Research

Abstract: 

This paper investigates how affirmative action quotas influence team decision-making through belief updating and confidence. While quotas increase diversity, they also create uncertainty about the role of merit versus demographics in selection. As a result, success conveys less information about candidates' relative abilities within a team when there is a quota. Yet, little is known about how such uncertainty influences participants' confidence and team dynamics. In a laboratory experiment, teams with a quota achieved, on average, 2 points higher scores on a 0–10 scale compared to teams without a quota, once controlling for any composition effect. While quotas do not differentially affect individual contributions by gender, they generate a gender gap in the confidence signals conveyed during interaction. When including the negative selection effect of the quota, I do not obtain a negative effect of quotas on team performance. This is due to women becoming less confident and contributing relatively less than men when advantaged by a quota. 

Abstract: 

Around the world today, women continue to earn less than men, a disparity largely attributed to parenthood. While the child penalty has been widely documented, the underlying mechanisms driving this phenomenon remain unclear. In particular, there is little evidence on whether the child penalty is driven by the supply side of the market (i.e., the behaviour of working parents of different gender) or by the demand side (i.e. the behaviour of employers towards working parents of different gender). This paper explores one potential factor contributing to the gender gap in career advancement at childbirth: the assignment and the acceptance of low-promotability tasks. Low-promotability tasks are defined as tasks that benefit the organization but are given relatively little weight in performance evaluations and promotion decisions.

The objective of this research is thus to address two questions. First, is there a gender gap among working parents in the assignment and acceptance of low promotability tasks? Second, does the allocation to low-promotability tasks contribute to the child penalty among working parents? To address these questions, we implement a survey experiment with high-skilled Italian workers.

Abstract:

Are female legislators more attentive to climate change issues than men? Despite evidence that female voters express greater concern for climate change and more frequently engage in pro-environmental behavior, it remains unclear whether these gender differences are mirrored in legislatures. Adopting a regression discontinuity design to study UK parliamentary speeches and voting records, we examine whether female legislators are more attentive to climate change and adopt more progressive climate positions than their male counterparts. We find no significant gender differences in attention to or positioning on climate issues, suggesting that, at least in the context of climate change topics, descriptive representation does not automatically translate into substantive representation. Supplementary analyses indicate that these null results are driven by party control and centralized agenda-setting on climate topics, which constrain individual MPs’ ability to express independent views. Our findings suggest that, rather than focusing on increasing women’s electability, greater attention should be given to encouraging parties to adopt more environmentally ambitious platforms.

Abstract: 

Conducting experiments online offers many advantages, but interactive designs requiring participants to engage simultaneously remain challenging: researchers cannot keep participants waiting indefinitely until enough are connected. This paper proposes practical guidelines and introduces a new method for running such experiments on online platforms. The method combines a maximum waiting time with brief, low-effort tasks for early arrivals, which slows down the fastest participants without reducing engagement. I show that this approach effectively increases the number of active participants at low cost and without impairing attention, performance, or social behavior. The paper also provides the accompanying Python code to implement the method.

Abstract:

The transfer of power through elections is a cornerstone of democracy. We examine the effect of elections on a key health measure, excess mortality. We argue that electoral turnovers can create frictions that delay policy implementation and thus harm health, especially in times of crisis. Our data come from France, where local elections were held at the beginning of the Covid-19 pandemic. Combining a conditional-on-observables strategy and a regression discontinuity design to analyze 5,000 electoral decisions, we find strong evidence supporting our theoretical argument. Municipalities where the incumbent mayor was defeated by an opposition candidate experienced 7-9 percentage points higher excess mortality, implemented fewer mask-wearing mandates, and issued significantly fewer emergency orders than municipalities with no turnover. Difference-in-differences analyses on a panel of 101 countries and territories suggest that these findings generalize to other contexts. While crucial to the functioning of democracy, holding elections during emergencies comes at a cost. 

Abstract: 

Trust is an important economic variable that may however be subject to measurement error, leading to econometric issues such as attenuation bias or spurious correlations. We use a test/retest protocol to assess idiosyncratic noise in the two main tasks that are used to elicit trust, namely survey questions and experimental games. We find that trust measures based on the trust game entail substantial measurement error (with up to 15\% of noise), while there is virtually no noise in stated trust measures. Given the specificity of our subject pool (students in a top Engineering school) and the short period of time between the test and the retest, we consider these percentages of noise as lower bounds. We also provide a sub-group analysis based on measures of cognitive ability and effort. We find substantial heterogeneity across sub-groups in trust-game behavior, but none for the survey questions. We finally discuss which measure of trust should be used, and the estimation strategies that can be applied to limit the effect of measurement error.

Abstract: 

How do voters react to an ongoing natural threat? Do voters sanction or reward incumbents even when incumbents cannot be held accountable because an unforeseeable natural disaster is unfolding? We address this question by investigating voters’ reactions to the early spread of COVID-19 in the 2020 French municipal elections. Using a novel, fine-grained measure of the circulation of the virus based on excess-mortality data, we find that support for incumbents increased in areas that were particularly hard hit by the virus. Incumbents from both left and right gained votes in areas more strongly affected by COVID-19. We provide suggestive evidence for two mechanisms that can explain our findings: an emotional channel related to feelings of fear and anxiety, and a prospective-voting channel, related to the ability of incumbents to act more swiftly against the diffusion of the virus than challengers.

Abstract: 

When committee members care about their reputation with a principal, making their choices transparent affects the outcome. In Parliaments, legislators care about their reputation with several principals, namely their constituents and Party leaders. It is thus unclear in which direction votes will move when they become observable, and moreover legislators may prefer to opt out of voting entirely in order to avoid conflict. This paper first uses French voting data to show that reputational concerns drive the decision to participate in a vote: in order to avoid blame legislators are less likely to vote when there is disagreement between the constituents and the Party. Second, making legislators' votes public increases their incentive to use voting for reputation-building, and, therefore, the distortion in group decision-making. The French transparency reform of 2014 provides a quasi-natural setting for a Difference-in-Differences analysis. Greater transparency led to less participation, as legislators preferred not to take sides.