23
grain procurement and the quality of village self-rule. At first, as the policy burden of
grain levy goes up, local officials have increasing bargaining power with higher levels
concerning their reluctance in the implementation of the Organic Law. This supports
our second hypothesis that the performance of grain procurement is negatively related
to the quality of elections. However, the story is not over here. As the policy burden
becomes heavier and heavier, the cost of enforcement will rise up rapidly in that the
resistance of villagers gets harder and harder. Now local officials have to consider
another policy task they must fulfill: social order. As we’ve discussed in the first
section, social order is the first priority of the central government and it has been
granted “veto power” in the cadre assessment system. This is a policy beyond
bargaining. That is, as rural resistance more and more threatens social stability, the
bargaining power of township cadres falls down dramatically. So they have to make
concessions to respect villagers’ rights, for instance, to improve village governance
through competitive elections.
For control variables, the year 1998 is significantly positive in most models. This
is consistent with our expectation that after the Organic Law has been formally passed,
the quality of village elections have been improved systematically. The percentage of
migrant labors is found to be negatively associated with village governance, which
shores up Oi and Rozelle’s observation that migrants generally are not interested in
political participation in their home community. Another significant control variable is
the percentage of agricultural income that has constantly positive influence on village
self-government. This also provides support for Oi and Rozelle’s finding that in
Convention