It’s Not Too Late to Prepare for the Next Housing Crisis

This article was originally published in Barron’s with Mark Calabria.

The housing cycle lives. Many economic models predict a housing crisis at some point in 2023 or 2024. The combination of high consumer debt, a slowing labor market, and inflated housing prices has created a ticking time bomb. It will explode – we just don’t know exactly when. But it is never too late to prepare. Luckily, policy makers can look back on past recessions to help refine approaches to mitigating the damage

Yes, some of the worst mortgage products behind the 2008 crisis are history. Unfortunately, they have been replaced with other troubling practices, such as record debt burdens among borrowers, as measured by debt-to-income. Significant numbers of recent borrowers are also already underwater on their mortgage or at risk of becoming so. Appropriately provided loan relief during the pandemic has also inflated borrower credit scores, making it significantly harder for lenders to accurately judge risk.

The federal state governments tried several strategies to mitigate the 2008 crisis, including the Home Affordable Modification Program. Those approaches were complex, convoluted, and often ineffective. New approaches were developed in 2020, in response to Covid, that can better serve both borrowers and the taxpayer. The means-testing of assistance in the 2008 crisis inadvertently resulted in massive disincentives for work. Delinquent borrowers often lost 31 cents of their mortgage assistance for every additional dollar earned. That contributed to one of the slowest job recoveries and biggest declines in homeownership in history.

In contrast, pandemic forbearance was based on time, not income. Borrowers were given a bridge to their expected unemployment insurance benefits, which can take weeks, if not months, to arrive. 

Moving away from means-testing also greatly reduced the paperwork required of both borrowers and lenders. Pandemic mortgage assistance provided through Fannie Mae and Freddie Mac assisted almost 3 million borrowers, twice that assisted under HAMP. Perhaps more importantly, the pandemic programs were set up six times faster. While HAMP helped mitigate the severity of the financial crisis on consumers, it only reached a third of its targeted indebted households because it overlooked important design elements in mortgage finance.

Speed and ease of use in government programs can be, and often is, an invitation to fraud and abuse. This was first addressed by requiring that any missed mortgage payments be repaid. Payments were paused, not forgiven, reducing the incentive to game the system.

Assistance was also paid for within the mortgage market, not by the taxpayer. The HAMP program cost well over $20 billion, not even including a $25 billion settlement with lenders or the $10 billion hardest-hit-fund. In contrast, pandemic assistance provided through Fannie and Freddie was largely paid for with a modest fee on high-income refinancings. The support provided to the mortgage market was paid for in the mortgage market, as it should be. Since the assistance was provided in a budget neutral manner, it also did not add to the inflationary pressures created by other pandemic assistance programs.

Policy matters because it not only directly shapes incentives and economic activity, but also indirectly influences consumer and homeowner expectations. Newly-released research from one of us, using data from the Federal Housing Finance Agency over the past three years, shows that homeowners who are more optimistic about the housing market at the time of their loan origination are roughly 2 percentage points more likely to enter forbearance compared with those who anticipated the housing market will stay the same. That’s significant given that the average rate of forbearance among homeowners was 4% in 2020.

But it’s not necessarily a bad thing – these homeowners who entered forbearance were likely more determined to get back on track with their payments, rather than lose it all and eventually get foreclosed upon. In fact, researchers from the Urban Institute released a report titled “normalizing forbearance,” suggesting that there should be expanded provisions that allow homeowners to share the requisite documentation about extreme events to their servicer and delay payments. In this sense, policy that promotes optimism also indirectly influences the efficacy of loan renegotiation.

There were comparable results among homeowners who anticipate future unemployment and income declines. Crucially, all these results control for a wide array of demographic factors, as well as other individual characteristics, such as credit scores and attitudes about risk.

The tight link between borrower expectations about the housing market at the time of origination and the likelihood of entering forbearance points to an important result: Homeowner attitudes matter and shape the decisions that ultimately might precipitate situations that require federal intervention. That means policy makers should not think about intervention from a static perspective, but rather a dynamic one where the choices they make today have consequences for how consumers and firms behave in the future and, in turn, impact the economy as a whole.

Christos A. Makridis is a research affiliate at Columbia Business School and Stanford University, and the founder/CEO of Dainamic, a financial technology startup. He holds dual doctorates and masters in economics and management science & engineering from Stanford University.

Mark Calabria is the former Director of the Federal Housing Finance Agency, and author of the forthcoming book, Shelter from the Storm: How a COVID Mortgage Meltdown was Averted.