Researchers have documented that female leaders tend to make different policy decisions than male leaders, especially in developing countries. They have been shown, for instance, to be less prone to corruption than men and to invest more in certain public goods, like health and education. But less is known about why female leaders make different choices. One possible explanation is that they have distinct policy preferences. In a recent paper, we explore an alternative mechanism: that female leaders seeking re-election adopt different policies because they face gender bias from voters and are judged differently than men, creating different electoral incentives.
Our research focuses on female and male mayors facing re-election in Brazil in November 2020, and the policies they enacted to combat the unfolding COVID-19 pandemic. A crisis, like the pandemic, is a good time to study electoral incentives. The stakes around different policies are high and extremely important to voters, and there is a direct link between policies and outcomes.
Brazilian Mayors During COVID
Brazil represents a good setting for such a study. Unlike other countries where decisions on containment policies are made at the regional or national level, mayors in Brazil’s federal system have jurisdiction over them. Brazilian mayors also face a two-term limit, meaning that mayors who were in their second term were not able to run again in November 2020, enabling us to compare mayors facing re-election incentives to term-limited mayors. In addition, the more than 5,000 municipalities provide a large sample of male and female mayors whose COVID policies can be compared.
To filter out other factors that could affect the spread of COVID-19 at the municipal level besides the gender of the mayor, we compare municipalities where a female candidate narrowly won against a male candidate in the 2016 election- the last one before the COVID outbreak – to those where a male candidate narrowly won against a female candidate. These municipalities (shown in Figure 1) are thus similar in every aspect other than the gender of their mayor. We also show that gender constitutes the chief difference between these narrowly-elected mayors, as they are comparable in terms of incumbency status, age, political orientation and other factors that could influence policy choices.
Figure 1. Municipalities with Close Mayoral Elections in 2016, in which the Top Two Candidates Were a Man and a Woman
The Different Evolution of COVID in Female-Led Municipalities
The results show clear differences between municipalities led by male and female leaders. Female-led municipalities had a 195% higher death toll at the beginning of the first wave of disease (April and May, 2020), almost three times that of male-led municipalities. But at the end of 2020 (November and December, 2020) they experienced a sharp reversal of fortune. In this period, female-led municipalities reported a 41% lower death toll relative to male-led municipalities.
We provide further evidence showing that these sharp differences in COVID mortality can be explained by different responses from female and male mayors to the crisis. Indeed, female mayors started closing non-essential businesses nearly a month later than their male counterparts in the pandemic’s early phase, when people still didn’t expect a major COVID impact. Instead, they were more likely to impose such measures later in the year – when COVID’s effect on health and mortality was clear.
Gender Biases and Electoral Incentives
Why the marked differences in these vital policy choices? A large body of literature provides evidence that voters are biased against female candidates and female leaders, believing them to be less capable of strength and assertiveness and judging women’s performance in office more harshly, both in business and politics. These biases could help explain why female mayors were less willing to intervene early in the pandemic when the COVID impact was still moderate and the toll on the economy potentially great. By contrast, voters’ biases could have pushed female leaders in the opposite direction when the threat of COVID to human life was glaringly evident later in the year. Voters’ belief that female leaders are less effective might have pushed them to implement more policies to achieve the same perceived results as their male counterparts.
This explanation appears to be supported by the data. The difference in COVID deaths in male versus female-led municipalities occurred only in those municipalities where mayors were not term-limited. They occurred, in other words, where leaders’ policy choices could impact their re-election chances. These differences were even greater in municipalities where the elections were particularly competitive and in municipalities where one could expect higher gender discrimination and voter bias, such as those with a higher gender wage gap in the local labor market.
Our findings speak directly to how we think of the choices made by male and female politicians. Despite the general view that female officials tend to care about different issues than male politicians, these results show that women’s choices in office may well be driven by electoral incentives, that is, by their need to respond to voters’ gender-biased assessment.