For decades economists have recognized (and often puzzled over the fact) that homeowners appear reluctant to tap home equity to finance retirement. Reverse mortgages are rare; with only about one million funded since 1989. In fact, Venti and Wise (1989; 2004)show that senior homeowners do not reduce home equity even as these homeowners age into their eighties and nineties. Yet, in the mid-2000s, things started changing, with a peak origination of more than 110,000 reverse mortgages in a single year.

Nonetheless, subsequent problems emerged. Sales plummeted. Almost 10 percent of borrowers face potential foreclosure due to failure to pay property taxes and insurance (aso-called T&I default). So rather than allow seniors to age in place, many seniors are facing the loss of their home at an especially vulnerable time. However, it is also possible that the existence of a reverse mortgage prevented borrowers from facing even more severe financial outcomes. That is, even more seniors might have lost their home without financial resources provided by a reverse mortgage.

We explore these issues having been granted unique access to origination and servicing records by HUD for 861,537 borrowers who have taken out a HECM reverse mortgage from 1989 to 2012. We merge the servicing data with deeds and assessment records on 90 million properties obtained from DataQuick, identifying all senior homeowners in at least 100 counties based on whether the owner applied for a senior property tax exemption.  Finally, we may access credit records on a subset of borrowers (and a control group of seniors without a reverse mortgage) from Equifax to measure other sources of debt that might have been paid off by a reverse mortgage such as forward mortgages, credit cards, and auto loans.

Our paper progresses in two steps.  First, we examine outcomes for borrowers with a reverse mortgages to similar senior households who have not taken out a reverse mortgages. We use proximity to a reverse mortgage lender, as well as time variation in the entry and exit of such lenders from the reverse mortgage market, tocreate instruments to identify exogenous variation in which borrowers take out a reverse mortgage. We then measure a variety of financial outcomes for senior homeowners including prepayment or default on a forward mortgage, credit cards, or other debt, a property tax delinquency, or a foreclosure, or property sale.

Next we examine potential design flaws in the product that might reduce demand by seniors. From a public policy perspective, HUD/FHA has committed to redesigning the product by considering financial assessment of senior borrowers to reduce the high rate of T&I default, with the target of designing a new required financial assessment tool for all potential reverse mortgage borrowers by October, 2013. Our study will help determine factors that might impact such an assessment.