Wednesday, December 30, 2009

Illness and Foreclosure (h/t Vanessa Perry)

Christopher T. Robertson , Richard Egelhof, and Michael Hoke have a paper:

In recent years, there has been national alarm about the rising rate of home foreclosures, which now strike one in every 92 households in America and which contribute to even broader macroeconomic effects. The "standard account" of home foreclosure attributes this spike to loose lending practices, irresponsible borrowers, a flat real estate market, and rising interest rates. Based on our study of homeowners going through foreclosures in four states, we find that the standard account fails to represent the facts and thus makes a poor guide for policy. In contrast, we find that half of all foreclosures have medical causes, and we estimate that medical crises put 1.5 million Americans in jeopardy of losing their homes last year.

Half of all respondents (49%) indicated that their foreclosure was caused in part by a medical problem, including illness or injuries (32%), unmanageable medical bills (23%), lost work due to a medical problem (27%), or caring for sick family members (14%). We also examined objective indicia of medical disruptions in the previous two years, including those respondents paying more than $2,000 of medical bills out of pocket (37%), those losing two or more weeks of work because of injury or illness (30%), those currently disabled and unable to work (8%), and those who used their home equity to pay medical bills (13%). Altogether, seven in ten respondents (69%) reported at least one of these factors.

If these findings can be replicated in more comprehensive studies, they will suggest critical policy reforms. We lay out one approach, focusing on an insurance-model, which would help homeowners bridge temporary gaps caused by medical crises. We also present a legal proposal for staying foreclosure proceedings during verifiable medical crises, as a way to protect homeowners and to minimize the negative externalities of foreclosure.

Did California Overbuild its Housing Stock?

It all depends on the relevant time frame. If we look at the 00's in isolation, California overbuilt.

But California has grew by 7 million people between 1990 and 2008 and added about 2.43 million housing units (all data are from US Census-- assume that 98 percent of units permitted are actually built). The average household in California has 2.9 people (which is the second highest in the country, and compares with 2.5 nationally), which means that even without removals from the stock and no change in household size, the state needed 2.41 million new housing units. So if we look at the 18 year horizon, California did not overbuild--there we almost surely more than 20,000 demolitions over an 18 year period.

What if we go back to 1980? California grew by 13 million people and added 4.4 million housing units. At 2.91 people per unit, California demanded 4.46 million units. Again, it is safe to assume 60,000 demolitions over 28 years, so it is hard to make a case for long-run overbuilding.

Certainly, some housing was built in the wrong places, or was the wrong type of housing for the place (Lancaster and Beaumont come to mind). But it is hard to make a case that in aggregate California now has too many housing units.

Monday, December 28, 2009

Memorial for Arthur Goldberger at AEA meetings.

Sunday, January 3, 2010: 10:15 am, Atlanta Marriott Marquis, Atrium Ballroom A Remembering Arthur S. Goldberger Presiding: James Heckman (University of Chicago) Speakers:
Lawrence Klein (University of Pennsylvania)
Glen Cain (University of Wisconsin)
Kate Antonovics (University of California-San Diego)
Gary Chamberlain (Harvard University)
Charles Manski (Northwestern University)
with remarks from Harry Kelejian (University of Maryland) read by James Heckman

Thursday, December 24, 2009

It's A Wonderful Life Fan Trailer

News item on Fannie/Freddie CEO pay

Bob Hagerty in the Wall Street Journal says they will make $6 million each. This is absurd--they are now government agencies. It would not be unreasonable to cap GSE CEO pay at the President's pay (by that I mean the US President). I would even let them have a rent-free house. I could even see giving them a bunch of shares, so if by some miracle the GSEs became going concerns again, the CEO would benefit. But the current arrangement is bizarre.

Wednesday, December 23, 2009

Paul Krugman should know better

Civility is important to me. I am all for fighting the other guys' arguments tooth and nail, but try not to engage in name calling, because name calling devolves into the sort of ugly spectacle we just saw in the US Senate, where a member (physician!) from Oklahoma encouraged prayer for the illness or death of a colleague.

I was thus disappointed when Paul Krugman, whose work I admire, encouraged the left to hang Joe Lieberman in effigy, and then issued a non-apology-apology is his column the other day. I am not crazy about Lieberman--he seems to have the Tartuffe-like combination of sanctimoniousness and naked ambition (ok, now I am name calling a bit)--but using violent imagery as part of a policy debate is not acceptable.

Tuesday, December 22, 2009

The real problem with John Taylor's paper... that it asks the wrong question. The issue is not what would happen to rates without the Fed backstop to Fannie/Freddie, but whether there would be fixed rate mortgages at all.

30-year-fixed rated-80-percent-loan-to-value mortgages beyond the conforming loan limit basically do not exist at the moment [update: by conforming loan limit, I mean the high cost area loan limit, which in some places is now $729,750]. A conversation I had today with two bank executives confirmed this. Some would argue that 30-year fixed rate mortgages are over-rated, but not I.

Monday, December 21, 2009

John Taylor says that the Fed could sell its MBS without having a material impact on interest rates

His blog post is here. I will download the paper when I get into the office tomorrow ( I can only download NBER papers while on campus).

But one thing from the blog post strikes me as strange: he uses spreads on agency debt as his measure of credit risk. But the very fact that the Fed has purchased MBS could produce a perception that the government is standing behind the debt--as the Fed exits, so too might this perception. I would also imagine that the low current interest rates mean expectations about prepayment are unusual at the moment, and that the most common methods for pricing the mortgage call option might not be appropriate either.

The point is that it is very difficult to separate total mortgage interest rates into the term-adjusted risk-free interest rate, the credit spread and the prepayment spread. More after I actually read the paper.

[update: I just read the paper. Taylor describes it well in his blog post, which means that I have not seen anything to change my mind that it is not very convincing. More to come soon].

Friday, December 18, 2009

Paul Carrillo, Stephanie Cellini and I have a new paper

Surfing for Scores:
School Quality, Housing Prices, and the Changing Cost of Information

The abstract:

n this paper we investigate the relationship between school quality and information disclosure in
housing markets. When presented with the option of identifying their local public school in a
real estate listing, we find that sellers with homes assigned to higher-performing schools are
more likely to provide this information, an effect that is driven by sellers of large single-family
units. Further, we find that controlling for school quality, information disclosure has no
independent effect on housing prices for single-family homes, implying that buyers with a high
willingness-to-pay for school quality will seek out information on school quality on their own.
On the other hand, we find that sellers’ disclosures about schools have a large positive impact on
the sale price of small multi-family residential units in 2001-02, but the effect disappears by
2006-07. Presumably, the increasing ubiquity of the Internet and the availability of new data
under No Child Left Behind dramatically reduced the cost of gathering information on school
quality over this period. Taken together the results reveal substantial heterogeneity in buyers’
willingness-to-pay for information on school quality, they support the findings of the education
literature on the importance of school quality in housing markets, and they confirm the
“unraveling” theory of information disclosure found in other markets.

Will post to SSRN soon.

Thursday, December 17, 2009

Very Sad News from Wisconsin--RIP Arthur Goldberger.

We lost one of our most influential econometricians and teachers of econometrics this week--Arthur Goldberger. The econ department's obit is here:

A week doesn't pass without me being grateful for my econometrics education at Wisconsin--in fact, there are times even now when it just starts to dawn on me what Charles Manski and Gary Chamberlain were trying to teach me. Professor Goldberger's (I never had the nerve to call him Art) class sticks with me whenever I even glance at data. His clarity of presentation allowed us to internalize thoroughly classical and "neoclassical" regression theory.

When I go to conferences on policy analysis, I am always struck by (and proud of) how many of the presenters are Wisconsin Ph.D.'s, and by how much of the work that they present is good. Arthur Goldberger deserves a large part of the credit for this. Professor Goldberger didn't only teach a terrific and influential class (his textbook, A Course in Econometrics, is essentially a set of his lecture notes). He infused the economics department at Wisconsin, along with its students, with an ethos that refused to tolerate sloppy empirical work. He shall be missed very much.

[Austin Kelly adds: "I never met the gentlemen, but I can't tell you how many times I have referred people to his classic pages on "micronumercy" - his slam at people who blame everything on multicollinearity."

He also taught me not to place too much faith in R-squared.]

Wednesday, December 16, 2009

How pro-transit people shoot themselves in the foot.

I read the following on the Seattle-based Alaska Viaduct project in the Infrastructurist:

McGinn’s statements suggest that this megaproject may still be defeated. For the sake of a city with exciting light rail plans and big-city potential, we hope this happens, since the tunnel is too expensive and will only advance the interests of car commuters, rather than encourage them to try transit.

Let me stipulate that the project may very well not pass a cost-benefit test. But the line "will only advance the interests of car commuters" reflects both snobbishness and detachment from reality. According to this blog, more than four-fifths of commuter trips and 85 percent of all trips in Seattle are made in private automobiles. Complaining that something advances the interest of auto commuters is like complaining about advancing the interest of, say, children--pretty much every one of us is one, or loves someone who is.

As the Onion so wisely headlined, "98 percent of US commuters favor public transportation for others."

Tuesday, December 15, 2009

Gregg Easterbrook has a good piece on college sports this week

It begins:

Charlie Weis and Bobby Bowden had to go -- Notre Dame and Florida State weren't winning every game! Get rid of the bums! All we heard from sports commentators, and from alums and boosters, was get rid of the bums, we gotta win, win, win. Sorry to interject, but why? Why does Notre Dame or Florida State or any university need to win every game? Is it now official that big colleges care more about sports than education?

If an NFL team, which is strictly a commercial enterprise in the business of providing entertainment, doesn't win, get rid of the bums. But a university exists to educate; winning football games is a secondary concern. Don't get me wrong. I attend way too many college football games, and I always like it when the school I'm rooting for wins. But I am not so misguided as to think that a college's winning games means more than a college's educating students, including athletes. Why is this distinction practically absent from sports commentary?....

I was harsh with him last week, so when I think his work is good...

Monday, December 14, 2009

Inquiry about Hyderabad

I will be making my second visit to the Indian School of Business this coming January. If anyone out there has expertise in the Hyderabad real estate market, I would very much appreciate hearing about it.

Sunday, December 13, 2009

Sameulson stands alone

From the time I was in high school, I was told of a symmetry in economics: that the world had two leading economists, one on the right (Milton Friedman)and one on the left (Paul Samuelson). But once I started reading their works in college and graduate school, I came to the view that there was no symmetry: that Friedman was a brilliant economist, but that Samuelson was at a different level altogether. Friedman was more like Gershwin, whereas Samuelson was more like Mozart.

Friedman's two masterpieces were A Monetary History of the United States 1867-1960 (which he coauthored with Anna Schwartz) and Essays in Positive Economics. The former is an astonishing work of scholarship,and rigorously uncovers data to make an important point about the quantity theory of money. The latter reads like a series of clever puzzles to be solved--it challenges you to figure out the underlying assumption that drives the result.

But Samuelson's Foundation of Economic Analysis was like Don Giovanni and the Magic Flute--it changed everything. I do not think it is an exaggeration to say that all economic thought since then has its (if you will excuse me for saying so) foundations in that work. At the same time, he was, like Mozart, astonishingly prolific, and had influence in pretty much every field of economics (macro, micro, public economics, trade, industrial organization, etc.), and at pretty much every level (as a scholar of original research, textbook author, popular press writer).

We will miss Paul Samuelson

Mark Thoma picks out a couple of nice articles.

How lending decisions produce sub-optimal social outcomes

The New York Times reports that many people who would like to refinance their mortgages into lower interest rate 30 year fixed rate mortgages are unable to do so.

Of course, were households able to refinance their mortgages, their probability of default would fall, because the present value of their mortgage balance would fall (effectively lowering the loan-to-value ratio) and payment-to-income ratios would also fall. At the same time, because the cost of capital for lenders is low, financial institutions would find the refinanced loans profitable.

But when a lender holds an underwater loan, it wants to earn as much profit as possible, and so hasn't sufficient incentive to lower the interest rate. The judgment of these lenders is that the profit gained by continuing to charge a high interest rate is greater than the potential losses from the increased probability of default. At the same time, other lenders do not want to take out a loan that is underwater. Hence, people are stuck.

This is surely socially less than optimal--keeping foreclosures as low as possible is in everyone's best interest at the moment. So here is a policy proposition--if a borrower has always been current on repaying their mortgage, they get to refinance at the current low rate of interest. Financial institutions are being subsidized with unnaturally low interest rates. Borrowers should get their share of those subsidies.

Thursday, December 10, 2009

The USC Casden forecast is on line

It is at

We handed it out today on flash drives instead of hard copy reports. The audience seemed to like this. Kudos to Lusk staffer Jennifer Frappier for coming up with the idea.

Wednesday, December 09, 2009

When Gregg Easterbrook writes about something I know about, he almost always gets his facts wrong

In his football column (which I like), he writes:

A generation ago -- a decade ago! -- home buyers were expected to have a 20 percent down payment; that made them unlikely to try to buy something they could not afford, and banks wouldn't be exposed if something went wrong, since they were lending only 80 percent of the value of the property. Now requiring 3.5 percent down is viewed as "toughening" standards. Isn't this an invitation for yet another cycle of mortgage problems?

Low downpayments for FHA loans have been around for awhile. When President Eisenhower signed the 1957 housing bill, Time Magazine reported:

The Federal Housing Administration, aiming to attract money for homebuilding, increased maximum interest rates on FHA-backed mortgages from 5% to 5¼%. And to woo more buyers from middle—and even low —income groups, it slashed down-payment requirements from 5% on the first $9,000 to 3% on the first $10,000 of a mortgage.

Yes, one could buy a house for $10K in 1957. A brand new 3-bedroom in Mansfield Ohio cost $15k at the time, so used homes would have cost less.

Buyers have also been able to use private mortgage insurance to buy houses with five percent down-payments for many years (at least since the 1980s--if any one has history on PMI down-payments before that, I would love to hear about it). VA loans have always had very low downpayment requirements.

Do I worry about FHA? Sure. But let's not pretend that we lived in some virtuous world before this decade in which everyone put 20 percent down on their house. It was just not true.

USC Student Jonathan Shum sends me to a Green Street Report on Capital Structure in the REIT Sector

Its summary:

A Cultural Affinity for Leverage: Until about twenty years ago, the structural make-up of the real estate industry (i.e. small players, fragmented ownership, no outside equity sources, etc.) dictated that debt, not equity, serve as the primary source of external capital. As a result, market participants grew accustomed to operating with far more leverage than is found in virtually any other industry. Now that financing options for major real estate companies are very similar to what
is available to other large corporations, there are several reasons to believe that less debt and more equity should be the norm going forward.

Financial Theory: There is no reason why a non-taxable entity (e.g. a REIT) should have any debt, yet the costs associated with credit crunches (both in the form of distress and missed opportunities) provide ample reason to limit leverage to relatively low levels. These costs have proven to be so high that optimal leverage targets for most REITs likely fall in the 0-30% (debt/
asset value) range.

Best Practices in Corporate America: Most corporations have a strong incentive to utilize debt –interest expense helps minimize their corporate tax bill – yet it represents less than one-quarter of the typical corporation’s capital structure. REITs, by contrast, have no reason to use debt, yet it typically comprises about half the capital structure. There is no justifiable reason why financing
practices should differ as much as they do.

The Real World Lab Experiment: Higher leverage should be accompanied by higher returns in order to compensate for its added risk. This has not been the case, however, in the REIT sector, as more levered REITs failed to provide meaningfully better returns even in the ten-year period preceding the peak of the asset valuation bubble. Lower levered REITs have substantially outperformed over the last fifteen years.

De-leveraging & Value Creation Ahead: The REIT sector has commenced what is likely to be a multi-year de-leveraging process. It should unfold in three stages: 1) de-lever to ensure survival; 2) de-lever to return to prior leverage targets; and 3) acknowledge that prior leverage targets were too high and de-lever to achieve new, lower targets. Much progress has been made on the first phase, yet most companies are not yet entirely out of the woods. The subsequent stages will entail
massive amounts of equity issuance, as leverage ratios need to decline by more than 1500 bps to return to prior norms, and a substantial reduction beyond prior targets is appropriate. At a time when other real estate market participants lack access to capital, REITs that aggressively de-lever as they articulate thoughtful strategic objectives with regard to their long-term capital structures will be well-rewarded.

Monday, December 07, 2009

For Universities, Leaders Matter

Steven Sample is retiring as president of USC this year. He is one of the reasons I wanted to join the USC faculty. I am guessing that no university president has done more over the past three decades or so to transform his institution than Sample; the only comparable academic leaders I can think of are Theodore Hesburgh at Notre Dame and John Sexton at NYU.

One indicator of Sample's influence is USC's ranking based on Avery, Glickman, Hoxby and Metrick's measure of student preferences. Their measure is simple: after controlling for such things as cost and legacy, they look at the propensity of students to choose one college over another. Based on this measure, USC is the 29th most favored college in the country. Every college ahead of USC based on this measure is an excellent institution (the places immediately above USC in this ranking are Chicago and Johns Hopkins). For those that get into both Berkeley and USC, 72 percent go to Berkeley, but for those that get into both USC and UCLA, 89 percent go to USC (once again, controlling for costs). When Sample took over as President, this would have been unimaginable.

Sample also writes gracefully about higher education.

Sunday, December 06, 2009

Ann Schnare and I write about the Rise and Fall of F&F

The paper is here and here.

Has anyone done a paper on network effects and restaurant ordering?

Some years ago, I was awaiting a flight from LA to Washington (I think). It was winter, and so the flight was delayed, so I went to the bar to have a drink. I ordered a Ketel One Dirty Martini; after I did so, the man two stools down from me said, "that sounds good," and ordered one, and then a women across the bar also said, "that sounds good," and ordered one, and so on, until 5 or six people at the bar had Ketel One Dirty Martinis in front of them.

I thought about this while in a restaurant last night, when one order of gumbo seemed to lead to a cascade of gumbo orders around the room. So I would like to know whether these were isolated incidents, or whether one order generally influences the surrounding orders. In the case of married couples (or significant others) who like to share, one order might have a negative influence on the probability of a similar order. When the boss is buying lunch, perhaps her order places a ceiling on the price of other orders.

Anyway, it seems like there is a paper in this, but until I finish the half-dozen papers that are currently between 10 percent and 80 percent done, I am not working on it.

Wednesday, December 02, 2009

My AEA Economists Calendar has Arrived!

While I need to double check this, it would appear that Locke is the oldest economist listed; Saez is the youngest.