Poverty seriously hinders the inclusive development of mankind and is closely related to economic growth, ecological protection, ecological restoration and sustainable use of resources. Based on the data of economic census and rural fixed observation point, a spatial econometric model is established to test the direct impact and spatial spillover effect of industrial clusters on rural poverty alleviation. The result of household-level is that the number of industrial clusters has a negative effect on poverty, namely the farmers who live in the county with more industrial clusters, may be less likely to become the poor. The number of industrial clusters in other regions also has a negative effect on poverty. By dividing farmers into the poverty and non-poverty group, the study finds that, for the poverty group, the number of industrial clusters has a positive direct and spillover effect on farmers' income. For the non-poverty group, the number of local industrial clusters has a positive direct effect on farmers' income, but the number of industrial clusters in other regions does not have any effects or has a negative direct effect on farmers' income. By classifying the industries, the study discovers that the labor-intensive industrial clusters, such as textiles, manufacture and processing of machinery parts and paper industries, have a positive effect on farmers' income.
Keywords: anti-poverty; farmer’s income; inclusive development; industrial clusters.