In 2016, Erin Maye Quade became only the third Black woman ever elected to the Minnesota House of Representatives. In 2022, she made a different type of history by seeking the Democratic-Farmer-Labor Party’s endorsement at their convention. Maye Quade was very pregnant and, 10 days shy of her baby’s due date, went into labor the morning of the convention. Nonetheless, Maye Quade shook hands and greeted delegates for hours before delivering a speech, pausing to steady herself as a contraction came over her. Her (imminent) motherhood on full display, she ultimately had to withdraw from balloting to go to the hospital when convention rules failed to accommodate her circumstance.Footnote 1 However, she reentered the primary race just weeks after her daughter’s birth and won the August primary, despite having to forgo the party endorsement. In November 2022, she became the first Black woman elected to Minnesota’s state senate.
Although research in the United States to date notes the potential pitfalls of political candidates who are parents, these studies largely examine how motherhood or maternal appeals impact the candidacies of white women (Bell and Kaufmann Reference Bell and Kaufmann2015; Deason, Greenlee, and Langer Reference Deason, Greenlee and Langer2015). American voters hold distinct stereotypes about women of color as political candidates (Brown and Lemi Reference Brown and Lemi2021; Cargile Reference Cargile2023; Gershon and Monforti Reference Gershon and Monforti2021), and these stereotypes intersect with maternal identity in ways that differ from stereotypes white women mothers face. We thus apply a framework of intersectional motherhood and candidate evaluation that incorporates US-based stereotypes about race, gender, and parental status to clarify how candidates like Maye Quade may be perceived by the electorate.
Previous research demonstrates that race and gender (McDermott Reference McDermott1998) and parenthood affect how Americans evaluate candidates (Elder Reference Elder and Greene2012). We examine the intersectional nature of these overlapping and interconnected identities. Our framework of intersectional motherhood has two parts. First, we argue that American Black women candidates will have different incentives to emphasize motherhood relative to their white women counterparts. Second, American voters will evaluate white women and non-white women differently when they emphasize their parental status candidates due to the intersectional nature of these women’s political identities. The existing literature in the United States suggests there are multiple theoretical reasons to believe that women of color running for office as mothers may receive more positive evaluations from voters than white women running for office, including the multiple in-group identity advantages that diverse candidates may accumulate (Bejarano Reference Bejarano2013; Gershon and Monforti Reference Gershon and Monforti2021).
We test our hypotheses using two different approaches. First, we use candidate advertising data to track patterns across candidate race and gender in the use of maternal appeals in the United States. Second, we use data from an original survey experiment using Black women to investigate how Americans’ candidate evaluations are impacted by the intersection of race and motherhood. We find, first, that Black women are just as likely to use motherhood messages relative to white women, and that Black women receive positive evaluations from voters from a message emphasizing motherhood. However, we also find that white and minority voters are equally likely to positively rate Black women who emphasize motherhood in US electoral contests. These findings contribute to the literature on women candidates in the United States and expand our understanding of voter evaluations at a time when American voters are presented with an increasing number of mothers, and mothers of color, seeking elective office. Ultimately, our results speak to the power and ability of Black women to run for and win political office in the United States, which has implications for the representation of Black women in elected leadership roles.
Gender Roles, Racial Stereotypes, and Political Leadership
Feminine stereotypes characterize women as caring, compassionate, beautiful, and intuitive (Diekman and Eagly Reference Diekman and Eagly2000), whereas masculine stereotypes characterize men as tough, competitive, and rational (Vinkenburg et al. Reference Vinkenburg, van Engen, Eagly and Johannesen-Schmidt2011). These stereotypes emerge from the social roles women and men perform with feminine stereotypes matching women’s communal social roles, including being a mother (Eagly and Karau Reference Eagly and Karau2002), whereas masculine stereotypes match men’s agentic, or power-oriented, social roles including being a political leader (Koenig et al. Reference Koenig, Eagly, Mitchell and Ristikari2011). Feminine stereotypes generally disadvantage women because they are incongruent with the masculine stereotypes that most strongly align with political leadership in the United States (Bauer Reference Bauer2015; Ditonto Reference Ditonto2017; Schneider, Bos, and DiFillippo Reference Schneider, Bos and DiFillippo2021; Sweet-Cushman Reference Sweet-Cushman2022). Indeed, American candidates who are mothers have been shown to be less desirable as candidates than candidates who are fathers (Campbell and Cowley Reference Campbell and Cowley2018; Stalsburg Reference Stalsburg2010).
Despite the perceived incongruence with the masculinity of political leadership, motherhood can be a strategic strength for women political leaders under certain conditions (Deason, Greenlee, and Langer Reference Deason, Greenlee and Langer2015; Greenlee Reference Greenlee2014; Thomas and Lambert Reference Thomas, Lambert, Thomas and Bittner2017). In fact, Stalsburg (Reference Stalsburg2010) and Bell and Kaufmann (Reference Bell and Kaufmann2015) each find no evidence that American voters inflict a “mommy penalty” that creates bias against mothers. Rather, candidates who are mothers are rated more highly than childless women, especially with regard to expertise on issues relevant to children, though they are also seen as having less time to dedicate to their political offices (Stalsburg Reference Stalsburg2010). This is particularly true for voters who are conservative or who hold more traditional gender attitudes around motherhood (Bell and Kaufmann Reference Bell and Kaufmann2015).
Past work on candidate presentation in the United States finds that women are less likely than men to display their families on their campaign websites (Bystrom, Kaid, and Robertson Reference Bystrom, Kaid and Robertson2004; Stalsburg Reference Stalsburg2010), which has also been found to be true in Canadian elections (Thomas and Lambert Reference Thomas, Lambert, Thomas and Bittner2017). Maternal appeals (which are not limited to mothers per se), however, are not necessarily helpful to women’s campaigns. As Deason (Reference Deason2011) finds, using both an experiment and evaluation of American campaign ads, maternal appeals are detrimental for female candidates and especially for Republican women. Bauer (Reference Bauer2015) reaches a similar conclusion examining traits that are commonly associated with motherhood and that provoke maternal stereotypes, like nurturing and compassionate.
Much of the extant scholarship on motherhood appeals in the United States does not vary the race of the female candidate and largely examines the implications of motherhood strategies for white women. However, motherhood appeals may affect Black women differently compared to white women. Black women who are running for office are more likely to be concerned about how they, their families, and especially their children may be treated or perceived. Black women may, as Shames (Reference Shames2015) finds, make a rational choice not to run partially because of how it can adversely affect their families and because they have concerns about how their family life may be perceived by voters. When Black women make the strategic choice to run for office, they may thus be more cautious about bringing attention to their status as a mother. However, as Thomas and Lambert (Reference Thomas, Lambert, Thomas and Bittner2017) concluded in Canada, and others have agreed, “there is no clear agreement on the strategic advantage or disadvantage for women in politics of displaying their parental status” (173). We aim to shed light on this question by investigating how Black women emphasize motherhood in their strategic messages, and how Americans respond to such messages from Black women.
A Framework of Intersectional Motherhood
Intersectional motherhood draws on research delineating the intersectional lens through which voters view US-based candidacies and through which Black women view their political candidacies. Black pregnancy and Black motherhood have always been uniquely politicized in the United States because Black women experience oppression from racism and sexism in ways that are distinct from other women (Caldwell Reference Caldwell1991; Hooks Reference Hooks1981; Roberts Reference Roberts1994). Political rights for Black women have thus been much slower to materialize. Whereas emancipation brought the promise (if not continued realization) of voting rights for Black men, Black women remained disenfranchised. Furthermore, both white male and white female abolitionists never advocated for social equality. Similarly, the efforts of white women that brought about women’s suffrage marginalized the contribution of Black women, failed to garner them the right to vote, and did little to eradicate patriarchal oppression given white women’s investment in a social order that preserved their status.
These threats to white women were tied to persistent stereotypes of Black women. Misogynoir (Bailey and Trudy Reference Bailey and Trudy2018) can manifest in a number of forms, such as the “mammy,” “jezebel,” and “angry black woman” labels discussed by Harris-Perry (Reference Harris-Perry2013) or “Sapphires” (Hooks Reference Hooks1981), who are considered evil or wicked. The stereotypes about Black women shape how these women view their position as political leaders. This unique stereotypes about Black women emerged from social and economic conditions that meant Black women had to occupy both masculine roles as an economic provider and feminine roles as a mother (Reynolds-Dobbs, Thomas, and Harrison Reference Reynolds-Dobbs, Thomas and Harrison2008), though this did not also mean that Black men and women lacked a patriarchal power structure (Hooks Reference Hooks1981).
Our framework thus contends that Black women have incentives to emphasize motherhood because of these unique intersectional stereotypes that are also held about Black women in political leadership roles in the United States. Gershon and Monforti (Reference Gershon and Monforti2021) argue that there are, in fact, unique trait stereotypes based on the race/ethnicity/gender of women candidates. For Black women, Goff, Thomas, and Jackson (Reference Goff, Thomas and Jackson2008) argue that “blackness” is associated with “maleness,” leading to assumptions that Black women are more masculine than their white counterparts. Stereotype content about American Black women includes both masculine and feminine qualities (Ford Dowe Reference Ford Dowe2020; Holder, Jackson, and Poterotto Reference Holder, Jackson and Poterotto2015; Smooth Reference Smooth2006).
Black women’s candidacies are frequently motivated by a desire to serve their communities, as opposed to being motivated solely by political ambition, and Black women aim to create policies that benefit not just their whole communities but also traditionally marginalized groups (Brown Reference Brown2014). In short, Black women view their role as a political leader as one that draws on their political agency, a stereotypically masculine quality, and their desire to fulfill communal goals, thereby drawing on both masculine and feminine stereotypes. Although women of color face stereotypes (Hancock Reference Hancock2004) and other obstacles that may limit their candidate emergence and electoral success in the United States (Brown and Lemi Reference Brown and Lemi2021; Philpot and Walton Reference Philpot and Walton2007; Simien Reference Simien2005), we argue that Black women will see their maternal identity as a net positive.
This positivity eschews the patriarchal burden of “motherhood” while embracing the liberating qualities of “mothering” (O’Reilly Reference O’Reilly and O’Reilly2004; Rich Reference Rich1976). Black women will use their roles as mothers to emphasize both the feminine and masculine qualities of mothering that make them well suited for American political leadership (Wright and McNeely Reference Wright, McNeely, Brown and Gershon2023). In this way, Black women can create an image of strong motherhood in their political campaign strategies, while still maximizing the benefits of the more communal aspects of mothering. Moreover, these women of color who run for political office may rationally decide that the context of races means authenticity is the strongest strategy and reveal and/or emphasize their parental status. Recent research on the effect of authenticity on candidate evaluation offers evidence that this is a wise strategy; candidates who are perceived as authentic are more positively evaluated by American voters (Kenny, Larner, and Lewis-Beck Reference Kenny, Larner and Lewis-Beck2021; Stiers et al. Reference Stiers, Larner, Kenny, Brieitenstein, Vallee-Dubois and Lewis-Beck2021). We thus pose the hypothesis: (H1) Women of color will be more likely to mention their motherhood in campaign ads relative to white women.
Black women surely know they must contend with the provocation of negative stereotypes. Conventional wisdom on campaign messaging suggests that emphasizing feminine stereotypes can be risky for candidates because feminine stereotypes do not align with the largely masculine perceptions that people hold of political leaders (Bauer Reference Bauer2015; Sweet-Cushman Reference Sweet-Cushman2022). Black women are also rated more negatively on societal stereotypes about sexual activity, sexual risk, motherhood status, and socioeconomic status (Rosenthal and Lobel Reference Rosenthal and Lobel2016). When pregnant, Black women are more likely to be perceived as being a single mother or needing of public assistance and are more likely to be evaluated more harshly than expecting white women (Rosenthal and Lobel Reference Rosenthal and Lobel2016). These stereotypes can also be reinforced when there is coverage of candidates’ family lives, which women tend to receive more of (Falk Reference Falk2010; Stalsburg and Kleinberg Reference Stalsburg and Kleinberg2016). For Black candidates in the United States, racist language is not overt (Burge, Hodges, and Rinaldi Reference Burge, Hodges and Rinaldi2020; Collins Reference Collins2004; Hill Collins Reference Hill Collins2002; Wingfield and Feagin Reference Wingfield and Feagin2012) but rather “largely filtered through stereotypes of the Black family” (Burge, Hodges, and Rinaldi Reference Burge, Hodges and Rinaldi2020, 1026). At times, Black women candidates may explicitly engage with racism, such as those who talk about police violence toward Black youth (Smith Reference Smith2022). Essentially, we expect that voters will prefer a certain type of motherhood appeal that is conditioned by race and other factors. Thus, overall, we hypothesize that there may also be shortcomings associated with this strategy: (H2) Black women candidates who reveal that they are mothers will be evaluated more negatively than non-mothers and white mothers.
We also suspect there may also, however, be an upside. Voter evaluations of Black women candidates may also be influenced by in-group bias (Brewer and Brown Reference Brewer, Brown, Gilbert, Fiske and Lindzey1998; Tajfel et al. Reference Tajfel, Billig, Bundy and Flament1971), where American voters prefer someone who matches their identity (Abney and Hutcheson Reference Abney and Hutcheson1981; Barreto Reference Barreto2010; Bejarano Reference Bejarano2013; Sanbonmatsu Reference Sanbonmatsu2002). This is most relevant when voters make choices in low-information contexts like races that do not receive much media coverage, where diverse candidates may be most likely to first emerge as candidates. When voters know little about candidates, group memberships such as gender or race function as heuristics because group stereotypes are most influential in perception when little or no other information is available about the individual (Golebiowska Reference Golebiowska2001). Thus, women of color using motherhood appeals may find favor among voters who share a racial, gender, or motherhood identity, or some combination of the three identities, with the candidates — which is ultimately a large swath of the electorate, especially among Democratic voters who are more likely to be younger, non-white, and women.
There is, in fact, reason to believe that Black moms who run for office in the American political system might not be “doubly disadvantaged,” but rather their gender and racial identities interact to form a distinct identity in candidate evaluation (Philpot and Walton Reference Philpot and Walton2007). This potentially aids in perceived stereotype fit for Black moms since masculinity overlaps more significantly with stereotypes of politicians than do feminine stereotypes (Schneider and Bos Reference Schneider and Bos2011). In other words, we tend to stereotype politicians as having masculine traits, and Black women are more likely, because of their race, to be stereotyped as masculine. Mothers may be seen as broadly competent in the political arena for their multitasking, diplomacy, and budget management skills (Deason, Greenlee, and Langer Reference Deason, Greenlee and Langer2015). Women are seen as more competent in several US issue areas in the realm of social welfare and ethics, as are Blacks on civil rights and helping the poor (McDermott Reference McDermott1998). Gershon and Monforti (Reference Gershon and Monforti2021) find that Black women are not necessarily evaluated differently than white women, and there are traits such as strong leadership and experience that advantage Black women over white women. Furthermore, some women of color will have access to broader voting coalitions than their white counterparts (Bejarano Reference Bejarano2013), which has been shown to result in a “multiple identity advantage” for American Latinas and Black women (Gershon and Monforti Reference Gershon and Monforti2021). This leads to the following possibility: (H3) Black women candidates who reveal that they are mothers will be evaluated more positively by minority voters than will candidates who are non-mothers and white mother candidates.
Data and Methods
We examine the use of motherhood in two ways: candidate strategy in the form of maternal appeals (Study 1) and candidate evaluation in the form of voter assessment of candidates (Study 2). To analyze candidate strategy, in Study 1 we rely on data from the Wesleyan Media Project (WMP) to better understand how women candidates vary in their use of motherhood appeals in the United States. This study tests H1 about the strategic choices women of color make to use or not use motherhood appeals. In Study 2, we test H2 and H3 examine candidate evaluation with an original experiment that allows us to assess how individuals respond to women’s use of motherhood appeals based on the race of the woman.
Study 1: Motherhood in Campaign Messages
We look at campaign ad data across eight US election cycles and include both congressional and gubernatorial elections (2010 to 2018) using data from the Wesleyan Media Project (WMP) (Fowler et al. Reference Fowler, Franz, Ridout and Baum2019; Fowler et al. Reference Fowler, Franz, Ridout and Baum2020; Fowler, Franz, and Ridout Reference Fowler, Franz and Ridout2014, Reference Fowler, Franz and Ridout2015, Reference Fowler, Franz and Ridout2017). This dataset records the characteristics of ads aired by political candidates in the 200 major media markets in the US, such as the name and party of the candidate, the media market where the ad was aired, the media outlet used by the candidate, the level of the race the candidate was in, the sponsor of the ad, and the tone of the ad.
There are 27 Black women in our dataset. Although we are primarily interested in Black women’s use of motherhood messages, we examine how women of other racial and ethnic minority backgrounds use motherhood messages to test for differences among women of color as a group but also within women of color as a group (Greene, Matos, and Sanbonmatsu Reference Greene, Matos and Sanbonmatsu2021). In total, there are 95 women of color in the dataset (including the 27 Black women) who ran for a House, Senate, or gubernatorial seat between 2010 and 2018 and also aired televised campaign ads. The numbers of women, by their race and the office they sought, can be found in Figure 1; we include a more detailed breakdown of the women of color who aired campaign ads in Table A1 in Appendix 1. It is important to note that our dataset only includes candidates who aired television ads. There are 137 unique ads aired by Black women in our dataset. Although this is a relatively low number (less than 1% of the total dataset), these analyses can still offer us original insights into how Black women incorporate motherhood into their strategic messages. The relatively low number of Black women, and women of color more generally, in our sample is a limitation of these data analyses. Many Black women run in majority-minority districts where the level of competition may not require them to air televised ads. Nevertheless, we use these data to determine how Black women and other women of color portray their motherhood status in messages that appeal to a broad audience. These data can give us a descriptive picture of differences in how Black women use motherhood in campaign messages.
Campaign ads aired by women of color, by year.
To identify ads that mention a candidate’s status as a mother, or a father, we transcribed each of the just over 28,000 campaign ads in the WMP datasets and provided the transcripts to two independent coders who recorded mentions of a candidate’s own family. We identified ads with motherhood appeals based on whether a candidate’s maternal or paternal status was explicitly disclosed with statements such as “I’m a mother” or “I’m a dad.” We also looked at whether ads clearly discussed the candidate’s children, such as “My mom/dad is great.” We excluded ads that mentioned other members of a candidate’s family, such as discussions about grandparents or siblings, from our parental status variable. We also excluded ads that discussed family or children in general terms, such as “families are struggling.” We provided more information on our coding system in the appendix. Our analyses use each unique ad as the unit of analysis.
To ensure a high level of coding reliability, 10% of the campaign ads were read and analyzed by a second coder blind to the purpose of the study. The intercoder reliability reaches a threshold level that indicates a high level of reliability: greater than or equal to 0.70 using Krippendorf’s alpha (O’Connor and Joffe Reference O’Connor and Joffe2020).
Our central outcome variable is a parental status variable coded as 1 if a candidate highlighted their parental status, and 0 otherwise. We term this variable a parental status variable as we coded for mentions of fatherhood for men along with mentions of motherhood for women.
Examples of ads using strategic motherhood include one from Senator Kelly Ayotte’s 2016 reelection campaign in New Hampshire, which featured her daughter saying the following:
My mom and I love shooting hoops. You know what? She can still learn a few things from me. Just like I’ve learned a lot from her. Mom helps with homework, she helped dad start his business, and she even fought to put bad guys in jail. Now, mom helps make laws that help people, especially when they need it most. I’m really proud of her. And she’s taught me that with hard work, you can do anything. I’m Kelly Ayotte and I’m Kate and we approve this message.
This ad from Senator Ayotte is classified as a motherhood message because the ad explicitly identifies Ayotte as a mother and includes her daughter discussing how Senator Ayotte performs her role as a mother.
Another example of a motherhood message is from Deb Haaland’s successful campaign for the House seat in New Mexico in 2018 where she said the following:
We’ve all faced struggles. Overcoming them made us fierce. I’m Deb Haaland, and I pushed through law school as a single mom to become a champion for our schools and our kids. I’m 30 years sober, and I know affordable healthcare that covers everyone is fundamental, and I will move us toward a clean energy future because when we build opportunity, we all rise. I approve this message because it’s a new day in New Mexico. Are you ready?
Haaland won her election and became one of the first Native American women ever elected to the US Congress in 2018,Footnote 2 and in 2021, she was appointed Secretary of the Interior, again as the first Native American person ever to hold that office.
We added a number of control variables to our data. We recorded parental status data for all the candidates in our dataset who ran in a general election for the House, Senate, or a governor’s seat. We found this data by looking for public disclosures of the candidate’s parental status through candidate run campaign websites, through official government websites for candidates who won, or through sources such as Ballotpedia that provide objective factual information about candidates. We used this variable to ascertain whether Black women mothers are more likely to mention motherhood are more or less likely to disclose this status in an ad compared to white women or Black men.
We obtained the candidate race and ethnicity data from congressional biographies, archived campaign websites, the database maintained by Who Leads America, minority caucus memberships, and other publicly available data about candidates. We created separate variables indicating whether a candidate identified as Black/African American, Latino, Asian/Pacific Islander, or another racial or ethnic minority group, including Native American. We also created an overall minority candidate variable in addition to the dichotomous variables for each specific race and/or ethnicity.
We also recorded whether the candidate was an incumbent, the candidate party, if the race was for an open seat, the level of competitiveness using the Cook Political Report ratings, the gender of an opponent, and whether a candidate ran in a majority-minority district. Our final set of variables records characteristics of campaign ads, including whether the candidate physically appeared in an ad, the logged cost of advertising in a media market (though this variable is not available in 2010), the tone of the ad (positive, attack, contrast), the length of an ad measured in seconds, and who sponsored the ad. We control for whether the ad was a general election ad or a primary election ad. We cluster all our models at the media market level to account for the unique characteristics of each political campaign that our set of controls might not capture (Banda and Windett Reference Banda and Windett2021; Kang et al. Reference Kang, Fowler, Franz and Ridout2018; Santia and Bauer Reference Santia and Bauer2023).Footnote 3
Motherhood in Campaign Ads
We start by examining differences across gender and based on a candidate’s minority status with logistic regression models that include an interaction between candidate gender and whether a candidate belonged to a racial or ethnic minority group. Our outcome variable is always the family mentions variable. We include our full set of controls to account for characteristics of the campaign ads, media markets, candidates, and their electoral contest. We cluster the errors by media market to account for unobserved heterogeneity across markets that affect whether a candidate airs an ad at all. Our data include just one observation for each unique ad, N = 27, 467.
Figure 2 displays the predicted probability of a candidate referencing their paternal status based on candidate gender and a candidate’s status as a racial or ethnic minority (See Appendix 1, Table A3 for full models). We find that minority women have a probability of 0.0299 (SE = 0.001) of mentioning motherhood in an ad, whereas white women have a 0.054 (SE = 0.006) probability of using the same strategy. These values significantly differ from one another, p = 0.022. White women more likely to mention motherhood in their campaign ads relative to women of color. This avoidance of motherhood mentions in ads from women of color reflects our expectation that women of color have incentives to avoid discussing their status as mothers. Looking across gender but within race, we find no differences in women’s and men’s use, based on their race/ethnicity, in the use of appeals to motherhood or fatherhood.
Predicted probabilities of candidate mentioning parental status in campaign ads.
Note: 95% confidence intervals included.
Most women of color who run for political office do so as Democrats. We estimated models with an interaction between candidate gender, race/ethnicity, and Democratic partisanship. Nevertheless, the number of Republican women of color, especially Republican Latinas and Republican Asian women, is slowly increasing. These results are included in Table A3, column 2, in the appendices. We do not find a significant effect for our interaction term. We estimated differences across party but within a woman’s race/ethnicity to see if Democratic women of color mention motherhood more than Republican women of color, and we find no significant effect, p = 0.169. Of course, the majority of women of color who run for political office are Democrats, and these comparisons should be taken with a note of caution.
We also estimated a series of models isolated to just women, and we included each of our specific race and ethnicity variables for the candidates (white women are the excluded category for these models), as opposed to an overall minority candidate variable. Figure 3 displays the predicted probabilities of candidate family references based on just the race/ethnicity of the woman (see Appendix 1, Table A3 for full models). Black women air ads that reference motherhood at a rate much higher than Latina and AAPI women, with a predicted probability of 0.070 (SE = 0.035). We find that the women in our “other minority” category have the highest probability of airing an ad that mentions family, with a predicted probability of 0.143 (SE = 0.061). This result, although striking, should be viewed cautiously as it included a very small number of women. There are four women in this “other” category who also ran campaign ads, and all these women ran in 2018: Deb Haaland, Sharice Davids, Rashida Tlaib, and Ilhan Omar. Most of the motherhood appeals come from Deb Haaland, who made frequent references to her experiences as a single mother in her campaign messages. We find that Latinas, predicted probability of 0.012 (SE = 0.007), and AAPI women, predicted probability of 0.010 (SE = 0.009), are the least likely to use motherhood appeals. Not unexpectedly given emphasis on traditional gender roles among Republican women (Deckman Reference Deckman2016; Wineinger Reference Wineinger2022), white women, compared to other women, have a 0.065 predicted probability of airing an ad that mentions family (SE = 0.005).
Motherhood references based on women’s race/ethnicity.
Note: 95% confidence intervals included.
Overall, we find that women of color are significantly less likely to use motherhood appeals relative to white women, and this does not fit with H1. However, our analyses focusing on just women show that Black women, in fact, use motherhood in messages more than Latina and AAPI women, and they are about just as likely to air motherhood messages than white women. The next section homes in more on how voters respond to messages about motherhood specifically based on candidate gender and race using a survey experiment.
Study 2: Responses to Motherhood Appeals
To track how individuals respond to motherhood appeals, we use a survey experiment to test H2 and H3. We rely on an experiment because the method offers the control needed to isolate how legislator gender affects the way voters respond to these strategies (McDermott Reference McDermott2002). Second, although we have a robust dataset on the use of familial appeals in candidate television ads, we do not have data on whether individuals saw those ads. For our study, the appropriate observational data does not exist, and in the absence of observational data, experiments are the only appropriate method (Morton and Williams Reference Morton and Williams2010).Footnote 4 We conducted our experiment on March 22, 2022, following a period with heightened attention to the role Black women play in the Democratic Party and in preserving democracy more broadly. This context gives our experiment increased ecological validity as it roughly mirrors public discussions around Black women, motherhood, and mothering more generally that took place in public discourse.
We conducted a 2x2 experiment manipulating the race of the candidate (a Black woman or a white woman) and whether they made a motherhood appeal or not. We embedded the manipulation in a short news clip about a candidate, Sandra Jones, running for an open House Seat. The motherhood treatment is below:
This past weekend the race for the open House seat continued as candidates gave speeches at a local rally. Sandra Jones spoke at the rally about how her experiences as a mother shaped her decision to run for office. Jones stated to the crowd: “As a mom, I know what it’s like to struggle. Like you, I worry about how safe it is for my children, about health care, and the economy.” Voters head to the polls on Tuesday to select a candidate.
The no motherhood treatment removes the mentions of the candidate’s status as a mother (both treatments are in Appendix 2).Footnote 5 The direct references to motherhood in the vignette come from our campaign advertising analysis. Anecdotally, the most common way that candidates reference their parenthood status is through the “As a mom” framework, and we aimed to emulate that structure here. We manipulated candidate race with a photo of the candidates (see Appendix 2). We chose a photo of women who were somewhat younger, less than 60 years old, so that voters would think of them as mothers rather than grandmothers. We chose a Black woman photo with a woman wearing her hair naturally in curls as opposed to chemically straightened to fit with how younger Black women are more likely to portray themselves in political campaigns (Brown and Lemi Reference Brown and Lemi2021).
We conducted our sample via Prolific, a survey platform where respondents are paid a nominal fee for participating in the study. We oversampled minority participants in our recruitment to track differences in how minority participants might respond to motherhood appeals made by a member of their racial/ethnic community relative to a person who does not belong to a racial or ethnic minority group (Campi and Junn Reference Campi and Junn2019). Our total sample size is N = 867 with n = 409 respondents identifying as Latino, Black, Asian/Pacific Islander, Native American, or another minority group, and n = 458 non-Hispanic whites.Footnote 6 Table A4 compares our full sample to the US Census. We also conducted a randomization check with a multinomial logit model to ensure that participants were randomly assigned into the conditions; see Appendix 3, Table A5. Table 1 breaks down our sample size by participant race/ethnicity for each of our four experimental conditions.
Experimental conditions and sample breakdown
We asked a series of questions of our study participants to assess how motherhood appeals can affect candidate evaluations. First, it is possible that motherhood appeals will reduce candidate ratings along qualities associated with political leadership and masculine stereotypes. To assess this affect, we asked the extent to which participants see the candidates as “strong leaders.” Second, it is possible that candidates making motherhood appeals will receive a boost on qualities that reflect communality, and here, we asked participants to rate how well the phrase “cares about people like me” described the candidate. For strong leadership and “cares about people like me,” the response options are “very well,” “somewhat well,” “somewhat unwell,” and “not very well at all.” Finally, to assess overall candidate support, we asked how likely a participant was to support the candidate with a vote likelihood question with response options ranging from “very likely to support,” “somewhat likely to support,” “somewhat unlikely to support,” and “very unlikely to support” the candidate. We deliberately excluded a “neither” option as this is likely to cue social desirability biases (Krupnikov, Piston, and Bauer Reference Krupnikov, Piston and Bauer2016). We recoded all the response options to range from 0 to 1; higher values indicate more positive evaluations for the candidate.
We start by conducting a series of two-tailed t-test comparisons.Footnote 7 First, we compare how each candidate’s evaluation changes from the treatment to the control group. Then, we compare differences across the motherhood treatment condition to assess how candidate race affects voter responses to motherhood messages. These first sets of comparisons focus on H2. Following these analyses, we then turn to assessing differences across participant race and ethnicity to test how the in-group relationship between a voter and a candidate affects responses to motherhood appeals to test H3.
Figure 4 displays the mean candidate evaluation for each experimental condition on strong leadership, “cares about people like me”, and vote likelihood (full means and comparisons in Appendix 3). Each variable ranges from 0 to 1; higher values indicate more positive evaluations. For the white woman, her evaluations do not significantly change from the treatment to the control condition on all three outcomes. For the Black woman, the difference between the motherhood condition and the control condition is more nuanced. The Black woman receives slightly worse evaluations in the motherhood condition relative to her control condition rating, consistent with H2. On strong leadership, her evaluation drops by 0.04 (SE = 0.02) p = 0.0643, or 4%, but on “cares about people like me” and vote choice, the difference from the control condition is insignificant (p = 0.2394 on “cares about” and p = 0.7159 on vote choice). These initial comparisons suggest that emphasizing motherhood may undercut Black women on qualities that match masculine stereotypes, but also Black women get no boost on qualities that align with feminine stereotypes with motherhood appeals.
Mean candidate evaluations across experimental conditions.
Note: 95% confidence intervals included.
Next, we compared directly across candidate race to see how participants responded to motherhood appeals. We compared the Black woman to the white woman in the control condition and the Black woman to the white woman in the motherhood condition. On strong leadership, the Black women receives more positive evaluations relative to the white women in both the treatment, an increase of 0.035, or 3.5%, (SE = 0.02), p = 00982, and in the control condition, an increase of 0.089, or 8.9%, (SE=0.02), p < 0.001. On “cares about people like me,” the pattern replicates. The Black woman receives an evaluation that is 0.17 points, or 17%, (SE = 0.03), higher than the white woman’s rating in the control condition. In the motherhood treatment, the Black woman’s rating is 0.097, or 9.7%, (SE = 0.03), higher than the white woman’s rating, p = 0.002. And the results are the same on vote choice. The Black woman’s level of vote support is higher than the white woman’s level in both the treatment and the control condition. These results suggest that contrary to H2, which predicted overall more negative evaluations, voters may be more supportive of Black women regardless of whether they make a strategic appeal to motherhood or not, especially when paired with the earlier finding that the Black women did slightly worse on strong leadership with a motherhood appeal relative to her control evaluation.
To test H3, which anticipates a more positive ingroup evaluation, we compare how participant race/ethnicity affected evaluations to the Black woman and white woman in the motherhood treatment condition. Figure 5 displays differences in the Black and white woman in the motherhood condition across participant race on all three of our key outcomes (full comparisons across all the groups are in Appendix 3).
Responses to motherhood appeals based on participant race.
Note: 95% confidence interval included.
On strong leadership, there are no differences across participant race/ethnicity in the ratings of the Black woman and the white woman in the motherhood conditions. On “cares about people like me,” there are, again, no differences in the Black woman’s rating in the motherhood condition comparing across participant race/ethnicity. For the white woman, white participants give her a more positive rating in the motherhood condition relative to minority participants by 0.11 points higher, or 11%, (SE = 0.04), p = 0.0054. And on vote likelihood, white participants report higher levels of support for the white woman relative to minority participants, by about 0.086 points higher, or 8.6%, (SE = 0.028, p = 0.0025). Our findings suggest that these patterns are due to the preferences of minority participants in our study. Table A7 in our appendix shows that white participants gave the Black and white women equal ratings, but minority participants gave the Black woman significantly higher ratings. This preference among minority participants for the Black woman candidate underscores the value that minority communities place on having representation with someone who shares their race/ethnicity (Campi Reference Campi and Junn2019). These equitable levels of support among white participants are important, not just for representation, but they provide a check against the possibility that social desirability pressures affect our responses. If participants felt pressured to report a higher level of support for the Black woman based on norms against racial prejudice, then we should see the Black woman receiving even higher levels of support from white participants.
We conducted several robustness checks with our experimental data based on characteristics of our study participants that can affect how they respond to messages from Black women and messages about motherhood. These results are in Appendix 3, Tables A8 and A9. First, we tested how participant gender affects responses to Black women and motherhood messages. We find, overall, that women rate the Black woman motherhood more positively than men on all three of our key outcomes; there is no difference across participant gender in the ratings of the white woman motherhood condition; and women rate the Black woman mother more positively than the white woman mother, but there are no differences in how men rate the Black and white women who are mothers. We also tested how parenthood status affects ratings of mothers. Only 32% of our sample indicated they had children under 18 living at home. This small subset of our sample makes it difficult to conduct t-tests as we have low power for parents, and unbalanced conditions. We estimated a series of regression models interacting our experimental conditions with the parenthood status of participants, and we do not find any significant effects with the three-way interaction. We also broke our experimental sample down by participant ideology to see if more liberal, and likely more Democratic voters, express a stronger preference for the Black woman motherhood appeal relative to the white woman motherhood appeal. We include a full set of comparisons broken down by participant ideology in Appendix 3, Table A10. We find that more liberal voters express a stronger preference for the Black woman mother relative to the white woman mother. We also conducted comparisons for those identifying as conservative or moderate, though we must issue a note of caution in interpreting these results as we have very low statistical power given that 61% of our sample identified as liberal. We find no differences in any of the experimental conditions for participants identifying as moderate, and we find that more conservative participants prefer the white woman motherhood appeal over the Black woman motherhood appeal. These patterns suggest that women, as a group, might be the group most responsive to messages about motherhood from Black women.
Results Summary
We apply our intersectional motherhood framework with two empirical tests. First, we examined how, in the United States, Black women use maternal appeals in their campaign messages. We found that women, regardless of their race, reference their families in campaign ads more than men, but Black women and white women are just as likely to use motherhood messages. These null effects do not fit with our prediction in H1. Second, we conducted an original experiment manipulating a woman’s race and whether she references motherhood in her campaign appeal. Our results suggest that Black women may receive lower ratings on masculine traits when they reference motherhood compared to the control condition, offering some support for H2. Third, we find that the race of the participant does not affect the Black woman’s evaluations in the motherhood condition, suggesting no support for H3.
Discussion/Conclusion
According to the Center for American Women in Politics (CAWP), a record number of Black women ran for and were elected to both national and state office in 2022 (Center for American Women in Politics 2023). More Black women and Black mothers than ever before are presenting themselves to voters, forcing Americans to contend with intersectional sets of stereotypes that have been relatively rare in the past — including the parental status of these distinctive candidates. Our results provide unique insight into the use of motherhood appeals in US-based campaign messages and American voter responses to such appeals. We find that Black women are just as likely as white women to incorporate motherhood into a campaign message and that Black women candidates can receive more positive evaluations relative to white women when using a motherhood appeal. Our study has critical implications for the campaign strategies Black women use to advance their electoral prospects. With Black women running for House, Senate, and gubernatorial seats in upcoming election cycles, these results speak to the ways voters may respond to these messages. We anticipate that there are contexts outside the United States where women of color are similarly marginalized but could also benefit from strategic use of motherhood appeals. Regardless, this would be a fruitful area of exploration for future research.
We examine Black motherhood by operationalizing motherhood based on whether a candidate had children or not. Mothering appeals, however, can be made by any candidate regardless of their parental status or gender. For example, the use of communal traits in campaign ads, such as describing yourself as caring or visually showing yourself with children, can each evoke ideas about mothering in messages. Past work considers the use of communal traits or communal images in campaign messages by women and how voters respond to such messages (Bauer Reference Bauer2015; Bauer and Carpinella Reference Bauer and Carpinella2018), but this work fails to consider the intersection of race and gender and how such types of messages can evoke mothering. Future work can expand on the research we have done here to adopt a more comprehensive approach to mothering appeals as opposed to explicit mentions of a candidate’s status as a mother.
One aspect unexamined in our experimental approach is that we only investigated voter responses to motherhood as opposed to mothering appeals for Black women, and voters may respond to the intersection of race or ethnicity and motherhood differently for other women of color. Our advertising results showed that Latinas may be less likely to use these messages compared to women of other racial and ethnic groups, but our results, at this point, cannot speak to whether Latinas, Asian American women, or other women of color will be helped or hurt by using these appeals in their campaign messages. Expanding studies of intersectionality and motherhood to vary voter responses to women of different racial and ethnic backgrounds is an important next step.
In 2022, Erin Maye Quade was not the only candidate to display her motherhood while campaigning. Certain electoral contexts can create conditions that increase or decrease the use of motherhood across election years and across voter responses. We did not examine how the electoral context gives rise to the use of motherhood messages and how voters process such information. A campaign climate that is more “women friendly” such as the 2018 “pink wave” election or the 1992 “Year of the Women” in the United States can shift how voters respond to motherhood messages. In short, there may be specific political contexts that lead voters to value motherhood appeals more than others. In times of unrest and uncertainty, voters may want to see comforting messages from candidates, and motherhood appeals are one way to accomplish this goal.
Electing women who are mothers to political office has ramifications for substantive representation. Bryant and Hellwege (Reference Bryant and Hellwege2018) find that women who are mothers or become mothers while in the US Congress are more likely to advocate for policies that benefit children and mothers as a group. The number of women who have young children while serving in the federal legislature is fairly sparse, as women tend to begin their political careers after their children are in high school and no longer quite so dependent on their mothers (Carroll and Sanbonmatsu Reference Carroll and Sanbonmatsu2013). Electing women from a diverse set of backgrounds, including younger women and women with young children, can shift the policy debates and outcomes in legislatures and produce greater substantive representation for women, children, and families.
Emphasizing motherhood in campaign messages has broader effects beyond whether a candidate wins political office. Although our focus has been parental status, stereotypes around mothering potentially offer advantages for women of color who emphasize communal attributes of caring for children (and others). Highlighting a distinctly feminine role as a benefit to holding political office may start to shift and reframe the masculine associations that voters have for political candidates. The increasing use of motherhood appeals among a diverse set of women may break down the masculine ways that voters evaluate candidates to change how voters think about political leaders. Doing away with notions of political leadership as a masculine pursuit can open more pathways for women’s political success in the future in the United States and beyond.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/S1743923X24000059.
Acknowledgments
The authors wish to thank Julia Hellwege and Lisa Bryant for sharing data and the LSU Gender and Politics Research Lab for assistance in data collection. We would also like to thank Marcruz Osorio whose feedback helped us sharpen the paper.
