Predicting the Effects of Permanent Programs from a Limited Duration Experiment.

TLDR: In this article, the authors examine and interpret experimental data in the context of a formal model of labor supply which permits an explicit consideration of the issue of what can be inferred from a limited duration experiment about the behavioral effects of a negative income tax if such a program were adopted permanently.

Abstract: Among the many problems involved in interpreting the results of the New Jersey-Pennsylvania experiment is the determination of what can be inferred from a limited duration experiment about the behavioral effects of a negative income tax if such a program were adopted permanently. In this paper we examine and interpret experimental data in the context of a formal model of labor supply which permits an explicit consideration of this issue. Section I summarizes a formal model of labor supply which indicates that the response of a "rational" individual to a temporary negative income tax would differ from his response to a permanent program, and that a limited duration experiment will correspondingly yield a biased prediction of "permanent" behavior.' Qualitative statements are made about the nature of the biases and their relationship to interest or discount rates, time horizons, and concepts of intertemporal substitution. Section II discusses strategies for empirical estimation of the predicted effects of a permanent negative income tax from what is, in principle, observable from an "ideally" constructed experiment; alternative strategies to handle data limitations implicit in the experiment are then proposed. Section III presents the empirical estimates required to project the labor-supply effects of a permanent negative income tax and interprets the conventionally reported results of the experiment in light of these estimates. Section IV summarizes the basic empirical results presented in the paper.