Abstract: Since the implementation of the poverty-stricken county system in China in 1986, whether and to what extent the establishment of national poverty-stricken counties can raise the real income level of local farmers requires an systematic and scientific empirical test. Based on propensity score matching-difference in difference (PSM-DID) methods, using the panel data of 296 counties in six provinces of China from 2007 to 2016, it is found that the establishment of national poverty-stricken counties failed to effectively improve the real income level of local farmers.On the contrary, it has a sustained negative effect. Further research has found that the development of secondary and tertiary industries in poverty-stricken areas can effectively raise the real income level of local farmers, but the policy of poverty-stricken counties has not optimized the local industrial structure; the government can effectively increase farmers' real income by intervening in the price of agricultural products. From the policy implication of the research conclusion, it is of great practical significance to promote precise poverty alleviation based on poor households, promote effective investment through attracting external investment, and optimize county industrial structure for enhancing farmers' income in poverty-stricken counties.

Key words: national poverty-stricken county, different-in-different, propensity score matching, farmers' real income