Endnotes
i The sample villages were selected randomly on the basis of a stratified random sampling procedure. The villages all
come from nine representative provinces (Zhejiang, Shandong, Hubei, Sichuan, Yunnan, Shaanxi, Hebei/Liaoning, and
Guangdong) which were randomly selected from each of China’s traditional geographic regions (East China--huadong,
North China--huabei, Central China--huazhong, Sichuan, Southwest China--xinan, Northwest China--xibei, Northeast
China--dongbei, and South China--huanan). Eight counties were selected from each province, two from each quartile of
a list of counties arranged in descending order of gross value of industrial output (GVIO). GVIO was used on the basis
of the conclusions of Rozelle, 1994 and Rozelle, 1996 that GVIO is one of the best predictors of standard of living and
development potential and is often more reliable than net rural per capita income. Two townships, one above the
median GVIO and one below were randomly selected from each county. Two villages in each township were selected
in the same manner. Data collection in Guangdong was so expensive that the study was never started. Due to the
exclusion of Guangdong, areas with high levels of off-farm employment may be under-represented in the sample.
ii The survey also attempted to estimate the number of permanent out-migrants. For the purposes of this study
permanent out-migrants are those who leave the village for employment purposes and have no intent to re-establish
residence in the village. Leaving the village permanently was such a rare event that the survey tabulated the total
number of workers leaving the village in the periods between 1989- 95 and found that the total number of permanent
out-migrants amounted to less than one percent of the labor force. For the remainder of the paper, migrants refer only to
long-term, not permanent, labor migrants. Due to the survey’s focus on emerging markets, we explicitly did not
consider those who moved for marriage, education, or retirement.
iii Leaders were able to provide information on the average daily wage for most of the categories in which workers
typically are paid wages. For the self employed, village leaders estimated the average daily earnings, which
incorporates the returns to labor as well as other fixed factors.
iv Although all villages in the survey are “rural,” some are highly industrialized and in several cases have very little
cultivated land.
v These estimates come from the percentage of rural labor going into villages (estimated by the survey) multiplied by
China’s total rural labor force as published by the State Statistical Bureau (SSB). The SSB reports 403 million rural
workers in 1988 and 446 million in 1995.
vi While only 27 percent of rural-to-rural commuters worked in light industry in 1995, this is a higher percentage than in
any other sector.
43