Jason Abrevaya

Misclassification of the dependent variable in a discrete-response setting

Abstract Misclassification of dependent variables in a discrete-response model causes inconsistent coefficient estimates when traditional estimation techniques (e.g., probit or logit) are used. A modified maximum likelihood estimator that corrects for misclassification is proposed. A semiparametric approach, which combines the maximum rank correlation estimator of Han (1987) (Journal of Econometrics 35, 303–316) with isotonic regression, allows for more general forms of misclassification than the maximum likelihood approach. The parametric and semiparametric estimation techniques are applied to a model of job change with two commonly used data sets, the Current Population Survey (CPS) and the Panel Study of Income Dynamics (PSID).

The Effects of Birth Inputs on Birthweight

Unobserved heterogeneity among childbearing women makes it difficult to isolate the causal effects of smoking and prenatal care on birth outcomes (such as birthweight). Whether a mother smokes, for instance, is likely to be correlated with unobserved characteristics of the mother. This article controls for such unobserved heterogeneity by using state-level panel data on maternally linked births. A quantile-estimation approach, motivated by a correlated random-effects model, is used to estimate the effects of smoking and other observables (number of prenatal-care visits, years of education, and so on) on the entire birthweight distribution.

Semiparametric Estimation with Mismeasured Dependent Variables: An Application to Duration Models for Unemployment Spells

This paper considers mismeasurement of the dependent variable in a general linear index model, which includes qualitative choice models, proportional and additive hazard models, and censored models as special cases. The monotone rank estimator of Cavanagh and Sherman [1998] is shown to be consistent in the presence of any mismeasurement process that obeys a simple stochastic-dominance condition. The emphasis is on measurement error which is independent of the covariates, but extensions to covariate-dependent measurement error are also discussed. We consider the proportional hazard duration model in detail and apply the estimator to mismeasured unemployment duration data from the Survey of Income and Program Participation (SIPP).

Leapfrog estimation of a fixed-effects model with unknown transformation of the dependent variable

Abstract This paper considers a fixed-effects panel version of the linear transformation model, in which the dependent variable is h(yt) for an unspecified, strictly monotonic h. Examples of the model include the multiple-spell proportional hazards model and dependent-variable transformation models (e.g., the Box–Cox model) with fixed effects. A semiparametric estimator, called the leapfrog estimator, is introduced and shown to be n -consistent and asymptotically normal. The leapfrog estimator allows for h to vary over time and for heteroskedasticity across observational units. Related semiparametric estimators are considered, and a general covariance result for estimators based on second-order U-processes is presented.

Rank estimation of a generalized fixed-effects regression model

Abstract This paper considers estimation of a fixed-effects version of the generalized regression model of Han (1987, Journal of Econometrics 35, 303–316). The model allows for censoring, places no parametric assumptions on the error disturbances, and allows the fixed effects to be correlated with the covariates. We introduce a class of rank estimators that consistently estimate the coefficients in the generalized fixed-effects regression model. The maximum score estimator for the binary choice fixed-effects model is part of this class. Like the maximum score estimator, the class of rank estimators converge at less than the n rate. Smoothed versions of these estimators, however, converge at rates approaching the n rate. In a version of the model that allows for truncated data, a sufficient condition for consistency of the estimators is that the error disturbances have an increasing hazard function.

The equivalence of two estimators of the fixed-effects logit model

Abstract This note shows that the scale-adjusted maximum likelihood estimator suggested by a result of [ Andersen, 1973 . Conditional inference and models for measuring. Mentalhygiejnisk Forlag, Copenhagen.] is equivalent to the conditional logit estimator of [ Chamberlain, 1980 . Analysis of covariance with qualitative data. Review of Economic Studies 47, 225–238.] for the fixed-effects logit model with two time periods.

Temper, Temperature, and Temptation Heat-Related Retaliation in Baseball

In this study, we analyzed data from 57,293 Major League Baseball games to test whether high temperatures interact with provocation to increase the likelihood that batters will be hit by a pitch. Controlling for a number of other variables, we conducted analyses showing that the probability of a pitcher hitting a batter increases sharply at high temperatures when more of the pitcher’s teammates have been hit by the opposing team earlier in the game. We suggest that high temperatures increase retaliation by increasing hostile attributions when teammates are hit by a pitch and by lowering inhibitions against retaliation.

Ladder tournaments and underdogs: lessons from professional bowling

Abstract This paper examines performance in ladder tournaments, using data from professional bowling competitions. Although there is a distinct advantage to starting with a higher rank in a ladder tournament, the results indicate that “underdogs” win more often than expected. The findings are explained by regression-to-the-mean and hot-hand theories, both of which are consistent with the particular structure of the ladder tournament.

Adoption of Technology-Mediated Distance Education: A longitudinal analysis

We investigated the factors that facilitated the fast adoption and utilization of Technology-Mediated Distance Education (TMDE) among higher education institutions. Our analysis was based on a rich data set on the utilization of TMDE between the 1997-1998 and 2000-2001 academic years. The analysis showed that size, public/private status, and location significantly predicted its actual adoption. Being in an urban location negatively affected enrollment in the courses at the undergraduate but not at the graduate level. While the intent to adopt TMDE correlated significantly with actual adoption, many schools that were not interested in TMDE in 1997-1998 adopted it by 2000-2001. Interestingly, late adopters utilized certain technologies as frequently as early adopters, such as synchronous Internet-based instruction and the use of CD-ROMs.

Fit to be tied: the incentive effects of overtime rules in professional hockey.

This article examines the incentive effects of different payoff structures in the National Hockey League. A rule change prior to the 1999-2000 season, which changed the way that teams were awarded points in overtime games, provides a natural experiment to test the reaction of teams to a change in the payoff structure. The rule change had the desired effect of increasing excitement during overtime play. This article shows, however, that the rule also had the effect of increasing the frequency of overtime games, an effect predictable from the change in incentives but one not intended by the league.