From Fannie Mae and Freddie Mac to the mortgage-interest tax deduction, the US federal government plays an enormous role in the country's housing market. But what effect is it having—and is there cause for it to be involved at all? Chicago Booth's Robert H. Topel and Eric Zwick discuss how government intervention affects prices, homeownership, home size, and more.
Did the housing stimulus program work?
Zwick: Together with David Berger at Northwestern and Nick Turner at the US Treasury, I looked into the First-Time Home Buyer Tax Credit, a temporary tax credit of $8,000 for first-time home buyers in the wake of the Great Recession to try and induce them to buy homes and shore up the housing market in 2009–10, when inventories in the housing market were at all-time highs.
We find that it was quite effective in inducing a large amount of demand. We estimated the program increased aggregate home sales by 7–14 percent during the policy period. There’s a lot of evidence that the program facilitated beneficial reallocation of underutilized assets from low-value owners, such as banks or builders, to constrained higher-value buyers previously unable to buy because of disruptions in the credit market or down-payment constraints. This demand didn’t immediately reverse. It was concentrated in the existing-home market, which implies that the direct GDP effects were quite modest and limited to realtor and origination fees, and, perhaps, complementary furniture purchases.
The market stabilized, and house-price growth stopped falling as rapidly, at least temporarily. When evaluated as a stabilizer, the program seems to be more effective than when thought of through the traditional demand-management-fiscal-policy view.
Topel: When the layperson hears you say “stimulus,” they’ll think of activities by the government that increase economic activity and increase total output. Eric finds that the program didn’t really increase economic activity in the housing market in the sense of creating a bigger housing stock. It didn’t have an effect that people would traditionally think of as a stimulus. What it did, on the best interpretation, is reallocate housing from some folks who were holding housing to other people who might have had higher value of the housing. So it’s totally reallocative. The stimulus impact in he traditional sense was really de minimis.
Zwick: That’s specific to a point in time with an unused inventory of houses in the millions. There’s a stock of extra homes that we’re trying to work down. It’s a classic investment boom, and the question is whether it makes sense to try and speed the process of using those assets. The stimulus effects were much smaller than the cost of the program. There’s a lot of evidence that the problems in the housing market would have been more severe without the program.
Topel: The program had different impacts in different markets, because there’s more potential for first-time buyers in different markets. The effects were to reallocate the ownership of existing housing from one group to another. Whether the failure of that happening [in the absence of the program] is a market failure is another question.
Zwick: Bob’s right to ask, what is the market failure this program is addressing? I’m trained as a public economist, and we always think about the role of the government as either correcting market failures or addressing concerns about redistribution. The redistributive case for this program is weak, because home buyers tend to be relatively wealthy. You have to evaluate it as addressing some market failure.
A lot of research suggests that, during the recession, foreclosure externalities were important—the effects on consumption, the effects on neighborhoods through crime and elevated vacancies. In other markets these kinds of spillovers are much less. So take things like unused battleships. I’m not sure that a first-time battleship-buyer program would be as good an idea. As to whether in peacetime we should have mortgage-interest subsidies, I agree with Bob that a lot of intensive margin subsidies in the housing market just cause people to buy larger homes, more square footage, more expensive places. I don’t really see the public-policy rationale there.
Topel: The existence of a foreclosure externality is fairly convincing. Whether it’s worthwhile having a government intervention in housing markets to deal with that is
If there is a market failure, can government help in reallocating resources?
Zwick: Foreclosure externalities should be solvable in the private market, but that doesn’t happen, and that’s a market failure that this program was specifically addressing. It’s not necessarily sufficient. You also need to have buyers. If there’s a development and no one actually wants to live there, moving people into those homes isn’t creating any welfare.
Topel: You need to make a case, even for that externality, that intervention to target that externality is a useful way of spending society’s resources. This foreclosure externality is true all the time. You drive through any neighborhood and the guy who doesn’t mow his lawn is creating an externality for the people next door. We have ways for the government to intervene in those cases, rules for how neighborhoods have to be kept up. An intervention in markets with an increase in government spending to internalize that particular externality is a leap of faith. It might be cheaper and more efficient to leave it alone, and recognize that externalities are everywhere: loud noise, unmowed lawns, and things like that. You can’t fix them all.
Do US government subsidies for the housing market—estimated at $70 billion–$150 billion annually—benefit the economy?
Topel: The evidence that the mortgage subsidy improves the operation of the housing market is meager. Would they need such a stimulus in Canada, where there is no mortgage subsidy? I’m not convinced we needed it here.
Zwick: We have very weak evidence of the social benefits. There’s an almost universal agreement among economists that this program should be modified, ranging from capped to replaced.
Topel: . . . to abolished.
Zwick: The question is, what would be the effect of such an abolition on house prices? It would be different in Houston than in New York. That would be an interesting area of research. Is there some transition cost for getting such a program implemented? The second question is, are there more-efficient ways—if we want to stimulate homeownership at all—to promote it? If downpayment constraints are the primary market failure, maybe we should think about a more cost-efficient program, potentially replacing [the mortgage subsidy] with a payment subsidy.
Topel: Economists are really good at saying that if there’s some subsidy that’s creating a large distortion—such as the mortgage subsidy, which pushes people toward consuming more housing services and less of other things—it would be a more efficient world if we didn’t have such distortions. Getting from A to B is another story, because getting rid of the mortgage tax deduction would affect housing prices and redistribute wealth from people who now own houses. The political support is unlikely to be there. It’s the same as the tax-preferred status of employer-provided health care. Americans overconsume health care because of that distortion, but getting rid of it is costly, because it’s so redistributive.
Who benefits and who is harmed by those subsidies?
Zwick: They seem to benefit owners relative to renters. They benefit high-priced areas relative to low-priced areas because of their structure. They benefit second-home buyers, because you’re allowed to deduct mortgage interest on two homes, up to about $1 million, in the amount of mortgage balance outstanding. Relative to other social programs, it’s actually quite regressive. Its value increases in your marginal tax rate, so the more income you make, the bigger the subsidy is. There’s a strong incentive for richer people in more expensive places to buy as much in housing services as they can, or to increase [spending] relative to what they would buy without a program.
Topel: Think of buying versus renting. It increases homeownership, and some politicians would say that a function of government is to do just that. Why we should have a larger proportion of people buying houses rather than renting homes escapes me.
Zwick: Or it changes the number of bedrooms you buy. Does a third bedroom really benefit society more than a second bedroom?
Fannie Mae and Freddie Mac—the government-sponsored agencies that buy mortgages and turn them into risk-free bonds—are supposed to help promote homeownership to low-income families. Does it work?
Zwick: There is an argument that support for the mortgage market reduces interest rates or stabilizes them across places in a way that makes some [rates] unnatural relative to the underlying risks. I think of that as less of a concern than the originate-to-distribute model. Freddie and Fannie buy mortgages for which they are not the lending officer. They’re not talking to the borrowers. It probably helps make sure that there’s some sort of financing available during bad times when credit markets are weaker, although those are exactly the times when affordability or the likelihood of repaying is low. We are trying to reevaluate as a society how we should intervene in the housing market. As house prices increase—by a lot in places where younger generations want to buy—we need to think about what role we want the government to have in the housing market, because there’s going to be increasing calls from relatively uninformed people for support and help in the housing market.
Topel: The premise of the question was that a larger fraction of the population should own houses rather than rent them. Is that a good public-policy goal? Is there some market failure that’s causing more people to be renting and fewer people to be owning houses than would occur in an efficient world? That’s got to be the rationale for some government intervention, but I don’t know that there is such a distortion. Some people should be renting, some people should be owning, depending on their preferences and their financial constraints.