Knowledge that Transforms

This paper shows how to estimate a model in which an unknown transformation of the dependent variable is a linear function of explanatory variables plus an unobserved random variable, U, whose distribution is unknown. The model nests many familiar parametric and semiparametric models, including models with Box-Cox transformed dependent variables and proportional hazards models with and without unobserved heterogeneity. The paper develops root-n consistent, asymptotically normal estimators of the transformation function, coefficients of the explanatory variables, and distribution of U. The results of Monte Carlo experiments indicate that the estimators work well in samples of size one hundred.

Typically, work on mechanism design has assumed that all private information can be captured in a single scalar variable. This paper explores one way in which this assumption can be relaxed in the context of the multiproduct nonlinear pricing problem. It is shown that the firm will choose to exclude some low-value consumers from all markets. A class of cases that allow explicit solution is derived by making use of a multivariate form of 'integration by parts.' In such cases the optimal tariff is cost-based.

This paper extends the spatial theory of voting to the case in which policy choices depend upon the interaction between executive and the legislature. Voters are strategic and to analyze equilibrium the authors apply 'coalition proof' type refinements. The model has implications consistent with voting behavior in the United States: (1) split-ticket with some voters choosing one party for the presidency and the other for Congress; (2) for some parameter values, a divided government with different parties controlling the executive and the majority of the legislature; and (3) the mid-term electoral cycle with the party holding the presidency always losing votes in mid-term legislative elections.

The aim of this paper is to explain the fact that certain properties of binary relations are frequently observed in natural language while others do not appear at all. Three features of binary relation are studied: (1) The ability to use the relation to indicate nameless elements. (2) The accuracy with which the vocabulary spanned by the relation can be used to approximate the actual terms to which a user of the language wishes to refer. (3) The ease with which the relation can be described by means of examples. It is argued that linear orderings are optimal according to the first criteria while asymmetric relations are optimal according to second. From among complete and asymmetric relations (tournaments) those which are transitive are optimal according to the third criterion.

Robin L. Lumsdaine, Consistency and Asymptotic Normality of the Quasi-Maximum Likelihood Estimator in IGARCH(1,1) and Covariance Stationary GARCH(1,1) Models, Econometrica, Vol. 64, No. 3 (May, 1996), pp. 575-596

The authors introduce and analyze 'multistage situations, ' which generalize 'multistage games' (which, in turn, generalize 'repeated games'). One reason for this generalization is to avoid the perhaps unrealistic constraints--inherent to noncooperative games--that the set of strategy tuples must be a Cartesian product of the strategy sets of the players. Another reason is that, in most economic and social activities (e.g., in sequential bargaining without a rigid protocol), the 'rules of the game' are rather amorphous; the procedures are rarely pinned down. Such social environments can, however, be represented as multistage situations and be effectively analyzed through the theory of social situations.

The SNP density estimator is a truncation (or sieve) estimator based upon a Hermite series expansion. Other nonparametric estimators based on orthogonal series expansions have been proposed.

This paper proposes some tests for parameter constancy in linear regression models with possible infinite variance.Both dynamic and trending regressors are allowed.The tests are based on the empirical distribution function of estimated residuals and are shown to have non-trivial local power against a wide range of alternatives.Within a certain class of alternatives including simple shifts, the tests have higher power for testing the simple shift alternatives.These tests are formulated in such a way that the limiting variables are distribution-free.The residuals may be obtained based on any root-n consistent estimator (under the null) of regression parameters.As part of these results, some weak convergence for weighted sequential empirical processes of residuals is established.

This paper develops procedures for inference about the moments of smooth functions of out-of-sample predictions and prediction errors when there is a long time series of predictions and realizations. The aim is to provide tools for analysis of predictive accuracy and efficiency and, more generally, of predictive ability. The paper allows for nonnested and nonlinear models as well as for possible dependence of predictions and prediction errors on estimated regression parameters. Simulations indicate that the procedures can work well in samples of size typically available.

Costas Meghir, Guglielmo Weber, Intertemporal Nonseparability or Borrowing Restrictions? A Disaggregate Analysis using a U.S. Consumption Panel, Econometrica, Vol. 64, No. 5 (Sep., 1996), pp. 1151-1181