Replication Data for: Polling Across Borders: The Promise and Pitfalls of Convenience Samples in a Cross-National Context

Harvard Dataverse · 2026-01-06

This study evaluates the performance of two widely used survey platforms, Lucid and Morning Consult, across six diverse national contexts: Brazil, India, Japan, Nigeria, the Philippines, and the United States. We assess the impact of platform choice on sample composition, response quality, political judgments, and treatment effect estimation, focusing on a randomized corruption treatment embedded within the survey. Attention filter passage rates were similar and generally high across countries and platforms, while the percentage of high frequency survey-takers varied greatly across countries. Our findings reveal significant demographic skews, with both platforms consistently overrepresenting college-educated respondents. Despite these differences, political assessments and estimated average and heterogeneous treatment effects remain broadly consistent across platforms, and in Brazil our estimates largely tracked those from past research with a probability sample. We find some evidence of cross-national variation in the magnitude of treatment effects, but these differences were often platform-specific. These results suggest that convenience samples can provide reliable estimates of causal effects even in diverse contexts. Taken together, our research highlights the trade-offs between cost, speed, and representativeness in global public opinion research, offering insights into the challenges and opportunities of online survey platforms.

Sustainability benefits and science knowledge shape consumer preferences for novel smartphone battery technology in the United States

Energy Research & Social Science · 2026-05-09

Public attitudes toward COVID-19 vaccination: The role of vaccine attributes, incentives, and misinformation

UNC Libraries · 2025-06-12

Legal constraint through political means? Legal foundations and public support for executive action

The Journal of Law Economics and Organization · 2025-02-27 · 1 citations

Abstract Many scholars question the extent to which presidents are legally constrained, but others argue that public opinion provides an indirect but important mechanism through which law checks unilateral power. Through thirteen survey experiments, we examine whether the legal foundations of executive action—whether framed as pursuant to delegated statutory authority or contra the will of Congress—affect public support for unilateralism. Legal frames can influence public support, particularly among those with the strongest attachments to the rule of law. However, these effects are highly concentrated in hypothetical vignettes or temporally distant cases. Legal frames have little effect on support for executive action by recent presidents, even when they shape public perceptions of an action’s legality. Our results inform debates about the conditions under which public opinion might serve as a backstop against democratic backsliding by checking presidential overreach, and the role of law in shaping public debates about presidential power.

Dynamic Democracy: Public Opinion, Elections, and Policymaking in the American States. By Devin Caughey and Christopher Warshaw. Chicago: University of Chicago Press, 2022. 248p. $95.00 cloth, $30.00 paper.

Perspectives on Politics · 2024-05-06

Dynamic Democracy: Public Opinion, Elections, and Policymaking in the American States. By Devin Caughey and Christopher Warshaw. Chicago: University of Chicago Press, 2022. 248p. 30.00 paper. - Volume 22 Issue 3

Replication Data for: The Potential Impact of Emerging Technologies on Democratic Representation: Evidence from a Field Experiment

Harvard Dataverse · 2023-02-15

Advances in machine learning have created natural language models that can mimic human writing style and substance. Here we investigate the challenge that machine generated content such as that produced by GPT-3 presents to democratic representation by assessing the extent to which machine-generated content can pass as constituent sentiment. We conduct a field experiment in which we send both hand-written and machine-generated letters (a total of 32,398 emails) to 7,132 state legislators. We compare legislative response rates for the human versus machine-generated constituency letters to gauge whether language models can approximate and scale up inauthentic constituency voice. Legislators were only slightly less likely to respond to AI-generated content than to human-written emails; the 2% difference in response rate was statistically significant but substantively small. Qualitative evidence sheds light on the potential perils that this technology presents for democratic representation, but also suggests potential techniques that legislators might employ to guard against AI-sourced astroturfing.

Treaty obligations and support for collective defence: Evidence from Italy after the invasion of Ukraine

European Journal of Political Research · 2023-12-15 · 5 citations

Abstract Whether countries and their publics are responsive to the international legal commitments they make is the source of long‐standing academic debate. Russia's February 2022 invasion of Ukraine brought real‐world significance to these debates. While Ukraine is not a member of the North Atlantic Treaty Organization (NATO), the invasion raised the prospect that other NATO members could be targeted and that Article 5 collective security commitments would be invoked. While recent research suggests that emphasizing mutual defence treaties can increase public support for defending an ally, prior work focuses on US opinion in a less fraught political environment. We constructed and fielded a survey experiment in Italy in the initial weeks of the Ukraine invasion to probe support for defending a NATO ally, the relevance of the Article 5 legal commitment on support for defending an ally, and the potential moderating influence of gender and political party. Our findings show that the Article 5 commitment significantly increased support for defending an ally. Consistent with past research, we find a significant gender gap, with men being more supportive of defending an ally than women; however, both men and women responded to the Article 5 commitment to virtually the same degree. The estimated treatment effect was larger for supporters of right‐wing parties than for the left; however, the difference was not statistically significant.

Executive politics in an era of democratic crisis

Presidential Studies Quarterly · 2023-05-29 · 3 citations

The author declares no conflict of interest.

Replication Data for: Treaty Obligations and Support for Collective Defense: Evidence from Italy after the Invasion of Ukraine

Harvard Dataverse · 2023-07-24

Whether countries and their publics are responsive to the international legal commitments they make is the source of long-standing academic debate. Russia’s February 2022 invasion of Ukraine brought real-world significance to these debates. While Ukraine is not a member of the North Atlantic Treaty Organization (NATO), the invasion raised the prospect that other NATO members could be targeted and that the Article 5 collective security commitments would be invoked. While recent research suggests that emphasizing mutual defense treaties can increase public support for defending an ally, prior work focuses on U.S. opinion in a less fraught political environment. We constructed and fielded a survey experiment in Italy in the initial weeks of the Ukraine invasion to probe support for defending a NATO ally, the relevance of the Article 5 legal commitment on support for defending an ally, and the potential moderating influence of gender and political party. Our findings show that the Article 5 commitment significantly increased support for defending an ally. Consistent with past research, we find a significant gender gap, with men being more supportive of defending an ally than women; however, both men and women responded to the Article 5 commitment to virtually the same degree. The estimated treatment effect was larger for supporters of right-wing parties than for the left; however, the difference was not statistically significant.