Binyamin Appelbaum has an article in the New York Times about the administration’s terrible response to the housing crisis. The administration “tried to finesse the cleanup of the housing crash, rejecting unpopular proposals for a broad bailout of homeowners facing foreclosure in favor of a limited aid program — and a bet that a recovering economy would take care of the rest.” This has several responses, including David Dayen at firedoglake, as well as Ezra Klein writing about the administration’s response from a balance-sheet recession and housing point of view. That got a response from Dean Baker arguing that this balance-sheet recession point of view, and the subsequent focus on mortgage debt reduction, is a distraction from better policy.
With President Obama pushing for a wider refinancing plan and the debate over refinancing and principal reduction back in the headlines due to the book Bailout and the fight over the GSEs, it might be useful to formalize the best liberal case against principal reduction. It’ll give us a set of arguments to wrestle with so that we can then work backwards toward better arguments. So what is the best case? I see three broad arguments.
1. Wealth Effect Means It Doesn’t Matter
This is the approach Dean Baker takes, and I think it is influential among many liberal wonks. The housing crashed destroyed a lot of housing value, leaving us feeling poorer, which means we spend less. An important way to understand this argument is that if every house during the housing bubble was paid for with cash instead of a mortgage, and we had the same housing bubble and crash but no mortgage debt overhang, our recession and slow recovery would look virtually identical. Reducing housing debt in our situation won’t help the economy as a whole (though it will help the individuals involved), because housing debt hanging on the economy isn’t the drag.
Foreclosures are still bad in this argument (and Dean has been at the forefront of fighting against foreclosures), but they only need to be stopped in the sense that all bad things should be stopped; housing crisis policy will help some and hurt some, but it isn’t a check on the recovery. It is not necessary and isn’t effective in getting us back to full employment.
I think there are some empirical problems with this argument. The elasticities people are finding are an order of magnitude bigger than realistic expectations. Declines in housing prices are nonlinear against wealth distribution. Something else is in play. See this interview or this paper for more on these arguments. The administration seems to be moving in this balance-sheet direction. Let’s say we reject this wealth effect argument — should we change policy?
2. Fiscal and Monetary Uber Alles
Christina Romer would say no. She, like many, would argue that housing debt is probably a drag on demand, but we should respond to it with fiscal and monetary stimulus. She would stay out of the policy in the purple circle above, which is the mapping I use around here to approach how people think of the recession. Romer, from September 2011:
[One argument is that the] bubble and bust in house prices has left households burdened with too much debt. Until we deal with this problem — perhaps by providing principal relief to the 11 million households whose mortgages are larger than the current value of their homes — we’ll never get the economy going.
The premise of this argument is probably true: recent evidence suggests that high debt is holding back consumer demand. But it doesn’t follow that the government needs to directly lower debt burdens to stimulate job growth.
Recent research shows that government spending on infrastructure or other investments raises demand even in an economy beset by over-indebted consumers. Another effective approach is to aim tax cuts and government payments at households that would like to spend, but can’t borrow because of their debt loads (such as the poor and the unemployed).
History actually suggests that the “tackle housing first” crowd may have the direction of causation backwards. In the recovery from the Great Depression, economic growth, which raised incomes and asset prices, played a big role in lowering debt burdens. I strongly suspect that fiscal stimulus will be more cost effective at speeding deleveraging and recovery than government-paid policies aimed directly at reducing debt.
There’s a general critique of the president’s stimulus program that argues it was too focused on tax cuts instead of long-term investments, which have a better bang for the buck. The same critique can be used on spending money on principal reduction. It’s money that by definition isn’t spent (it was already spent), so you need second-order effects for it to go. We’d prefer just giving people money (tax cuts) over principal reduction in the same sense that we’d prefer infrastructure over tax cuts.
And one doesn’t need to be a conservative worried about helping the “losers” or someone who is uncomfortable with the fairness of mortgage debt reduction to think there are better ways to spend this money. Consider having $250 billion dollars to spend, one benchmark put forward as the amount of money that could have been spent from TARP. You could hand it out in some manner to pay off underwater debts, perhaps a matching scheme with the banks. That wouldn’t reduce overall mortgage debt that much because there is a lot of it.
Meanwhile, with $250 billion dollars, you could build 5,000 miles of high-speed rail. You could fund universal pre-K for a decade. You could take the 13 million people unemployed under the traditional unemployment measure and give them a basic income of almost $10,000 for two years. You could build infrastructure, create social goods designed to foster egalitarianism, or tackle poverty. These are all better investments for us to make, plus they build a better society and they get us to full employment faster. Tackling mortgage debt produces none of these benefits.
When Geithner’s argued against principal reduction, saying that it would be “dramatically more expensive for the American taxpayer, harder to justify, [and] create much greater risk of unfairness,” he followed it up by saying “The whole foreclosure crisis across the country now is really driven by what happened to unemployment and what happened to the income of Americans. The best things we can do now to help mitigate that risk is to help get the economy. growing again, bring unemployment down as quickly as we can, put people back to work.” I view that as in line with Romer’s argument.
By itself, I think this is correct. But one important response to that is that principal reduction can often pay for itself, especially in situations where a borrower is at risk. A lender will want a consistent, if lower, payment stream rather than to take ownership of an abandoned house in a depressed market. As Lew Ranieri said, “You are almost always better off restructuring a loan in a crisis with a borrower than going to a foreclosure.” So it is good economics, especially in a distressed market. Another response is that few people propose just giving money away, but instead want to tie it to some sense of risk and reward, or reaccounting of the banks’ balance sheets. So how does that play out?
3. Upsides and Downsides
One reason giving away money to pay off underwater debts is a bailout, and thus politically unpopular, is that there would be a disconnect between who absorbed the costs on the downside and who gains the potential value from the upside. If taxpayers just paid off mortgage debts, banks and homeowners would gain a windfall that isn’t directly shared with taxpayers. One way to deal with this is either to force creditors to eat a cost upfront — they absorb the downside and then can benefit from the upside. The other is for taxpayers to gain from the upside, usually through the mass purchase and/or refinancing of mortgages. Let’s look at the first way.
Why aren’t bank servicers doing writedowns? There’s a mix of bad incentives and poor resources that result in bad practices. The administration hasn’t been aggressive with using financial fraud, like the range of practices including robosigning and documentation fraud, to force reform here, instead focusing on removing legal liabilities from the banks. Maybe that task force will someday do something, but from my read even sympathetic observers think it was a wasted opportunity.
But even if policy is centered on forcing servicers to clean up their fraud, there’s a lot of creditor free-riding in ad hoc debt writedowns that becomes problematic. Is writing down first mortgages good policy even if junior mortgages, often held by the biggest banks, are untouched? If home equity lines of credit are acting as a last line of income maintenance and credit for households in this weak recovery, is it wise to push policy to extinguish them to adjust first mortgages? If you wipe out both, isn’t that a giant transfer to other creditors like auto lenders, private student loans, and credit card companies? Should we be concerned about moral hazard from the debtor’s side? You need some mechanism to coordinate and bind the collective behavior of creditors while preventing free riding and also bringing in impartial adjudication, which is a traditional function of bankruptcy. Bankruptcy reform was famously not pushed by the administration, and to me that was its biggest mistake.
The other approach to avoid a bailout is for the government to gain a share of the upside for taking on the downside. This is one reason writedowns for the GSEs make sense: we gain the upside, as we own the GSEs, and we’re already on the hook for the downside, so the risk on the downside isn’t a “bailout” but prudent policy.
When it comes to dealing with the broader housing market, a lot of the programs proposed, like revitalizing HOLC or Senator Merkley’s plan on refinancing, would have taxpayers put up money but gain in the upside. Even the IMF is now encouraging the United States and other countries to investigate bringing back something like an HOLC. The two counter-arguments would be that HOLC still had a high redefault rate, a rate that would have a lot of people crying foul. The second is the problem of what to pay for the mortgages. Recent attempts to use eminent domain to purchase mortgages at below-market rate in order to compensate taxpayers for absorbing these risks in a terrible market also have a lot of people crying foul.
My general thought is that moral hazard can be a problem, but the misery and wasted lives of mass unemployment is a much bigger problem. That said, bankruptcy and these government programs eliminate most moral hazard concerns. Bankruptcy can be done in such a way to hit homeowners as well; for the government program you’d want people to be trying to take advantage of them. That’s why so many people have been shocked that the administration hasn’t pushed on either.
What I find interesting is that all these articles about what could have been done with housing take the way TARP played out as given. But starting a HOLC program, rebooting the broken servicing model, or otherwise writing down mortgage principal would have been significantly easier if the banks were put into a receivership in early 2009. TARP policy, which was to protect the banks’ balance sheets at all costs, worked counter-productively, putting administration resistence to enacting even the lowest-hanging policy fruit. Receivership would have cost more upfront, but it would have been significantly easier to tackle these problems. There is a major debate to have on this topic.