We study model selection strategies based on penalized empirical loss minimization. We point out a tight relationship between error estimation and data-based complexity penalization: any good error estimate may be converted into a data-based penalty function and the performance of the estimate...

Human capital, productivity and physical capital are considered the main factors in the economies’ GDP per capita determination. According to the neoclassical approach, human capital accumulation explains about a third of the variation in per capita income across countries. However, there is...

In this paper, we evaluate the income convergence hypothesis in Latin America and East Asia between 1960 and 2000 through the use of quantile regressions to estimate growth equations. This approach allows us to assess how the effect of policy variables on per worker income growth rate can vary...

This paper studies the determination of the surpluses of the Brazilian central government. For quarterly data from 1996 through 2011, the empirical strategy included: i) structural break models and nonlinear estimates on the level of public debt, to handle the inertia after 1999, for high levels...

This paper presents new econometric specifications for the Brazilian exports in the period 1995-2009 using data from the Quarterly Accounts and allowing for nonlinearities. In the cointegrating vector, there is evidence of a level shift, but the elasticities have not changed significantly. The...