Revisiting Public Income Replacing Private Transfers: A Regression Discontinuity Design with Specification Errors

Abstract: I estimate the extent that a provision of a public pension scheme displaces, or ‘crowds out’, remittance that the elderly recipients originally collected from family members or other private sources. I explore the introduction of the Farmer’s Pension Program (FPP) of Taiwan, using an estimation strategy based on a regression-discontinuity (RD) design with specification errors. The results suggest that a dollar of the FPP benefit has displaced around 31 to 42 cents of private transfers. Despite the moderate level of pension benefit and the significant crowding out, the pension appears to increase recipients’ medical expense and utilization of medical service.