Nonlinear Effect of Household Debt on Consumption: Evidence from Household-level Panel Data in Korea

This paper attempts to answer the question of whether there exist thresholds for two household debt ratios, debt-to-disposable income and debt service ratios, beyond which consumption can be hampered by an additional increase in the ratios, using household-level panel data from 2000 to 2014 in Korea. The results of the panel estimation indicate that households begin to reduce their consumption spending when the debt-to-disposable income ratio exceeds 210% of their disposable income or when the debt service ratio exceeds 27% of their disposable income. In addition, the nonlinear effects become less significant in the estimations for consumption growth. Finally, the estimation results for consumption level according to sub-income groups indicate that the nonlinear effects of debt service ratio are more significant for higher income groups while the estimated threshold for the ratio is smaller for lower income households.