¿Ciencia económica o el arte de hacer economía? Metodología científica y replicaciones en economía

Abstract

The aim of this article is to argue that for the economics to be fully settled as a science, and not simply as an art based on intuition, it has to follow certain scientific principles. The article contributes to the debate on the matter, specifically on two methodological aspects: 1) the proper analysis of data to confirm the hypothesis of a research, and 2) the replication of scientific findings in order to evaluate them. The first aspect, based on Haavelmo's monograph of 1944 entitled “The probabilistic approach to econometrics”, implies that economists should make sure that the probabilistic assumptions of the statistical models used in their research are met. The second aspect suggests that the replications are a means to show the fragility (arose by errors of mishandling databases or inadequate specification of statistical models), or the strength (arose from the robustness of the results from changes in time period, the economy under analysis, the specification of the variables, the estimation methods, etc.) of the results reported in a published research. The article contributes with an illustration of our own methodological proposal: it is a replication of an article by Kakkar (2001), entitled “Long run actual exchange rates: Evidence from Mexico”, published in the journal Economics Letters.