Based on the data of China’s household tracking survey, the relative poverty scale and government subsidies are used to measure the impact of relative poverty scale and government subsidies on restraining the continued relative poverty of families through the relative poverty measurement index and decomposition method, and the binary regression model is used to analyze family characteristics, factor allocation, risk protection, income. The relative poverty reduction effect of structure and further analyze the interaction of household urban-rural heterogeneity. The study found that the mismatch of household production factors will cause relative poverty to persist, but government guidance and the optimization of labor factor allocation can restrain the negative impact of factor mismatch; in the long run, the relative poverty reduction effect of household production factor allocation is dependent on the initial wealth of the family. low; the allocation of labor factors is relatively deprived by the heterogeneity of households in urban and rural areas. Based on this, policy suggestions are put forward, such as promoting the transformation of the government’s poverty governance capacity, making up for the shortcomings of the factor market, and accelerating the process of urban-rural integration.