Knowledge that Transforms

We provide easy to verify sufficient conditions for the consistency and asymptotic normality of a class of semiparametric optimization estimators where the criterion function does not obey standard smoothness conditions and simultaneously depends on some nonparametric estimators that can themselves depend on the parameters to be estimated. Our results extend existing theories such as those of Pakes and Pollard (1989), Andrews (1994a), and Newey (1994). We also show that bootstrap provides asymptotically correct confidence regions for the finite dimensional parameters. We apply our results to two examples: a ‘hit rate’ and a partially linear median regression with some endogenous regressors.

We consider the problem of constructing a portfolio of finitely many assets whose returns are described by a discrete joint distribution.We propose mean-risk models that are solvable by linear programming and generate portfolios whose returns are nondominated in the sense of second-order stochastic dominance. Next, we develop a specialized parametric method for recovering the entire mean-risk efficient frontiers of these models and we illustrate its operation on a large data set involving thousands of assets and realizations.

This paper provides deterministic approximation results for stochastic processes that arise when ¯nite populations recurrently play ¯nite games.The deterministic approximation is de¯ned in continuous time as a system of ordinary di®erential equations of the type studied in evolutionary game theory.We establish precise connections between the long-run behavior of the stochastic process, for large populations, and its deterministic approximation.In particular, we show that if the deterministic solution through the initial state of the stochastic process at some point in time enters a basin of attraction, then the stochastic process will enter any given neighborhood of that attractor within a ¯nite and deterministic time with a probability that exponentially approaches one as the population size goes to in¯nity.The process will remain in this neighborhood for a random time that almost surely exceeds an exponential function of the population size.During this time interval, the process spends almost all time at a certain subset of the attractor, its so-called Birkho® center.We sharpen this result in the special case of ergodic processes.

This paper analyzes the linear regression model y = x + with a conditional median assumption Med( j z) = 0 where z is a vector of exogenous random variables. Added complication arise due to the censoring of the outcome y. We treat the censored model as a model with interval-observed outcomes thus obtaining an incomplete model with inequality restrictions on conditional median regressions. This allows us to use the estimator introduced by Manski and Tamer (2000) to analyze the information contained in these inequality restrictions. We give identication conditions in the absence of censoring and introduce a p N-consistent estimator based on the minimum distance method. We then give suÆcient conditions for global identication of with censored y and endogenous x. In the case of interval data on y and endogenous x, we provide a set-consistent estimator that is based on a modied minimum distance method. In the case where we have point identication, we show that the estimator is p N-normal and derive its asymptotic distribution with a feasible asymptotic variance. A Montecarlo analysis illustrates our estimator. We thank Bo Honore for comments and the Econometrics Research Program at Princeton for support.

We are interested in estimating the average effect of a binary treatment on a scalar outcome. If assignment to the treatment is exogenous or unconfounded, that is, independent of the potential outcomes given covariates, biases associated with simple treatment-control average comparisons can be removed by adjusting for differences in the covariates. Rosenbaum and Rubin (1983) show that adjusting solely for differences between treated and control units in the propensity score removes all biases associated with differences in covariates. Although adjusting for differences in the propensity score removes all the bias, this can come at the expense of efficiency, as shown by Hahn (1998), Heckman, Ichimura, and Todd (1998), and Robins, Mark, and Newey (1992). We show that weighting by the inverse of a nonparametric estimate of the propensity score, rather than the true propensity score, leads to an efficient estimate of the average treatment effect. We provide intuition for this result by showing that this estimator can be interpreted as an empirical likelihood estimator that efficiently incorporates the information about the propensity score.

This paper builds a dynamic industry model with heterogeneous firms that explains why international trade induces reallocations of resources among firms in an industry. The paper shows how the exposure to trade will induce only the more productive firms to enter the export market (while some less productive firms continue to produce only for the domestic market) and will simultaneously force the least productive firms to exit. It then shows how further increases in the industry's exposure to trade lead to additional inter-firm reallocations towards more productive firms. These phenomena have been empirically documented but can not be explained by current general equilibrium trade models, because they rely on a representative firm framework. The paper also shows how the aggregate industry productivity growth generated by the reallocations contributes to a welfare gain, thus highlighting a benefit from trade that has not been examined theoretically before. The paper adapts Hopenhayn's (1992a) dynamic industry model to monopolistic competition in a general equilibrium setting. In so doing, the paper provides an extension of Krugman's (1980) trade model that incorporates firm level productivity differences. Firms with different productivity levels coexist in an industry because each firm faces initial uncertainty concerning its productivity before making an irreversible investment to enter the industry.

Methods are proposed for testing stochastic dominance of any pre–specified order, with primary interest in the distributions of income. We consider consistent tests, that are similar to Kolmogorov–Smirnov tests, of the complete set of restrictions that relate to the various forms of stochastic dominance. For such tests, in the case of tests for stochastic dominance beyond first order, we propose and justify a variety of approaches to inference based on simulation and the bootstrap. We compare these approaches to one another and to alternative approaches based on multiple comparisons in the context of a Monte Carlo experiment and an empirical example.

This paper develops a general method for constructing exactly similar tests based on the conditional distribution of nonpivotal statistics in a simultaneous equations model with normal errors and known reduced-form covariance matrix. These tests are shown to be similar under weak-instrument asymptotics when the reduced-form covariance matrix is estimated and the errors are non-normal. The conditional test based on the likelihood ratio statistic is particularly simple and has good power properties. Like the score test, it is optimal under the usual local-to-null asymptotics, but it has better power when identification is weak.

Many refinements of Nash equilibrium yield solution correspondences that do not have closed graph in the space of payoffs or information. This has significance for implementation theory, especially under complete information. If a planner is concerned that all equilibria of his mechanism yield a desired outcome, and entertains the possibility that players may have even the slightest uncertainty about payoffs, then the planner should insist on a solution concept with closed graph. We show that this requirement entails substantial restrictions on the set of implementable social choice rules. In particular, when preferences are strict (or more generally, hedonic), while almost any social choice function can be implemented in undominated Nash equilibrium, only monotonic social choice functions can be implemented in the closure of the undominated Nash correspondence.

This paper considers tests for structural instability of short duration, such as at the end of the sample. The key feature of the testing problem is that the number, m, of observations in the period of potential change is relatively small—possibly as small as one. The well-known F test of Chow (1960) for this problem only applies in a linear regression model with normally distributed iid errors and strictly exogenous regressors, even when the total number of observations, n+m, is large. We generalize the F test to cover regression models with much more general error processes, regressors that are not strictly exogenous, and estimation by instrumental variables as well as least squares. In addition, we extend the F test to nonlinear models estimated by generalized method of moments and maximum likelihood. Asymptotic critical values that are valid as n→∞ with m fixed are provided using a subsampling-like method. The results apply quite generally to processes that are strictly stationary and ergodic under the null hypothesis of no structural instability.