Climate Sensitivity and the Value of Agricultural Production in the Brazilian Northeast: An Approach Using Spatial Panel Data

This work analyzes the sensitivity of agricultural production in relation to changes in precipitation and temperature levels in the Northeast region of Brazil. For that purpose, data from 2006 to 2016 were used for 952 municipalities in the area. The econometric methodology derived from Kunwar and Bohara (2017) and is applied to the production value of municipalities in the Brazilian Northeast, where it is assumed that the production value is also impacted by the interactions between precipitation and temperature. Thus, time and space data are used, applied to a spatial econometric methodology. The results suggest that the agricultural production of the municipalities of the Brazilian Northeast is spatially autocorrelated. There is indication that there are municipalities with high (or low) levels of production, which have neighbors with these same characteristics. It was also verified that, from 2006 to 2011, the agricultural production was more sensitive to changes in temperature levels than to changes in average precipitation. However, after the years 2012, 2013, 2015 and 2016, agricultural production has become more sensitive to changes in the precipitation levels. It was also noted that in the analyzed period there was an increase in the average sensitivity of the agricultural production in relation to the precipitation levels, while the average temperature sensitivity showed a decrease.