Outside of the Old Boys Club? Gender Differences in Outside Groups’ Advertising Support for U.S. Senate Candidates

Abstract

Since 2010, televised campaign ads have proliferated, raising questions about gender differences in the number of ads outside groups air. Assuming incumbent Democratic women and Democratic women with prior office-holding experience run with support from their party and interest groups, we expect Democratic-leaning groups air more favorable ads supporting Democrats in races that include experienced or incumbent Democratic women than they do in races that include experienced or incumbent Democratic men. Conversely, we assume Republican women receive less support from their party and interest groups. We expect outside groups air more favorable ads supporting Republicans in races featuring experienced or incumbent Republican men than races including experienced or incumbent Republican women. Given the potential for backlash to attack ads, we expect Democratic- and Republican-leaning groups air more ads in races including experienced or incumbent male opponents than in races that include experienced or incumbent female opponents. We test these hypotheses by focusing on 2010–2018 U.S. Senate races and combining original data with data from the Wesleyan Media Project, the U.S. Census, and the Cook Political Report. Democratic- and Republican-leaning groups both sponsor significantly fewer ads in races featuring experienced female Democratic candidates compared to races including experienced male Democratic candidates.

Keywords

super PACs 527s 501cs/dark money PACs female Senate candidates campaign advertising

Since EMILY’s List was established in 1985, it has helped Democratic women win elections (Burrell, 2018; Carroll and Sanbonmatsu, 2013; Crespin and Deitz, 2010; Day and Hadley, 2005; Dittmar, 2015; Kitchens and Swers, 2016; Thomsen and Swers, 2017). Today, EMILY’s List (2023) claims their SuperPAC created 49 unique television ads during the 2020 elections, helping them mobilize “millions of voters in 48 states.” More broadly outside interest groups aired 1.5 million ads during the 2020 election (Wesleyan Media Project [WMP] 2021). These data reveal outside groups’ television ads are critical components of American elections raising questions about how outside groups’ decisions about how many television ads to air help or hurt female candidates. We ask: do Democratic- and Republican-leaning outside groups air more ads on behalf of female or male candidates during U.S. Senate races?

Outside groups include four types of campaign finance organizations American interest groups establish to influence elections: traditional political action committees (PACs), SuperPACs, 527s, and 501c “dark money” groups. All four types help groups elect their preferred candidates to office, and most have clear partisan or ideological leanings (Boatright, 2007; CRP, 2023a; Dwyre and Braz, 2014; Herrnson, 2017; Sides et al., 2019). Though the Senate is becoming less male-dominated (Center for American Women and Politics [CAWP] 2023a; 2023b), women continue to face unique opportunities and challenges when they run for the Senate. Incumbent Democratic women and Democratic women with prior office-holding experience often run with the enthusiastic support of their party establishment and allied interest groups. Meanwhile, incumbent Republican women and Republican women with prior experience in elected office run with less support from party elites and allied interest groups. Bringing these insights together, we expect Democratic-leaning groups should run more favorable ads when their preferred Democratic candidate is a female incumbent Senator rather than a male incumbent Senator, or an experienced female challenger (i.e., one who has held prior elected office, but is not a sitting Senator) rather than an experienced male challenger. Conversely, Republican-leaning groups should run more favorable ads when their preferred Republican candidate is a male incumbent Senator rather than a female incumbent Senator, or an experienced male challenger rather than an experienced female challenger. We also expect both Democratic- and Republican-leaning groups should run more attack ads when the opposition party candidate is a male incumbent rather than a female incumbent or an experienced male challenger rather than a female experienced challenger. Finally, we expect there will be few gender differences in the number of ads that both Democratic- and Republican-leaning groups air in races that include inexperienced challengers who do not have prior office-holding experience.

We test these hypotheses using WMP data on 135,115 favorable and attack ads that 304 outside groups aired on television during Senate general elections from 2010 to 2018; an original dataset on U.S. Senate candidates’ gender, incumbency status, prior office-holding experience, and partisanship; state-level Census data; and the 2010–2018 Cook Political Reports. We find Republican - and Democratic-leaning groups air more favorable ads for their party’s preferred candidate in races that include experienced male Democratic challengers as one of the two contenders compared to races that include experienced female Democratic challengers as one of the two contenders. Republican-leaning outside groups also air significantly more attack ads in races that include experienced male Democratic challengers than races that include experienced female Democratic challengers. Democratic-leaning outside groups run significantly more favorable ads in races that include inexperienced female Democratic challengers compared to races that include inexperienced male Democratic challengers. At all levels of experience, there are no differences in the number of favorable or attack ads that Democratic- or Republican-leaning groups air based on whether the Republican candidate is a man or a woman.

Outside Groups and Senate Campaign Ads

Outside groups change election outcomes by supporting their preferred candidates’ campaigns. The term “outside group” distinguishes campaign finance organizations associated with interest groups from campaign finance organizations associated with the political parties or the candidates themselves. Typically, outside groups establish one (or more) of four types of campaign finance organizations: a traditional PAC, a SuperPAC, a 527, or a 501c. Traditional PACs are “groups that work toward the election of candidates, to pass or defeat legislation, or to advance a political agenda” (Sides et al., 2019). Traditional PACs primarily contribute funds to candidates they endorse. They face the largest number of campaign finance restrictions. Donations to traditional PACs are legally capped; in the 2023–2024 cycle, the maximum donation an individual can make to a PAC is $3,300 (Federal Election Commission [FEC], 2023). Traditional PACs are also required to publicly identify donors who have contributed at least $200 (Sides et al., 2019).

SuperPACs make “independent expenditures” which primarily fund communications, such as television ads, that explicitly advocate “for the election or defeat of a clearly identified candidate” and “that are not made in cooperation with, consultation or concert with, or at the request or suggestion of any candidate” (FEC, 2019). SuperPACs are allowed to “solicit and accept unlimited contributions from individuals, corporations, labor organizations, and other political committees,” which means they can raise and spend far more money airing ads than traditional PACs (FEC, 2019).

527s primarily support candidates by using independent expenditures to fund political communications. They are “organized and operated primarily for the purpose of directly or indirectly accepting contributions or making expenditures or both” for voter mobilization campaigns and issue advocacy (FEC, 2019). Unlike SuperPACs, 527s are not allowed to explicitly advocate for or against specific candidates, but their ads can focus on candidates’ issue positions (Meier, 1999). Consequently, 527s were largely replaced by SuperPACs.

501c “dark money” groups are three types of tax-exempt non-profits: 501c4s, 501c5s, and 501c6s. 501c4s are “social welfare organizations” which “must not be organized for profit and must be operated exclusively to promote social welfare” (IRS, 2019c). 501c5s promote “the betterment of conditions of those engaged in the pursuits of labor, agriculture, or horticulture” (IRS, 2019b; Sides et al., 2019). 501c6s are “business leagues, chambers of commerce, real estate boards, boards of trade, and professional football leagues” which do not provide earnings to shareholders or individuals (IRS, 2019a). 501cs are not allowed to endorse candidates or tell voters which candidates to support and participating in politics cannot be their “primary” purpose (Meier, 1999; Potter and Morgan, 2013). 501c groups are appealing to political donors because three important legal loopholes allow them to air campaign ads. First, the IRS weakly enforces primary purpose requirements, allowing 501cs to air large numbers of ads (Potter and Morgan, 2013; Sides et al., 2019). Second, 501cs are allowed to air issue advocacy ads that support or criticize a candidate’s stance on a given issue relevant to the 501c’s mission as long as they do not use “magic words” that tell voters to specifically vote for or against a candidate (Meier, 1999). Third, 501cs do not have to publicly disclose their donors (Sides et al., 2019), so they allow donors to fund air campaign ads with little public scrutiny.

Given the different opportunities and challenges associated with each type of organization, outside groups often set up multiple campaign finance arms that operate simultaneously. For example, Planned Parenthood (PP) has a traditional PAC, a SuperPAC, and a 501c4 (Center for Responsive Politics [CRP] 2023b). Its traditional PAC endorses candidates and directly contributes to them. However, those campaign contributions are capped and publicly disclosed, so other arms help circumvent those limitations. PP’s SuperPAC spends unlimited funds on ads that explicitly tell voters who to vote for. For donors who do not want their contributions publicly disclosed, PP’s 501c dark money arm spends unlimited funds on issue ads, giving donors an anonymous way to shape elections.

Outside groups air approximately one-third of the ads that appear during competitive Senate races, running more ads than the political parties in recent years (Fowler et al., 2016). The large number of outside group ads aired during Senate races is important for four reasons. First, SuperPACs, 527s, and 501cs cannot legally contribute directly to candidates, and SuperPACs and 527s use independent expenditures to fund political communications. Consequently, airing ads is one of the primary ways those groups influence Senate elections.

Second, though women’s media coverage is becoming more balanced (Banwart et al., 2003; Bystrom et al., 2004), gendered challenges remain. Coverage of female candidates continues to draw disproportionate attention to women’s appearances (Andrich et al., 2023; Bystrom et al., 2004; Carlin and Winfrey, 2023; van der Pas and Aaldering, 2020), their viability or “electability” (Gibbons, 2022; van der Pas and Aaldering, 2020), likability or warmth (Carlin and Winfrey, 2023; Cassese et al., 2021; Godble et al., 2019), the “novelty” of women in office (Bauer, 2020; Bauer and Taylor, 2023; Gibbons, 2022; O’Regan and Stambough, 2011), and/or women’s personal lives (Banwart et al., 2003; Bauer, 2020; Bauer and Taylor, 2023; Gibbons, 2022; van der Pas and Aaldering, 2020). Some trait coverage continues to come at the expense of issues coverage (Dunaway et al., 2013). Women of color receive less coverage overall and more frequent negative coverage than their competitors (Gershon, 2012; Ward, 2016). Therefore, we expect women’s outside groups, particularly Democratic ones, should air large numbers of favorable ads to help female candidates counteract the gendered stereotypes and frames that are prevalent in the media.

Third, because Senate campaigns must turnout voters across entire states, television ads are one of the most important ways voters learn about the candidates and participate in Senate elections (Fowler et al., 2016; Jacobson and Carson, 2020). The more ads groups air (and the more different times, programs, and channels they are on), the more likely outside groups will be able to precisely micro-target specific groups of voters and get their message out to as many voters as possible (Fowler et al., 2016). Advertising expenditures alone cannot account for the number of chances voters have to be exposed to each group’s ads on television.

Fourth, the large number of outside group ads has changed the tone of Senate elections, making them into “no holds barred contests” that feature personal attacks and/or inflammatory language, sometimes without even naming the opposition (Dowling and Miller, 2014; Dowling and Wichowsky, 2015; Herrnson, 2017; Kahn and Kenney, 2004, xi–xii). Thus, by focusing on the number of ads aired, we can analyze the factors associated with outside groups’ decisions about how many favorable and attack ads to air.

Hypotheses

The majority of outside groups focus on winning elections for their partisan or ideological allies, making them well-positioned to influence races in today’s hyper-partisan Senate (Abramowitz, 2012; Boatright, 2007; CRP, 2023a; Dwyre and Braz, 2014; Herrnson, 2017; Lee, 2012; Sides et al., 2019; Sinclair, 2012). Wealthy mega-donors who are narrowly focused on achieving their partisan/ideological goals fund many outside groups (especially SuperPACs; Boatright, 2007; Dwyre and Braz, 2014; Garrett, 2013; Herrnson, 2017). Consequently, we assume outside groups often function as informal extensions of the parties (or ideological factions within the parties) and their decisions about the number of ads to air during Senate elections will be closely related to their desire to help candidates from one party (or one ideological wing of one party) win Senate seats. In addition to considering candidates’ partisanship/ideology, the decisions outside groups make about how many ads to air should also be influenced by the candidates’ gender, incumbency status, and prior office-holding experience.

First, we focus on advertising campaigns focused on female Democratic incumbents (i.e., sitting U.S. Senators) and female experienced challengers (i.e., women who have held prior elected office, but who were not U.S. Senators when they ran for the Senate). Though early research indicated partisan gatekeepers were biased against some female candidates or recruited from masculine “old boys” networks (Niven, 2006; Sanbonmatsu, 2006), gender-biased recruitment patterns are changing for Democrats. Female Democrats greatly benefit from their connections to outside groups and to women’s PACs, such as EMILY’s List and Planned Parenthood (Burrell, 2018; Carroll and Sanbonmatsu, 2013; Crespin and Deitz, 2010; Day and Hadley, 2005; Dittmar, 2015; Kitchens and Swers, 2016; Thomsen and Swers, 2017). Those groups have become powerful because the Democrats’ culture and structure is open, decentralized, and supportive of protecting or expanding rights for members of marginalized groups (Crowder-Meyer and Cooperman, 2018; Elder, 2021). The Democrats’ donor base also contains many “women’s representation policy demanders” (WRPDs) who are interested in increasing the number of women in office (Crowder-Meyer and Cooperman, 2018). WPRD’s create pressure for Democrats to elect women, and they are much better integrated into the Democratic Party than the Republican Party (Crowder-Meyer and Cooperman, 2018). Democratic women’s PACs are also well-known among Democratic donors, and they receive more support from Democratic donors than Republican women’s PACs receive from Republican donors (Crowder-Meyer and Cooperman, 2018). In fact, EMILY’s List is so powerful that it is now seen as an extension of the Democratic party, helping fill gaps during primary elections and races for safe seats (Elder, 2021).

Unlike Republican women, Democratic women also benefit from connections to female mega-donors, such as Karla Jurvetson, who contributed $5.4 billion to EMILY’s List/Women Vote between 2011 and 2018 (FEC, 2020; Herrnson and Heerwig, 2021). Moreover, during our 2010-2018 study period, 22 of the 31 women who served in the Senate were Democrats, and Democratic women secured leadership positions in the Senate, including Assistant Majority Leader and SCCC Chair (Patty Murray, D-WA) and the Appropriations Committee Chair (Barbara Mikulski, D-MD) (CAWP, 2023a; US Senate, 2023). When women secure leadership positions, they have been able to put more pressure on their parties to recruit and support female candidates (Crowder-Meyer, 2013; Elder, 2021; Geras, 2021; Sanbonmatsu, 2006). With all of these structures to support them, it is no surprise Democratic female candidates raise more from individual donations than their male counterparts (Crespin and Deitz, 2010), female Democratic donors give more to female Democratic candidates (Thomsen and Swers, 2017), female Democrats receive more support during primaries their male counterparts do (Hassell and Visalvanich, 2019; Thomsen and Swers, 2017), and female Democratic nominees attract more donors and raise more money from them than male nominees do (Fraga and Hassell, 2021).

Because many outside groups exist to achieve partisan/ideological goals, outside groups should be more likely to support incumbents or experienced challengers whose legislative records align with their preferred party’s “brand.” Democrats should support candidates who align with the party’s “brand” as the women-friendly party and allow the party to take advantage of the gender gap in its favor (Sanbonmatsu, 2004; Swers, 2013). Thus, Democratic leaders seek out female Senators for press conferences and efforts to communicate their work on behalf of women’s issues and social welfare legislation to showcase the party’s reputation on those issues (Swers, 2013). Following their lead, Democratic-leaning outside groups should air more favorable ads for candidates in races that include female Democratic incumbents or experienced women challengers who have records focused on sponsoring feminist, women’s issue, and social welfare bills (Swers, 2002; 2013). Women and other donors who contribute to female Democratic candidates are also particularly fond of contributing to liberal Democratic women (Crespin and Deitz, 2010; Kitchens and Swers, 2016; Thomsen and Swers, 2017), creating incentives for outside groups to air ads highlighting Democratic women’s legislative records. Altogether, we expect that among Democrats, the party’s women-friendly brand, its increased number of women in the Senate, increased number of donations on behalf of female Democrats, and strong party support for women’s groups mean there should also be a great deal of pressure for Democratic-leaning groups to use their funds to air ads on behalf of Democratic female candidates. As partisan/ideological groups, and groups that often exist to make and air ads (or other political communications), we expect Democratic-leaning outside groups should also be eager to follow the party’s lead as it makes decisions about how many ads to air on behalf of female candidates. Thus, our first hypothesis states:

H ₁ Democratic Group Preference for Supporting Female Democratic Candidates : Democratic-leaning outside groups will air more favorable ads for female Democratic incumbents and female experienced challengers than for male Democratic incumbents or male experienced challengers.

Second, we focus on gender differences in the number of ads Republican-leaning outside groups air on behalf of Republican candidates, where the gendered context is quite different. Unlike on the Democratic side, WRPDs and Republican women’s groups, such as Maggie’s List, She PAC, Susan B. Anthony List, Republicans for Choice, and View PAC, are not well-funded or well-integrated into the party structure and there is little demand for increasing women’s descriptive representation (Carroll and Sanbonmatsu, 2013; Crowder-Meyer and Cooperman, 2018; Elder, 2021). Republican female candidates hesitate to get involved with Republican women’s groups because they believe party leaders sideline and ignore them for being out-of-step with the party’s brand (Elder, 2021). Since the 1970s and 1980s, Republican policy demanders have also focused on promoting individualism, merit, and traditional gender roles centered on women’s roles as wives and mothers, rather than identity politics (Deckman, 2016; Elder, 2021; Schreiber, 2008; Swers, 2013; Thomsen and Swers, 2017; Wineinger, 2022). Groups that exist to elect Republican women, including newer groups like Representative Elise Stefanik’s (R-NY) E-PAC, challenge both of those core party values, so they have received little support from party leaders (Crowder-Meyer and Cooperman, 2018; Elder, 2021; Kurtzleben, 2020).

Republican women’s groups also raise far less money than their Democratic counterparts, as many Republican donors are unaware that Republican women’s groups exist or are unlikely to donate to candidates based on their sex or gender (Crowder-Meyer and Cooperman, 2018; Dittmar, 2015; Thomsen and Swers, 2017). Over the course of 20 years, VIEW PAC (one of the largest Republican women’s groups) raised $6.5 million for candidates, while EMILY’s List raised $100 million in 2018 alone (Elder, 2021, 151). Many Republican women’s groups also go out of business after only a couple of election cycles (Dittmar, 2015; Elder, 2021). With little emphasis on supporting women’s campaigns, Republican women (including incumbents and experienced challengers) raise less money than their male counterparts (Kitchens and Swers, 2016; Thomsen and Swers, 2017). Moderate Republican women, who are often perceived as more liberal than the party as a whole, also struggle to raise funds from “a polarized, congressional donor pool” and to receive funds from the party itself (Crespin and Deitz, 2010, 590; Carroll and Sanbonmatsu, 2013). Even female Republicans prefer to donate to conservative candidates (Thomsen and Swers, 2017). These gendered disadvantages in financing Republican women’s campaigns suggest Republican-leaning groups will also struggle to raise money to support Republican women, and they will have less money available and less donor pressure to air ads on behalf of Republican women. Consequently, we expect Republican-leaning outside groups have few incentives to air ads that support Republican women running as women who legislate on behalf of women.

We also expect Republican-leaning outside groups will be more likely to support Republican male candidates than female candidates for two gatekeeping reasons. First, with only 9 Republican women serving in the Senate between 2010 and 2018, few Republican women serving in leadership positions, and poorly integrated women’s groups there is little pressure from Republican women in office to challenge gendered structures within the party, particularly given the party’s strong emphasis on hierarchy and loyalty (CAWP, 2023a; Elder, 2021; US Senate, 2023). Thus, Republican leaders and party elites may continue to support men who they know from their existing networks of largely male office-holders and/or business contacts (Crowder-Meyer, 2013; Sanbonmatsu, 2006). Second, many mega-donors to Republican-leaning outside groups are male and are primarily focusing on helping Republicans or party extremists win elections (Boatright, 2007; Dwyre and Braz, 2014; Herrnson, 2017; Sides et al., 2019). Conservative female mega-donors, such as Miriam Adelson (wife of Sheldon), also contribute to achieve partisan/ideological goals (Boatright, 2007; Bryner and Weber, 2013). Since those mega-donors play an important agenda-setting role for their groups, they are focused on winning elections for their partisan/ideological teams, and party elites hesitate to support female candidates in competitive elections (Crowder-Meyer, 2013; Niven, 1998; Sanbonmatsu, 2006), we expect Republican-leaning outside groups will be more eager to air ads that support Republican men rather than Republican women. Altogether, these insights lead us to our second hypothesis:

H ₂ Republican Group Preference for Supporting Male Republican Candidates : Republican-leaning outside groups will air more favorable ads for male Republican incumbents and experienced challengers than for female Republican incumbents and experienced challengers.

Third, we expect that Democratic- and Republican-leaning outside groups will carefully consider the gender of the opposition party’s candidate when they decide how many negative attack ads to air. Negative ads have traditionally played an important role in Senate campaigns because they influence the outcomes of competitive races, helping voters recall and evaluate candidates and turnout (Engstrom and Kenny, 2002; Kahn and Kenney, 2004). Thus, outside groups have incentives to air more ads attacking their opponents’ policy agendas, issue positions, personal traits, and leadership qualities as the race tightens (Kahn and Kenney, 1999). Outside groups ads have a particularly important role in competitive Senate elections because they provide candidates with cover by running the kinds of negative attack ads that voters may punish candidates for running (Brooks and Murov, 2012; Dowling and Wichowsky, 2015; Fowler et al., 2016; Herrnson, 2017).

Although airing attack ads is effective, we expect outside groups carefully consider the optics associated with airing ads that attack female candidates. Because the Democratic party includes groups and donors that provide explicit support for increasing women’s descriptive representation and the party’s brand is associated with being “women-friendly” (Crepsin and Deitz, 2010; Crowder-Meyer and Cooperman, 2018; Elder, 2021; Fraga and Hassell, 2021; Sanbonmatsu, 2004; Kitchens and Swers, 2016; Thomsen and Swers, 2017), we expect Democratic-leaning outside groups will be hesitant to air attack ads targeting women. Airing such attack ads could open them to charges that Democrats are actually “anti-woman” or unwilling to support all women. Republican-leaning outside groups that air attack ads against Democratic women risk retrenching arguments that the Republican Party is anti-woman (Elder, 2021; Sanbonmatsu, 2004). Republican groups that air attack ads on behalf of Republican women could also encounter tensions with Republican positions that focus on preserving women’s traditional roles as caregivers or moral protectors (Deckman, 2016; Elder, 2021; Schreiber, 2008; Wineinger, 2022). It could be difficult for Republican-leaning groups (and candidates they support) to claim they are caring or have the moral high ground if they air large numbers ads that attack other female candidates. Hence, our third hypothesis states:

H ₃ Preference for Attacks on Male Candidates : Both Democratic- and Republican-leaning outside groups will air more attack ads when their preferred party’s candidate is running against a male incumbent or a male experienced challenger than when their preferred party’s candidate is running against a female incumbent or a female experienced challenger.

Finally, we expect outside groups will consider whether candidates have prior office-holding experience when they decide how many television ads to air. Because candidates with prior office-holding experience are more likely to win than those who have never held elected office (Carson et al., 2007; Cox and Katz, 1996; Jacobson and Carson, 2020; King and Gelman, 1991), we expect outside groups will air fewer ads in races that include an inexperienced challenger as one of the two major party contenders. Therefore, our final hypothesis states:

H ₄ Inexperienced Challengers : The numbers of favorable and attack ads that outside groups air will not be associated with the gender of inexperienced challengers in the race.

Data and Methods

We utilize three types of data to examine gender differences in the number of television ads that outside groups air. First, we use WMP data on television ads purchased in all 210 American media markets during 2010–2018 Senate elections. Second, we examine data on the gender, partisanship, incumbency status, and prior office-holding experience of each candidate in the 2010–2018 general election races for the U.S. Senate.¹ Third, we control for state-level demographic information with U.S. Census data and use Cook Political Report data to control for the partisan lean and competitiveness of each race.

The WMP data identify the sponsor of each ad, indicating whether the sponsor was the candidate, one of the political parties, or an interest group, and we used those codes to limit the analysis to interest group ads. WMP’s “interest group” category does not distinguish between the different types of outside groups, so it contains ads aired by traditional PACs, SuperPACs, 527s, and 501c dark money groups. Second, the WMP data identify whether each ad favored a candidate that an outside group preferred or attacked a candidate an outside group opposed. These data allow us to identify the number of ads outside groups ran supporting Democratic and Republican candidates and categorize the groups based on whether they lean Democratic or Republican. Third, the WMP data contain information about the type of ad aired (i.e., favorable or attack). Thus, we examine the following scenarios: (1) the number of Republican-leaning outside group ads aired (favorable and attack) based on the gender, incumbency status, and prior experience of the group’s Democratic opposition candidate; (2) the number of Democratic-leaning outside group ads aired (favorable and attack) based on the gender, incumbency status, and prior experience of the group’s preferred Democratic candidate; (3) the number of Republican-leaning outside group ads aired (favorable and attack) based on the gender, incumbency status, and prior experience of the group’s preferred Republican candidate; and (4) the number of Democratic-leaning outside group ads aired (favorable and attack) based on the gender, incumbency status, and prior experience of the group’s Republican opposition candidate.

The dependent variables in our analyses are the number of favorable and attack ads that Democratic- and Republican-leaning groups aired based on the gender, partisanship, and prior office-holding experience of the candidates in each race.² Favorable ads support an outside group’s preferred candidate, and attack ads express opposition to an outside group’s opposition candidate. The number of favorable ads Republican-leaning groups aired per race ranges from 0 to 21,405 (mean = 2593.94), and the number of attack ads Republican-leaning groups ran per race ranges from 0 to 42,209 (mean = 5671.96). The number of favorable ads Democratic-leaning groups ran per race ranges from 0 to 26,162 (mean = 3675.65), and the number of attack ads Democratic-leaning groups ran per race ranges from 0 to 45,339 (mean = 5952.11). These data indicate that Democratic groups, on average, aired 1,081.71 more favorable and 280.15 more attack ads than Republican leaning groups did.³

To explicitly account for partisan differences in the number of ads outside groups air, the dependent variables in Table 1 are (1) the number of favorable ads and (2) the number of attack ads that outside groups aired based on the Democratic candidate in the general election race . Thus, the dependent variables in the first two models in Table 1 reflect the number of favorable and attack ads Republican-leaning outside groups ran based on the gender, incumbency status, and prior experience of the group’s Democratic opposition candidate . The dependent variables in the third and fourth models in Table 1 reflect the number of favorable and attack ads Democratic-leaning outside groups aired based on the gender, incumbency status, and prior experience of the group’s preferred Democratic candidate .

Table 1.

Outside Group Ads by Gender, Incumbency Status, and Prior Office-Holding Experience of the Democratic Candidate.

	Rep Leaning Group Ads: Dem Candidate				Dem Leaning Group Ads: Dem Candidate
	# Favor		# Attack		# Favor		# Attack
Key variables
Female inexp	1.199*	(0.512)	−0.473	(0.762)	1.705**	(0.641)	1.275	(0.915)
Male exp	0.396	(0.358)	1.774**	(0.571)	1.261*	(0.627)	0.627	(0.756)
Female exp	−0.659 ^†	(0.380)	−0.018	(0.568)	0.744	(0.660)	0.089	(0.734)
Male inc	−0.702*	(0.339)	1.116*	(0.527)	2.012**	(0.647)	1.723*	(0.689)
Female inc	−0.769 ^†	(0.445)	0.925 ^†	(0.562)	2.236***	(0.617)	1.592 ^†	(0.825)
Controls
Ad cost	−0.001 ^†	(0.001)	0.000	(0.000)	0.000	(0.000)	0.000	(0.000)
Inc leader	0.517	(0.735)	0.811**	(0.313)	0.534	(0.734)	0.849 ^† ***	(0.450)
% urban	−0.001	(0.012)	−0.018	(0.014)	0.013	(0.010)	−0.014	(0.011)
Solid race	−1.805***	(0.332)	−3.190***	(0.350)	−1.533***	(0.350)	−3.493***	(0.411)
Toss-up race	0.639**	(0.203)	1.674***	(0.280)	0.194	(0.256)	1.310***	(0.320)
2010	0.445	(0.396)	−0.738 ^†	(0.441)	−0.824**	(0.322)	−0.722	(0.517)
2012	−0.124	(0.337)	−0.143	(0.356)	−0.741**	(0.262)	−0.378	(0.415)
2014	0.220	(0.301)	0.067	(0.506)	−0.446	(0.341)	−0.451	(0.595)
2016	0.644 ^†	(0.355)	0.502	(0.511)	−0.274	(0.268)	−0.592	(0.650)
Constant	8.801***	(0.895)	8.499***	(1.154)	6.310***	(0.723)	8.655***	(1.754)
N Cases	169		169		169		169
Wald χ²	125.26***		247.09***		118.26***		400.51***

***p < .001; **p < .01; *p < .05; ^†p < .10. Standard errors are clustered at the Senate race level.

The dependent variables in Table 2 are (1) the number of favorable ads and (2) the number of attack ads outside groups aired based on the Republican candidate . Thus, the first two models in Table 2 reflect the number of favorable and attack ads Republican-leaning outside groups aired based on the gender, incumbency status, and prior experience of the groups’ preferred-party Republican candidate . The third and fourth models in Table 2 reflect the number of favorable and attack ads Democratic-leaning outside groups ran based on the gender, incumbency status, and prior experience of the Republican opposition candidate .

Table 2.

Outside Group Ads by Gender, Incumbency Status, and Prior Office-Holding Experience of the Republican Candidate.

	Rep Leaning Group Ads: Rep Candidate				Dem Leaning Group Ads: Rep Candidate
	# Favor		# Attack		# Favor		# Attack
Key variables
Female inexp	0.536	(0.700)	0.147	(0.482)	0.046	(0.394)	−0.760	(0.483)
Male exp	1.061***	(0.311)	−0.415*	(0.511)	−0.293	(0.319)	−0.026	(0.445)
Female exp	1.132**	(0.371)	−0.860	(0.646)	−0.511	(0.384)	−0.458	(0.579)
Male incumbent	1.293***	(0.351)	−0.628	(0.555)	−1.564**	(0.510)	−0.692	(0.566)
Female incumbent	1.544*	(1.544)	0.774 ^†	(0.468)	−0.924	(0.682)	−0.839	(0.583)
Controls
Ad cost	0.000	(0.001)	0.000 ^†	(0.000)	0.000	(0.000)	0.001	(0.000)
Inc leader	−0.075	(0.655)	1.327*	(0.668)	0.918	(0.746)	1.337**	(0.444)
% urban	−0.014	(0.014)	−0.029 ^†	(0.017)	0.011	(0.013)	−0.015	(0.018)
Solid race	−1.076***	(0.278)	−3.213***	(0.501)	−1.389***	(0.386)	−3.011***	(0.355)
Toss-up race	0.666***	(0.203)	1.228***	(0.281)	0.458 ^†	(0.246)	1.174***	(0.266)
2010	0.561	(0.411)	−0.883 ^†	(0.495)	−0.904*	(0.390)	−1.145*	(0.453)
2012	−0.176	(0.370)	−0.799 ^†	(0.426)	−0.752*	(0.343)	−0.948*	(0.402)
2014	0.120	(0.357)	−0.834 ^†	(0.445)	−0.814*	(0.375)	−0.958 ^†	(0.511)
2016	0.528	(0.466)	0.068	(0.514)	−0.241	(0.388)	−0.248	(0.698)
Constant	7.633***	(0.885)	11.196***	(1.310)	8.466***	(0.962)	10.129***	(1.277)
N Cases	168		168		168		168
Wald χ²	127.58***		273.44***		99.88***		314.76***

***p < .001; **p < .01; *p < .05; ^†p < .10. Standard errors are clustered at the Senate race level.

The first five rows of Tables 1 and 2 present the results of our analyses focused on our primary independent variables: the candidate’s gender, incumbency status, and prior office-holding experience. We account for whether a candidate is an incumbent or has held prior elected office because incumbents and experienced candidates are more likely to win (Carson et al., 2007; Cox and Katz, 1996; Jacobson and Carson, 2020; King and Gelman, 1991). Thus, outside groups should be more likely to air ads that support or attack incumbents and experienced challengers than to air ads on behalf of inexperienced newcomers. Our models include the key independent variables for the gender, incumbency status, and prior office-holding experience of the candidate that each ad targeted using the five dummy variables listed at the top of Tables 1 and 2: inexperienced female challenger (IFC) , experienced female challenger (EFC), experienced male challenger (EMC), male incumbent (MI), and female incumbent (FI) .⁴ The excluded baseline category is an inexperienced male challenger (IMC). For all of the analyses, incumbents are men and women who were currently serving in the U.S. Senate when an ad aired, experienced challengers are men and women who were not currently serving in the U.S. Senate, but had prior office-holding experience when an ad aired, and inexperienced challengers are men and women who were not serving in the Senate and did not have prior office-holding experience when an ad aired. Among Republican candidates, 19% are IMCs, 5.75% are IFCs, 35.06% are EMCs, 6.9% are EFCs, 29.31% are MIs, and 4.02% are FIs. Among Democratic candidates, 14.86% are IMCs, 4.0% are IFCs, 24.0% are EMCs, 13.71% are EFCs, 29.14% are MIs, and 14.29% are FIs.

The models also include several controls: Ad Cost , Toss-Up Race , Solid Race , Incumbent Party Leader , Percent Urban , and election year dummy variables. Ad Costs are the estimated media market level costs in 2010–2018 WMP data, which are based on information provided by Kantar/Campaign Media Analysis Group (CMAG). Ad costs vary depending on the market, the timing, and sponsor of each ad, but the WMP data do not include the actual price of each ad that each group aired. Thus, data on each group’s actual television advertising expenditures are not available and the WMP’s estimated media market level costs are the best available proxy for how much each ad cost. We expect that outside groups have limited and fixed advertising budgets. Thus, as ad costs increase, outside groups should air fewer ads. As estimated ad costs increase, fewer groups should be able to afford to air ads, further decreasing the number of ads expected to air. The ad cost variable ranges from $39.90 to $2,804.15 with a mean of $566.83.

Next, we include two dichotomous variables that reflect the level of competition in each race: Toss-up Race and Solid Race . We utilized the Cook Political Report’s ratings for each U.S. Senate race in each election year to construct these variables. The Cook Report categorizes races as “solid Republican,” “likely Republican,” “leans Republican,” “toss-up Republican,” “toss-up Democrat,” “leans Democrat,” “likely Democrat,” or “solid Democrat.” Based on the last Cook Report assessment before Labor Day in a given election year, Toss-Up Race is coded “1” if a race is a toss-up Republican or toss-up Democratic race and “0” otherwise. In the observed time-period, 26.74% of the races were toss-up races. Solid Race is coded “1” if the race is a solid Democratic or solid Republican race. In the observed time-period, 47.19% of the races were solid races. The baseline category reflects races that are likely or leaning Republican or Democrat. We expect outside groups will air fewer ads in solid races compared to the toss-up, likely, and leaning races that are more competitive.

The model also includes control variables for whether an incumbent party leader is running and urbanicity . Incumbent party leader is included because outside groups prioritize contributing to incumbents with seniority and/or leadership roles (Brunell 2005; Grier and Munger, 1993; Hall and Deardorff, 2006). Incumbent Party Leader is a binary variable coded “1” if the incumbent is a party leader. In the observed time-period, only 1.1% of the races included an incumbent party leader. Urbanicity ( % Urban ) is the proportion of the state population that resides in an urban area, and it ranges from 38.66% to 94.95% with a mean of 74.09%. Party leaders have expressed concerns that rural areas will not elect women (Sanbonmatsu, 2006) and we assume that outside groups with partisan/ideological motivations will follow suit. Thus, we include urbanicity because we expect outside groups will be more likely to air more ads supporting female candidates and fewer ads attacking them in urban districts. Finally, the models include binary dummy variables for 2010 , 2012 , 2014 , and 2016 with 2018 as the baseline. These year variables control for differences in ad use across elections.

Ads Aired Based on Gender, Incumbency Status, and Experience of the Democratic Candidate

Because the dependent variables are all counts (number of favorable and attack ads), we use negative binomial regression to estimate the models. Table 1 presents the results for the number of favorable and attack ads , contingent on the gender, incumbency status, and prior experience of the Democratic candidate in the race. The results on the left display the findings for the number of favorable and attack ads Republican-leaning groups aired as a function of the gender, incumbency status, and prior experience of the Democratic opposition candidate. The results on the right show the findings for the number of favorable and attack ads Democratic-leaning groups aired as a function of the gender, incumbency status, and prior experience of their preferred Democratic candidate.

Given the categorical nature of the gender and experience independent variables, it is not easy to discern the impact of gender, incumbency status, and prior experience on the number of ads aired simply by examining negative binomial coefficients. Therefore, we present the predicted counts for number of favorable and attack ads that both Democratic- and Republican-leaning outside groups aired based on the gender, incumbency status, and prior experience of Democratic candidate in Table 3.

Table 3.

Differences in Outside Group Ad Use: Democratic Candidate.

	Rep Leaning Group Ads: Dem Candidate		Dem Leaning Group Ads: Dem Candidate
	# Favor	# Attack	# Favor	# Attack
Male incum	1755.18	8,915.46	5,914.15	15,179.65
Female incum	1,641.51	7,364.66	7,397.45	13,306.34
∆	−113.67	−1,550.80	1,483.30	−1873.20
Male exp	5,262.13	17,209.83	2,791.68	5,072.36
Female exp	1830.96	2,868.24	1,664.50	2,961.89
∆	−3,431.16*	−14341.68*	−1,127.19*	−2,110.47
Male inexp	3,540.54	2,919.81	790.79	2,708.83
Female inexp	11,739.89	1819.62	4,350.48	9,697.99
∆	8,199.36	1,100.19	3,559.69*	6,989.17

The findings in Table 3 indicate Republican-leaning outside groups aired significantly fewer favorable ads (3,431.16 fewer ads) in races that include experienced female Democratic challengers as one of the two general election candidates (1,830.96 ads) compared to races that include experienced male Democratic challengers (5,262.13 ads). Republican-leaning outside groups also aired significantly fewer attack ads (14,341.68 fewer ads) in races that include experienced female Democratic challengers (2868.24 ads) compared to races that include experienced male Democratic challengers (17,209.83 ads).

Next, Democratic-leaning outside groups air significantly fewer favorable ads (1,127.19 fewer ads) in races that include experienced female Democratic challengers (1,664.50 ads) compared to races that include experienced male Democratic challengers (2,791.68 ads). However, Democratic-leaning outside groups run significantly more favorable ads (3,559.69 more ads) in races that include inexperienced female Democratic challengers (4,350.48 ads) compared to races that include inexperienced male Democratic challengers (790.79 ads).

Together, both Republican-leaning and Democratic-leaning outside group s air significantly more favorable ads in races that include experienced male Democratic challengers compared to races that include experienced female Democratic challengers. These results reject our first hypothesis, which suggests Democratic-leaning groups should be more likely to run favorable ads in races that include experienced or incumbent Democratic women than experienced or incumbent men. However, they partially confirm our second hypothesis, which states Republican-leaning groups should air more favorable ads in races that include experienced or incumbent Democratic men. Republican-leaning outside groups also air significantly more attack ad in races that include experienced male Democratic challengers compared to races that include experienced female Democratic challengers, partially confirming our third hypothesis that outside groups should be more likely to air attack ads when the opposition party candidate is a man.

Ad Aired Based on Gender, Incumbency Status, and Experience of the Republican Candidate

The results displayed on the left of Table 2 present the negative binomial model results for the number of favorable and attack ads Republican-leaning outside groups ran as a function of the group’s preferred-party Republican candidate . The right-hand side of Table 2 presents the results for the number of favorable and attack ads Democratic-leaning groups aired as a function of the gender, incumbency status, and prior experience of the group’s opposition Republican candidate. Table 4 presents predicted counts for the number of favorable and attack ads outside groups aired, this time based on the gender, incumbency status, and prior experience of the Republican candidate.

Table 4.

Differences in Outside Group Ad Use: Republican Candidate.

	Rep Leaning Group Ads: Rep Candidate		Dem Leaning Group Ads: Rep Candidate
	# Favor	# Attack	# Favor	# Attack
Male incum	3,574.09	6,143.73	1,353.06	4,367.55
Female incum	4,593.35	24,975.90	2,564.55	3,770.52
∆	1,019.26	18,832.17	1,211.49	597.03
Male exp	2,834.84	7,601.78	4,824.18	8,494.26
Female exp	3,041.94	4,868.50	3,878.09	5,515.35
∆	207.10	−2,733.28	−946.08	−2,978.91
Male inexp	980.80	11,514.52	6,466.96	8,721.86
Female inexp	1,676.36	13,340.56	6,774.04	4,077.79
∆	695.56	1826.03	307.08	−4,644.07

The left panel of Table 4 shows there are no significant differences in the number favorable ads or attack ads Republican-leaning groups aired when comparing races that include male Republican incumbents to races with female Republican incumbent candidates, races that include male experienced Republican challengers to races with female experienced Republican challengers, or races that include male inexperienced Republican challengers to races with female inexperienced Republican challengers. These findings indicate that, contrary to expectations, Republican-leaning groups do not air more favorable ads for their experienced or incumbent preferred-party male Republican candidates than they do for experienced or incumbent female Republicans.

The right panel of Table 4 provides the results for the Democratic-leaning outside groups. We unexpectedly find there is no significant difference in the number favorable or attack ads when comparing races that include male Republican incumbents to races with female Republican incumbents, races that include experienced Republican male challengers to races with experienced female Republican challengers, or races that include inexperienced Republican male challengers to races with inexperienced female Republican challengers. Together, the findings indicate the gender, incumbency status, and prior experience of Republican party candidate is not significantly related to the number of ads Republican- or Democratic-leaning outside group air in Senate races. These results stand in stark contrast to the results focusing on the number of ads outside groups air based on the gender and experience of Democratic candidates.

Discussion

Despite recent increases in the number of women and women of color serving in the Senate (CAWP 2023a; 2023b) and the increasing prominence of women’s issues on the agenda during the 2010s, our analyses suggest the proliferation of outside group ads created some new gendered challenges for experienced female candidates who ran for Senate seats between 2000 and 2018. Though we expected Democratic-leaning groups would air more favorable ads supporting female candidates, Democratic-leaning groups air significantly more favorable ads in races that include experienced male Democratic challengers compared to races that include experienced female Democratic challengers.

Democratic-leaning groups’ unexpected preference for airing favorable ads in races that include experienced male Democratic challengers is surprising, because Democrats provide more support for female candidates and increasing women’s descriptive representation than their Republican counterparts do (Crowder-Meyer and Cooperman, 2018; Fraga and Hassell, 2021). Why aren’t outside groups producing similar advantages for Democratic women by airing more favorable ads during their races? One possibility is that Democratic women’s groups (e.g., EMILY’s List) prioritize candidate training, technical assistance, sharing information about politics and policy, and/or bundling campaign contributions, over running ads (Crowder-Meyer and Cooperman, 2018; Day and Hadley, 2005; Fraga and Hassell, 2021; Hassell and Vislvanich, 2019; Kitchens and Swers, 2016; Kritzer and Osborn, 2019; Thomsen and Swers, 2017). EMILY’s List/Women Vote prioritizes supporting incumbent women (Malcolm, 2016; Schriock, 2021). Thus, it is possible that EMILY’s List/Women Vote and other women’s groups run more ads in races that include women incumbents than races with experienced challengers because prioritize maintaining continued access to incumbents (Grier and Munger, 1993; Hall and Deardorff, 2006). If that is the case, then the overall number of ads Democratic-leaning groups air for experienced women challengers would drop, potentially giving experienced male Democratic challengers an advantage over experienced female Democratic challengers on the airwaves. Another possibility is that Democratic-leaning groups not explicitly focused on increasing women’s descriptive representation do not run as many ads for Democratic women as the Democratic women’s groups do, because male Democratic donors do give more to male candidates than female candidates (Thomsen and Swers, 2017). Thus, those donors may also push broadly-focused Democratic-leaning groups to use those funds to air more ads for experienced male challengers rather than experienced female challengers. Future research should test these hypotheses.

Though our results suggest experienced Democratic women challengers face some advertising disadvantages, there are some ways in which the increasing number of outside group ads in Senate races is helping, or at least not harming, female candidates. Most notably, our results indicate outside groups are not airing more attack ads against female candidates as compared to their male counterparts. We only found one gender difference associated with the number of attack ads outside groups air, and it poses a challenge for men (not women). Republican-leaning outside groups are significantly more likely to air attack ads in races that include experienced male Democratic challengers as one of the two contenders than races that include experienced Democratic women as competitors. Our findings that outside groups do not air more negative ads targeting women may be good news for women who see negative campaigning as a deterrent to running (Lawless and Fox, 2010). Moreover, Republican-leaning groups may shy away from airing ads targeting experienced Democratic women challengers to avoid accusations of unfairly attacking women (Kahn and Kenney, 1999). Instead, they air ads in races that include experienced male Democratic challengers as one of the two competitors because those races are less likely to produce gendered backlashes.

The last notable gender difference we found indicated Democratic-leaning outside groups air more favorable ads in races that include inexperienced female challengers than they do in races that include inexperienced male challengers. This may be intriguing evidence that Democratic groups are starting to prioritize supporting new types of Democratic candidates, including women of color, who are more likely to run for office without prior office-holding experience or without coming from the kinds of traditional recruiting networks that often benefit white candidates and men (Doherty et al., 2019; Niven, 2006; Ocampo, 2018; Ocampo and Ray, 2020; Sanbonmatsu, 2006). Given the record number of women of color that ran and won in 2018 and an increase in the number of candidates without prior office-holding experience (CAWP, 2018a; 2018b; Dittmar, 2018; 2020; Walter and Jones, 2022), we expect that Democratic-leaning groups’ preference for running ads on behalf of inexperienced female candidates rather than inexperienced male candidates will only grow over time.

In contrast to our findings related to the number of ads outside groups air based on the gender and experience of the Democratic candidate in a given race, there are no significant differences in the number of favorable and attack ads that outside groups run as a function of the gender, incumbency status, and prior experience of the Republican candidate in the race. Republican-leaning groups, surprisingly, do not seem to favor their own preferred-party candidates when they are men. These null findings could be driven by the small number of Republican incumbent women and challengers in our dataset that outside groups could have supported with ads. Our findings should be revisited as the number of Republican female Senators and candidates increases.

As the number of women of color serving in Congress reached record highs in 2018 and 2020 (CAWP, 2018a; 2018b; Dittmar, 2018; 2020), more research should examine how the number of ads that outside groups air is associated with candidates’ race/ethnicity, in addition to their gender, incumbency status, and prior experience. Though women’s groups exist to recruit and support female candidates on both sides of the aisle, only 31 out of nearly 400 groups that support female candidates explicitly focus on women of color (Kritzer and Osborn, 2019). Women of color running in House races also raise only 70 to 80 percent of the amounts raised by other candidates, and Democratic donors and party committees provide more support for white women than they do for women of color (Fraga and Hassell, 2010; Hassell and Visalvanich, 2019; Sorenson and Chen, 2022, 745). In interviews, Democratic leaders, office-holders, and party elites claim the party “could and should” do more to recruit and support women candidates of color (Elder, 2021). Our findings, which show Democratic groups run more favorable ads in races with inexperienced female challengers, suggest these groups may already be trying to support more women candidates of color without prior office experience. However, this was not explicitly tested in this project and should be in future research.

Outside groups that support candidates of color more broadly may also struggle to air large numbers of ads on behalf of candidates who have been marginalized due to their race or ethnicity. Black candidates receive fewer PAC contributions than their white counterparts, and Black House members raise fewer campaign funds than their white counterparts, making it more difficult for them to run for Senate seats (Johnson et al., 2012; Wilhite and Theilman, 1986). Democratic and Republican leaders also believe that Black and Latino/a primary candidates are less likely to win their races, and party leaders provide fewer endorsements for Latino/a candidates, excluding them from important partisan networks (Doherty et al., 2019; Ocampo, 2018; Ocampo and Ray, 2020). In contrast, Fraga and Hassell (2021) unexpectedly found that minority Republican House candidates received more donations from Republican donors than the white candidates that ran in the previous cycle, potentially due to Republicans’ attempts to appeal to more minority voters or win over white moderates.

Additional research on the number of ads outside groups air on behalf of candidates of color might help further untangle these mixed results. Because some SuperPACs exist to pull the parties to their ideological extremes (Dwyre and Braz, 2014), we expect Fraga and Hassell’s (2021) finding that Republicans provide more support to minority candidates than white ones will not generalize to the number of ads that Republican-leaning outside groups air. In other words, we do not expect to find that Republican outside groups try to appeal to moderate (white) voters by airing more ads on behalf of candidates of color. Ultimately, we expect that outside group donors contribute less to groups that air ads on behalf of male and female candidates of color, making those groups less likely to air favorable ads for them. Though outside groups tend to air ads that are more negative and aggressive than those aired by parties or candidates (Dowling and Miller, 2014; Dowling and Wichowsky, 2015; Herrnson, 2017), we also expect that they will air few attack ads against candidates of color to avoid potential charges of racism.

Conclusion

Our results provide an important first step towards understanding gender differences in the number of television ads outside groups air during Senate campaigns. We show Republican- and Democratic-leaning groups approach candidates’ gender, incumbency status, and prior experience similarly when they decide how many ads to air during Senate races, as both types of groups air more favorable ads in races that include experienced male Democratic challengers as one of the two contenders compared to races that include experienced female Democratic challengers as one of the two contenders. Republican-leaning outside groups also air significantly more attack ads in races that include experienced male Democratic challengers than races that include experienced female Democratic challengers. Democratic-leaning outside groups also run significantly more favorable ads in races that include inexperienced female Democratic challengers compared to races that include inexperienced male Democratic challengers. At all levels of experience, there are no differences in the number of ads Democratic- or Republican-leaning outside groups air based on whether the Republican candidate is a man or a woman. Republican-leaning groups do not air more favorable ads on behalf of Republican men than Republican women. Therefore, we show that gender differences in the number of ads outside groups air are likely to have the largest consequences for Democratic women—hurting experienced Democratic women running for the Senate for the first time, but potentially helping inexperienced Democratic women.

Supplemental Material

Supplemental Material - Outside of the Old Boys Club? Gender Differences in Outside Groups’ Advertising Support for U.S. Senate Candidates

Supplemental Material Outside of the Old Boys Club? Gender Differences in Outside Groups’ Advertising Support for U.S. Senate Candidates by Ashley English, Regina Branton, and Amy Friesenhahn in Political Research Quarterly

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

Ashley English

Regina Branton

Amy Friesenhahn

Supplemental Material

Supplemental material for this article is available online.

Notes

References

Abramowitz

Alan

. 2012. “US Senate Elections in a Polarized Era.” In The US Senate from Deliberation to Dysfunction, 27–48. Washington, DC: CQ.

Andrich

Aliya

Bachl

Marko

Domahidi

Emese

. 2023. “Goodbye, Gender Stereotypes?” Journalism and Mass Communication Quarterly. doi:10.1177/10776990221142248.

Banwart

Mary Christine

Bystrom

Dianne G.

Robertson

Terry

. 2003. “From the Primary to the General Election.” American Behavioral Scientist 46 (5): 658–676.

Bauer

Nichole

. 2020. The Qualifications Gap. Cambridge: Cambridge University.

Bauer

Nichole

Taylor

Tatum

. 2023. “Selling Them Short?” Political Research Quarterly 76 (1): 308–322.

Boatright

Robert

. 2007. “Situating the New 527 Organizations in Interest Group Theory.” The Forum 5 (2).

Brooks

Deborah Jordan

Murov

Michael

. 2012. “Assessing Accountability in a Post-Citizens United Era.” American Politics Research 40 (3): 383–418.

Bryner

Sarah

Weber

Doug

. 2013. Sex, Money, and Politics. https://www.opensecrets.org/downloads/CRP_Gender_Report_2013.pdf. (July 20, 2023).

Burrell

Barbara

. 2018. “Political Parties and Women’s Organizations.” In Gender and Elections, edited by Carroll

Susan

Fox

Richard

, 220–249. New York: Cambridge University.

10.

Bystrom

Dianne

Banwart

Mary Christine

Lee Kaid

Lynda

Robertson

Terry

. 2004. Gender and Candidate Communication. New York: Routledge.

11.

Brunell

Thomas L

. 2005 “The relationship between political parties and interest groups: Explaining patterns of PAC contributions to candidates for congress.” Political Research Quarterly 58(4): 681–688.

12.

Carlin

Diana

Winfrey

Kelly

. 2023. “Have You Come a Long Way, Baby, Since 2008?” Communication Studies 74 (2): 131–146.

13.

Carroll

Susan

Sanbonmatsu

Kira

. 2013. More Women Can Run. Oxford: Oxford.

14.

Carson

Jamie

Engstrom

Erik

Roberts

Jason

. 2007. “Candidate Quality, the Personal Vote, and the Incumbency Advantage in Congress.” American Political Science Review 101.

15.

Cassese

Erin

Conroy

Meredith

Mehta

Dhrumil

Nestor

Franchesca

. 2021. “Media Coverage of Female Candidates’ Traits in the 2020 Democratic Primary.” Journal of Women, Politics & Policy 43(1): 42–63.

16.

CAWP . 2023a. “Women Officeholders by Race and Ethnicity.” https://cawp.rutgers.edu/facts/women-officeholders-race-and-ethnicity. (July 17, 2023).

17.

CAWP . 2023b. “Women in the US Senate 2023.” https://cawp.rutgers.edu/facts/levels-office/congress/women-us-senate-2023. (July 14, 2023).

18.

CAWP . 2018a. “Data Point: 2018.” https://cawp.rutgers.edu/sites/default/files/resources/data-point-compare-1992-2018.pdf. (June 1, 2021).

19.

CAWP . 2018b. “Results.” https://cawp.rutgers.edu/sites/default/files/resources/results_release_5bletterhead5d_1.pdf. (November 18, 2020).

20.

Cox

Gary

Katz

Jonathan

. 1996. “Why Did the Incumbency Advantage in U.S. House Elections Grow?” American Journal of Political Science 40 (2): 478–497.

21.

Crespin

Michael

Deitz

Janna

. 2010. “If You Can’t Join ‘Em, Beat ‘Em.” Political Research Quarterly 63 (3): 581–593.

22.

Crowder-Meyer

Melody

. 2013. “Gendered Recruitment without Trying.” Politics & Gender 9 (4): 390–413.

23.

Crowder-Meyer

Melody

Cooperman

Rosalyn

. 2018. “Can’t Buy Them Love.” Journal of Politics 80 (4): 1211–1224.

24.

CRP . 2023a. “All Disclosed Outside Spending by Spenders’ Industries.” https://www.opensecrets.org/outside-spending/spenders-industries. (July 14, 2023).

25.

CRP . 2023b. “Planned Parenthood Outside Spending.” https://www.opensecrets.org/outside-spending/detail/2020?cmte=Planned+Parenthood&tab=summary. (August 1, 2023).

26.

Day

Christine

Hadley

Charles

. 2005. Women’s PACs. Upper Saddle River: Prentice Hall.

27.

Deckman

Melissa

. 2016. Tea Party Women. New York: NYU.

28.

Dittmar

Kelly

. 2020. “What You Need to Know about the Record Numbers of Women Candidates in 2020.” https://cawp.rutgers.edu/blog/what-you-need-know-about-record-numbers-women-candidates-2020. (July 17, 2023).

29.

Dittmar

Kelly

. 2018. “Putting the Record Numbers of Women’s Candidacies into Context.” https://cawp.rutgers.edu/footnotes/putting-record-numbers-womens-candidacies-context. (November 18, 2020).

30.

Dittmar

Kelly

. 2015. Navigating Gendered Terrain. Philadelphia: Temple.

31.

Doherty

David

Dowling

Conor

Miller

Michael

. 2019. “Do Local Party Chairs Think Women and Minority Candidates Can Win?” Journal of Politics 81 (4): 1282–1297.

32.

Dowling

Conor

Miller

Michael

. 2014. SuperPAC! New York: Routledge.

33.

Dowling

Conor

Wichowsky

Amber

. 2015. “Attacks without Consequence?” American Journal of Political Science 59 (1): 19–36.

34.

Dunaway

Johanna

Lawrence

Regina G.

Rose

Melody

Weber

Christopher R.

. 2013. “Traits versus Issues.” Political Research Quarterly 66 (3): 715–726.

35.

Dwyre

Diana

Braz

Evelyn

. 2014. “SuperPAC Spending Strategies in the 2012 Federal Elections.” In Presented at the APSA Conference, Washington, DC, August 28-31.

36.

Elder

Laurel

. 2021. The Partisan Gap. New York: NYU.

37.

EMILY’s List . 2023. “Women Vote.” https://emilyslist.org/women-vote-emilys-list/. (July 18, 2023).

38.

Engstrom

Richard

Kenny

Christopher

. 2002. “The Effects of Independent Expenditures in Senate Elections.” Political Research Quarterly 55 (4): 885–905.

39.

FEC . 2023. “Contribution Limits for 2023-2024.” https://www.fec.gov/updates/contribution-limits-for-2023-2024/. (July 14, 2023).

40.

FEC . 2020. “Individual Contributions.” https://www.fec.gov/data/receipts/individual-contributions/?two_year_transaction_period=2020&min_date=01%2F01%2F2019&max_date=12%2F31%2F2020. (September 16, 2020).

41.

FEC . 2019. Federal Elections 2018. https://www.fec.gov/resources/cms-content/documents/federalelections2018.pdf. (March 11, 2021).

42.

Fowler

Erika Franklin

Franz

Michael

Ridout

Travis

. 2016. Political Advertising in the United States. Boulder: Westview Press.

43.

Fraga

Bernard

Hans

Hassell

. 2021. Are Minority and Women Candidates Penalized by Party Politics? Political Research Quarterly 74(3): 540–555.

44.

Garrett

R. Sam

. 2013. SuperPACs in Federal Elections: Overview and Issues for Congress. Washington, DC: Congressional Research Service.

45.

Grier

Kevin

Munger

Michael

. 1993. “Comparing Interest Group PAC Contributions to House and Senate Incumbents, 1980-1986.” Journal of Politics 55 (3): 615–643.

46.

Geras

Matthew

. 2021. “Women Running for the Party and Women Running for Congress.” Party Politics 27 (5): 942–952.

47.

Gershon

Sarah

. 2012. “When Race, Gender, and the Media Intersect.” Journal of Women, Politics & Policy 33 (2): 105–125.

48.

Gibbons

Stephanie

. 2022. “Gender on the Agenda.” Newspaper Research Journal 43 (1): 102–123.

49.

Godble

Maya

Malvar

Noelle

Valian

Virginia

. 2019. “Gender, Modern Sexism, and the 2016 Election.” Politics, Groups and Identities 7 (3): 700–712.

50.

Hall

Richard

Deardorff

Alan

. 2006. "Lobbying as Legislative Subsidy." American Political Science Review 100 (1): 69–84.

51.

Hassell

Hans

Visalvanich

Neil

. 2019. “The Party’s Primary Preferences.” American Journal of Political Science 63 (4): 905–919.

52.

Herrnson

Paul

. 2017. “The Impact of Organizational Characteristics on SuperPAC Financing and Independent Expenditures.” In Presented at the Bipartisan Policy Center, Washington, DC.

53.

Herrnson

Paul

Heerwig

Jennifer

. 2021. “The Super Women and the Super Men Behind Super PACs.”

54.

Internal Revenue Service . 2019a. “Business Leagues.” https://www.irs.gov/charities-non-profits/other-non-profits/business-leagues. (August 12, 2019).

55.

Internal Revenue Service . 2019b. “Labor and Agricultural Organizations.” https://www.irs.gov/charities-non-profits/other-non-profits/labor-and-agricultural-organizations. (August 12, 2019).

56.

Internal Revenue Service . 2019c. “Political Activity and Social Welfare.” https://www.irs.gov/charities-non-profits/other-non-profits/political-activity-and-social-welfare. (August 12, 2019).

57.

Jacobson

Gary

Carson

Jamie

. 2020. The Politics of Congressional Elections. New York: Rowman & Littlefield.

58.

Johnson

Gbemende

Oppenheimer

Bruce

Selin

Jennifer

. 2012. "The House as a Stepping Stone to the Senate." American Journal of Political Science 56 (2): 387–399.

59.

Kahn

Kim Fridkin

Kenney

Patrick

. 2004. No Holds Barred. Upper Saddle River: Pearson Prentice Hall.

60.

Kahn

Kim Fridkin

Kenney

Patrick

. 1999. The Spectacle of US Senate Campaigns. Princeton: Princeton University.

61.

King

Gary

Gelman

Andrew

. 1991. “Systematic Consequences of Incumbency Advantage in U.S. House Elections.” American Journal of Political Science 35 (1): 110–138.

62.

Kitchens

Karin

Swers

Michele

. 2016. “Why Aren’t There More Republican Women in Congress?” Politics & Gender 648–676.

63.

Kritzer

Rebecca

Osborn

Tracy

. 2019. “The Emergence and Activities of Women’s Recruiting Groups in the U.S.” Politics, Groups, and Identities 7 (4): 842–852.

64.

Kurtzleben

Danielle

. 2020. “How A Record Number Of Republican Women Got Elected To Congress.” NPR, November 13. https://www.npr.org/2020/11/13/934249216/how-a-record-number-of-republican-women-got-elected-to-congress

65.

Lawless

Jennifer

Fox

Richard

. 2010. It Still Takes a Candidate. Cambridge: Cambridge University.

66.

Lee

Frances

. 2012. “Individual and Partisan Activism on the Senate Floor.” In The US Senate from Deliberation to Dysfunction, 110–131. Washington, DC: CQ.

67.

Malcolm

Ellen

. 2016. When Women Win. New York: Houghton Mifflin Harcourt.

68.

Meier

Robert Paul

. 1999. “The Darker Side of Nonprofits: When Charities and Social Welfare Groups Become Political Slush Funds.” University of Pennsylvania Law Review 147 (4): 971–1008.

69.

Niven

David

. 2006. “Throwing Your Hat Out of the Ring.” Politics & Gender 2 (4): 473–489.

70.

Niven

David

. 1998. The Missing Majority. Westport, CT: Praeger.

71.

Ocampo

Angela

. 2018. “The Wielding Influence of Political Networks.” Political Research Quarterly 71 (1): 184–198.

72.

Ocampo

Angela

Ray

John

. 2020. “Many Are Called But Few Are Chosen.” Politics, Groups, and Identities 8 (4): 738–761.

73.

O’Regan

Valerie

Stambough

Stephen

. 2011. “The Novelty Impact.” Journal of Women, Politics & Policy 32 (2): 96–113.

74.

Potter

Trevor

Morgan

Bryson

. 2013. “The History of Undisclosed Spending in U.S. Elections & How 2012 Became the Dark Money Election.” Notre Dame Journal of Law, Ethics, and Public Policy 27 (2): 383–479.

75.

Sanbonmatsu

Kira

. 2006. Where Women Run. Ann Arbor: University of Michigan.

76.

Sanbonmatsu

Kira

. 2004. Democrats, Republicans, and the Politics of Women’s Place. Ann Arbor: University of Michigan.

77.

Schriock

Stephanie

. 2021. Run to Win. Dutton.

78.

Sides

John

Shaw

Daron

Grossman

Matt

Lipsitz

Keena

. 2019. Campaigns and Elections. New York: W.W. Norton.

79.

Schreiber

Ronnee

. 2008. Righting Feminism. Oxford: Oxford University.

80.

Sinclair

Barbara

. 2012. “Senate Parties and Party Leadership, 1960-2010.” In The US Senate from Deliberation to Dysfunction, 85–109. Washington, DC: CQ.

81.

Sorenson

Ashley

Chen

Philip

. 2022. “Identity in Campaign Finance and Elections.” Political Research Quarterly 75 (3): 738–753.

82.

Swers

Michele

. 2013. Women in the Club. Chicago: University of Chicago.

83.

Swers

Michele

. 2002. The Difference Women Make. Chicago: University of Chicago.

84.

Thomsen

Danielle

Swers

Michele

. 2017. “Which Women Can Run?” Political Research Quarterly 70 (2): 449–463.

85.

U.S. Senate . 2023. “About Parties and Leadership.” https://www.senate.gov/about/origins-foundations/parties-leadership.htm. (July 20, 2023).

86.

van der Pas

Daphne J.

Aaldering

Loes

. 2020. “Gender Differences in Political Media Coverage.” Journal of Communication 70: 114–143.

87.

Walter

Amy

Jones

Peter

. 2022. “Rookie Politics.” https://www.cookpolitical.com/analysis/national/national-politics/rookie-politics-congress-becoming-less-experienced#:∼:text=Theresultisasmallin1981to34.6percent25today. (July 18, 2023).

88.

Ward

Orlanda

. 2016. “Seeing Double.” Politics & Gender 12: 317–343.

89.

WMP . 2021. “Political Ads in 2020.” https://mediaproject.wesleyan.edu/2020-summary-032321/. (July 18, 2023).

90.

Wilhite

Allen

Theilman

John

. 1986. “Women, Blacks, and PAC Discrimination.” Social Science Quarterly 67 (2): 283.

91.

Wineinger

Catherine

. 2022. Gendering the GOP. New York: Oxford 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

0.04 MB