Nowcasting Poverty in Central America, Panama, and the Dominican Republic: A Micro-simulation Approach

Assessing the development of countries relies on poverty rates as a key indicator. However, official poverty rates are derived from household surveys that often have limited frequency, unexpected gaps due to field work constraints, and substantial delays in processing and publication. This paper presents a novel micro-simulation method for estimating poverty, which introduces changes in demographic and labor variables into the surveys, that can be derived from just a few observed or forecasted macroeconomic indicators. We apply this method for the case of Central America, Panama, and the Dominican Republic (CAPDR) and show that it outperforms the fit of other methods that solely rely on direct imputations from GDP to households income. Lastly, our approach can be easily replicated across countries and in different time periods, which is not the case for the majority of other micro-simulation techniques.