We analyze the correlation between farm productivity and market participation
using comparable household data from Tanzania, Vietnam and Guatemala. Each
farm’s input use and output levels provide a within-sample measure of relative
productivity, which we relate to that household’s level of participation in local
markets using a wide range of agricultural, demographic and infrastructural
variables as controls and as instruments in two-stage regressions. Results indicate
that, controlling for differences in market access and the underlying determinants
of market participation, households with higher productivity have greater
participation in agricultural markets. In contrast, households with greater rates of
market participation do not consistently demonstrate higher levels of relative
productivity. This result holds only in Vietnam and Guatemala, however. In the
Tanzania sample we find no significant correlation in either direction. Combining
household surveys in this way offers a promising approach to testing the
robustness of key hypotheses across countries and over time.