29
induce structural change in income-generating activities if participants do not have adequate education
for off-farm work. These findings add yet another piece of empirical evidence suggesting that China will
have to expand its investment in education to achieve its goals.
In the 2005 survey, we asked participating household members what they would do if the
government stopped payments after five to eight years (Figure 5). More than 20 percent of the
respondents wanted to find work off the farm. If they are unable to do so without a certain level of
education, they are at risk for being trapped in poverty when program compensation ceases.
VIII. Conclusion
In our study, we consistently find that, on average, the Grain for Green program has a positive
(although only moderately strong) effect on off-farm labor participation. In other words, households that
participate in the program are increasingly shifting their labor endowment from on-farm work to the
off-farm labor market. This shift occurs not only in absolute terms but is statistically significant when
compared to similar shifts in nonparticipating households. In terms of program intensity, we find that
program impacts increase as the ratio of a household’s retired plots to total land holdings grows. These
findings are different from those of previous studies that evaluated Grain for Green. The results also
indicate that households with less liquid assets are more affected (positively) by the program. This result
supports the view that the compensation paid by Grain for Green for setting aside cultivated land may
be relaxing the liquidity constraint for participating households, allowing participants to more readily
move into the off-farm employment sector (relative to nonparticipants).