Card-Carrying Members on the Ballot: Group Affiliations and Their Electoral Implications

Abstract

How members of Congress align with and promote the interests of their coalitional allies is a concern for academics and the public alike. To understand the relationship between interest groups and officeholders, we must examine group ties prior to a candidate’s electoral victory. From there, we can begin to evaluate the partisan dynamics and electoral implications of candidate-group alignments. How do Democrats and Republicans differ in terms of group affiliation? Moreover, what are the consequences of these affiliations for a candidate’s performance at the ballot box? To answer these questions, we turn to a novel dataset on the recorded professional and civic group affiliations of all US congressional candidates from 2008–2014. This dataset, built on Vote Smart data, allows us to comprehensively assess the role of these group affiliations in recent elections. To complement these observational data, we conducted a survey experiment to test whether specific group affiliations influence officeholder evaluations. Ultimately, we find that few partisan groups exert any significant effect on evaluations. We find some evidence that non-partisan groups with positive reputations can mitigate outgroup hostility among opposing partisans.

Keywords

US congress interest groups civic groups congressional elections

Introduction

In 2018, Lucy McBath (D-GA) flipped a long-held Republican district in the Georgia suburbs. The seat, once held by former Speaker Newt Gingrich, represented a significant victory for the Democratic Party and its coalitional allies. Despite little prior experience in electoral politics, McBath’s candidacy was bolstered by establishment figures and the grassroots alike. In the wake of her son’s tragic death by gun violence, McBath rose to national prominence as an activist and spokesperson for groups such as Mothers of the Movement, Moms Demand Action, and Every Town. McBath’s candidacy was inextricably tied to these groups, and her role as a leading gun control activist. Few members of Congress can chart their rise in politics in such a clear trajectory from group activist to elected official. Nonetheless, would-be candidates often cut their teeth in civic organizations long before they ever consider running for office. Most officeholders belong to or affiliate with numerous civic groups, including many with clear issue positions. Moreover, as Fenno (1978) observed with his travel companions, politicians are quick to highlight these affiliations as evidence of their grassroots support or to demonstrate real-world policy expertise. What do these group affiliations signal to voters and how do they affect the electoral fortunes of candidates?

Scholarship on party coalitions contends that policy demanders play a key electoral role in deciding not only who runs for office but also in determining the positions candidates take along the way (Cohen et al., 2008; Karol, 2009). In addition to shaping the policy priorities of the candidates themselves, interest groups may explicitly offer and advertise endorsements of candidates to signal their policy positions (Arceneaux & Kolodny, 2009; Gimpel, 1998; McDermott, 2006). While these endorsements do matter, does membership in a particular group signal a stronger commitment to its cause? Presumably, membership in a well-known group should indicate the candidate’s sympathies with the organization, providing a clear sense of their legislative priorities to voters.

Group affiliations, particularly social groups, are believed to be central to how voters think about parties and officeholders (Barber & Pope, 2022; Converse, 2006; Grossmann & Hopkins, 2015, 2016; Kane et al., 2021; Mason & Wronski, 2018). While social groups differ from civic organizations and issue advocacy groups, these associations nevertheless likely help color voter perceptions of the candidates. Research on heuristics finds that individuals often use their knowledge or affect towards different groups to make an “informed” decision when casting a ballot or choosing between policy proposals (Arceneaux & Kolodny, 2009; Brady & Sniderman, 1985; Lupia, 1994). Yet, recent work by Broockman et al. (2021) raises important questions as to whether voters possess the requisite knowledge about politically active groups to use these heuristics effectively. What, then, do interest group ties signal to voters about a candidate’s ideology or policy priorities? Moreover, how does group affiliation predict or influence success at the ballot box?

Despite considerable effort to answer these questions, existing work has been unable to document a consistent link by which group membership or association affects the electoral fortunes of candidates. This project endeavors to understand how groups shape legislative priorities by first identifying how these linkages connect to parties, and how they persist throughout an officeholder’s electoral career. In adopting a broader lens, we can begin to evaluate the electoral implications of candidate-group alignments.

This paper’s empirical study proceeds in three steps. First, taking a more descriptive approach, we identify which candidates belong to, or affiliate with, various groups. To map out this complicated interest group landscape, we created a novel dataset using biographical information from Vote Smart. Using these data, we explore how professional and civic group affiliations differ by party among all general-election US congressional candidates from 2008–2014. This analysis represents an important contribution in its own right by providing a comprehensive look at the partisan dynamics of candidate-group linkages. We find partisan differences that are consistent with perceptions of the respective party coalitions. For instance, Democratic candidates are much more likely to be members, or former members, of labor unions and environmental groups. Republican candidates, by contrast, tended to associate more with Christian groups, business associations, and gun-rights organizations.

Second, after mapping out these candidate-group connections, we then examine whether any group affiliations predict congressional elections. To account for candidate and district-level variation, we used a series of multivariate models to study whether winning candidates are more likely to belong to specific groups. Compared to other well-known electoral predictors, we find relatively modest effects of group membership on candidate vote share. Indeed, our findings suggest that very few groups, as Broockman et al. (2021) contend, carry an electoral benefit or penalty. Finally, building on the observational analyses, we deployed a survey experiment to study how specific group affiliations shape evaluations of officeholders. By manipulating both party and group, we are better able to test the effects of group ties on candidate evaluations. We find that only the most well-known interest groups wield a significant influence over evaluations. For instance, the National Rifle Association (NRA) boosts support among Republican subjects and opposition among Democrats. While some partisan groups (e.g. the AFL-CIO) appear to depress support among opposing partisans, these results are less stable and suggest the limited informational value of group affiliations. That said, we find preliminary evidence that positively viewed non-partisan groups can help soften a candidate’s image among opposing partisans. We conclude with a brief discussion of the project’s implications for a wide range of literature, from candidate strategies and elections to interest group politics and legislative effectiveness.

Groups, Candidates, and Voters

From card-carrying members of the NRA to local Parent Teacher Association (PTA) presidents, candidates for office hail from countless groups of which the costs and responsibilities of membership are as varied as the groups themselves. At a more basic level, considerable research has examined the role that social groups, particularly race and religious affiliation, play in shaping how voters understand candidates, parties, and issues (Converse, 2006; Green et al., 2002; Campbell et al., 2011; Mason & Wronski, 2018; Elder & O’Brian, 2022). When an office seeker belongs to one of these well-known social groups, voters are capable of inferring where the candidate stands on a host of issues. Few scholarly works, however, explicitly study whether civic and advocacy group membership affect these outcomes. While membership or association with these groups may be less salient to voters, it may actually convey more information about a candidate’s issue priorities or positions. What does it mean for prospective officeholders to not just align with an interest group, but count themselves as a member? For the individual candidate, these connections have the potential to shape policy positions, structure their priorities during a campaign, and constrain legislative behavior upon entering office. Shifting the lens to constituents, how do voters interpret these affiliations? Do group memberships signal any additional political or ideological information to a representative’s constituents?

Partisan Cues and Group Heuristics

Public opinion research has long held that groups serve an important role in anchoring voters’ understanding of the political world (Converse, 2006; Goggin et al., 2020). Indeed, the images of parties are often tied to social groups, especially with these groups better sorted along partisan lines (Campbell et al., 2011; Mason, 2018; Mason & Wronski, 2018). For instance, Brady and Sniderman (1985) found that individuals can use their own affective disposition towards a group to successfully identify the attitudes of the group. This “likability heuristic,” as Brady and Sniderman contend, suggests that voters infer group positions with relatively little information. Indeed, recent work by Elder and O’Brian (2022) suggests that voters possess extensive knowledge of where social groups stand, and that such information plays a key role in promoting ideological coherence. Consequently, candidates are likely to be viewed in very different ways depending upon their individual positioning within these groups. Indeed, existing work finds that a candidate’s race, gender, and religion can indicate a candidate’s ideology to voters (McDermott, 1998, 2007). Additionally, we might expect these group affiliations to matter differentially between the two parties. If groups are central to how voters understand the Democratic party and relatively less important to views about of the GOP (Grossmann & Hopkins, 2015, 2016), then the effects of candidate affiliations are likely conditional on the party of both constituents and elites.

These studies, while compelling, largely focus on large, well-known social groups within American society. Race and religion are potent signals to voters, but what of the myriad groups that are active in politics? Moreover, how does the public perceive relatively apolitical civic organizations, let alone advocacy groups that actively lobby government?

Though less salient than large social groups, civic and issue advocacy groups are similarly entangled with how voters understand parties and issues, and should arguably provide more meaningful political signals. While civic group membership may lack significant political meaning, issue advocacy groups have clear, often explicit, associations with policy positions for voters and can use these explicitly to affect candidates’ electoral fortunes with their endorsements (McDermott, 2006). For individuals who lack an “encyclopedic” knowledge of politics, locating an interest group’s position is often sufficient for making an “informed” decision. Lupia (1994) finds that group-based heuristics can be powerful tools in even the most complex decision-making environments, for example choosing between several competing ballot initiatives. If a group’s reputation is well-known, then their support or opposition can also shape how voters see candidates. In a field experiment deployed during a competitive state election, Arceneaux and Kolodny (2009) found that some voters, for example low-info Republicans, used an abortion group’s support for the candidate to inform their choice at the ballot box. Moving beyond endorsements, one would think that a more direct affiliation with an organization, for example membership, would signal an even tighter bond with the group’s values and legislative priorities.

While these studies present a fairly optimistic view of the democratic competence of the electorate (Lupia & McCubbins, 1998), more recent work suggests that voters often misperceive not only the positions but the very composition of groups (Ahler, 2014; Ahler & Sood, 2018). If voters are unaware of how groups relate to parties, let alone policy issues, then a candidate’s connection is unlikely to shift attitudes in any meaningful way. Indeed, Lau and Redlawsk’s (2001) experimental work suggests that heuristics are utilized most effectively by politically sophisticated voters. Low-information voters, by contrast, are less adept at using endorsements as a shortcut to reach the “correct” decision. Moreover, Broockman et al. (2021) find that voters rarely possess the requisite knowledge to effectively utilize a group’s position to evaluate a member of Congress. Instead, citizens routinely engage in a process of “heuristic projection” whereby individuals assume that a group’s positions reflect their own. Beyond merely casting doubt on the value of heuristics, Broockman et al.’s results suggest that interest group ratings can serve to mask a politician’s actions and are, thus, not just ineffective, but harmful to democratic accountability.

Cheap Affiliations and Noisy Signals

Despite the inconsistent findings with endorsements and group ratings, existing scholarship hints at conditions in which group affiliations are likely to matter to voters, and why candidates would highlight their association. Like an endorsement, a candidate’s group membership serves as a meaningful signal to strong partisans, activists, and other policy demanders. In particular, we argue that group affiliations can convey both organizational and ideological information to attentive voters.

First, group affiliations, like endorsements, can signal a candidate’s potential electoral support to voters (McKelvey & Ordeshook, 1985). By linking the candidate to a mass-based or local civic association, an affiliation informs voters and other potential supporters that the candidate has a credible chance of winning. Membership in a large, grass-roots group suggests that the candidate has a reserve of organizational support to draw upon come campaign season (Rosenstone & Hansen, 1993). In addition to help with mobilization, these ties may indicate a higher quality candidate who has more experience in the public eye. Special interest organizations and professional associations, by contrast, indicate that the candidate has the backing of wealthy groups, a key indicator of candidate quality and electoral viability. In short, group affiliations suggest that a candidate has strong electoral prospects to strategic actors (e.g. donors, party leaders) and constituents alike.

Second, depending upon the group’s policy reputation, a candidate’s affiliation communicates ideological information to the electorate. For groups with well-known positions, such as the NRA or labor unions, membership with the group signals to voters that the candidate sympathizes with the group’s policy positions and would support their agenda while in office. If voters like or dislike the group, then an affiliation can help voters decide where they stand towards the candidate (Brady & Sniderman, 1985). With affective polarization on the rise (Iyengar et al., 2012, 2019), voters are predisposed to be hostile to opposing partisans even with some potential ideological agreement. Yet, an association with a positively-viewed group may soften an out-party candidate. Conversely, affiliations with a well-known, negatively-viewed group would exacerbate outgroup hostility. For instance, Democrats should evaluate Republican candidates associated with positive groups, like the PTA, more favorably than a candidate affiliated with the NRA.

Given these potential tradeoffs, candidates have strategic incentives to either advertise or downplay their group memberships in campaigns. Indeed, it is not inconceivable that candidates strategically develop these associations, at least partly, for electoral reasons. When a group’s position is consistent and well-established, the connection to a politician sends a clear signal to potential supporters (or opponents). While attentive voters are unlikely to view many partisan groups of the opposing coalition favorably, candidates can and do tout their connection to civic organizations with neutral or positive reputations. Compared to partisan groups, these affiliations are unlikely to hurt a candidate’s prospects, but may offer fewer positive benefits in return.

Before considering whether candidates benefit from membership in a specific group, we first need to understand the group affiliations themselves. To that end, the next section details our process of identifying and classifying candidate-group memberships. This descriptive analysis examines the presence of group memberships themselves without considering the advertising or endorsements of these associations in electoral campaigns. Through this first look we can see the full map of organizational associations, including many that may not be salient to voters. Indeed, it is likely that voters are unaware of many of these group affiliations. Nevertheless, this first analysis helps us map the full picture of these associations that campaigns can draw on in their campaign advertising and messaging strategies.

Candidate-Group Membership Data

How members of Congress align with and promote the interests of their coalitional allies is an issue of concern for academics and the public alike, and yet we know relatively little about the early development of these linkages. Rather than starting with officeholders, scholarship must consider how candidates, both winners and losers, develop and advertise their links to groups at the local and national level. In order to ascertain candidates’ membership in civic organizations, we collected candidate biographies from Vote Smart, https://www.votesmart.org. Vote Smart collects these biographies by reaching out to ballot-qualified candidates with the “Political Courage Test,” which asks candidates a broad array of personal and policy-related questions. While not all candidates reply, Vote Smart employees and volunteers make efforts to achieve as high a response rate as possible. An example biography is shown in Figure 1.

To systematize this information, we utilized the Vote Smart API (VoteSmart, 2015) and the R package ‘pvsR’ (Matter, 2014) to download this information for all general election Democratic & Republican US House and Senate candidates in 2008, 2010, 2012, and 2014.¹ Particularly, we focus on the “candidate.orgMembership” field, which contains responses from candidates about all “Religious, Civic, and other Memberships” they hold.² The candidates are allowed to list as many organizations as they wish in this field, with the average being 3.72. A histogram of the number of organizations listed can be seen in Figure 2. Notably, 32.16% of candidates did not list any civic organizations whatsoever.³ This is a slightly higher missing data rate than for many other fields in the Vote Smart data. For example, “candidate.family,” which lists a candidate’s familial information, has a completion rate of 87.2%. Not unexpectedly, candidates who listed no civic organizations in this field tend to be lower quality and less likely to win.⁴

This project utilizes two distinct datasets. First, to examine the relationship of these civic organizations with partisanship and other candidate-level variables, we analyze a dataset where each Democratic & Republican, House & Senate, general election candidate from 2008–2014 is an observation. This dataset contains 2267 unique candidates. Second, to examine the electoral prospects of these candidates, we build a dataset containing candidate-election observations for Democratic & Republican, House & Senate, general election candidates in 2008, 2010, and 2012. This dataset contains a wide variety of FEC data and other electoral variables, including a candidate’s two-party vote share in each election.⁵

In order to examine how these listed organizations are related to partisanship, electoral success, and behavior within the House and Senate across these datasets, we rely on a combination of both automated and manual text analysis. First, the data field containing candidates’ organizational affiliations was preprocessed, removing capitalization, punctutation, numbers, stopwords, and whitespace. Second, words were stemmed to render pluralized words and those with suffixes comparable. Next, a document-term matrix was generated for each unit of observation, and to reduce sparsity, only words that appeared a minimum of ten times in the full dataset were kept. This leaves us with 506 words in the candidate dataset and 819 in the candidate-election dataset. While we analyze the full set of terms in some descriptive analyses, we code several meaningful categories for easier analysis. Table 1 shows twenty categories of terms that were coded. If any one of the component strings were found in this field, the variable was coded as 1, else 0.

Table 1.

Categorization of word strings.

Variable	Stemmed Component strings
Chamber of Commerce	chamber, commerc
K-12	elementari, school, teacher, pta
College	academi, alumni, colleg, univers
Environment	environ, environment, natur, park, sierra
Gun	gun, nra, rifl
Scouts	scout, scoutmast
Legal	attorney, bar, law, lawyer, legal
Rep party org	republican
Dem party org	democrat
Military	militari, naval, navi, veteran, vfw, war
Christian	baptist, bibl, christ, christian, lutheran, episcop, methodist, minist, ministri, presbyterian, cathol, cathedr
Jewish	jewish
Sports	athlet, basebal, basketbal, coach, soccer, sport
Public safety	crime, fire, polic, polici, safeti, sheriff
Unions	union
Health	health, cancer, hospit, medic, mental, nurs, organ
Civic	auxiliari, civic, club, communiti, downtown, elk, greater, homeown, kiwani, knight, legion, neighborhood, rotari, societi, ymca, ywca
Charity	angel, fund, habitat, homeless, literaci, salvat, servic, shelter, volunt
Business	busi, chamber, commerc, corpor, develop, econom, financ, financi, industri
Farm	agricultur, farm, farmer

In deciding on both the categories and the component strings, we tried to select only words that would indicate with extremely high certainty that a candidate was a member of a particular group or group type. Consequently, while we may miss some candidates in this manner, we chose to err on the side of exclusion of candidates unless it was clear they were a member of a given category or organization. For robustness, we also conducted a more comprehensive, manual approach integrating not just single words or word stems but phrases. This alternative grouping, compiled from manual inspection of the data, involves searching for 519 different words or phrases (case insensitive, punctuation removed) across 24 categories.⁶

Group Affiliations and Partisanship

Descriptively, we know little about how the group affiliations of candidates vary by party. Are Democratic candidates indeed more likely to belong to progressive groups, and do Republicans hail from conservative organizations? Are there partisan differences between professional and other civic organizations with less clear issue or ideological valence? To study the relationship between party and organizational affiliation, Figure 3 shows the log odds predicting whether a candidate is Democratic or Republican from the top 50 most predictive words, out of 506, in the document-term matrix.⁷ The words on the right are associated with higher probabilities of the candidate being a Democrat, while those on the left are associated with higher probabilities of the candidate being a Republican. Because the axis is log odds, the terms “democrat” and “republican” suggest, unsurprisingly, that the odds that the candidate is of that partisan affiliation are 100:1. More substantively, the odds that a candidate is Republican are 10:1 if they claim “nra” membership in their biography. For a plot of the full list of words (N = 506), see Figure 7 in the Appendix.⁸

While many of the group affiliations conform strongly to partisan stereotypes and expectations, not all of the words shown in Figure 3 are meaningful to our understanding of candidates’ civic group affiliations. To better understand the relative membership of common categories, Table 2 displays the percentage of Democratic and Republican candidates belonging to each category shown in Table 1.⁹ Of note are the relatively low percentages for both parties among many of the categories, which is partially due to the lower ceiling given that over 30% of candidates listed zero group memberships. Even with this lowered ceiling, there appears to be remarkable variation across categories with some organizations, for example civic, Christian organizations, far more likely to be mentioned. We observe statistically significant and expected partisan differences for many politically relevant categories. For instance, candidates with membership in environmental groups and labor unions are more likely to be Democrats. Conversely, Republican candidates overwhelmingly make up gun-rights organizations. We also find striking differences among religious groups and professional associations. Democrats are more likely to list Jewish groups and legal organizations. Meanwhile, Republican candidates represent a greater share in Christian, farm, military, and business organizations.¹⁰

Table 2.

Percentage of candidates in categories, by party.

Variable	% of Democrats	% of Republicans	p-value of difference
Chamber of Commerce	7.84	13.69	0.001
K-12	10.96	10.82	0.916
College	14.92	14.34	0.694
Environment	5.48	3.98	0.091
Gun	0.93	8.33	0.001
Scouts	4.05	6.85	0.004
Legal	16.02	8.60	0.001
Party org	9.27	11.75	0.056
Military	3.63	6.66	0.001
Christian	15.51	24.14	0.001
Jewish	1.69	0.74	0.038
Sports	3.71	7.03	0.001
Public safety	4.64	5.64	0.280
Unions	4.05	1.30	0.001
Health	14.92	14.62	0.837
Civic	28.67	35.99	0.001
Charity	14.59	12.30	0.111
Business	15.68	20.35	0.004
Farm	1.85	5.27	0.001

The Electoral Fortunes of Group Affiliations

The previous analysis suggests that while many group affiliations appear weakly associated with partisanship, some groups are composed disproportionately of one party or the other. Groups with wide partisan gaps have the potential to serve as heuristic anchors to informed voters, but do they? Are voters aware of these associations? Moreover, what should we make of the numerous groups that carry less obvious partisan differences? Though lacking a partisan cue, these memberships could demonstrate other social ties that help with fundraising or electoral success more broadly. For instance, civic group membership can signal the candidate’s local roots or commitment to the community. To assess the relationship between various groups and electoral success, we employ our second dataset, containing all 2008–2012 Democratic & Republican, US House & Senate, general election candidates, with the unit of observation being a candidate-election.

To systematically examine the relationship between group membership and electoral success, Table 3 displays two OLS regressions of a candidate’s two-party vote share (scaled 0–100) regressed on various control variables and the meaningful group categories shown in Table 1.¹¹ Because we might expect some group memberships to mean different things for Democrats versus Republicans, we display separate regressions for candidates of each party.¹²

Table 3.

Electoral vote share regressed on organization membership (OLS), 2008–2012.

	Dependent variable: Vote share (0-100)
	Democrats	Republicans
2004 vote share	0.656***	0.660***
	(0.044)	(0.037)
Incumbent	13.013***	13.097***
	(1.236)	(0.968)
Total $ spent	0.050	−0.760
	(0.331)	(0.600)
Total unique givers	0.205**	1.159*
	(0.103)	(0.686)
CF-score	2.834***	−2.914**
	(0.940)	(1.304)
Female	0.964	−2.307**
	(0.914)	(1.007)
No orgs listed	1.074	1.307
	(1.159)	(1.145)
# of orgs listed	0.376**	0.364***
	(0.147)	(0.130)
Chamber of Commerce	0.396	−1.287
	(1.305)	(1.560)
K-12	−2.380**	−0.228
	(1.182)	(1.223)
College	0.010	−0.647
	(1.043)	(1.096)
Environment	−1.639	−0.576
	(1.501)	(1.463)
Gun	1.560	−0.596
	(2.185)	(1.125)
Scouts	−0.840	0.868
	(1.596)	(1.453)
Legal	0.999	−0.661
	(1.007)	(1.311)
Dem party org	0.778
	(1.108)
Rep party org		−1.227
		(1.177)
Military	0.737	0.394
	(1.198)	(1.116)
Christian	0.646	1.560*
	(0.982)	(0.900)
Jewish	−4.412	4.560
	(2.865)	(5.609)
Sports	0.923	−1.287
	(1.725)	(1.117)
Public Safety	−2.146	−1.562
	(2.371)	(1.687)
Unions	3.194*	2.564
	(1.640)	(3.255)
Health	0.364	0.263
	(0.876)	(1.104)
Civic	0.508	−0.795
	(0.837)	(0.887)
Charity	1.481	0.204
	(0.961)	(0.998)
business	−0.165	2.007
	(0.986)	(1.446)
Farm	−3.864	0.050
	(2.528)	(1.387)
2010 cycle	−6.944	9.004***
	(0.902)	(0.915)
2012 cycle	−2.360**	5.990***
	(1.095)	(1.021)
Constant	14.419***	4.489*
	(2.457)	(2.454)

Note. $N_{dems} = 1721$ , $N_{reps} = 1531$ . Adj. $R_{dems}^{2} = 0.63$ , Adj. $R_{reps}^{2} = 0.59$ .

SE’s clustered at the level of the candidate.

*p $<$ 0.1; **p $<$ 0.05; ***p $<$ 0.01.

First, as expected, many of our control variables significantly predict candidates’ electoral success for both Democratic and Republican candidates.¹³ The ideological tilt of the district, measured by the party’s vote share in 2004, and the candidate’s status as an incumbent are both strongly associated with electoral success. Ideological extremity, as measured by Bonica’s CF-score, is also highly predictive for candidates, with the models suggesting that less extreme candidates perform better in general elections.¹⁴ To rule out the possibility that results are driven by whether candidates completed the organization field in the Vote Smart data, we control for the field being left blank.¹⁵ We also included a variable to capture the candidate’s overall number of listed organizations, as shown in Figure 2. Unsurprisingly, we find that listing more groups is positively associated with increased vote share. More specifically, the data suggest that each additional group is associated with an increase of 0.3-0.4%. While this result is causally suspect, it nonetheless suggests that candidates with more group affiliations tend to do better at the ballot box. This is consistent with the argument that candidates embedded in interest group networks have more resources (e.g. donations, campaign volunteers) to rely upon and perform better accordingly.

With respect to the various group categories, we find that several organizational types are associated with electoral performance.¹⁶ That said, the results cast doubt on a universal group effect for Democrats and Republicans. In other words, group affiliations appear conditional on the partisanship of the candidate. For example, we find that Democratic candidates with an affiliation to a K-12 education group tend to receive 2.4% lower vote share; we find no significant effect for Republicans. On the positive-side, Democrats who are members of a labor organization experience 3.2% higher returns. For Republicans, we see only one positive and statistically significant effect of group membership. GOP candidates who listed a Christian affiliation are associated with an expected 1.6% increase in two-party vote share.¹⁷

Altogether, the regression results in Table 3 indicate that group memberships, while far from deterministic, are related to how candidates perform. These preliminary findings are hardly conclusive, however. Many well-know political organizations (e.g. NRA and other gun groups) appear unrelated to electoral performance, despite strong theoretical reasons to expect otherwise. To make sense of these disparate findings, we need to consider the pathway by which these memberships shape voting behavior. Moreover, we must explicitly test whether voters incorporate group memberships into their evaluations of officeholders.

Affiliations Experiment

The observational data suggest that, depending on the group and party, affiliations can benefit or hurt an officeholder at the ballot box. Yet, these analyses do not offer insight into the mechanism, rather, just associations between the presence or absence of particular associations and partisanship and electoral outcomes. These analyses also do not rule out alternative explanations arising from associations between other candidate characteristics and particular civic or issue advocacy groups. Indeed, we should expect such a problem if candidates purposely join groups and strategically highlight some memberships more than others to voters. To establish the causal effect of these affiliations, we fielded a survey experiment to measure how subjects evaluate candidates on the basis of group membership. In this experiment, we ask voters to respond directly to candidates knowing their partisanship and a group affiliation. While this may not perfectly translate to the information environment in a real-world campaign, it allows us to test the mechanism in its most simple form – can knowledge of group membership help inform choice even in the presence of partisanship, an incredibly strong cue?

Design

The survey experiment was fielded as part of the 2018 Cooperative Congressional Election Study (CCES). The CCES is an online survey conducted by more than fifty universities, and consists of both Common Content and individual team modules. The survey experiment was included on a single team module, with a sample size of 1588.¹⁸ After reading a short statement, subjects were asked to evaluate a hypothetical officeholder on the basis of their party and a group affiliation. Subjects evaluated the candidate from strongly oppose (−2) to strongly support (2). Figure 4 provides an example of the experimental interface.

Subjects were randomly assigned to one of eight conditions (See Table 4). For a set of baseline conditions, we wanted to select non-partisan groups that met two conditions. First, candidate affiliations with the groups should be common among both parties. Second, the groups should register as either neutral or positive to respondents. Drawing on our observational data of group affiliations, we identified “civic organization” and the Parent Teacher Association (PTA) as meeting these criteria. The very generic term “civic organization” carries little weight, but it is common in the data and may confer a degree of organizational support to such a candidate. Similarly, membership in the PTA indicates a commitment to the community and carries a generally positive connotation to most voters.

Table 4.

Experimental conditions.

Condition	Party and group affiliation
Non-partisan groups
D-Civic	Democrat; a civic organization
R-Civic	Republican; a civic organization
D-PTA	Democrat; the parent teacher association (PTA)
R-PTA	Republican; the parent teacher association (PTA)
Partisan groups
D-AFL	Democrat; the AFL-CIO
R-CoC	Republican; the U.S. Chamber of commerce
D-Sierra	Democrat; the sierra club
R-NRA	Democrat; the national rifle association (NRA)

We then selected four additional groups, two for each party that are active in national politics. First, these groups were conventionally seen as being aligned with either Democratic or Republican candidates. Indeed, Figure 3 supports this classification as all four groups are strong predictors of partisanship. Second, in contrast to the baseline groups, these organizations are explicitly engaged in lobbying and other political activities. The first set, the AFL-CIO and Chamber of Commerce, are major economic groups with longstanding ties to the respective party coalitions. In addition to their deep connections with the parties, the AFL-CIO and Chamber have a long history of advocating for policy and reputations of taking liberal or conservative positions, generally speaking. That said, as peak organizations they are also more heterogeneous in terms of both membership and the scope of issues that concern them. The second set of groups, the Sierra Club and National Rifle Association (NRA), are similarly aligned with one party of the other. Compared to organized labor and big business, the Sierra Club and NRA are more focused on the environment and gun rights, respectively. As such, these groups may signal a clearer message to voters in terms of the candidate’s ideology and legislative priorities.

Experimental Results

Given the prominent role party plays in candidate evaluation, the results are divided by both candidate party and respondent party identification. Figure 5 presents the experimental means for the Democratic candidate conditions, while Figure 6 displays Republican candidate conditions. The figures are centered at 0, indicating neither opposition (negative values) nor support (positive values). For the precise estimates and statistical significance, Table 5 presents the experimental means by condition and subject party.

Table 5.

Experimental means — by condition and subgroup.

	Subject Party
Condition	Democrat	Independent	Republican
D-Civic	0.680	−0.037	−0.547
D-PTA	0.844	0.077	−0.208**
D-AFL	0.683	−0.419	−0.953**
D-Sierra	0.698	0.000	−0.646
R-Civic	−0.434	0.000	0.171
R-PTA	−0.076**	0.143	0.623**
R-CoC	−0.600	0.135	0.521**
R-NRA	−1.277***	0.593	1.065***

Compared to the D-Civic or R-Civic, condition significant at *p $<$ 0.1; **p $<$ 0.05; ***p $<$ 0.01. P-values corrected for multiple testing within partisan groupings (Benjamini and Hochberg, 1995).

Looking at the Democratic conditions, we find large differences across the subjects’ party identification, demonstrating that the party cue does much of the work. That said, when looking closer, we find that group affiliations convey meaningful information within partisan subgroups as well. First, the PTA seems universally positive and a “safe” way to raise evaluations among both Democrats and Republicans. While this effect is statistically insignificant among Democrats and Independents, the PTA candidate is nonetheless at the top of each panel. Moreover, for Republican subjects, the difference between a PTA affiliation and a generic civic attachment, as well as the more partisan Democratic groups, is statistically significant. This result suggests that membership in the parent group can mitigate hostility of opposing partisans. Second, for Democratic subjects, organized labor and environmental ties produced no boost over that of a generic civic association, let alone the PTA affiliation. This null result is surprising given the ideological information signaled by these partisan group affiliations. What is more, in the regression results in Table 3, the union category appeared to be positively associated with Democratic election returns. As a fairly subtle experimental manipulation, the additional group connection may not offer much above and beyond the candidate’s Democratic label.

On the other hand, we find that Republican subjects significantly punish candidates affiliated with an AFL-CIO connection. Indeed, union candidates are rated lower than those with positive associations (PTA) and other advocacy groups (Sierra Club) alike. Altogether, we find that a hypothetical Democratic candidate sees little incentive to highlight membership in a partisan group — such connections do little to improve their standing with co-partisans or independents. Furthermore, to the extent that a candidate seeks to appeal to opposing partisans, these results suggest that ties to left-leaning groups, such as unions, are likely to backfire among Republican constituents.

Turning to the Republican conditions in Figure 6, we again find that group affiliations often convey very different messages depending on the subjects’ party. First, the PTA label is, once again, the most positive affiliation among opposing partisans. Republican candidates affiliated with the PTA were significantly more appealing to Democrats. It seems that the attachment to a generally positive group like the PTA can soften the image of an out-party candidate. Moreover, Republican subjects were more favorable of GOP candidates with a PTA affiliation than the generic civic association, suggesting a net positive across the board. Second, Democratic subjects viewed the GOP candidates with partisan group affiliations, especially the NRA, far more negatively. This pattern is very similar to the reactions that Republican subjects had to the Democratic candidate with ties to the AFL-CIO. Third, candidates with an NRA affiliation received the most positive evaluation among Republican subjects.¹⁹ Indeed, the NRA affiliation appears to be an extremely potent signal as to a candidate’s ideology, and conservative subjects responded accordingly. While Chamber of Commerce connections produce a boost over the civic association baseline, this improvement is vastly overshadowed by the NRA label. Ultimately, the Republican results again suggest that candidates pay a cost among opposing partisans when aligning with well-known partisan advocacy groups. That said, Republican candidates also benefited tremendously from these attachments with their base.

Conclusion

Candidate evaluation and the effectiveness of group-based heuristics have received considerable attention by scholars (Brady & Sniderman, 1985; Broockman et al., 2021; Lupia & McCubbins, 1998). Far less scholarly attention addresses the implications of a candidate’s membership in a group and its association with partisan electoral outcomes.

Before establishing whether groups hold sway over candidates and voters, we first sought to descriptively map out the linkages between parties and groups. We find expected variation in group membership across parties, owing to the personal backgrounds and policy priorities often associated with each party. Candidates associated with organized labor and environmental groups tend to be Democrats. Conversely, candidates who belong to Christian organizations, gun groups, and business associations are more likely to be Republicans. Interestingly, while strong predictors of party, the differential share of candidates belonging to these groups were not as stark as one would expect. Whether this is due to strategic candidate selection, including picking candidates that “trespass” on opposite party stereotypes, or is merely reflective of broad non-ideological membership in many groups remains a question for future research. These descriptive results suggest that candidates’ group membership and affiliation do vary across partisan lines, and that, in some cases, these group affiliations can affect evaluations of the candidates. Because of the strength of partisanship in shaping vote choice, the additional effect of group membership may be small alongside party cues. However, in some circumstances, particularly when group membership does not match party stereotypes, we may expect some memberships to cross-pressure partisans (Hillygus & Shields, 2008).

We then turned to the electoral implications of these group ties. When controlling for various measures of candidate ideology and quality, we find few group affiliations have significant associations with electoral outcomes. These preliminary findings suggest groups wield little direct impact on outcomes. But such a conclusion would be premature for two reasons. First, civic group associations are likely related to other measures of candidate quality. That is, candidates with party-stereotypical and more numerous associations are likely to be more interconnected in their community and be better fundraisers and vote-getters. Second, the finding of minimal effects may simply be due to voter information. Are constituents even aware that candidates belong to these groups? While these biographies are publicly available, there is no guarantee that voters learn about group memberships. More importantly, as Broockman et al. (2021) argue, there is good reason to expect that voters lack the requisite knowledge about groups to effectively use this information as a heuristic when voting. As with any heuristic, informational limitations and the presence of other strong information can undermine their use, especially when studied using observational data. Moreover, those same engaged voters that are likely to receive a group’s signal also hold strong priors, and can be less malleable (Zaller, 1992).

To better understand the mechanism in voters’ minds, we fielded a survey experiment that manipulated group membership alongside partisanship. We find that some affiliations, when directly presented to subjects, indeed affect how voters think about a candidate. This is particularly effective when a group, like the NRA, sends a strong signal to partisan voters. In short, both the group’s ideological reputation and attachment to a candidate must be known to register an impact on election outcomes. The fact that gun groups did not register a statistically significant effect in the observational analysis but did in the experimental context further supports this notion. That said, we also found that educational memberships (K-12) in the observational data were associated with a decrease in support among Democratic candidates — this result is starkly at odds with the net positive reaction that PTA candidates received by experimental subjects. While impossible to reconcile these competing results, this finding may stem from unobserved characteristics about the quality of candidates with K-12 ties. Moreover, in scrutinizing the experimental results closer, it seems that the popularity of PTA candidates is strongest among less engaged voters.²⁰ In other words, respondents who report the highest level of political interest, those most likely to turn out, tended to be less swayed by these non-partisan group memberships.

This study has important implications for not just the literatures at hand, but many bigger debates. Studies into the relationship between interest groups and officeholders should begin prior to the candidate’s electoral victory. Here we examine this relationship among general election candidates. It would also be a fruitful avenue of research in future work to examine how group membership varies between successful and unsuccessful primary candidates — dynamics the present study could not evaluate due to data limitations. Beyond merely getting one’s “foot in the door,” the group affiliations of candidates may leave a lasting imprint upon the candidate’s time in office for both more broad civic organizations and issue advocacy groups alike. Civic and interest group ties demonstrate a high degree of inter-connectedness between a party’s office-seekers and the coalition’s “policy demanders” (Cohen et al., 2008). To the extent that groups are likely to shape officeholder behavior, we should expect that explicit membership in a group will exert tremendous influence over their legislative priorities.

Moving forward, scholars should continue to examine the long-term legislative consequences of these affiliations. The candidate-group data are relevant to fundamental questions on the role interest groups play in agenda-setting and the legislative process (Baumgartner & Leech, 2001; Gilens & Page, 2014). Are candidates culled from the ranks of civic and interest groups more likely to introduce bills favored by those groups? Furthermore, how do such candidates compare against their colleagues in other context? For instance, are candidates affiliated with interest groups more effective legislators, do they have more dense networks to rely upon (Tam Cho & Fowler, 2010; Volden & Wiseman, 2014)?

Supplemental Material

Supplemental Material - Card-Carrying Members on the Ballot: Group Affiliations and Their Electoral Implications

Supplemental Material for Card-Carrying Members on the Ballot: Group Affiliations and Their Electoral Implications by Stephen Goggin, and Travis Johnston in American Politics Research

Footnotes

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.

ORCID iDs

Stephen N. Goggin

Travis M. Johnston

Supplemental Material

Supplemental material for this article is available online.

Notes

References

Ahler

D. J.

(2014). Self-fulfilling misperceptions of public polarization. The Journal of Politics, 76(3), 607–620. https://doi.org/10.1017/s0022381614000085

Ahler

D. J.

Sood

(2018). The parties in our heads: Misperceptions about party composition and their consequences. The Journal of Politics, 80(3), 964–981. https://doi.org/10.1086/697253

Arceneaux

Kolodny

(2009). Educating the least informed: Group endorsements in a grassroots campaign. American Journal of Political Science, 53(4), 755–770. https://doi.org/10.1111/j.1540-5907.2009.00399.x

Barber

Pope

(2022). Groups, behaviors, and issues as cues of partisan attachments in the public. American Politics Research, Article 21083831.

Baumgartner

F. R.

Leech

B. L.

(2001). Basic interests: The importance of groups in politics and in political science. Princeton University Press.

Benjamini

Hochberg

(1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society: Series B, 57(1), 289–300. https://doi.org/10.1111/j.2517-6161.1995.tb02031.x

Bonica

(2013). Database on ideology, Money in politics, and elections: Public version 1.0 [dataset]: Stanford University Libraries.

Brady

H. E.

Sniderman

P. M.

(1985). Attitude attribution: A group basis for political reasoning. American Political Science Review, 79(4), 1061–1078. https://doi.org/10.2307/1956248

Broockman

D. E.

Kaufman

A. R.

Lenz

G. S.

(2021). Heuristic projection: How interest group cues can harm voters’ judgements. Technical Report. Working Paper. https://osf.io/6yskq

10.

Campbell

D. E.

Green

J. C.

Layman

G. C.

(2011). The party faithful: Partisan images, candidate religion, and the electoral impact of party identification. American Journal of Political Science, 55(1), 42–58. https://doi.org/10.1111/j.1540-5907.2010.00474.x

11.

Cohen

Karol

Noel

Zaller

(2008). The party decides: Presidential nominations before and after reform. University of Chicago Press.

12.

Converse

P. E.

(2006). The nature of belief systems in mass publics (1964). Critical Review, 18(1–3), 1–74. https://doi.org/10.1080/08913810608443650

13.

Elder

E. M.

O’Brian

N. A.

(2022). Social groups as the source of political belief systems: Fresh evidence on an old theory. American Political Science Review, 116(4), 1407–1424. https://doi.org/10.1017/s0003055422000326

14.

Fenno

R. F.

(1978). Home style: House members in their districts: Little.

15.

Gilens

Page

B. I.

(2014). Testing theories of american politics: Elites, interest groups, and average citizens. Perspectives on Politics, 12(3), 564–581. https://doi.org/10.1017/s1537592714001595

16.

Gimpel

J. G.

(1998). Packing heat at the polls: Gun ownership, interest group endorsements, and voting behavior in gubernatorial elections. Social Science Quarterly, 634–648.

17.

Goggin

S. N.

Henderson

J. A.

Theodoridis

A. G.

(2020). What goes with red and blue? Mapping partisan and ideological associations in the minds of voters. Political Behavior, 42(4), 985–1013. https://doi.org/10.1007/s11109-018-09525-6

18.

Green

Palmquist

Schickler

(2002). Partisan hearts and minds: Political parties and the social identities of voters. Yale University Press.

19.

Grossmann

Hopkins

D. A.

(2015). Ideological republicans and group interest democrats: The asymmetry of american party politics. Perspectives on Politics, 13(1), 119–139. https://doi.org/10.1017/s1537592714003168

20.

Grossmann

Hopkins

D. A.

(2016). Asymmetric politics: Ideological republicans and group interest democrats. Oxford University Press.

21.

Hillygus

D. S.

Shields

T. G.

(2008). The persuadable voter: Wedge issues in presidential campaigns. Princeton University Press.

22.

Iyengar

Lelkes

Levendusky

Malhotra

Westwood

S. J.

(2019). The origins and consequences of affective polarization in the United States. Annual Review of Political Science, 22(1), 129–146. https://doi.org/10.1146/annurev-polisci-051117-073034

23.

Iyengar

Sood

Lelkes

(2012). Affect, not ideology: A social identity perspective on polarization. Public Opinion Quarterly, 76(3), 405–431. https://doi.org/10.1093/poq/nfs038

24.

Kane

J. V.

Mason

Wronski

(2021). Who’s at the party? Group sentiments, knowledge, and partisan identity. The Journal of Politics, 83(4), 1783–1799. https://doi.org/10.1086/715072

25.

Karol

(2009). Party position change in American politics: Coalition management. Cambridge University Press.

26.

Lau

R. R.

Redlawsk

D. P.

(2001). Advantages and disadvantages of cognitive heuristics in political decision making. American Journal of Political Science, 45(4), 951–971. https://doi.org/10.2307/2669334

27.

Lawless

J. L.

Fox

R. L.

(2005). It takes a candidate: Why women don’t run for office. Cambridge University Press.

28.

Lupia

(1994). Shortcuts versus encyclopedias: Information and voting behavior in California insurance reform elections. American Political Science Review, 88(1), 63–76. https://doi.org/10.2307/2944882

29.

Lupia

McCubbins

M. D.

(1998). The democratic dilemma: Can citizens learn what they need to know? Cambridge University Press.

30.

Mason

(2018). Uncivil agreement: How politics became our identity. University of Chicago Press.

31.

Mason

Wronski

(2018). One tribe to bind them all: How our social group attachments strengthen partisanship. Political Psychology, 39(S1), 257–277. https://doi.org/10.1111/pops.12485

32.

Matter

(2014). pvsR: An R package to interact with the Project Vote Smart API for scientific research. R package version 0.3.

33.

McDermott

M. L.

(1998). Race and gender cues in low-information elections. Political Research Quarterly, 51(4), 895–918. https://doi.org/10.2307/449110

34.

McDermott

M. L.

(2006). Not for members only: Group endorsements as electoral information cues. Political Research Quarterly, 59(2), 249–257. https://doi.org/10.1177/106591290605900207

35.

McDermott

M. L.

(2007). Voting for Catholic candidates: The evolution of a stereotype. Social Science Quarterly, 88(4), 953–969. https://doi.org/10.1111/j.1540-6237.2007.00512.x

36.

McKelvey

R. D.

Ordeshook

P. C.

(1985). Elections with limited information: A fulfilled expectations model using contemporaneous poll and endorsement data as information sources. Journal of Economic Theory, 36(1), 55–85. https://doi.org/10.1016/0022-0531(85)90079-1

37.

Miratrix

L. W.

Sekhon

J. S.

Theodoridis

A. G.

Campos

L. F.

(2018). Worth weighting? How to think about and use weights in survey experiments. Political Analysis, 26(3), 275–291. https://doi.org/10.1017/pan.2018.1

38.

Rosenstone

S. J.

Hansen

J. M.

(1993). Mobilization, participation, and democracy in America. Macmillan.

39.

Tam Cho

W. K.

Fowler

J. H.

(2010). Legislative success in a small world: Social network analysis and the dynamics of congressional legislation. The Journal of Politics, 72(1), 124–135. https://doi.org/10.1017/s002238160999051x

40.

Volden

Wiseman

A. E.

(2014). Legislative effectiveness in the United States congress: The lawmakers. Cambridge University Press.

41.

VoteSmart (2015). Vote smart. Accessed on January 26, 2015.

42.

Zaller

(1992). The nature and origins of mass opinion. Cambridge University Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.14 MB