Inflation and long term growth: bad predictions, good models?

  • Santiago Javier Fucci Universidad de Buenos Aires

    Docente auxiliar de Crecimiento Económico y Cuentas Nacionales en la Universidad de Buenos Aires. Especialista en tópicos de econometría y en data mining aplicado al sector financiero.

  • Martín Grandes Universidad de Buenos Aires

    Doctor en Economía por la Ecole des Hautes Etudes en Sciences Sociales de Paris (2004). Investigador del CONICET. Docente titular en la Universidad de Buenos Aires, Universidad Católica Argentina y en la Escuela de Gobierno UTDT. Profesor invitado en otras universidades de Francia, Estados Unidos, Colombia, México y China.

Keywords: Inflation, economic growth, macroeconomic modelling


It is a fact well documented by economic theory that high and durable inflation has a negative effect on economic growth, but that as the effects on the latter decline, they are non-linear. This work answers two research questions related to the works that have studied this empirical relationship in the last three decades. The first is about its external validity: How useful are they to predict the economic evolution of emerging countries? Simulating the growth of a sample of emerging markets for the past 20 years from these models, we found high and disparate prediction errors and therefore certain drawbacks when generalizing their conclusions. Understanding why this happens gives rise to the second question, linked to the internal validity of the models: What unresolved problems exist in the estimates carried out by the authors? the answer is that there are problems of specification, simultaneity and omission of variables, which generate endogeneity and hinder the causal identification of the effect of inflation on growth in the type of regressions under analysis, even in models with thresholds and non-linear.


Download data is not yet available.
How to Cite
Fucci, S. J., & Grandes, M. (2018). Inflation and long term growth: bad predictions, good models?. Economia, 41(82), 83-123.