Abstract
Latin America has elected more female presidents than any other region in the world, yet dominant theories on campaigning tend to ignore gender. In addressing this lacuna, this article argues that the widespread belief that women are better at mobilizing women means that female candidates tend to invest more significant effort into cultivating a core constituency of women on the basis of gender identity. In contrast, male candidates tend to delegate women-mobilization tasks to female surrogates. An analysis of approximately 1,000 newspaper articles reveals that the “most different” female candidates in Chile and Brazil consistently met with female voters early in their campaigns, evoked gender identities and promised pro-women change. The “most different” male candidates enlisted their wives and female politicians to target women, defend their pro-women promises, and deflect accusations of sexism. The theory illuminates multiple ways in which viable female candidates’ entry into the political arena can improve women's representation.
Introduction
Despite its reputation for machismo and profound gender inequalities, Latin America leads the world in terms of democratically electing female presidents, having done so seven times since 2006. 1 A broader, underlying trend is that women are more often, and more competitively, running for the presidential office. Female candidates from ideologically diverse parties have made viable runs 29 times in all but four Latin American countries (Reyes-Housholder and Thomas 2018). 2 Women are voting at similar, or even higher, rates as men (Desposato and Norrander 2009), and constitute a numerically large and electorally lucrative group. Why and how might female presidential candidates differ from male candidates in targeting of the female constituency?
I thank Jennifer Piscopo and Magda Hinojosa for their comments on earlier versions of this study. A 2015 Fulbright-Hays Doctoral Dissertation Fellowship funded fieldwork in Brazil and Chile. Support was also provided by the Center for Social Conflict and Cohesion Studies (COES)-Universidad Diego Portales through the CONICYT/FONDAP/15130009 grant.
A viable candidacy here means the candidate won at least 15 percent in the first (or only) round of voting.
It is intuitively plausible that female and male candidates will employ divergent strategies vis-à-vis female voters. Yet, dominant theories of Latin American campaigning are gender blind, with the conventional wisdom predicting that female and male candidates from the same party or country will adopt similar strategies (Boas 2016; Dix 1989; Luna 2014; Roberts 2015). Research from outside the region instead shows that gender stereotypes motivate women running for legislative and sub-national posts to employ innovative mobilization tactics (Herrnson 2003; Herrnson and Lucas 2006; Kahn 1996).
Studies looking at the impact of candidates’ gender identities on campaign strategies have not explored presidential elections in Latin America. Female voters in this region tend to disproportionately support female presidential candidates (Morgan 2015), but no study explains or shows exactly why this might be the case. We know little about what specific kinds of electioneering tactics female candidates may employ to attract female voters, how such tactics might differ from those used by their male competitors, and how these competitors may react.
In addressing this lacuna, this article introduces a theory of gender-strategic mobilization, which posits a prevalent belief that women are more effective than men at mobilizing female voters. I argue that this belief powerfully shapes the extent and ways in which presidential candidates target women. Female candidates tend to invest early efforts in targeting a core constituency of women, specifically on the basis of gender identity. Acting on the same belief, male candidates target women less overall on the basis of gender identity. However, when they do so, which is more likely to occur when they are campaigning against a woman, they tend to delegate the task to female surrogates.
I illustrate this theory with a series of comparisons of “most different” female and male candidates. Examining candidates from Chile and Brazil allow me to maximize difference in the conventional predictors of campaign strategies: Party ideology and country culture. Women democratically won the presidency for the first time in 2006 and 2010, respectively. Other women ran unsuccessful, but viable, campaigns during these same elections. I base the analysis on about 1,000 newspaper articles covering these elections, as well as the ones immediately preceding them.
I found that “most different” female candidates strategically sought to cultivate a core constituency of women on the basis of gender identity by employing three common tactics: They met with women early on in their campaigns; evoked gendered experiences as mothers or victims of sexism to establish a shared identity with female voters; and leveraged these shared identities to more credibly promise pro-women reforms. I also found that “most different” male candidates rarely sought to target women on the basis of gender identity, but they seemed to do so more often when competing against viable female candidates. In contrast to their female competitors, male candidates often tried to mobilize female voters by recruiting spouses or prominent women from their parties, who were then responsible for talking about gender issues and deflecting accusations of sexism.
Presidential campaigns are important determinants of the quality of political representation because this is when candidates make policy commitments and citizens more often tune in to political news (Boas 2016). The theory of gender-strategic mobilization has multiple implications for women's representation in Latin America, which is a topic of increasing scholarly interest (Schwindt-Bayer 2010, 2018). It suggests that the entry of viable female candidates, even if they lose, can at least draw more attention to issues affecting this historically marginalized group, such as maternal health, day care for young children, and gender discrimination. Female candidates often seem to spur male candidates to raise the profile of women in their campaigns, possibly positioning them to serve as high-level appointees.
Dominant theories on campaigning in Latin America tend to focus on party systems and electioneering cultures. Classic theories from the latter category argue that historically salient social cleavages determine which groups’ parties will target and which campaign promises they will make (Dix 1989; Lipset and Rokkan 1967). According to these theories, candidates make few attempts to mobilize women on the basis of gender identity because, unlike class and religion, gender is not a cleavage around which parties traditionally have organized themselves (Roberts 2015). 3 Luna's (2014) argument that parties make segmented appeals to attract multiple social classes further extends the gender-blind partisan perspective. These theories expect that male and female candidates from the same parties will use similar tactics to target similar groups.
Some women's parties have formed in Latin America, but they have never been electorally viable.
A second line of research from Latin America analyzes electioneering cultures rather than parties. Boas (2016) argues that presidential candidates’ strategies will converge on the demonstrated success of previous presidential campaigns in the same country, regardless of party and ideology. According to contagion theory, the most important differences in campaign appeals and styles will manifest themselves cross-nationally, and female presidential candidates will adopt the same strategies as those of winning campaigns led by males.
Historical research on Latin America does suggest that gender has influenced how male leaders discursively shore up support. Kampwirth (2010) shows how male populists have leveraged shared understandings of masculinity and femininity to obtain power. Other research demonstrates that organized women and female politicians have used maternal-ism to broadly justify their political participation (Franceschet, Piscopo, and Thomas 2016; Miller 1991). However, this research has not explained how and why female presidential candidates might try to leverage their identities as mothers to specifically galvanize female voters and how male candidates might respond to this tactic.
Research conducted outside the region has shown that whether candidates are female appears to affect the types of issues they emphasize, the negativity of their tone, and how they portray themselves (Kahn 1996). Some follow-up studies conclude that female and male candidates in the U.S. differ in terms of their issue focus, tone and style (Herrnson and Lucas 2006; Panagopoulos 2004), but others have yielded null results (Bystrom and Kaid 2002; Dolan 2005). Finally, some scholars maintain that campaigning as women can boost women's success rates (Herrnson 2003). This suggests that female candidates often leverage their gender identity to their electoral advantage.
The scholarship on how candidate gender affects campaign strategies outside of Latin America generally focuses on legislative or sub-national races. Yet, studies of female candidates for chief executive offices worldwide have examined determinants of women's successes and failures, rather than their campaign strategies (Adams 2016; Murray 2010; Thomas and Adams 2010; Valenzuela and Correa 2009). This research emphasizes how double binds, sexist expectations, and the media's pro-male bias often impede women's chances for success.
This article alternatively seeks to understand how presidential candidates’ gender leads to differences in how they seek to mobilize a historically marginalized but electorally lucrative group: women. Results from modeling the Latin American Public Opinion Project data shows that female voters are statistically more likely to say that they voted for a female presidential candidate. Morgan (2015) explains this relationship by suggesting that female voters believe female presidents are more likely to act on their behalf. However, scholars have yet to unpack this black box between presidential candidates’ gender and female support. This study confronts this challenge by digging deep into presidential campaigns to see if women running for president in Latin America target female voters on the basis of gender identity, or if, as conventional theories would predict, female candidates largely campaign in ways similar to their male counterparts from the same party or country.
It is a well-established assumption that viable presidential candidates seek the most efficient, effective way to earn popular support. Literature on constituencies shows that candidates tend to go after potential loyalists or core voters, rather than swing voters, early in their campaigns, (Cox 2010; Cox and McCubbins 1986; Kriner and Reeves 2015). Candidates’ own social characteristics influence which groups they perceive as potential core constituencies (Fenno 1978).
Following this, I posit that political elites in Latin America tend to believe that shared gender identity can generate trust and, eventually, votes. Women are thought to be more effective and efficient than men at mobilizing female voters specifically on the basis of gender identity. For the purposes of this study, strategic mobilization of female voters on the basis of gender identity involves three specific tactics: (1) Meeting with groups of women; (2) evoking women's multiple gendered identities; and (3) promising pro-women change.
Why might political elites believe that female candidates enjoy a comparative advantage over their male counterparts in employing each of these tactics? First, female candidates are more likely than male candidates to have affinities with both unorganized and organized women. Female candidates will likely network more with women, due to the sociological principle of gender homophily (McPherson and Smith-Lovin 1986). Organized women's and feminist groups prefer to see more women elected to political office, particularly the most powerful and historically male-dominated office of the presidency. Therefore, these groups are more likely to support female candidates than their male counterparts.
Concerning the second tactic, any candidate – female or male – can evoke women's gender identities, such as their identities as (potential) mothers or victims of sexism. Nevertheless, female candidates in Latin America almost always can reasonably claim shared gender identity with female voters by pointing to common experiences. For example, only female candidates can convincingly argue to have personally experienced motherhood or sexism, while male candidates almost never can. Female candidates therefore enjoy a credibility advantage in claiming to understand challenges related to gender-specific roles and stereotypes.
Regarding the final tactic, any candidate can promise pro-women change, but female candidates likely possess another credibility advantage. Having establishing shared gender identity, these women can then draw on notions of linked fate, or the idea that an individual's destiny depends in many ways on the destiny of their group. Female voters may trust female candidates more than their male counterparts to defend these interests because (in theory, but usually not in practice) all women would benefit. Therefore, female candidates are more likely to evoke shared identities with female voters and promise pro-women change in an effort to elicit trust and political support.
All of the above reasons sustain the theory's premise that political elites tend to believe that women are better at mobilizing women, specifically on the basis of gender identity. I argue that this belief powerfully shapes female candidates’ strategies vis-à-vis female voters. In seeking the most effective way to galvanize popular support, female candidates are more likely to employ the three tactics specified above to target women on the basis of gender identity.
Candidates who dedicate significant attempts, rather than haphazard ones, to mobilize women on the basis of gender identity generally pursue this strategy by employing all three of these tactics. Timing also is important. Targeting women early on suggests that the candidate views female voters as a potential loyalist base, or core constituency. Candidates who prioritize their female mobilization strategy – who I contend are more likely to be female than male – tend to employ these specific tactics at least several months before the election, or right when they jump into the race.
The same belief concerning female candidates’ comparative advantage vis-à-vis female voters generates theoretical expectations for male candidates’ strategies. Overall, they are less likely to target women on the basis of gender identity, but when they do, they are more likely to use female surrogates to campaign on their behalf. These female surrogates could be the male candidates’ wives, co-partisan female politicians, or other well-known women. Some of the female surrogates’ specific tasks in mobilizing women on the basis of gender identity overlap with those of female candidates. Such tasks include: Meeting with groups of women, making gendered appeals to female voters, and promoting the male candidates’ pro-women promises. At other times, how these surrogates, particularly female campaign managers or spokeswomen, are supposed to attract female voters remains ambiguous. The implication is that a female face is deemed sufficient. Male candidates are also more likely to target women when they are campaigning against a female candidate, sometimes because they fear that the female candidate will siphon off women from their own core constituencies, which may be mobilized on the basis of class or religion. Female surrogates in these situations may also be tasked with criticizing the female candidate.
“Most Different” Case Comparisons
This study seeks to inductively generate the theory of gender-strategic mobilization, rather than definitively test it. To this end, I will leverage a series of purposefully selected case comparisons (Gerring 2007; Tarrow 2010). The primary unit of analysis is the presidential campaign: Either a female or male candidacy beginning about one year before the election. The independent variable of interest is candidates’ identities as men or women, and the dependent variable is the way, and extent to which, candidates target female voters.
A “most different” case study design compares campaigns that vary in theoretically relevant ways, but manifest the same values on the dependent variable (Gerring 2007). Chile, Argentina, Costa Rica, and Brazil all democratically elected female presidents from 2006 to 2010. Of these countries, Chile and Brazil allow the most variation in the independent variable of candidates’ gender, as well as other theoretically salient factors specifically campaign cultures and party ideologies. 4 Beginning with the former, Chilean campaigns tend toward personalism and emotion, with presidential candidates emphasizing direct linkages with voters and mentioning few concrete proposals (Boas 2016). Brazilian campaigns tend to be more technocratic and policy-focused, with candidates portraying themselves as results-oriented executives. Concerning party ideologies, I compared candidates who were “most different” in terms of the social (rather than economic) dimension of ideology, categorized as either progressive or conservative. Social progressives tend to advocate for greater individual freedom in terms of gender, while social conservatives generally promote traditional gender roles.
Elisa Carrió in Argentina ran against Cristian Fernández de Kirchner in 2007, but she maintained a similar center-left ideology as Fernández. Laura Chinchilla ran the only viable female campaign in Costa Rica during this period (Reyes-Housholder and Thomas 2018).
The empirical analysis includes all viable presidential campaigns for those elections that women won for the first time in Chile (2005–2006) and Brazil (2010), as well as the immediately preceding elections. Candidates are considered viable if they were perceived at some point in the campaigns as potentially advancing to the second round. Table 1 lists the presidential campaigns included in this study. Four women and six men ran competitive campaigns in these countries Joaquín Lavín campaigned twice, so he appears twice in the table, making a total of 11 campaigns.
Presidential Campaigns
Winner.
Following “most different” logic, I sought to maximize variation on the variables of culture and ideology, which were not the study's main focus, while holding constant the independent variable of interest, candidate gender. Therefore, I could separately analyze the “most different” female and “most different” male candidates. Table 2 categorizes the same candidates from Table 1 according to the explanatory variables of gender, country, and ideology. Gender was also the main independent variable of interest for Table 2.
“Most Different” Female and Male Candidates
The first two rows of Table 2 show that female candidates within each country ideologically differed. Michelle Bachelet and Dilma Rousseff's social ideologies are classified as progressive. The Socialist Party (PS) and Party for Democracy (PPD) were Bachelet's strongest supporters, and the Workers’ Party (PT) provided Rousseff's strongest support. These socially progressive parties have sometimes challenged traditional norms (Kitschelt et al. 2010).
Soledad Alvear and Marina Silva represent social conservatives, as shown in the right-hand columns. Alvear was backed by the conservative Christian Democrats (DC), which has historically balked at advancing issues important to women's interests (Ríos Tobar 2009). Although the Green Party (PV) is considered progressive on other dimensions, Marina Silva qualifies as a social conservative, primarily because of her affiliation with evangelical churches, which tend to espouse traditional gender roles in Brazil. 5 Therefore, the female candidates paired on the cross-diagonals qualify as “most different”. Bachelet vs. Silva, and Rousseff vs. Alvear vary in terms of ideology and culture, as well as their success at winning the presidency. Finding that these diverse female candidates displayed common female-mobilization strategies will further strengthen my argument's external validity.
Silva has often advocated for greater state intervention, particularly in environmental issues, so might be considered more left-leaning on other ideological dimensions.
The bottom two rows of Table 2 show the viable male candidates selected for this study. Starting with the socially progressive candidates, Ricardo Lagos hailed from the Socialist Party (PS), and Lula led the Worker's Party (PT). Concerning the socially conservative candidates, Lavín represented the far right Independent Democratic Union (UDI) party, which has historically embraced social conservatism. Sebastián Piñera, subscribing to the National Renovation (RN) party, has also expressed socially conservative views. Finally, the ideological heterogeneity of Brazil's Social Democratic Party (PSDB) complicates classifying Geraldo Alckmin and José Serra (Samuels and Zucco 2014). I categorized them as conservatives because both campaigned to maintain the status quo on reproductive rights, rather than liberalizing Brazil's abortion ban.
Therefore, the “most different” male candidates were again located on the cross-diagonals of Table 2's bottom rows. Of these candidates, the socially progressive Lula, who won in Brazil, appears “most different” when paired with either Lavín or Piñera, who both lost in Chile. The socially progressive Lagos who won in Chile is “most different” when paired with either Alckmin or Serra who lost in Brazil. 6
The way these candidates emerged as presidential contenders did not seem to consistently vary along gendered lines. All candidates had significant political experience as ministers, governors, mayors, or legislators. None were outsiders, with the partial exception of Silva, who nevertheless had served as a minister in the Lula administration. Lula was the only candidate running for re-election, and one of the men (Lagos) and two other women (Bachelet, Rousseff) ran as incumbent party candidates.
Although many campaign studies examined advertisements, I used media reports to observe how candidates targeted female voters. Despite their sexist bias, these reports document campaign events, and speech excerpts, as well as interviews with candidates, advisors, and analysts. I specifically examined national newspapers because political analysis, one-on-one interviews, and special reports in newspapers often provide greater in-depth coverage than those from television or radio.
For the Chilean analyses, I examined the first and third most read newspapers, La Tercera and El Mercurio. I did not examine the sensationalist newspaper La Cuarta, which ranks second, as it offers less reliable coverage. The Biblioteca Nacional in Santiago features complete collections of these newspapers in microfilm, paper and electronic archives. I also examined two of the most widely circulated newspapers in Brazil – O Globo and Folha de São Paulo. Representing Brazil's largest cities (Rio de Janeiro and São Paulo), these publications offer the most comprehensive national coverage of presidential campaigns. 7 Proquest houses a complete online database of these newspapers.
Some evidence came from Valor Econômico, another reputable national newspaper that focuses on economics and politics.
After collecting more than 1,000 articles, 8 I then reviewed them in chronological order to understand sequence of events. I was particularly interested in understanding which tactics (if any) candidates used to target women before, during, and at the end of the official campaign season. I looked for any evidence related to the candidates’ electoral strategies, public images, and promises. My focus was not on the effectiveness of the campaigns vis-à-vis women, but rather on the candidates’ specific tactics.
I conducted keyword searches for the period from January to November of 2006 and 2010. Keywords included “mulher*” (woman/women), “campanha,” (campaign) “presiden*” (president/presidential) and “eleitorado feminino” (female electorate), as well as the candidates’ names.
Viable female candidates in Chile and Brazil targeted women on the basis of gender identity early on, and many times, throughout their campaigns. Their specific tactics revealed a similar pattern. First, these female candidates met with groups of women months before the election, suggesting they believed female voters could coalesce into a base of loyalist support, or a core constituency. Second, they tried to establish common identities with women by evoking gender-specific experiences, most prominently motherhood and sexism. Third, they also aimed to earn women's votes by promising pro-women change. Although Bachelet vs. Silva, and Rousseff vs. Alvear constitute the “most different” pairings, I analyzed their campaigns in rough chronological order and by country, beginning with Chile.
Meeting with Female Voters
As Minister of Foreign Affairs, Alvear led the polls until 2004. In July, more than a year before the election, she visited the town of Doñihue to meet with female artisans. An El Mercurio journalist captured the sense of excitement among women waiting for the minister to arrive. “At least in this town, we perceived ‘feminine strength’ (fuerza femenina). Almost spontaneously there emerged among those present a kind of pride for having a woman occupy a ministerial position and moreover for the possibility of having a presidenta,” the journalist wrote. “We are going to vote for a woman, that is clear,” said one of the female town-dwellers (Lagos V. 2004). That same year, Alvear participated in a program for low-income families at a school in Santiago (Marré V. 2004). She appeared at the school's gym, where hundreds of women were waiting to see her, and Chilean news reported that she seemed exceptionally skilled when interacting with women (Lagos V. 2004).
As Health and then Defense Minister, Bachelet eventually became even better known than Alvear for her success in rallying women behind her presidential candidacy (Ríos Tobar 2008). Similar to Alvear, she began targeting female voters more than a year before the official campaign season. In April 2004, she met with Puente de Chile Solidario program beneficiaries, many of whom were single mothers (Marré V. 2004). In visiting the El Arenal sector, she sought to build a loyalist base of mostly unorganized female voters.
Both Silva's and Rousseff's efforts to target women in Brazil early in the race were notable, since women historically were not core constituencies of the PV or the PT. Eager to become a more electorally competitive party, Silva's team was enthusiastic about her perceived potential among female voters. “Marina has good access to poor women, a segment which the PV has never dreamed of getting close to,” said Alfredo Sirkis, general coordinator of her pre-campaign. “They identify with her” (Folha De S. Paulo 2010a). Polls suggesting the ability of this former senator and minister to grow among female voters reportedly motivated her efforts to reach out to women early in 2010. She met with Green Party women and female leaders from the Federation of Rural Workers of the State of São Paulo in March 2010, several months before the official campaign season (Delgado 2010a; Ribeiro 2010; Scolese 2010). She scheduled another event with female rural workers the following month (Delgado 2010b).
As Lula's chief of staff, Rousseff similarly sought to energize and broaden her party's female base by interacting with female union workers in São Paulo, a historic stronghold of the Workers’ Party (Delgado 2010c). Rousseff and Lula opened the Second Women's Metalworker Congress of ABC together at the end of March 2010. She clearly sought to generate enthusiasm among women from the left by personally meeting with them early in the electoral contest.
Evoking Gender Identities
In addition to interacting with female voters, female candidates repeatedly evoked women's gender identities, particularly their identities as (potential) mothers and victims of sexism. Both candidates in Chile highlighted their identities as mothers to cultivate their support more than a year before the election. Alvear spoke in detail about her children to directly relate to female artisans in Doñihue.
I am always connected. I know exactly the day that one has a test or a job interview […] Gutenberg is very well, in the fourth year of civil engineering at the Universidad Católica. He is a spectacular son, supportive and a very good brother. Maybe one of the strengths that is generated with a working mother is the link among the siblings. (Lagos V. 2004)
Alvear added that she would not become a grandmother soon because her oldest married daughter was finishing her PhD. “There are so many hours of studying, and maternity is not yet in her plans. Thanks to email, we are in contact various times a day” (Lagos V. 2004). Alvear therefore seemed to strategically discuss her life as a working mother to establish shared identity with the female artisans with whom she was personally interacting.
While Alvear sought to come off as easily balancing her professional and family responsibilities as a married mother, Bachelet candidly related her struggles as a divorced mother, which enabled her to relate to other single working mothers:
I spend very little time with her and with my two older children, who are 20 and 25. I suppose I'm calm about it, because I feel that they've been educated on solid bases. With my youngest girl it hurts when she gets angry with me for how little we see each other, and when I get the impression that she doesn't need me, that she's learned to live without me. That's something I hate. The help that my mother's provided has been fundamental. When one works and is separated, the presence of another adult is crucial. […] I will try to protect my loved ones, so that they aren't hurt because their mother is a public figure. (The Santiago Times 2004)
Although Rousseff was also a single mother, she took an alternative, sometimes more figurative approach in leveraging her maternal identity. She vowed to care for the Brazilian people and use a maternal leadership style to get results (faz cobranças). “I am like a mother,” Rousseff said.
Mothers order (their children) to brush their teeth or do household chores. In the government, I was a kind of mother. I enforced deadlines but I gave support […] I am firm, but I take care. I protect. (Scolese and Rocha 2010)
Folha interpreted these remarks as strategically targeting female voters.
Rousseff also seized opportunities, such as holidays, to symbolically evoke her motherhood. Surrounded by women, she broached the problem of crack cocaine in a television appearance on Mother's Day. This issue was thought to interest mothers because it especially affected Brazilian youth (Fraga 2010). Rousseff claimed to share the concerns of many mothers. “We are going to conquer this fight, and we mothers are going to be on the front lines,” she declared (Fraga 2010).
In addition to motherhood, female candidates can establish shared identity with women by celebrating advances in gender equality and relating experiences with sexism. Both Alvear and Bachelet symbolically promoted women's political progress while meeting Chilean citizens during the government's 2004 celebration of International Women's Day (Salinas 2004). The slogan for the celebrations was “And why not? Chile needs women's power (fuerza de la mujer).” On a separate occasion, Bachelet shared an example of gender discrimination with female voters. For some political leaders the fact that she and Alvear were performing well in the polls for the presidency was considered “a traffic accident; something unexpected that the women did not know how to handle” (Marré V. 2004). Retelling a personal story of sexism was a way for her to directly relate to female constituents who may have personally experienced machismo.
Like Alvear and Bachelet, part of Silva's female mobilization strategy involved citing her own struggles against sexism. She discussed in a television interview how she overcame gender prejudice in deciding to leave her rural hometown and study at the university (Delgado 2010b). Another common tactic female candidates in Brazil used was a first-female discursive frame to evoke shared gender identities. Such framing emphasized the possibility of electing a woman to the presidency for the first time in national history, marking a victory for all women. High-profile politicians endorsed Rousseff by drawing on shared gender identity and linked fate (O Globo 2010a). One female PT senator declared: “I believe that every woman in this country is going to feel like the President of the Republic with a woman in the presidency” (Ceolin and Ramalho 2010). In other words, a victory for a first-time presidenta would benefit all Brazilian women.
Female candidates often leveraged women's multiple identities. For example, Bachelet's final comments during a presidential debate hinged on her identities as a middle-class single mother, female doctor, and potential victim of sexism:
The first thing is that I am a woman of the middle class. I am a woman and moreover now I am a single mother and I have worked my whole life. […] I go shopping at the supermarket, I go to leave my daughter at school, that is to say, effectively, I am a professional, I had opportunities to study and that is why I have been able to be a female doctor and have a dignified life […] I know that as a woman, people are going to be looking at me through a magnifying glass to see if I am doing it well or not, and I have a tremendous responsibility, not just with the people who vote for me, with those who believe in me, but also with the women of this country, to demonstrate also that women can do it […]. (Archivo Chile 2005)
Bachelet's remarks began with her status as a woman and then layered on other gendered identities. She talked about her everyday activities as a middle-class mother and her career as a female doctor. She finally acknowledged that, as a female politician, she expected greater scrutiny than her male competitors. This alluded to double standards, which is a classic manifestation of sexism. In short, Bachelet seized a high-stakes opportunity to sell her candidacy to the Chilean public by evoking multiple gendered identities.
As a black, evangelical woman who grew up in abject poverty, Silva also strategically deployed her intersecting identities to elicit the trust and political support of specific subsets of women. Folha reported in March 2010 that the Green Party candidate was ready to target three (often overlapping) demographics: Women, Christians, and the lower classes (Delgado 2010d). Silva's campaign manager confirmed that her candidacy presented potential for growth among poor women (Folha De S. Paulo 2010a). The Green Party reportedly had crafted a strategy to target popular sectors by stressing her impoverished childhood and evangelical faith. Her pre-campaign coordinators remarked that these sectors had been difficult to capture in previous elections, but they predicted that Silva could perform better among these groups thanks to commonly held social identities.
Silva's religious affiliation thereby motivated and shaped her targeting of not only low-income women, but also socially conservative women (O Globo 2010b). In August 2010, campaign coordinators reportedly hoped that their candidate would lure away support from poor women who seemed to back Rousseff or Serra. Silva and her team designed television advertisements featuring her biography as an impoverished black woman from the Northern state of Acre (Roxo 2010). According to O Globo, Green Party advisers expected Silva's life story to elicit empathy from lower class women, who had identified with President Lula in the past.
A final way that female candidates leverage their shared identities with female voters is by emphasizing positively valued qualities that they claim universally characterize women, such as their capacity to serve as inclusive, consensus-seeking leaders. 9 This tactic of affirmatively essentializing women's identities, or claiming that all women possess similar traits, serves at least three distinct purposes. First, it unites women behind a shared, favorable identity. Second, it reinforces commonalities between the female candidate and female constituents. Third, it helps justify greater participation for women in the historically male-dominated executive branch.
This study is agnostic as to whether women possess such qualities, but rather focuses on the discursive use of these claims.
One example of this positive essentialization for electoral gain occurred during a high-profile Sabatina Folha interview in June 2010. Silva was asked about the significance of a female president potentially governing the country. She responded: “Women are more inclusive. They have more negotiation ability. They tend much more to consensus than to disputes” (Folha De S. Paulo 2010b). Generalizations such as these assume that all women possess a comparative advantage in leadership and were designed to unite women behind Silva's candidacy. At the same time, these generalizations give more reasons to elect a female president, rather than a male one. Rousseff deployed a similar tactic at an event to galvanize female metallurgical workers. She emphasized women's leadership competence, affirming that women are “capable of making decisions, of directing, of being good leaders and constructing an environment of understanding and comprehension” (Delgado 2010c).
A final way that female presidential candidates in Chile and Brazil attempted to mobilize women on the basis of gender identity is by promising change benefitting women. In Chile, early childhood education was one of the most prominent pro-women promises in Bachelet's platform: “We will implement a system of childhood protection destined to equal opportunities for the development of Chilean children in their first eight years of life” (Bachelet 2005: 13). 10 She specifically pledged to “guarantee kindergarten access to all children.” Preschools would enable mothers with small children to work outside the home, thereby targeting this subset of women.
Due to low polling numbers, Alvear dropped out of the race before her coalition's primaries. She never widely distributed a formal platform, as candidates generally do so toward the middle and end of the official campaign season. Therefore, I found very little documentation of Alvear's campaign pledges.
Similarly catering to (potential) mothers in Brazil, Rousseff's signature pro-women proposals focused on pregnant women's health and infant care (Suwwan 2010). She pledged to expand public services for these women and their babies by establishing specialized clinics, new maternities, intensive care services, and emergency ambulances. “We are going to articulate this network to the SAMU-Stork so that women do not stay here, having children in the middle of the street,” she announced.
In addition to that, the SAMU (emergency ambulances) with mini-intensive care babies, so that children that are born and run risks in the first months of life can be transported with security. (Suwwan 2010)
She also promised to improve state childcare. “We have a very important thing in the PAC 2 for women and their children. We are going to construct crèches” (Delgado 2010c). She then defended proposals to extend maternity leaves.
Female candidates, who have emphasized their experiences as impoverished women, rather than as mothers, may make pro-women promises that target this subgroup of women. For example, Silva promised to improve existing programs for low-income women. “The housing deficit is very serious,” she said during a presidential debate.
In the last 16 years, investment for the poorest families left something to be desired. The program Minha Casa Minha Vida is very good, but it does not reach the population that earns up to two minimum wages. We are going to maintain this program and give it a new quality, so that it reaches even poor women. (O Globo 2010c)
Men running for the presidency in Chile and Brazil from 2000–2010 targeted female voters less overall, and when they did, they employed different strategies than those of their female opponents. I found little evidence of viable male candidates targeting women on the basis of gender identity early on in their campaigns. However, each male candidate used a female surrogate to court female voters, particularly if they campaigned against a viable female candidate. The analysis moreover reveals that their female surrogates were assigned to assigned to three tasks, which overlap with the those of the female candidates described above: meeting with groups of female voters, evoking gender identities in speeches and interviews, and promoting the male candidates’ pro-women promises.
Male candidates campaigning against women often delegated a fourth task to their female surrogates – deflecting accusations of sexism and criticizing the female candidate. Since women are more often viewed as victims of gender discrimination then are men, male candidates sometimes worry that they will be perceived as sexist or taking advantage of gender stereotypes to the detriment of the female candidates. To better react to machismo accusations or pre-empt them, male candidates deploy their female surrogates to defend the male candidates and attack the female opponent. The evidence suggests that female surrogates focus on only one or two of these tasks. As with the analysis of female-led campaigns, I reviewed the male-led campaigns in rough chronological order, beginning with Chile.
Male candidates in the 1999–2000 campaigns employed a female-surrogate strategy to mobilize women during the second round of campaigning to either compensate for their weakness (in the case of Lagos) or to expand one of their core constituencies (in the case of Lavín). Lagos’ team believed that women constituted his Achilles’ heel, so they would be an important group to target (Arriagada and Navia 2005). 11 He responded by naming then-Justice Minister Alvear as his spokesperson just two days after the first round of campaigning.
Alvear's female mobilization tasks were not specified in the Chilean press, possibly suggesting that merely a female face was deemed sufficient.
Former mayor Lavín looked to counterbalance Lagos’ female campaign manager and solidify his core constituency of women by deploying his wife, María Estela de León. She made her first official campaign speech to more than 4,000 women at the Cultural House of Santiago's Ñuñoa neighborhood (El Mercurio 4 1999). She later initiated a parallel campaign with her children, as a way to court female voters (El Mercurio Online 1999).
Facing female competition in the next elections, Lavín seemed even more concerned with maintaining his constituency of low-income women. He appeared to expand his female surrogate strategy in an attempt to counter-balance Bachelet, whose growing popularity among women was viewed as a threat to his female base (Errázuriz L. 2004). 12 One of Lavín's advisors said that naming a female campaign manager would “break with the axis of the 2005 presidential election that it is a matter of gender, that is to say, that they will have to choose between a man and a woman” (Errázuriz L. 2004). After considering other female politicians from the right, 13 Lavín tapped Cristina Bitar as his generalísima (Aránguiz 2005). Analysts noted that one of her assets was that, at the age of 35, she was a young woman, suggesting that these dual identities could help capture both young and female voters (El Mostrador 2005).
Lavín also sought to erode Bachelet's advantage by suggesting that she was, in effect, playing the female card. He commented: “She beats me because she is a woman” (Durán 2005). His advisers believed these statements helped “make explicit that Bachelet's support is artificial and that women back her because of gender” (Durán 2005).
Concepción's mayor, Jacqueline Van Rysselberghe, was considered a potentially good pick in May 2004, since she was recently named vice president of the UDI (Errázuriz L. 2004).
Lavín was not the only male candidate worried about Bachelet's popularity among women. El Mercurio predicted in October 2005 that, regardless of whether Lavín or Piñera advanced to the second round, either candidate would likely use female surrogates since “a female face is essential to compete against Bachelet” (Aravena Bolívar 2005). Billionaire businessman and former senator Piñera focused on competing against Lavín until the first round on 11 December 2005, which eliminated his conservative, intra-coalition rival. 14 Piñera and his team then adjusted their strategy to defeat Bachelet.
Lavín earned 23.2 percent of the vote, while Piñera advanced to the second round with 25.4 percent.
A main plank of this renewed strategy was to counterbalance Bachelet with female surrogates. To this end, Piñera invited RN Deputy Lily Pérez to serve as his campaign manager. She also assumed responsibilities of rallying female voters. Piñera's wife also served as a female surrogate, meeting with groups of women at the end of December 2005 (Campusano S. 2005).
Male candidates competing in the 2006 Brazilian race displayed similar patterns as those of their male counterparts in Chile. Although they rarely targeted women on the basis of gender identity, both Lula and his main opponent Geraldo Alckmin, former governor of São Paulo and PSDB president, used female surrogates to attract female voters in 2006. In October of that year (relatively late in the campaign), Lula's wife, Marisa Letícia and the wife of his vice-presidential pick, Mariza Alencar, led a women's march with about 500 female party members and ministerial workers through a mall in the capital of Brasília (Folha De S. Paulo 2006; O Globo 2006; Lima 2006). Women from Lula's re-election team had organized the event, and another march in the Federal District was scheduled to take place a few days later.
In addition to Letícia and Alencar, Lula employed São Paulo's former mayor and one of his campaign coordinators, Marta Suplicy, to target female voters on his behalf in 2006 (Teodoro 2006). Suplicy dedicated her Oct. 21st agenda to targeting women and people living in the northeastern part of the country. She was scheduled to participate in a march in Santo André in São Paulo's ABC district that morning, along with Letícia and Alencar (Paulo 2006). Finally, Alckmin's wife, “Dona Lu,” played a comparable surrogate role (Bautzer 2006). The PSDB's women's section of Maranhão organized a march in October 2006 through the center of São Luís, where she greeted female voters.
Campaigning against two viable female candidates in 2010, former governor of São Paulo José Serra deployed his wife, Mônica Serra, as his surrogate to counterbalance with Rousseff, who was leading in the polls. “The PSDB's idea is to make Mônica the counterpoint to Dilma – that exploits the fact of being a woman in her campaign,” Folha reported (Guerreiro and do Valle 2010). Mônica Serra met with PSDB women at the beginning of August 2010. She also scheduled time with women's groups in Ceará, Rio Grande do Norte, Paraíba and Pernambuco, and campaigned alongside Fernanda Richa, wife of the PSDB's candidate for Paraná's governorship.
Similar to the female candidates themselves, female surrogates campaigning on behalf of male candidates often evoked shared identities as women to establish trust with female voters. They also sought to promote male candidates’ pro-women reforms. For example, Lavín's wife, María Estela León, “is the typical Chilean woman,” according to one leader in the Lavín camp,
She is a housewife; she has lots of kids; she is strong-willed (tiene empuje), and her physical appearance is typically Chilean. According to our studies, the female sex feels very identified with her. (La Tercera 6 1999a)
Seeking to leverage her perceived comparative advantage, León evoked maternal themes during her first campaign speech in Santiago (El Mercurio 4 1999). A mother of seven, she once was asked to explain Lavín's exceptional performance among female voters:
I believe that women always think about their children. They are always concerned about their ‘puppies’ and what they want is work for their husband, or for themselves to maintain their children, so that they are raised well and have the possibility to go to the doctor. They look over the well-being of their family and they have seen in Joaquín a person that has done many concrete things. Women are not guided by ideologies, but more by the feelings of who gives them the most security. (El Mercurio 4 1999)
León first defined women according to their roles as mothers and wives, and then implied that women were concerned with concrete, everyday issues, a focal point of her husband's campaign. She also characterized women's decision-making as intuitive and guided by their emotions, rather than political ideas. León moreover made essentialist claims concerning women's identities and interests. Given her own identity as a woman, mother and wife, León seemed to believe she possessed the authority to speak on behalf of female voters. Her message: Trust me because I also am a woman.
Female surrogates in Brazil also sought to mobilize female voters by drawing on shared gender identity. For example, at the end of the 2006 Women's March, Mariza Alencar shouted: “I make a special appeal to women miners and also to all Brazilian women” (O Globo 2006). Later that month, Letícia gave her first televised speech during Lula's advertising time. Her appearance ran on the night dedicated to Lula's pro-women achievements, including the creation of the women's ministry (Teodoro 2006). “He has worked so much in favor of Brazilian women,” Letícia said. “And this work is going to grow even more in the next four years. Palavra da mulher (women's word).” Here, Letícia re-appropriated the more common, patriarchal phrase palavra de homem (men's word) to emphasize that women can trust her word because she is a woman.
Because women are thought to be more credible at promising pro-women reforms, male candidates often delegated the task of promoting their pro-women proposals to their female surrogates. Supporting Serra, Serrano defended her party's record on women's issues by pointing to the pro-women achievements during the 1995–2003 administration of co-partisan President Fernando Henrique Cardoso. She cited the creation of the Women's Special Secretary and the fact that Cardoso's conditional cash-transfer programs directed maternal benefits to low-income women (Delgado 2010b). O Globo also reported that Serrano emphasized health issues to cultivate Serra's female vote. “Women feel a lot of security,” she said.
They have a strong recollection of him as Health Minister and the general revolution that he did with the creation of Family Health. We have a very clear notion that health is an issue intimately linked to women who end up being responsible for taking care of the health of the parents, the children and the husband. (O Globo 2010e)
Finally, men vying for the presidency often used female surrogates to defend themselves and attack female opponents. Piñera worried about being perceived as sexist during the 2005–2006 campaigns, and Pérez became campaign spokesperson
so that she be the one to confront Michelle Bachelet and therefore avoid that the criticisms and rebuttals that the Alianza candidate could dedicate would not be seen as an ‘act of machismo‘. (Guerra 2005)
El Mercurio reported that the Alianza was concerned with the Bachelet camp's allegations of sexism “because this could activate a spirit of solidarity among women” (Guerra 2005).
In 2010, Serra enlisted his spouse to attack Rousseff. On the web site PSDB Mobilizes, Mônica Serra accused the female PT candidate of ignoring a controversy over Brazilian intervention in a stoning of an Iranian woman charged with infidelity. “And the woman candidate? She is mute. She doesn't say anything” (Guerreiro and do Valle 2010). In order to downplay the significance of Rousseff's gender identity, Mônica Serra later told reporters: “This is not a gender election. The gender issue is overcome.” Serra's wife also accused the PT candidate of being “in favor of killing little children” because of her falsely alleged pro-choice stance (Cervellini, Giani, and Pavanelli 2011). Serra also used a high-profile actress to attack Rousseff on television in October (O Globo 2010d). In one of Serra's commercials, the actress claimed that Rousseff called upon Lula every time a problem arose. In another pro-Serra advertisement, the actress further criticized the future presidenta by reminding viewers of a scandal involving Rousseff's right-hand female aide.
The recent rise of female presidential candidates brings greater attention to the incompleteness and inadequacies of dominant theories on presidential campaigning in Latin America. They especially fall short of explaining how the entry of viable female candidates affects the extent, and ways in which, candidates compete for women's support. This article systematically explains the mechanisms linking presidential candidates’ gender identities to the extent and ways in which they target female voters. The analyses empirically support the theory of gender-strategic mobilization in showing how “most different” female candidates all used three tactics early on in the race: meeting with groups of women, evoking gendered identities and promising pro-women change. The “most different” male candidates did not tend to prioritize targeting of female voters in these ways, but when they did – often responding to viable female candidates – they tended to delegate the tasks to female surrogates.
Although party factors may still account for many aspects of presidential campaigns, this study shines new light on how women running for president in Chile and Brazil often broke with their parties’ traditional campaign tactics and the groups their parties traditionally targeted. With the exception of the DC, the parties of each of these female candidates tended to perform better among women than men. The PS in Chile, and the PV and PT in Brazil all tended to attract more male than female voters (Hunter 2010; Lewis 2004). More specifically, although Lagos and Bachelet were from the same party in the same country, Bachelet made much earlier, and more consistent, attempts to mobilize women on the basis of gender identity. Lagos attempted this by using a female surrogate only at the very end of his 1999–2000 campaign. Lula and Rousseff were also from the same party in Brazil. Rousseff, however, made earlier and greater attempts than Lula to mobilize women on the basis of gender identity in 2010, which is consistent with this article's theory. Lula used female surrogates, particularly his wife, to mobilize women on his behalf, as did other viable male candidates in Chile and Brazil.
To conclude, the rise of competitive female presidential candidates might be considered an example of how the entry of groups representing historically marginalized interests, such as the emergence of class-based parties (Kitschelt et al. 2010; Roberts 2015), can deepen democracy by diversifying the supply of campaign promises and voices. This study focuses on countries marked by growing dissatisfaction among citizens and perceived failures of the political class in channeling citizen demands. While it is unrealistic to expect female candidates to solve such endemic problems, characteristic of “crises of representation” (Hagopian 2016; Luna 2016), their candidacies could at least marginally help counteract them by enhancing women's political presence, as well as policy-making on behalf of women (that is, women's descriptive and substantive representation). The recent entry of viable female candidates seems to motivate male candidates to raise the profiles of women in their campaigns (thereby potentially enhancing descriptive representation). Some of these women, in turn, could find themselves better positioned to serve in ministerial or other high-level posts if their candidate wins. The emergence of competitive female candidates also seems to move their male competitors to focus more on crafting proposals to target this demographic (thereby potentially improving substantive representation). Therefore, even if viable female candidates ultimately lose a presidential race, other female politicians, as well as female citizens, may reap auxiliary benefits.