Small Donors in Congressional Elections

Abstract

Candidates raise substantial sums of money to compete in federal elections. Scholars and election observers are concerned by potential corruption related to the reliance on donors who make significant contributions. One reform effort to counterbalance large donors is encouraging small donor participation. Still, some worry that ideologically extreme candidates are best able to raise small donations. We analyze internal U.S. Federal Elections Commission data to examine small donor giving in the 2006 through 2010 U.S. House elections. We find small donors may expand the scope of participation, in that the supply of small donors is unrelated to income and that all types of candidates—incumbents, their challengers, and open seat candidates—are equally adept at attracting small donors. Candidates in the most competitive races raise the most in small contributions. We temper reformers’ enthusiasm, finding that ideologically extreme incumbents tend to raise more money from small donors.

Keywords

campaign finance small donors elections

Introduction

Donating to political campaigns is a form of political participation. Unlike voting, where each eligible citizen has only one vote in a geographically constrained race, citizens can donate to more than one candidate, across multiple jurisdictions, up to regulated limits. Much has been made of the political inequality resulting from a donor class of “fat cats” who underwrite American elections. Early research noted that “the bulk of political money is supplied by a relatively small number of people” (Heard, 1962, p. 48). If, as studies find, wealthy individuals are more likely to donate, and if elected officials are responsive to these “monetary surrogate” constituents (Mansbridge, 2003, p. 523), then policy outcomes may favor the wealthy (e.g., Bartels, 2005), and distort representation in favor of a minority with more resources (e.g., Farrar-Myers & Dwyre, 2008). To counter the potential corrupting effect of large donors, some in the reform community have revived arguments to empower small and local donors thus “ . . . expanding the range of citizens who can participate in the political process” (Bauer & Potter, 2013; see also Corrado, Malbin, Mann, & Ornstein, 2010; Potter & Morgan, 2013).

Public concern over large donations and the merits of empowering small donors date to the Progressive Era (Mutch, 2014; Weyl, 1913). Parties and candidates have pursued the “small donor” over time, even if for little more than good public relations. As then-Senator John Warner (R-VA) noted, “Of all the sources of private monies, the most desirable and least controversial is that contributed by instate individuals in small amounts” (cited in Rosenberg, 2002, p. 1). Despite the desirability of small donors, some argue candidates’ pursuit of small contributors is an inefficient use of campaign time and resources (Gais, 1998; Heard, 1962; Sorauf, 1988). Candidates for federal office must raise large sums of money to fund their expensive campaigns, so it is more economical to raise money from a large donor than many small donors. Advances in technology have perhaps changed this calculus by reducing candidate costs to cultivate small donors (Christenson, Smidt, & Panagopoulos, 2014; Malbin, 2013). Furthermore, a handful of places such as New York City and Los Angeles have enacted reforms such as matching small donations to encourage more small donor activity (Malbin, 2013; Malbin, Brusoe, & Glavin, 2012); others advocate for publicly funded vouchers that citizens may donate to candidates (i.e., Hasen, 2016).

What might be expected if reforms are adopted to encourage small donations and if candidates more often pursue them? Wilcox (2008) finds both ideologically extreme and moderate candidates successfully use the Internet to attract small donors. Moreover, ideologically extreme individuals are well known to participate at high rates in voluntary campaign activities, including the infrequent activity of contributing to candidates (e.g., Ensley, 2009; Francia, Green, Herrnson, Powell, & Wilcox, 2003) and responding to direct mail solicitation (Brown, Powell, & Wilcox, 1995; Wilcox, 2008). Thus, Ezra Klein (2013) argued in the Washington Post, “Just as big money is corrupting, small money is polarizing.”

Although having ample campaign resources does not guarantee victory, fundraising is a critical campaign activity. Given the reform activity and detractors’ cautions, we believe a clearer understanding of which candidates are successful at raising small donor money is in order, else unintended consequences may follow. Candidates often must motivate donors to give to their campaigns, particularly small donors who may not have aspirations of access to the candidate if they win office. Because candidates often must solicit directly or make it easier for small donors to participate, we focus on the campaign perspective as to which candidates are successful at attracting small donors, unlike prior research that investigates who donors in general are. If small donors give more money to extremist candidates, increasing candidates’ reliance on small donors may favor ideologically extreme candidates who can inflame the passions of ideologically proximate small donors. In contrast, if small donors give money to influence election outcomes, we may expect their contributions to flow to the most competitive races, regardless of candidate ideology and whether they are an incumbent, challenger, or open seat contests.

Although the normative appeal of increased small donor participation may seem self-evident, the scientific evidence regarding the consequences of small donations is incomplete. The extant literature focuses on incumbents, large donors, survey self-reports, or small donors in presidential elections. By analyzing small donor patterns to all House campaigns, we increase the scope of analysis beyond presidential elections or congressional incumbents, and can thus add to the debates as to the effect of small donors on politics. To do so, we analyze internal U.S. Federal Elections Commission (FEC) campaign finance data to investigate patterns of small donor contributions in the 2006 to 2010 House elections. We use Bonica’s (2013) estimates of candidate ideology drawn from patterns of campaign donations, which permit us to fully explore small donation patterns among incumbents, their challengers, and open seat candidates rather than restricting our analysis to incumbents’ ideology alone (Poole & Rosenthal, 1997). We find small donors can be democratizing by expanding the scope who receives money, as small donors do not appear to be constrained by their income and give their support fairly evenly across incumbents, their challengers, and open seat candidates. Electoral competition also factors prominently in small donor fundraising, particularly with respect to challengers and open seat candidates. However, for those concerned about polarization, we observe that ideologically extreme incumbents raise more small donations. Interestingly, only incumbents benefit from extremism, whereas their opponents or open seat candidates do not similarly benefit from extremism nor are challengers able to raise money off an extreme incumbent opponent. Our findings do not have a clear resolution to the small donor debate, rather that reformers and scholars should take a more nuanced approach.

Who Gives to Whom?

Generally speaking, most donations do not occur spontaneously—donors must be asked and motivated to give. Scholars have approached the study of donors from both angles: from the perspective of the donor and from the perspective of the campaigns. Scholars focus primarily on large donors, those who give higher amounts of money to campaigns. There are several reasons why this is true. Large donors are disclosed in FEC reports thereby enabling surveys of this class of donors (e.g., Barber, Canes-Wrone, & Thrower, 2017; Francia et al., 2003), although some scholars investigate donor activity through general surveys like the American National Election Study (e.g., Johnson, 2010, 2013; Lipsitz & Panagopoulos, 2011). Other survey analysis of small donors focuses on presidential races (i.e., Graf, Reeher, Malbin, & Panagopoulos, 2006; Wilcox, 2008). Large donors are of primary interest to scholars interested in their potential distortive effects on elections and policy outputs, and only recently have reformers and scholars become interested in systematically studying small donors’ potential countervailing effects.

Scholars are generally in consensus as to who donors are. Graf et al. (2006, p. 51)¹ find large donors tend to be wealthy; in contrast small donors, while still elite, tend toward the middle class. Others arrive at similar results regarding large donors in presidential elections (e.g., Brown et al., 1995; Malbin, 2013; Wilcox et al., 2003). Internet donors in 2000 were also similar, though Democratic Internet donors were younger and less wealthy (Wilcox, 2008). Gimpel, Lee, and Pearson-Merkowitz (2008) call these contributors a “donor class” (p. 374). This makes sense as the poor less frequently have the resources available to donate, much less participate in other forms of political participation (Brown et al., 1995; Francia et al., 2003; Schlozman, Verba, & Brady, 2012). These patterns may be shifting toward greater inequality: In an analysis of American National Election Survey (ANES) respondents who contributed money to a candidate, party, or organization, Lipsitz and Panagopoulos (2011) find that donors in general were less representative of the general population in the 2008 presidential election than in the past.

An analysis of aggregate donation patterns cannot directly infer individual behaviors due to the ecological inference problem; however, a campaign level analysis such as ours may still reasonably include constituency level variables. Scholars find the donor class is geographically concentrated in a few fundraising markets (Gimpel et al., 2008) or within preexisting dense geographic networks (Cho & Gimpel, 2007). If campaigns predominantly cultivate small donors within their districts, demographic variation across congressional districts may correlate with small donor activity to candidates in those districts, as do election outcomes (Koetzle, 1998). From the analyses of individual donors, we expect campaigns with more affluent and educated constituencies to raise more money. However, if small donors give to candidates regardless if a donor is a constituent, similar to big donors (Gimpel et al., 2008), then little or no discernible patterns linking candidates and their geographic constituencies may be observed. In a similar vein, we might expect no relationship between income and small donations if small donors are less constrained by their economic situation to contribute than large donors.

Donor Motivations

Extant research on donor motivations relies on collective action theory—identifying the material, solidary, and purposive incentives that motivate donors (e.g., Brown et al., 1995; Francia et al., 2003; Johnson, 2010, 2013). In their examination of “significant” or big donors in 1996, Francia et al. (2003) create four categories of donors: Investors (material) who seek access or personal gain, Ideologues (purposive) who pursue particular issue agendas, Intimates (solidary) who are motivated by the social aspects of giving, and Incidentals who have inconsistent giving patterns. These categories are not mutually exclusive, as contributors may exhibit multiple types of donation patterns. Malbin (2013) notes wealthier donors may be more inclined to material and solidary motives than poorer individuals who are more likely to pursue purposive benefits. In a survey of presidential Internet donors in 2000, which includes small donors, Wilcox (2008) similarly finds these donors favor material and solidary incentives, with some partisan differences. In contrast, Barber et al. (2017) find policy agreement and committee membership were related to donations, but access did not motivate giving when investigating significant donors to senate incumbents. Likewise, Johnson’s (2013) analysis of FEC data and ANES self-reports of incumbents suggests purposive giving may be the main donor motivation. In his examination of total receipts to candidates in 1996, Ensley (2009) finds ideological candidates fare well at fundraising. La Raja and Wiltse (2011) cite evidence that the proportion of ideologues in the donor pool has increased, but fall short of concluding this motivates their giving.

We find the theoretical framework of material and purposive individual motivations to be useful in structuring our theory and analysis. To these motivations, we add instrumental motivations to influence election outcomes. We would like to be clear that in adopting this framework, we cannot distinguish with certainty individual motivations from campaign strategies from an aggregate analysis. Indeed, as we shall see, campaigns may adopt a strategy to stoke particular individual motivations to increase contributions to their campaign.

Material Incentives: Access

By definition, all candidates are either incumbents, challengers, or those vying for an open seat. Quality candidates, those with previous legislative experience, are known to be effective fundraisers (Jacobson, 2013), so reasonably these candidates may exhibit prowess in raising money from small donors. Incumbents by definition are quality candidates, and have little difficulty raising funds from multiple sources even when faced with an opponent (Alexander, 1984; Herrnson, 1992; Jacobson, 1980; Sorauf, 1988). Materialistic donors may be especially attracted to incumbents for the access they provide to the policymaking process. Scholars find large donors of all types give primarily to incumbents, eschewing their challengers (Austen-Smith, 1995; Francia et al., 2003; Sorauf, 1992). Materialistic donors attempt to gain access, so party leaders would be desirable targets for contributions.

On the contrary, a small donor may not have the same expectation as a large donor regarding access as the donation amount might not reach a perceived cost of access. To borrow from the rational choice perspective, the probability that a single small donation even in a competitive election will decisively tilt the balance is minuscule (Riker & Ordeshook, 1968), so the benefit of a small donation to a candidate is trifling. Scholars find in more recent studies limited or no support for access-oriented strategies even among large donors (Barber et al., 2017; Gimpel et al., 2008). Furthermore, incumbents are usually electorally safe and can attract large donations more easily, so they may find it inefficient to raise money from small donors (Gais, 1998; Heard, 1962; Sorauf, 1988). From this perspective, we may expect incumbents, their challengers, and open seat candidates to do equally well in raising money from small donors.

Instrumental Incentives: Competitive Elections

Previous research indicates large donors act instrumentally to support campaigns where contributions matter the most: the competitive elections (Cox & Munger, 1989; Duan, 1983; Ensley, 2009; Francia et al., 2003; Gimpel et al., 2008). A number of factors may identify a competitive electoral environment, the most obvious being a balance of electoral support for the candidates. Other indicators of a competitive electoral environment are the presence of a potentially vulnerable incumbent freshman and the presence of a quality candidate or an opposing quality candidate who can raise money to mount an effective campaign challenge (Jacobson, 2013).

Scandal also has political and electoral consequences, affecting vote share, competition, the emergence of challengers, and fundraising. Research on scandals, particularly the 1991 House banking scandal (Jacobson & Dimock, 1994), finds scandal induces strategic retirements and increased opportunities for primary and quality challengers to emerge (see also Boatright, 2013). Basinger (2013) finds that scandal costs incumbents up to 5% of their vote share on average, but only slightly increases the odds of a contested election. Challengers may capitalize on the negative attention accorded the incumbent by raising and spending more money, generally, in the presence of an incumbent’s scandal (Basinger, 2015). We might thus expect challengers to scandal-ridden incumbents should thus be more successful at raising money from all donor types, including small donors.

Purposive Incentives: Ideologues

Malbin (2003) notes unsolicited money pulled into elections is often from contributors maintaining access to incumbents, whereas contribution appeals push money into campaigns from donors with ideological agendas. Prior research indicates the purposive incentive of ideology is an important donor motivation (Brown et al., 1995; Francia et al., 2003) and that candidates position themselves to attract such funds (Aldrich, 1983; Barber et al., 2017). Candidates may make purposive appeals based on ideology because, “[t]he donor pool, as a whole, is fairly polarized. Donors tend to congregate at the ends of the political spectrum, whereas most of the general public congregates in the middle” (Graf et al., 2006, p. 1). La Raja and Wiltse (2011) argue candidates “selectively mobilize ideological donors” (p. 20). Interviews with consultants in the heyday of direct mail point out the necessity of scare tactics, emotion, and controversy: “People are only motivated to give when they feel threatened. They have to believe their contribution will go to some candidate who has to be elected or the country will be in deep trouble” (Terry Dolan in Luntz, 1988, p. 159). Brown et al. (1995) note small donors are more likely to be solicited via direct mail and that candidates with more extreme ideologies tend to rely on this resource.

We may expect absent any changes to the structure of fundraising that donation patterns would persist. Candidates find fundraising from habitual donors a cost-effective fundraising strategy (Brown et al., 1995; Francia et al., 2003; Heard, 1962; Herrnson, 1992). Research suggests candidates rely heavily upon their habitual donors, a stable pool of donors that was predominated by large donors (Brown et al., 1995; Francia et al., 2003). Habitual donors may also include the smaller donor, but broadening the fundraising base requires finding these donors, which in the past often was possible only through direct mail appeals relying on messages of values and conflict to motivate donors (Brown et al., 1995; Luntz, 1988).

Technology has changed the ways by which campaigns make appeals to donors (Panagopoulos, 2017) and the Internet can be a particularly useful tool to fundraise from small donors (Christenson et al., 2014; Karpf, 2013). Recent research has found that although Internet donors are more moderate (Wilcox, 2008), electronic solicitations often exhibit the same qualities as traditional direct mail targeting small donors by appealing to fear, values, and ideology (Panagopoulos, 2017). Internet donors appear to respond to ideologically extreme candidates (e.g., Bonica, 2011; Karpf, 2013), although there exists some disagreement on this perspective (cf. Malbin, 2013). Some candidates are simply polarizing. For example, after Michele Bachmann (R-MN) made strong accusatory comments about investigating Congress for “anti-Americans,” she raised 1 million dollars; moreover, more than 1 million dollars from individuals flowed into her opponent’s campaign in addition to another million from the Democratic National Committee.² Anecdotal evidence such as this suggests social media could be used to motivate impulsive, emotionally driven donors to contribute both to a candidate and to their challenger (Vaccari & Nielson, 2013).

Summary

In sum, the previous literature, which largely focuses on significant donor motivations and candidate appeals, provides intuition for small donor activity. First, small donors are likely not primarily motivated for material incentives, to gain access to politicians, because a small donation is likely too small to reach a threshold to purchase access. Second, small donors may be motivated to give by purposive or ideological reasons, and is reinforced by campaigns that make ideologically extreme appeals either through traditional direct mail or “urgent” email appeals. Third, candidates in competitive electoral environments are the most motivated to raise money from all sources, and can draw support from donors most interested in affecting the partisan composition of Congress.

Methodology and Data

We are interested in explaining patterns of congressional candidate fundraising success among small donors. We analyze aggregate contributions from individuals totaling no more than US$200 per election, commonly known as unitemized donations, and what we refer to as small donors. In contrast, significant or large donors contributing US$200 or more to a candidate committee in a single election is itemized and information about that donor is disclosed; this includes the donor who gives multiple small donations to the same candidate in a general or primary election period that total US$200 or more. Others have noted how some candidates are particularly adept at turning their small donors into repeat donors who cross the reporting threshold into significant donors, especially with email solicitation and subscriptions (Lipsitz & Panagopoulos, 2011; Malbin, 2011). As most campaign committees report small donor data in bulk as unitemized donations (undisclosed), we cannot perform an individual level analysis, only an aggregate analysis of fundraising success.

We examine aggregate donations from small donors to House candidates in the 2006 through 2010 election cycles, which include primary and general elections.³ We examine all major party congressional candidates who ran against a major party opponent in the general election and who filed complete FEC reports. The combined total of all such contributions amounts to US$245 million dollars. Candidates are raising an increasing amount of small donor contributions. In 2006, House candidates raised US$66.2 million in small donor contributions; in 2008, they raised US$75.2 million; and in 2010, they raised US$103.5 million. This is consistent with increasing small donor money activity in presidential elections (Lipsitz & Panagopoulos, 2011). However, contributions are rising from all sources, such that the percentage of small donor contributions from total receipts remains small and fairly constant across all three electoral cycles, ranging from 6.4% (2008) to 8.4% (2010). Differences exist across the types of candidates. Incumbents, on average, raise only 6.7% of their receipts from small donors, whereas open seat contenders average 11.8%. Challengers raise the highest percent of their receipts on average from small donor contributions, at 21.0%.⁴

The dynamics of elections and hence, candidate behavior, varies on the basis of candidacy status. Incumbents, challengers to incumbents, and open seat candidates all face varied conditions during their campaigns that have differing effects on their small donor-fundraising success. Moreover, some of the available independent variables are only applicable to one type of candidate. For these reasons, and to follow the methodology used in previous research of large donors (Francia et al., 2003), we analyze three statistical models using linear regression. In Model 1, we examine only incumbents who had a major party opponent. In Model 2, we examine only challengers to incumbents. In Model 3, we examine only open seat candidates and we simultaneously analyze both open seat candidates to the same seat. We provide descriptive statistics for all our variables in the appendix.

Operationalizing the Dependent Variable

Our unit of observation is the candidate, and our dependent variable is the total amount of small donor contributions received by each candidate. The campaign committee summary data publicly available from the U.S. FEC proved insufficient for our analysis. We discovered some campaign committees disclose some or all their contributors giving below the aggregate US$200 mandatory reporting threshold. We call these committees “serial reporters” for their penchant to disclose all donors, a practice that is consistent with federal law which provides committees discretion as to whether or not they itemize contributions for donors who did not reach the aggregate US$200 itemization threshold.⁵ We discovered that when candidate committees choose to report their donations in this manner, the FEC does not tabulate these itemized small donors in the publicly available campaign committee summary reports and neither are these individuals reported in publicly available individual contributor data files. Furthermore, some campaign committee summary reports available from the FEC website apparently (and inconsistently) include other categories of money within the aggregate individual donor categories, such as candidate self-loans. When we compute small contributions by subtracting all itemized donors from a committee’s total individual donations, we produce obvious nonsensical donor statistics for some candidates, as much as over negative 1 million dollars. These clear outliers led us to contact the FEC, and through their generous cooperation, we obtained internal FEC databases of candidate summary reports and itemized individual donations for all campaign committees, including the serial reporters.⁶ We calculate a campaign’s aggregate small contributions as the sum of the unitemized donations from these internal FEC candidate summary reports and databases of itemized donors for the serial reporters, selecting donors who gave an aggregate of US$200 or less to a single candidate committee.⁷

Our dependent variable is the aggregate small donations given to a candidate. Following Jacobson (1978, p. 471), we transform small donations from actual dollars to their natural log to account for potential outliers and to capture the diminishing returns of campaign contributions on electoral success. Of the 1,862 candidates in our analysis, 849 are incumbents (45.6%), 831 are challengers to incumbents (44.6%), and 182 are open seat candidates (9.8%). In all, candidates in our analysis received a total of US$244.8 million in small donations, or an average of US$131,516. The maximum amount raised was US$8 million and the minimum was US$20. Incumbents raised an average of US$129,630; challengers to incumbents raised an average of US$109,544; and open seat candidates raised an average of US$148,510.⁸ These statistics do not include small donor activity for all candidate committees; we exclude candidates in “top two” primaries and some holdover committees still active from previous cycles or from candidates who left office for whatever reason mid-cycle.

Material Incentives

We measure material incentives through the overall distribution of small donor activity among incumbents, their challengers, and open seat candidates. Furthermore, party leaders are in a prominent position to influence policymaking. If small donors are driven by material incentives, we expect incumbents who are in House leadership to raise a greater amount of small donor money. We score the dichotomous Leader variable as a “1” for incumbents who hold a formal role in the House hierarchy with either party, including committee chairs. Leaders are prodigious fundraisers, so we might expect their capacity to raise large donor money to spill over to small donors. Conversely, if small donors are not motivated for access, we may observe leaders do not raise any more money than other incumbents. Indeed, if there are opportunity costs of raising money from large or small donors, we may even expect leaders to spend more resources pursuing large donors over small donors.

Instrumental Incentives

If small donors give for instrumental reasons, to affect the outcome in competitive elections, we expect indicators of competitive elections to correlate with small donor activity. One measure of electoral Competitiveness in the general election is the vote margin between the two major party candidates.⁹ We operationalize competitiveness by subtracting the margin from 100 (and dividing through by 100). The new variable ranges from 0 to 1, with larger values indicating greater electoral competition. We expect a positive relationship between this variable and small donor contributions.

As FEC candidate summary data are reported for both general and primary election periods in an election cycle, we may expect candidates are pressured by a competitive primary election to raise small donor money. If candidates use the primary period to fundraise in anticipation of a competitive general election contest, then candidates will raise money irrespective of the intensity of their primary challenge. We measure the pressure to raise donations during the primary cycle by the Primary Competitiveness, measured similarly to competitiveness above.¹⁰

Freshmen members of Congress are frequent electoral targets as they may have yet to solidify their reelection constituency. We may expect freshmen to feel more pressure to raise small donor money than their senior colleagues; however, there may be a noninstrumental reason to expect this pattern, too. Freshmen are still building their fundraising networks and may more willingly embrace new fundraising strategies, especially newer technologies that facilitate small donor giving to broaden their donor pool.

The capacity of candidates to run effective campaigns is expected to be related to fundraising success. Higher quality candidates, those who have held prior elective office, may be better at raising small donor money because they are generally more adept at running campaigns (Jacobson, 2013). We wish to draw a clear distinction with Jacobson’s familiar terminology, who refers to quality challengers campaigning against incumbents, as we are interested in both the capacity for quality candidates to raise money and the reaction of candidates, particularly incumbents, to the presence of a quality challenger opponent. We expect that a Quality Candidate in the general election is capable of raising more small donor money, and that a candidate facing a Quality Opponent will activate the breadth of their donor resources when facing greater competition. Both measures are dichotomous variables. Incumbents are, by definition, quality candidates as they currently hold elective office. As there is no variation among incumbents, we drop the Quality Candidate variable from the model predicting incumbent fundraising and drop the Quality Opponent variable from the model predicting challengers to incumbent fundraising. In contrast, challengers to incumbents and both candidates to open seats may have previously held elective office, so challengers to incumbents are classified by their level of quality, as are both the candidate and their opponent in an open seat.

Incumbent scandals may also factor into a competitive electoral environment when quality challengers emerge to contest a vulnerable incumbent. An Incumbent Scandal can improve the odds for challengers and improve their fundraising success when challengers raise their profile and attract more donors on the perceived misdeeds of the current officeholder, capitalizing on the negative evaluations of the electorate (Basinger, 2013). We make use of a Basinger (2015) comprehensive database on congressional scandals, which identifies those incumbents with an ethics issue, including financial, political, or sex scandals. We include a dichotomous variable scandal and using this broad measure, we find that 3.22% of incumbents (n = 42) in our study were scandal-tainted and a smaller percentage (2.45, n = 32) of challengers faced such incumbents.

Purposive Incentives

If small donors give for expressive reasons, to aid ideologically proximate candidates or to oppose extremely dissimilar candidates, we expect candidate ideology to correlate with small donation patterns. Scholars commonly use NOMINATE scores as a measure of incumbents’ ideology, which are based on incumbents’ voting records in Congress (Poole & Rosenthal, 1997). Candidates who have not served in Congress, such as the challengers to incumbents and open seat candidates who are central to our analysis, do not have NOMINATE scores. Excluding these candidates’ ideology scores limits our analysis to incumbents and may thus result in omitted variable bias. We thus turn to another candidate ideology measure, Bonica’s (2013, 2014) CFScores (i.e., Campaign Finance Scores). Bonica calculates CFScores from campaign finance data, leveraging the insight that contributors’ decisions to give to candidates are analogous to the binary choices confronted by legislators in roll call votes, and thus candidates’ ideologies may be constructed from contributions using methodologies similar to Poole and Rosenthal’s to construct ideologies from roll call votes. We do not believe that using contributor behavior to predict small donor contributions introduces bias as CFScores only locate candidates’ ideology and do not predict donation amounts. As we are interested in the ability for extreme candidates to raise small donations, these scores are transformed by calculating their absolute values to create two variables, Candidate CFScore (Folded) and Opponent CFScore (Folded). As a robustness check, in the appendix, we replicate our models using NOMINATE scores, with no notable differences in our substantive conclusions.

Control Variables

Contribution patterns of small donors among serial reporting committees suggest small donors are more likely to give to candidates running for elective office from the district in which they live, though earlier research suggested that there was little difference in inter-district between large and smaller donors (Gimpel et al., 2008). Thirty-seven percent of small donations reported by serial reporting campaigns are from contributors who reside in zip codes that overlay candidates’ congressional districts, although we caution that unknown biases may lurk in this statistic, for example, arising from differences between serial and nonserial reporting campaigns.¹¹ As at least some small donations come from within districts, we expect district demographic characteristics may correlate with aggregate small donations to candidates, and therefore include district-level control variables.

Although we cannot directly infer individual behavior with our aggregate data, given what we know from surveys, we include control district-level demographic variables of characteristics known to correlate with donor activity. We obtain most of our district-level demographic variables from Proximity 1, such as estimates of the natural log of the 2007 Median Household Income and 2008 Percent High School Educated.¹² We also include the Percent Rural population of districts and Percent White population, as reported by the Census Bureau.¹³ Previous research (Cho & Gimpel, 2007) suggests that the returns on the fundraising to dispersed populations are not worth candidates’ efforts. However, the Internet reduces the costs of contacting these constituencies (Karpf, 2013; Malbin, 2013; Panagopoulos, 2017; Wilcox, 2008), and incumbents have more time to develop such fundraising networks. We generally expect these variables to measure a general supply of small donors, with more potential small donors found in wealthier, highly educated, urban, and Whiter districts consistent with prior research (e.g., Francia et al., 2003; Graf et al., 2006). District-level demographics may also help control for the inherent differences in the costs of elections across districts, which may affect candidates’ demand for contributions.

Finally, we control for the party of the member as it may be possible that candidates of the two political parties have different strategies to raise money (Francia et al., 2003, 2005; Karpf, 2013; Wilcox, 2008). We measure candidate partisanship as a binary variable, with Democrats coded as 1. To measure potentially increasing small donor activity across the three elections we study, and variation across midterm and presidential elections, we include two dichotomous year fixed-effect variables for 2008 and 2010, with the omitted reference category being 2006.

Results

We present the results predicting the aggregate small donor contributions from our three statistical models of incumbents (Model 1), challengers to incumbents (Model 2), and candidates in open seats (Model 3) in Table 1. We report coefficients with robust standard errors to account for the presence of heteroscedacity.

Table 1.

House Candidate Small Donor Contributions, 2006-2010.

Variables	(1)	(2)	(3)
Variables	Incumbents	Challengers	Open seats
Material
Leader	0.014
Leader	(0.114)
Instrumental
Competitiveness	1.600***	6.359***	3.902***
Competitiveness	(0.235)	(0.409)	(0.643)
Primary Competitiveness	−0.048	−0.161	−0.083
Primary Competitiveness	(0.154)	(0.123)	(0.184)
Quality Candidate		0.512***	0.398***
Quality Candidate		(0.100)	(0.143)
Quality Opponent	0.204***		−0.252*
Quality Opponent	(0.074)		(0.138)
Freshman	0.123	0.187
Freshman	(0.099)	(0.126)
Scandal Incumbent	0.155	1.127***
Scandal Incumbent	(0.162)	(0.337)
Purposive
Candidate CFScore (folded)	1.331***	0.217	−0.349
Candidate CFScore (folded)	(0.114)	(0.157)	(0.569)
Opponent CFScore (folded)	−0.051	−0.031	0.021
Opponent CFScore (folded)	(0.084)	(0.118)	(0.113)
Controls
Percent White		−0.015***	−0.006
Percent White	(0.003)	(0.006)	(0.008)
Median Household Income (log)	0.337	−0.070	−0.581
Median Household Income (log)	(0.209)	(0.282)	(0.515)
Percent High School	0.015**	0.033***	0.108***
Percent High School	(0.007)	(0.010)	(0.032)
Percent Rural	0.005**	0.001	0.004
Percent Rural	(0.002)	(0.003)	(0.005)
Democrat	0.144*	0.069	0.185
Democrat	(0.074)	(0.108)	(0.138)
2008	0.010	0.084	−0.397**
2008	(0.084)	(0.121)	(0.171)
2010	0.254***	0.445***	−0.205
2010	(0.086)	(0.124)	(0.175)
Constant	2.858	4.162	6.040
Constant	(2.011)	(2.740)	(4.588)
Observations	849	832	182
R ²	.350	.433	.453

Note. Robust standard errors in parentheses. CFScore = Campaign Finance Score.

p < .1. **p < .05. ***p < .01.

The dependent variable is the natural logarithm of small contributions, so the independent variables’ coefficients must be transformed using the exponential function to estimate the substantive effect of an independent variable on the raw aggregate amount of small donor contributions. The natural logarithmic transformation of the dependent variable implies a nonlinear relationship between the (untransformed) dependent and independent variables, which means that the effect of one independent variable is contingent on the values of the other independent variables. Following Wooldridge (2013, p. 213), we compute the predicted value of aggregate small donations, $\hat{y}$ , as follows:

\hat{y} = {\hat{a}}_{0} \exp (\hat{\log y}),

Where $\hat{\log y}$ is the fitted values from the linear regression with the natural logarithm of small contributions as the dependent variable and ${\hat{a}}_{0}$ is the smearing estimate (Cox & Munger 1989; Duan, 1983), which is the simple average of the exponent of the residuals, ${\hat{u}}_{i}$ , given by,

{\hat{a}}_{0} = n^{- 1} \sum_{i = 1}^{n} \exp ({\hat{u}}_{i}),

Because ${\hat{a}}_{0} \geq 1$ , failure to use Equation 1 will typically result in an underestimate of the effect of a change of an independent variable on aggregate small donations. To estimate substantive effects of an independent variable of interest, we apply this adjustment and compute the average predicted effects at the means of all other variables.

Material Incentives

Small donors do not appear to be motivated by material incentives. As noted above, we observe relative parity among the candidate types, in contrast to prior research that finds that on average, incumbents consistently outraise challengers from all sources or just large donors (e.g., Francia et al., 2003; Herrnson, 2008; Krasno, Green, & Cowden, 1994). Furthermore, an incumbent who is a high profile party Leader is no more likely to raise small donations than their colleagues.

Instrumental Incentives

We observe general election Competitiveness figures strongly in small donor contribution patterns. For incumbents, the effect of being in the most competitive election compared with the least is worth on average US$110,933; for challengers, it is US$584,921; and for open seat candidates, it is US$233,583.¹⁴ All these relationships are statistically significant (p < .001). Compared with incumbents, the substantive effects of general election competition are more than 5 times greater for challengers to incumbents and twice for open seat candidates, suggesting to that small contributions to incumbents are less sensitive to the general election electoral environment than the other two candidate types.¹⁵ In contrast, Primary Competitiveness is unrelated to small donor fundraising by any candidate type and is signed in the wrong anticipated direction. For incumbents, this behavior is consistent with studies that find incumbents continually anticipate electoral challenges, even when none is currently present (Boatright, 2013; Fenno, 1977). This may similarly be true of other candidates and donors who anticipate the general election regardless of their primary environment.

Candidates’ competence, as measured by the presence of a quality candidate or challenger who has prior experience as an elected official, is related to small donor activity. All incumbents are quality candidates or opponents by definition, so we cannot estimate the effect of incumbents as candidates or incumbents as opponents. Challengers to incumbents who are Quality Candidates raise US$66,175 on average more from small donors (p < .01). Similarly well-qualified open seat candidates raise US$48,154 (p < .01). Jacobson (2013) finds quality candidates are more likely to contest an election when their opportunities for success are higher. Thus, the presence of a quality candidate may be an a priori indicator that an election will be competitive, spurring incumbents to raise more money anticipating a competitive election. We find the presence of a Quality Opponent to an incumbent results in modestly higher small donor fundraising by US$21,934 (p < .01). The effect is signed in the opposite direction as anticipated for open seat candidates, but is not statistically significant at conventional levels (p < .07). This finding may be confounded by the fact that open seat elections are highly competitive in general, thereby attracting more quality candidates and opponents.

Where we observe challengers raising small donations from characteristics of their incumbent opponents is where an incumbent is embroiled in a scandal. These challengers raise US$220,694 more than others (p < .001). Yet, at the same time, scandals have no discernable effect on incumbents’ fundraising prowess among small donors. We suspect scandals are a focus of personality-driven news, thereby raising the profile of challengers who would otherwise languish in obscurity. Finally, we observe Freshmen incumbents are no more or less capable of raising small donations than their colleagues or that their challengers raise more small donations to combat them with.

Purposive Incentives

We find support that ideologically extreme incumbents are able to raise more in small donor contributions. The most ideologically extreme incumbents raise US$306,557 more than their most moderate colleagues (p < .001). Yet, incumbents are the only candidate type where extremism correlates with small donor activity. Extremist challengers to incumbents and open seat candidates do not raise a greater amount of small donor contributions. We observe no reactive effect whereby candidates are able to raise money from their opponent’s extremism.

Control Variables

Among the district demographic variables, we find that the education level of a district’s constituents matters. When varied across its observed valid range, incumbents in districts with a highest Percent High School Education raise US$54,604 more than those in the least educated districts (p < .05); for challengers to incumbents, the differential is US$117,654 (p < .001); and for open seat candidates, the differential is US$308,027 (p < .001). In contrast, we observe no relationship between district median income and small donations as Median Household Income is not statistically significant in any of our three models. These relationships between education, income, and small donations are potentially susceptible to the ecological fallacy, but have construct validity, as they are consistent with the idea that income does not constrain donations, while education is generally associated with greater participation of all types (Schlozman et al., 2012). We cannot infer with certainty individuals’ motivations. There could be other reasons why we observe these patterns, such as more costly media markets are found in more highly educated districts.

The racial character of a district inconsistently correlates with small donor patterns. Incumbents in the highest Percent White districts raise US$68,401 more in small donations compared with those with the most non-White populations (p < .01). However, challengers to incumbents in the least Percent White districts raise US$183,785 more compared with those with the greatest White populations (p < .01). For open seat races, the racial composition is uncorrelated with small donor activity. For incumbents, this is consistent with previous research describing the donor pool as largely White (e.g., Francia et al., 2003). The pattern among challengers is intriguing, suggesting that challengers to incumbents in non-White districts may more often tap alternative fundraising networks when it comes to small donors, though this is an area for future research.

We observe the rural character of a district matters only for incumbents. Incumbents raised a modest average of US$41,671 more in the greatest Percent Rural districts compared with the most urban (p < .03). The effects for challengers to incumbents and open seat candidates is far from statistical significance. This finding is consistent with our expectation that incumbents more often build and support small donor networks in rural areas.

We find no statistically significant difference in small donor contributions to Democratic candidates compared with their Republican candidates, although the coefficients are positive in all three models and near statistical significance for Democratic incumbents who raise a modest US$14,610 more than Republican incumbents (p < .06). This suggests candidates of both parties are capable of tapping into the small donor pool.

Finally, the year fixed effects suggest an increase in small donor activity in 2010 compared with 2008 and 2006. However, the substantive effect is modest. Incumbents raised US$1,064 in 2010 compared with the 2006 baseline (p < .01) and challengers to incumbents raised US$9,397 more (p < .01). The effect for open seat candidates is signed in the opposite direction and not statistically significant. Together these modest and inconsistent results suggest to us changing campaign dynamics are largely responsible for differences in small donor activity between the 2006 and 2010 elections, even though total amounts of small contributions are increasing over this time period.

Discussion

Encouraging citizen participation is and should be a priority in any democracy. It is necessary to ensure the continued “responsiveness of the government to the preferences of its citizens” (Dahl, 1971, p. 1). But when such a small portion of the population contributes to campaigns and an even smaller portion of the population accounts for the majority of the money in campaigns, one might question to whom candidates will be responsive. Some believe small donors can counterbalance big money and broaden the base of citizen participation, whereas others worry that the potential ideological extremity of small donors may substitute one evil for another.

Our analysis of small donor contributions in the 2006 to 2010 congressional elections finds that small donors are motivated primarily by instrumental and purposive incentives, but not material incentives. With respect to material incentives, whereas Francia et al. (2003) conclude that “contributions [from large donors] flow primarily to incumbents, reinforcing the incumbent bias in the system” (p. 121), we observe small contributions flow fairly uniformly to all candidate types, rather than just to incumbents who can affect policy. Furthermore, the most influential incumbent party leaders raise similar small contributions compared with their peers.

With respect to instrumental motivations, we observe small contributions flow to all candidate types locked in competitive general elections. Another indicator of a competitive general election is the presence of a quality candidate, which stimulates small donation activity both by such candidates and their opponents who react to their presence. Competitive primary elections appear not to matter, suggesting that candidates in uncontested primary elections raise money in anticipation of the general election. Incumbents’ scandals matter as well, but only to aid fundraising for their challengers.

With respect to purposive incentives, Malbin (2013) argues against polarizing effects of small donors on two accounts: first, that those who argue for an effect in the popular press (Bonica, 2011; Klein, 2013) incorrectly draw inferences from a few outliers, and second, that a bivariate relationship of incumbent ideology and small donations reveals no effect. In addition, La Raja and Wiltse (2011) argue that candidates do not become more ideological to attract donors, rather they activate preexisting ideologues in the donor pool.

In contrast, we find the most ideologically extreme incumbents raise more small contributions. The effect of incumbents’ extremism is nearly 3 times as great as the effect of being in a highly competitive election. Yet, challengers and open seat candidates do not monetarily benefit from extremism. This suggests to us that candidates must first establish a record of their ideological extremism, and the most effective manner to do so is through congressional votes. Prospective ideologically motivated contributors can learn through media attention of extremist incumbents as to which incumbents to support. These findings are consistent with research finding that challengers to incumbents and open seat candidates do not enjoy the same level of public awareness as incumbents (Jacobson, 2013, p. 132), and thus do not attract the same level of attention. Competition and ideology thus matter, but in interesting ways. Small donors are not intrinsically polarizing as Karpf (2013) argues nor as benign as Malbin (2013) suggests. Rather, the results support the hypothesis La Raja and Wiltse (2011) advance, but only with respect to incumbents, as only extremist incumbents are successful at mobilizing ideologically motivated donors.

Our findings suggest that promoting greater participation by small donors may help with democratization of the electoral process by expanding participation in campaign financing, with money flowing more equally to all candidate types. We cannot at this time predict how the relative parity of small donor money flowing to incumbent, their challengers, and open seat candidates will affect the ideological composition of Congress in the long run. It may be that once new members win office, they change their behavior to be more ideologically driven to appeal to ideologically motivated donors. The study of longitudinal effects is a next logical step for investigations of small donors. Also of interest will be to contrast candidates’ small and large donor activities to better understand the trade-offs of encouraging candidates to pursue fundraising from these different types of donors.

Campaign finance reform, as a whole, is a complex policy question. Any change that limits money is likely to face a hostile Supreme Court in the near term. The same is true of any reform that enhances the speech of one group at the expense of another (Farrar-Myers & Dwyre 2008; Ortiz, 2005; Smith, 2001). Yet, the political parties may reach consensus on reforms to encourage small donors for philosophical reasons. Republicans generally favor increasing the role of money in elections and Democrats generally favor greater citizen participation in the electoral process. Since the Progressive Era, both parties have included small donor targeting in their fundraising strategies; historically, the Republican Party has been more effective at targeting small donations (Mutch, 2014). Still, strategic interests may intrude when one party is perceived to benefit from an electoral reform more than the other. We find candidates of both parties are equally capable of raising small donations, therefore, removing perhaps the largest political obstacle to future reform efforts.

Footnotes

Appendix

This appendix provides descriptive statistics for the variables in our analysis and to demonstrate that the results presented in Table 1 are robust to the measurement of candidates’ ideology. In Table A1, we substitute Bonica’s Campaign Finance Scores (CFScores) with incumbents’ first dimension NOMINATE scores. As challengers and candidates for open seats do not have NOMINATE scores, we drop nonincumbent ideology measures from Model 1 (incumbents) and Model 2 (challengers), and we do not report results for Model 3 (open seats). Table 1 results indicate that these variables are not statistically significant so there is no reason to believe that removing these variables from these models will change the estimated effects of the other variables, vis-à-vis omitted variable bias.

A comparison of the results presented in Table 1 and Table A1 confirms that the use of CFScores instead of NOMINATE scores likely does not affect our results. As in Table 1, small donor activity correlates for incumbents only, in a similar statistically strong direction. No variables of statistical significance in Table 1 change their direction or status in Table A1. Two variables that were not statistically significant in Table 1 have greater statistical significance, but their signs are not changed.

The variable indicating an incumbent is a Democrat is statistically significant at the p < .10 level in the Table 1 incumbent model and at the p < .001 in the Table A1 incumbent model. The substantive predicted effect for Democratic incumbents from the Table 1 model is US$14,610 and in Table A1, it is US$23,962, a difference of US$9,352.

The variable identifying the presence of a Freshman incumbent is not statistically significant in any Table 1 model; for the incumbent model, it is p < .22 and for the challenger model, it is p < .14. In Table A1, the coefficients are in the same direction and statistically significant at the p < .001 level in the Table A1 incumbent model and at the p < .10 level in the challenger model. For incumbents, the predicted effect in Table 1 is US$13,057 and now it is US$36,701, a difference of US$23,644. For challengers to incumbents, the predicted effect in Table 1 is US$22,041 and in Table A1, it is US$27,423, a difference of US$5,382.

These differences for these variables appear to be balanced on the knife-edge and fluctuate around statistical significance when measurement of the ideology variable is changed and when a measure of challengers’ ideology is omitted from the models. As our main substantive findings regarding instrumental and purposive motivations remain the same, we do not believe that these changes warrant favoring the use of NOMINATE scores in our statistical models.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Tyler Culberson received his MA in Political Science at George Mason University. He has worked at the Campaign Finance Institute and the U.S. Federal Election Commission.

Michael P. McDonald is an associate professor of Political Science at the University of Florida. He received his PhD in Political Science from the University of California, San Diego.

Suzanne M. Robbins is an assistant professor of Political Science at the University of Florida. She received her PhD in Political Science from Stony Brook University.

References

Aldrich

(1983). A Downsian spatial model with party activism. American Political Science Review, 77, 974-990.

Alexander

H. E.

(1984). Financing politics: Money, elections and political reform. Washington, DC: CQ Press.

Austen-Smith

(1995). Campaign contributions and access. American Political Science Review, 89, 566-581.

Barber

M. J.

Canes-Wrone

Thrower

(2017). Ideologically sophisticated donors: Which candidates do individual contributors finance. American Journal of Political Science, 61, 271-288.

Bartels

L. M.

(2005). Economic inequality and representation. Princeton, NJ: Princeton University.

Basinger

S. J.

(2013). Scandals and congressional elections in the post-Watergate era. Political Research Quarterly, 66, 385-398.

Basinger

S. J.

(2015). The electoral effects of congressional scandals. In Dagnes

Sachleben

(Eds.), Scandal! An interdisciplinary approach to the consequences, outcomes and significance of political scandal (pp. 67-82). New York, NY: Bloomsbury Press.

Bauer

Potter

(2013, May 9). A new recipe for election reform. The Washington Post. Retrieved from https://www.washingtonpost.com/opinions/a-new-recipe-for-campaign-finance-reform/2013/05/09/bfb2a1d0-b80b-11e2-b94c-b684dda07add_story.html?utm_term=.e96a7dbf5f3d

Boatright

R. G.

(2013). Getting primaried: The changing politics of congressional primary challenges. Ann Arbor: University of Michigan Press.

10.

Bonica

(2011, July 22). Small donors and polarization. Boston Review. Retrieved from http://www.bostonreview.net/bonica-small-donors-polarization

11.

Bonica

(2013). Database on ideology, money in politics and elections. Stanford, CA: Stanford University Libraries.

12.

Bonica

(2014). Mapping the ideological marketplace. American Journal of Political Science, 58, 367-386. doi:10.1111/ajps.12062

13.

Brown

C. W.

Powell

L. W.

Wilcox

(1995). Serious money: Fundraising and contributing in presidential nomination campaigns. New York, NY: Cambridge University Press.

14.

Cho

W. K.

Gimpel

J. G.

(2007). Prospecting for (campaign) gold. American Journal of Political Science, 51, 255-268.

15.

Christenson

D. P.

Smidt

C. E.

Panagopoulos

(2014). Deus ex Machina: Candidate web presence and the presidential nomination campaign. Political Research Quarterly, 67, 108-122.

16.

Corrado

Malbin

M. J.

Mann

T. E.

Ornstein

N. J.

(2010). Reform in an age of networked campaigns: How to foster citizen participation through small donors and volunteers. Washington, DC: Campaign Finance Institute.

17.

Cox

G. W.

Munger

M. C.

(1989). Closeness, expenditures, and turnout in the 1982 elections. American Political Science Review, 83, 217-231.

18.

Dahl

R. A.

(1971). Polyarchy: Participation and opposition. New Haven, CT: Yale University Press.

19.

Duan

(1983). Smearing estimate: A nonparametric retransformation method. Journal of the American Statistical Association, 78, 605-610.

20.

Ensley

M. J.

(2009). Individual campaign contributions and candidate ideology. Public Choice, 138, 221-238.

21.

Farrar-Myers

V. A.

Dwyre

(2008). Limits and loopholes: The quest for money, free speech, and fair elections. Washington, DC: CQ Press.

22.

Fenno

R. F.

Jr. (1977). US house members in their constituencies: An exploration. American Political Science Review, 71, 883-917.

23.

Francia

P. L.

Green

J. C.

Herrnson

P. S.

Powell

L. W.

Wilcox

(2003). The financiers of congressional elections: Investors, ideologues and intimates. New York, NY: Columbia University Press.

24.

Gais

(1998). Improper influence: Campaign finance law, political interest groups and the problem of equality. Ann Arbor: University of Michigan Press.

25.

Gimpel

J. G.

Lee

F. E.

Pearson-Merkowitz

(2008). The check is in the mail: Interdistrict funding flows in congressional elections. American Journal of Political Science, 52, 373-394.

26.

Graf

Reeher

Malbin

M. J.

Panagopoulos

(2006). Small donors and online giving: A study of donors to the 2004 presidential campaign. Washington, DC: Institute for Politics Democracy and the Internet in Collaboration With the Campaign Finance Institute.

27.

Hasen

R. L.

(2016). Plutocrats united. New Haven, CT: Yale University Press.

28.

Heard

(1962). The costs of democracy: Financing American political campaigns. Garden City, NY: Doubleday.

29.

Herrnson

P. S.

(1992). Campaign professionalism and fundraising in congressional elections. The Journal of Politics, 54, 859-870.

30.

Herrnson

P. S.

(2008). Congressional elections: Campaigning at home and in Washington (5th ed.). Washington, DC: CQ Press.

31.

Jacobson

G. C.

(1978). The effects of campaign spending in congressional elections. American Political Science Review, 72, 469-491. doi:10.2307/1954105

32.

Jacobson

G. C.

(1980). Money in congressional elections. New Haven, CT: Yale University Press.

33.

Jacobson

G. C.

(2013). The politics of congressional elections (8th ed.). New York, NY: Pearson/Longman.

34.

Jacobson

G. C.

Dimock

M. A.

(1994). Checking out: The effects of bank overdrafts on the 1992 elections. American Journal of Political Science, 38, 601-624.

35.

Johnson

(2010). Individual contributions: A fundraising advantage for the individually extreme? American Politics Research, 38, 890-908.

36.

Johnson

(2013). Political giving: Making sense of individual campaign contributions. Boulder, CO: Lynne-Reiner Publishers.

37.

Karpf

(2013). The internet and American political campaigns. The Forum, 11, 413-428.

38.

Klein

(2013). Big money corrupts Washington. Small donors polarize it. Retrieved from http://www.washingtonpost.com/blogs/wonkblog/wp/2013/05/10/big-money-corrupts-washington-small-donors-polarize-it/

39.

Koetzle

(1998). The impact of constituency diversity upon the competitiveness of US house elections, 1962-1996. Legislative Studies Quarterly, 23, 560-573.

40.

Krasno

J. S.

Green

D. P.

Cowden

J. A.

(1994). The dynamics of campaign fundraising in house elections. The Journal of Politics, 56, 459-474. doi:10.2307/2132148

41.

La Raja

R. J.

Wiltse

D. L

. (2011). Don’t blame donors for ideological polarization of political parties: Ideological change and stability among political contributors, 1978-2008. American Politics Research, 40, 501-530.

42.

Lipsitz

Panagopoulos

(2011). Filled coffers: Campaign contributions and contributors in the 2008 elections. Journal of Political Marketing, 10, 43-57.

43.

Luntz

(1988). Candidates, consultants, & campaigns. New York, NY: Basil Blackwell.

44.

Malbin

M. J.

(Ed.). (2003). Life after reform: When the Bipartisan Campaign Reform Act meets politics. New York, NY: Rowman & Littlefied.

45.

Malbin

M. J.

(2011). Small donors, large donors, and the internet: Rethinking public financing for presidential elections after Obama. In Panagopoulos

(Ed.), Public financing in American elections (pp. 36-61). Philadelphia, PA: Temple University Press.

46.

Malbin

M. J.

(2013). Small donors: Incentives, economies of scale, and effects. The Forum, 11, 385-411.

47.

Malbin

M. J.

Brusoe

P. W.

Glavin

(2012). Small donors, big democracy: New York City’s matching funds as a model for the nation and states. Election Law Journal, 11, 3-20.

48.

Mansbridge

J. J.

(2003). Rethinking representation. American Political Science Review, 97, 515-528.

49.

McDonald

M. P.

Tolbert

C. J.

(2012). Perceptions v. Actual Exposure to Electoral Competition and Effects on Political Participation. Public Opinion Quarterly, 76, 538-554.

50.

Mutch

R. E.

(2014). Buying the vote: A history of campaign finance reform. New York, NY: Oxford University Press.

51.

Ortiz

D. R.

(2005). The first amendment and the limits of campaign finance reform. In Corrado

Mann

T. E.

Ortiz

D. R.

Potter

(Eds.), The new campaign finance sourcebook (pp. 91-122). Washington, DC: Brookings.

52.

Panagopoulos

(2017). Political campaigns: Concepts, context, and consequences. New York, NY: Oxford University Press.

53.

Poole

K. T.

Rosenthal

(1997). A political-economic history of roll call voting. New York, NY: Oxford University Press.

54.

Potter

Morgan

B. B.

(2013). Campaign finance: Remedies beyond the court. Democracy: A Journal of Ideas, 27, 38-45.

55.

Riker

W. H.

Ordeshook

(1968). A theory of the calculus of voting. American Political Science Review, 62, 25-42.

56.

Rosenberg

(2002). Broadening the base: The case for a new federal tax credit for federal contributions. Washington, DC: American Enterprise Institute.

57.

Schlozman

K. L.

Verba

Brady

(2012). The unheavenly chorus: Unequal political voice and the broken promise of American democracy. Princeton, NJ: Princeton University Press.

58.

Smith

B. A.

(2001). Unfree speech: The folly of campaign finance reform. Princeton, NJ: Princeton University Press.

59.

Sorauf

F. J.

(1988). Money in American elections. Glenview, IL: Scott, Foresman/Little, Brown College Division.

60.

Sorauf

F. J.

(1992). Inside campaign finance: Myths and realities. New Haven, CT: Yale University Press.

61.

Vaccari

Nielson

R. K.

(2013). What drives politicians’ online popularity? An analysis of the 2010 midterm elections. Journal of Information Technology & Politics, 10, 208-222.

62.

Weyl

W. E.

(1913). The democratization of party finances. American Political Science Review, 7, 178-182.

63.

Wilcox

(2008). Internet fundraising in 2008: A new model. The Forum, 6(1), 1-16.

64.

Wilcox

Cooper

Francia

P. L.

Green

J. C.

Herrnson

P. S.

Powell

L. W.

. . . Webster

B. A.

(2003). With limits raised, who will give more? The impact of BCRA on individual donors. In Malbin

M. J.

(Ed.), Life after reform: When the Bipartisan Campaign Reform Act meets politics (pp. 61-79). Washington, DC: Campaign Finance Institute.

65.

Wooldridge

J. M.

(2013). Introductory econometrics: A modern approach (5th ed.). Mason, OH: South-Western Cengage Learning.