ARSC November 23rd-24th 2019 Applied Regional Science Conference

Nov 22, 2019 12:00 AM

Felipe Santos and Carlos Mendez

prepared for the 2019 ARSC at Saga Univeristy


This paper studies convergence and spatial autocorrelation of homicide rates in municipalities and departments in Colombia. Indicators of sigma convergence are found only at the state level. In contrast, at both levels significant beta coefficients were reported with different magnitudes. A distributional analysis approach also presents different patterns; at the department level four convergence clubs are found, while the conformation of at least two convergence clubs is present at the lower level. Lastly, robust spatial autocorrelation is found only at the municipality level. These results state the importance of the disaggregation level when studying convergence and the influence of neighboring spatial units. The findings of the distribution and spatial autocorrelation are milestones for further research and indicate the possibility of applying spatial filtering techniques and other Spatial econometric tools in order to study patterns of social development and diffusion.

link to the slides

Research Associate - Doctoral Candidate

Research Associate and Doctoral Candidate at TU Dresden, Dev Econ Master and Physics BSc, my research interests include spatial econometrics, machine learning, remote sensing, convergence and economic growth.