In this paper we aim to disentangle how sectoral economic growth affects the size of the middle class, using state-level data of Bolivia from 2000 to 2017 and breaking the three main economic activities  into  subsectors  to  attain  more-specific  results.   Because  the  data  from  Bolivia  are limited, we utilize a Bayesian hierarchical longitudinal model for small samples.  We find that the  commerce  and  services  sectors  have  the  biggest  impact  on  the  size  of  the  middle  class in Bolivia and that mining and agriculture have a similar though smaller effect.  Our results also suggest that both formality and public social investment do have significant effects, albeit smaller ones. 

ACCESS PAPER