AERoPUP Program

There is some preliminary research that suggests the importance of social capital in affecting mortality trends related to “deaths of despair” but this has not been examined empirically.

I will implement a multi-level model to understand the relationship between various measurers of social capital—in particular religious social capital—and mortality across demographic, regional, and educational cross-sections of the U.S. CDC mortality data will provide measures of mortality across both various causes of death and population cross-sections. I will merge the mortality data with social capital proxies from the Gallup Daily Survey, DDB Needham Life Style Surveys, and U.S. Census. I will draw on the Bureau of Labor Statistics, Census Bureau, and National Cancer Institute’s Surveillance, Epidemiology, and End Results Program for economic and demographic controls.

While there is broad evidence that prompts (e.g., events, advertising) which provoke emotional responses in individuals can correlate with political engagement, there is little work on the mechanisms underlying this process. Moreover, it is unlikely that aggregate data will permit us to get very far in mapping these relationships.

This study devises an experiment in which a set of prompts are offered to experimental subjects, and their emotional responses are observed on a variety of dimensions (anger, anxiety, etc.). Another set of questions ascertains the subject’s level of political participation (both actual and contemplated). At this stage of research, it is not at all clear whether emotion moves participation, or whether past/contemplated participation shapes emotional reaction to given stimuli. My hope is that the data gathered in this survey experiment may shed some light on several questions, e.g., whether there are measurable differences between self-identified Republicans vs Democrats in the patterns which emerge.

A combination of conjoint and framing experiments will be employed to understand American citizens’ multidimensional preferences on immigration policy reforms. I have carefully followed the current immigration policy debate in the U.S., and I seek to understand the degree to which respondents respond to various components (i.e., specific issues) of the debate as well as how the description of immigrants -- as unauthorized, illegal, or undocumented immigrants -- might influence American citizens' policy preferences.

Although many studies are examining multidimensional preferences on immigrants based on conjoint analysis, to the best of my knowledge, this study is the first to apply conjoint analysis to understanding respondents' immigration policy preferences.

Americans express support for institutions playing an active role in promoting human rights, but previous observational studies are prone to desirability biases. The debate over how to respond to Jamal Khashoggi’s assassination by Saudi Arabia, suggests considerable uncertainty about the public’s support for human rights promotion. In this research project, we explore the extent to which the American public cares about human rights in other countries by conducting a survey with three embedded experiments for testing support for human rights promotion and the salience of the issue for the public. The survey will utilize a vignette treatment, a conjoint experiment, and mouse tracking experiment in an online survey of a national-representative academic sample collected by Luc.id.

This research will provide experimental evidence for public support (or lack thereof) for human rights promotion in the context of international trade policy and sanctions. By engaging a nationally representative sample, I hope to further explore the topics set forth by Carnegie and Gaikwad’s (2017) and extend on this research by including human rights. This method will not only provide a more solid foundation for understanding current public opinion on international human rights, but also of how they are likely to react to the tradeoffs faced in making foreign policies related to human rights.

Though studied for decades, different authors, each with high-quality data sets and rigorous methodologies, have purported to offer differing dispositive claims on the impacts of majority-minority districts on minority substantive representation. This study contends that the previous literature has suffered from four different problems, 1. The heterogeneity of individual districts, 2. Important theoretical impacts beyond the raw partisan vote shares, 3. Election-specific factors that limit generalizability, and, most importantly, 4. Selection bias into which states create majority-minority districts. To solve these problems, this study exploits a rare bit of randomness in the states that faced pressure to create majority-minority districts, under the test established by the Supreme Court in Thornburg v. Gingles.

All states must (re)define representational districts on a regular basis both for their state legislatures and their Congressional seats. Observing elections between 1982 and 2016 across all states, we can see variation in the competitiveness of elections for state-level as well as Congressional positions. Since there is variation in whether states have a) engaged in Independent Redistricting; b) when it was instituted; c) the design of the Independent Commission, we have the makings of a quasi-experiment which will permit us to discover whether there are measurable differences within states over time and between states who have and have not adopted Independent Redistricting Commissions.

Are law-makers vulnerable to manipulation by the framing of issues when they are deciding how to vote on a piece of legislation? Using a survey experiment instrument with both state-level and federal legislators as subjects, this study seeks to determine whether this is true. Following the work of Kahneman and Tversky, which has found empirical support in several settings, experiments constructed will test hypotheses that legislators are more willing to support “risky” proposals when an issue is framed in the domain of losses. If the hypotheses are supported, the results will suggest a) legislators can make mistakes like anyone else, but more consequentially b) they are vulnerable to lobbyists who frame issues in this particular way, and c) may vote for bad policies based more on framing than merits. While no formal model is offered here, the results could feed into existing formal models which employ Prospect Theory, or possibly a formal model specifically applied to legislator voting preferences.

Fears within the U.S. electorate of declining living standards, possible job loss, as well as perceptions of crime and threats to homeland security tied to illegal immigration created a solid base for the candidacy of Donald Trump in 2016. Using a factorial vignette survey, this study seeks to gain insight into the relative weight given by voters to these two sources of anxiety in their decision calculus.

A highly debated question within literature on education administration is whether efforts to reduce segregation by school zoning has been successful. Largely absent from this debate thus far are: 1) consideration of the role played by income segregation (not only racial segregation) and 2) application of more recent spatial/GIS techniques to a debate in which space plays a central role. This study employs shapefile data from the National Center of Education Statistics and the U.S. Census Bureau to offer new insights into the debate surrounding school zoning.

Three randomized survey experiments will be administered to large samples of French natives to ascertain whether high vs low contact with immigrants affects: 1) their acceptance of false information in certain French regarding immigrants; 2) the extent to which immigrants are perceived as ‘threatening’; 3) characteristics of the types of immigrants considered ‘acceptable’

Despite winning nominal independence from European powers (decades ago, in several cases) this study searches for systematic patterns between the ‘imperial variants’ of particular European colonial powers and institutional features – such as corruption and racist practices – found within many of these post-colonial states today. One important contribution is creation of a simple measure of ‘mercantile’ vs ‘settler’ colonial style, which permits observation of variation in governance within metropoles, not simply between metropoles (e.g., Britain vs France) which is the usual method. The ‘mercantile’ vs ‘settler’ classification is shown to be a significant predictor of present day levels of corruption. Several other characteristics will be examined for their hypothesized effects on present day democratization and racism.

There are several reasons why an election outcome can deviate from a pollster’s prediction. But many pollsters would agree that the most bedeviling is discerning which individuals in their sample will actually show up at the polls. This is why the formulae used by most pollsters to answer this pivotal question remains proprietary. This study will employ the very richly detailed Cooperative Congressional Election Study (CCES) to build from scratch a predictive model of individual turn-out.

This study uses the UCSD supercomputer to gather and organize data on DOS attacks during the first quarter of 2017.  Patterns are then assessed regarding originators, targets and possible ends. This is largely an exploratory study, given that so little is understood about DOS attacks

Does female leadership make it less likely a nation will engage in armed conflict? There is a good-sized literature regarding this question, but no consensus. This study takes up the debate and cuts into the data in new ways.

The study tests the hypothesis that donors respond to a ‘winner-loser’ gap over time (while taking account of candidate’s distance from their ideological ideal point). Adam Bonica’s Database on Ideology and Money in Elections (DIME) covering of all federal candidates from 1979 through 2012, is employed.

The study employs text-processing methods to glean statements made in the U.S. Senate regarding the U.K, and the House of Commons about the U.S. during the 19th Century. Event-coding techniques, well-known in the field of international relations are used to ascribe degrees of positive/negative affect. From this, a data-based description of the evolving nature of the U.S.-U.K. relationship is built.

This study attempts to sort the effects of different forms of ignorance. Three questions are addressed: 1) do uninformed voters behave differently from those who are actively misinformed, specifically with respect to turn-out and tolerance of others? 2) There is policy-specific ignorance, and ignorance about American institutions. Can we find any patterns in behavior which correlate with type of ignorance, or is it the degree of ignorance which is the predictor? 3) Is ignorance so broad and deep that we should question the health of American democracy?

Most all research asks whether U.N. operations succeed in stopping the inter- or intra-state conflicts they were sent to prevent. This study uses data from the Sexual Violence in Armed Conflict dataset (Cohen and Nordås 2014) to assess levels of violence against local populations and the extent to which U.N. forces help (or not) ameliorate problems.

An experimental design is used to examine attitudes toward the ‘out’ group in three types of Cypriot High Schools: All-Greek, All-Turkish, and deliberately integrated.

Text from major dailies in several Chinese cities is scraped and analyzed to find extent to which the Party line is followed from Beijing down to provincial and local levels.

This study compares practices used over time in the American National Election Survey (ANES) and the General Social Science Survey (GSS) to assess why turn-out over-reporting is generally higher in the former. This is not of sake of criticism, but as an exploration into the effects of different polling practices.