. His relationship with his state representative would also be contentious due to factors related to the conservative nature of Orange County. His efforts to get a Latino state legislator in the county to speak to the Latina MC did not go over well, as he told me that the legislator “wanted no part of it” (Personal 41 Two of the interviewees with experience asked that the Orange County legislators not be named and only spoke to me about their experiences if I agreed to that condition. 121 Interview 01/2015). Robert’s description above indicates the lack of mutual recognition between undocumented Latino immigrants, especially the DREAMers, and the OC Latino representatives. Robert also recalls that the two Latino state legislators of Orange County were very selective about which immigrant rights bills to support. This was in contrast to the experiences he had with this Latino state legislator’s office in 2011. “It was Immigrant Day,” he recalls, “and I went there to lobby for the California Dream Act and the SEIU people were there to lobby for driver’s license. The office told the woman that they were not going to do it.” This experience illustrates how staff can represent the will of the legislator leading Robert to interpret their actions as not being supportive of the bill. He would later tell the office that the actions of the legislator was “hypocritical” for supporting the CA DREAM Act but not driver’s licenses considering that the students, and their parents, would still be at risk. He was especially indignant that the legislator opposed this bill as he said that it was something that could help the parents of the DREAMers as well. The legislator’s selective attitude, as dictated by his office, shows how legislators can take stands to help certain segments of an intersectionally stigmatized groups as he was only willing to support a bill that only helped “the good immigrant…the one that goes to college” (Personal Interview, 01/07/2014). Robert said that he would like his future work to focus on holding “progressive officials” accountable to the undocumented immigrant community. Two of my other interviewees also have experience lobbying in Orange County. Irene, a white woman with a long history of working for progressive causes, began working as a community organizer for her organization in the late 2000s. She worked mostly with Latino immigrants, many of them undocumented, and would essentially help train people in the community to be leaders, which included learning how to lobby. Irene recalls that the approach 122 of her organization was to allow the undocumented immigrants to define their interests and then work to place it on the agenda of decision-makers (Personal Interview 12/12/14). For Irene, meetings with the Latino state legislators often left her frustrated. She recounts one instance where she accompanied some leaders of the undocumented community to the office of a Latino state legislator and how he was dismissive towards them: It was like they came from this sentiment of; well of course I already know what you're going to tell me. I'm a Latino. It was as if they just thought they knew everything that our people were going through. I think what it was like; of course I'm going to support immigrant legislation. Of course I'm going to do that. Then he dismissed us. Yes, he was a Latino Democrat. We knew we had his vote but that wasn't the only point. It was about these people coming out of the woodwork, telling their story for the first time, trying to talk to their state legislator. Explain the details of the challenges they face. He was sort of like, "yeah whatever, you already have my vote. Now go away" (Personal Interview, 12/14/14). The experience Irene describes above highlights a few potential pitfalls when a legislator represents an intersectionally marginalized group. The first is that the legislator conflates his ethnicity with national status and claims that his understanding of marginalization on a single- axis (“being Latino”) means that he can effectively represent undocumented immigrants. His attitude can be interpreted as a negative consequence of gyroscopic representation, the kind of representation where a legislator looks to her “own values and judgement” (Mansbridge 2003). By looking within, the legislator is able to privilege his own beliefs, or his perceived understanding of the issues facing undocumented immigrants, and then uses them to dismiss those very constituents. It is also worth contemplating what the role race played in the interactions between Irene and the Latino legislator as his comments might have been in the context of him reacting to a White woman’s presence as an advocate for undocumented Latinos. It is possible that the situation Irene described was the result of him asserting to her that he had the appropriate background to represent undocumented Latinos. 123 This is an example of the failure of mutual recognition as Irene and the undocumented Latino legislator recognized him as a descriptive representative, but it appears that he did not recognize them as partners. While driver’s licenses are an important issue, the immigrant leaders wanted the chance to open a dialogue with the legislators about the specific challenges they face as residents of his district. Having his vote was not enough because it would allow him to believe that he has effectively represented undocumented immigrants without feeling the need to effectively communicate with them, or reach out to them. According to Irene, the immigrant leaders, and advocates, wanted to be partners in legislating with the state legislator because they wanted to work on issues together as they arose, and not be tied to a potentially static agenda. Irene’s negative experience with this Latino state legislator is an interesting contrast to her experience with a White Republican legislator who was also a self-described libertarian. While Irene made it clear that Orange County Republicans were usually not supportive of undocumented immigrants, and had often been hostile, she believes that this particular legislator was open to having a conversation with the undocumented immigrants because of his libertarian leanings. “His own party did not like him because he challenged them a lot,” she says (Personal Interview 12/12/2014). She felt that he gave the undocumented Latino leaders the most respect, and was the most interested in listening to their stories. However, this required Irene and the immigrant leaders to be creative in how they approached him, most specifically on the issue of towing cars owned by undocumented immigrants driving without a license (Personal Interview 12/12/14): As a Libertarian the way we framed it was the system is about going after people who are driving drunk. Why are we targeting those communities who are trying to get home from church? From work? That doesn't seem like it's a good use of the system's money, right? We didn't even make it about the more human aspect. We made it about government spending. Guess what? We got his vote. We got a Libertarian Republican's vote, and he was one of the key swing votes (Personal Interview, 12/12/14). 124 This specific case could be an example of gyroscopic representation as the White Republican legislator looked to his Libertarian ideology to understand the issue of how towing cars without licenses negatively impacted the undocumented community. Irene’s comment suggests that he did not vote with his party, and given that he is a Republican, it is unlikely that this voting base is supportive of, or possibly even aware, of this particular issue. However, it could also be an example of anticipatory representation in that he might have engaged in this dialogue with the immigrant leaders in order to learn about “latent issues” in his district in order to please some of his constituents. The data does not make it clear as to which one is occurring or if both are occurring simultaneously, but what is obvious is that he did engage in building a relationship with the undocumented community. Lanh, an Asian American activist with an immigrant background, also worked in Orange County for a faith-based organization. Lanh grew up in Orange County and remembers that he was inspired to become politically involved by both the election of a second- generation Latina to Congress and the connection between his own immigrant background and the general climate of nativism that expressed hostility toward Latino immigrants. As an organizer, Lanh would work to help community leaders to develop an agenda and take them to advocate for the undocumented community. He tells the following story of one of his experiences with a Latino state legislator: [The state legislator] had preferred a bill that…I think it was basically completely ineffective. He was touting that. We wanted something more comprehensive statewide [for] long term English learners. He had his bill, which really didn't do anything, [but] he did listen to us and was very cordial, but I don't think there was any action that was going [to]happen. I was a little disappointed. The thing is with developing grassroots leaders, we're cultivating an understanding of politics and how legislation works, but the thing that we're also cultivating is the relationship that they have with their 125 representatives. It wasn't complete failure in any sense…It takes time I guess (Personal Interview 12/16/14). While his experience was not as negative as Irene’s as this was a different legislator that she met with, he notes that he did leave the meeting feeling disappointed. The legislator did actually listen to the community leaders, which he identified as a positive, and as Lanh notes, this was a first step to building a relationship with that legislator. Lanh made it clear, however, that the state legislator never really addressed the main issue that brought him and the community leaders to that meeting: increased funding for English-language learners. It is also possible that the legislator was only meeting with Lanh and the community leaders because it is a courtesy extended to all constituents, and that he did not have an interest in actually incorporating their views into his legislation. Enrique is an undocumented immigrant who first became active in college, where he was a member of student organizations founded by undocumented immigrants for the expressed purpose of providing support for them. In 2011, he would find a job as a community organizer with a grassroots immigrant rights coalition located in the Inland Empire. It was his job to help build the coalition by reaching out to non-traditional partners who could help provide services of undocumented immigrants in the area like health care providers and law firms. One of his jobs was to coordinate advocacy at all levels of government. He recalled how he repeatedly met with a Republican MC, but to no avail: We've met with [Republican MC] over last two years at least eight or nine times…and every time he was just non-responsive or non-committing to anything. We would say "Hey, there's all these talk about immigration reform, can we make sure that you vote yes?" that kind of thing or "Can you look what we think?" and he would just…shove us off and [say] "Well, I'm hearing from both sides and I'm not going to commit to anything…(Personal Interview, 12/15/14). Enrique’s assessment of the Republican as non-committal is somewhat reminiscent to that of 126 Lanh’s assessment of his meeting with the Latino state legislator in Orange County. Whereas Lanh was diplomatic about the meeting, Enrique is much less forgiving and essentially describes such meetings as fruitless. He does not believe that such meeting, and he had a good number of them, did not lead to the building of any particular relationship. He says he often had the same relationship with Republican state legislators, who he says his coalition has had issues in getting meetings with, as opposed to Democratic state legislators who are more willing to meet with them. He specifically mentions his meeting with a Republican who was known for not supporting immigrant rights. The fact that he was able to meet with this Republican, who Enrique says is known for making comments about carrying a gun in order to protect himself from DREAMers, only adds weight to the idea that perhaps legislators meet with constituents out of some professional courtesy or sense of obligation. “I’ve met with him at least twice and a couple other coalition folks…they'd met with him as well, and like I said ... we'd get up there…he would escort us rather promptly most of the time,” he says. “He was not very friendly,” he added (Personal Interview 12/15/14). Enrique attributed the differences in his experiences with the state legislators to party. The Republicans, he said, had a base that was anti-immigrant and that gave them incentives to ignore the issues facing undocumented immigrants, as well as not listen to the advocates of undocumented immigrants. The easier access to the Democratic state legislators he attributes to the party’s support for “social justice and civil rights,” as well as to possibly appeal to the Latino citizens who could vote. This electoral incentive is important to note as Enrique also said that his coalition engaged in civic engagement and voter mobilization. Due to time constraints, he did not go into specifics about this particular aspect of his work, but it sounded as though he supported these efforts in order to pressure legislators to act. 127 Diana is an undocumented Latina who became involved in the immigrant rights movement when she was in high school and realized that she did not know what to after graduation because of her undocumented status. She ended up meeting with a high school guidance counselor who was actually familiar with AB 540 and he recommended that she attend a local seminar for undocumented youth on applying to college being held by an immigrant rights organization in Los Angeles County, where she lived. She attended the event and would eventually become involved with their youth leadership program. Eventually, she would become a coordinator of grassroots lobbying as she became involved with the statewide campaign to press for federal legislation on immigration reform (Personal Interview 12/22/14). She became active in the mid to late 2000s and her goal was to build “immigrant power in moderate and conservative areas” of California. She was to help organize immigrant communities and train youth leaders to advocate for immigration reform, as the agenda at that time was focused on the national DREAM act, which she and others believed to be a stepping- stone to comprehensive immigration reform. She would “drive a thousand miles a week” to what she called “moderate Republican” district to organize the undocumented immigrant youth, help them find resources, and eventually lobby federal officials (Personal Interview 12/2014). Diana recalls that the issue she faced most often was getting access to the legislators, both at the local and state level. “For Republicans,” she says, “it’s hard to get meetings with them if you are not a constituent. It’s all about constituency.” This was her primary experience when dealing with Republican state legislators. She remembers organizing in the Congressional District of Republican Kevin McCarthy, who she recalls she and the immigrant activists of that district “pressed on immigration reform for five years” (Personal Interview 12/22/2014). She met him and described him as very cordial. She was organizing and lobbying around the same time 128 that other groups in the area were attempting to grow Latino voting power in his district through voter registration drives. The exact relationship between the lobbying and the voter registration is not clear, but it is possible that they could have been related. Meeting with one Republican MC proved even more difficult as she would try to organize a visit with him and his office with some of his constituents who were undocumented youth. She remembers that his office kept putting them off, even though these youth were his constituents. Eventually she recalls, they forced him to meet with them by showing up at his office. The MC’s staff called the cops on them, but she says, their attitude changed after the local media heard about the ordeal. “The next day, his staff told us that he wanted to video-conference with us and they invited us to the office where there had cookies and bread. It’s small, but it’s a big victory for communities that had never heard of the [Member of Congress] before, and he had been there for over 25 years,” she said. For Diana, finally getting this long-serving MC to see his undocumented constituents was seen as a small step forward as it granted them a visibility that they never had before. This recognition was seen as politically empowering for the communities of undocumented Latinos that she was helping to organize. Political power, as Diana expresses here, is seen as holding elected officials accountable, and as she later told me, voting serves as a powerful sanction. As she explained it to me, undocumented immigrants and their communities have power, but are structurally (and legally) barred from exercising formal political power: The immigrant community can’t give to political campaigns…[but] they can go knock on doors, right? The motto is like, “I can’t vote, but I’m going to get ten other people to vote for me.” For a long time, that’s just the way that it has to be, so educating Latino to Latino and really training immigrant workers and immigrant youth to be civically engaged through door knocking…It’s really powerful in terms of getting the Latino community to the next level (Personal Interview, 12/22/14). The above passage is quite loaded and demonstrates two things: the importance of the vote, and 129 of building relationships between undocumented Latinos and Latino citizens. She went on to talk about the Latino community and expressed the desire to see its various members come together to talk about legislation and how it affects them. Ethnicity, she seems to be saying, is a powerful organizing force for undocumented Latino immigrants. Latino voting power seems to be one way for undocumented immigrants to receive representation as Latinos have the power to sanction or influence legislators on behalf of undocumented immigrants, and more specifically, their co- ethnics who happen to be undocumented. 42 Unlike the previous sections, this chapter features advocates who had a more contentious relationship with the legislators they lobbied. Their interactions with these legislators mostly marked them as outsiders, as most of the legislators were not as invested in building the kinds of relationships that invited active participation of the advocates. The Latino legislators, as described here, were examples of what happens when there is failure of mutual recognition as none of the advocates who met with them felt like they were being heard. Their actions are also an example of how intersectional stigmas can limit the representation of advanced marginalized groups. Robert, the young Latino male, and his group of activists even resorted to pressure tactics in order to get a Latino MC to support the federal DREAM Act. Both Diana and Robert would ultimately endorse a sanction-based model of representation as they believed that community organizations and advocates must hold legislators accountable. It is also important to note the time dynamics at play here as the general time frame for the experiences of these interviews was the mid to late 2000s to around 2011. Undocumented Latino students who gained access to public universities through AB 540 became more empowered and set out to make new legal claims based on their quasi-legitimized status, and 42 The importance of Latino voter engagement was also brought up explicitly by some of the interviewees, including Irene and Lanh, but was not mentioned extensively in this analysis. 130 found new modes of activism (Abrego 2008; Abrego 2011; Seif 2011). This was a new phenomenon in the immigrant rights movement, as in the past, such actions were considered too risky by undocumented Latino youth and their advocates (Seif 2004; Abrego 2008). Their activism also occurred in the context of increased Latino voting power in California, a fact that Diana (as well as Lanh and Irene) mentioned in their interviews. Both Lanh and Irene (but this was not included in the text) spoke of increased efforts to register and turn out Latino voters in support of immigrant rights because as Irene political interests, such as labor unions and progressive organizations “figured out that Orange County was majority-minority now” (Personal Interview, 12/14/14). While they all agreed that Latinos voting was necessary, and therefore implied that a sanctions-based model of representation was necessary in order for undocumented immigrants to gain political power, I included Diana’s comment specifically because she spoke most eloquently of the need for an alliance between Latino citizens who can vote and undocumented Latinos. She seems to understand that a sense of linked fate between Latino citizens and undocumented Latinos is necessary in order for them to achieve representation. 8. Grassroots Organizations and Gaining Access to Legislators The analysis so far has highlighted two interesting narratives: one, being that intersectional stigma is not always a barrier to substantive representation by descriptive representatives as described by the Latina advocates who were active in education, while the other narrative agrees with Strolovitch (2007) that strategic imperatives do serve to impede their representation of undocumented Latino immigrants by Latino legislators. This section features data from my interviews with two undocumented Latinas who are affiliated with grassroots community organizations, and highlights that there are cases when collaborations with other 131 organizations are necessary for gaining access to state legislators. They are both active in Los Angeles county. Both women also highlighted the importance of gender, and how they gave the impression that issues related to it were often left off the political agenda. An undocumented youth who became involved in activism was Carolina, and she explicitly discusses gender and how it influenced her activism. Carolina was born in Puebla, Mexico and came to the United States at a young age in the early 1990s, settling with her family in West Los Angeles. Carolina became an activist in high school with an established Chicano organization after she learned about how being undocumented status impacted her college choices. She continued her activism when she went to community college, but encountered great difficulty with the advisor of the Chicano activist club. “We disagreed on certain methods to the point where he couldn’t even say nice things about me when I was not there,” she said. Carolina felt that her problems with the advisor largely stemmed from the fact that the advisor, who was an older Chicano male, might “have felt threatened by a younger female with a new perspective” (Personal Interview 01/09/15) She left that group and worked exclusively with another group that focused solely on issues facing undocumented immigrants. She then transferred to a four- year college and remembered being burnt out on student organizing, which she had done for five years at that point. She took a job with a relatively new, grassroots immigrants rights organization as an ESL teacher. She described herself as being very jaded on organizing at that point, partially due to the elevated status of the DREAMers: I felt that in the narrative of the undocumented student, the DREAMers were put on a pedestal, and the rest of the community was put on the backburner and I didn’t like it because my parents would drive me to anything in the first place. So I felt like my parents needed to be a part of some sort of reform, or some part of the law. (Personal Interview, 01/09/15). 132 She enjoyed her time away from organizing and took up folklorico, a Mexican style of dance at her college. This was important to her, and I mention it here because it helped her understand that she had identities outside of organizing and politics, and in her words, helped her became aware of her own womanhood. She made friends there that allowed her to speak about her gender identity and what it meant to her, something that she felt had been suppressed during her previous activism. She felt that her previous mode of activism split people into “good” and “bad” based on Marxist theory, but her education at the public university opened her up to the idea that people are complex, and this in turn made it easier for her think about feminism and womanhood. This awakening helped her realize that even though she was an activist, that she was not fully empowered. As a young undocumented woman, she remembers the risk of sexual assault associated without having a driver’s license. She could not get a driver’s license and did not want a ticket or to have her car towed, but also had to worry about the vulnerability that can happen when one is dependent on some other mode of transportation. This concern about sexual assault is a legitimate one for undocumented women and served to either prevent her from her activism, or do so and be put at risk in order to do perform her duties as an activist. The grassroots immigrant rights organization that she joined was founded by two White women who were sociology professors, and they provided, a more welcoming space for her. She was happy to find positive female role models that she could embrace (besides her own mother), but admitted that it took time to get used to there style of management. She described it as “sweet” and “tender,” and as a space where she did not have to “feel ashamed of her womanhood” (Personal Interview 01/09/15). This new environment and organizing style allowed her to grow and she became a full time intern, and embraced a new kind of politics that 133 was not just concerned with passing policy, but changing the fundamental dynamics of society. She then became lead organizer and set out to work in the community. She currently sets up presentations regarding relevant state and local programs for the undocumented community, and also helps undocumented immigrants with issues related to residency and the police. “I do a variety of things,” she told me, “we are very low capacity.” They only become involved with lobbying when it is in collaboration with other ethnic or immigrant-serving organizations. One example she gave me was when she and the executive director met with the staffer of their Congress Member, a White Democrat, as part of a coalition of organizations that were concerned about the detention of undocumented immigrants. She also had met with the staff of her state legislator, a Latino Democrat, who has collaborated with her organization to organize forums in the community regarding driver’s licenses for undocumented immigrants. My interview with Carolina explicitly concerned gender and whether it was left of the political agenda for undocumented immigrants. “I often think that housewives are not a constituency or a group, that we look at…we have laborers and workers and students, but the moms…they are the ones we have the hardest time finding documentation for,” she told me. This was an issue of great importance to her as her daily work brought her into contact with mostly women who would could to seek aid for their families, yet whose issues are mostly invisible. She applauds these women and the advocacy on behalf of their families, including their husbands, but tries to find ways to talk about self-empowerment. Carolina also uses her position to advocate for the women on issues of domestic violence. She notices that mothers will often come to seek aid for their daughters regarding domestic violence, but rarely come for themselves. She recalls that sometimes she has to be proactive in getting women to seek help for their own situation, which means sometimes asking more than 134 once if they need help. Carolina told me that the undocumented women also usually don’t come out and tell her right away, but that the conversation usually begins on another topic and then eventually gets to the topic of domestic violence or sexual assault. This was not an infrequent issue, as she told me that once she had seven cases of gendered violence reported to her in a three-week span. (Personal Interview 01/09/15). Two of the undocumented women in the sample came to the United States as adults and settled in Los Angeles County. Their experiences are different from many of the other interviewees, as they have dealt with issues of gender, language, and parenthood in addition to being undocumented. However, they are like Carolina in that they are members of smaller grassroots organizations geared towards service, and they also, in their own way, recognize the importance of gender. Both women were Spanish-dominant and required the use of a translator in order to be interviewed. Due to the similarities of the interviews, I will share the interview of only one of them below. Virginia came to the United States twenty years ago to be reunited with her husband, settling in Los Angeles County. She told me that everything she does is very personal to her, and that part of her motivation for her activism stemmed from her experience of giving birth to her first son here in the United States. She remembered that there were complications during the pregnancy but that she and the hospital officials were unable to communicate due to the language barrier. She had to call her sister, who spoke English but lived in Texas, and have her serve as a translator. While the situation turned out all right for her and the baby, she remembers that the event “woke her up” and inspired her to get active to make sure others do not have to go through the same situation (Personal Interview, 12/10/14). She joined a local organization that provided English language classes and access to other 135 community organizations geared towards family or youth. She was able to take classes at the organization as well as do leadership training to be an advocate for other parents. Virginia expressed gratitude that the organization that offered the class allowed her to bring her children to the class, which made it easier for her to attend the leadership training. The need for daycare, or programs that recognize the need for such a service at their meeting, has also emerged in other research (Milkman and Terriquez 2012) as a necessary condition for the empowerment of undocumented women. Virginia recognized this fact as well and attributed it to her ability to obtain the required leadership skills. She eventually founded her grassroots group after the original organization that provided the leadership training ran out of funding, and stayed connected to the small network of service provider organizations geared towards families and children. She, and her organization, relied on them extensively to connect them to city services and elected officials due to their lack of English language skills. She says being a part of this network helped them build relationships with city services to the point where members of city departments would come to them asking if they needed help (Personal Interview, 12/10/14). Her first experience with an elected official, her local city council member, a White woman, was met with difficulty because of the language issue. The first time they met with someone from the office, they had issues with the staffer. Virginia did not go into specifics about what happened, but did mention that eventually the city councilor replaced that staffer with another one who she said “knew the community very well.” This one was Latina and a Spanish- speaker, and her ability to build relationships with Virginia and her group improved relationships between her and the councilmember (Personal Interview, 12/10/2014). In assessing her interests, Virginia was able to locate herself in the matrix of domination 136 (Collins 1990). She described to me that she and her group, which was mostly made up of immigrant mothers, faced disadvantages related to having low incomes, being a woman, being Latino, being Spanish-dominant, and being undocumented. When I asked her how she identified her groups’s “interests,” she told me it was because she lived them, and that every time her group attempted to gain access to resources, they encountered barriers related to some aspect described above. This includes going to community events where the Spanish language translation is of poor quality, or dealing with medical service providers who she says, are dismissive of her as a low-income, Spanish speaking woman. “I recognized that women do not receive…respect,” she said in regards to her attempts to get hospitals to take her son’s asthma seriously. Virgina learned to advocate state legislators initially through a coalition concerned about air quality and its effect on the inhabitants, particular children, in the LA region. She remembers meeting with her state senator, a White Democrat, in about 2004, but nothing fruitful came out of the interaction. As she puts it, he listened and did offer them some assistance, although she was not too specific about how he was helpful. When it came to the state legislators, she felt more at ease in dealing with the new Latino state legislator. She said that her relationship with the new Latino state legislator is “totally different than it was with the White Democrat.” While she thought the White Democrat was nice, she felt that being Latino gave him an understanding of the problems facing her community. For her descriptive representation meant that there was someone in office with a particular insight about what issues poor Latino immigrants face in their neighborhoods. Descriptive representation is important to her Virginia also because of the language barrier. She mentioned that she needed a translator to communicate with the White Democrat, and she did not need one with the Latino Democrat because he spoke Spanish, and also because 137 he had more than one Spanish-speaker on his staff. This made it easier for them to communicate and allowed her to express herself fully in Spanish. She also has a direct relationship with the Latino legislator’s staff, and remembers speaking with him directly about issues affecting immigrant children. “Whenever I have an issue now,” she says, “I call his staff directly, and they work with me” (Personal Interview 12/10/15). It is interesting to note how very different the relationship between Virginia’s organization and the Latino state legislator’s office was compared to her relationship with the White Democrat. While members of the same party, Virginia reported that the Latin state legislator’s office was more responsive to her than the White Democrat, but the question is, is ethnicity by itself a sufficient explanation? This particular Latino state legislator has a known passion for immigrant right, and his interest with having a working relationship with Virginia’s group could be related to the fact that he is the son of Mexican immigrants who were once undocumented. The importance of the Latino state senator’s childhood and his parents’ story was not lost on Virginia who said, “it is incredible when he talks about his mom…he recognizes the work of the women…and he knows the problems we face.” she said. The importance of descriptive representation for Virginia seems to be in that not only does the Latino state legislator have firsthand knowledge, through being the son of an undocumented Latina mother, but also that through talking about his mother’s experience, he grants women like Virginia visibility in the public space. Yet, much like Carolina’s analysis of gender and its near invisibility on the political agenda of the immigrant rights movement, Virginia spoke of specific gender issues, such as domestic violence and the need to empower women: The thing is the real issue in the community is domestic violence. [It] is complicated. It is not only that the man abuses the woman. The issue is because low-income people do not receive 138 support…especially the women and [they are told] you are the woman, you keep quiet, and you never say anything (Personal Interview, 12/10/14). Virginia links domestic violence to the low socioeconomic status of her community and the absence of resources to deal with this problem and the other social ills her community faces. She identifies this as a problem and highlights how the immigrant women she knows deal with their circumstances in the United States, including being a part of the community. “In this country, we understand that the women need to go out…and advocate, I think for respect…and for our kids,” she said. She also says that the women must work for change, and said that her organization had programs designed to help the women and their children. Like Carolina, she did not mention these specific gender issues in regards to lobbying state legislators or other elected officials. The two Latinas in featured in this section spoke about organizing with grassroots organization that have more limited resources. In terms of representation, their experiences were more limited as they sometimes relied on other coalitions to get access to legislators. Both women shared that having a Latino state legislator meant that their relationship with his office was more collaborative, with Virginia saying that his background as the child of undocumented immigrants was very important. The two women also spoke about how gender as a category of difference affected the lives of undocumented Latina women, primarily through the issue of domestic violence. While Carolina was more explicit about how she felt that gender was not something that the immigrant rights movement has dealt with yet, Virginia seemed to intimate the same thing as she discussed about the need of women to advocate on their own behalf for respect. 9. Discussion and Implications This chapter used the grounded theory method to uncover some of the ways undocumented immigrants receive representation from state legislators. The data and analysis 139 from the interviews with the advocates finds that descriptive representation is a necessary, but not always sufficient condition for representing undocumented immigrants. This finding indicates that, at least in California, issues associated with being undocumented are thought of in racialized terms, specifically, related to Latinos. The fact that undocumented immigration is seen as a Latino issue probably should not come as a big surprise since the overwhelming majority (76%) of the undocumented immigrants in the United States are believed to be Latino (Passel and Cohn 2009). What is interesting is that many of the advocates believed that this was an issue that Latino legislators should be active on, regardless of the electoral climate. Some of the advocates did mention Asian American undocumented immigrants, but never in a substantial manner that could be analyzed. However, much like Strolovitch’s study (2007), I find that in certain cases, representation along one-dimension, in this case Latino ethnicity, is not enough to provide sufficient representation, as strategic concerns might prevent some Latino state legislators from partnering with undocumented Latino immigrants. Descriptive representation certainly is an important part of the story regardless, and Dovi’s (2002) analysis of the “preferable descriptive representative” seems to apply here. Part of the frustration of these activists was the Latino legislators were not receptive to working with the undocumented Latino community leaders to craft an agenda based on the experiences of the former, but instead, as one activist put it, just wanted to vote on a bill and follow their party. This is in contrast to the experience of the advocates who worked with Polanco, Firebaugh, Norma, and the Latino state senator described by Virginia and Carolina. These descriptive representatives worked closely with the advocates, with the latter two being known for having personal connections to the immigrant struggles. One thing that should be considered for these three is that they were from largely safe Latino-majority districts in the more liberal Los Angeles 140 County. It is probable that advocating for undocumented immigrants was not seen as electorally dangerous in their districts; however, I do not think we should assess these candidates based solely on electoral concerns, as there is evidence that all of them worked on these issues throughout their lives without much public scrutiny. This findings corresponds to the finding in Chapter 3 that Latino voting age population is not connected to whether legislators responded to an undocumented Latina constituent. What were the factors that helped lead these Latino legislators to advocate for the undocumented immigrants? One possible factor is that they were involved in collaborative efforts with Latino professionals who cared about this issue; there is also some evidence from the interviews to suggest that these relationships began before some of the legislators became state legislators themselves. However, this might be a phenomenon specific to southern California. The more relevant factor here for scholars of representation is that Latino state legislators like Firebaugh and Norma brought their personal experience, and knowledge, with them into office and this this helped them to legislatively advocate for undocumented Latino students (e.g., Burden 2007; Butler 2014). There is also the fact that these two legislators in particular might serve as examples of gyroscopic representation as some of the advocates who worked with them believed that their experiences as the children of immigrants influenced their desire to advocate for undocumented immigrants. The paradigm intersectionality approach helped illuminate some of the contextual factors that influenced the representation of undocumented immigrants. The dimension of Time Dynamics helped show the changes of advocacy in the wake of the passage of AB 540. Some of the advocates pointed to the growing political power of Latinos, and posited that building a sense of linked fate between Latino citizens and undocumented Latinos. 141 The paradigm intersectionaliy approach’s categorical multiplicity helped illuminate that the various identity categories, and their corresponding intersection, were discussed by the advocates mostly in terms of the the inegalitarian traditions that accompanied them. Most interestingly, many of the respondents spoke of the way the inegalitarian traditions met, even the Latino citizens discussed ethnicity largely in terms related to structural racism. Structural racism was also deeply entwined with issues of class, especially when the Latina education advocates would talk about issues of student financial aid. It was interesting that class, as related to low socio-economic status, also discussed as a disadvantage by some of the later advocates like Virginia. Gender also emerged as a salient domain of disadvantage for the undocumented Latinas as three of those that I interviewed (only two of which were featured in the text) also mentioned the importance of empowering women, and some of the obstacles they have faced. Age was also salient as the most often discussed public policy domain was related to higher education, meaning that the proposed remedies would have primarily benefited undocumented Latino youth. All of these categories remained relevant, even if the primary analysis of the categorical intersection affecting undocumented immigrants remained that of race and national status. The dimension of Institutional-Individual Relationships explored the various ways that the advocates built relationships with legislative officials, or expressed their political power in relation to legislators. This dimension also most aptly demonstrated the diversity within the Latino community and how that affected the meaning that the advocates gave to their interactions. The Latina education advocates were able to influence the specific Latino legislators they worked with without the threat of sanction and built a working relationship based on not only mutual recognition, but also mutual goals. The Latina advocates found that descriptive representation was crucial to helping them work towards their policy aims. A 142 documented Latino immigrant youth like Robert did not find that descriptive representation necessarily helped him build relationships with Latino legislators, and only served to heighten the sense of marginalization that he felt undocumented Latino immigrants faced. The study showed that the diversity of Latino political actors in the discussion of immigrant rights advocacy included education professionals, community organizers, students, and elected officials. Their interactions with each other, as well as some non-Latino elected officials, helped determine whether they felt like political insiders or outsiders. 10. Conclusion Due to the small N and highly localized nature of this chapter, this cannot be considered an exhaustive exploration of the reasons why state legislators provide representation to the undocumented. More research is certainly needed on the experiences of immigrant rights advocates and the representational process that are concerned with other states and local contexts, such as rural areas; other types of organizations, like unions; and non-Latino undocumented immigrants. The insights offered by study this study, however, do correspond with the earlier chapters of my dissertation as they potentially identify specific mechanisms that might explain why Latino legislators might represent undocumented Latinos. The intersectional analysis also helps highlight the need for political science to investigate both the issues facing undocumented immigrants, and the need to attend to the various ways in which they are marginalized. For example, who sets the political agenda for undocumented immigrants, and, are gender-related issues left undiscussed? These are questions that can only be answered by more thoroughly investigating the relationship, and potential conflicts, between the immigrant rights movement and the state legislators. 143 Appendix A: Letters Undocumented constituent treatment Dear ______, My name is Maria Fernanda Garcia and I need help finding financial aid for college. I would be the first person in my family to attend college, which is very meaningful since we are undocumented immigrants. As a resident of your district, I was hoping you could help me find information on state, federal, or community grants. Thank you for your time. Sincerely, Maria Fernanda Garcia Citizen constituent treatment Dear ______, My name is Maria Fernanda Garcia and I need help finding financial aid for college. I would be the first person in my family to attend college, which is very meaningful since I am a third- generation Mexican American. As a resident of your district, I was hoping you could help me find information on state, federal, or community grants. Thank you for your time. Sincerely, Maria Fernanda Garcia Control Group Dear ______, My name is Maria Fernanda Garcia and I need help finding financial aid for college. I would be the first person in my family to attend college. As a resident of your district, I was hoping you could help me find information on state, federal, or community grants. Thank you for your time. Sincerely, Maria Fernanda Garcia 144 Appendix B1: Robustness Checks for Field Experiment on National Status Dep. Var.: 1=legislator responded;0=did not Independent variables Coefficient Std. Err. P-value Undocumented Treatment -0.079 0.025 0.002 Citizen Treatment -0.020 0.025 0.426 Constant 0.331 0.018 0.000 Note: Reference category is control group (national status not mentioned). N=1871 Appendix B2: Robustness Checks with State Dummy Variables Dep. Var.: 1=legislator responded; 0=did not Independent variables Coefficient Std. Err. P-value Undocumented Treatment -0.081 0.025 0.001 Citizen Treatment -0.168 0.025 0.511 Constant 0.377 0.040 0.000 Note: Reference category is control group (national status not mentioned). The estimation in this table also included dummy variables for states, which are not displayed. N=1871 145 Appendix B3: Robustness Checks: Regression Models with Interactive/Conditional Models Independent Variables Coef. (s.e.) Coef. (s.e.) Coef. (s.e.) Undocumented Treatment -0.083(0.027)** -0.129(0.032)*** -0.131(0.032)*** Citizen Treatment -0.008 (0.027) -0.060 (0.035) -0.060 (0.035) Undoc. Treat. x Latino legislator .029 (0.082) -0.014 (0.084) -0.015 (0.084) Citizen Treat. x Latino legislator -0.125 (0.088) -0.176 (0.090) -0.176 (0.090) Latino legislator 0.037 (0.068) 0.070 (0.069) 0.068 (0.069) Undoc. Treatment x Dem. Leg -------------------- 0.104 (0.038)** 0.107 (0.038)** Citizen treatment x Dem. Leg -------------------- 0.116 (0.053)* 0.117(0.053)** Democratic Legislator -------------------- -0.101 (0.038)** -0.104 (0.037)** Latino VAP -0.009 (0.001) -0.006 (0.001) -0.004 (0.001) Squire Index (2007) 0.038 (0.119) 0.067 (0.122) 0.131 (0.073) Legislative Staffer 0.003 (0.004) 0.002 (0.004) ----------------------- Constant 0.308 (0.027)*** 0.350 (0.031)*** 0.354 (0.0308)*** Note: Reference category is undocumented treatment group. N=1871 ***p<0.001 **p<0.01; *p<0.05 146 Appendix C1: Balance Check Across Co-Variates for Experiment Disentangling Ethnicity and National Status Undoc. Latina Treatment Undoc. East. Euro. Latina Citizen East. Euro. Citizen Latina Control East. Euro. Control Democratic Legislators (%) 46.9% 49.5% 46.2% 48.3% 48% 49.3% Women Legislators (%) 25.6% 23.1% 24% 26.7% 25.1% 24.3% Latino Legislators (%) 4.7% 4.2% 5.3% 4.7% 3.7% 2.7% Appendix C2: Robustness Checks Dependent Variable: 1=legislator responded; 0=did not Independent Variables Coefficient Std. Error P-value Undoc. East. Euro. -0.009 0.018 0.298 Latina Citizen 0.111 0.018 0.000 East. Euro. Citizen 0.060 0.018 0.001 Latina Control 0.028 0.018 0.059 East. Euro. Control 0.128 0.018 0.000 Constant 0.162 0.012 0.000 Note: Reference Category is Undocumented Latina Treatment. I estimated an OLS model with legislator response as the dependent variable. The results are similar to those presented as crosstabs in Table 1. 147 Appendix C3: Robustness Checks with State Dummy Variables Dep. Variable 1=legislator responded; 0=did not Independent Variables Coefficient Std. Error P-value Undoc. East. Euro -0.007 0.017 0.331 Latina Citizen 0.117 0.017 0.000 East. Euro. Citizen 0.066 0.017 0.000 Latina Control 0.027 0.017 0.063 East. Euro. Control 0.130 0.017 0.000 Constant 0.101 0.039 0.005 Note: Reference Category is Undocumented Latina Treatment Group. The estimation in this table also included dummy variables for states, which are not displayed. 148 Appendix C4: Robustness Checks: Regression Models with Interactive/Conditional Models Based on Latino Ethnicity Independent Variables Coef. (s.e.) Coef. (s.e.) Coef. (s.e.) Undoc. East Euro. -0.108 (0.018) ----------------- -0.008 (0.018) Latina Citizen 0.109 (0.018)*** ----------------- 0.114 (0.018)*** East. Euro. Citizen 0.062 (0.018)** ----------------- 0.067 (0.018)*** Latino Control 0.049 (0.018)* ----------------- 0.033 (0.018)* East. Euro. Control 0.128 (0.018)*** ----------------- 0.126 (0.018)*** Undoc. East. Euro. x Latino Leg. ----------------- 0.016 (0.082) ---------------------- Latina Citizen x Latino Leg. ------------------ 0.139 (0.077)* ---------------------- East. Euro. Citizen x Latino Leg. ------------------- 0.017 (0.079) ---------------------- Latina Control x Latino Leg. ------------------- -0.037 (0.082) ---------------------- East. Euro. X Latino Leg. ------------------- 0.154 (0.098) ---------------------- Democratic Legislator ------------------ --------------- 0.001 (0.001) Undoc. East. Euro. x Dem. Leg. ------------------- --------------- -0.001 (0.002) Latina Citizen x Dem. Leg. ------------------- ---------------- -0.001 (0.001) East. Euro. Citizen x Dem. Leg. ------------------- ---------------- -0.002 (0.001)* Latina Control x Dem. Leg. -------------------- ----------------- -0.001 (0.001) East. Euro. Control x Dem. Leg. -------------------- ----------------- 0.000 (0.001) Constant 0.162 (0.013)*** 0.145 (0.056)* 0.161 (0.013)*** Note: Reference category is undocumented Latina treatment group. N=6,082 ***p<0.001 **p<0.01; *p<0.05 149 Appendix D: Questions for the semi-structured interviews Consent Questions: Is your participation voluntary? Do you understand the goals of this research project? Do you understand that you do not have to answer any question that makes you feel uncomfortable and that you can terminate the interview at anytime? Substantive Questions: Can you please identify yourself, and tell me a little about yourself, such as where you were born, where you grew up, your age, and your highest level of education? What is the name of the organization you work for? How did you become involved in the immigrant rights movement? Have you always been interested in immigrant rights? How did you become interested in this issue? How did you become aware of the struggle of undocumented immigrants? How long have you been involved with this particular organization? How did you become involved with this organization? What are the issues that you have found to be among the most important to undocumented immigrants? Are you involved in crafting policy for the organization? How closely does your organization work with other immigrant rights groups? Are there any associations or groups that have trained you in activist tactics, or offered support for your endeavors? How would you define your role in your organization? For example, what are some of the duties you have performed for this group? Could you describe to me your previous experiences in social justice advocacy? Do you have any previous experiences in politics? If so, how did you become involved in politics? 150 Have you ever engaged in lobbying of elected officials, specifically, state legislators? Who have you lobbied? Do you meet with the elected official or the elected official’s staff? What is the most positive experience you have had while lobbying an elected official? (I will reference any elected official by their salient characteristics. I will not record names). Can you share some negative experiences you have had while lobbying? (I will ask the interviewee to not specifically name any individuals involved in the negative experience). Do you have any previous experiences in politics? If so, how did you become involved in politics? Who are some elected officials that you feel are allies of undocumented immigrants? Here is a scenario I would like you to consider: California elects a governor who holds anti- immigrant views. You need to mobilize quickly against this governor. Name three people or organizations that you would call immediately. Have you ever thought of running for political office yourself? 151 References: Abrego, Leisy. 2008. "Legitimacy, social identity, and the mobilization of law: The effects of Assembly Bill 540 on undocumented students in California." Law & Social Inquiry 33 (3): 709- 734. Abrego, Leisy J. 2011. "Legal Consciousness of Undocumented Latinos: Fear and Stigma as Barriers to Claims‐Making for First‐and 1.5‐Generation Immigrants." Law & Society Review 45, (2): 337-370. Acuna, Rodolfo. 1972. “Occupied America: The Chicano’s Struggle Toward Liberation.” San Francisco: Canfield Press. Acuna, Rodolfo. 1996. “Anything But Mexican: Chicanos in Contemporary Los Angeles.” New York: Verso. Anzaldua, Gloria and Cherrie Morgara. 1987. This Bridge Called My Back: Writings by Radical Women of Color. New York: Aunt Lute Books. Ansolabehere, Stephen, and Philip Edward Jones. 2010. "Constituents’ Responses to Congressional Roll‐Call Voting." American Journal of Political Science 54 (3): 583-597. Anzia, Sarah F. and Christopher R. Berry. 2011. “The Jackie (and Jill) Robinson Effect: Why Do Congresswomen Outperform Congressmen.” American Journal of Political Science, 55 (3): 478- 493. Arnold, R. Douglas. 1990. The Logic of Congressional Action. Yale University Press: New Haven. Ariely, Dan, Anat Bracha, and Stephan Meier. 2009. “Doing Good or Doing Well? Image Motivation and Monetary Incentives in Behaving Prosocially.” American Economic Review, Vol. 99 (1): 544-55. Alvarez, R. Michael, and Lisa Garcia Bedolla. 2003. “The Foundations of Latino Voter Partisanship: Evidence from the 2000 Election.” Journal of Politics 65 (1): 31-49. Barreto, Matt and Jose Munoz. 2003. "Re- examining the 'Politics of In-Between': Political Participation Among Mexican Immigrants in the United States." Hispanic Journal of Behavioral Sciences 25 (4): 427-444. Baretto, Matt. 2005. “Latino Immigrants at the Poll: Foreign-born Voter Turnout in the 2002 Election.” Political Research Quarterly 58 (1): 79-86. Barreto, Matt A., Sylvia Manzano, Ricardo Ramirez, and Kathy Rim. 2009. "Mobilization, Participation, and Solidaridad Latino Participation in the 2006 Immigration Protest Rallies." Urban Affairs Review 44 (5): 736-764. 152 Baretto, Matt and Gary Segura. 2014. Latino America: How America’s Most Dynamic Population is Poised to Transform the Politics of the Nation. New York: Public Press. Bejerano, Christina E. 2013. The Latina Advantage: Gender, Race, and Political Success. Austin: University of Texas Press. Beltran, Cristina. 2010. The Trouble with Unity: Latino Politics and the Creation of Identity. Oxford: Oxford University Press. Bergan, Daniel E. 2009. “Does Grassroots Lobbying Work? A Field Experiment Measuring the Effects of an E -mail Lobbying campaign on Legislative Behavior.” American Politics Research 37 (2): 327-52. Bertrand, Marianne and Sendhil Mullainathan. 2004. “Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination.” American Economic Review 4: 991-1013. Bosniak, Linda. 2006. The Citizen and the Alien: Dilemmas of Contemporary Membership. Princeton: Princeton University Press. Branch-Brioso, Karen. 2008. “UCLA Study Reveals Growing Gender Gap Among Hispanic College Students.” Published October 16, 2008. Accessed April 20, 2014. http://diverseeducation.com/article/11836/. Bloemraad, Irene, and Christine Trost. 2008. "It's a Family Affair Intergenerational Mobilization in the Spring 2006 Protests." American Behavioral Scientist 52 (4): 507-532. Brand, John A. 1969. “Concillers, Activists, and Electors: Democratic Relationships in Scottish Cities. In Comparative Legislative Behavior: Frontiers of Research, ed. Samuel C. Patterson and John Wahlke. University of Iowa Press. Bratton, Kathleen A. 2002. “The Effect of Legislative Diversity on Agenda Setting: Evidence from Six State Legislatures.” American Politics Research 30 (2): 115-142. Bratton, Kathleen A. 2005. “Critical Mass Theory Revisited: The Behavior and Success of Token Women in State Legislatures.” Politics and Gender 1 (1): 97-125. Bratton, Kathleen. 2002. “The Effect of Legislative Diversity on Agenda Setting: Evidence from Six State Legislatures.” American Politics Research 30 (2): 115-142. Bratton, Kathleen A. 2006. “The Behavior and Success of Latino Legislators: Evidence from the States.” Social Science Quarterly, Vol. 87, Issue 5, pp. 1136-1157. Bratton, Kathleen A. and Kerry L. Haynie. 1999. “Agenda Setting and Legislative Success in State Legislatures: The Effects of Gender and Race.” Journal of Politics 61 (3): 658-679. 153 Broockman, David. 2013. “Black Politicians Are More Intrinsically Motivated to Advance Blacks’ Interests: A Field Experiment Manipulating Political Incentives.” American Journal of Political Science 57 (3): 521-36. Brown, Nadia. 2014. Sisters in the Statehouse. New York: Oxford University Press. Brown, Anna and Patten, Eileen. “Statistical Portrait of the Foreign-Born Population in The United States, 2012.” Pew Research Center. Published April 29, 2012 <http://www.pewhispanic.org/2014/04/29/statistical-portrait-of-the-foreign-born-population-in- the-united-states-2012/>. Bullock, Charles S., III 1981. “Congressional Voting and the Mobilization of a Black Electorate in the South.” Journal of Politics, 39: 662-682. Burden, Barry C. 2007. Personal Roots of Representation: Princeton: Princeton University Press. Butler, Daniel. 2014. Representing the Advantaged: How Politicians Reinforce Inequality. New York City: Cambridge University Press. Butler, Daniel and David Broockman. 2011. “Do Politicians Racially Discriminate Against Constituents? A Field Experiment on State Legislators.” American Journal of Political Science, 55 (3): 463-475. Butler, Daniel, Christopher F. Karpowitz, and Jeremy C. Pope. 2012. “A Field Experiment on Legislator’s Home Styles: Service versus Policy.” Journal of Politics, 74 (2): 474-86. Cain, Bruce, John Ferejohn, and Morris Fiorina. 1987. The Personal Vote: Constituency Service and Electoral Independence. Cambridge: Harvard University Press. Cameron, Charles, David Epstein, Sharyn O’Halloran. 1996. “Do Majority-Minority Districts Maximize Black Substantive Representation In Congress?” American Political Science Review 90: 794-812. Canes-Wrone, Brandice, David W. Brady, and John F. Cogan. 2002. "Out of step, out of office: Electoral accountability and House members' voting." American Political Science Review 96 (1): 127-140. Canon, David. 1999. Race, Redistricting, and Representation: The Unintended Consequences of Black Majority Districts. Chicago: University of Chicago Press. Carroll, Susan. 1994. Women as Candidates in American Politics. Bloomington: Indiana University Press. Chavez, Maria, Jessica L. Lavariega Monforti, and Melissa R. Michelson. 2014. New Immigration Policies and the Lives of Undocumented Latino Youth. Boulder: Paradigm Publishers. 154 Carnes, Nicholas. 2012. “Does the Numerical Underrepresentation of the Working Class in Congress Matter?” Legislative Studies Quarterly 37 (1):5-34. Casellas, Jason P. 2009b. “Coalitions in the House? The Election of Minorities to State Legislatures and Congress.” Political Research Quarterly 62 (1): 120-131. Casellas, Jason P. 2011. Latino Representatives in State Houses and Congress. Cambridge: Cambridge University Press. Chong, Dennis and Reuel Rogers. 2005. “Racial Solidarity and Political Participation.” Political Behavior, 27(4): 347-374. Citrin, Jack and Donald P. Green. 1990. “The Self-Interest Motive in American Public Opinion.” Research in Micropolitics 3:1-28. Choo, Hae Yeon and Myra Max Feree. 2010. “Practicing Intersectionality in Sociological Research: A Critical Analysis of Inclusions, Interactions, and Institutions in the Study of Inequality.” Sociological Theory, 28 (22): 129-49. Clausen, Aage R. 1977. “The Accuracy of Leader Perceptions and Constituency Views.” 2 (4): 361-384. Cohen, Cathy. 1999. The Boundaries of Blackness: AIDS and the Breakdown of Black Politics. Chicago: University of Chicago Press. Collins, Patricia Hill. 2000. Black Feminist Thought: Knowledge, Consciousness, and the Politics of Empowerment. New York: Routledge. Combs, Michael W., John R. Hibbing, and Susan Welch. 1984. “Black Constituents and Congressional Roll Call Votes.” Western Political Quarterly. 37 (3): 424-434. Cordero-Guzmán, Hector, Nina Martin, Victoria Quiroz-Becerra, and Nik Theodore 2008. "Voting With Their Feet: Nonprofit Organizations and Immigrant Mobilization." American Behavioral Scientist 52 (4): 598-617. Crenshaw, Kimberle. 1989. “Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Anti-Discrimination Doctrine, Feminist Theory, and Anti-Racist Practices.” University of Chicago Legal Forum, 139. Crenshaw, Kimberle. 1991. “Mapping the Margins: Intersectionality, Identity Politics, and Violence Against Women of Color.” Stanford Law Review 43 (6): 1241-1299. Cuadraz, Gloria Holguin and Lynett Uttal. 1999. “Intersectionality and In-depth Interviews: Methodological Strategies for Analyzing Race, Class, and Gender.” Race, Gender, and Class 6 (3): 156-186. Dahl, Robert. 1961. Who Governs? Democracy and Power in an American City. Yale University 155 Press. Darcy, Robert, and Charles D. Hadley. 1988. "Black women in politics: The puzzle of success." Social Science Quarterly, 629-645. Davis, Angela Y. 1981. Women, Race, and Class. New York: Random House. Dawson, Michael. 1994. Behind the Mule: Race and Class in African American Politics. Princeton: Princeton University Press. Dhamoon, Rita Kaur. 2011. "Considerations on Mainstreaming Intersectionality." Political Research Quarterly, 64 (1): 230-243. Dovi, Suzanne. 2002. “Preferable Descriptive Representatives: Will Just Any Woman, Black, or Latino Do?” American Political Science Review 96 (4): 729-743. Downs, Anthony. 1957. “An Economic Theory of Democracy.” New York: Harper and Row. Ellickson, Mark C., and Donald E. Whistler. 2001. "Explaining State Legislators' Casework and Public Resource Allocations." Political Research Quarterly 54 (3): 553-569. Erickson, Robert S. 1978. “Constituency Opinion and Congressional Behavior: A Re- examination of the Miller-Stokes Representation Data.” American Journal of Political Science 71 (3): 883-912. Eulau, Heinz and Paul D. Karps. 1977. “The Puzzle of Representation: Specifying Components of Responsiveness.” In The Politics of Representation. Edited by Heinz Eulau and John Walke. Beverly Hills, Calif.: Sage, pp. 55-71. Fenno, Richard. 1978. Home Style: House Members in Their Districts. Boston: Little Brown. Fenno, Richard. 2003. Going Home: Black Representatives and Their Constituents. Chicago: University of Chicago Press. Fiorina, Morris. 1974. Representatives, Roll Calls, and Constituencies. Lexington Books. Fiorina, Morris P. 1989. Congress: Keystone of the Washington Establishment. New Haven: Yale University Press. Fogg-Davis, Hawley. 2008. “Theorizing Black Lesbianism within Black Feminism: A Critique of Same-Race Street Harrassment.” Politics & Gender 2 (1): 57-76. Fraga, Luis, and Ricardo Ramírez. 2003. “Latino Political Incorporation in California, 1990– 2000.” In Latinos and Public Policy in California: An Agenda for Opportunity, eds. David López and Andrés Jiménez. Berkeley: Institute for Governmental Studies, University of California. 156 Fraga, Luis R., Valerie Martinez-Ebers, Linda Lopez, and Ricardo Ramirez. 2007. “Gender and Ethnicity: Patterns of Electoral Success and Legislative Advocacy among Latina and Latino State Officials in Four States.” Journal of Women, Politics, & Policy 28 (3-4): 121-145. Frymer, Paul. 1999. Uneasy Alliances: Race and Party Competition in America. Princeton: Princeton University Press. Garcia Bedolla, Lisa. 2005. Fluid Borders: Latino Power, Identity, and Politics in Los Angeles. Berkeley: University of California Press. Garcia Bedolla, Lisa. 2007. “Intersections of Inequality: Understanding Marginalization and Privilege in the Post-Civil Rights Era.” Politics & Gender 3 (3): 323-48. Garcia Bedolla, Lisa and Melissa R. Michelson. 2012. Mobilizing Inclusion: Transforming the Electorate through Get-out-the-Vote Campaigns. New Haven: Yale University Press. Gerber, Alan and Donald P. Green. 2000. “The Effects of Canvassing, Telephone Callas, and Direct Mail on Voter Turnout: A Field Experiment.” American Political Science Review 94 (3): 653-663. Gerber, Alan and Donald P. Green. 2008. “Social Pressure and Voter Turnout: Evidence from a Large-Scale Field Experiment.” American Political Science Review 102 (1): 33-48. Geron, Kim. 2005. Latino Political Power. Boulder: Lynne Rienner Publishers. Greenwood, Ronni Michelle. 2008. “Intersectional Political Consciousness: Appreciation for Intragroup Differences and Solidarity in Diverse Groups.” Psychology of Women Quarterly 32: 36-47. Griffin, John D. and Brian Newman. 2005. “Are Voters Better Represented?” Journal of Politics. 67 (4): 1206-1227. Griffin, John D. and Brian Newman. 2007. “The Unequal Representation of Latinos and Whites.” Journal of Politics 69 (4): 1032-1046. Gonzales, Roberto G. 2008. "Left out but not shut down: Political Activism and the Undocumented Student Movement." Northwestern Journal of Law & Social Policy, 3: 219. Grofman, Bernard, Lisa Handley, Richard G. Niemi. 1992. Minority Representation and the Quest for Voting Equality. New York: Cambridge University Press. Grose, Christian. 2005. “Disentangling Constituency and Legislator Effects in Legislative Representation: Black Districts or Black Legislators?” Social Science Quarterly, 86 (2): 427-43. Grose, Christian. 2011. Congress in Black and White: Race and Representation in Washington 157 and at Home. Cambridge: Cambridge University Press. Grose, Christian R. 2014. “Field Experimental Work on Political Institutions.” Annual Review of Political Science 17. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2335222. Guillen, Liz. 2003. “Undocumented immigrant students: A very brief overview of access to higher education in California.” In Teaching to Change LA's School Report Card 3, ed. J. Rogers, Los Angeles, CA: UCLA's Institute for Democracy, Education, & Access. Guinier, Lani. 1994. The Tyranny of the Majority. New York: Free Press. Hancock, Ange-Marie. 2004. The Politics of Disgust: The Public Identity of the Welfare Queen. New York: New York University Press. Hancock, Ange- Marie. 2007. “When Multiplication Does Not Equal Quick Addition: Examining Intersectionality as a Research Paradigm.” Perspectives on Politics 5 (1): 63-79. Hancock, Ange-Marie. 2009. "An Untraditional Intersectional Analysis of the 2008 Election." Politics & Gender 5(1): 96-105. Hancock, Ange-Marie. 2011. Solidarity Politics for Millennials: A Guide to Ending the Oppression Olympics. New York: Palgrave MacMillan. Hancock, Ange-Marie. 2012. “Trayvon Martin, Intersectionality, and the Politics of Disgust.” Theory & Event, 15 (3). Hancock, Ange-Marie. 2013. “Empirical Intersectionality: A Tale of Two Approaches.” UC Irvine Law Review 3(2): 259-96. Hancock, Ange-Marie. 2014. “Intersectional Representation or Representing Intersectionality? Reshaping Empirical Analysis of Intersectionality.” In Representation: The Case of Women, ed. Maria C. Escobar-Lemmon and Michelle M. Taylor-Robinson. Oxford: Oxford University Press. Hawkesworth, Mary. 2003. “Congressional Enactments of Race-Gender: Toward a Theory of Race-Gendered Institutions.” American Political Science Review, 97 (4): 529-550. Hayes, Joseph and Laura Hill. 2013. “Just the Facts: Undocumented Immigrants.” Public Policy Institute of California. Published February 2013. Accessed April 20 th 2015. http://www.ppic.org/main/publication_show.asp?i=818. Hero, Rodney, F. Chris Garcia, John Garcia, and Harry Pachon. 2000. “Latino Participation, Partisanship, and Office Holding.” PS: Political Science and Politics 33 (3): 529-534. Ingram, Helen, Anne L. Schneider, and Peter DeLeon. 2007. "Social Construction and Policy Design." Theories of the Policy Process 2: 93-126. 158 Hero, Rodney E. and Caroline Tolbert. 1995. “Latinos and Substantive Representation in the U.S. House of Representatives: Direct, Indirect, or Nonexistent?” American Journal of Political Science, 39 (3): 640-652. Hondagneu-Sotelo, Pierrette. 2008. God’s Heart Has No Borders: How Religious Activists Are Working for Immigrant Rights. Berkeley: University of California Press. Hooks, bell. 1981. “Ain’t I a Woman: Black Women and Feminism.” Boston: South End Press. Huerta, Juan Carlos and Adolfo Santos. 2006. “Latino Representation in the US House: How Much and By Whom?” American Review of Politics 27: 115-128. Jacobson, Gary C., Morris P. Fiorina, and David W. Brady. 2003. Politics of Congressional Elections. Longman Publishing Group. Jones, Bryan D. 1973. “Competitiveness, Role Orientations, and Legislative Orientations.” Journal of Politics, Vol. 35, Issue 4, pp. 924-947. Kasinitz, Philip, John H. Mollenkopf, and Mary Waters. 2010. Inheriting the City: The Children of Immigrants Come of Age. New York: Russell Sage Foundation. Kerr, Brink and Will Miller. 1997. “Latino Representation, It’s Direct and Indirect.” American Journal of Political Science, 41 (3): 1066-1071. Kingdon, John. 1989. Congressmen’s Voting Decisions. University of Michigan Press. Kousser, J. Morgan. 1974. The Shaping of Southern Politics: Suffrage Restriction and the Establishment of the One-Party South, 1880-1910. New Haven: Yale University Press. Kymlicka, Will. 1993. “Group Representation in Canadian Politics.” In Equity and Community: The Charter, Interest Advocacy, and Representation, ed. F.L. Seidle. Montreal: Institute for Research on Public Policy. Lawless, Jennifer and and Richard L. Fox. 2010. It Still Takes a Candidate: Why Women Don’t Run for Office. New York: Cambridge University Press. Leal, David and Luis Ricardo Fraga. 2004. “Playing the Latino Card”: Race, Ethnicity, and National Party Politics. Dubois Review 1(2): 297-317. Leal, David and Jason Casellas. 2013. “Partisanship or Population? House and Senate Votes in the 109 th and 110 th Congress.” Politics, Groups, and Identities, 1 (1): 48-65. Lilley, Sandra. “Poll: 1 out of 3 Americans Inaccurately Think Most Hispanics are Undocumented.” NBC Latino. Published 09/12/2012. <http://nbclatino.com/2012/09/12/poll-1-out-of-3-americans-think 159 most-hispanics-are-undocumented/> Lublin, David. 1997. The Paradox of Representation: Racial Gerrymandering and Minority Interests in Congress. Princeton: Princeton University Press. Mansbridge, Jane. 1999. “Should Blacks Represent Blacks and Women Represent Women? A Contingent Yes.” Journal of Politics, 61 (3): 628-57. Mansbridge, Jane. 2003. ‘Rethinking Representation.” American Political Science Review, Vol. 97 (4): 515-528. Marimow, Anne. “Maryland Senate Approves In-State Tuition Bill.” Washington Post. Published March 14, 2011. http://www.washingtonpost.com/local/politics/maryland- senate- approves-in-state-tuition-bill/2011/03/14/ABxs2IW_story.html. Mariscal, George. 2005. Brown-eyed Children of the Sun: Lessons from the Chicano movement, 1965-1975. University of New Mexico Press. Marquez, Timothy and Scot Schraufnagel. 2013. “Hispanic Population and Growth and State Immigration Policy: An Analysis of Restriction (2008-12). Publius: The Journal of Federalism. 43 (3): 347-67. Medina, Jennifer. “Legislature in California Set to Pass a Dream Act.” The New York Times. Published August 13, 2011. http://www.nytimes.com/2011/09/01/us/politics/01dream.html. Mayhew, David R. 1974. Congress: The Electoral Connection. New Haven: Yale University Press. McConnell, Michael W. 2000. “The Redistricting Cases: Original Mistakes and Current Consequences.” Harvard Journal of Law and Public Policy, 24 (103): 104-117. Michelson, Melissa R. 2003. “Getting Out the Latino Vote: How Door-to-Door Canvassing Influences Voter Turnout in Rural Central California.” Political Behavior 25(3): 247-263. Miller, Warren and Donald Stokes.1963. “Constituency Influence in Congress.” American Political Science Review 57 (1): 45-56. Miller, Arthur H., Patricia Gurin, Gerald Gurin, and Oksana Malanchuck. 1981. “Group Consciousness and Political Participation.” American Journal of Political Science, 25(3): 494- 511. McDermott, Rose. 2002. “Experimental Methods in Political Science.” Annual Review of Political Science 5 (1): 31-61. Michelson, Melissa R. 2003. “Getting Out the Latino Vote: How Door-to-door Canvassing Influences Voter Turnout in Rural Central California.” Political Behavior 25 (3): 247-63. 160 Milkman, Ruth and Veronica Terriquez. 2012. “We Are the Ones Who Are Out in Front: Women’s Leadership in the Immigrant Rights Movement.” Feminist Studies, 38(3): 723-749. Minta, Michael D. 2011. Oversight: Representing the Interests of Blacks and Latinos in Congress. Princeton: Princeton University Press. Morton, Rebecca and Kenneth C. Williams. 2010. Experimental Political Science and Causality: From Nature to the Lab. New York: Cambridge University Press. Nicholls, Walter. 2013. The Dreamers: How the Undocumented Youth Movement Transformed the Immigrant Rights Debate. Stanford University Press. Nyhan, Brendan, Eric McGhee, John Sides, Seth Masket, and Steven Greene. 2012. "One Vote out of Step? The Effects of Salient Roll Call Votes in the 2010 Election." American Politics Research 40(5): 844-879. Panagopolous, Costas. 2011. “Thank You for Voting: Gratitude Expression, and Voter Mobilization.” Journal of Politics 73(3): 707-717. Page, Benjamin, Robert Shapiro, Paul Gronke, and Robert Rosenberg. 1984. “Constituency, Party, and Representation in Congress.” Public Opinion Quarterly 48(4): 741-756. Pantoja, Adrian D., Ricardo Ramirez, and Gary M. Segura. 2001. “Citizens by Choice, Voters by Necessity: Patterns in Political Mobilization by Naturalized Latinos.” Political Research Quarterly 54(4): 729-50. Parker, Frank R. 1990. Black Votes Count. Chapel Hill: University of North Carolina Press. Passel, Jeffrey S. and D’Vera Cohn. 2009. “A Portrait of Unauthorized Immigrants in the United States.” Pew Hispanic Center Report. 14 April. Poole, Keith T. and Howard Rosenthal. 1997. Congress: A Political-Economic History of Roll Call Voting. New York: Oxford University Press. Preuhs, Robert R. and Eric Gonzalez Juenke. 2012. “Irreplaceable Legislators? Rethinking Minority Representation in the New Century.” American Journal of Political Science 56(3): 705- 715. Preuhs, Robert R. 2005. “Descriptive Representation, Legislative Leadership, and Direct Democracy: Latino Influence on English Only Laws in the States, 1984-2002.” State Politics & Policy Quarterly 5: 203-224. Preuhs, Robert R. 2007. “Descriptive Representation as a Mechanism to Mitigate Political Backlash.” Political Research Quarterly, Vol. 60 2: 277-292. Prestage, Jewell. 1977. “Black Women State Legislators.” In a Portrait of Marginality: The Political Behavior of the American Woman, ed. Marianne Githens and Jewell Prestage. New York: David McKay. 161 Ramakrishnan, S. Karthick. 2005. Democracy in Immigrant America: Changing Demographics and Political Participation. Stanford: Stanford University Press. Ramakrishnan, S. Karthick and Thomas J. Espenshade. 2001. “Immigrant Incorporation and Political Participation in the United States.” International Migration Review 35(3): 870-909. Ramirez, Ricardo. 2013. Mobilizing Opportunities: The Evolving Latino Electorate. Charlottesville: University of Virginia Press. Reingold, Beth. 2000. Representing Women: Sex, Gender, and Legislative Behavior in Arizona and California. Chapel Hill: University of North Carolina Press. Reingold, Beth and Adrienne Smith. 2013. “Welfare Policymaking and Intersections of Race, Ethnicity, and Gender in U.S. State Legislatures.” American Journal of Political Science 56 (1): 131-47. Reingold, Beth, and Kerry L. Haynie. 2014. “Representing Women’s Interests and Intersections of Race, Gender, and Ethnicity in U.S. State Legislatures.” In Representation: The Case of Women, ed. Maria C. Escobar-Lemmon and Michelle M. Taylor-Robinson. New York: Oxford University Press. Riach, Peter A. and Judith Rich. 2004. “Deceptive Field Experiments of Discrimination: Are They Ethical?” Kyklos 57(3): 457-470. Rincon, Alejandra. 2010. Undocumented Immigrants and Higher Education. LFB Scholarly Publishing LLC. Roberts, Sam. 2007. “In the U.S. Name Count, Garcias are Catching up with the Joneses.” The New York Times. Published November 17, 2007. http://www.nytimes.com/2007/11/17/us/17surnames.html?pagewanted=print. Rocha, Rene R., and Tetsuya Matsubayashi. 2013. "Latino immigration and representation in local politics." Urban Affairs Review 49(3): 353-380. Romero, Mary. 2008. "The Inclusion of Citizenship Status in Intersectionality: What Immigration Raids Tells Us About Mixed-Status Families, the State, and Assimilation." International Journal of Sociology of the Family 34(2):131-152. Rouhani, Setareh. 2014. “Intersectionality-informed Quantitative Research: A Primer.” Institute for Intersectionality Research and Policy. Rosenthal, Cindy. 1998. When Women Lead: Integrative Leadership in State Legislatures. New York: Oxford University Press. Rouse, Stella. 2013. Latinos in the Legislative Process: Interests and Influence. Cambridge: Cambridge University Press. 162 Rumbaut, Rubén G. 2004. "Ages, Life Stages, and Generational Cohorts: Decomposing the Immigrant First and Second Generations in the United States." International Migration Review 38 (3): 1160-1205 Sanchez, Gabriel R. and Natalie Masuoka. 2010. “Brown-Utility Heuristic: The Presence and Contributing Factors of Linked-Fate.” Hispanic Journal of Behavioral Scientists 32(4): 519-531. Schmidt Sr., Ronald, Yvette Alex-Assensoh , Andrew L. Aoki, and Rodney E. Hero. 2010. Newcomers, Outsiders, and Insiders: Immigrants and American Racial Politics in the Early 21 st Century. Ann Arbor: University of Michigan Press. Seif, Hinda. 2004. “Wise Up! Undocumented Latino Youth, Mexican American Legislators, and the Struggle for Higher Education Access.” Latino Studies 2(2): 210-230. Seif, Hinda. 2011. "Unapologetic and Unafraid”: Immigrant Youth Come Out from the Shadows." New directions for child and adolescent development Vol. 134, pp. 59-75. Serra, George, and Albert D. Cover.1992. "The Electoral Consequences of Perquisite Use: The Casework Case." Legislative Studies Quarterly, 233-246. Serra, George, and David Moon. 1994. "Casework, issue positions, and voting in congressional elections: A district analysis." The Journal of Politics 56(1): 200-213. Simien, Evelyn. 2006. Black Feminist Voices in Politics. Albany: State of University of New York Press. Smooth, Wendy. 2006. "Intersectionality in Electoral Politics: A Mess Worth Making." Politics & Gender 2(3): 400-414. Strolovich, Dara. 2007. Affirmative Advocacy: Race, Class, and Gender in Interest Group Politics. Chicago: University of Chicago Press. Swain, Carol. 1995. Black Faces, Black Interests: The Representation of African Americans in Congress. Harvard University Press. Swers, Michele L. 1998. “Are Women More Likely to Vote for Women’s Issue Bills than Their Male Counterparts?” Legislative Studies Quarterly 23(3): 435-448. Swers, Michele. 2005. “Connecting Descriptive and Substantive Representation: An Analysis of Sex Differences in Cosponsorship Activity. Legislative Studies Quarterly 30(3): 407-433. Swers, Michele L. and Stella Rouse. 2011. “Descriptive Representation: Understanding the Impact of Identity on Substantive Representation of Group Interests. In The Oxford Handbook of the American Congress, ed. Eric Shickler and Francis E. Lee. New York: Oxford University Press. Tarrow, Sidney. 1998. Power in Movement: Social Movements and Contentious Politics. New 163 York: Cambridge University Press. Tate, Katherine.1998. From Protest to Politics: The New Black Voters in American Elections. Cambridge: Harvard University Press. Tate, Katherine. 2003. Black Faces in the Mirror: African Americans and their Representatives in the United States Congress. Princeton: Princeton University Press. Thomas, Sue. 1991. “The Impact of Women on State Legislative Policies.” Journal of Politics. 53(4): 958-976. Truman, David B. 1951. The Governmental Process. New York: Knopf. Top Baby Names in Mexico. 2010. Baby Name Facts. http://www.babynamefacts.com/popularnames/countries.php?country=MEX. U.S. Census Bureau, 2009 American Community Survey, B04006, “People Reporting Ancestry,” http://factfinder.census.gov, January 2011. Verba, Sidney and Norman Nie. 1972. Participation in America: Political Democracy and Social Equality. New York: Harper and Row. Verba, Sidney, Kay L. Schlozman, and Henry E. Brady. 1995. Voice and Equality: Civic Voluntarism in American Politics. Cambridge: Harvard University Press. Wallace, Sophia, Chris Zepeda-Millán, and Michael Jones-Correa. “Spatial and Temporal Proximity: Examining the Effects of Protests on Political Attitudes.” American Journal of Political Science 58(2): 433-448. Wallace, Sophia. 2014. “Representing Latinos: Examining Descriptive and Substantive Representation in Congress.” Political Research Quarterly, published online July 2014, pp. 1- 13. DOI: 10.1177/1065912914541795 Welch, Susan and John R. Hibbing. 1984. “Hispanic Representation in the U.S. Congress.” Social Science Quarterly 65(2): 328-335. Whistler, Donald E., and Mark C. Ellickson. 1999. "The incorporation of women in state legislatures: a description." Women & Politics Vol. 20, Issue 2, pp. 81-97. Whitby, Kenny. 1997. The Color of Representation. Ann Arbor: University of Michigan Press. Williams, Melissa S.1998. Voice, Trust, and Memory: Marginalized Groups and the Failings of Liberal Representation. Princeton: Princeton University Press. Wilson, James Q. [1974] 1995. Political Organizations. Princeton: Princeton University Press. 164 Wilson, Walter C. 2013. “Latino Congressional staffers and policy responsiveness: an analysis of Latino interest agenda setting.” Politics, Groups, and Identities, Vol.1, No. 2, pp. 164-180. Wong, Janelle. 2006. Democracy’s Promise: Immigrants and American CivicInstitutions. Ann Arbor: University of Michigan Press. Wong, Janelle and Vivian Tseng. 2008. “Political Socialization in Immigrant Families.” Journal of Ethnic and Migration Studies, Vol. 34, Issue 1, pp. 151-168. 28. Yanow, Dvora. ed. 2000. Conducting interpretive policy analysis. Sage. Yanow, Dvora. 2003. "Interpretive empirical political science: What makes this not a subfield of qualitative methods." Qualitative Methods Vol.1, Issue 2, pp. 9-13. Young, Iris Marion. 1990. Justice and the Politics of Difference. Princeton: University of Princeton Press. Young, Iris Marion. 1997. Intersecting Voices: Dilemmas of Gender, Political Philosophy, and Policy. Princeton: Princeton University Press. Zepeda-Millan, Chris. 2014. “Perceptions of Threat, Demographic Diversity, and the Framing of Illegality: Explaining (non)Participation in New York’s 2006 Immigrant

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Achieving the dream: undocumented Central American Latinas in college

PDF

La eleccion de la pandemia: analyzing Latino political behavior during the 2020 election

PDF

Working with la familia: a study of family work relations among Latina/o children and adolescents who work with their parents as street vendors in Los Angeles

PDF

Finding home: the migration and incorporation of undocumented, unparented Latino youth in the US

PDF

The representation of LGBT interests in the United States: a multimethod approach

PDF

Descriptive representation in a multi-racial America

PDF

The political representation of Kurdish, Kemalist, and conservative Muslim women in Turkey (1990-2010)

PDF

Translating race, class, and immigrant lives: the family work of children language brokers

PDF

Hope and higher education: undocumented students and the legacy of U.S. immigration policy

PDF

Undocumented student organizations: navigating the sociopolitical context in higher education

PDF

Pernicious racism and the impacts of intentional racist in a color-blind society

PDF

The role of skin tone on candidate evaluation: an analysis of racial socialization among Latino immigrants

PDF

Changing political attitudes and behavior in a diverse America: incorporating individual and contextual determinants

PDF

Voices unheard, stories untold: Black women, police violence and political participation

PDF

Colonial brews: café and power in the Américas

PDF

Immigrants at a loss: the need for services that promote child well-being among Latino families with child welfare contact

PDF

Politics is something we do together: identity and institutions in U.S. elections

PDF

The role of student affairs professionals: serving the needs of undocumented college students

PDF

Can election reforms improve democracy?

PDF

Examining parent involvement activities in two mmigrant-impacted schools: a comparative case study

Asset Metadata

Creator Mendez, Matthew Synclair (author)

Core Title In/visible constituents: the representation of undocumented immigrants

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Politics and International Relations

Publication Date 07/30/2016

Defense Date 05/13/2015

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

Tag empirical intersectionality,field experiments,Latino politics,OAI-PMH Harvest,representation,undocumented immigrants

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Hancock, Ange-Marie (committee chair), Grose, Christian R. (committee member), Terriquez, Veronica (committee member)

Creator Email matthew.mendez85@gmail.com,msmendez@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-622883

Unique identifier UC11303902

Identifier etd-MendezMatt-3783.pdf (filename),usctheses-c3-622883 (legacy record id)

Legacy Identifier etd-MendezMatt-3783.pdf

Dmrecord 622883

Document Type Dissertation

Format application/pdf (imt)

Rights Mendez, Matthew Synclair

Type texts

Source 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 a...

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