6. Conclusions
Using our measure of corruption from the China Corruption Database as the values of bribes received by government officials, we have found no consistent evidence for the fair-wage hypothesis, which requires comparison of the public-sector wage with an external wage reference. We find evidence that the relation between the absolute value of the public-sector wage and our measure of corruption as the value of individual government officials’ bribes is U-shaped. Increasing public-sector wages, ceteris paribus, reduces corruption when provincial public-sector wages and income levels are low. The corruption-deterrent effect of a higher public-sector wage diminishes and eventually comes to an end as provincial per capita income grows. Carrots and sticks can be used in fighting corruption. A bribe is attractive to underpaid civil servants, especially when they are struggling for subsistence. In this case, a few more carrots (increases in the public-sector wage) will work. When income grows to higher levels, sticks seem to be required as carrots alone do not satisfy insatiable people.