significant determinant of the demand for incoming labor, especially for incoming
migrants, as indicated by its coefficient which is significant at the 10 percent level for the
OLS specification (column 11).viii
Restrictions in village labor markets and systematic preferences for hiring own
villagers, appear to become less binding over the period. The coefficient on the private
enterprise dummy is positive and significant for all incoming labor and for incoming
migrants in 1988 (row 3, columns 1, 2, 9, 10). In 1988, villages with a private enterprise
were more likely to have incoming labor and tended to have higher concentrations of
incoming labor, especially migrants, than villages with only collective enterprises. The
coefficient on the size of the village labor force is significantly negative in the 1988
estimates for incoming labor, implying that villages with large labor forces were less
likely to hire incoming labor from nearby villages (row 5, columns 1-4). By 1995, rural
industries had changed their behavior and were hiring workers more according to
economic criteria. The coefficient on the private enterprise dummy is no longer
significant in the estimates for 1995 (row 3, columns 3, 4, 7, 8, 11, 12). Because villages
with private enterprises do not show significantly different hiring practices from villages
with only collective enterprises in 1995, these results indicate that collective enterprises
may have abandoned their restrictive hiring practices and were behaving more like
private enterprises. The size of the village labor force also was no longer negatively
correlated with the proportion of incoming labor (row 5, columns 3 and 4), which
provides further evidence that incoming labor was not restricted to villages with an
insufficiently small local labor force in 1995.
23