Research

Journal Articles

Siow, Jeremy. "Bilingual Education Reduces Ethnic Outgroup Discrimination." Journal of Politics. Forthcoming. [Paper][Pre-Analysis Plan][Replication Data and Code]

Extant research on prejudice reduction often focuses on short-term interventions that yield inconclusive results. Instead, this article focuses on people's formative years and argues that bilingual education reduces ethnic outgroup discrimination through perspective-taking. I test my expectations in three distinct studies. In Study 1, cross-national analyses reveal that bilingual speakers are more sensitive to the unequal treatment of other ethnic communities in their country compared to monolinguals. Study 2 leverages a quasi-experimental variation in the language of instruction used in Malaysian schools and finds that individuals who received bilingual education display more inclusive attitudes than those exposed to monolingual instruction. In Study 3, results from a survey experiment indicate that participants who were instructed to switch between two languages, rather than using just one language, expressed lower levels of ethnic outgroup discrimination. The findings imply that education and language play important roles in reducing intergroup prejudice from an early age.

Damann, Taylor J., Jeremy Siow, and Margit Tavits. 2023. "Persistence of Gender Biases in Europe." Proceedings of the National Academy of Sciences (PNAS). 120(12), e2213266120. [Paper][Replication Data and Code]

Prior work suggests that modern gender bias might have historical roots but has not been able to demonstrate long-term persistence of this bias due to a lack of historical data. We follow archaeological research and employ skeletal records of women’s and men’s health from 139 archaeological sites in Europe dating back, on average, to about 1200 AD to construct a site-level indicator of historical bias in favor of one gender over the other using dental linear enamel hypoplasias. This historical measure of gender bias significantly predicts contemporary gender attitudes, despite the monumental socioeconomic and political changes that have taken place since. We also show that this persistence is most likely due to the intergenerational transmission of gender norms, which can be disrupted by significant population replacement. Our results demonstrate the resilience of gender norms and highlight the importance of cultural legacies in sustaining and perpetuating gender (in)equality today.

Working Papers

Siow, Jeremy. "Does Civic Education Improve Ethnic Tolerance?" Under Review. [Paper]

While current research shows that civic education improves multiple political outcomes, its impact on ethnic tolerance remains unclear. Civic education may foster outgroup tolerance by teaching respect for diversity and promoting a shared national identity, but it may be counterproductive when emphasizing ethnic nationalism. I investigate this question by analyzing (1) a reform in Singapore that quasi-randomly assigned students to a new civic education course and (2) the International Civic and Citizenship Education Study (ICCS) data. Both studies reveal no improvement in ethnic tolerance from civic education. Given these findings, I explore whether the effectiveness of civic education depends on various contextual factors, including country-level differences (regime type, centralization, and indoctrination), classroom openness, and students' ethnic backgrounds. Only classroom openness appears to be an important moderator in amplifying civic education's impact on ethnic tolerance. These results highlight the pivotal role of pedagogy in shaping civic education's effects on ethnic tolerance.

Schleiter, Petra, Jonathan Homola, Jeremy Siow, Joelle Tasker, and Margit Tavits. "Does Mandatory Voter ID Suppress Turnout? Empirical Evidence from the UK Voter ID Reform." Under Review.

Mandatory voter identification (ID) laws are increasingly adopted in democracies without national ID systems, sparking debate about their effects on electoral participation. Most prior research focuses on the United States, a uniquely polarized context with a history of racialized voter suppression, thus limiting the generalizability of the findings. We investigate the electoral consequences of the voter ID reform in the United Kingdom, a less polarized democracy, through three studies: (1) local-authority-level turnout comparisons across English and Northern Irish local elections from 1993 to 2023, (2) individual-level turnout changes using the British Election Study Internet Panel (2014-2025), and (3) a regression discontinuity design exploiting an age-based threshold in photo driving license ownership. Across all three studies, we find consistent turnout reductions in the short term. However, these effects dissipate over time, most likely due to voters switching to postal voting. Our findings underscore the need for safeguards protecting electoral access for all voters.

Schleiter, Petra, Jonathan Homola, Jeremy Siow, Joelle Tasker, and Margit Tavits. "Does Promoting Free Voter ID Increase Participation Among Voters Without Qualifying Identification? An Experimental Evaluation." Under Review.

When voter identification (ID) laws are introduced to enhance electoral integrity, they risk disenfranchising voters who lack qualifying documents, particularly in countries without universal national ID. This paper offers a novel study of a policy designed to mitigate these adverse effects: official information promoting free voter ID. Using a preregistered survey experiment conducted during the 2024 UK general election, the first to require ID nationwide, we study a hard-to-reach population of 1,005 eligible voters without qualifying ID. Participants were randomly assigned to receive official Electoral Commission information about the free Voter Authority Certificate (VAC). The intervention primarily improved perceptions of voting accessibility but did not reliably increase willingness to apply for the VAC and turnout intentions. These findings suggest that official information about free voter ID alone, without additional mitigating measures, has limited effectiveness in promoting access and participation.

Siow, Jeremy, Minh D. Trinh, and Tony Zirui Yang. "Transnational Propaganda and Ethnic Polarization: Experimental Evidence from Malaysia." [Paper][Pre-Analysis Plan]

Authoritarian regimes increasingly engage in transnational information campaigns to expand their political and cultural influence beyond national borders. We argue that such operations can exacerbate ethnic polarization in target countries, specifically between diaspora communities and ethnic majority populations. Drawing on Social Identity Theory, we theorize that transnational propaganda exacerbates affective polarization by reducing intergroup social and political tolerance, while also increasing polarization in foreign policy preferences in racially diverse settings. To test these expectations, we conduct a preregistered survey experiment in Malaysia and examine whether China's transnational propaganda affects intergroup and foreign policy attitudes among ethnic Chinese and Malays. The results show that China's propaganda decreases Malays’ social tolerance toward Chinese Malaysians while strengthening Chinese Malaysians' -- but not Malays' -- agreement with China's foreign policy positions in the region. These findings highlight how transnational propaganda can undermine social cohesion and create divergent policy preferences within multicultural societies.

Liu, Amy H., Jeremy Siow, Charles K.S. Wu, and Yao-Yuan Yeh. "The Sound of Exclusion: Foreign-Accented Mandarin Speakers and Welfare Chauvinism in Taiwan."

Do foreign accents induce welfare chauvinism? Prior work has investigated the exclusionary impact of visual and auditory cues, but it remains uncertain whether these effects stem from the perceived foreignness of the cues or are limited to certain outgroups. Our research addresses this question by focusing on accents. We argue that natives should become less supportive of extending social benefits to immigrants when exposed to a foreign accent associated with a politicized origin. We conducted two experiments using different Mandarin accents on Taiwanese nationals in Taiwan and the US. Our findings reveal that participants in Taiwan exhibited increased welfare chauvinism when hearing an accent from a politicized foreign origin -- the People's Republic of China. However, we found no evidence that the foreignness of an accent influenced attitudes towards immigration policies. These results underscore the importance of considering the political context and accent origin when studying their impact on immigration attitudes.

Moy, Bryant and Jeremy Siow. "Who's at the Cutoff? Characterizing the Analytical Sample of Regression Discontinuity Designs."

Regression discontinuity designs rest on local comparison, yet most studies do not describe the units that determine their estimates. When researchers omit that description, readers cannot know the empirical scope conditions or judge how far the findings travel. This failure to address scope conditions can stall theory development and misguide policy advice. We outline a procedure that computes kernel distance-weighted covariate means for near-cutoff units in order to describe the sample. We demonstrate our approach by re-analyzing three published studies using sharp, multi-cutoff, and fuzzy designs. In each case the analytical sample differs from the population the study purports to inform. Characterizing scope conditions facilities proper contextualization and a broader theory-informed discussion of results. An accompanying R package implements the calculations. The procedure makes sample composition explicit, clarifies empirical scope conditions, and strengthens the descriptive foundation on which causal inference rests.

Ong, Elvin, Jeremy Siow, and Tricia Yeoh. "Public Opinion Towards Political Reforms and Compromise in a New Democracy." [Pre-Analysis Plan]

In new democracies, democratic consolidation necessitates political reforms that even the playing field for democracy. What shapes public opinion towards such democratizing reforms? We conducted a survey experiment in Malaysia, a new democracy since 2018, to test the effect of framing on public opinion towards democratizing reforms, and provide further observational evidence on public opinion towards proposed compromise solutions. We document two findings: First, public opinion towards democratizing reforms is sensitive to perceptions of threat to the political dominance and privileges of ethnic groups under the previous authoritarian regime. Such perceptions can originate from frames of direct threat or indirect loss. Second, proposed compromise solutions can increase support for democratizing reforms, but only for those reforms which have divided public opinion. Where reforms enjoy broad support, proposed compromise solutions depress support for reforms. Our results contribute to the renewed scholarship on the challenges of democratization, democratic consolidation, and precarity of backsliding in new democracies.

Siow, Jeremy. "Short-Term Pains, Long-Term Gains: Education and Economic Effects of Malaysia's Language-of-Instruction Policy Reform."

There is a longstanding debate about whether instruction in the mother tongue or colonial languages yields better outcomes in postcolonial societies. Critics argue that teaching in a colonial language imposes learning costs and worsens academic performance, while proponents contend that exposure to a globally relevant language like English improves long-term socioeconomic outcomes. This article offers evidence on this question by studying a reform in Malaysia where English replaced mother-tongue instruction for Mathematics and Science. Study 1 examines the reform's short-term impact on academic achievements using data from the Trends in International Mathematics and Science Study (TIMSS), while Study 2 analyzes longer-term socioeconomic outcomes through an original survey. Both studies reveal that students who switched to English instruction initially experienced lower test scores but subsequently achieved better educational and occupational outcomes, especially among Malays. These findings suggest that learning in a non-native language creates temporary growing pains that yield future gains.

Page updated

Google Sites

Report abuse