Machine learning (ML) methods have gained a great deal of popularity in recent years among public administration scholars and practitioners. These techniques open the door to analysis text, image other types data that allow us test foundational theories develop new theories. Despite excitement surrounding ML methods, clarity regarding their proper use potential pitfalls is lacking. This article attempts fill this gap literature through providing an "guide practice" for Here, we take view...

Matching for causal inference is a well-studied problem, but standard methods fail when the units to match are text documents: high-dimensional and rich nature of data renders exact matching infeasible, causes propensity scores produce incomparable matches, makes assessing quality difficult. In this paper, we characterize framework documents that decomposes existing into (1) choice representation (2) distance metric. We investigate how different choices within affect both quantity matches...

Delegation of powers represents a grant authority by politicians to one or more agents whose are determined the conditions in enabling statutes. Extant empirical studies this problem have relied on labor-intensive content analysis that ultimately restricts our knowledge how delegation has responded politics and institutional change recent years. We present machine learning approach estimation constraint European Union (EU) legislation, demonstrate its ability accurately generate same...

In this paper, we introduce a scalable machine learning approach accompanied by open-source software for identifying violent and peaceful forms of political protest participation using social media data. While protests are statistically rare events, they often shape public perceptions movements. This is, in part, due to the extensive disproportionate coverage which receives relative participation. past, when small number conglomerates served as primary information source about movements,...

Local governments operate in contexts that significantly vary their complexity, turbulence, and munificence. Such variations context have important implications for organizational outcomes practices, including budgetary orientations. To evaluate public sector we focus on budget functions California county proposed budgets during 2012–2017. These documents present a wealth of untapped information, which shed light number key variables interest. Computational text analysis methods offer highly...

We use the Conference Board’s Help-Wanted Index (HWI) to document how immigrant supply shocks change number of job vacancies. Our analysis reveals a sizable drop in Miami’s HWI relative comparable cities first few years after Mariel shock, followed by recovery afterward. An text help-wanted ads also documents significant decline low-skill vacancies advertised Miami Herald. Beveridge curve shifted inward mid-1980s, suggesting more efficient local labor market, contrast outward nationwide...

How do migration and immigration shape the political geography of American cities? In this article, we propose a mechanism partisan sorting demographic change which is tested using mass African Americans from New Orleans to Houston, Texas in aftermath Hurricane Katrina. We argue that differences residential choice preferences among partisans combined with changes increase diversity can induce by triggering flight (migration) ideological conservatives. Using Katrina evacuee data schools...

Machine learning methods have gained a great deal of popularity in recent years among public administration scholars and practitioners. These techniques open the door to analysis text, image other types data that allow us test foundational theories develop new theories. Despite excitement surrounding machine methods, clarity regarding their proper use potential pitfalls is lacking. This paper attempts fill this gap literature through providing “guide practice” for Here, we take view describe...

We present experimental results demonstrating how ideology shapes evaluations of technology aimed at detecting unconscious biases: (1) liberals supported use the to detect racism but not anti-Americanism, whereas conservatives showed reverse pattern, (2) and opposed punishing individuals for bias organizations failing root out, respectively, or (3) concerns about researcher false accusations mediated effects on support technology, (4) participants taking strong initial stands were likelier...

Organizations produce copious volumes of written documents, including position papers, meeting summaries, minutes from hearings, presentations, and budget justifications. These documents present a wealth untapped information, which can shed light on variety organizational factors-individual group behaviors, managerial policy choices, other key inter- intra-organizational dynamics that are great interest to public managers administration scholars. Computational text analysis methods offer...

While members of Congress now routinely communicate with constituents using images on a variety internet platforms, little is known about how are used as means strategic political communication. This due primarily to computational limitations which have prevented large-scale, systematic analyses image features. New developments in computer vision, however, bringing the study within reach. Here, we develop framework for understanding visual communication by extending Fenno's analysis home...

The adoption and use of emerging smart technologies like robotics is a key concern as the advanced economies undergo latest technological revolution. While firms factories often adopt manufacturing techniques at fast rate, government agencies are seen lagging in their such tools. We offer evidence about from case American crime laboratories. Using data census labs, we show that depends on traditional drivers agency capacity demand: budgets, agency's task environment, its relative level...

Matching for causal inference is a well-studied problem, but standard methods fail when the units to match are text documents: high-dimensional and rich nature of data renders exact matching infeasible, causes propensity scores produce incomparable matches, makes assessing quality difficult. In this paper, we characterize framework documents that decomposes existing into: (1) choice representation, (2) distance metric. We investigate how different choices within affect both quantity matches...

Regression discontinuity designs are frequently used to estimate the causal effect of election outcomes and policy interventions. In these contexts, treatment effects typically estimated with covariates included improve efficiency. While including improves precision asymptotically, in practice, a small number observations, resulting considerable fluctuations magnitude depending upon chosen. This practice thus incentivizes researchers select which maximize statistical significance rather than...

Does racial threat motivate support for anti-immigration laws? I answer this question by manipulating the skin-tone and geographical proximity to American citizens of a fictional undocumented Mexican immigrant. find that when respondents are exposed non-Caucasian immigrant, laws increases relative an otherwise identical Caucasian These reactions immigrant's observed only believe immigrant resides in their city state, suggesting triggers threat.

Do partisans sort? If so, why? I address both questions in this paper by developing the Migration-Assimilation-Polarization (MAP) theory of partisan sorting which is tested using Hurricane Katrina migration to Houston, Texas as a natural experiment. According MAP Theory, conservative flight induced changes diversity provides mechanism for can at least partially account trends geographic polarization. Using variety empirical tools including synthetic controls, demonstrate that led increases...

Recent research suggests that suburbanization precipitated by the development of Interstate Highway System drew Republican residents out urban centers and gave rise to geographic political polarization between cities suburbs (Nall 2015). In recent years, however, rate has more than doubled while suburban growth declined construction ceased. this paper, we argue these trends can be explained a complimentary "push migration" mechanism demographic change facilitated existing infrastructure in...

Innovative natural experiments, observational research and theories of racial threat suggest that skin tone is a determinant nativist sentiment, yet experiments which include immigrant as treatment find little connection between the two. We argue these contradictory findings can be partially explained by experimental designs exclude information about geographic context, an essential component threat. To address issues, we design survey experiment in context are randomly manipulated. has...