Explore the vast range of titles available.

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 practiced importance to also explain why such impacts occur. In this paper, we introduce causal mediation analysis, a statistical framework for analyzing causal mechanisms that has become increasingly popular in social and medical sciences in recent years. The framework enables us to show exactly what assumptions are sufficient for identifying causal mediation effects for the mechanisms of interest, derive a general algorithm for estimating such mechanism-specific effects, and formulate a sensitivity analysis for the violation of those identification assumptions. We also discuss an extension of the framework to analyze causal mechanisms in the presence of treatment noncompliance, a common problem in randomized evaluation studies. The methods are illustrated via applications to two intervention studies on preschool classes and job-training workshops.

The regression discontinuity (RD) design is a popular quasi-experimental design for causal inference and policy evaluation. The most common inference approaches in RD designs employ \"flexible\" parametric and nonparametric local polynomial methods, which rely on extrapolation and large-sample approximations of conditional expectations using observations somewhat near the cutoff that determines treatment assignment. An alternative inference approach employs the idea of local randomization, where the very few units closest to the cutoff are regarded as randomly assigned to treatment and finite-sample exact inference methods are used. In this paper, we contrast these approaches empirically by re-analyzing the influential findings of Ludwig and Miller (2007), who studied the effect of Head Start assistance on child mortality employing parametric RD methods. We first review methods based on approximations of conditional expectations, which are relatively well developed in the literature, and then present new methods based on randomization inference. In particular, we extend the local randomization framework to allow for parametric adjustments of the potential outcomes; our extended framework substantially relaxes strong assumptions in prior literature and better resembles other RD inference methods. We compare all these methods formally, focusing on both estimands and inference properties. In addition, we develop new approaches for randomization-based sensitivity analysis specifically tailored to RD designs. Applying all these methods to the Head Start data, we find that the original RD treatment effect reported in the literature is quite stable and robust, an empirical finding that enhances the credibility of the original result. All the empirical methods we discuss are readily available in general purpose software in R and Stata; we also provide the dataset and software code needed to replicate all our results.

Theory predicts that regression discontinuity (RD) provides valid causal inference at the cutoff score that determines treatment assignment. One purpose of this paper is to test RD's internal validity across 15 studies. Each of them assesses the correspondence between causal estimates from an RD study and a randomized control trial (RCT) when the estimates are made at the same cutoff point where they should not differ asymptotically. However, statistical error, imperfect design implementation, and a plethora of different possible analysis options, mean that they might nonetheless differ. We test whether they do, assuming that the bias potential is greater with RDs than RCTs. A second purpose of this paper is to investigate the external validity of RD by exploring how the size of the bias estimates varies across the 15 studies, for they differ in their settings, interventions, analyses, and implementation details. Both Bayesian and frequentist meta-analysis methods show that the RD bias is below 0.01 standard deviations on average, indicating RD's high internal validity. When the study-specific estimates are shrunken to capitalize on the information the other studies provide, all the RD causal estimates fall within 0.07 standard deviations of their RCT counterparts, now indicating high external validity. With unshrunken estimates, the mean RD bias is still essentially zero, but the distribution of RD bias estimates is less tight, especially with smaller samples and when parametric RD analyses are used.

Evaluations of public programs in many fields reveal that different types of programs— or different versions of the same program—vary in their effectiveness. Moreover, a program that is effective for one group of people might not be effective for other groups, and a program that is effective in one set of circumstances may not be effective in other circumstances. This paper presents a conceptual framework for research on such variation in program effects and the sources of this variation. The framework is intended to help researchers—both those who focus mainly on studying program implementation and those who focus mainly on estimating program effects—see how their respective pieces fit together in a way that helps to identify factors that explain variation in program effects, and thereby support more systematic data collection. The ultimate goal of the framework is to enable researchers to offer better guidance to policymakers and program operators on the conditions and practices that are associated with larger and more positive effects. © 2014 by the Association for Public Policy Analysis and Management.

The sharp regression discontinuity design (RDD) has three key weaknesses compared to the randomized clinical trial (RCT). It has lower statistical power, it is more dependent on statistical modeling assumptions, and its treatment effect estimates are limited to the narrow subpopulation of cases immediately around the cutoff, which is rarely of direct scientific or policy interest. This paper examines how adding an untreated comparison to the basic RDD structure can mitigate these three problems. In the example we present, pretest observations on the posttest outcome measure are used to form a comparison RDD function. To assess its performance as a supplement to the basic RDD, we designed a within-study comparison that compares causal estimates and their standard errors for (1) the basic posttest-only RDD, (2) a pretest-supplemented RDD, and (3) an RCT chosen to serve as the causal benchmark. The two RDD designs are constructed from the RCT, and all analyses are replicated with three different assignment cutoffs in three American states. The results show that adding the pretest makes functional form assumptions more transparent. It also produces causal estimates that are more precise than in the posttest-only RDD, but that are nonetheless larger than in the RCT. Neither RDD version shows much bias at the cutoff, and the pretest-supplemented RDD produces causal effects in the region beyond the cutoff that are very similar to the RCT estimates for that same region. Thus, the pretest-supplemented RDD improves on the standard RDD in multiple ways that bring causal estimates and their standard errors closer to those of an RCT, not just at the cutoff, but also away from it.

Evaluations of the impact of social programs are often carried out in multiple sites, such as school districts, housing authorities, local TANF offices, or One-Stop Career Centers. Most evaluations select sites purposively following a process that is nonrandom. Unfortunately, purposive site selection can produce a sample of sites that is not representative of the population of interest for the program. In this paper, we propose a conceptual model of purposive site selection. We begin with the proposition that a purposive sample of sites can usefully be conceptualized as a random sample of sites from some well-defined population, for which the sampling probabilities are unknown and vary across sites. This proposition allows us to derive a formal, yet intuitive, mathematical expression for the bias in the pooled impact estimate when sites are selected purposively. This formula helps us to better understand the consequences of selecting sites purposively, and the factors that contribute to the bias. Additional research is needed to obtain evidence on how large the bias tends to be in actual studies that select sites purposively, and to develop methods to increase the external validity of these studies.

The ability of nonexperimental estimators to match impact estimates derived from random assignment is examined using data from the evaluation of two interdistnct magnet schools. As in previous within-study comparisons, nonexperimental estimates differ from estimates based on random assignment when nonexperimental estimators are implemented without pretreatment measures of academic performance. With comparison groups consisting of students drawn from the same districts or districts with similar student body characteristics as the districts where treatment group students reside, using pretreatment test scores reduces the bias in nonexperimental methods between 64 and 96 percent. Adding pretreatment test scores does not achieve as much bias reduction when the comparison group consists of students drawn from districts with different student body characteristics than the treatment group students' districts. The results suggest that using pretreatment outcome measures and comparison groups that are geographically aligned with the treatment group greatly improves the performance of nonexperimental estimators.

We use experimental survey methods in a nationally representative survey to test alternative ways of identifying (1) individuals in the population who would be better able to work if they received workplace accommodation for a health condition; (2) the rate at which these individuals receive workplace accommodation; and (3) the rate at which accommodated workers are still working four years later, compared to similar workers who were not accommodated. We find that question order in disability surveys matters. We present suggestive evidence of priming effects that lead people to understate accommodation when first asked about very severe disabilities. We also find a sizeable fraction of workers who report they receive a workplace accommodation for a health problem but do not report work limitations per se. Our preferred estimate of the size of the accommodation-sensitive population is 22.8 percent of all working-age adults. We find that 47 to 58 percent of accommodation-sensitive individuals lack accommodation and would benefit from some kind of employer accommodation to either sustain or commence work. Finally, among accommodation-sensitive individuals, workers who were accommodated for a health problem in 2014 were 13.2 percentage points more likely to work in 2018 than those who were not accommodated in 2014.

Randomization bias occurs when the random assignment used to estimate program effects influences the types of individuals that participate in a program. This paper focuses on a form of randomization bias called \"applicant inclusion bias,\" which can occur in evaluations of discretionary programs that normally choose which of the eligible applicants to serve. If this nonrandom selection process is replaced by a process that randomly assigns eligible applicants to receive the intervention or not, the types of individuals served by the program—and thus its average impact on program participants—could be affected. To estimate the impact of discretionary programs for the individuals that they normally serve, we propose an experimental design called Preferred Applicant Random Assignment (PARA). Prior to random assignment, program staff would identify their \"preferred applicants,\" those that they would have chosen to serve. All eligible applicants are randomly assigned, but the probability of assignment to the program is set higher for preferred applicants than for the remaining applicants. This paper demonstrates the feasibility of the method, the cost in terms of increased sample size requirements, and the benefit in terms of improved generalizability to the population normally served by the program.

The doubly randomized preference trial (DRPT) is a randomized experimental design with three arms: a treatment arm, a control arm, and a preference arm. The design has useful properties that have gone unnoticed in the applied and methodological literatures. This paper shows how to interpret the DRPT design using an instrumental variables (IV) framework. The IV framework reveals that the DRPT separately identifies three different treatment effect parameters: the Average Treatment Effect (ATE), the Average Treatment Effect on the Treated (ATT), and the Average Treatment Effect on the Untreated (ATU). The ATE, ATT, and ATU parameters are important for program evaluation research because in realistic settings many social programs are optional rather than mandatory and some people who are eligible for a program choose not to participate. Most of the paper is concerned with the interpretation of the research design. To make the ideas concrete, the final section provides an empirical example using data from an existing DRPT study.