This study provides empirical estimation results concerning the degree of production coupling for the fixed and variable direct payment program designed for stabilizing rice farmers’ income. Based on a set of producer’s expectation formulations, we analyzed the impacts of direct payments on the rice cultivation area, using a panel data estimation technique. To consider the risk preference of farmers, we used an expected utility maximization model under the expected utility hypothesis. Also, we explicitly included a variable capturing truncated variance of rice price related to variable direct payments. Our estimation results indicate that both fixed and variable direct payments tend to have production coupling effects. Specifically, we found that the production coupling effect of variable direct payments was higher than that of fixed direct payments. We also discovered that the rice cultivation area has been affected by these direct payments via the insurance effects (variable direct payments) and wealth effects (fixed direct payments).