University of Southern California Dissertations and Theses

Campaign financing for the U.S. House of Representatives: an interactive Web map

(USC Thesis Other)

Campaign financing for the U.S. House of Representatives: an interactive Web map

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content i

CAMPAIGN FINANCING FOR THE U.S. HOUSE OF REPRESENTATIVES:

AN INTERACTIVE WEB MAP

by

Jamen Dean Underwood

A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(GEOGRAPHIC INFORMATION SCIENCE AND TECHNOLOGY)

December 2015

Copyright 2015 Jamen Dean Underwood

ii

iii
DEDICATION
This thesis is dedicated to all the campaign finance reformers and outside the box
thinkers.

iv
ACKNOWLEDGMENTS
There are many people who have supported and inspired my work on this thesis.
Many thanks to my thesis committee Dr. Yao-Yi Chiang, Dr. Robert Vos, and Dr.
Daniel Warshawsky for their valuable feedback. As my Committee Chair, Dr. Chiang
has been especially patient and supportive. Dr. Mitra Parineh and Dr. Mandy
Hobmeier gave me feedback on my thesis drafts and helped to greatly improve my
writing. The faculty and staff of the SSI have been so great to work with throughout
my Master’s degree journey. What a wonderful opportunity it has been to learn from
them.
I would like to thank Randy Steiner for taking time to get me through a
number of hurdles in my map code. He gave me great help in debugging and made
up for my limited web development experience. I would like to thank Mike Rothwell
who helped me figure out how to use table views.
Finally, I could not have completed my thesis or degree without the support
and sacrifice of my wife, Lisa, and my children. Lisa spent many days keeping it all
together minus a husband. She encouraged me to keep going and do difficult things.
I am inspired by my kids smiles and hugs, and look forward to spending more time
with them.

v
TABLE OF CONTENTS
DEDICATION ................................................................................................................... iii
ACKNOWLEDGMENTS ................................................................................................. iv
TABLE OF CONTENTS .................................................................................................... v
LIST OF TABLES ............................................................................................................ vii
LIST OF FIGURES ......................................................................................................... viii
LIST OF ABBREVIATIONS ............................................................................................. x
ABSTRACT ....................................................................................................................... xi
CHAPTER ONE: INTRODUCTION ................................................................................. 1
1.1 Motivation ................................................................................................................. 2
1.2 Thesis Contribution ................................................................................................... 6
1.3 Web map Overview .................................................................................................. 6
CHAPTER TWO: BACKGROUND AND RELATED WORK ........................................ 9
2.1 Campaign Finance Regulatory Structure .................................................................. 9
2.1.1 Regulatory Agencies ........................................................................................ 10
2.1.2 Classification and regulation of contributions and expenditures ..................... 10
2.1.3 Outside Groups ................................................................................................ 14
2.2 Campaign Finance Law .......................................................................................... 16
2.3 Review of Websites and Visualizations .................................................................. 21
2.4 Choice of Technologies .......................................................................................... 27
CHAPTER THREE: METHODOLOGY ......................................................................... 28
3.1 Workflow for Campaign Finance Web Map .......................................................... 28
3.2 Data Download ....................................................................................................... 29

vi
3.3 Create Database Diagram ....................................................................................... 31
3.4 Format Data ............................................................................................................ 33
3.5 Create Database in PostGIS .................................................................................... 34
3.6 Upload Data to GeoServer ...................................................................................... 37
3.7 Build Web Application Using Leaflet .................................................................... 39
CHAPTER FOUR: EVALUATION................................................................................. 50
4.1 Background Knowledge ......................................................................................... 50
4.2 Comparison to the FEC website ............................................................................. 52
4.2.1 Contributions Comparison ............................................................................... 53
4.2.2 Independent Expenditures Comparison ........................................................... 55
4.2.3 General Comparison ........................................................................................ 57
4.3 Web Map Feedback ................................................................................................ 60
CHAPTER FIVE: DISCUSSION AND CONCLUSION ................................................ 62
5.1 Results of the Campaign Finance Web Map .......................................................... 62
5.2 Web Map Improvements ........................................................................................ 63
5.3 Future Work ............................................................................................................ 64
REFERENCES ................................................................................................................. 66
APPENDIX A: GOOGLE EVALUATION SURVEY .................................................... 70

vii
LIST OF TABLES

Table 1 Contribution Limits for 2013-2014 11

viii
LIST OF FIGURES

Figure 1 Map Screen Shot 7

Figure 2 Map of Campaign Contributions for Individual Legislators 22

Figure 3 Per Capita Contributions 24

Figure 4 Example of FEC House and Senate Campaign Finance Map 25

Figure 5 Political Moneyball by The Wall Street Journal 25

Figure 6 Workflow Diagram 29

Figure 7 Database Diagram 32

Figure 8 PostGIS Database 35

Figure 9 Layers Published Directly from the Database 38

Figure 10 SQL View for Independent Expenditures Layer 39

Figure 11 User Input Controls 40

Figure 12 Map with Base Layer 41

Figure 13 Data Flow Diagram 42

Figure 14 Map State Selected 43

Figure 15 Map with Congressional District Selected 43

Figure 16 Map with Candidate Selected 44

Figure 17 Overlapping Markers Before and After Clicking 45

Figure 18 Popup with Contribution Information 46

Figure 19 List of Committees During User Input 47

Figure 20 Expenditures Displayed After a Committee was Selected 47

Figure 21 Popup with Expenditure Information 48

Figure 22 Chart of Survey Question #1 Responses 51

ix

Figure 23 Chart of Survey Question #2 Responses 51

Figure 24 Chart of Survey Question #3 Responses 52

Figure 25 Graph of Survey Question #4 Responses 53

Figure 26 Graph of Survey Question #6 Responses 54

Figure 27 Graph of Survey Question #8 Responses 54

Figure 28 Graph of Survey Question #5 Responses 55

Figure 29 Graph of Survey Question #7 Responses 56

Figure 30 Graph of Survey Question #9 Responses 57

Figure 31 Graph of Survey Question #10 Responses 58

Figure 32 Graph of Survey Question #11 Responses 59

Figure 33 Graph of Survey Question #12 Responses 59

x
LIST OF ABBREVIATIONS
BCRA Bipartisan Campaign Reform Act of 2012
CRP Center for Responsive Politics
CSS Cascading Style Sheets
CSV Comma Separated Value
FEC Federal Election Commission
FECA Federal Election Campaign Act
FIPS Federal Information Processing Standards
GIST Geographic Information Science and Technology
IRS Internal Revenue Service
OMS Overlapping Marker Spiderfier
PAC Political Action Committee
SIC Standard Industrial Classification
SCOTUS Supreme Court of the United States
SQL Structured Query Language
SSI Spatial Sciences Institute
UI User Interface
USC University of Southern California
USPS United States Postal Service
ZCTA Zip Code Tabulation Area

xi
ABSTRACT
It is expensive to get elected to the U.S. House of Representatives, and in the past several
decades the increase in spending has been very steep. In 2012, candidates spent an
average of nearly $1.2 million (Ornstein, et al 2013). However, that includes only direct
candidate or party expenditures, and does not included money spent by outside (i.e.,
“independent”) groups. Lessig (2011) argues that the way campaigns are funded, and the
dependence members of Congress have on a relatively small number of donors is a form
of corruption in our political system. This thesis produces an interactive web map
showing the geographic distribution of campaign contributions and independent
expenditures made for members of the U.S. House of Representatives. Campaign finance
data are most commonly displayed in tables and graphs. They are useful and important
for those seeking to investigate the details of campaigns or needing to answer specific
question, but a map is more accessible and engaging for the general public. There are
numerous other visualizations available on the internet, but many have not been updated
since 2012 election cycle (or earlier), or may not include all sources of spending. The
web map created as a part of this thesis enables a user to select a candidate and view
contributions summed by zip code using graduated symbols. The geographic origin of
contributions is apparent, whether within or outside the district. A user can also search for
groups that made independent expenditures and see the congressional districts where
money was spent. An evaluation of the web map by a small sample of people showed the
effectiveness of visualizing campaign finance data to better inform the public about
money used in elections.

1

CHAPTER ONE: INTRODUCTION
Successful candidates seeking office at the federal level are required to raise large sums
of money to run their campaign organizations and communicate with voters through mass
media. Most voters probably do not think much about how that money is raised or where
it comes from, but it is central to the way our political system functions (Lessig 2011).
Politicians at the federal level spend a great deal of time fundraising and interacting with
donors from across the country (Cho and Gimpel 2007). Campaign finance regulation
requires candidates to disclose contributions (Briffault 2007). The public can then be
informed about money raised and spent in elections, but it requires some thought on how
to effectively communicate the complexities of the campaign finance system.
The best way to make campaign finance data accessible to voters is through data
visualization. The visual representation of data is a simple, efficient, and powerful form
of communication because large amounts of data can be condensed in a way that utilizes
the brain’s capacity for processing visual information and recognizing patterns (Krum
2013). Ideally, a visualization of campaign finance data should engage people in a way
that motivates them to be more involved in the political process.
This thesis project aims to visualize campaign contributions and independent
expenditures through the use of an interactive web map showing direct contributions to
candidates for the U.S. House of Representatives, as well as spending by outside groups.
Although data for Senate candidates could have also been mapped, the House seemed
better suited for a thesis project because all seats are up for election during a two year
election cycle. Senators are elected every six years and only a portion of them are up for
re-election in a given two year election cycle. There are simply more candidates and

2
more data to examine for the House, and it is easier to compare contributions and
spending for different candidates.
1.1 Motivation
It has been argued that the increasing amount of money spent on elections undermines the
integrity of our political system, but not in a way that the ties between candidates and
funders are easily perceived (Lessig 2011). Those seeking to be elected to Congress are
often dependent on a relatively small number of wealthy donors for much of their funds;
raising doubts about whose interests are being served. In the 2010 election cycle 48% of
contributions were from those who gave $200 or more, but the number of people making
those contributions were less than 0.5% of country’s population (Center for Responsive
Politics 2010). This type of corruption is not blatant bribery or quid pro quo, as it has
sometimes been in the past, and its effects are subtle.
There are few instances of a member of Congress taking money for voting a
certain way or supporting specific legislation, but it does happen occasionally. Two
recent examples of quid pro quo corruption were Representative Randall “Duke”
Cunningham of California and Representative William J. Jefferson of Louisiana.
Cunningham was convicted on bribery charges in 2006 after taking $2.4 million in
exchange for assistance in acquiring Defense Department contracts. Jefferson similarly
sought and took large bribes and was convicted on corruption charges in 2009 (Lessig
2011). The few legislators that do break the law do not improve people’s view of
politicians.
Although outright bribery is not the norm in Washington D.C, many people feel
that large campaign contributions are close to bribes. A number of polls have shown a

3
clear majority of people perceive large contributions as a corrupting influence and that
members of Congress give preference to contributors (Persily and Lammie 2004). The
public consensus is that campaign contributions and money from lobbyists buys access
and influence. Lessig (2011) argues that wealthy donors or potential donors have a much
easier time getting their views heard by members of Congress. It is difficult to measure
the exact effects that contributions have on gaining access, but one field experiment has
shown that when individuals identify themselves as political donors they are three to four
times as likely to meet with the Congressperson or his or her Chief of Staff compared to
those who identify themselves as constituents (Kalla and Broockman 2014). More
evidence is needed to prove conclusively that campaign contributions are corrupting, but
they certainly appear to have some affect.
Burke (1997) refers to the influence of campaign contributions on policymaking
as distortion corruption because contributions do not represent the views of the broader
public. In other words the politicians’ policy views are more closely aligned with their
donors than their constituents (Stephanopolous 2014). This distortion may be one reason
why the public has such a low opinion of Congress. In polls conducted by Gallup
between January and August 2014 the average Congressional approval rating was 14%
(Jones 2014). Voters may feel that Congress is not listening to them, but is listening to
those giving them money.
Most people have the perception that campaign contributions influences
politicians’ behavior, whether or not there is direct evidence of this. There is a lack of
trust in our government because of the presence of so much money in the political
process (Lessig 2011). One aspect of this mistrust is the difficulty of tracking campaign

4
finance. It’s hard to really know what is going on without becoming familiar with all the
details. Complex regulations, outside spending, and the flow of huge sums of money
make the whole system challenging to understand.
While it is important to be explicit when regulating money in politics, the
complexity of the regulations creates a lot of jargon. Take the example of political action
committees (PACs). They are groups independent of political parties or candidates that
raise and spend money to influence elections through advertisements and direct
contributions to candidates and parties. Federal election law limits how much individuals
may contribute to PACs and how much PACs may contribute to candidates and parties
(Center for Responsive Politics 2014b). The Federal Election Commission (FEC)
classifies a PAC as being either a separate segregated fund (SSF) or a non-connected
PAC. An SSF has a sponsoring organization such as a corporation or labor union, while a
non-connected PAC does not. They also have different requirements for reporting
administrative expenses and how they conduct fundraising (FEC 2014b). The term PAC
applies to many different groups, but the distinction between SSF and non-connected
PAC is not often discussed.
Money used in elections is classified by whether or not it is subject to federal
campaign finance regulation. Contributions made directly to candidates are referred to as
“hard money” and are subject to regulation. Contributions made to party committees and
outside interests groups involved in federal elections is referred to as “soft money”
(Center for Responsive Politics 2014a). Much of the legislation, court rulings, and debate
regarding campaign finance since 2000 has focused on soft money and other outside
spending. Despite efforts to curb outside spending it increased dramatically between 2002

5
and 2010; going from $27.7 million to $309.8 million (Center for Responsive Politics
2014e). The increase in spending has increased the complexity of campaign finance.
More types of organizations have been allowed to spend more money (Tokaji and Strause
2014).
People usually encounter campaign finance data through news media. It is likely
that most do not examine it any further. For those that do there are a number of online
resources. Contributions and expenditures reported to the FEC by committees and outside
groups are available on the FEC website. They are searchable, but are mostly viewed in
tabular form. For example, a person could search for a particular candidate and see a
table of all his or her contributions as well as the sum of the contributions. A number of
organizations, including the Center for Responsive Politics and the Sunlight Foundation,
are working to inform citizens about money in politics. Many of them produce high
quality visualizations; primarily tables, charts, and graphs. The Sunlight Foundation
created an excellent series of maps depicting political contributions by county (Sibley,
Lannon, and Chartoff 2013). The advantages and disadvantages of these resources are
discussed further in Chapter 2. It is the objective of this thesis project to complement the
work already being done by showing both contributions to candidates and outside
spending in a single map. Having both these sources of money in the same visualization
allows for easier comparison and may provide new insights into campaign finance.
Spatial analysis of campaign contributions has proven useful in revealing patterns
of participation in politics. Studies indicate that while wealth is a factor in the distribution
of contributions, local social networks are also important in both volunteer involvement
in political campaigns and donations to them (Cho and Gimpel 2010; Gimpel, Lee, and

6
Kaminski 2006). Analysis through tables or graphs, such as a table listing the sum of
contributions from census tracts, would not have found these spatial relationships. A web
map showing the distribution of contributions and independent expenditures will be more
interesting to the general public than other visualizations because the data can be easily
associated with a physical location. Seeing the spatial patterns will provide new insight,
and tell the story of campaign finance in a impactful way.
1.2 Thesis Contribution
The intent of this project is to make campaign financing more comprehensible
(than traditional tabular forms or static maps) in a visual way. The web map provides an
interactive tool for users to understand the geographic attributes of campaign financing.
The main contribution includes the visualization tools for understanding:
- the locus of money raised by candidates and
- where outside groups spend money to influence elections
1.3 Web map Overview
The interactive web map created as part of this thesis project shows the
geographic distribution of campaign finance data for candidates of the U.S. House of
Representatives in the 2014 election cycle. Figure 1 shows an example of a candidate’s
contribution data. The development process of the map is further discussed in Chapter 3
including data downloading and formatting, database design and creation, layer
publishing, and application coding. The map supports basic navigation functions such as
panning and zooming that have become common to web maps since the advent of Google
Maps. It is designed to be easily navigated and responsive to user input.

7

Figure 1 Map Screen Shot

The two kinds of data the map can display are direct contributions to candidates
and independent expenditures. For the direct contributions users select a state by clicking
on the state button. Then radio buttons with the congressional districts for that state
appear, and the user clicks the button for a congressional district to see a list of
candidates to choose from. Once a candidate is selected, the contributions are
summarized by zip code and displayed with symbols sized according to the amount
donated. The user can then click on a symbol to get more information about the donations
from a given zip code.

8
Independent expenditures are viewed by selecting an organization from a
dropdown list. The expenditures are displayed by Congressional District. The user can
then click on a symbol to see which candidates the organization was supporting or
opposing and how much money they spent. Viewing expenditure data on the map is a
very efficient way of finding out what races a committee was seeking to influence.
Campaign finance data are most commonly displayed in tables and graphs. They
are useful and important for those seeking to investigate the details of campaigns or need
to answer a specific question, but a map may be more accessible and engaging for the
general public. Furthermore, the interactivity provided by the technologies used in the
web map allow the user to easily control what data are displayed and enable in depth
exploration of the data.
This chapter discussed the motivation, impact, and general functions of the web
map created for this thesis project. The remainder of this thesis is organized into four
chapters. Chapter 2 examines the background of campaign finance regulation and law as
well as campaign finance data visualizations. Chapter 3 presents the method for building
the interactive campaign finance web map. Chapter 4 describes the evaluation of the web
map by a sample of volunteers. Chapter 5 concludes and

9
CHAPTER TWO: BACKGROUND AND RELATED WORK
The first section of this chapter provides an explanation of campaign finance regulatory
agencies, classifies contributions and expenditures, and outlines the types of groups that
attempt to influence elections. Section 2.2 presents an overview of the history campaign
of finance regulation at the federal level and discusses changes due to recent court
decisions. Section 2.3 describes various online visualizations of campaign finance data.
2.1 Campaign Finance Regulatory Structure
Running for public office almost always requires some money. Candidates may choose to
self-fund their campaigns, especially those that are very wealthy. In most cases however
candidates rely on contributions from other people to finance the costs of campaigns. It is
expensive to get elected to the U.S. House of Representatives, and in the past several
decades the increase in spending has been very steep. Between 1982 and 2010 total
spending by Senate and House candidates rose from $343 million to $1.8 billion (Garrett
2011). In the 2012 election, House candidates spent an average of nearly $1.2 million
(Ornstein, et al 2013). But that includes only direct candidate or party expenditures, and
does not include money spent by outside (i.e. “independent”) groups. Much of the
expense of campaigns can be attributed to television and other media advertisements as
well as fundraising costs (Cantor 2009). The natural tendency for those who are wealthy
to have or seek political influence requires some intervention to prevent the corruption of
a democratic political system. The campaign finance regulatory structure attempts to do
this in a number of ways, but its complexity requires some explanation.

10
2.1.1 Regulatory Agencies
There are two agencies that are involved with campaign finance regulation. The primary
agency is the Federal Election Commission (FEC), which enforces campaign finance
laws, discloses campaign finance information, and oversees public funding of
Presidential elections (FEC). Most political organizations are required to report
contributions and expenditures to the FEC, but loopholes in campaign finance law create
some important exceptions discussed in section 2.1.3.
The Internal Revenue Service (IRS) determines the taxes paid by political
organizations under Section 527 of the Internal Revenue Code (IRC). A political
organization’s primary function must be “influencing or attempting to influence the
selection, nomination, election or appointment of an individual to a federal, state, or local
public office or office in a political organization” to receive tax exemptions under Section
527 (IRS 2015). The IRS and the FEC have different functions, but are both important to
regulating political fundraising and spending.
2.1.2 Classification and regulation of contributions and expenditures
The main tool the FEC has for regulating campaign finance is disclosure. Disclosure is a
two-step process in which contributions and expenditures are reported to the FEC then
the information is made accessible to the public (Briffault 2010). Contributions include
“gifts of money, goods and services (in-kind contributions), loans (other than bank loans
meeting certain conditions), and guarantees or endorsements of bank loans” (FEC 2013).
The broad range of contributions must be carefully tracked for accurate record keeping.
Committees report all contributions to the FEC and are required to provide the names and
addresses of individuals who contributed over $200 in a calendar year (FEC 2013).

11
Individuals and PACs may contribute a maximum of $2600 and $5000 respectively to
each candidate per election (e.g. primary and general). There are also limits on how much
may be given to party committees, see Table 1. Disclosure of contributions ensures that
these limits are maintained.
Table 1 Contribution Limits for 2013-2014 (Center for Responsive Politics 2014b)
To each
candidate or
candidate
committee
per election
To national
party
committee
per
calendar
year
To state,
district &
local party
committee
per calendar
year
To any
other
political
committee
per
calendar
year
1

Special
Limits
Individual may
give
$2,600* $32,400* $10,000
(combined
limit)
$5,000 None
National Party
Committee may
give
$5,000 No limit No limit $5,000 $45,400* to
Senate
Candidates
per
campaign
3

State, District &
Local Party
Committee may
give
$5,000
(combined
limit)
No limit No limit $5,000
(combined
limit)
None
PAC
(multicandidate)
4

may give
$5,000 $15,000 $5,000
(combined
limit)
$5,000 None
PAC (not
multicandidate)
may give
$2,600* $32,400* $5,000
(combined
limit)
$5,000 None

12
Authorized
Campaign
Committee may
give
$2,000
5
No limit No limit $5,000 None
Source: Center for Responsive Politics, http://www.OpenSecrets.org
* These contribution limits are increased for inflation in odd-numbered years.
(1) A contribution earmarked for a candidate through a political committee counts
against the original contributor’s limit for that candidate. In certain circumstances, the
contribution may also count against the contributor’s limit to the PAC. 11 CFR 110.6.
See also 11 CFR 110.1(h).
(2) No more than $46,200 of this amount may be contributed to state and local party
committees and PACs.
(3) This limit is shared by the national committee and the national Senate campaign
committee.
(4) A multicandidate committee is a political committee with more than 50 contributors
which has been registered for at least 6 months and, with the exception of state party
committees, has made contributions to 5 or more candidates for federal office. 11 CFR
100.5(e)(3).
(5) A federal candidate's authorized committee(s) may contribute no more than $2,000
per election to another federal candidate's authorized committee(s). 11 CFR
102.12(c)(2).
An expenditure is broadly defined as “a purchase, payment, distribution, loan,
advance, deposit, or gift of money or anything of value made for the purpose of
influencing a federal election (FEC 2013, 162).” Voters are most familiar with
expenditures directed at them in the form of TV, radio, and internet advertisements,
mailers, signs, and bumper stickers. Expenditures may be classified as either coordinated
with a candidate’s campaign or independent of the candidate. Independent expenditures
are specifically for communications that expressly advocate the “election or defeat of a
clearly identified candidate and which is not made in cooperation, consultation, or
concert with, or at the request or suggestion of any candidate, or his or her authorized

13
committees or agents, or a political party committee or its agents” (Code of Federal
Regulations Title 11). The definition of independent expenditures is clearly intended to
prevent candidates from using contributions to outside groups to augment their own
campaign spending.
Independent expenditures then fall into the category of express advocacy, defined
as communications that “explicitly advocate for the defeat or election of a clearly
identified federal candidate” (Center for Responsive Politics 2014a). Obviously
candidates are likely to use express advocacy themselves since the whole function of
their campaigns is to ask constituents to vote for them. In Buckley v. Valeo the Supreme
Court gave examples of language that constituted express advocacy including “ ‘vote
for,’ ‘elect,’ ‘support,’ ‘cast your ballot for,’ ‘Smith for Congress,’ ‘vote against,’
‘defeat,’ ‘reject’ ”(Briffault 2011). Such phrases are a very narrow form of speech and it
is not difficult to discuss a candidate without using them. Communications that do not
meet the standard of express advocacy are treated as issue advocacy, which focuses on a
particular matter voters may be concerned about such as gun control or abortion (Center
for Responsive Politics 2014a). Advertisements may praise or criticize candidates
without directly calling for their election or defeat and avoid being regulated by
campaign finance limits (Briffault 2011). Increasing spending on issue advocacy to
influence elections during the 1990s eventually led Congress to attempt to limit this form
of political speech.
The Bipartisan Campaign Reform Act of 2002 (BCRA) implemented a number of
changes in campaign finance law, and sought to regulate issue advocacy spending by
creating a new class of communications called electioneering communications (Briffualt

14
2011). These are television or radio advertisements that identify a federal candidate, are
directed at least 50,000 members of the candidate’s electorate, and “air within 30 days of
a primary election or 60 days of a general election” (Center for Responsive Politics
2014a). The BCRA prohibited the use of unregulated soft money from outside group for
electioneering communications (Center for Responsive Politics 2014a). The BCRA did
not go unchallenged and court rulings have led to significant weakening of its reforms.
Section 2.2 provides further discussion of the court decisions and their consequences.
2.1.3 Outside Groups
There are several different types of outside groups that try to influence elections.
Traditional PACs are those that contribute to candidate’s campaigns; some of which also
make independent expenditures. Individuals may contribute up to $5000 per year to a
PAC. FEC regulations permit a PAC to contribute $5000 per election to a candidate as
well as $15000 annually to a national political party (FEC 2013). Committees called
Super PACs, or independent expenditure-only committees, sound similar to traditional
PACs, but operate very differently. They are allowed to spend unlimited funds on
elections as long as it is not in coordination with any candidate and they do not make any
contributions directly to candidates (Center for Responsive Politics 2014a). Super PACs
have become major players in elections, but they are not the only kind of group that has
been used to avoid FEC regulations since the BCRA.
While the IRS broadly defines political organizations under section 527of the
IRC, groups commonly referred to as 527s are a subset of organizations that operate
outside of FEC regulations. Many 527 organizations are focused on state or local
elections. Others get involved in federal elections, but do not claim it as their main

15
purpose (Tokaji and Strause 2014). They avoid reporting to the FEC by not expressly
advocating for or against candidates (Center for Responsive Politics 2014a). 527
organizations were prominent during the 2004 presidential election, but have been less
influential since that time.
Groups known as political nonprofits or 501(c)s receive tax-exemptions under
section 501(c) of the IRC. There are three subtypes of 501(c)s that may make
expenditures and contributions as long as it is not their primary purpose. There is some
ambiguity in determining their primary purpose, but it must be less than half of their
activities (Center for Responsive Politics 2014a). 501(c)(4)s are advocacy groups that
promote “social welfare” (Tokaji and Strause 2014). Such groups include the National
Rifle Association (NRA), the Sierra Club, Crossroads GPS, and Patriot Majority (Center
for Responsive Politics 2014d). 501(c)(5)s are labor unions and agricultural groups.
501(c)(6)s are chambers of commerce and trade associations (Center for Responsive
Politics 2014a). The different classifications allow additional avenues for a wide range of
interest groups to influence elections.
Political nonprofits have become increasingly controversial because they are not
required to disclose their donors and the 2010 Supreme Court Citizens United ruling
allows them to spend unlimited funds. 501(c)4 and 501(c)6 groups are seen as tools for
pooling corporate resources while hiding who is seeking to influence elections (Briffault
2010). Also, some Super PACs receive contributions funneled through 501(c)s, allowing
the Super PAC to hide the identity of the true source of the money.

16
2.2 Campaign Finance Law
Campaign finance regulation has long been a part of the U.S. political system. The first
regulations requiring some level of campaign finance disclosure at the federal level were
enacted early in the 20th century. The 1907 Tillman Act banned candidates from
accepting contributions from banks and corporations for use in general elections, and the
first limited disclosure requirements were enacted in 1910 (Tokaji and Strause 2014).
They were created among other Progressive Era reforms that sought to curb the power of
corporations and special interests to influence politics and had corrupted many areas of
government (Lessig 2010). Since then there has been a trend toward more disclosure and
regulation of campaign finance (Briffault 2010). Laws passed through the 1960s were far
from comprehensive and there was no agency to ensure enforcement.
Campaign finance laws were significantly strengthened in the 1970s. First, in
1971 Congress passed the Federal Election Campaign Act (FECA) which authorized
limits on “contributions from candidates and their families”, regulated expenditures on
media, and required public disclosure of financial activity (Tokaji and Strause 2014). The
Watergate scandal revealed serious financial abuses by the 1972 Nixon presidential
campaign. This motivated Congress to amend FECA in 1974 by extending contribution
limits to include those from individuals, parties, and PACs, set maximum spending limits
for congressional and presidential candidates, and further strengthened disclosure. The
FEC was also formed by the 1974 amendments to carry out the regulations (Tokaji and
Strause 2014). Such comprehensive reform was certainly not without detractors and legal
challenges have occurred over the years. Despite some changes in response to court
rulings the FECA remained mostly intact for over 30 years.

17
The first and most influential court case was the Supreme Court’s 1976 ruling in
Buckley v. Valeo. It invalidated the limits on expenditures, but left in place those for
contributions. The decision centered on whether the regulations violated First
Amendment freedom of speech rights, and if they were needed to avoid “corruption or
the appearance of corruption”. Limiting expenditures were viewed as being too restrictive
on free speech, while contributions deserved less First Amendment protection because of
their potential for corruption (Burke 1997). All subsequent court decisions regarding
campaign finance have framed their arguments around Buckley’s drawing of corruption,
but have differed in their interpretation of its meaning.
There has been a struggle to balance the need to limit corruption with the
protection of free speech. Supreme Court decisions often revolve around the definition of
corruption. They have at times taken a wider view and expanded the definition of
corruption. One of the Court’s broadest delineations came in the 1990 Austin v. Michigan
Chamber of Commerce case in which it upheld a state ban on corporations making
independent expenditures. Here the Court went so far as to assert that corporate wealth
essentially had too much power to distort elections and would be an unfair advantage in
promoting political ideas (Briffault 2011). The Court viewed corruption to include the
concept of equality in the political process. Critics say this violates free speech and would
allow a campaign finance system in which money could only be spent if it were in line
with public opinion (Burke 1997). Although court rulings continued to generally maintain
the standards of corruption set forth in Buckley through the mid-2000s, most used a more
narrow definition of corruption than Austin.

18
As mentioned before, the BCRA of 2002, also called McCain-Feingold, made the
biggest changes to campaign finance regulation since the original FECA. Besides the
limits on independent expenditures for issue advocacy by defining them as electioneering
communications and banning corporations and unions from funding those types of ads, it
also prohibited candidates and national parties from raising soft money (Tokaji and
Strause 2014). Initial challenges to the reforms of BCRA were upheld by the Supreme
Court in McConnell v. FEC with the Court affirming the constitutionality of its limits on
corporate campaign spending. The ruling also supported the electioneering
communications measures and felt they were “properly tailored to regulate campaign
messages” (Briffault 2011). The opinion of the Court has since changed, and has led to a
rejection of a number of campaign finance regulations.
The first indication of the Court’s new stance on campaign finance came in 2007.
The composition of the Court had been altered with the retirements of Chief Justice
William Rehenquist and Justice Sandra Day O’Connor and their replacement by Chief
Justice John Roberts and Justice Samuel Alito. The case involved Wisconsin Right to
Life, a 501(c)(4) organization, using TV advertisements urging citizens to ask Wisconsin
Senators “to oppose filibusters of President Bush’s judicial nominees” during the 2004
election (Tokaji and Strause 2014). In Wisconsin Right to Life v. FEC the Court
acknowledged Congress’ authority to regulate spending on campaign ads, but effectively
invalidated the definition of electioneering communications. They ruled that if an ad
could not be reasonably interpreted as “an appeal to vote for or against a specific
candidate” then corporations could not be prohibited from funding the ads (Briffault

19
2011). Outside groups spending money independently to influence elections was not seen
as potentially corrupting.
The erosion of limits on independent expenditures continued with the 2010
Citizens United v. FEC ruling. Citizens United is another 501(c)(4) organization that in
2008 sought an exception to electioneering communications regulations for ads that
promoted its film Hillary: The Movie, which was highly critical of then presidential
candidate Hillary Clinton. Briffault (2011) argues that the Court could have made a
narrow interpretation of campaign finance rules in granting Citizens United an exception,
Instead, the majority saw that the electioneering communication statutes violated the First
Amendment and gave corporations permission to independently spend unlimited funds
supporting or opposing candidates. The decision was major setback for supporters of
campaign finance reform.
The majority opinion in Citizens United relied on a limited definition of
corruption, and saw that the only legitimate reason for regulation of campaign finance
would be to prevent quid pro quo corruption (Tokaji and Strause 2014). Since
independent expenditures are by definition not coordinated with candidates, in the
Court’s view there was no danger of corruption. The majority also felt that disclosure
requirements for independent expenditures were a sufficient form of regulation. The
ability to quickly and more fully make campaign finance data available to the public
through the use of Internet was seen by the Court to greatly reduce the potential for
improper use of finances (Briffault 2010). It may be a small consolation that disclosure
was upheld.

20
A D.C. Circuit Court of Appeals ruling, which came just two months after
Citizens United, also had important consequences for campaign finance. In
SpeechNow.org v. FEC PACs that wished to make expenditures but not contribute to
candidates were granted permission to raise unlimited funds (Tokaji and Strause 2014). .
The D.C. Circuit Court almost seemed to have no choice but to follow the Supreme
Courts lead in Citizens United, and take a position that expenditures by independent
groups were protected by the First Amendment and should not be restricted. The FEC did
not appeal the decision, and set up guidelines for committees now known as Super PACs
(Briffault 2011). Super PACs quickly became prominent in federal elections.
Corporations have long been banned from directly contributing to candidates’
campaigns. The consensus that allowing direct corporate or labor union contributions has
significant potential for corruption has remained in place. However, limits on
contributions from individuals have been loosened. The 2014 McCutcheon v. FEC ruling
further undermined campaign finance regulation. In the 2012 cycle an individual was
limited to $46,200 in contributions total. The Supreme Court struck down the aggregate
limits on individual campaign contributions, while upholding limits on how much could
be given to individual candidates (Tokaji and Strause 2014). Donors could now
contribute to as many candidates as they wanted.
Disclosure is now more important than ever. Yet a large gap in the regulations
remains. Political non-profits are not required to disclose any of their donors, but are
required to report expenditures. (Tokaji and Strause 2014). There are occurrences of
501(c) organizations contributing to Super Pacs, which makes it nearly impossible to
know the true source of the money (Briffault 2011). But even better disclosure may not

21
prevent financial abuses. Briffault (2010) argues that before BCRA there were those
willing to make six and seven figure soft money contributions to national parties despite
requirements of disclosure. However, efforts to improve disclosure should be
encouraged.
2.3 Review of Websites and Visualizations
Organizations working to enhance disclosure have provided some spatial analysis of
campaign finance data. MapLight.org created a visualization in 2008 that used
proportional symbols to map contributions to members of Congress (Figure 2).
Contributions were summed by congressional district, and the size of the symbols
depended on the total amount for each district. The visualization was available online and
allowed the user to select a member of Congress and see the map for that member. Their
report revealed that on average 79% of campaign funds for House members came from
out of district, and 57% came from out of state (MapLight.org 2008). Successful
candidates must be able to raise funds from areas outside their district.

22

Figure 2 Map of Campaign Contributions for Individual Legislators
MapLight.org

One difficulty in trying to map finance data is there are some sources of
contributions and expenditures that do not have accurate locations. Locations used to map
contributions are determined by some form of geocoding, where address information is
translated to coordinates such as latitude and longitude. Contributions from a national
party committee, either the Republican National Committee or Democratic National
Committee, to a candidate could have originated from individuals anywhere in the United
States. MapLight.org (2008) cited this reason for excluding contributions to candidates
from political parties, other candidates, and leadership PACs. Expenditures made by

23
organizations that do not disclose donors pose similar problems. Using the headquarters
of these organizations is the only way to include the data.
The Sunlight Foundation has created a series of maps depicting campaign
contributions in several different ways. One map shows the concentration of individual
campaign contributions to PACs and candidates by county (Figure 3). A person can move
the mouse pointer over individual counties to view the amount of money contributed per
person. At the bottom of the map users can scroll over a timeline to see the changes in
contribution over the last several election cycles, beginning with 1992 and continuing to
2012 (Sibley, Lannon, and Chartoff 2013). One unique feature is links that allow the
maps to be embedded on other web pages. The maps are well designed and intuitive, but
they do not associate contributions to specific candidates.

24

Figure 3 Per Capita Contributions
Sunlight Foundation

Since 2007 the Federal Election Commission (FEC) has provided map based tools
for viewing contribution and expenditure data (Anonymous 2008). The House and Senate
map in Figure 4 allows the user to select a state then district or candidate and see the
funds they have received categorized by source; whether individual, PAC, party,
candidate, or other. Independent expenditures are accessed through separate maps (FEC

25
2014a). They provide easy to understand information, but the user interface (UI) lacks
functions common to most web maps such as zoom.

Figure 4 Example of FEC House and Senate Campaign Finance Map

News websites also visualize campaign finance data. The Wall Street Journal’s
Political Moneyball web app for the 2012 presidential election (Figure 5) was built using
Tulip data visualization software. It shows proportional dot symbols representing money
raised spaced relative to political ideology (e.g. organizations supporting liberal
candidates are located close to each other) and are connected by lines to the committees
to which they made contributions. It is very interesting visually, but because of its
complexity it is not easy to navigate. The visualization is also not location based, so the
geographic distribution of contributions cannot be discovered .

26

Figure 5 Political Moneyball by The Wall Street Journal

The application for this thesis bears the most resemblance to the MapLight.org
project discussed earlier, at least on the user interface. The map has not been updated and
there are no plans to do so (Philip Minnitte, March 4,2014, email message to the author).
It is not known exactly what web technologies were used, but this project likely uses
different ones due to rapid changes since the 2008 election.
While there are a number of good campaign finance maps and data visualizations
accessible on the web, this project offers a couple of advantages and unique features.
First, users are able to see a better approximation of where each candidate’s contributions
come from. Other visualizations either don’t map data for specific candidates, or only
provide tables and graphs of contributions. Second, the application is unique in that both
contributions to candidates and outside spending can be viewed on the same map. Being
able to look at contributions and outside spending provides a more complete picture of
the money involved in House elections.

27
2.4 Choice of Technologies
The web application for this thesis is built on the open source geospatial platform
OpenGeo Suite installed on a Dell Inspiron N7110 laptop running on the Windows 7
operating system. OpenGeo includes a database component, PostGIS, to store and
manage spatial data, and a server component, GeoServer, to publish data over the web
(OpenGeo Suite). These are used to control access to the data being displayed. The client
side of the application uses the open source JavaScript libraries Leaflet and jQuery to
control the display, styling, and user interaction of the web map (Leaflet). The main
advantages of using all open source software are cost and ease of installation.
Additionally, anyone else could easily set up their own web map project based on this
thesis.

28
CHAPTER THREE: METHODOLOGY
This chapter discusses the development process of the web map as part of this thesis
project. Section 3.1 details the data sources and how they were downloaded. Section 3.2
presents the database diagram. Section 3.3 discusses data formatting and how it was
prepared for entering into the database. The creation of the database is described in
Section 3.4. Section 3.5 outlines how the data was uploaded to GeoServer. Section 3.6
explains how the web map was built.
3.1 Workflow for Campaign Finance Web Map
The process of developing the campaign finance web map included a number of separate
tasks (Figure 6). The initial step was to find and download the source data which included
a database of OpenSecrets.org campaign finance data, zip code shapefiles, and a
congressional districts shapefile. A database diagram was then created to guide the design
of the database and determine how to format the data. Formatting the data included
exporting the campaign finance data from the OpenSecrets.org database and editing the
attributes of the shapefiles. Then the database for the web map was created and the
formatted data was imported in the respective tables. The data was made available for use
in the web map by uploading to GeoServer. The final step in the development process
was to build the web application with JavaScript code.

29

Figure 6 Workflow Diagram

3.2 Data Download
Two types of datasets were needed for the web map: campaign finance and geospatial
data. The main dataset was the campaign finance data, which included both contributions
to candidates and independent expenditures. The only spatial information with these data
was the addresses of individuals and PACs that made contributions. Mapping the
contribution data required a way for the address of the contribution source to be
translated into coordinates. The geospatial data used to do this was a zip code layer.
Coordinates of the zip codes were joined to the contributions and provided their location
on the map. The independent expenditures of PACs and other outside organizations were
placed on the map using the coordinates of the congressional district in which the money
was spent.

30
The Center for Responsive Politics (CRP) provides detailed campaign finance and
lobbying data through its website (http://www.opensecrets.org). Anyone can download
the data in compressed comma separated values (CSV) format if they create an account
for the site. The files for each election cycle were downloaded in a single compressed
file. A database, including all the tables and fields, would have to be created and then the
CSV files would be imported into the tables one a time. A GitHub repository was found
that included a PostGIS database backup file called campaign_finance.dump containing
all the campaign finance data for election cycles starting in 1990 through February 2014.
It was determined that using the .dump file was easier than the CSV files since all the
data could be imported into PostGIS with a single, simple command. It was also faster
because the database did not have to be created manually and the data for each election
cycle did not have to be downloaded separately.
The one drawback to the campaign_finance.dump file was that it included only
partial data for the 2014 cycle. It was desirable to include as much campaign finance
information as possible for the map. Files that included contributions and expenditures
reported through September 5, 2014 were downloaded from OpenSecrets.org. Although
they were not the final numbers, they provided a more complete picture of campaign
finance for the 2014 cycle.
The geospatial data needed for the project included four shapefiles. State and zip
code boundary shapefiles as well as a zip code point shapefile were downloaded from
ArcGIS Online via ArcMap. Congressional District boundaries were downloaded from
the U.S. Census Bureau. Initially a Census Bureau zip code tabulation area (ZCTA)

31
shapefile was downloaded , but it did not meet the needs of the project because not all zip
codes used by the U.S. Postal Service (USPS) were included.
3.3 Create Database Diagram
A database diagram was created early in the process to clearly define the data needs of
the web application. The diagram enabled the relationships between the tables to be seen,
and served as a guide for building the database. The design of the database was easily
understood in this format. Adjustments were made to the diagram as the project
progressed and changes were made to the database to better meet the functionality
requirements of the web map. Figure 7 represents the final design of the tables in the
database. Each table has primary key fields, highlighted in red, to ensure each row has a
unique identifier and foreign key fields, highlighted in blue, that link the tables.

32

Figure 7 Database Diagram
The table most central to the database was the Candidates table. Candidates
receive contributions from both individuals and PACs, so the Individuals to Candidates
and PACs to Candidates tables, which contained fields with detailed the contribution
information, were linked to the Candidates table by the Candidate ID. The Candidates
table was also linked to the Congressional Districts table by the District ID- State foreign
key. The PACs to Candidates table contained both contributions and independent
expenditures. If the value of the Direct/Indirect field was “D”, then the record represented
a contribution. A value of “I” meant it was an independent expenditure. The Committee

33
table held more detailed information about the PACs including the full name of the PAC,
address, and industry category code. The Individuals table had similar data for individual
contributors. The Individuals and Committees tables were linked to the Zip Codes and
Industry Codes tables by the Zip-State and Industry Code foreign keys respectively. The
Zip Codes and Congressional Districts tables each had point coordinates for the
centroids of the features which were used to place the contributions and independent
expenditures on the map.
3.4 Format Data
Before working with the campaign finance data it was necessary to install and configure
OpenGeo Suite. The data in the campaign_finance.dump file was imported into PostGIS
using the RESTORE command. This created a database called “campaign_finance”. It
included tables defined by Openecrets.org from which data for the web application was
selected (Center for Responsive Politics 2015 ). The more complete 2014 data were then
added to the database. There were more fields in the tables of the database than were
needed for the web map. Since the database included candidates for all federal offices the
data were filtered so that only rows containing information for House candidates were
included in the output. The selection of the fields and filtering of the rows was
accomplished using Structured Query Language (SQL) queries performed on the
campaign_finance database, and the results were exported to CSV files.
Shapefiles also had to be processed to eliminate unnecessary data. They were
modified using ArcMap. The zip code boundaries contained census data attributes that
were not needed, so they were deleted from the attribute table. One unusual characteristic
of zip codes is that they do not truly correspond to a geographic area. Instead they

34
represent a set of roads or a specific address serviced by the USPS (Grubesic 2008). The
zip code boundaries file did not include all zip codes, but the zip code points file did. The
Merge tool was used to give all the points a polygon representation, although that created
some overlapping polygons.
Congressional District attributes were changed to match the format used by CRP.
The original shapefile used the numeric Federal Information Processing Standards (FIPS)
code to identify the state of the congressional district. For example the FIPS code for
California is 06, so the 13
th
congressional district had a district ID of 0613. The CRP
data had district IDs composed of the two letter state abbreviation and district number, so
the example district ID was CA13. The Field Calculator in ArcMap was used to create
the CRP formatted district ID.
3.5 Create Database in PostGIS
The first step in building the database was to create an new, empty database called
House_campfin (Figure 8). The tables were created using the database diagram as a
guide. They were then populated with the data exported from the campaign_finance
database. The pgShapeLoader tool was used to import the zip code and congressional
district polygon shapefiles. It was later determined that point geometry better met the
needs of the map, so new tables were created for the zip codes and congressional
districts. Point coordinates for the features were calculated using the ST_centroid
function.

35

Figure 8 PostGIS Database

After the data had been imported primary keys and foreign keys were added to the
tables. These keys are a type of constraint on the data that help ensure there are not errors
in the data (PostSQL). Constraints were necessary for the maintenance of the database,
but a great deal of time had to be spent fixing problems with the data before they could be
added.
One challenge that took considerable effort to overcome was the errors in the
address information for individuals. The foreign key linking the individual table to the
zipcode_state table could not be created until all the errors had been eliminated. Several
different types of errors related to the zip code and state fields had to be dealt with. First,
there were data entry errors. The state may have been entered incorrectly or the zip code

36
may have had numbers mixed up. The state or zip code may also have been missing
completely. If the street address and/or city fields were correct, then the incorrect data
was determined by entering the address in the ZIP Code Lookup tool on USPS website or
by searching for the address in Google Maps. Some street address information was also
incorrect and erroneous zip codes could not be fixed. There were also individuals with
foreign addresses, although they weren’t always obvious. These had to be deleted since
the database could not handle foreign addresses.
The percentage of individuals with errors in their address information was
probably less than one percent. However, that was roughly 2,000 records that needed to
be corrected or deleted. It took many hours to search for the zip codes and make the edits.
Data in the Committees and Candidates tables had errors as well. There were
some zip code errors in the Committees table, but there were only about 20 that had to be
corrected. The Candidates table included candidates that were not actually running for
House seats in 2014. It was not certain why, but it seemed to that at least some candidates
that had run in previous cycles still had committees that received or distributed funds
even though they weren’t running in 2014. A number of these candidates were found
because the district they were in did not exist. Others were not found until the web map
was being built and more candidates were being listed than had actually been running for
office.
As development of the map began, some other additions were made to the
database. Queries had to be written that would return the necessary data to the users as
they interacted with the map. When the queries were being tested in PostGIS, some of
them had significant performance issues. They would run indefinitely without returning

37
any data. The problems were likely caused by the way the queries were looking for data
across multiple tables. As the queries were executed they had to run through loops that
took far too long to process. The solution to the looping was to create “views.”
A view is a query stored in the database. The results of the query are not stored or
“materialized”, but they could be queried just as a table would be (PostgreSQL).
Returning contributions summarized by zip code was one query that was greatly
simplified by views. It required creating three views. One view was created that summed
the contributions from PACs to candidates by zip code, and another did the same for
contributions from individuals. Then the two summary views were queried to make a
single view with total contributions per zip code per candidate. More views were created
as the map was developed, and they became the main way the map returned data from the
database. They allowed data to be de-normalized for specific uses, which helped improve
performance. They also simplified the queries needed for user interaction with the map.
3.6 Upload Data to GeoServer
Once the data were imported into the tables and all the constraints were implemented,
uploading to GeoServer could begin. The first step for this was to create a Data Store,
which is the connection between GeoServer and the PostGIS database. GeoServer could
then see all the tables and views in the database and make them available to the web map
through publishing.
The data were published using two different techniques. If all the data in a
published layer was to be displayed on the map or another part of the page, then the table
or view was published directly by GeoServer (Figure 9). The attributes and geometry
columns were read and interpreted by GeoServer, and could then be added to a web map.

38
The congressional district polygons table and some of the views were published in this
manner. If only a portion of the data from a table or view was to be displayed, then SQL
View layer was created.

Figure 9 Layers Published Directly from the Database

The SQL View method used a SQL query to return a portion of the data one or
more tables. For example, the SQL View settings in Figure 10 show the query for
independent expenditures. Using a SQL query allowed for more flexibility in what data
were added to the map as well as enabling user interaction through input parameters. The
parameter for the independent expenditure layer was called commname. The value for
commname came from a committee name selected by the user. The query then returned
the expenditures made by that committee. Parameters were used in all the SQL View
layers.

39

Figure 10 SQL View for Independent Expenditures Layer

3.7 Build Web Application Using Leaflet
Once some of the layers were published through GeoServer, the iterative process of
developing the code for the map began. The map was built using the Leaflet and jQuery
JavaScript libraries. They were downloaded and the html code linked to these locally
stored copies along with their respective Cascading Style Sheets (CSS). The html
included elements for user interaction (Figure 11) and a div element containing the map.

40

Figure 11 User Input Controls

The first layer added to the map was a basemap. The basemap for the application
was obtained from MapBox. A free MapBox account was created so that an access token
could be used to authorize access to the basemap. The Light basemap was chosen
because the colors of the data markers stood out well against its grayscale color scheme.
Next the Congressional Districts layer was added as an additional reference, which the
user could toggle on and off with a check box control on the map (Figure 12).

41

Figure 12 Map with Base Layers

User interaction with the map was controlled by the same basic process (Figure
13). Input made by the user triggered a request to a GeoServer Layer via a jQuery ajax
function (jQuery.ajax 2015). When a candidate or committee was selected then the
relevant campaign finance data was displayed on the map. Specific ways users could
interact with the map are detailed below.

42

Figure 13 Data Flow Diagram

The map could display two types of campaign finance data; direct contributions to
candidates and independent expenditures. The user displayed contributions to a candidate
by making three selections. First they selected a state and then radio buttons with the
congressional districts were displayed (Figure 14). After the button for a district was
clicked, the candidates from that district appeared (Figure 15). The button for the desired
candidate was clicked and the map refreshed to show markers for all the zip codes from
which the candidate received contributions (Figure 16).

43

Figure 14 Map State Selected

Figure 15 Map with Congressional District Selected

44

Figure 16 Map with Candidate Selected

The markers for the contributions were sized proportionally based on the amount
contributed from the zip code. When a candidate had contributions from a large number
of zip codes, the markers frequently overlapped and where sometimes directly on top of
other markers. A plugin for Leaflet called Overlapping Marker Spiderfier (OMS) was
used to handle overlapping markers. When a user clicked on a group of overlapping
markers the OMS code caused them to separate so that a single marker could be selected
more easily (Figure 17). If a marker was directly on top of another, the bottom marker
could not be selected without the OMS.

45

Figure 17 Overlapping Markers Before and After Clicking

Additional information about the contributions appeared in a popup when the user
clicked on a marker (Figure 18 ). The zip code and total contributions for the selected
marker were listed first. The contributions were then listed by industry category. The
industry categories were created by CRP based on the federal government’s Standard
Industrial Classification (SIC) codes and cover over 400 specific industries and
ideological interests. Categories for contributions from individuals were determined by
the person’s occupation or employer (Center for Responsive Politics 2014c). The various
categories were different for each zip code and were partly a reflection of the local
economy and interest groups.

46
Figure 18 Popup with Contribution Information

Independent expenditures were selected in a different way from the contributions.
A jQuery User Interface (UI) autocomplete widget was used for user input. After the user
typed at least three letters in the search box the autocomplete widget returned all the
committee names that included the typed letters (Figure 19). A committee could then be
selected from the list, and the user then clicked the submit button to display the
expenditures of the selected committee (Figure 20). The expenditure markers were sized
proportionally, and color coded green if money was spent supporting a candidate or
orange if opposing a candidate. If there were any overlapping markers, the OMS would
separate them just as it did with the contribution markers.

47

Figure 19 List of Committees During User Input

Figure 20 Expenditures Displayed After a Committee was Selected

48
More specific information about the expenditures was made available in a popup;
similar to the contribution data but simpler (Figure 21). When a user clicked on a marker,
the popup appeared. The congressional district in which the money was spent was given
along with the total spent and the candidate the committee was supporting or opposing.

Figure 21 Popup with Expenditure Information

As with any application development process, building the web map was an
iterative process. A number of challenges had to be overcome. The ajax requests had to
be formatted correctly for GeoServer to accept them. Otherwise no data would be
returned. The data were returned in a GEOJSON format, which was difficult to parse for

49
the needs of the user interface. The SQL View layer settings included regular expression
validation which help prevent security breaches. They also had to be set correctly for data
requests to be accepted by GeoServer.
Designing the user interface (UI) was another challenging aspect of the project.
Some options that were explored for selecting a candidate were not able to be
implemented. Time was spent researching a way to show the states, districts, and
candidates in a tree view, which would have put them in hierarchal list structure.
However, all the tools and plugins found for tree views required very specific data
structures that were too difficult to achieve for this project. The UI that was created may
not be the most ideal, but it had adequate controls for the user to explore the data. It was
also hoped that the styling of the UI would be more polished. A jQuery UI theme was
applied to the state selection buttons, but there wasn’t time to apply it to the rest of the
page elements. Overall the web map worked well and was very responsive to the user.

50
CHAPTER FOUR: EVALUATION
The campaign finance web map was evaluated by 10 volunteers to assess how well it met
the goals, discussed in Chapter 1, of visualizing where contributions to candidates come
from and where outside groups spent money. The evaluators were a small sample of co-
workers and friends of the author. Five evaluators were employees at Valley Air Photos
in Caldwell, Idaho, and five were friends of the author residing in Meridian, Idaho. A
Google Survey Form was used to ask questions about the map and store the evaluators’
answers. The questions fell into three categories: evaluator background knowledge,
comparison of the campaign finance web map to the FEC website, and open ended
feedback.
4.1 Background Knowledge
The evaluators were asked three questions about their general knowledge of politics and
experience with web maps. The answers gave an indication of their perceptions of
campaign finance. The first question asked “How often do you watch, listen, or read
about national politics?” with 50% answering daily or weekly (See Figure 22). The rest
of the responses were in the monthly to rarely/never range. This may reflect how much
the evaluators follow national news more generally. It is likely that someone who follows
national news regularly would be exposed to stories about politics.

51

Figure 22 Chart of Survey Question #1 Responses

The next question dealt with the evaluators’ attitude about money in politics. It
asked, “How important do you feel campaign finance and money spent in elections is?”
No one felt that money was unimportant. There were 7 that answered it was very
important and 3 answered somewhat important (Figure 23. The responses were in line
with polls mentioned in Chapter 1 showing a majority of people perceiving money
having an influence on politicians.

Figure 23 Chart of Survey Question #2 Responses

52
The third background question gave some idea of the evaluators’ experience with
web maps. In answering the question “How often do you use web maps?”, 1 responded
with daily, 4 with weekly, and 2 with monthly. The other 3 answered occasionally or
rarely/never. Figure 24 shows a chart of the responses. Most of the evaluators used web
maps at least monthly, so most of them would be comfortable using web maps to find
locations and information.

Figure 24 Chart of Survey Question #3 Responses

4.2 Comparison to the FEC website
Most of the questions in the survey asked the evaluators to compare how the
campaign finance web map presented data with how pages on the FEC website presented
data. Three candidates, all of whom were incumbents, where chosen for the evaluators to
find contributions and independent expenditures. The names of some candidates were
more likely to be recognized by the evaluators than others. Mike Simpson is the
representative for Idaho’s 2
nd
Congressional District. The evaluators were likely to be
most familiar with him since they all lived in Idaho. John Boehner, who is Speaker of the
House, would be familiar to anyone that follows politics. For this group of respondents,

53
Barbara Lee, from California’s 13
th
Congressional District, was likely to be less
recognized. The results for the contribution questions will be discussed first, then the
independent expenditure questions, followed by some general comparison questions.
4.2.1 Contributions Comparison
The survey asked evaluators, “Using the Federal Election Commission (FEC) 2014
House and Senate Campaign Finance page and this web map, which tool better informed
you about contributions made to Representative Mike Simpson?” The same question was
asked for John Boehner and Barbara Lee. The evaluators were answer on scale of 1 to 5,
with 1 being “FEC page is much better” and 5 being “ Web map is much better”. They
were instructed to answer the question by finding the contributions for each candidate on
the web map, and then find contribution data on the FEC page. Figure 25 shows the
responses for contributions made to Mike Simpson. The evaluators felt better informed
by the web map than the FEC page, with 60% answering with a 4 or 5. Two responded
with a 3, one gave a 2, and one gave a 1.

Figure 25 Graph of Survey Question #4 Responses

54
The evaluators more strongly favored the web map for contributions to John
Boehner (Figure 26). Only one felt better informed by the FEC page, while 80%
preferred the web map.

Figure 26 Graph of Survey Question #6 Responses

The evaluators again felt better informed by the web map for contributions to
Barbara Lee, with 60% giving a 4 or 5 (Figure 27). Three evaluators responded with at 3
and one gave a 1.

Figure 27 Graph of Survey Question #8 Responses

55
The responses to the contribution questions were quite similar for all three
candidates. However, it was interesting to see stronger preference for the web map in the
responses for John Boehner. He received far more contributions than the other two
candidates, and the visualizing the data on the web map made a stronger impression.
4.2.2 Independent Expenditures Comparison
The questions for independent expenditures were very similar to the contribution
questions and used the same scale. Evaluators were asked, “Using the FEC 2014 House
Independent Expenditures page (http://www.fec.gov/disclosureie/ienational.do) and this
web map, which tool better informed you about money spent by outside groups
supporting or opposing Representative Mike Simpson?” The same question was asked for
the other two candidates. Figure 28 shows the responses to the independent expenditure
question for Mike Simpson. The evaluators felt somewhat better informed by the web
map, with five answering with a 4 or 5. Three evaluators gave a 2 or 1, and two answered
with 3.

Figure 28 Graph of Survey Question #5 Responses

56
The evaluators said they were somewhat better informed by the FEC page in their
responses to the independent expenditures question for John Boehner (Figure 29). Four
responded with a 1 or 2. Four were neutral; responding with at 3. Only two thought web
map better informed them.

Figure 29 Graph of Survey Question #7 Responses

Asking about independent expenditures for Barbara Lee was problematic. After
four people had completed the survey it was discovered that there were not any
independent expenditures in Barbara Lee’s district. The last six evaluators were informed
of the situation. Half the evaluators answered with a 3, two felt better informed by the
FEC page, and three felt better informed by the web map (Figure 30). The responses for
this question are of questionable value, since there really wasn’t any data to find.

57

Figure 30 Graph of Survey Question #9 Responses
Comparing the web map and the FEC page for independent expenditure data was
more difficult than for the contribution data. The web map was designed to look at
independent expenditures by separate outside groups, not the total spent by all groups
supporting or opposing a candidate. The FEC page could better answer the independent
expenditure question in aggregate because users could see the total spent in a particular
district as well as what was spent supporting or opposing a candidate.
4.2.3 General Comparison
After looking at data for specific candidates the survey asked evaluators to compare the
web map and the FEC pages more generally. The survey asked, “After exploring the
contribution data in the web map are you better informed about the geographic
distribution of money donated to U.S. House candidates in general compared to how the
data is presented on the FEC House and Senate Campaign finance page?” The responses
corresponded well with those for the candidate specific contribution questions (Figure
31). Six of the ten evaluators felt better informed by the web map and three felt they were
about the same. Only one evaluator felt better informed by the FEC page.

58

Figure 31 Graph of Survey Question #10 Responses

The survey also asked about the overall independent expenditure presentation.
The question said, “After exploring the independent expenditure data in the web map are
you better informed about where money is spent by outside groups in U.S. House
elections compared to how the data is presented on the FEC Independent Expenditure
page?” The responses were less consistent with the candidate specific independent
expenditure questions (Figure 32). Six of the ten evaluators felt better informed by the
web map when asked about the independent expenditures in general, with none being
better informed by the FEC page. Whereas they had somewhat favored the FEC page
when asked about independent expenditures for specific candidates. The results indicated
that while the web map didn’t let users see all the independent expenditures in a district,
it was still informative.

59

Figure 32 Graph of Survey Question #11 Responses

Finally, the survey asked evaluators to compare the ease of use of the tools
(Figure 33). The question said, “How easy is the campaign finance web map to navigate
and search compared to the FEC pages?” Half of the evaluators thought web map was
easier to use, three thought they were about the same, and two thought the FEC page was
easier to use.

Figure 33 Graph of Survey Question #12 Responses

60
4.3 Web Map Feedback
The last section of the survey asked evaluators to answer four open-ended questions.
These questions helped provide a little more insight to how they reacted seeing the
campaign finance data and give suggestions for improving the web map. The first
question asked, “How did the web map change your understanding of campaign
contributions to U.S. House candidates?” In general, the evaluators had a better
understanding of where candidates get contributions from. One was not aware that the
information could be accessed before doing the survey. Three evaluators mentioned they
hadn’t realized how much of contributions come from outside the candidate’s district.
The way the data was visualized made a strong impression and was easy to understand.
The second open-ended question asked, “How did the web map change your
understanding of outside spending in U.S. House elections?” The evaluators were
interested in seeing where money was spent by outside groups. Some were surprised or
shocked by how much outside spending there was. One said it should be a crime. There
seemed to be some misunderstanding about what constituted independent expenditures. A
few of the responses used the word contribution or talked about giving money to
candidates. The concept of outside spending can be confusing and perhaps it was not
explained well enough in some instances. The third open-ended question asked, “How
can the campaign finance web map be improved?” A number of useful suggestions were
made. Two evaluators thought that the total contributions for a candidate should be
shown. One person would have preferred a graph of contributions by category in the
popup. This respondent asked for something more visually appealing than the list in the
popup. It was also suggested that the user should also be able to search for candidates by

61
name, so if a user did not know the state or district of a candidate, there could still be way
to find contributions.
Last of all, a space was provided for the evaluators to make additional comments.
It was not required and not everyone left a response. One evaluator felt the web map was
easy to navigate and showed the contribution and expenditure data in a “distinctive
manner” Another would have liked to have more time to explore the map and wanted a
more direct way of finding specific candidates. One suggestion made was to link to the
more in-depth data found on the FEC website. The web map was good for quick
searches, but additional useful information could be made available. In a conversation
after completing the survey, one evaluator stated that the web map was better “at telling a
story.”

62
CHAPTER FIVE: DISCUSSION AND CONCLUSION
Chapter 5 describes the contribution of the campaign finance web map to improve the
public’s understanding of campaign contributions and independent expenditures and how
well it met the contributions set forth in Chapter 1. This chapter then concludes with a
discussion of improvements that should be made to the web map and potential future
work.
5.1 Results of the Campaign Finance Web Map
The complexities of the campaign finance system can be difficult to follow. Large sums
of money and the number of outside groups seeking to influence elections make the task
of informing voters increasingly challenging. Data visualization can be an effective way
of presenting campaign finance data. This thesis project created an interactive web map
with the goal of providing tools for viewing the spatial patterns of contributions to
individual candidates and expenditures made by outside groups. Visualizing the data in
this way reveals aspects of campaign finance that charts and tables do not.
There are other campaign finance data visualizations available, but the application
is unique in mapping both contributions and expenditures. It also better approximates the
locus of contributions and outside money spent opposing or supporting candidates. The
approach used in this thesis can inspire others to try new ways of presenting campaign
finance data. Voters need to be better informed about the campaign finance system and
the more tools available for visualizing the data the better they will understand how
campaign finance affects the political process. They will then be able to make more
informed choices when voting.

63
Based on the responses in the evaluation survey, the web map met the goals of
creating tools for visualizing contributions to candidates and independent expenditures.
Those who evaluated the web map better understood where candidates’ received
contributions from, and where outside groups spent money. Overall, they preferred the
way the data was presented on the web map compared to how it was presented on the
FEC website. The sample size of evaluators was small, so their responses may not
correlate well with a larger population. However, the survey results indicate that the web
map is a more engaging way of showing campaign finance data than the usual tables and
charts. Future work should include larger samples of users evaluating the web map in a
way similar to what was done for this thesis.
The accuracy of the locus of campaign contributions is limited by the accuracy of
the address information. While trying to correct errors in the zip codes, it was discovered
that some individuals had made contributions from different addresses. Most of the
different addresses were from previous elections cycles. It is possible that they had
moved, but they also might have more than one residence. Another possibility is that a
person used a business address instead of a home address. When there were errors in the
address information, the address that most closely matched was used. If a correct zip code
could not be determined then the contributor had to be deleted. These errors were a small
percentage of the total, so the overall accuracy was high.
5.2 Web Map Improvements
The web map is adequate as it is currently designed, but a few changes would make it
more useful. The suggestion to show the total amount of contributions a candidate has
received is a good one. Seeing the total amount would allow someone to better compare

64
candidates and give more context to all the markers displayed on the map. Another
relatively easy improvement would be to add the ability to search for candidates by name.
The same autocomplete widget used for the committee name search would make finding
a specific candidate quick and easy. The user could also avoid having to select a state,
then a district, and then select from a list of candidates.
The number of clicks it takes to see the contributions can be a bit awkward. A
more extensive redesign could reduce the number of clicks and have users interact more
directly with the map. Users could click on a congressional district in the map and see a
list of candidates. There is more than one way this functionality might be implemented,
and further research is needed to determine the best approach.
Refining the style of the UI is another area that would improve the user
experience.
5.3 Future Work
The campaign finance web map is currently implemented in a development setting on a
laptop. A number of steps will need to be taken to make it available on the internet. The
first step will be to install OpenGeo Suite on a server and transfer the House_campfin
database to it. The layers will then need to be published in GeoServer. The code for the
web map will also need to be copied to the server. A domain name will have to be
acquired, or use an existing organization’s website to host the map.
There are no firm plans to make the map available on the internet, but a couple of
the evaluators offered to help set it up on their personal servers. One of them also had a
couple of domain names that could be used. It may be worthwhile to contact some
organizations such as CRP or the Sunlight Foundation to see if they would be interested

65
in hosting the map. This would give greater exposure to the map than using a domain
name that few people are familiar with.
The web map does not currently have the complete data from the 2014 election
cycle. Before it is made publicly available the most up to date data will need to be
downloaded from OpenSecrets.org and imported into the database. However, this raises
several issues that need to be addressed for the long term maintenance of the web map.
The most immediate problem will be dealing with the errors in the addresses of
individuals and PACs. When new data is imported to the tables, it overwrites the existing
data and all the errors previously corrected will return. It is not practical to manually
correct the errors every time new data is available. The simplest way to deal with them is
to delete the records with address errors, but even that would be a long process to carry
out manually. An automated process for dealing with the address errors must be created
for the web map is to be updated regularly. One possible method could be to write a
Python script that reads the data in CSV format. It would look at the zip code field and
try to find a match in a list of the actual zip codes. If a match was not found the record
would be deleted. A SQL query might also be written that does the same thing inside the
database.
There are also issues with being able to use the web map for more than one
election cycle. The 2014 election cycle is over and the 2016 election cycle has begun.
People are likely to be more interested in what is happening campaign finance right now
than in the past. The web map could be modified to display data from different cycles.
However, there are changes to zip codes and congressional districts over time that will
have to be dealt with.

66
REFERENCES
Anonymous. 2008. “FEC Launches Enhanced Presidential Campaign Finance Map.” PR
Newswire, May 21 2008.
http://search.proquest.com.libproxy.usc.edu/docview/447426755.

Briffault, Richard. 2010. Campaign Finance Disclosure 2.0. Election Law Journal, 9(4):
273+.

Briffault, Richard. 2011. “Nonprofits and Disclosure in the Wake of Citizens United.”
Election Law Journal, 10 (4): 337–61.

Burke, Thomas F. 1997. “Concept of Corruption in Campaign Finance Law, The.”
Constitutional Commentary, 14: 127.

Cantor, Joseph E. 2009. “Campaign Finance: An Overview” In. Campaign Finance :
Background, Regulation and Reform, edited by Thomas P Kallen, 107-129. New
York, NY, USA: Nova Science Publishers, Inc. Accessed February, 19 2015.
http://site.ebrary.com/lib/alltitles/docDetail.action?docID=10660203.

Center for Responsive Politics. 2010 “The Top 10 Things Every Voter Should Know
About Money-in-Politics” OpenSecrets.org. Accessed April 24, 2015.
https://www.opensecrets.org/resources/dollarocracy/04.php

Center for Responsive Politics. 2014a. “Academic Resources: Glossary.”
OpenSecrets.org. Accessed March 14, 2015.
http://www.opensecrets.org/resources/learn/glossary.php

Center for Responsive Politics. 2014b. “2014 Campaign Contribution Limits”
OpenSecrets.org. Accessed March 14, 2015.
https://www.opensecrets.org/overview/limits.php

Center for Responsive Politics. 2014c. “Follow the Money, A Handbook.”
OpenSecrets.org. Accessed July 26, 2015.
https://www.opensecrets.org/resources/ftm/ch12p1.php

Center for Responsive Politics. 2014d. “Outside Spending: Frequently Asked Questions
About 501(c)(4) Groups.” OpenSecrets.org. Accessed March 18, 2015
http://www.opensecrets.org/outsidespending/faq.php

Center for Responsive Politics. 2014e. “Total Outside Spending by Election Cycle,
Excluding Party Committees.” OpenSecrets.org. Accessed March 6 2015.
http://www.opensecrets.org/outsidespending/cycle_tots.php.

67
Center for Responsive Politics. 2015. “OpenSecrets OpenData User’s Guide.”
OpenSecrets.org. Accessed June 23, 2015.
http://www.opensecrets.org/resources/datadictionary/UserGuide.pdf

Cho, Wendy K. Tam, and James G. Gimpel. 2007. Prospecting for (Campaign) Gold.
American Journal of Political Science, 51 (2): 255–68.

Cho, Wendy K. Tam, and James G. Gimpel. 2010. Rough Terrain: Spatial Variation in
Campaign Contributing and Volunteerism. American Journal of Political Science
54 (1): 74–89.

Code of Federal Regualtions. Title 11 Section 100.16. Accessed January 29, 2015.
http://www.gpo.gov/fdsys/pkg/CFR-2014-title11-vol1/xml/CFR-2014-title11-
vol1-sec100-16.xml

FEC. “FEC Mission and History.” Accessed February 21, 2015.
http://www.fec.gov/info/mission.shtml

FEC. 2013 “Political Party Committees.” Accessed February 24, 2015.
http://www.fec.gov/pdf/partygui.pdf

FEC. 2014a. “House and Senate Campaign Finance.” Federal Election Commission.
Accessed April 27, 2014. http://www.fec.gov/disclosurehs/hsnational.do

FEC. 2014b. “SSFs and Nonconnected PACs.” Accessed January 29, 2015.
http://www.fec.gov/pages/brochures/ssfvnonconnected.shtml

IRS. 2015. “Exempt Function – Political Organization” Accessed February 21, 2015.
http://www.irs.gov/Charities-&-Non-Profits/Political-Organizations/Exempt-
Function-Political-Organization

Garrett, R. Sam. 2011 “The State of Campaign Finance Policy: Recent Developments and
Issues for Congress.” Quoted in Lessig, Lawrence. 2011. Republic, Lost: How
Money Corrupts Congress--and a Plan to Stop It. 1st edition. New York: Twelve.

Gimpel, James G., Frances E. Lee, and Joshua Kaminski. 2006. “The Political Geography
of Campaign Contributions in American Politics.” Journal of Politics 68 (3): 626–
39. doi:10.1111/j.1468-2508.2006.00450.x.

Grubesic, Tony H. 2008. “Zip Codes and Spatial Analysis: Problems and Prospects.”
Socio-Economic Planning Sciences 42 (2): 129–49.
doi:10.1016/j.seps.2006.09.001.

68
Jones, Jeffrey M. 2014. “Congressional Job Approval Stays Near Historical Low.”
Accessed September 19, 2014.
http://www.gallup.com/poll/174806/congressional-job-approval-stays-near-
historical-low.aspx.

“jQuery.ajax().” 2015. jQuery API Documentation. Accessed August 8, 2015.
http://api.jquery.com/jQuery.ajax/.

Kalla, Joshua L., and David E. Broockman. 2014. Congressional Officials Grant Access
to Individuals Because They Have Contributed to Campaigns: A Randomized
Field Experiment. Accessed March 11, 2015.
https://www.ocf.berkeley.edu/~broockma/kalla_broockman_donor_access_field_
experiment.pdf.

Krum, Randy. 2013. Cool Infographics : Effective Communication with Data
Visualization and Design. Somerset, NJ, USA: John Wiley & Sons. Accessed
March 16, 2015. ProQuest ebrary.

Leaflet. “Leaflet API.” Last accessed March 19, 2015. http://leafletjs.com/reference.html.

Lessig, Lawrence. 2011. Republic, Lost: How Money Corrupts Congress--and a Plan to
Stop It. 1 edition. New York: Twelve.

OpenGeo Suite. http://boundlessgeo.com/solutions/opengeo-suite/

Ornstein, Norman J., Thomas E. Mann, Michael J. Malbin, and Andrew Rugg. 2014.
“Vital Statistics on Congress.” The Brookings Institution. Accessed April 19,
2014. http://www.brookings.edu/research/reports/2013/07/vital-statistics-
congress-mann-ornstein.

Persily, Nathaniel, and Kelli Lammie. 2004. Perceptions of Corruption and Campaign
Finance: When Public Opinion Determines Constitutional Law. Scholarship at
Penn Law upenn_wps-1033. University of Pennsylvania Law School. Accessed
March 15, 2014. http://ideas.repec.org/p/bep/upennl/upenn_wps-1033.html.

“Political Moneyball, from WSJ.com.” 2012. The Wall Street Journal. Accessed April 6,
2014. http://graphics.wsj.com/political-moneyball.

PostgreSQL 9.4.4 Documentation. 2015. “CREATE VIEW.” Accessed July 21, 2015.
http://www.postgresql.org/docs/9.4/static/sql-createview.html

PostgreSQL 9.4.4 Documentation. 2015. “5.3. Constraints.” Accessed July 21, 2015.
http://www.postgresql.org/docs/9.4/static/ddl-constraints.html

69
Sibley, Ryan, Bob Lannon, and Ben Chartoff. 2013. “Political Influence by County: A
New Way to Look at Campaign Finance Data.” Sunlight Foundation. Accessed
February 20, 2014. http://sunlightfoundation.com/blog/2013/10/23/political-
influence-by-county-a-new-way-to-look-at-campaign-finance-data/

Stephanopoulos, Nicholas O. 2014. “Elections and Alignment.” Columbia Law Review
114: 283–365. Accessed September 9, 2014. http://columbialawreview.org/wp-
content/uploads/2014/03/Stephanopoulos-N..pdf

Tokaji, Daniel P, and Renata E.B. Strause, Moritz College of Law, and Election Law @
Moritz. The New Soft Money: Outside Spending in Congressional Elections,
2014. Accessed September 11, 2014.
http://moritzlaw.osu.edu/thenewsoftmoney/wp-
content/uploads/sites/57/2014/06/the-new-soft-money-WEB.pdf

“ZIP Code.” 2015. Wikipedia, the Free Encyclopedia. Accessed August 8, 2015.
https://en.wikipedia.org/wiki/ZIP_code

70
APPENDIX A: GOOGLE EVALUATION SURVEY

71

72

Abstract (if available)

Abstract It is expensive to get elected to the U.S. House of Representatives, and in the past several decades the increase in spending has been very steep. In 2012, candidates spent an average of nearly $1.2 million (Ornstein, et al 2013). However, that includes only direct candidate or party expenditures, and does not included money spent by outside (i.e., “independent”) groups. Lessig (2011) argues that the way campaigns are funded, and the dependence members of Congress have on a relatively small number of donors is a form of corruption in our political system. This thesis produces an interactive web map showing the geographic distribution of campaign contributions and independent expenditures made for members of the U.S. House of Representatives. Campaign finance data are most commonly displayed in tables and graphs. They are useful and important for those seeking to investigate the details of campaigns or needing to answer specific question, but a map is more accessible and engaging for the general public. There are numerous other visualizations available on the internet, but many have not been updated since 2012 election cycle (or earlier), or may not include all sources of spending. The web map created as a part of this thesis enables a user to select a candidate and view contributions summed by zip code using graduated symbols. The geographic origin of contributions is apparent, whether within or outside the district. A user can also search for groups that made independent expenditures and see the congressional districts where money was spent. An evaluation of the web map by a small sample of people showed the effectiveness of visualizing campaign finance data to better inform the public about money used in elections.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Wake County District Overlay: an online electoral data visualization application

PDF

A fire insurance map geocoder for pre-earthquake San Francisco

PDF

A user study of GIS infused genealogy with dynamic thematic representation and spatiotemporal control

PDF

GIS data curation and Web map application for La Brea Tar Pits fossil occurrences in Los Angeles, California

PDF

Visualizing email response data to improve marketing campaigns

PDF

The geography of voter power in the U.S. electoral college from 1900-2012

PDF

Semi-automated visualization of spatial information in unstructured text

PDF

Philly Bike Report: a mobile app for mapping and sharing real-time reports of illegally blocked bike lanes in Philadelphia

PDF

Guiding business oriented volunteered geographic information through geotigger services: a case study of CrossFit affiliates

PDF

Mapping uniformity of park access using cadastral data within Network Analyst in Wake County, NC

PDF

Development of a Web GIS for urban sustainability indicators of Oakland, California

PDF

Validation of volunteered geographic information quality components for incidents of law enforcement use of force

PDF

Investigating electoral college reform: geography's impact on elections, and how maps influence our perception of election outcomes

PDF

Evaluating machine learning tools for humanitarian road network mapping

PDF

Drawing better lines: comparing commissions to legislatures on compactness and coterminosity

PDF

Local votes and outside money: campaign contribution geographic origins and their impact on Los Angeles City Council election outcomes

PDF

Spatiotemporal visualization and analysis as a policy support tool: a case study of the economic geography of tobacco farming in the Philippines

PDF

Integrating spatial visualization to improve public health understanding and communication

PDF

Using pedestrian accessibility indicators to locate schools: a site suitability analysis in Greenville County, South Carolina

PDF

Social media canvassing using Twitter and Web GIS to aid in solving crime

Asset Metadata

Creator Underwood, Jamen Dean (author)

Core Title Campaign financing for the U.S. House of Representatives: an interactive Web map

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 09/16/2015

Defense Date 08/21/2015

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

Tag campaign contributions,campaign finance,independent expenditures,OAI-PMH Harvest,web map

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Chiang, Yao-Yi (committee chair), Vos, Robert (committee member), Warshawsky, Daniel (committee member)

Creator Email jamen.underwood@gmail.com,jdunderw@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-182990

Unique identifier UC11272532

Identifier etd-UnderwoodJ-3914.pdf (filename),usctheses-c40-182990 (legacy record id)

Legacy Identifier etd-UnderwoodJ-3914.pdf

Dmrecord 182990

Document Type Thesis

Format application/pdf (imt)

Rights Underwood, Jamen Dean

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