Luke John Keele

Luke John Keele
Georgetown University | GU · McCourt School of Public Policy

Ph.D.

About

124
Publications
52,948
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
11,373
Citations
Citations since 2017
44 Research Items
8194 Citations
201720182019202020212022202305001,0001,500
201720182019202020212022202305001,0001,500
201720182019202020212022202305001,0001,500
201720182019202020212022202305001,0001,500
Additional affiliations
July 2011 - present
Pennsylvania State University
Position
  • Professor (Associate)
January 2006 - July 2011
The Ohio State University
Position
  • Professor (Assistant)
October 2003 - December 2005
University of Oxford
Position
  • PostDoc Position

Publications

Publications (124)
Article
Background: Little is known about the impact of multimorbidity on long-term outcomes for older emergency general surgery patients. Study design: Medicare beneficiaries, age 65 and older, who underwent operative management of an emergency general surgery condition were identified using Centers for Medicare & Medicaid claims data. Patients were cl...
Article
Importance: A surgical consultation is a critical first step in the care of patients with emergency general surgery conditions. It is unknown if Black Medicare patients and White Medicare patients receive surgical consultations at similar rates when they are admitted from the emergency department. Objective: To determine whether Black Medicare p...
Article
Objective: To determine the effect of operative versus nonoperative management of emergency general surgery conditions on short-term and long-term outcomes. Background: Many emergency general surgery conditions can be managed either operatively or nonoperatively, but high-quality evidence to guide management decisions is scarce. Methods: We in...
Article
Background: Little is known about the impact of multimorbidity on outcomes for older emergency general surgery patients. Objective: The aim was to understand whether having multiple comorbidities confers the same amount of risk as specific combinations of comorbidities (multimorbidity) for a patient undergoing emergency general surgery. Researc...
Article
Full-text available
We discuss some causal estimands that are used to study racial discrimination in policing. A central challenge is that not all police–civilian encounters are recorded in administrative datasets and available to researchers. One possible solution is to consider the average causal effect of race conditional on the civilian already being detained by t...
Article
We investigate the efficacy of surgical versus non-surgical management for two gastrointestinal conditions, colitis and diverticulitis, using observational data. We deploy an instrumental variable design with surgeons’ tendencies to operate as an instrument. Assuming instrument validity, we find that non-surgical alternatives can reduce both hospit...
Preprint
We discuss some causal estimands used to study racial discrimination in policing. A central challenge is that not all police-civilian encounters are recorded in administrative datasets and available to researchers. One possible solution is to consider the average causal effect of race conditional on the civilian already being detained by the police...
Article
Full-text available
Clustered observational studies (COSs) are a critical analytic tool for educational effectiveness research. We present a design framework for the development and critique of COSs. The framework is built on the counterfactual model for causal inference and promotes the concept of designing COSs that emulate the targeted randomized trial that would h...
Preprint
Instrumental variable (IV) studies seek to emulate randomized encouragement designs where patients are assigned a random nudge towards a particular treatment. Unfortunately, IV studies may fall short of their experimental ideal due to hidden bias affecting both the proposed instrument and the outcomes. While sensitivity analyses have been developed...
Article
Objectives: Thiamine deficiency may propagate lactate production by limiting pyruvate dehydrogenase activity, and studies suggest benefit for thiamine administration in septic adults. We studied the effect of thiamine on physiologic and clinical outcomes for children with septic shock and hyperlactatemia. Design: Retrospective matched cohort stu...
Article
Full-text available
Importance: It is unknown which deteriorating ward patients benefit from intensive care unit (ICU) transfer. Objectives: To use an instrumental variable (IV) method that assesses heterogeneity and to evaluate estimates of person-centered treatment effects of ICU transfer and 28-day hospital mortality by age and illness severity. Design, setting...
Article
Background: Instrumental variable (IV) analysis can estimate treatment effects in the presence of residual or unmeasured confounding. In settings wherein measures of baseline risk severity are unavailable, IV designs are, therefore, particularly appealing, but, where established measures of risk severity are available, it is unclear whether IV met...
Preprint
Applied analysts often use the differences-in-differences (DID) method to estimate the causal effect of policy interventions with observational data. The method is widely used, as the required before and after comparison of a treated and control group is commonly encountered in practice. DID removes bias from unobserved time-invariant confounders....
Article
Background: Instrumental variable (IV) methods are becoming an increasingly important tool in health services research as they can provide consistent estimates of causal effects in the presence of unobserved confounding. However, investigators must provide justifications that the IV is independent with any unmeasured confounder and its effect on t...
Article
Instrumental variable methods, subject to appropriate identification assumptions, enable consistent estimation of causal effects in the presence of unobserved confounding. Near–far matching has been proposed as one analytic method to improve inference by strengthening the effect of the instrument on the exposure and balancing observable characteris...
Article
Full-text available
Confounding by indication is a critical challenge in evaluating the effectiveness of surgical interventions using observational data. The threat from confounding is compounded when using medical claims data due to the inability to measure risk severity. If there are unobserved differences in risk severity across patients, treatment effect estimates...
Article
Many observational studies of causal effects occur in settings with clustered treatment assignment. In studies of this type, treatment is applied to entire clusters of units. For example, an educational intervention might be administered to all the students in a school. We develop a matching algorithm for multilevel data based on a network flow alg...
Preprint
Clustered randomized trials (CRTs) are popular in the social sciences to evaluate the efficacy of a new policy or program by randomly assigning one set of clusters to the new policy and the other set to the usual policy. Often, many individuals within a cluster fail to take advantage of the new policy, resulting in noncompliance behaviors. Also, in...
Preprint
Regression discontinuity (RD) designs are viewed as one of the most credible identification strategies for program evaluation and causal inference. However, RD treatment effect estimands are necessarily local, making the extrapolation of these effects a critical open question. We introduce a new method for extrapolation of RD effects that exploits...
Article
We analyze a geographic natural experiment during the 2010 Colorado primary election in the USA, when counties in the state of Colorado had the option to have an all-mail election or retain traditional in-person voting on Election Day. The town of Basalt, in the southwestern part of the state, is split in half by two counties that chose different m...
Preprint
Many policy evaluations occur in settings with randomized assignment at the cluster level and treatment noncompliance at the unit level. For example, villagers or towns might be assigned to treatment and control, but residents may choose to not comply with their assigned treatment status. For example, in the state of Andhra Pradesh, the state gover...
Preprint
Matching methods have become one frequently used method for statistical adjustment under a selection on observables identification strategy. Matching methods typically focus on modeling the treatment assignment process rather than the outcome. Many of the recent advances in matching allow for various forms of covariate prioritization. This allows a...
Article
Do minority voters respond to co-racial or co-ethnic candidates? That is does the increased chance of substantive representation translate into increased participation? Here, we focus on this question among African American voters. While much of the empirical literature on this question has produced conflicting answers, recent studies suggest that...
Article
The method of instrumental variables provides a framework to study causal effects in both randomized experiments with non-compliance and in observational studies where natural circumstances produce as if random nudges to accept treatment. Traditionally, inference for instrumental variables relied on asymptotic approximations of the distribution of...
Article
We retrospectively studied the effect of introducing procalcitonin into clinical practice on antibiotic use within a large academic pediatric intensive care unit. In the absence of a standardized algorithm, availability of the procalcitonin assay did not reduce the frequency of antibiotic initiations or the continuation of antibiotics for greater t...
Article
Importance Important metrics of residency program success include the clinical outcomes achieved by trainees after transitioning to practice. Previous studies have shown significant differences in reported training experiences of general surgery residents at nonuniversity-based residency (NUBR) and university-based residency (UBR) programs. Object...
Article
Presidents often campaign on behalf of candidates during elections. Do these campaign visits increase the probability that the candidate will win? While one might attempt to answer this question by adjusting for observed covariates, such an approach is plagued by serious data limitations. In this paper we pursue a different approach. Namely, we ask...
Article
Full-text available
While randomization inference is well developed for continuous and binary outcomes, there has been comparatively little work for outcomes with nonnegative support and clumping at zero. Typically, outcomes of this type have been modeled using parametric models that impose strong distributional assumptions. This article proposes new randomization inf...
Article
Objective: We sought to compare postoperative outcomes of female surgeons (FS) and male surgeons (MS) within general surgery. Summary of background data: FS in the workforce are increasing in number. Female physicians provide exceptional care in other specialties. Differences in surgical outcomes of FS and MS have not been examined. Methods: W...
Chapter
We study research designs where a binary treatment changes discontinuously at the border between administrative units such as states, counties, or municipalities, creating a treated and a control area. This type of geographically discontinuous treatment assignment can be analyzed in a standard regression discontinuity (RD) framework if the exact ge...
Article
What effect does a candidate’s race have on coracial voter turnout? Recent studies have found mixed results, largely because it is difficult to separate the effect of candidate race from other factors that drive voter turnout. We argue that viability is a key element in the theory of turnout among coracial voters that has been overlooked in the ext...
Article
Full-text available
Pre-treatment selection or censoring (`selection on treatment') can occur when two treatment levels are compared ignoring the third option of neither treatment, in `censoring by death' settings where treatment is only defined for those who survive long enough to receive it, or in general in studies where the treatment is only defined for a subset o...
Article
A distinctive feature of a clustered observational study is its multilevel or nested data structure arising from the assignment of treatment, in a non-random manner, to groups or clusters of units or individuals. Examples are ubiquitous in the health and social sciences including patients in hospitals, employees in firms, and students in schools. W...
Preprint
Social scientists use the concept of interactions to study effect dependency. Such analyses can be conducted using standard regression models. However, an interaction analysis may represent either a causal interaction or effect modification. Under causal interaction, the analyst is interested in whether two treatments have differing effects when bo...
Article
Objective: Delayed antimicrobial therapy in sepsis is associated with increased hospital mortality, but the impact of antimicrobial timing on long-term outcomes is unknown. We tested the hypothesis that hourly delays to antimicrobial therapy are associated with 1-year mortality in pediatric severe sepsis. Design: Retrospective observational stud...
Article
Objective: To test the hypothesis that resuscitation with balanced fluids (lactated Ringer [LR]) is associated with improved outcomes compared with normal saline (NS) in pediatric sepsis. Study design: We performed matched analyses using data from 12 529 patients <18 years of age with severe sepsis/septic shock at 382 US hospitals between 2000 a...
Article
In this article, we highlight three points. First, we counter Grant and Lebo’s claim that the error correction model (ECM) cannot be applied to stationary data. We maintain that when data are properly stationary, the ECM is an entirely appropriate model. We clarify that for a model to be properly stationary, it must be balanced. Second, we contend...
Article
This issue began as an exchange between Grant and Lebo (2016) and ourselves (Keele, Linn, and Webb 2016) about the utility of the general error correction model (GECM) in political science. The exchange evolved into a debate about Grant and Lebo’s proposed alternative to the GECM and the utility of fractional integration methods (FIM). Esarey (2016...
Article
Full-text available
In randomized experiments, subjects often fail to comply with the assigned treatment assignment. When such non-compliance occurs, the method of instrumental variables provides a framework to study causal effects for those who actually received the treatment. In this paper, we compare various finite sample methods of inference used in instrumental v...
Article
In randomized controlled trials with non-adherence, instrumental variable (IV) methods are frequently used to report the complier average causal effect. With binary outcomes, many of the available IV estimation methods impose distributional assumptions. We develop a randomization-inference-based method of IV estimation for binary outcomes. The meth...
Article
In a natural experiment, treatment assignments are made through a haphazard process that is thought to be as if random. In one form of the natural experiment, encouragement to accept treatment rather than treatments them-selves are assigned in this haphazard process. This encouragement to accept treatment is often referred to as an instrument. Inst...
Article
We consider a regression discontinuity (RD) design where the treatment is received if a score is above a cutoff, but the cutoff may vary for each unit in the sample instead of being equal for all units. This multi-cutoff regression discontinuity design is very common in empirical work, and researchers often normalize the score variable and use the...
Article
Full-text available
Objectives: Pediatric severe sepsis remains a significant global health problem without new therapies despite many multicenter clinical trials. We compared children managed with severe sepsis in European and U.S. PICUs to identify geographic variation, which may improve the design of future international studies. Design: We conducted a secondary...
Article
Full-text available
In randomized controlled trials, the evaluation of an overall treatment effect is often followed by effect modification or subgroup analyses, where the possibility of a different magnitude or direction of effect for varying values of a covariate is explored. While studies of effect modification are typically restricted to pretreatment covariates, l...
Article
Full-text available
In parameter determination for the heteroskedastic probit model, both in simulated data and in actual data, we observe a failure of traditional local search methods to converge consistently to a single parameter vector, in contrast to the typical situation for the regular probit model. We identify features of the heteroskedastic probit log likeliho...
Article
Full-text available
In policy evaluations, interest may focus on why a particular treatment works. One tool for understanding why treatments work is causal mediation analysis. In this essay, I focus on the assumptions needed to estimate mediation effects. I show that there is no “gold standard” method for the identification of causal mediation effects. In particular,...
Article
Full-text available
Many areas of political science focus on causal questions. Evidence from statistical analyses is often used to make the case for causal relationships. While statistical analyses can help establish causal relationships, it can also provide strong evidence of causality where none exists. In this essay, I provide an overview of the statistics of causa...
Article
Causal analysis in program evaluation has primarily focused on the question about whether or not a program, or package of policies, has an impact on the targeted outcome of interest. However, it is often of scientific and practical importance to also explain why such impacts occur. In this paper, we introduce causal mediation analysis, a statistica...
Article
Full-text available
Political scientists often attempt to exploit natural experiments to estimate causal effects. We explore how variation in geography can be exploited as a natural experiment and review several assumptions under which geographic natural experiments yield valid causal estimates. In particular, we focus on cases where a geographic or administrative bou...
Article
Full-text available
This essay is a comment on whether there are tensions between big data, theory (mostly formal), and the statistics of causal inference. Of these three areas, big data is clearly the newest. However, the idea that political methodology should focus on causal inference is also a relatively new idea. I have been attending the Annual Political Methodol...
Article
Full-text available
Mediation analysis has been extensively applied in psychological and other social science research. A number of methodologists have recently developed a formal theoretical framework for mediation analysis from a modern causal inference perspective. In Imai, Keele, and Tingley (2010), we have offered such an approach to causal mediation analysis tha...
Article
Full-text available
Political scientists often turn to natural experiments to draw causal inferences with observational data. Recently, the regression discontinuity design (RD) has become a popular type of natural experiment due to its relatively weak assumptions. We study a special type of regression discontinuity design where the discontinuity in treatment assignmen...
Article
Full-text available
Princeton Abstract In this paper, we describe the R package mediation for conducting causal mediation analysis in applied empirical research. In many scientific disciplines, the goal of researchers is not only estimating causal effects of a treatment but also understanding the process in which the treatment causally affects the outcome. Causal medi...
Article
In some observational studies of treatment effects, matched samples are created so treated and control groups are similar in terms of observable covariates. Traditionally, such matched samples consist of matched pairs. However, alternative forms of matching may have desirable features. One strategy that may improve efficiency is to match a variable...
Article
Full-text available
A distinctive feature of a clustered observational study is its multilevel or nested data structure arising from the assignment of treatment, in a non-random manner, to groups or clusters of individuals. Examples are ubiquitous in the health and social sciences including patients in hospitals, employees in firms, and students in schools. What is th...
Article
Ballot initiatives allow the public to vote directly on public policy. A literature in political science has attempted to document whether the presence of an initiative can increase voter turnout. We study this question for an initiative that appeared on the ballot in 2008 in Milwaukee, WI, using a natural experiment based on geography. This form o...
Article
Scholars of state politics are often interested in the causal effects of legislative institutions on policy outcomes. For example, during the 1990s a number of states adopted term limits for state legislators. Advocates of term limits argued that this institutional reform would alter state policy in a number of ways, including limiting state expend...
Article
Political scientists are often interested in estimating causal effects. Identification of causal estimates with observational data invariably requires strong untestable assumptions. Here, we outline a number of the assumptions used in the extant empirical literature. We argue that these assumptions require careful evaluation within the context of s...
Article
There is growing interest in natural experiments in political science. Natural experiments are often analyzed with instrumental variable estimators reflecting a belief that combining the power of natural random assignment with an instrumental variable approach will solve many of the research design problems endemic to social science. Here, we highl...
Article
Experiments have become an increasingly common tool for political science researchers over the last decade, particularly laboratory experiments performed on small convenience samples. We argue that the standard normal theory statistical paradigm used in political science fails to meet the needs of these experimenters and outline an alternative appr...
Article
Full-text available
Identifying causal mechanisms is a fundamental goal of social science. Researchers seek to study not only whether one variable affects another but also how such a causal relationship arises. Yet commonly used statistical methods for identifying causal mechanisms rely upon untestable assumptions and are often inappropriate even under those assumptio...
Article
Full-text available
The 1965 Voting Rights Act has been a central part of Federal efforts to increase minority participation in the U.S. The latest phase of enforcement under the Voting Rights Act has been the creation of majority-minority Congressional districts. To fight vote dilution, these districts are drawn so that a majority of the voting age population are min...
Article
Full-text available
Causal mediation analysis is routinely conducted by applied researchers in a variety of disciplines. The goal of such an analysis is to investigate alternative causal mechanisms by examining the roles of intermediate variables that lie in the causal paths between the treatment and outcome variables. In this paper we first prove that under a particu...
Article
Full-text available
Traditionally in the social sciences, causal mediation analysis has been formulated, understood, and implemented within the framework of linear structural equation models. We argue and demonstrate that this is problematic for 3 reasons: the lack of a general definition of causal mediation effects independent of a particular statistical model, the i...
Article
Full-text available
Political scientists are often interested in understanding whether state laws alter individual level behavior. For example, states often alter their election procedures, which can increase or decrease the cost of voting. In this example, it is important to understand whether these changes alter turnout since changes in costs may disproportionally a...
Article
Full-text available
I simulate a mixture process that generates individual preferences that, when aggregated into precincts, have counts whose second significant digits approximately satisfy Benford's Law. By deriving sincere, strategic, gerrymandered and coerced votes from these preferences under a plurality voting rule, I find that tests based on the second digits o...
Article
Full-text available
The Cox proportional hazards model is widely used to model durations in the social sciences. Although this model allows analysts to forgo choices about the form of the hazard, it demands careful attention to the proportional hazards assumption. To this end, a standard diagnostic method has been developed to test this assumption. I argue that the st...
Article
Full-text available
Why do people practice citizenship in a partisan rather than in a deliberative fashion? We argue that they are not intractably disposed to one type of citizenship, but instead adopt one of two different modes depending on the strategic character of current circumstances. While some situations prompt partisan solidarity, other situations encourage p...
Article
Full-text available
Abstract will be provided by author.
Article
Full-text available
Causal mediation analysis is widely used across many disciplines to investigate possible causal mechanisms. Such an analysis allows researchers to explore various causal pathways, going beyond the estimation of simple causal eects. Recently, Imai et al. (2008) (3) and Imai et al. (2009) (2) devel- oped general algorithms to estimate causal mediatio...
Article
Full-text available
It has long been understood that the presence of the ballot initiative process leads to different outcomes among states. In general, extant research has found that the presence of ballot initia-tives tends to increase voter turnout and depress state revenues and expenditures. I reconsider this possibility and demonstrate that past findings are an a...
Article
Presidential campaign visits for members of the House are important events during cam- paigns. The role of the president in House elections serves a dual purpose. First, it may increase the probability of the president being of the majority party in Congress. Second, it ensures that members of Congress owe some debt to the president. Assessing the...
Article
When will people become ambivalent about politics? One possibility is that the roots of ambivalence lie within the individual, with differences in political knowledge and attitude strength predicting whether a person internalizes the conflicts of politics. Alternately, attitudinal ambivalence could result from structural differences in the way poli...
Article
Full-text available
Deliberative theorists emphasize that citizensÕ capacity to be-come informed when given a motive and the opportunity to participate in politics is important for democratic citizenship. We assess this capacity among citizens using a deliberative field experiment. In the summer of 2006, we conducted a field experiment in which we recruited twelve cur...
Article
If ignored, non-compliance with a treatment and nonresponse on outcome measures can bias estimates of treatment effects in a randomized experiment. To identify treatment effects in the case where compliance and response are conditioned on unobservables, we propose the parametric generalized endoge-nous treatment (GET) model. As a multilevel random...
Article
An introductory guide to smoothing techniques, semiparametric estimators, and their related methods, this book describes the methodology via a selection of carefully explained examples and data sets. It also demonstrates the potential of these techniques using detailed empirical examples drawn from the social and political sciences. Each chapter in...
Chapter
Generalized Linear ModelsEstimation of GAMSStatistical InferenceExamplesDiscussionExercises