Data and Code for: Can incorrect beliefs about the racial composition of welfare and unemployment insurance beneficiaries be changed

Principal Investigator(s): View help for Principal Investigator(s) Jeffrey Carpenter, Middlebury College; Jakina Debnam Guzman, Amherst College; Peter Hans Matthews, Middlebury College; Erin Wolcott, Middlebury College

Version: View help for Version V1

Name	File Type	Size	Last Modified
Carpenter_et_al_2025_Papers_and_Proceedings.do	text/plain	6.6 KB	02/12/2025 10:49:AM
Carpenter_et_al_2025_Papers_and_Proceedings.dta	application/x-stata-dta	405 KB	02/12/2025 10:49:AM
Carpenter_et_al_2025_Papers_and_Proceedings_Experimental_instructions_appendix.pdf	application/pdf	955.5 KB	02/12/2025 10:49:AM
README.pdf	application/pdf	80.1 KB	06/10/2025 07:23:AM

Project Citation:

Carpenter, Jeffrey , Guzman, Jakina Debnam, Matthews, Peter Hans, and Wolcott, Erin. Data and Code for: Can incorrect beliefs about the racial composition of welfare and unemployment insurance beneficiaries be changed. Nashville, TN: American Economic Association [publisher], 2025. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2025-06-10. https://doi.org/10.3886/E219150V1

Project Description

Summary: Some argue that support for the social safety net in the United States is influenced by beliefs about the beneficiaries’ race. Information treatments have the potential to change these beliefs, but for them to be policy relevant, their effects must last beyond the intervention. Our findings from two parallel experiments that exploit the different racialized histories of welfare and unemployment insurance indicate that racial beliefs do predict stated support for the racially stigmatized welfare program but not for the less stigmatized unemployment program. We also find these beliefs are stable if uncorrected and that they can be persistently corrected.

Funding Sources: Amherst College; Middlebury College

Scope of Project

Subject Terms: Information; Experiment; Race; Labor; Redistribution; Welfare; unemployment insurance

JEL Classification:
      D90 Micro-Based Behavioral Economics: General
      J08 Labor Economics Policies
      J15 Economics of Minorities, Races, Indigenous Peoples, and Immigrants; Non-labor Discrimination

Geographic Coverage: United States

Time Period(s): 6/10/2024 – 8/6/2024

Collection Date(s): 6/10/2024 – 8/6/2024

Universe: A nationally representative sample of adults living in the United States. Sample participants engage with the experiment through Connect, a crowdsourcing platform run by CloudResearch which provides participants for online surveys.

Data Type(s): experimental data

Collection Notes: This dataset includes some of the data collected in each of two parallel experiments studying the causal effects of providing information about the number of Black people receiving benefits from two labor market policies (Temporary Assistance for Needy Families and Unemployment Insurance) on support for these policies.

Methodology

Response Rate:

The experiment proceeded in three stages. In the first stage, of the 3,029 participants who began the experiment, 3,022 completed the experiment. 94% of the participants who completed Stage 1 returned to complete Stage 2 of the experiment. 82% of the Stage 2 participants returned to complete the third stage of the experiment.

Sampling: The data collection proceeded in three stages. In the first stage, a nationally representative sample of U.S. adults was collected through Connect, a crowdsourcing platform run by CloudResearch. We collected the sample using a "census matched template” that targets a sample matched to the U.S. census along basic demographic characteristics. For the remaining two stages, first stage participants were invited back to participate again.

Data Source: Data collected by the authors within a novel online information experiment.

Collection Mode(s): web-based survey

Scales: Data contain results of one Likert-style question: "Which statement best describes you?" With a scale that ranges from 0 ("I prefer Black people to White people") to 10 ("I prefer White people to Black people), with the middle of the scale anchored at "I like White people and Black people equally". The results of this question are recorded in the variable "explicit_binary".

Weights: We collected the sample using a "census matched template” within the Connect research platform. This setting targets a sample matched to the U.S. census along basic demographic characteristics.

Unit(s) of Observation: Individuals

Geographic Unit: Whether the participants live in a southern state in the U.S. or in a northern state in the U.S.

Related Publications

Download this project

Published Versions

V1 [2025-06-10]

Export Metadata

Dublin Core

DDI 2.5

Report a Problem

Found a serious problem with the data, such as disclosure risk or copyrighted content? Let us know.

This material is distributed exactly as it arrived from the data depositor. ICPSR has not checked or processed this material. Users should consult the investigator(s) if further information is desired.