Abstract
Agricultural support policies, which increase production to excessive levels, have been heavily criticized for their distortional effects on farmers' decisions. High input consumption, which harms the environment, and the distribution of supports, which causes income inequality among farmers, are also the other factors that economists criticize these support policies. However, in recent years, the use and dependence on production and land have been increasing globally due to fears of food insecurity in several countries. In this article, the impact of a support policy on cotton farmers' land in Turkey is evaluated using a detailed farm-level micro dataset. We use the dynamic panel model to consider the production dynamics, unobservable individual heterogeneity, and the endogeneity of the support variable. Our results show that considering the support variable as endogenous significantly increased its impact on farmers' land. A farmer's past land allocation decisions also affected their current lands holdings. However, the results varied considerably between regions and between different land sizes. Therefore, the findings show the importance of heterogeneity in the impact of support policies. We conclude that due to the heterogeneous impact of support policies, it might not be possible to discuss the effect of support policies using an average estimate applicable for all farmers. Therefore, a different support policy considering regional characteristics or farm size is necessary to obtain the desired policy outcomes.