Comment on: ‘‘Welfare implications of the transition to high household debt’’ by Campbell and Hercowitz

The purpose of Campbell and Hercowitz (2009) is to evaluate quantitatively the macroeconomic effects of deregulation of the mortgage market in the early 1980s. This deregulation made it easier for consumers to use houses as collateral for loans. The paper focusses on two questions. First, can this deregulation help to explain the increase in real interest rates in the early 1980s? Second, and more importantly, what effect did this deregulation have on household welfare? Campbell and Hercowitz do a good job of motivating their quantitative analysis, citing facts about changes in household debt and mortgage refinancing in the U.S. during the past quarter-century. The percentage of mortgagees who had ever refinanced their homes, for example, increased from 10% in 1983 to 44% in 2001, and housing equity as a fraction of house value (for newly purchased homes with a loan-to-value ratio of at least 50%) decreased from 23% of 1983 to 16% in 2001. Moreover, household debt increased from 43% of U.S. GDP in 1982 to 62% in 2000. Finally, throughout this period, homes and cars collateralized roughly 90% of household debt. These are big numbers that underscore the need for a coherent quantitative analysis of the macroeconomic effects of deregulation in the mortgage market. Campbell and Hercowitz do an excellent job of providing precisely such an analysis. The main findings of the paper are twofold. First, the sudden loosening of collateral (or, more generally, borrowing) constraints increases interest rates and lowers wages during the transition to a new steady state. Second, in a reasonably calibrated quantitative model, deregulation hurts borrowers: although they benefit directly from looser borrowing constraints, the indirect (general equilibrium) effect of deregulation on prices more than offsets these benefits, leading to a net welfare loss. Lenders, on the other hand, gain. I will use the bulk of the rest of my discussion to illustrate these effects in a drastically simplified version of Campbell and Hercowitz’s model. By stripping away extraneous details, this simplified version of their model suggests that the forces at work in Campbell and Hercowitz’s model are robust to changes in the exact details of how housing, housing debt, and mortgage financing are modelled. Moreover, even in this simple model there are opposing effects on household welfare, emphasizing again the need for a quantitative analysis such as Campbell and Hercowitz’s in order to measure their net effect.