Principios básicos de microeconometría y del uso de variables instrumentales para la inferencia causal

Abstract

Microeconometrics with its causal identification logic and the microdata use, leads to a set of specific measurement and modeling techniques in applied economics. This paper shows basic elements of microeconometrics, taking account that the peak of its application goes together with availability rise of cross-sectional and panel-type data. It describes three representative features of microeconometry: i) strategies of causal identification for isolating the correlation from omitted variables or from sample selection; ii) non-linear modeling for discrete and censored variables such as Logit, Probit or Tobit models; and iii) reduction of assumptions about the distributions and the flexibilization of functional forms. In addition, a brief exposition is made of the frequent problem of endogeneity and the use of instrumental variables as a simple but powerful solution strategy; and finally, it is recalled that the microeconometric techniques without the rest of the elements that shape the research set, loses its potential. (English) [ABSTRACT FROM AUTHOR]