What candidate will fight corruption? Gender and anti-corruption stereotypes across European countries

Abstract

What candidates do voters perceive as best to combat corruption? While recent studies suggest that parties recruit women in order to restore legitimacy, we know less about whether voters believe that women candidates are better equipped than male candidates to fight corruption. This study suggests that women mayors are seen as more likely to fight corruption, yet that the credibility of both male and female politicians increases if they are ascribed traits traditionally seen as ‘female,’ including being risk averse or specializing in the provision of welfare services. Leveraging the diverse levels of socio-economic development, corruption, and gender equality across 25 EU member countries, our unique conjoint experiment shows support for these claims. Both women and male candidates benefit from being described as risk averse and prioritizing social welfare issues, while outsider status has no effect. Male candidates, however, have a consistent disadvantage, particularly among women voters. Moreover, the effects of candidate gender are strongest in areas of Europe with the highest levels of political gender equality.

Introduction

Competitive elections, an institutional cornerstone of democracy, should in theory provide the most important means through which citizens can remove corrupt politicians from office and replace them with representatives more inclined to serve the public good. However, public dissatisfaction with democratic institutions and disenchantment with competitive elections can at least partly be traced to the limited supply of candidates that are seen as able to credibly fight corruption (Della Porta and Vannucci, 2007; Persson et al., 2013; Agerberg, 2020; Vera, 2022). Recent research suggests that voters tend to trust women candidates more than their male counterparts and perceive female candidates as less likely to be implicated in corruption scandals (Barnes and Beaulieu, 2014, 2019). Moreover, building on the seminal works of Swamy et al. (2001) and Dollar et al. (2001), a sizable literature has shown a strong association between the share of women in office and lower levels of corruption. This has been attributed not only to women being more honest, but also more risk averse and therefore less likely to engage in corruption if it risks being detected and punished (Esarey and Chirillo, 2013; Esarey and Schwindt-Bayer, 2018). Moreover, scholars also indicate that women are, more often, marginalized outsiders with more limited opportunities to engage in corruption (Goetz, 2007; Bjarnegård, 2013). With few exceptions, however, these theoretical frameworks have mainly been used to explain why women candidates are less likely to engage in or be implicated in corruption scandals rather than explaining the extent to which women are seen as ‘cleaners’ by voters – i.e., as more capable of fighting corruption, post-scandals.

Although political parties and elites sometimes strategically assume that women candidates signal honesty and trustworthiness (see Valdini, 2019), we know far less about the extent to which voters believe that women candidates are better suited to fight corruption. This study addresses this lacuna and investigates if women are seen as ‘cleaners’ by ordinary voters, and if so, why, across diverse political and cultural contexts. We argue that voters believe that women candidates are better suited to fighting corruption compared to male candidates. Further, the credibility of both male and female politicians increases if they are ascribed traits that scholars use to explain the advantage enjoyed by women candidates when it comes to reducing corruption, including being risk averse (Esarey and Schwindt-Bayer, 2018; Barnes and Beaulieu, 2019) or prioritizing the delivery of public services (Bauhr et al., 2019; 2024). In other words, while women candidates are generally seen as more honest and trustworthy than male candidates when it comes to fighting corruption, activating counter stereotypical traits may also improve the credibility of male candidates.

Our research design uses newly collected data from the most recent European Quality of Government Index survey fielded online, which includes 48,080 respondents across 25 EU countries (see Charron et al., 2022), and which, therefore, expands our claims of generalizability in comparison to single-country source data. In order to elucidate causal effects, we embedded a conjoint experiment in which respondents were asked which candidate, based on a list of randomized candidate attributes such as gender, professional background (insider or outsider), personality characteristics (risk taker or cautious), and campaign issues (social welfare issues or infrastructure), along with partisan affiliation, they believed would be best to fight corruption ¹ .

With this unique experimental data, we investigate voter preferences regarding which candidates they perceive to be most likely to combat corruption in a mayoral race in their municipality. First, we investigate if women candidates, overall, have an advantage over male candidates when it comes to combatting corruption, and, if so, the extent to which this perceived advantage travels across Europe. We investigate this as expressed both through candidate preference and expressed abstention. We find that while co-partisanship is far and away the most salient factor in voter choice, there is, nonetheless, a clear and consistent effect of candidate gender, in that the respondents favor female over male candidates on the issue of fighting corruption. In addition, having two male candidates on the ballot significantly increases abstention, implying that none of the male candidates are seen as suitable when it comes to reducing corruption (Birch, 2010; Bauhr and Charron, 2018; Agerberg, 2020).

Second, we investigate whether common theoretical frameworks in gender and corruption research explain the perceived female advantage for reducing corruption in the eyes of voters. We find that respondents prefer candidates that are cautious as opposed to risk-acceptant, and that issue area specialization also matters in candidate choice, but that whether a candidate is an outsider, or a well-connected politician is negligible. Interestingly, while voters prefer candidates assigned these qualities, they do so regardless of the candidate’s gender, and the ‘women’s advantage’ we found is not amplified by being assigned these characteristics. Finally, we elucidate several heterogeneous treatment effects. Our individual level interactions show that the overall preference for women candidates to fight corruption is most pronounced among women respondents. Furthermore, in looking at the effects across the diverse sample of 25 EU member states, cross-level interactions reveal that the gender (female) treatment is strongest in contexts with a higher share of women in office.

Our study thereby seeks to make several contributions. First, we seek to directly investigate if voters prefer women candidates in the aftermath of a corruption scandal. Most previous studies have focused on investigating if and why voters see women as less likely to be directly implicated in a corruption scandal (Eggers et al., 2018; Barnes and Beaulieu, 2019; Wiesehomeier and Verge, 2020). Considering the challenges involved in initiating and implementing anti-corruption reforms (see Della Porta and Vannucci, 2007; Persson et al., 2013; Johnston, 2014; Bauhr, 2017), and that fighting corruption poses different challenges compared to refraining from being implicated in a corruption scandal, investigating candidate qualities that are seen as suitable to this task is of heightened importance. Our study investigates which candidate profiles voters believe will be best equipped to fight political corruption given that it has already occurred. We propose that women candidates signal honesty and are perceived as the most trustworthy option for fighting corruption even when male candidates are assigned traits and characteristics that should help them to be perceived as anti-corruption candidates, and that the female candidate’s advantage when it comes to fighting corruption cannot be explained by only a general preference for women candidates.

Second, this paper also adds new insights into how all candidates, regardless of gender, may benefit from characteristics traditionally associated with women, such as being risk averse (Charness and Gneezy, 2012) or focusing on social welfare issues (Alexander and Andersen, 1993). While studies tend to place particular attention on how women candidates may or may not benefit from masculine traits (Dolan, 2010; Bauer, 2017; Bauer and Santia, 2022), our findings add additional evidence as to how male candidates may also benefit from stereotypically female traits when it comes to fighting corruption. This is important, since voters rely on candidate’s characteristics as heuristics for vote-choice (Popkin, 1991; McDermott, 1998; Webster and Pierce, 2019). In other words, our study adds new insights, both on how traditional explanations for the gender and corruption link resonates with voters in the context of fighting corruption and the extent to which these expectations are gendered.

Despite the prominence of marginalization theory in explaining why women are excluded from participating in corrupt transactions (e.g., Goetz, 2007; Bjarnegård, 2013), and women being perceived as less corrupt (Barnes and Beaulieu, 2019; Wiesehomeier and Verge, 2020), we suggest that outsider status is not necessarily seen as a desirable trait in an effective anti-corruption candidate. Rather, candidates that are cautious and careful and that presumably effectively manage the important risks involved in fighting corruption are seen as more effective. Furthermore, while some previous studies propose that women’s policy priorities matter for reductions in corruption levels (Alexander and Ravlik, 2015; Bauhr et al., 2019; 2024) our study provides unique evidence on voters linking a social welfare policy agenda to effectiveness when it comes to fighting corruption.

Finally, most previous experimental studies on voter preferences regarding corruption build on single-country studies (Incerti, 2020). Candidate choice experiments on gender and corruption are no exception, and most studies to date also make use of single-country cases, such as Spain (Wiesehomeier and Verge, 2020), the UK (Eggers et al., 2018), or the USA (Barnes and Beaulieu, 2014, 2019). Very few studies seek to compare experimental results across countries (but see Batista Pereira, 2021; Le Foulon and Reyes-Housholder, 2021). This means that we, thus far, have very limited knowledge about the extent to which the preference for female candidates is generalizable across a diversity of contexts in terms of socio-economic conditions, levels of corruption, and gender equality. Such geographic diversity allows for testing for heterogeneous treatment effects via cross-level interactions across 25 EU countries.

Corruption, anti-corruption, and voter preference for women candidates

Recent experimental studies find that voters tend to prefer women political candidates, i.e., that there is a ‘pro-woman bias’ among the electorate (Aguilar et al., 2015; Schwarz and Coppock, 2022). ² An adjacent literature suggest that citizens hold gendered stereotypes (Bauer, 2017; Ellemers, 2018; Sweet-Cushman, 2022). Citizens sometimes hold stereotypes about women being more honest, compassionate, and better at dealing with education and health care, or other ‘women’s issues’, and less competent when it comes to issues such as the military, policing, and the economy (e.g., Alexander and Andersen, 1993; Huddy and Terkildsen, 1993; Dolan, 2010). Traditional traits and competences associated with female candidates have also been viewed as incongruent with office holding. However, as the proportion of women in politics has increased, research has started to question whether stereotypes consistently harm female candidates (Dolan, 2017) and suggest that parties and other elites instead seek to consciously exploit stereotypes about women being less corrupt in order to gain electoral advantage and to restore trust in the face of a corruption scandal (Funk et al., 2021; Armstrong et al., 2022).

Some studies focus directly on investigating whether voters perceive women candidates as being less involved in misconduct and corruption scandals (Barnes and Beaulieu, 2014; Eggers et al., 2018; Barnes et al., 2018; Wiesehomeier and Verge, 2020; Batista Pereira, 2021; Le Foulon and Reyes-Housholder, 2021). For instance, Batista Pereira (2021) finds that Mexican voters support women accused of corruption less than their male counterparts, while no gender differences were found in the Brazilian sample. Relatedly, Eggers et al. (2018) find that male voters do not punish corrupt women more harshly in general but that women voters do so. In contrast, Le Foulon and Reyes-Housholder (2021) find that voters prefer allegedly corrupt women candidates over male candidates in Uruguay, but not in the more corrupt contexts of Argentina and Chile. Interestingly, this study does not find that women candidates who have fought corruption have an advantage. Still, studies in the USA confirm that voters expect women in office to be less corrupt (Barnes and Beaulieu, 2019) and that women candidates reduce suspicion of fraud (Barnes and Beaulieu, 2014).

However, most studies to date investigate the extent to which voters support a woman candidate allegedly implicated in a corruption scandal. Fewer studies investigate if women are seen as political cleaners by voters, i.e., that they would also be perceived as fighting corruption levels more broadly, post-hoc. Effectively fighting corruption is, in part, different from refraining from being implicated in corruption scandals, and a candidate may therefore benefit from partly different qualities.

Furthermore, most experimental studies to date build on single-country studies (Barnes and Beaulieu, 2014; Barnes et al., 2018; Eggers et al., 2018; Wiesehomeier and Verge, 2020), which make the findings less comparable across different contexts. Although some recent studies compare data from several countries (e.g., Batista Periera, 2021; Le Foulon and Reyes-Housholder, 2021; Schwindt-Bayer et al., 2018), experimental studies on gender and corruption follow the general approach of the wider field of candidate choice experiments (Incerti, 2020) and are largely single-country based, and sometimes based on convenience samples. This means that we know little about the extent to which stereotypes about women and corruption are generalizable across diverse contexts.

However, there is growing evidence on the link between a larger share of women in office and lower levels of corruption (e.g., Alexander, 2021), and recent evidence that women in office may actually cause such a change (Esarey and Schwindt-Bayer, 2019; Bauhr and Charron, 2021). Furthermore, research indicates that some voters believe that women candidates will be less likely to be implicated in corruption scandals (Barnes and Beaulieu, 2014; 2019; Eggers et al., 2018; Wiesehomeier and Verge, 2020; Batista Pereira, 2021; Le Foulon and Reyes-Housholder, 2021), Drawing together these insights, we hypothesize that voters also see women as being more likely to combat corruption and that this effect may at least partly travel across contexts.

Following the broader literature on corruption and electoral accountability (Davis et al., 2004; Birch, 2010; Sundström and Stockemer, 2015; Bauhr and Charron, 2018), we propose that voter preferences can be expressed in at least two different ways, both as an active choice (voice) and as a more passive abstention (exit) from choosing any of the available candidates (cf. Hirschman, 1970). How citizens express discontent with candidates has wider implications for the functioning of democratic institutions. Ideally, citizens should be able to distinguish between two candidates and decide which candidate they prefer. This would allow for the best available candidate to gain support and hopefully influence outcomes in the preferred direction. Disillusionment with the ability of any of the available candidates to deal with corruption is instead associated with resignation, alienation, and disengagement (Bauhr and Grimes, 2014; Chong et al., 2015).

Much in line with previous work, we thus seek to investigate these two alternative responses, since they both have implications for citizens’ demand for accountability. We hypothesize that candidate gender is used by voters as a consequential heuristic for both of these electoral responses. We expect participants to prefer a female candidate over a male candidate in order to fight corruption, but also that two male candidates on the ballot will make participants more likely to perceive that neither of the available candidates will successfully fight corruption and therefore refrain from choosing any of the available candidates. This leads to the following hypotheses:

H1a) Voters are more likely to select a women candidate compared to a male candidate to combat corruption, ceteris paribus (voice).

H1b) The presence of only male candidates on the ballot will increase abstention (exit) in selecting candidates to combat corruption, ceteris paribus.

How can we understand women candidate advantage? In the following, we depart from our baseline assumption that voters simply see women as more honest and trustworthy when it comes to fighting corruption and investigate the explanatory power of three alternative theoretical frameworks: risk aversion, outsider status, and a social welfare policy profile. These are all well-established and popular explanations for the link between the share of women in office and lower levels of corruption. Drawing on documented gender differences in risk aversion (Bord and O’Connor, 1997; Slovic, 1999; Watson and McNaughton, 2007; Charness and Gneezy, 2012), studies advance that the link between more women in office and lower levels of corruption may at least partly be attributed to women being more risk averse, and therefore less likely to engage in corruption, particularly if there is a risk of corruption being detected and punished (Esarey and Schwindt-Bayer, 2018). ³ Alternatively, women candidates may have fewer opportunities to be implicated in corruption scandals since they are ‘outsiders’, who, on average, are less embedded in the collusive and often male dominated networks where corrupt transactions are made (Goetz, 2007; Bjarnegård, 2013; Bauhr and Charron, 2021).

Finally, building on studies that show a link between women’s descriptive representation and policy outcomes (Schwindt-Bayer, 2006; Wängnerud, 2009; Clayton, 2021), recent studies suggest that women’s policy priority on delivering social welfare services also contributes towards reducing corruption (Alexander and Ravlik, 2015; Bauhr et al., 2019; 2024; Gao and Mahutga, 2023). This could partly be attributed to women constituents being disproportionately affected by dysfunctional and corruption ridden public service delivery (Kubbe and Merkle, 2021).

Building on this work, we propose three candidate characteristics that may potentially partly explain why voters prefer women: risk aversion (personality), outsider status (background), and campaigning on social welfare issues (policy priority) – and investigate the effect of activating these stereotypes on voters’ candidate preferences. We explore whether female candidate advantage persists even if male candidates are also attributed these characteristics. In other words, we seek to contrast these explanations with our baseline expectation that voters simply see women candidates as more honest and trustworthy than male candidates (cf. Dollar et al., 2001; Wiesehomeier and Verge, 2020). Thus, given the difficulties involved in investigating the honesty/trustworthy stereotype directly, we follow the approach of Wiesehomeier and Verge (2020) and investigate the extent to which activating these alternative stereotypes can ‘explain away’ the general preference for women candidates. ⁴

Some recent experimental studies suggest that activating stereotypes about women candidate’s being risk averse or having outsider status make voters perceive women as less corrupt (Barnes et al., 2018; Barnes and Beaulieu, 2019). Similarly, Wiesehomeier and Verge (2020) suggest that women stand for honesty, but women candidates that are described as being as equally embedded in corruption prone networks as men are seen as just as corruptible as their male counterparts. Interestingly, male candidates may also benefit from being depicted as risk averse (Barnes and Beaulieu, 2019). Studies also provide more indirect support for the risk aversion theory by showing that women voters punish women candidates more harshly, electorally, for engaging in corruption (Eggers et al., 2018). However, this does not necessarily suggest that voters perceive women as more risk averse, but rather that it is particularly risky for women candidates to be implicated in corruption scandals because they are held to higher standards.

Fewer studies investigate if women candidate advantage can be explained by a social welfare policy profile. Studies suggest that citizens are more likely to perceive that women candidates are more competent in dealing with public service provision that directly affects citizens, such as health care and education (e.g., Alexander and Andersen, 1993; Huddy and Terkildsen, 1993; Dolan, 2010), and that women are also more caring and have a stronger communality compared to men (Ellemers, 2018). ⁵ We do not know, however, if campaigning on social welfare issues explains women candidate advantage in fighting corruption in the eyes of voters.

Thus, recent studies provide some direct or indirect evidence for the link between these candidate characteristics and a candidate’s expected capacity to contain corruption. Our suggested shift of dependent variable, from investigating whether candidates are seen as likely to be implicated in corruption scandals to whether candidates are seen as effective when it comes to fighting corruption more broadly, may have implications for preferred candidate characteristics. The numerous risks involved in fighting corruption are well documented (Okonjo-Iweala, 2018), not least since fighting corruption often comes at the cost of gaining powerful enemies among corrupt elites. Voters may, therefore, expect that candidates that take on this task may, in part, benefit from networks and connections to the elite if they also aim to stay in office (cf. Bauhr and Charron, 2021). It is therefore far from given that voters will see risk averse candidates as more likely to fight corruption, nor that they will be seen to benefit from having outsider status. Thus, despite the prominence of marginalization theory in explaining why women are excluded from participating in corrupt transactions (e.g., Goetz, 2007; Bjarnegård, 2013) and women being perceived as less corrupt (Barnes and Beaulieu, 2019; Wiesehomeier and Verge, 2020), outsider status is not necessarily seen as a desirable trait for a candidate that sets out to fight corruption.

However, we anticipate that voters prefer candidates that are seen as having the right profile to enact change, but that such candidates also possess the capacity to remain in office while doing so. In other words, if voters are sensitive to the numerous insights and accounts of the important personal and political risks involved in tackling corruption (Persson et al., 2013; Okonjo-Iweala, 2018; Bauhr and Charron, 2021), and therefore expect that only candidates that are careful and cautious will be able to fight corruption effectively while also remaining in office, risk aversion can be seen as a desirable candidate trait. Furthermore, candidates that campaign on social welfare issues may be seen as more likely to combat corruption, partly because this signals a commitment to improving access to public services and caring about the welfare of citizens (c.f. Bauhr et al., 2019). Expectations on the role of outsider status when it comes to explaining women candidate advantage are more mixed, however. On the one hand, an outsider may be more willing to challenge insider elites. Yet, being too much of an outsider can also impede a candidate’s capacity to challenge the status quo and also remain in office.

Thus, building on the last few decades of work on the link between gender and corruption, we investigate the extent to which the theoretical frameworks proposed by this literature contribute towards explaining voters’ perceptions of women candidate advantage in combating corruption. If this is the case, we expect the pro-woman bias to be reduced (or perhaps even disappear) when such stereotypes are activated. Conversely, if women candidates are preferred over male candidates, irrespective of what attributes a candidate is assigned on average, this would mean that these theories do not fully explain the advantage of women candidates. Given important evidence on voters perceiving women candidates to be generally more honest and trustworthy, we expect that these frameworks may not necessarily fully explain women candidate advantage.

Thus, we hypothesize:

H2a) The preference for women candidates over male candidates decreases if both candidates are depicted as risk averse as opposed to risk takers.

H2b) The preference for women candidates over male candidates decreases if both candidates are depicted as specializing in social welfare issues as opposed to other political issues.

H2c) The preference for women candidates over male candidates decreases if both candidates are depicted as being less networked outsiders as opposed to insiders.

A sizable literature also propounds that voter stereotypes and expectations about women candidates may vary across contexts as well as respondents. The positive effects of elected bodies more closely mirroring the population from which they are drawn has been widely observed (Phillips, 1998; Mansbridge, 1999) and studies suggest that women’s descriptive representation can improve perceptions of women’s ability to govern (Alexander, 2012). Furthermore, although findings are somewhat mixed, several studies confirm that women show a stronger aversion towards corruption than men (see Swamy et al., 2001; Torgler and Valev, 2010; Rivas, 2013). Relatedly, recent studies investigating gender differences in voter responses to corruption find that women are more likely to refrain from voting for a party and candidate involved in corruption scandals (Alexander et al., 2020). Studies also show that women voters, in particular, may have higher standards for female candidates (Eggers et al., 2018).

Drawing from this literature, we anticipate that women citizens may be more likely to perceive and prefer women candidates as ‘cleaners’. Furthermore, we also anticipate this effect to be particularly strong in contexts of greater political gender equality, where citizens would also be more likely to see women as possessing the necessary characteristics to effectively enact change. Although the notion of whether a critical mass of women is needed for women to effectively enact policies has been debated for decades (Childs and Krook, 2006; Funk et al., 2022), voters in contexts where there is a higher share of women representatives may be less likely to hold stereotypical perceptions of women being unfit for leadership or unable to enact change, i.e., to be independent agents rather than token women. Studies also advise that higher gender equality may be associated with larger gender differences in preferences and values (Schwartz and Rubel-Lifschitz, 2009; Falk and Hermle, 2018). It is also possible that citizens that live in contexts with a higher share of women representatives may have more experiences with women acting with integrity, and therefore expect more of women candidates. ⁶ Thus, building on this previous work, we hypothesize that women voters and voters in contexts of higher gender equality will be more likely to believe that women candidates will combat corruption.

H3a) Compared with male voters, women voters prefer women candidates on the issue of fighting corruption to a larger degree.

H3b) Voters are most likely to select a women candidate compared to a male candidate to fight corruption in areas where political gender equality is high.

Data and experimental design

The three hypotheses are tested using recently collected data from the European Quality of Government Index (EQI) Survey (see Charron et al., 2022). The survey was launched in October 2020 and concluded in the first week of February 2021 in all EU27 countries. A total of 129,991 respondents from all EU-27 member countries were included, with 50% being randomly selected and administered by computer assisted telephone interviews (CATI), while the other 50% responded via online administration, with our study using only the online sample (see Table A1 in the online Appendix for the full list and sample size by country). The survey’s main purpose is to gauge citizen attitudes about their local public services (such as health care and education) and the degree to which they rate them as high quality and administered with impartiality and low corruption. There are also a number of questions about citizens’ experiences with corruption, as well as some questions on elections and their own personal partisan preferences, along with standard demographic questions.

We embedded a conjoint survey experiment following the main EQI questions. For practical purposes, the experiment was only given to the online respondents, where a total of 48,080 completed the survey and answered the conjoint experiment without a ‘don’t know’ response, and each respondent was given the experiment twice, equaling 96,160 observations in our main analysis. Smaller countries Malta and Cyprus are not included in our analyses, thus our analyses have up to 25 EU member countries. As online respondents self-select into the survey, they tend to over-sample higher educated and younger respondents, inter alia (see Pew Research, 2018). However, treatment effects between online, convenience and randomly drawn, nationally representative samples tend to be similar (Coppock, 2019). In addition, the EQI’s target sample is between 500–600 respondents at the NUTS 1 and NUTS 2 regional (sub-national) levels within countries, ⁷ and thus respondents from less populated, rural regions have a much higher probability of being included in the survey compared with respondents from urban regions. We thus employ post-stratification and design weights, which adjust differences between the sample and population by gender, age, education and regional population. More on the survey and the respondents can be found in the online Appendix.

The research design employs a dual profile conjoint experiment (Hainmueller et al., 2014). Similar to a factorial vignette, the experiment allows researchers to identify causal effects of several variables of interest concurrently and has been used in several recent publications on the effects of corruption on voting preferences (e.g., Breitenstein, 2019; Agerberg, 2020; Mares and Visconti, 2020). A key difference in our approach is that rather than testing the degree to which corruption matters in candidate choice by randomizing the presence or absence of a scandal in the candidate profiles, we imply to the respondents that the corruption scandal has already occurred, which allows us to focus on which candidates are most stereotypically seen as cleaners. The respondents are presented with two different mayoral candidate profiles, from which they are asked which candidate they perceive would be best able to combat corruption in their municipality. The question reads as the following:

Finally, let’s say that two politicians in a city or town similar to yours are running ⁸ for city mayor, and there have been several recent corruption scandals. The election is close and these are the only candidates with a chance to take office. Based on what you see below, which of these candidates do you believe would be best to reduce municipal level corruption?

The candidate profiles consisted of six different factors. The first was the candidate’s gender, male or female. We randomize three factors to capture the mechanism elucidated from the literature – risk aversion, issue specialization, and insider/outsider status. Admittedly, these are not straightforward concepts to operationalize, yet we take a pragmatic approach and make simple binary or categorical variables to capture these concepts to provide respondents with simple contrasts, to test effects from these factors. The ‘risk aversion’ mechanism is operationalized on a line labeled ‘personality type’ in the conjoint set up, where we randomize two traits ‘cautious and careful’ and ‘willing to take risks’. The insider/outsider mechanism is captured on a line titled ‘background’, where a candidate is randomly assigned either ‘well-connected politician’ or ‘political outsider’. As per the issue specialization factor, we randomly assign each candidate with one of four main issues on the line ‘main election issue’ – education, health care, infrastructure or economic development.

Additionally, we include two other factors. The first is candidate age, for which we limit variation to just three years – 51, 52 and 53 ⁹ – as we are less interested in this factor. The final factor is the one in which the literature on corruption and voting highlights as most salient – that of partisanship (Anduiza et al., 2013; Eggers, 2014). We include a randomized list of the largest 2–5 parties in each country and assign them randomly to each candidate. We then take advantage of the respondent’s previous answer on the question of which party they would support ‘if the election were held today’ and create a variable for a ‘partisan match’ – whereby the variable equals ‘1’ if the candidate belongs to the party that the respondent claimed to support, and ‘0’ if otherwise. This allows us to compare the magnitude of the effects of candidate gender to the most salient heuristic on which voters rely. The full visual of the conjoint table is found in the Appendix, section 2.

With this test, whether gender has an independent effect on candidate choice given these three factors, along with partisanship, are simultaneously accounted for and thus we can compare the relative effects across all treatments simultaneously. The outcome question distinguishes between ‘voice’ (selecting one of the available candidates, H1a) and ‘exit’ (refraining from selecting any of the available candidates, H1b, c.f. Hirschman, 1970; Bauhr and Charron, 2018). We argue that allowing respondents to abstain from choosing any of the candidates is a more realistic scenario and provides more valid estimates (see Haaijer et al., 2001; Agerberg, 2020 for more on this point). In the sample, 34.2% selected ‘candidate a’, 34.6% selected ‘candidate b’ while 31.2% selected ‘neither’. The survey was run twice for each respondent, randomizing the candidate profiles each time.

Our empirical strategy of the main conjoint experiment is to estimate the average marginal component effect (AMCE) of the quantities of interest (de la Cuesta et al., 2022). The AMCE is conceptually similar to the average treatment effect (ATE) and provides the causal effect of a single candidate characteristic, while averaging over the remaining characteristics. Our estimation relies on the assumptions of the AMCE, which are that there are a) no carryover effects across profiles, b) no profile order effects, and c) profiles are completely and independently randomized (Hainmueller et al., 2014). To mitigate any biases from violations for the latter two assumptions, we randomized the ordering of the candidate attributes, and an algorithm was used to completely randomize candidate profiles across and within respondents. While some critique the use of the aggregated AMCE in analyzing electoral preferences in conjoint experiments (Abramson et al., 2022), we are not interested in the electoral results per se, but about anti-corruption stereotypes among the general population. Thus, for H1a and H1b, the AMCE is the best available estimand. H2–H3 require additional estimation over the AMCE.

The additional advantages of this design are that we can estimate two types of interaction effects that are of interest to our theoretical propositions. One, we can estimate the average component interaction effect (ACIE), which tests whether the causal effect of one attribute (in our case candidate gender) varies based on other characteristics, for example a candidate’s background or personality type. In this case, we specify interactions between candidate characteristics of interest to test H2. Two, to test H3 we rely on the conditional AMCE (Hainmueller et al., 2014), in which we are interested in the interaction between fixed respondent characteristics (their own gender) and randomized candidate characteristics.

Results

Candidate sex and vote choice: H1a

We begin by testing H1, whether respondents on average favor female candidates when it comes to reducing corruption levels, all things being equal. We first estimate the model by including all candidate A and B characteristics simultaneously, and run a multinomial logit regression, with candidate B as the baseline outcome. The full results can be found in Appendix, section 2, Table A2. ¹⁰ For purposes of presentation, we use a simple binary dependent variable with the outcome being ‘1’ vote for candidate A and ‘0’ if vote for B (or for hypothesis H1b ‘1’ for abstention and ‘0’ for vote for either candidate), and estimate the causal effects of the candidate characteristics using a linear probability model to elucidate marginal effects. The results are nonetheless very similar to the multinomial estimation and are summarized in Figure 1. ¹¹

Figure 1. Treatment effects on choosing candidate A – marginal means.

Note: Dots show the average marginal component effect (AMCE) of the ‘candidate A’ via marginal means (MM) from Logit estimation, whereby the difference in predicted probability for the various treatments (x-axis) represents the AMCE. The baseline support for ‘candidate A’ is 0.339 is highlighted via the dashed vertical line. All candidate B characteristics are also included in the model along with age (not shown). Issue: ‘Inf’ = infrastructure, ‘Edc’ = education, ‘E.D.’ = economic development, ‘H.C.’ = health care. 48,080 individuals and 96,160 total profiles analyzed. Estimates adjusted for design and post-stratification weights and standard errors are clustered by respondent.

Figure 1 above summarizes the results and shows the relative magnitude of each of the candidate characteristics. We observe several interesting findings: First, as expected from the literature and clear in our results, partisanship is the most salient factor in a respondent’s choice. When the candidate’s party is the same as the respondent’s preferred party, this increases the probability that the respondent will choose that candidate by over 40% – that is to say respondents go from roughly a 29% chance of selecting ‘candidate A’ to 70%; which is a similar magnitude to the effect of party identification found in experimental studies in the USA among partisans (Bonneau and Cann, 2015). Clearly overall, voters believe their own party is best equipped to combat corruption relative to other parties. Yet despite the large effect of partisanship, we still observe an independent effect of candidate gender in the model. When a candidate is presented as a female (compared with a male), the respondent is roughly 3% more likely to choose that candidate, all things being equal. Thus, the results provide initial empirical evidence for H1.

Moreover, Figure 1 shows that among the other mechanisms, we find support for our contention that voters prefer a candidate that is ‘careful and cautious’ over one that is a risk taker when it comes to reducing corruption levels. Additionally, we see that issues that candidates specialize in also play a role in guiding voters to determine which candidate might be best suited to combat corruption. The results show that when candidates run on infrastructure (‘Inf.’), they are least preferred by respondents, possibly due to the perceived opportunities for corruption in local infrastructure spending. However, while economic development (‘E.D.’) is the most preferred by respondents, respondents believe that candidates that specialize on heath care (‘H.C.’) and also, but to a somewhat lesser extent, education (‘Edc.’), are better at reducing corruption levels. Finally, presenting a candidate as a political outsider relative to a connected insider, ¹² does not affect a respondent’s choice.

Candidate sex and voter abstention: H1b

Our next hypothesis deals with whether the candidate’s sex affects respondents’ choice to abstain from electing a candidate altogether. In this case, we shift our dependent variable to the ‘exit’ or ‘abstain’ option and collapse the candidate gender options into three groups – 1) two males, 2) split gender, 3) two females. We anticipate that seeing two males on the ballot will reduce turnout compared with the other constellations of candidate sex.

Figure 2 elucidates our findings. We find that respondents are significantly more likely to abstain only when two male candidates are shown, and that rates of predicted abstention are equal for one or two females. We interpret this as a small, but significant ‘male penalty’ that voters have against candidates in the face of a corruption scandal, demonstrating support for H1b.

Figure 2. Candidate sex and abstention.

Note: Dots show the predicted probabilities of Abstention from Logit estimation with 95% confidence intervals. All candidate A and B characteristics are also included in the model (not shown). 48,080 individuals and 96,160 total profiles analyzed. Estimates adjusted for design and post-stratification weights and standard errors are clustered by respondent.

Risk-aversion, outsider status and prioritizing social welfare issues: testing H2

To test our second set of hypotheses, we begin with a series of interaction terms between the gender of the candidate and each of the three variables capturing the mechanisms (ACIE). Building on three broad theoretical frameworks derived from the larger literature on the gender corruption link-outsider status, risk aversion and prioritizing social welfare policy, we are interested in whether the gender gap decreases or persists when a candidate is assigned these traits and if candidates ascribed these traits are preferred among the electorate. The findings are presented in Figure 3 for the main interactions of interest: those between candidate’s sex, and the three mechanisms of personality, insider/outsider background and election issue. In this case, we provide marginal means (Leeper et al., 2020), where the level of support for males (blue triangles) and females (black circles) are shown over the range of attributes across the other three variables from three separate interaction models. The horizontal dashed line represents the mean level of support for ‘candidate A’ compared with the other two alternatives.

Figure 3. Marginal mean plot of ACIEs – summary of H2.

Note: Blue triangle and black circles represent male and female candidates respectively with 95% confidence intervals from respondent-clustered standard errors. Estimates from three separate interaction models with candidate gender – (1) personality type, (2) background status, (3) campaign issue. Dashed line represents the mean, baseline support for ‘candidate A’ (0.3405). Design and post-stratification weights are included.

Overall, we find no significant two-way interaction terms, meaning that respondents prefer females over males at a consistent rate across all values of the other variables in the model on average, which is clearly shown in Figure 3. Thus, contrary to H2 a–c, the gender gap persists despite the activation of these candidate traits, suggesting that they cannot explain women candidate advantage when it comes to fighting corruption and also providing strong, albeit indirect, support for our contention that women are simply seen as more honest and trustworthy. The analysis also shows that respondents on average prefer more cautious candidates, and those who do not campaign on infrastructure regardless of the gender of the candidate, while insider/outsider is negligible. However, irrespective of what other attributes a candidate is assigned, we observe a significant ‘female advantage’ in all cases compared with the hypothetical male one; the gender gap favoring the female candidate is roughly 3% across all attributes. ¹³ Thus, the ‘female advantage’ is not washed away by ascribing a female candidate any of the above characteristics.

Heterogeneous treatment effects by respondent gender: testing H3a

We now move to testing whether the treatment items in the conjoint experiment have varying effects based on respondent characteristics. Namely, we are interested in whether female respondents respond differently to the various treatments compared with males in our survey, since previous studies find gender differences in perceptions of candidates (e.g., Eggers et al., 2018). We are therefore interested in this case in the conditional ACME, and include standard age, education and population of residence when estimating interactions with a respondent’s gender.

Figure 4 summarizes our findings from several interaction models of interest. ¹⁴ To investigate the effect of the respondent’s gender for H3a, we simultaneously interact respondent’s gender with each of the treatments to test whether females are more inclined to support certain types of candidates to combat corruption relative to males. In our first model, we find that while both male and female respondents prefer a hypothetical female candidate over a male one, the treatment effect is three times larger among females than among males (0.046 compared with 0.015, P = 0.001). Thus, the findings show that the conditional effect is largely driven by the fact that female respondents are simply less inclined to support male candidates. Similarly, in model 2, we find that both females and males prefer candidates that are ‘cautious’ compared with ‘risk takers’ on average, yet the effect of this personality distinction is only significant among females (0.023, P < 0.05). Thus, in models 1 and 2, females respond more strongly to the candidate attributes of sex and personality more so than male respondents. Both interactions are statistically significant at the 0.05 threshold.

Figure 4. Conditional treatment effects by respondent’s gender – marginal means.

Note: Predicted probability of ‘voting candidate A’ shown on the x-axis, with 95% confidence intervals from respondent-clustered standard errors. Results from interaction models, with marginal mean plots by respondent gender. Design and post-stratification weights are included.

The interaction model shows no conditional treatment effect regarding outsiders or insiders. However, with respect to the issue specialization, we do find mainly consistent treatment effects for both female and male respondents. Females, however, only respond negatively to ‘infrastructure’ as an issue compared to the other three, while male respondents favor candidates who run on health care and economic development when considering combating corruption, compared with education and infrastructure.

Are the effects driven by context of gender equality? Testing H3b

As noted in the hypothesis, we anticipate that the effect of female candidacy will be strongest in areas where the levels of political gender equality are highest, both because more women in office may improve citizens’ beliefs in women’s ability to govern (Alexander, 2012) and because citizens in these contexts may share beliefs and experiences of women’s capacity to enact integrity reforms. To test H3b empirically, we thus ran several cross-level interactions with candidate gender and variables that serve to proxy for contextual political gender equality, including share of women in sub-national parliaments. As political gender equality varies significantly within countries, we report the cross-level interaction with the sub-national measure to account for the within-country variation of the moderating variable and to increase the number of second-level observations in the model from 25 countries to 200 ¹⁵ .

Figure 5 reports the results of the interaction model, whereby we highlight the preference for ‘candidate A’ on the y-axis as a function of the randomized gender of candidate interacted with the proportion of female MPs in sub-national, regional parliaments as a moderating variable (x-axis). In this case, we observe clear moderating effects of gender equality; where female representation is low – roughly less than 20% – there is no observable effect of candidate gender on voter’s preferences to combat corruption. This constitutes approximately 19% of the sample. Yet at 20% of female representation and higher, there is a significant difference in line with H1 whereby respondents prefer female candidates over male ones. At the highest levels of gender equality there is a roughly 5.1%-point difference in the preference for female candidates over male ones.

Figure 5. Women’s representation and preferences for women candidates.

The results in Figure 5 are robust to the alternative measure of female share of MPs in national parliaments (Figure A2) as well as alternative estimation methods such as hierarchical estimation whereby we specify random regional intercepts and random slope for the candidate gender variable. In Figure A3, we show the AMCE test of H1 by country to further elucidate the degree of context-level effects of the main variable. In sum, we find that respondents prefer women candidates across most countries studied, yet that the effect is insignificant in some of the smaller countries in our sample. In no case do we find that respondents in any country significantly prefer male candidates over females to fight corruption.

Anti-corruption, or simply a preference for female candidates in general?

While our results support the main hypotheses, one might question whether the observed findings are simply driven by a general preference for female candidates. In order to investigate whether our results are driven by a general ‘pro-women bias’ (Aguilar et al., 2015; Schwarz and Coppock, 2022) rather than a preference for women candidates when it comes to corruption specifically, we provided an alternative, more neutral question framing to 25% of the sample in countries with a sufficient number of respondents. ¹⁶ The alternative framing inquired only about candidate preference in a mayoral election, and there were no mentions of corruption or corruption scandals (i.e., the phrase ‘and there have been several recent corruption scandals’ is taken out.) Full results of this analysis are reported in Appendix Table A4 and Figure A1. In general, we find some evidence for a general pro-woman bias in our sample. However, the significant interaction term (P = 0.03) indicates that the ‘female effect’ is roughly 2.3 times larger (0.034 versus 0.015) when corruption is specified, compared with voting in general, which supports the notion of a female ‘cleaner’ stereotype in response to corruption. Thus, our results cannot exclusively be explained by a preference for women candidates in general. In terms of the other attributes concerning H2, we do however find that the respondents prefer cautious candidates versus risk-taking candidates, and dislike infrastructure as an issue compared to the others in general, across both treatment groups.

Conclusions and discussion

This study investigates voters’ perceptions of women candidate advantage in fighting corruption. Using a unique conjoint experiment fielded in 25 European countries, with 48,080 respondents, we investigate if voters believe that women candidates will fight corruption across diverse political and cultural contexts. We elucidate three main findings. First, we find clear and consistent support for our contention that women mayors are seen as more likely to fight corruption compared to male mayors. Notably, voters express support for women candidates as corruption fighters, both by selecting women candidates for this task (voice) and by refraining from selecting any candidate if there are only two males on the ballot (exit). Women candidates obtain roughly 3% greater support from voters, and this support cannot be explained by a preference for women candidates in general. Second, while activating stereotypes about risk aversion and social welfare policy priority leads to greater electoral support in general, outsider status has no effect, on average. Interestingly, however, women candidates do not benefit more than male candidates from being assigned these traits, and women candidates are seen as the preferred anti-corruption fighters regardless. Third, while voters across Europe find that women are the better choice when it comes to fighting corruption, effects are more pronounced among women voters (cf. Eggers et al., 2018; Barnes and Beaulieu, 2019) and in contexts where the share of women representatives is relatively large.

Overall, we find that voters prefer women candidates when it comes to fighting corruption, and that this effect is particularly strong among female respondents. In looking at ‘ideal types,’ we find that women candidates, described as cautious and careful and specializing in improving public service delivery and the economy, are preferred over male candidates that are risk takers or who specialize in infrastructure. Naturally, one caveat in this study is that one can certainly question whether a response to a hypothetical question such as the one used here translates into the real world (Incerti, 2020). While there is some evidence that conjoint survey experiments compare relatively well with real world behavior (Hainmueller et al., 2015), which speaks to the external validity of our study, we are indeed most interested in voter stereotypes, for which we believe this hypothetical question provides an ideal setting. On that note, it is also worth highlighting that our study is about voter perceptions of desirable candidate characteristics, and that these perceptions do not necessarily predict actual candidate effectiveness when it comes to combating corruption. Consequently, our study suggests several interesting avenues for future research. Follow-up studies could include developing new experiments, varying the type of manipulation ¹⁷ or studying how actual candidate characteristics have implications for their effectiveness in fighting corruption.

Our findings have several implications for our understanding of voter preferences for women candidates and the credibility of anti-corruption candidates more broadly. Our study highlights important new insights into how theoretical frameworks typically used to explain the link between women in office and lower levels of corruption, resonate with voters. Voters generally prefer candidates that are cautious and careful, as opposed to risk takers, and who focus on social welfare issues. Potentially, this reflects that voters are sensible to the important personal and political risks involved in fighting corruption or implementing other important changes while also remaining in office (Bauhr and Charron, 2021). Interestingly, while some studies suggest that women’s focus on social welfare issues can contribute towards explaining why women candidates reduce corruption (Alexander and Ravlik, 2015; Bauhr et al., 2019; Gao and Mahutga, 2023), our study also provides micro level evidence that voters prefer candidates with this policy profile, in particular compared with candidates that focus on infrastructure, which has been notoriously exposed to corruption risks (OECD, 2016; Bauhr et al., 2020).

Ultimately, our study suggests that voters perceive women candidates as more honest and trustworthy when it comes to actively fighting corruption, and not only as a candidate that should refrain from being implicated in scandals. This is amplified by female voters and in areas with greater levels of gender equality. Our results have implications for our understanding of voters’ expectations of women candidates, since most previous studies focus on voter punishment of allegedly corrupt women candidates (Eggers et al., 2018; Barnes and Beaulieu, 2019; Wiesehomeier and Verge, 2020; Batista Pereira, 2021). If women are also preferred as anti-corruption fighters, this may add to the electoral advantage of women candidates, but also potentially to the challenges that women candidates face in meeting voter expectations.