The Home Purchase Effects of GSE Redistribution: Evidence from Difference-in-Discontinuity Estimates

Abstract

The guarantee fee (g-fee) pricing of government-sponsored enterprises (GSEs) plays an important role in redistributing credit. We study the distributional effects of g-fee pricing on home purchases across borrower credit score and income distributions. Using a difference-in-discontinuities design based on recent g-fee adjustments, we find that, within low credit score groups, elasticities are larger for higher-income borrowers, suggesting that GSE redistribution benefits them more. Within high credit score groups, lower-income borrowers reduce their home purchases in response to a rise in g-fees, whereas higher income borrowers avoid the impact of g-fees by accessing non-agency loans. We develop a structural model, fit to the reduced-form elasticities, to analyze counterfactual g-fee policies. We find that the GSEs' redistribution toward lower credit score borrowers is regressive in the income dimension, in line with our reduced-form findings. Furthermore, we find that recent g-fee adjustments did not increase home buying among lower income borrowers. These results highlight the central role of heterogeneity in household elasticities when evaluating GSE reform efforts.

Keywords: Government Sponsored Enterprises (GSEs), Securitization, Redistribution, US Mortgage Market, Difference-in-Discontinuity JEL classification: G18

JEL Classification: G18, G21, G51