Measuring Identity, Factual Beliefs, and Policy Views
How to cite
Gennaioli, Nicola; Tabellini, Guido. 2021. "Measuring Identity, Factual Beliefs, and Policy Views." AEA RCT Registry. AEARCTR-0008707. https://doi.org/10.1257/rct.8707-1.3000000000000003
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
In Bonomi, Gennaioli, and Tabellini 2021 we develop a theory of identity politics that builds on two ideas. First, when policy conflict renders a certain social divide (economic or cultural) salient, a voter’s beliefs are tainted by the stereotype of the group he identifies with. Second, culture influences beliefs across many domains. One key prediction of the group is that voters’ beliefs about facts and their policy views are slanted toward the distinctive opinions of the group a person identifies with and in away from the opinions of the outgroup. By this mechanism, identity is correlated with specific forms of incorrect beliefs and polarization in policy views. In this study, we plan to ask respondents to answer questions on their: i) identity, ii) factual beliefs, and iii) policy preferences on a range of domains (economic domains: redistribution; cultural domains: immigration, abortion, affirmative action. We also inquire factual beliefs and policy views on trade policy and environmental policy). With such data, we want to test some hypotheses and predictions of the model.
Some of the questions we would like to answer are:
1. Does identity predict disagreement in terms of beliefs and policy preferences across economic and cultural domains, and also the extent to which disagreement is correlated across different domains?
2. Is disagreement stronger for individuals with a strong identity?
3. Does identity with political parties play a distinct role from identity with social groups (e.g. religious groups or economic classes)?
Intervention
The study does not involve any treatment. We only vary the order in which factual beliefs and policy views are measured (some subjects first answer about facts and then about policy views, others the other way around).
Experimental design
We collect data from a sample of the US population. The survey contains questions that allow us to (i) gather general demographic information on the respondents, (ii) elicit the social or economic identity of the respondents, (iii) measure the beliefs of the respondents related to different economic and cultural phenomena, and (iv) measure the respondents’ policy preferences on the same issues. In order to control for potential framing effects, half of the sample is presented the questions on beliefs first, while the remaining half of the sample is presented the questions on policy preferences first.
The answers to the questions allow us to divide the sample in different groups based on identity. Each group is divided in sub-groups, based on the degree of progressiveness or conservatism. We then study how beliefs, policy preferences, and demographics vary between and within each group and sub-group. Furthermore, using an index of the strength of attachment to one’s identity, we look at the relationship between beliefs and policy preferences on the tails of the distribution.
Design details
Cited by
No citation data available for this registration yet.