Read my research statement (click here).
Dissertation: Citizen and Elite Responses to Ethnic Representation
- Why is there so much variation in the ethnic composition of cabinets? (Published in Politics, Groups, and Identities)
AbstractWhy is there so much variation in the ethnic composition of cabinets? Previous work has focused on cabinet policy alignment, largely overlooking the role of identity-based characteristics like ethnicity. I theorize that country leaders make ethnic cabinet appointments to gain political support, both when country-level ethnic diversity is high and when ethnic groups rely on leader decisions in order to receive resources. Cabinet appointments offer leaders a way to distribute resources while creating a credible commitment to remove these resources if ethnic group support wanes. I introduce a new cross-national time-series dataset of cabinet minister names from 149 countries from 1967 to 2017. Using novel methods from computer science, I code the ethnicity of cabinet ministers’ names and construct a cabinet diversity index for each country-year. After validating this measure, I find support for my hypothesis. Country leaders in reliant societies increase ethnic cabinet appointments seven to twenty-five percent over leaders in non-reliant societies. The results provide the first large scale cross-national analysis of the strategic ways in which leaders use cabinet seats to manage the distribution of resources.
- Can country leaders improve citizens’ views of the ethnic outgroup when making political appointments? (Under Review)
AbstractCan country leaders improve citizens’ views of the ethnic outgroup by changing the ethnic composition of their government? Years of pressure from the international community calling for leaders to ethnically diversify their governments — particularly cabinets — seems to suggest that ethnic representation is key to improving citizens’ views of the outgroup. I argue that citizens’ views improve when their level of ethnic representation increases. Increased ethnic representation provides perceived benefits to coethnic citizens which lead to a more favorable view of ethnic outgroups. When non- dominant group ethnic representation increases, non-dominant citizens’ views of the outgroup will become more favorable, while dominant citizens’ views will worsen. Using a vignette experiment with ethnic Albanians and Macedonians in North Macedonia, I show that ethnic representation does not provide the improvements in outgroup relations that many have hoped. Both groups’ a↵ect toward and perceptions of the cabinet change somewhat, but changing ethnic representation does not alter overall outgroup attitudes.
- How do political elites respond to ethnic diversity?
AbstractHow do political elites’ views of ethnic outgroups change in response to increased ethnic diversity? Political elites serve critical roles as elected representatives and public figures such that is is not clear whether existing work on the effects of ethnic diversity among citizens can be extended to include elites. I argue that political elites work in a competitive environment wherein increased ethnic diversity promotes ethnic animosity and worsens views of ethnic outgroups. Frequent interethnic contact can serve to counteract this relationship, but its effects may be stifled by rhetorical ethnic cues. Using new data and a survey experiment with local government committee members in India, I show that neither increased committee diversity nor ethnic cues negatively impact elites’ views of the outgroup while interethnic contact improves outgroup views. Increased ethnic diversity is, therefore, a promising avenue for improving views of the outgroup if leaders can encourage interethnic contact between elites.
- What impact do ethnically diverse political elites have on substantive representation?
AbstractWhat impact does an ethnically diverse political elite have on substantive representation? I design the first study to test the effect of ethnic diversity on substantive representation impacting citizens. Focusing on cabinet appointments, I argue that increasing ethnic cabinet diversity results in a power struggle that cabinet ministers attempt to win by providing substantive representation. I test this theory using state cabinet data and social welfare spending from India, where an exogenous shock to cabinet size provides an instrument for cabinet diversity. The shock circumvents the cabinet leader’s role in strategically appointing cabinet ministers, thus testing my intra-cabinet power hypothesis. Difference-in-differences results show that increasing cabinet diversity has a positive effect on social welfare spending, supporting the hypothesis. I suggest that individual cabinet ministers have a more important role in providing substantive representation than previously thought and are not always co-opted by the cabinet leader.
- Who participates in public meetings to increase ethnic representation?
AbstractWhen do citizens participate in public meetings aimed at increasing ethnic representation? Existing literature emphasizes how citizens vote to increase ethnic representation, but voicing opinions at public meetings are an alternative way to prompt change. I argue that citizens who value substantive representation over descriptive representation will be particularly motivated to participate in public meetings to increase ethnic representation. Drawing on an original survey about community meeting participation in North Macedonia, I find that ethnic Macedonians act as expected, whereas non-dominant Albanians who value descriptive representation are more likely to participate in public meetings. I confirm the non-dominant ethnic group findings using survey data on consultation and protest participation among New Zealand Maori. These results show that political participation about representation is not necessarily motivated by citizens demanding substantive representation. Country leaders must understand heterogeneity in citizen motivations in order to effectively respond to public meeting comments and political participation more generally.
Violence and Critical Events
- Is Muslim radicalization a byproduct of low socioeconomic status or identity threat? (Published in British Journal of Political Science)
AbstractDoes anti-Muslim rhetoric by Western politicians breed radical attitudes among European Muslims? We explore this question by conducting an experimental study in Bosnia - a European democracy, where, unlike the rest of Europe, Muslims are neither immigrants nor socio-economically disadvantaged. This helps us clearly identify the radicalization potential of Western rhetoric alone, absent contextual factors such as social inferiority. Experimental evidence with Bosnian Muslims from five surveys (with a total N=2,608) suggests that rhetorical attacks on Islam by Western politicians do not strengthen individuals’ Muslim identity, cause higher levels of animosity toward the West, or lead to condoning the use of violence. We also find that pro-Muslim rhetoric, while increasing positive views of the West, does not affect individuals’ strength of Muslim identity or their radical sympathies. These results provide important implications for our understanding of sources of radicalization and for efforts to curb radical tendencies.
- What is the relationship between natural resources and conflict? Meta-analysis. (Published in Research and Politics)
AbstractThe relationship between natural resource wealth and civil conflict remains unclear, despite prolonged scholarly attention. Conducting a meta-analysis — a quantitative literature review — can help synthesize this broad and disparate field to provide clearer directions for future research. Meta-analysis tools determine both the aggregate effect of natural resources on conflict and whether any particular ways in which variables are measured systematically bias the estimated effect. I conduct a meta-analysis using sixty-nine studies from sixty-two authors. I find that there is no aggregate relationship between natural resources and conflict. Most variation in variable measurement does not alter the estimated effect. However, measuring natural resource wealth using Primary Commodity Exports and including controls for mountainous terrain and ethnic fractionalization all do significantly impact the results. These findings suggest that it may be worth exploring more nuanced connections between natural resources and conflict instead of continuing to study the overall relationship.
- How does the presence of natural resources influence wartime casualties? (Published in Journal of International Studies)
AbstractThe 1992 to 1995 Bosnian War was not a war begun over a conflict for natural resources. Instead, this study hypothesizes that the intensity of fighting during the War was positively influenced by the presence of high levels of natural resource wealth distributed throughout Bosnia. Drawing upon self-coded data for myriad measures of the intensity of fighting and natural resources in a given area of the country, we use multiple regression techniques as well as factor analysis to support the hypothesis and conclude that natural resource wealth was an important factor in influencing the course of the War. Natural resources both strategically deprived the opposing army from adequate land and water during the War and also ensured that high quality natural resource infrastructure and forest resources were available after the War ended. This research is unique in that it examines the importance of natural resources where fighting takes place.
- Do wealthy regions actively attempt to become more autonomous in order to keep resources for themselves? (Published in Caucasus Survey)
AbstractPrevious models of individuals’ preferences for decentralization have focused on either economic or identity based motivations. In some cases, however, elites in rich regions with high inequality prefer decentralization contrary to their economic preferences. This paper proposes a model to explain these instances by focusing on the preferences of self-interested elites who manipulate a strong regional identity for personal gain. I develop the hypothesis that a strong regional economy, combined with a regional identity, provides the incentive for and the mechanism by which elites prefer decentralization. I test this hypothesis using the puzzling case of Adjara, Georgia, a rich region with high inequality and a strong regional identity where elites, contrary to expectations, pushed for decentralization. Results suggest that regional identities are used by elites in order to seize on a strong economy for personal gain. This motivates future study on the interaction between regional economies and identity.
- What impact does tourism, particularly small regional airport infrastructure, have on regional government demands for decentralization? (Published in Tourism Economics)
AbstractThe relationship between increased tourism and increasing regional autonomy is quite nuanced and understudied. This paper hypothesizes that only an increase in both regional air traffic and in international tourism will impact the level of regional autonomy. Using the period after ten countries were admitted to the European Union in 2004 as an example of a dramatic tourism increase, the paper finds that countries with increased tourism, but without regional airports, did not experience a sudden increase in their regional autonomy. In Poland, however, the large number of growing regional airports and increased tourism did provoke regions to argue with the central government for more regional autonomy. These findings contribute to a better understanding of how international interventions impact regional decentralization preferences.
- Are unelected officials differentially responsive to critical events? (Published in British Politics)
AbstractWhat motivates Peers to attend legislative sittings? Sitting attendance is a symbolic way for legislators to show citizens that they are being productive and hence is often explained by electoral motivations that Peers lack. I argue that Peers make decisions to attend sittings when critical events threaten their position in the legislature. Attending at these times --- namely after scandals and House of Lords reform debates --- is an attempt to counteract negative impressions about the House and its members. Other critical events that may impact elected legislators such as terrorist attacks and natural disasters should have no impact on Peers attendance. Using a newly compiled dataset on attendance and critical events, I show that Peers respond by increasing attendance only after House of Lords reform debates in either House; attendance after scandals, natural disasters, and terrorist attacks is unchanged. This suggests that Peers are responsive in only the most urgent cases: when they are in the spotlight and the future of the House is on the line. More broadly, I offer the first empirical investigation of symbolic responsiveness among unelected legislators and show that there are some situations where said legislators feel the need to respond.
- How does the public talk about autonomous regions compared to de facto states? (Under Review)
AbstractPublic discourse about autonomous regions and de facto - or unrecognized - states plays an important role in the policies parent states adopt to respond to these entities. I theorize that public discourse about autonomous regions will underscore commonalities with the parent state. Public discourse about de facto states will treat the state as a separate entity. Changes in the relationship between the separatist entity and parent state may also alter public discourse. I employ newspaper data as a measure of public discourse about Adjara and Abkhazia, Georgia. Using sentiment analysis and topic models, I show that public discourse about autonomous regions discusses domestic politics whereas discourse about de facto states emphasizes international relations. This analysis of public discourse offers insights into the ways in which parent state citizens discuss separatist entities and how leaders may wish to steer public discourse about these regions in the future.
- How do country leaders use electoral system change to influence the international community’s evaluation of election fraud? (Under Review)
AbstractWhat prompts leaders in post-Communist states to alter electoral system proportionality? We propose that the international community rewards leaders for increasing proportionality. Our formal model characterizes leaders as either truly interested in democratic reforms or pseudo-reformers, who increase proportionality to receive international community benefits while engaging in electoral fraud. We hypothesize that the international community will be more (less) likely to detect fraud when leaders decrease (increase) proportionality regardless of whether fraud actually occurs. We first test our hypotheses in Ukraine, finding that fraud occurs, but that the international community is less likely to detect fraud following an increase in proportionality and vice versa. Using cross-national data from 20 Eastern European countries, we find again that decreased proportionality is associated with more fraud reports. We suggest that the international community relies too heavily on proportionality as a signal of election quality, frequently missing fraud in more proportional systems.
Measurement and Event Data
- Does the process of collecting news influence how riots are reported? Newspaper text analysis in India. (Published in Asian Journal of Communication)
AbstractRiot reporting is one aspect of newspaper coverage that can drive people into the streets in acts of collective protest or violence. Media observers and scholars have proposed that the language of Indian newspapers, be it English or vernacular, partially dictates the kinds of riot events reported and the quality of those reports. I tested whether this conventional wisdom holds by investigating the content of Indian riot coverage in the English Times of India and Hindu Hindustan. While Hindustan emphasized official statements and interviews with political parties, neither newspaper accurately represented the actual number of riots in their reporting. In fact, coverage in both papers followed predictable patterns likely driven by a new focus on selling newspapers at any cost in order to increase advertising revenue. This study contributes to a growing literature highlighting the similarities between media outlets and the degree to which their reporting is removed from actual events.
- Measuring protest event size using social media reports. (Published in Studies in Indian Politics)
AbstractEvents are a form of geographically and temporally focused collective action aimed at making broad social or political claims. Event attendance matters for how much impact the event has, but estimating the number of event attendees is quite difficult. Political scientists have recently developed methods for detecting event size using social media data from technology-based sources. These methods have been used to estimate event size in some contexts, but the Indian case presents special challenges that make using technology-based data to estimate event size particularly difficult. Drawing on fieldwork during the 2019 Citizenship Amendment Act protests in Delhi, I find that strategic choices on behalf of both protesters and the government made estimating event size using technological data quite challenging. I then discuss some ways that event size measurement techniques can be adapted for the Indian context.
- How do two sources of riot information - newspapers and government data - compare?
AbstractHow does newspaper-based event data compare to a government data source? While scholars have long recognized the importance of and biases present in newspaper-based event data, few studies have been able to compare newspaper reports with official government data to better understand the severity and impact of such biases. We develop this comparison in the context of riots, a form of violent collective action that represents an important middle ground between peaceful protests and protracted civil conflict. Using newly collected police precinct-level government data from India, we compare these data to a high-quality newspaper source. Though similar at the aggregate level, newspaper riot reports correlate poorly with government data at the local level. In order to better understand the discrepancies between these two sources, we develop models to explain the frequency of newspaper and government riot reports based on literacy, location, and other demographic characteristics. Literacy and location influence newspaper riot reports, but not government riot reports. We conclude that newspaper riot data does partially reflect aggregate riot trends, but the newspaper editorial process also plays an important role. Government data is better for within country comparisons and for analyzing event trends over time. Our findings suggest that using collective action event data from both sources may help ensure that results are not driven by biases in either data source.
- What is an effective and transparent way to code individuals’ caste?
AbstractHow diverse are political elites? Elite caste diversity has the potential to ease interethnic tensions and to provide substantive representation to minority groups in India, the world’s largest democracy. Yet, we lack reliable and replicable methods for coding caste diversity and for using this information to calculate a diversity measure for a group of elites. Prior work has utilized time consuming archival methods or various name-based approaches, but has not considered the strengths and weaknesses of these techniques. We develop and apply a caste coding protocol which blends name-based caste coding with archival work. Using an application with Indian state cabinet ministers, we show that our method simplifies the ethnic categorization process while still retaining accuracy. We compare our method with three name-based approaches to caste coding and identify the most appropriate situations in which to use each method. The balance of accuracy and efficiency of our method helps better equip scholars to more easily study questions about caste diversity among a wider array of political elites. Our results have implications for how comparative politics scholars conceptualize and measure ethnic diversity.
- Defining political space in competitive authoritarian regimes. (Accepted with Springer Verlag)
AbstractCompetitive authoritarian regimes are those in which electoral competition is allowed, but elections are not free and fair. Dozens such regimes exist around the world, but their political space is unexplored. We hypothesize that voters in these regimes define political space in two dimensions, where these dimensions are different from those in developing democracies. Using World Values Survey data from Kyrgyzstan in two time periods, we show the development of voter preferences along two dimensions: fondness for tradition and trust of political institutions. We explore why parties faced with this political space do not run on these issues even though their party platforms nominally try to appeal to them. Finally, we argue that these findings extend to regimes of a similar type by mapping political space in Kazakhstan, Georgia, and Hungary. Providing incentives for parties to run on their stated platforms may help institutionalize political competition in competitive authoritarian regimes.
- Assessing unexpected responses among Indian Amazon Mechanical Turk respondents. (Under Review)
AbstractWhat can researchers do to address unexpected survey and experimental responses on Amazon Mechanical Turk (MTurk)? Much of the unexpected responses problem has been traced to India, and several survey and technological techniques have been developed to detect foreign workers accessing U.S.-based surveys. We survey Indian MTurkers and find that 26% pass common, survey-based methods of detecting foreign workers. Additionally, 3% of our respondents claim to be located in the United States, implying that these respondents circumvent IP address blocking techniques designed to prevent foreign workers from accessing U.S.-based surveys. Foreign workers have strong monetary incentives for seeking out U.S.-based MTurk tasks. We suggest that restricting respondents to Master Workers and removing the U.S. location requirement is a cost-effective way to more effectively filter out unexpected responses.