Consumption based estimates of urban Chinese growth

- We estimate income growth implied by food Engel curves in urban China in 1993-2005.
- Implied income growth is 1% per year higher than measured growth.
- We attribute this discrepancy to a small bias in the CPI.
- That bias is comparable to CPI bias estimates for the United States.
- CPI Bias was higher for poorer households.

This paper estimates the household income growth rates implied by food demand in a sample of urban Chinese households in 1993-2005. Our estimates, based on Engel curves for food consumption, indicate an average per capita income growth of 6.8% per year in 1993-2005. This figure is slightly larger than the 5.9% per year obtained by deflating nominal incomes by the CPI. We attribute this discrepancy to a small bias in the CPI, which is of a similar magnitude to the one often associated with the CPI in the United States. This result supports the view that Chinese price statistics are reliable. Our estimates indicate stronger gains among poorer households, suggesting that urban inflation up to 2005 in China was “pro-poor,” in the sense that the increase in the cost of living for poorer households was smaller than for the average one.