The long-term impact of land certification on factor reallocation and household welfare in rural China

Well-defined and secure land tenure rights are critical for promoting factor reallocation and improving household welfare. Yet, available studies on land tenure security-enhancing programs generally do not consider time lags in their implementation and provide mixed evidence on the impact on rural household welfare. This paper estimates the long-term effects of China's 2009–2018 land certification program on factor reallocation and household welfare, using survey data collected in 2019 from three Chinese provinces. We exploit village-level variation in the number of years since program completion, addressing potential selection bias in program rollout. The results show that longer program exposure significantly increases households' participation in land rental markets, driven primarily by increased probability of land renting-in rather than renting-out. However, we find no statistically significant effect on cross-sectoral labor reallocation. Changes in agricultural investment patterns and operational farm sizes further support the observed evidence in land and labor reallocation. Regarding household welfare, while agricultural income rises with longer program exposure, total household income does not. Further analysis of income dispersion suggests that the program has a pro-poor distributional effect at the sample level but not within villages. These findings underscore the importance of understanding the discrepancies between policy design and implementation in evaluating program impact and provide novel evidence on the sustained effects of land tenure reforms. • The long-term effects of China's 2009–2018 land certification on factor reallocation and household welfare are estimated. • Survey data collected in 2019 shows a large variation in program timing across villages, even within the same county. • Longer program exposure boosts participation in land rentals primarily via land renting-in, but not in labor migration. • Household agricultural income rises with program exposure duration, but total income does not. • Longer program exposure has a pro-poor distributional effect between villages, but not within villages.