While Amazon's Mechanical Turk (MTurk) has reduced the cost of collecting original data, in 2018, researchers noted the potential existence of a large number of bad actors on the platform. To evaluate data quality on MTurk, we fielded three surveys between 2018 and 2020. While we find no evidence of a “bot epidemic,” significant portions of the data—between 25 and 35 percent—are of dubious quality. While the number of IP addresses that completed the survey multiple times or circumvented location requirements fell almost 50 percent over time, suspicious IP addresses are more prevalent on MTurk than on other platforms. Furthermore, many respondents appear to respond humorously or insincerely, and this behavior increased over 200 percent from 2018 to 2020. Importantly, these low-quality responses attenuate observed treatment effects by magnitudes ranging from approximately 10 to 30 percent.

Many believe that weaker legal constraints make unilateral action easier to reverse than legislation. Yet in some cases, unilateralism survives a successor's determined opposition. I argue that legislative persistence and unilateral transience may arise not only from differences in legal status but also from a selection effect: unilateralism and subsequent rescission can both stem from weak underlying policy-demanding groups. But if unilateralism itself alters the landscape of group power through policy feedback effects, it can survive a purportedly opposed new president. An example is the Trump Administration's failure to depart from the Obama Justice Department's lax stance on cannabis. Weak constituents’ policy demands may thus need to rely on a sympathetic president's unilateral action to begin a feedback loop, subsequently softening opposition. This paper therefore identifies unilateral action as a key tool for shifting policy in an enduring way, explaining its success or failure in terms of interest group power.

Since the end of World War II, the US government has spent nearly $4 trillion on humanitarian, economic, and military assistance to other countries. Despite the myriad benefits of such programs, mass support has long been lacking. Here, I argue that low citizen trust in government can help us to understand why. Using cross-sectional and panel survey data from the United States, I find a positive and substantively significant relationship between political trust and support for government spending on foreign aid. Overall, these findings underscore the relevance of political trust and further illustrate the drivers of U.S. public opinion toward foreign aid, something that has implications for whether America should turn inward or continue its long-standing role of global leadership.

How does a candidate's racial background affect the inferences voters make about them? Prior work finds that Black candidates are perceived to be more liberal. Using two survey experiments, we test whether this effect persists when candidate partisanship and issue positions are specified and also consider other consequential voter perceptions. We make two contributions. First, we show that while Black candidates are perceived to be more liberal than White candidates with the same policy positions, this difference is smaller for Black candidates who adopt more conservative positions on race-related issues. Second, we find that voters, both Black and White, believe Black candidates will prioritize the interests of Black constituents over those of White constituents, regardless of candidate positions.

What impact do party leaflets and canvass visits have on voter turnout? Get Out The Vote (GOTV) experiments consistently find that campaigning needs to be personal in order to be effective. However, the imbalance between United States and European-based studies has led to recent calls for further European GOTV experiments. There are also comparatively few partisan experiments. I report the findings of a United Kingdom-based field experiment conducted with the Liberal Democrats in 2017. Results show that party leaflets boost turnout by 4.3 percentage points, while canvassing has a small additional effect (0.6 percentage points). The study also represents the first individual level experiment to compare GOTV effects between postal voters and in-person voters outside the United States.

We demonstrate how few-shot prompts to large language models (LLMs) can be effectively applied to a wide range of text-as-data tasks in political science—including sentiment analysis, document scaling, and topic modeling. In a series of pre-registered analyses, this approach outperforms conventional supervised learning methods without the need for extensive data pre-processing or large sets of labeled training data. Performance is comparable to expert and crowd-coding methods at a fraction of the cost. We propose a set of best practices for adapting these models to social science measurement tasks, and develop an open-source software package for researchers.

Do policy priorities that candidates emphasize during election campaigns predict their subsequent legislative activities? We study this question by assembling novel data on legislative leadership posts held by Japanese politicians and using a fine-tuned transformer-based machine learning model to classify policy areas in over 46,900 statements from 1270 candidate manifestos across five elections. We find that a higher emphasis on a policy issue increases the probability of securing a legislative post in the same area. This relationship remains consistent across multiple elections and persists even when accounting for candidates' previous legislative leadership roles. We also discover greater congruence in distributive policy areas. Our findings indicate that campaigns provide meaningful signals of policy priorities.

Various robust communication effects have been identified, but evidence is overwhelmingly based on artificial survey treatments with limited real-world insight. I conducted a natural experiment on the impact of the European–Turkey statement closing the Balkan route during the 2015/16 European refugee crisis in Germany. This design tests the lasting effect of the statement's framing on public sentiment. I identify treatment and control groups based on timing to demonstrate its effect on perceptions of the crisis, asylum attitudes, and policy preferences. Effects are largest immediately following the announcement but decline rapidly. This shows political communication can significantly change opinion within a limited time frame. This study enhances our understanding of real-world communication effects and offers a broadly applicable methodology.

Economic sanctions are a policy tool that great powers frequently use to interfere with domestic politics of another state. Regime change has been a primary goal of economic sanctions over the past decades. This article studies the relationship between leader-contingent sanctions—sanctions that are designed to impede the flow of revenue to a specific leader—and violent political conflict in target countries. I build a theoretical model to illuminate two mechanisms by which leader-contingent sanctions destabilize a regime—the Depletion Mechanism and the Instigation Mechanism. The Depletion Mechanism works when sanctions mechanically deplete the government's resources so that it becomes unable to buy off domestic opposition even by making the largest possible offer. The Instigation Mechanism implies that as sanctions decrease the benefit of negotiated settlement relative to war, the government may strategically choose to repress rather than buy off the opposition even when it is able to do so. Leader-contingent sanctions lead to bargaining failure by rewarding the opposition for revolt while reducing the government's ability and willingness to appease the opposition.

Technological change often increases demand for high-skilled jobs, with low-skilled losers turning to the populist right in response. The political effects of technological change that increases demand for low-skilled workers are largely unknown. The growth of the salmon fish-farming industry in rural Norway improved the labor-market situation for low-skilled workers, and we find that support for the populist right-wing party increased in municipalities that benefitted from the industry growth. The electoral change is due to a right-wing shift on the economic, but not the cultural dimension. Our results support political economy frameworks that point to lower demand for state interventions after positive labor market shocks, but raise the question of in what contexts support for populism will decline.

Can democracies attract and elect politicians who are both high-ability and from diverse backgrounds? Using data covering the entire Danish population, including every candidate for local and national elections in 1990–2015, we explore the selection of political candidates. We show that Danish candidates and elected politicians have higher ability than the voters they represent, that selection on ability reflects individual skills rather than social background, that politicians are selected from economically diverse backgrounds, and that there is no substantial trade-off between ability and representation. Furthermore, we utilize a major structural reform, which significantly reduced the number of municipalities in Denmark, to show that increased political competition did not affect politicians’ ability, economic background, or the trade-off between the two.

The revolving door is a potential mechanism of private influence over policy. Recent work primarily examines the revolving of legislators and their staff, with little focus on the federal bureaucracy. To analyze decisions to turnover into lobbying, we develop an argument emphasizing the (1) policy expertise acquired from federal employment; (2) the proximity of employees to political decision-making; and (3) the agency policymaking environment. Leveraging federal personnel and lobbying data, we find the first two factors predict revolving whereas the policymaking environment has an inconsistent impact. We highlight the importance of studying selection into lobbying for estimating casual effects of lobbyist characteristics on revenue and contribute to the literature on bureaucratic careers and the nature of private influence in policymaking.

In this paper we test how much party labels influence vote choices between candidates when voters have access to realistically distributed information about candidate positions and demographics. We do not seek to adjudicate a long-running debate about the role of party labels, but present some nuances on the two archetypal theoretical views on vote choices. We use data from the Representative Audit of Britain (RAB) and the British Elections Study (BES) to generate electoral match-ups between randomly selected Conservative versus Labour candidates, with only half of respondents seeing party labels in addition to candidates’ positions and demographics. For our experiment fielded in October 2021, we find negligible to moderate effects of party labels on vote choices. Our results suggest the information on candidate positions and party labels largely act as substitutes for one another, with only modest changes when party labels are made explicit.

Inverse probability weighting is a common remedy for missing data issues, notably in causal inference. Despite its prevalence, practical application is prone to bias from propensity score model misspecification. Recently proposed methods try to rectify this by balancing some moments of covariates between the target and weighted groups. Yet, bias persists without knowledge of the true outcome model. Drawing inspiration from the quasi maximum likelihood estimation with misspecified statistical models, I propose an estimation method minimizing a distance between true and estimated weights with possibly misspecified models. This novel approach mitigates bias and controls mean squared error by minimizing their upper bounds. As an empirical application, it gives new insights into the study of foreign occupation and insurgency in France.

Social connections between individuals can profoundly impact their political behavior. A growing body of research on legislative politics examines how spatial proximity to fellow legislators affects voting behavior within the institution. However, studies that examine this question often suffer from a fundamental identification problem in which proximity effects may reflect actual behavioral diffusion between members or, instead, homophily, in which legislators of a similar political feather flock together. We overcome this observational equivalence by exploiting a unique random seating lottery for seating assignments in the world's oldest existing parliament, Iceland's national legislature, Alþingi. Utilizing this naturally occurring randomization, we employ spatial analyses of more than 20,000 estimates of spatial dependence and find little evidence that seating proximity leads to similar voting behavior by members in this legislative context.

How well do existing survey instruments differentiate between opinions that affect individual behavior and opinions that don't? To answer this question, we randomly assigned U.S. respondents to one of three survey instruments: Likert items (Likert), Likert items followed by personal importance items (Likert+) and Quadratic Voting for Survey Research (QVSR), which gives respondents a fixed budget to buy “favor” or “oppose” votes, with the price for each vote increasing quadratically. We find that, relative to Likert, both Likert+ and QVSR better identify people who care enough about an issue to act in opinion-congruent ways, with QVSR offering the most consistent improvement overall. Building on these results, we show how conclusions regarding the relationship between policy opinions and self-interest can differ across measurement strategies.

Stance detection is identifying expressed beliefs in a document. While researchers widely use sentiment analysis for this, recent research demonstrates that sentiment and stance are distinct. This paper advances text analysis methods by precisely defining stance detection and outlining three approaches: supervised classification, natural language inference, and in-context learning. I discuss how document context and trade-offs between resources and workload should inform your methods. For all three approaches I provide guidance on application and validation techniques, as well as coding tutorials for implementation. Finally, I demonstrate how newer classification approaches can replicate supervised classifiers.

We estimate how asset prices respond to a range of political shocks, including changes in a country's economic stewardship, national elections, coup d'états, wars, and terrorist attacks. Using an event study approach and daily prices from the Buenos Aires exchange (Argentina) between 1967 and 2020, we find that stock-market volatility increases in the days immediately following major policy-shifting events. These results hold irrespective of whether returns are measured in nominal terms, in local consumption units, or in US dollars. The most significant increase in post-event risk is associated with irregular government turnovers. Volatility also increases in the days immediately following a defeat in an international war, national elections, and changes in the country's economic stewardship. No changes in stock-market volatility occur, however, after terrorist attacks or when the date of a new administration's inauguration is publicly known and determined sufficiently far in advance.

Across a series of experiments, I show that racial threat from a stereotypically nonthreatening racial minority group, Asian Americans, has a direct impact on white Americans' views of discrimination toward the group. When white Americans learn the group is growing, they feel a distinct racial threat which decreases support for the idea that Asian Americans experience discrimination while simultaneously increasing support for policy which actively discriminates against Asian Americans. I show this concept to be portable over context, examining support for discriminatory policy toward Asians in education policy and COVID-19 policy. I conclude by discussing the implications for how racial threat can drive more racial conflict while simultaneously decreasing the perception that this discrimination is occurring.

We study how the electoral success of radical right populist parties (RRPPs) affects mainstream parties' defense policy positions. The success of RRPPs threatens the credibility of established left-wing parties with coalition and international partners due to substantive overlap between their and RRPPs' defense-skeptical position. We argue that left parties adopt more assertive defense positions to distinguish themselves from RRPPs, thus increasing mainstream consensus on defense policy. Examining 27 European countries between the end of the Cold War and Russia's occupation of Crimea (1990–2013), we test this argument based on a regression discontinuity design around electoral thresholds for obtaining parliamentary seats. We find that, in response to RRPP success, left parties adopt more assertive defense policy positions, whereas center-right parties stand their ground. This study yields evidence for an adversarial response to the radical right, often thought to have lost out to accommodation, and for mechanisms other than electoral incentives, in a highly consequential domain.