This study uses detailed household level data to investigate income
and activity diversication among households in rural Tanzania. Unlike
previous research in the area I explicitly evaluate marginal returns
within dierent activities, aiming to assess whether households are
able to allocate labor so as to maximize their incomes. The ndings
indicate that specialization in agriculture does not seem to be correlated
with household welfare, and that agricultural wage work is a last
resort option, as those engaged in agricultural wage work also allocate
labor to their own farms to such an extent that marginal returns
become lower than among others. Furthermore, wage rates are much
higher than the marginal returns in agriculture, implying that there
may be large gains to be made from expanding the non-farm side of
the rural economy. There is also some evidence that both credit and
social networks are important determinants of a household's freedom
of activity choice.