The boom and subsequent decline of the U.S. housing market has many economists and the general public asking: What happened? Joseph Gyourko, a professor of real estate at the University of Pennsylvania's Wharton School of Business, has spent a lot of time examining that issue. In part, he says, the run-up was caused by irrational, speculative behavior by private lenders and borrowers. But there were other causes, too, such as land-use regulation that limited building in some cities and thus drove up prices. Gyourko argues that in areas where it is relatively easy to build, the correction will likely persist until prices are again driven by fundamentals — meaning, by production costs. In those areas where regulation has effectively capped the supply of housing, such as New York City, prices will be determined almost solely by demand, which has fallen recently as numerous Wall Street firms have encountered troubles.
Gyourko also has looked at the problems Fannie Mae and Freddie Mac have experienced. He argues that the implicit subsidy those companies received came with a costly catch — to provide risky loans to marginal applicants in the name of "affordable housing." While the provision of affordable housing may be a laudable policy goal, it should be done transparently, with the budgetary costs clear to everyone. Ultimately, Gyourko argues, Fannie and Freddie should be shrunk and privatized.
As an urban economist, Gyourko has studied what drives economic growth and vitality. In today's economic environment the most important factor is human capital. Cities with innovative, high-skilled work forces will continue to thrive while many of the giant industrial cities will face a steady, if slow, decline.
Gyourko joined the Wharton faculty in 1984, after earning his Ph.D. that same year at the University of Chicago. He has served as co-editor of Real Estate Economics, been a visiting scholar at the Philadelphia Fed and a nonresident senior fellow at the Brookings Institution, and since 2006 been a research associate with the National Bureau of Economic Research (NBER). He currently is director of Wharton's Zell/Lurie Real Estate Center and co-author with Edward Glaeser of the recently published Rethinking Federal Housing Policy: How to Make Housing Plentiful and Affordable. Aaron Steelman interviewed Gyourko at his office at Penn on Sept. 25, 2008.
Gyourko: I think there are benefits to homeownership. That said, expanding it the way we have done is clearly not worth enduring systemic risk. So, given how we did it and what we got, the answer is no.
This is how I think we should consider the issue: When you become an equity owner, you have a stake in your community. You have a stronger incentive to make the community better than if you were transient. That is the standard economic argument in favor of potentially subsidizing homeownership. But there is no evidence in the literature that relies on truly experimental or exogenous variation which shows there is any benefit. We all believe there is but it is very hard to find. You can identify correlations of being a homeowner and better outcomes for children, of being a homeowner and being more publicspirited in terms of becoming informed about public issues. But it’s hard to show causality. Yes, I think economists believe there are positive externalities to being a homeowner. But there is no way that those positive externalities justify the extent we have intervened in the market through the huge subsidies that have gone to Fannie Mae and Freddie Mac and created systemic risk.
Gyourko: It's a political sacred cow. But I don't think it has done much good and I am not in favor of it. Ed Glaeser has done some work that shows its effect on the homeownership rate is very small. Consider my own case. My decision to buy a house is not affected at all by the home mortgage interest deduction. I have two children who go to public schools in a nice area. And to go to those public schools, you have to own a home because there is not much rental stock. So that's a big reason why I am a homeowner, not because I can deduct the mortgage interest.
What the policy does affect, though, is the type of homes people own. I probably own a bigger home than I would otherwise. But it's not at all clear to me why you would want to subsidize middle- and upper-income households. I don't think it is affecting the homeownership rate much and it is a subsidy to the relatively better off.
Gyourko: This is an incidence issue. The tax expenditures are in the tens of billions of dollars. That's the legal incidence. Does that mean I receive the subsidy? It depends on where I live. If I live in a land-constrained market like San Francisco, the answer is no. I am probably just going to bid up the price of a home that I purchase and the original landowner gets the subsidy. But if I live in Houston, where you can build wherever you want, I may get to keep the subsidy. This is a classic question of legal versus economic incidence. There is a subsidy. Who actually reaps it depends on relative supply and demand elasticities, just like standard incidence theory tells us.
Gyourko: We are a little unclear on sharp estimates. But Ed Glaeser and I did some work in 2005 which shows that cheap housing in general is really attractive to the poor in declining areas. The argument goes as follows: If you are a low-skilled worker, your wage is pretty much the same in Detroit as it is in, say, Charlotte. However, prices in declining areas like Detroit can fall dramatically below construction costs but not in Charlotte. So you might actually be relatively better off in Detroit, even though your employment opportunities are less, because you can consume a lot of great housing at a cheap price.
More broadly, some work I recently did that is now out in the NBER Working Papers series shows that if you go negative in terms of your home equity, your two-year mobility rate falls by half. Why? Mainly, I think, because those people are capital constrained. Also, I think there is some loss aversion that is going on. But it's clear that there are large mobility effects if you have negative equity in your home. I think they are much smaller if you do not have negative equity.
We do know that there are big differences in mobility between renters and owners. But, again, that is not necessarily causal. I am an owner and I am much less mobile than my research assistant who is a renter. But that's because he is young and doesn't have kids, whereas I want to stay in the same place. I don't want to move my kids out of school. People who want to stay in one place do tend to own a home, but it's not necessarily the home that locks them in.
Gyourko: Actually, the nonprime market reached 50 percent in 2006. That includes subprime, Alt-A, home equity, and FHA/VA. The sum of those four reached 50 percent of mortgage volume issued in 2006. In a typical year, it would range somewhere between 10 percent and 20 percent. It skyrocketed. I think the 10 percent to 20 percent range is more normal.
Gyourko: There were a couple of factors. First, a long period of rising prices led both borrowers and lenders to believe that, with relatively little risk, you could have very high leverage and very low rates. That, combined with the very large fees that these loans generated, led to an increase in market share. Second, I believe the government strongly encouraged it. Fannie and Freddie provided important liquidity for this market.
Why did they do that? I doubt that it was because they thought it was in their shareholders' interests. They probably thought these were risky loans. They had Congress telling them that their implicit subsidy comes with this mandate to provide affordable housing, and this was one of the ways that they did it. Our political system very much encouraged them to provide liquidity and extend these risky loans to very marginal buyers.
Gyourko: I would like to see them shrunk, privatized, and spun off. That means the affordable housing component of their reason for being has to be taken over by somebody else, because I don't think we should do away with all affordable housing programs.
But I do think those programs ought to be brought onbudget and placed in the Department of Housing and Urban Development or a similar entity and made an explicit cost to the government. The way to get these programs managed properly is to have them truly transparent, and then we can decide which ones we think are worth it and which ones are not. I think the reason these programs got so big was that they were off-budget, and politicians could nudge Fannie and Freddie management because there was no cost in terms of actual budgetary expenditures, even though there was obviously a very large cost in terms of risk.
Gyourko: I don't think it is wrong per se, but I think the real driver of urban growth is broader. It is human capital. I will reinterpret the "creative class" to mean "bohemian." I don't think you have to be bohemian to have an urban area grow. What you need is a bunch of innovation. If those innovators are bohemian in spirit, alright, but I think there are a lot of innovators who are not. And what you need is high-skilled, innovative people to grow. So I would reinterpret Florida's work more broadly.
Gyourko: I think you are likely to see more of it down the road, particularly if energy costs get high or if we impose a carbon tax. In theory, well-designed new urbanist places could cut down on the need to drive. But in the absence of higher energy costs, I'm not so sure they will develop. People like open space. They like their own yards.
Your question noted the presence of subsidies. The question is whether new urbanist communities are going to need an ongoing subsidy. I think the only way they won't need it is if energy costs get very high.
Gyourko: Glaeser took a leave from Harvard and visited Penn for a year. We started talking and one of the things we noted was that the dispersion in house prices across markets is going way up over time. However, the high land price areas are relatively few and they are almost all on the coasts of the United States. The question is: Why is that the case? One answer would be, if you only believe in horizontal development, you have an ocean as a natural barrier to growth. But you can also build up, and New York was the classic case.
Another reason might be income. The reason prices are higher in New York, Boston, San Francisco, and Los Angeles is that incomes are higher. And then you get another boost for San Francisco and Los Angeles because the climates are great. People are willing to pay for that.
But supply is an important factor. It became clear that unless you have differences in supply, you could not explain why prices were very high in those markets. There has been very little new construction. And the reason why is that there have been restraints on growth. Developers would like to build but they can't because of regulation. We have now convinced ourselves, and I hope others, that supply-side restrictions play a big role in how those markets have changed and are a big reason why their prices are so high. The existing residents of those areas have been very successful at limiting growth.
Gyourko: We don't know. As economists, our first thought is that the people of New York and San Francisco are just protecting capital gains. But let's assume that the people of Richmond and Atlanta are not stupid, and obviously they are not. They could figure this out, too, and place restrictions on growth. But they haven't. Clearly there is something else going on. It could be social. It's a huge research question.
Gyourko: In the long run, that trend will continue. But in the short run, they are going to fall. New York's housing supply is inelastic so the prices are determined by demand. When Wall Street is booming, housing prices are going to rise. When it is having trouble, like now, housing prices are going to fall, I think substantially. But in the long run — let's say 10 years or more — prices are going to continue to rise. They are attractive places to live — they have a lot of amenities — and they have high human capital. So as long as people want to live there and they are fundamentally productive, prices will go up in the long run if you restrict supply.
Gyourko: Yes, it does. But the city builds like mad. Chicago has been a relatively free market. Whenever prices get above physical construction costs, land and land assembly costs, and entrepreneurial profit, you see a surge in new supply in Chicago. That tells you that the locality is not restricting supply very much.
Gyourko: In terms of growth, there's not much difference. Atlanta, Dallas, and Houston have all grown well above the national average in terms of population expansion. And if you are growing in population, it means you are building homes. The correlation between the change in population and the change in housing units is almost one. So they are all growing at fairly similar rates.
What differentiates Houston is that it is more of a hodgepodge, because it is unique in its relative lack of zoning policies. So you get one type of development right next to a completely different type. Personally, I don't like that. I think there actually are negative externalities to that type of growth.
While Glaeser and I have argued that the social costs of development congestion are not nearly as high as the price increases associated with excessive limitations, I am not a believer in no zoning and no regulation. There really are some social costs, and I think Houston probably goes too far. It doesn't internalize some of those costs. You should think about traffic flows. You should think about pollution spillovers. It is reasonable to try to internalize the costs of those things. It is quite legitimate for government to congregate certain types of activities in select areas. That's a far step from, say, having huge minimum lot size requirements, so that only the super rich could live in an area. In short, I think cities are better off with some regulation, much less than New York, but probably more than Houston.
Gyourko: It's been huge. It's hard to quantify precisely, but I have no doubt that it has been very important. People like to live in those climates but also want to be able to get out of the heat. So air conditioning and the relatively low cost of land have been two of the really important drivers in the growth of the Sun Belt.
Also, when you think about growth in various areas of the country, you have to remember that most producers now work in high competitive industries. And that means you cannot pay your workers high wages to compensate for high prices. Your average manufacturer needs to be in a place with low factor costs — and to have low factor costs, you have to have cheap housing.
Gyourko: I think it is highly unlikely that we will see a rebound in those cities' housing markets. Again, what drives modern growth in the modern era — ever since manufacturing deurbanized — is skills. You need high-skilled people, and there is no reason for those people to go to Buffalo. They can go to Charlotte, Atlanta, or Dallas, where the climates are better and there tend to be more amenities.
Now, is it possible that one of those slumping cities will have the next Bill Gates in it, just by serendipity, and that he sprouts a huge new industry there? Yes, it's possible. But it would require a lot of luck.
Gyourko: They are important. Those things help to explain why urban decline is so long and so slow. This wealth that they accumulated when they were great industrial towns is very durable. It has thrown off things that have lasted a long, long time. The Cleveland Clinic is still with us, it is still one of the great medical centers in the world, and that’s not going to go away. But the rest of the place will slowly decline because very little of our economy is tied to having cheap water access, which means that the business value of being on the Great Lakes has declined sharply.
Gyourko: The bulk of the evidence is that, in a purely fiscal sense, these things don't work. The numbers don't add up. So to justify them, you have to make different arguments. When Ed Rendell, the former mayor of Philadelphia and now the governor of Pennsylvania, argued in support of subsidies for stadiums for the Phillies and the Eagles he said that there are valuable social spillovers. There's a feeling of community that is generated by having nice stadiums where fans can go and have a good time. I think that is the only argument that is plausible.
I personally don't like these subsidies. Consider the Eagles. The National Football League operates a type of monopoly. There is not free entry. So when you subsidize stadiums, you are providing a subsidy to players and management. I see no reason why the median taxpayer in this town should subsidize Jeffrey Lurie, the owner, or multimillion dollar players. I didn't favor such subsidies. I don't favor them. And I have never seen any hard, good evidence, which is replicable, that shows they are positive in a fiscal sense.
Rendell made a noneconomic argument. I don't think it's a compelling argument. But it is a far stronger argument than claiming that there will be a net revenue gain by subsidizing stadium construction. One of the reasons that the economic argument is weak is that people have a relatively fixed budget for leisure. People who don't go to the football stadium and spend money there are likely to spend those funds elsewhere, say, at the movies. But they are not likely to spend money both at the game and at the movies. They are choosing between available options.
Gyourko: Yes, that has been a justification for some stadium subsidies. But when people say that they want to improve an area, I suspect that what they really want to do as a society is to help the people living in that area. A much cheaper and more efficient way to do that is to directly give them money. So when you think about the subsidy for the Washington Nationals' new ballpark in Anacostia, think about the money that was spent on that. Then take that same sum, divide it by the number of poor people living around the stadium, and that's the size of the check you could have given them. If you really want to help them, that's the way to do it.
Gyourko: I haven't seen economists address that question. But I think people understand quite well what the trade-offs are. And, increasingly, people who work in the downtown core do not live in the exurbs. Remember, the reason you have exurbs is that businesses follow people out there. So those people's commuting costs are often much lower than you might think.
I will give you an example: the Philadelphia metro area. The largest office node is not downtown Philadelphia. It's the King of Prussia office node northwest of the city. It has 10 million more square feet of office space than downtown Philadelphia. People live around there and have pretty short commutes. So employment has suburbanized and that's why commuting times are not as severe as a lot of people believe.
I think people understand how much they are going to have to drive. The real question is: Do we pay too little for gasoline? The answer is almost certainly yes, at least until recently. The true social cost of driving was higher than the price we were paying at the pump.
Gyourko: I think you should always just try to figure out the right tax. I don't think we know what that level is, but we should aim for a system that is transparent. Complex schemes often wind up obfuscating the real costs and lead to bad outcomes. It's like hiding the costs of affordable housing policies by making them part of the mandates of Fannie and Freddie. Instead, you make those costs clear to everyone so that they can be evaluated properly.
Gyourko: I am not a big fan of the popular ones because they don't do it through revealed choice. Quality of life should be measured by how much you are willing to pay to live in an area. Instead, a lot of these indices put places near the top of the list with cheap housing. As an economist, I know that housing must be expensive somewhere because it is really attractive or really productive. High prices signal high quality of life, not low prices. The classic urban spatial equilibrium models say that housing prices are the fees you pay to access the amenities and productivities of an area. I built some indices in the early 1990s based on that assumption. I think most of the business community gets this backward, because they like low prices, not unsurprisingly. From their point of view, Green Bay looks really attractive. Do you really think the quality of life is higher there? My argument would be no, because people are not willing to bid up the price.
Gyourko: That paper is forthcoming in the Quarterly Journal of Economics. There is a big political economy literature out there which says that at the federal and state levels, political parties matter. That is, partisanship is important because the policy outcomes are quite different if, say, the Republicans are in control rather than the Democrats. My colleague Fernando Ferreira and I thought that this probably isn't true at the local level. We suspected that the reason is that there is much more competition at the local government level. For instance, if I don't like policy at the federal level, what am I going to do? I could move to Canada. But that's very costly. It's also usually costly to move to another state. But if I don't like the policies of my hometown of Swarthmore, it's pretty cheap to go next door. So we thought that competition would restrain partisanship.
We spent a couple of years collecting data on mayoral elections: which party won and by how much. In this paper, we looked at close elections and compared local fiscal policy outcomes — how big was the government, what they spent their money on, things like that. The reason we relied on close elections is that it doesn't appear that the populace of those jurisdictions have extreme preferences. And so it becomes possible to test whether a politician is able to impose his views in favor of either smaller or bigger government. It turned out that it didn't matter whether a Republican or a Democrat won a close election. They do basically the same thing.
This, in my view, is basically a validation of Anthony Downs' median voter theory. At the local level — perhaps not at the federal or state level — what really matters are the preferences of the median voter. It's not the preferences of the politicians who get elected. The politicians are driven by the competitive forces of their environment to do what the median voter wants. Mobility is so high and there are so many competitive districts within a metropolitan area where people can move to, that the politicians must respond to the median voter's wishes. A politician might be ideological but he can't act on it.
To give you an example, in Swarthmore there are a bunch of single-family homes owned by relatively high-income people who are willing to tax themselves a lot to get good public schools. That's the median voter. Any politician who said he was going to cut taxes and not fund schools would get run out of town on a rail. He might believe that personally, but he would never declare it publicly.
Gyourko: Ed Glaeser, Albert Saiz, and I have this paper, "Housing Supply and Housing Bubbles," that just came out in the Journal of Urban Economics. Basically, it shows that in inelastically supplied housing markets, volatility really is higher. It's true not just on the blackboard, it also shows up in real-world data. Inelastically supplied housing markets have much bigger housing booms and busts, which is a bit foreboding right now. When fundamentals change, price could adjust or quantity could adjust. But quantity can't adjust in an inelastic market. All of the adjustment from the change in fundamentals is in prices. They can really boom in good times and really bust in bad times. So as we think about this housing debacle, one of the ways to at least lower volatility in the next downturn — and there will be another one — is to think about increasing the elasticity in these markets.
The other thing in that paper which I think is interesting really goes back to work by Sherwin Rosen and Jennifer Roback. In a free market, prices are pinned down by production costs. And those production costs are physical construction costs, land and land assembly costs, and entrepreneurial profit. In that paper, we compute each of those costs for each market. You can build an 1,800 square foot home in any market in the United States for less than $200,000 in today's dollars.
If you look prior to 2003, throughout the Sun Belt, in any unconstrained market where you think supply elasticity is high, actual prices never deviate by more than 10 percent. It appears to work really well. Then, after 2003, in Florida, Phoenix, Las Vegas, the Inland Empire of California, prices started to deviate considerably. That appears to be the beginning of the real mispricing of housing. Those markets start to look like they are inelastic in supply, but the number of permits never went down. As an old Chicago School guy, it's the closest I have ever come to saying we simply mispriced this asset. Theory tells us that prices should be pinned down by the sum of those three factors, and for 20 years they were. To me, this also suggests where prices are going back to in those markets.
Gyourko: Certainly, Sherwin Rosen at Chicago provided the foundation for the urban economics that I have done for a long time. Also at Chicago would be Arnold Harberger, who was chair of my thesis committee, and Dennis Carlton, who does a lot of law and economics now but had a strong interest in urban economics then. Since I left graduate school, Ed Glaeser has been a guy who has influenced me a lot both through his individual work and through our many collaborations. I have been at Wharton for 25 years, and here the two people would be Anita Summers and Peter Linneman. Anita and I have done a lot of work together on issues with public-policy relevance. Peter was the guy who got me interested in real estate issues.
Order single copies of or subscribe to Econ Focus (formerly Region Focus) and other publications from the Federal Reserve System.