Efficiency versus Equality: A Survey Experiment with French Economists

In this project we conduct a survey experiment with French economists who express opposing political orientations. Our experiment includes both a social task and a risk task. In the social task, respondents are presented with a series of decisions involving a trade-off between equality and efficiency under the veil of ignorance. In the risk task, respondents face a series of decisions involving a trade-off between risk and expected return.

The experiment comprises two types of tasks: a social task and a risk task. A short demographic survey always comes last, in which we collect respondents’ basic informa- tion, including age, gender, country of work, profession, research areas, and a set of be- liefs related to income inequality in France, its causes, and the policy trade-offs between equality and efficiency. In the social task, respondents engage with two scenarios—the leaky bucket, as described by Okun (2010), and the reverse leaky bucket. In the leaky bucket scenario, respondents are asked to envision a world with two income classes: the rich and the poor. Following Rawls’ veil of ignorance concept (Rawls, 1971), they assume an equal chance of belonging to either class. The rich earn n times more than the poor, and we assess respondents’ willingness to transfer 1,000 =C from the rich to the poor, considering potential inefficiencies. Specifically, we aim to identify the maximum loss respondents are willing to accept for such transfers. In the reverse leaky bucket scenario, we evaluate their willingness to support a policy that increases the income of the rich by 1,000 =C at the expense of the poor, thereby eliciting the maximum loss to the poor that respondents are willing to accept.
The risk task mirrors the social task but removes the social aspect, focusing on in- dividual risk-taking decisions. Respondents are presented with a lottery that can yield either a high or a low outcome, each with equal probabilities. The high and low out- comes correspond to the high and low incomes in the social task. This task includes two scenarios: the hedging scenario and the leveraging scenario, which parallel the leaky bucket and reverse leaky bucket scenarios, respectively. In the hedging scenario, re- spondents make insurance-related decisions, allowing them to transfer 1,000 =C from the high-earning state to the low-earning state, acknowledging some efficiency loss that reflects deadweight loss. We assess the maximum loss respondents are willing to accept in response to such policies. In the leveraging scenario, respondents can agree to a pol- icy that increases the high earning by 1,000 =C while decreasing the low earning, thus raising the associated risk. Here, we elicit the maximum reduction in the low earning that respondents are willing to accept.

2024-11-11

2024-12-31

The maximum efficiency loss that respondents are willing to accept when transferring €1000 from the rich to the poor in the “leaky bucket” scenario.
The maximum loss to the poor that respondents are willing to accept when a policy benefits the rich by €1000 in the “reversed leaky bucket” scenario.
The maximum efficiency loss that respondents are willing to accept when transferring €1000 from the high-earning state to the low-earning state in the “hedging” scenario.
The maximum loss in the low-earning state that respondents are willing to accept when a policy adds €1000 to the high-earning state in the “leveraging” scenario.

This survey experiment targets two groups of economists: those who openly support the NFP (National Front Populaire) and those who do not. We hand-collected the email addresses of 111 economists in the ‘Against NFP’ group and 269 in the ‘For NFP’ group, based on publicly available information from social media. Both groups were invited to participate in an online survey experiment through individual emails.
The experiment comprises two types of tasks: a social task and a risk task. A short demographic survey always comes last, in which we collect respondents’ basic informa- tion, including age, gender, country of work, profession, research areas, and a set of be- liefs related to income inequality in France, its causes, and the policy trade-offs between equality and efficiency. In the social task, respondents engage with two scenarios—the leaky bucket, as described by Okun (2010), and the reverse leaky bucket. In the leaky bucket scenario, respondents are asked to envision a world with two income classes: the rich and the poor. Following Rawls’ veil of ignorance concept (Rawls, 1971), they assume an equal chance of belonging to either class. The rich earn n times more than the poor, and we assess respondents’ willingness to transfer 1,000 =C from the rich to the poor, considering potential inefficiencies. Specifically, we aim to identify the maximum loss respondents are willing to accept for such transfers. In the reverse leaky bucket scenario, we evaluate their willingness to support a policy that increases the income of the rich by 1,000 =C at the expense of the poor, thereby eliciting the maximum loss to the poor that respondents are willing to accept.
The risk task mirrors the social task but removes the social aspect, focusing on in- dividual risk-taking decisions. Respondents are presented with a lottery that can yield either a high or a low outcome, each with equal probabilities. The high and low out- comes correspond to the high and low incomes in the social task. This task includes two scenarios: the hedging scenario and the leveraging scenario, which parallel the leaky bucket and reverse leaky bucket scenarios, respectively. In the hedging scenario, re- spondents make insurance-related decisions, allowing them to transfer 1,000 =C from the high-earning state to the low-earning state, acknowledging some efficiency loss that reflects deadweight loss. We assess the maximum loss respondents are willing to accept in response to such policies. In the leveraging scenario, respondents can agree to a pol- icy that increases the high earning by 1,000 =C while decreasing the low earning, thus raising the associated risk. Here, we elicit the maximum reduction in the low earning that respondents are willing to accept.

We control for potential ordering effects between different types of tasks. Our baseline ordering ($SR_1$) is as follows: respondents start with the social task, beginning with the leaky bucket scenario, followed by the reverse leaky bucket scenario. They then proceed to the risk task, starting with the hedging scenario and concluding with the leveraging scenario. An alternative ordering ($RS_1$) is as follows: respondents begin with the risk task, starting with the hedging scenario and concluding with the leveraging scenario, before moving to the social task, which starts with the leaky bucket scenario and concludes with the reverse leaky bucket scenario. There are two additional orderings ($SR_2$ and $RS_2$), in which we reverse the order of scenarios within each type of task. In $SR_2$ and $RS_2$, the reverse leaky bucket scenario and the leveraging scenario are presented first in their respective tasks.

At the individual level.

Yes

We hand-collected the email addresses of 111 economists in the ‘Against NFP’ group and 269 in the ‘For NFP’ group, based on publicly available information from social media.

We hand-collected the email addresses of 111 economists in the ‘Against NFP’ group and 269 in the ‘For NFP’ group, based on publicly available information from social media.

Given the nature of our study, we are currently uncertain about how many economists will respond to our survey. We will need 63 participants in each group based on our power analysis.

Given the nature of our study, we have a limited sample size compared to other survey studies that use representative populations. The power analysis suggests that we would need approximately 63 participants per group to detect a moderate effect size with 80% power in a two-tailed, two-sample t-test.