Bounds on Average and Quantile Treatment Effects on Duration Outcomes Under Censoring,Selection, and Noncompliance

Abstract

We consider the problem of assessing the effects of a treatment on duration outcomes using data from a randomized evaluation with noncompliance. For such settings, we derive nonparametric sharp bounds for average and quantile treatment effects addressing three pervasive problems simultaneously: self-selection into the spell of interest, endogenous censoring of the duration outcome, and noncompliance with the assigned treatment. Ignoring any of these issues could yield biased estimates of the effects. Notably, the proposed bounds do not impose the independent censoring assumption—which is commonly used to address censoring but is likely to fail in important settings—or exclusion restrictions to address endogeneity of censoring and selection. Instead, they employ monotonicity and stochastic dominance assumptions. To illustrate the use of these bounds we assess the effects of the Job Corps (JC) training program on its participants’ last complete employment spell duration. Our estimated bounds suggest that JC participation may increase the average duration of the last complete employment spell before week 208 after randomization by at least 5.6 log points (5.8%) for individuals who comply with their treatment assignment and experience a complete employment spell whether or not they enrolled in JC. The estimated quantile treatment effects suggest the impacts may be heterogeneous, and strengthen our conclusions based on the estimated average effects.     

OPTIMAL IV ESTIMATION OF SYSTEMS WITH STOCHASTIC REGRESSORS AND VAR DISTURBANCES WITH APPLICATIONS TO DYNAMIC SYSTEMS

This paper considers the general problem of Feasible Generalized Least Squares Instrumental Variables (FGLS IV) estimation using optimal instruments. First we summarize the sufficient conditions for the FGLS IV estimator to be asymptotically equivalent to an optimal GLS IV estimator. Then we specialize to stationary dynamic systems with stationary VAR errors, and use the sufficient conditions to derive new moment conditions for these models. These moment conditions produce useful IVs from the lagged endogenous variables, despite the correlation between errors and endogenous variables. This use of the information contained in the lagged endogenous variables expands the class of IV estimators under consideration and thereby potentially improves both asymptotic and small-sample efficiency of the optimal IV estimator in the class. Some Monte Carlo experiments compare the new methods with those of Hatanaka (1976). For the DGP used in the Monte Carlo experiments, asymptotic efficiency is strictly improved by the new IVs, and experimental small-sample efficiency is improved as well.     

Revised GRASP with path-relinking for the linear ordering problem

The linear ordering problem (LOP) is an NP\mathcal{NP}-hard combinatorial optimization problem with a wide range of applications in economics, archaeology, the social sciences, scheduling, and biology. It has, however, drawn little attention compared to other closely related problems such as the quadratic assignment problem and the traveling salesman problem. Due to its computational complexity, it is essential in practice to develop solution approaches to rapidly search for solution of high-quality. In this paper we propose a new algorithm based on a greedy randomized adaptive search procedure (GRASP) to efficiently solve the LOP. The algorithm is integrated with a Path-Relinking (PR) procedure and a new local search scheme. We tested our implementation on the set of 49 real-world instances of input-output tables (LOLIB instances) proposed in Reinelt (Linear ordering library (LOLIB) 2002). In addition, we tested a set of 30 large randomly-generated instances proposed in Mitchell (Computational experience with an interior point cutting plane algorithm, Tech. rep., Mathematical Sciences, Rensellaer Polytechnic Institute, Troy, NY 12180-3590, USA 1997). Most of the LOLIB instances were solved to optimality within 0.87 seconds on average. The average gap for the randomly-generated instances was 0.0173% with an average running time of 21.98 seconds. The results indicate the efficiency and high-quality of the proposed heuristic procedure.     

The work process setting and situational contexts based on socially distributed cognition: an interactive,cognitive and social proposal of analysis

To carry out an ethnographic study on the work process in the sterilization unit of a hospital in Catalonia (Spain), we found the socially distributed cognition approaches of Hutchins and Kirsh useful. However, these approaches lack sufficient explanation on three important issues: (a) the pragmatic criteria for identifying and delimiting a relevant unit of analysis and therefore the setting and contexts of the work process; (b) the mechanisms and results of reciprocal influences between these levels of analysis; and (c) the relation between these levels. Therefore, we added several new elements to these approaches, in addition to Layder's model of contexts with some important modifications, with the aim of offering an interactive, cognitive and social proposal of analysis that clarifies these three issues, allowing a more exhaustive and broader empirical analysis that captures better the ‘social’ dimension embedded in the work processes.     

What is Recklessness in Scientific Research? The Frank Sauer Case

On May 22, 2017, administrative law Judge Leslie Rogall of the Department of Health and Human Services’ Departmental Appeals Board, Civil Remedies Division, ruled in favor of the Office of Research Integrity (ORI) concerning its decision to charge former University of California at Riverside biochemistry professor Frank Sauer with research misconduct for fabricating or falsifying digital image data included in three papers and seven grant applications submitted to the National Institutes of Health. More specifically, Sauer was deemed responsible for manipulating, reusing, and falsely labeling images of autoradiograms and gels in his research in epigenetics. One month after this decision, ORI announced its final ruling concerning Sauer, which barred him from serving in any advisory capacity to the Public Health Services and required him to retract affected papers. The case raises some interesting and important questions concerning research integrity because it focused on the legal issue of what constitutes recklessness in scientific research.     

On sufficient conditions for the comparison of some quantile-based measures

In this article, we focus on characterizations and sufficient conditions for the comparison of some quantile-based measures. Mainly we focus on crossing conditions and changes of monotonicity of the quantile functions. The results are given for the comparison of the tail value at risk, total time on test, and the expected proportional shortfall measures.     

Prevalence of Problem Gambling Among Adolescents: A Comparison Across Modes of Access,Gambling Activities,and Levels of Severity

The high availability and accessibility of online gambling have recently caused public concern regarding the potential increase of gambling-related problems among young people. Nonetheless, few studies among adults and none among adolescents have explored specific characteristics of gamblers as a function of gambling venues to date. This study sought to analyze the prevalence of gambling among a sample of adolescents in the last year, as well as sociodemographic and gambling-related characteristics as possible predictors of at-risk and problem gambling. The sample comprised 1313 adolescents aged 14–18 years. Participants were asked to respond to several questions regarding their gambling behavior. Chi square and ANOVA tests were performed in order to explore differences between groups, and a set of multinomial regressions established significant severity predictors. The prevalence of at-risk and problem gambling was 4 and 1.2 %, respectively. Regression analyses showed that having a relative with gambling problems predicted at-risk gambling. Both living with only one parent or not living with parents at all, and the prevalence of Electronic Gambling Machines in the last year were associated with problem gambling. Mixed-mode gambling was a predictor of both at-risk and problem gambling. Our findings extend previous research on gambling among adolescents by exploring gambling behavior according to different modes of access. Although the prevalence of exclusive online gambling among the total sample was low, these results support the need to consider specific subgroups of gamblers and their concrete related features when conducting both indicated prevention and treatment protocols for adolescents.     

Spanish validation of the Servant Leadership Short Scale for young adults and adolescents / Validación de la Escala de Liderazgo de Servicio Versión Abreviada en español para jóvenes adultos y adolescentes

The aim of the current study was to validate a shortened Spanish version of the Servant Leadership Profile (SLP). In Study 1, a number of experts evaluated and reduced the original scale to 14 items and named the revised version the Servant Leadership Short Scale (SLSS). An exploratory factor analysis conducted with a sample of 1,001 students revealed two possible structure solutions, a one-factor solution and a two-correlated factor solution. Study 2 used a different sample of 456 students to examine the explored structures using confirmatory factor analysis. A one-factor solution was proposed (χ²/df = 3.4, CFI = .91, TLI = .90, RMSEA = .07, SRMR = .06). Finally, in Study 3, servant leadership was used as a predictor of self-efficacy (b = .685, p < .001) and demonstrated good predictive validity. This model explained 41.55% of the variance on self-efficacy. Reliability analyses of the three studies showed that this 14-item Spanish version had high internal consistency. Together, the results of the three studies support the application of the SLSS in future research.     

The role of right-wing authoritarianism,sexism and culture of honour in rape myths acceptance / El papel del autoritarismo de derechas,del sexismo y de la cultura del honor en la aceptación de los mitos sobre la violación

The purpose of this work is to analyse how some ideological variables (ambivalent sexism, culture of honour, and right-wing authoritarianism) affect rape myths acceptance (RMA). Two hundred and fourteen university students (83 men and 131 women), mean age 20.39 years old, participated. The results show that there is a high correlation between RMA and hostile sexism, benevolent sexism, culture of honour, and right-wing authoritarianism. Hostile sexism and right-wing authoritarianism predict RMA in both male and female samples. Benevolent sexism predicts RMA in the female sample. In addition, in the male sample right-wing authoritarianism interacts with hostile sexism to predict RMA, while in the female sample right-wing authoritarianism interacts with benevolent sexism to predict RMA. These results show the close relationship existing between certain ideological variables and RMA.     

Employment Quality: Are There Differences by Types of Contract?

Measuring welfare state generosity in developing and transitional welfare states is often challenged not only by lack of comparative quantitative data, but also by issues of conceptual stretching. This paper demonstrates and discusses the use of one of the key measures of welfare entitlement generosity developed in the comparative welfare state research in the context of post-communist countries of Central and Eastern Europe (CEE). With the new time series data provided by the Comparative Welfare Entitlements Dataset CWED2, comparative approaches including the CEE countries have become feasible. This paper first discusses quantitative measures of welfare entitlement generosity in the tradition of the social rights of citizenship approach and how they can be applied for cross-country comparisons. It then demonstrates empirically how the emerging CEE welfare states’ generosity compares to mature “old” OECD welfare states. Finally, the paper shows the potential and the pitfalls of quantitative measures of welfare state generosity by discussing, to what extent do indicators of social security scheme generosity measure the same in established and emerging welfare states, which functional equivalents may be relevant in the context of emerging welfare states and how far can we stretch our theoretical concepts.