Knowledge that Transforms

Recently there has been a surge in econometric work focusing on estimating average treatment effects under various sets of assumptions. One strand of this literature has developed methods for estimating average treatment effects for a binary treatment under assumptions variously described as exogeneity, unconfoundedness, or selection on observables. The implication of these assumptions is that systematic (for example, average or distributional) differences in outcomes between treated and control units with the same values for the covariates are attributable to the treatment. Recent analysis has considered estimation and inference for average treatment effects under weaker assumptions than typical of the earlier literature by avoiding distributional and functional-form assumptions. Various methods of semiparametric estimation have been proposed, including estimating the unknown regression functions, matching, methods using the propensity score such as weighting and blocking, and combinations of these approaches. In this paper I review the state of this literature and discuss some of its unanswered questions, focusing in particular on the practical implementation of these methods, the plausibility of this exogeneity assumption in economic applications, the relative performance of the various semiparametric estimators when the key assumptions (unconfoundedness and overlap) are satisfied, alternative estimands such as quantile treatment effects, and alternate methods such as Bayesian inference.

This paper examines the importance of format in aid programs, focusing on state appropriations to public postsecondary institutions. These funds subsidize costs for in-state students, but they may also influence choices between institutions due to their in-kind format. Using the conditional logistic choice model and extensive match-specific information, the paper approximates the choice between nearly 2700 college options to examine the effect of several dissimilar state systems. The results suggest that the level and distribution pattern of subsidies strongly affect decisions. If the aid could instead be applied to any in-state college, up to 29% more students would prefer to attend private four-year colleges.

In previous research, a substantial gap in test scores between white and black students persists, even after controlling for a wide range of observable characteristics. Using a newly available data set (the Early Childhood Longitudinal Study), we demonstrate that in stark contrast to earlier studies, the black-white test score gap among incoming kindergartners disappears when we control for a small number of covariates. Real gains by black children in recent cohorts appear to play an important role in explaining the differences between our findings and earlier research. The availability of better covariates also contributes. Over the first two years of school, however, blacks lose substantial ground relative to other races. There is suggestive evidence that differences in school quality may be an important part of the explanation. None of the other hypotheses we test to explain why blacks are losing ground receive any empirical backing.

Over fifty years ago, Abram Bergson, Janet Chapman, and others sought to assess the large gap in Soviet statistics between the published wage bill and the full wage bill; and Frank Lorimer drew attention to the related gap between the employment data in the 1939 population census and in the annual employment returns. Soviet archives which have recently been declassified reveal the considerable extent to which the findings of the Western authors were accurate.

This paper develops a model of intrahousehold allocation with endogenous fertility, which captures the relationship between birth order and investment in children. It shows that a birth order effect in intrahouse hold allocation can arise even without assumptions about parental preferences for specific birth orders of children or genetic endowments varying by birth order. The important contribution is that fertility is treated as endogenous, a possibility that other models of intrahousehold allocation have ignored. The implications of the model are that children with higher birth orders (that is, who are born later) have an advantage over siblings with lower birth orders, and that parents who are inequality-averse will not have more than one child. The model furthermore shows that not taking account of the endogeneity of fertility when analyzing intrahousehold allocation may seriously bias the results. The effects of a child's birth order on its human capital accumulation are analyzed using a longitudinal data set from the Philippines that covers a very long period. We examine the effects of birth order on both number of hours in school during education and completed education. The results for both are consistent with the predictions of the model.

Although the theoretical importance of expectations in decision-making is well known to economists, only a few empirical papers investigate the impact of individual subjective expectations on economic outcomes. This paper examines the link between expectations of future job losses and the subsequent impact that these expectations have on household consumption behavior. The first part of the paper documents the empirical relationship between job loss expectations and subsequent job losses. Subjective job loss expectations have significant predictive power in explaining future job losses even when standard demographic information known to be associated with the prevalence of job displacement is included in the analysis. Furthermore, higher subjective job loss probabilities are correlated with an increased expectation of future earnings declines. Overall, these results indicate that the variable for subjective job loss expectations is a meaningful predictor of subsequent displacement. Since a job displacement results in large and persistent earnings losses, job loss expectations should have an important impact on household consumption smoothing following a job loss. The second part of the paper finds that although a job loss significantly reduces household consumption, there is little evidence that the degree to which households anticipate job losses reduces the impact of displacement on consumption. Alternative models of interpreting responses to expectations questions and of household consumption behavior that may explain these results are discussed.

In the many years since their introduction, positive theories of inflation have rarely been tested. This paper documents a negative relationship between inflation and the welfare state (proxied by the parameters of the unemployment benefit program) that is to be expected in such theories. Because unemployment benefits make the monetary authority less concerned about the plight of the unemployed, building a welfare state has a similar effect to appointing a conservative central banker. The relationship holds in a panel of 20 OECD countries over the period 1961–1992, a region where Romer finds no evidence of commitment problems. It holds controlling for country and time fixed effects, country specific time trends, other covariates, and using a decadal panel. Interpreted as causal, the estimated effect is economically large: a 1-standard deviation decrease in benefit duration is predicted to add 1.4 percentage points onto inflation, or 31% of the standard deviation in inflation.

A panel data estimation finds a high speed of income convergence among the U.S. states. However, initial incomes show a pattern which is difficult to explain by the estimated model. A simulation study shows that this pattern can be explained much more naturally when we assume that true convergence is slow.

Utilizing a utility-maximizing, Roy-type, discrete choice model of worker location in Canadian provinces and U.S. states that incorporates returns to skill, amenities, fixed costs, distance, language, and border effects, we find that individuals with higher skills migrate to areas with higher returns and that the 49th parallel attenuates migration. Simulations indicate that equalizing returns in the two countries has a modest effect on cross-country migration; however, reductions in border effects tend to have large nonlinear effects on it. Our results confirm the qualitative results of previous research emphasizing the importance of returns to skill and border effects in migration decisions.

Most panel data studies of intertemporal labor supply assume classical measurement error. Recent validation studies refute this assumption. In this study I address nonclassical measurement error explicitly. I use data on males from the Panel Study of Income Dynamics Validation Study to purge measurement error from the Panel Study of Income Dynamics. I find a large amount of predictable wage variation in the data, even after allowing for measurement error. However, there is almost no labor supply response to these predictable wage changes. Therefore, failure to control for nonclassical measurement error cannot explain the low estimated labor supply elasticities in other papers.