​​Desarrollamos innovación metodológica en técnicas econométricas y aprendizaje de máquina fundamentados en un sólido marco teórico y computacional, aplicada a la identificación de efectos causales y pronóstico.​​

Proyec​​to 1

Where do undercover criminals live? A​ Spatial one-sided errors model

We analyze the rates of arrested individuals living per neighborhood in Medellín (Colombia) to identify neighborhoods where relatively more uncaptured criminals potentially live. We exploit a novel dataset consisting of confidential reports of residential addresses of individuals at the moment of being arrested by the police. We find that undercover hot regions depend both on the crime typology and time. Our model takes into account persistent and transient one-sided error components and spatial effects within a panel data structure. Simulation exercises suggest that our proposal has good finite sample performance. In particular, we find that the predictive intervals have accurate coverage, which is robust to endogeneity. Moreover, our proposal has good predictive accuracy since the hot regions are well-classified, even in the presence of endogeneity.

Notes: Probability of rejecting the null hypothesis of evenly distributed rates of criminals living in neighborhoods in Medellín. Potential hot regions of neighborhoods of criminals per typology.

Proyecto​​ 2

A semi-parametric specification of the incomplete EASI demand system: Welfare implications of a tax

Resumen: We provide a novel inferential framework to estimate the exact affine Stone index (EASI) model. Our inferential framework is based on a non-parametric specification of the stochastic errors using Dirichlet processes mixtures that allows handling non-normal errors, gaining efficiency, identifying clusters, and taking into account, microeconomic restrictions, censoring, simultaneous endogeneity, and non-linearity. We perform a welfare analysis due to a tax on electricity consumption using a novel data set in the​​​e Colombian economy. We find that 95% of the households belong to one cluster and that there is a 95% probability that the equivalent variation of the representative household is between USD cents 34.1 and USD cents 34.3, given an approximately 1% tariff increase (USD cents 0.12), that is, the welfare loss is approximately 300 times the actual tariff increase.

Figure. Price and income elasticities: Utilities in Colombia. 

Panel A: Hicksian price share semi-elasticities. Panel B: Marshallian price share semi-elasticities.
Panel C: Marshallian price quantity elasticities. Panel D: Income elasticities.​

Notes: Circles are posterior mean values, and bars are 95% credible intervals. Notation wupj (qupj) indicates the effect of percentage change in price of good j on share (quantity) for good u.​