Tuesday, July 26, 2016

Thoughts from California about a year in Washington

The nice people I worked with at HUD asked me to write a short piece reflecting on my year there.

Five Things I have learned from a year in Government (The things I have written below represent no one’s views other than my own).

1. The federal government is staffed by some remarkable people.

I have been most fortunate throughout my career to work with intelligent, committed and

ethical people. My time at HUD has been no different. I have had the privilege of working

closely with remarkable people at HUD and at other agencies. I would name names, but worry

about leaving someone out. Many people who choose to work in government could receive far

more compensation doing something else, but are motivated by a desire to make people’s lives


Government service is honorable service—this was a commonplace as recently as when I was in

college, and remains so in places like France and Japan. We as a society should respect

excellent government work more than we do right now.

2. The federal government relies too much on obsolete technology.

We see throughout the United States examples of under-maintenance of infrastructure, from a

bridge falling into the Mississippi River to tracks in metro systems catching fire. Less visible, but

just as problematic, is an unwillingness to invest in modern technology systems. Within HUD,

for example, the FHA program relies on systems that are driven by coding in COBOL, a

mainframe (!) language developed in 1959 (!!). Because almost no one uses COBOL anymore,

our university computer science departments don’t train students in its use. As COBOL

programmers retire, it will become impossible to find people to maintain the system.

On a more personal level, I was stunned to learn that my HUD PC had a 32-bit operating system

in a world where 64-bit system have been around for PCs for 13 years. As a practical matter,

32 bit systems are limited in the amount of data they can analyze, whereas 64 bit systems are

nearly unlimited. Many doing HUD work rely on large data sets (for example the Public Use

Microsamples of the Census and the American Community Survey). The current standard for

operating systems makes it relatively easy to use these datasets; the old standard requires


3. Academics teaching policy issues should spend some time in government, if for no other

reason than to appreciate the importance of details.

Before I joined HUD, I thought I was an expert on mortgage backed securities (I even wrote a

book about them). Spending time with the good people of Ginnie Mae revealed to me that I

really wasn’t. I did understand how to evaluate cash flows from MBS, but I didn’t really

understand how Ginnie Mae operated at all. What I learned is that mortgage default risk is not

the only risk that needs to be managed; issuer risk needs to be managed as well (while

FHA/VA/Rural housing insures mortgages, Ginnie Mae insures the issuers of mortgage backed

securities that fund mortgages). Suppose a mortgage goes into default and so a Ginnie Mae

issuer needs to buy it out of a pool. That issuers needs to have enough cash on hand to survive

until it receives an insurance payment from FHA/VA/Rural Housing. For non-bank lenders, this

could be a problem in times of low liquidity.

We academics are good at thinking about analytics; we are not so good at thinking about

operations. Yet without good operations, analytics lose much of their value.

4. Regulators care too much about details

Regulations about disclosures make the point. Government sometimes worries too much

about the details of disclosures and not enough about their effectiveness.

Consider disclosures for the price of a long-term fixed rate mortgage. For consumers to be well

informed about what they are getting themselves into, they need to know two numbers: total

upfront cash payment, and the all-in interest rate (which might include a mortgage insurance

payment). Armed with these two numbers, consumers can comparison shop in a

straightforward manner.

The first page of the new TRID closing form does this well, and lenders should absolutely be

held responsible for presenting this page accurately. But the details on the following pages are

essentially irrelevant to consumers and, by increasing the length of the form five-fold, make it

more complicated and confusing than necessary.

5. Few people know who the third most powerful person is in the Federal Government.

My guess is that the name Shaun Donovan is not well known outside the Beltway. But pretty

much nothing gets done without the approval of the OMB director.

Wednesday, February 17, 2016

Housing now

From HUD's The Edge.

This past fall, the state of housing reached something approaching normalcy in some dimensions (new construction and price) but continued to worsen in others (rental affordability and the homeownership gap between underrepresented minorities and others).

When President Obama took office in January 2009, residential construction in the United States was at its lowest level since World War II; only 490,000 units were built that month (on a seasonally adjusted annualized basis). By November 2015, that total had risen to nearly 1.2 million units — an increase of 139 percent over the course of the administration. This total has still not, however, reached the average level of new construction over the past 55 years of 1.4 million units.

The below normal (if substantially improving) levels of construction explain why housing prices have recovered substantially from their troughs. Prices in all 20 Case-Shiller cities are well above their troughs, in part because the paucity of construction has led to falling vacancy rates nearly everywhere. In two cities, Dallas and Denver, prices are at all-time highs, and Portland, San Francisco, and Boston have recovered all of their losses. A particularly noteworthy fact is that prices have recovered while the homeownership rate has declined. The price story is an absence of supply story.

Although the demand for owner housing has been stagnant, the demand for rental housing has soared, pushing up rents even in the face of strong multifamily construction. Rental demand has risen sharply for several reasons.

First the marriage rate in the United States has been falling steadily. According to Pew, 65 percent of the “greatest generation” were married by the age of 35; among millennials, the marriage rate is only 26 percent. After taking into account age, education, race, ethnicity, and geography, married couples are 22 percentage points more likely to be owners than singles. If millennials continue to postpone (or avoid) marriage, the ownership rate will continue to fall.

Second, racial and ethnic minorities, again after taking into account the standard list of demographic and economic characteristics, have lower ownership rates than non-Hispanic whites. The population of African Americans, Asians, and Hispanics is growing much faster than the population of non-Hispanic whites. African Americans, for instance, have a homeownership rate that is 17 percentage points lower after controls than it is for non-Hispanic whites. If homeownership rates among the groups whose population is growing fastest continue to lag, the pressure on the rental market will become even greater.

The reasons for lagging ownership among minorities are doubtless varied and complex, but part of the gap almost certainly results from continued discrimination in housing markets and issues with access to credit. Turner and Yinger have demonstrated the continued existence of discrimination, but we will say a few words about access to credit here.

One group of Americans, now very large, does not have access to mortgage credit at the moment: those whose homes went into foreclosure during the global financial crisis. RealtyTrac puts the number of homeowners who were foreclosed upon at [nearly 7] million, or about 5.5 percent of U.S. households. These households overwhelmingly became rental households (some doubled up with other families or moved back to their parents’ homes), and this phenomenon alone put sudden pressure on rental markets.

They also became ineligible for mortgage debt for at least 3 years (the number of years the Federal Housing Administration requires to have followed foreclosure for borrowers to become eligible for a loan) to 7 years (the minimum number of years post-foreclosure government-sponsored enterprises require before issuing a loan). Many of these potential borrowers are about to become eligible again for mortgages, and should thus relieve pressure a bit from rental markets. But many, having been traumatized by the homeowning experience, might decide to remain renters. As it happens, minorities bore a disproportionate share of the foreclosure burden.

The other access to credit issue involves access to wealth and credit scoring. Many researchers have shown that children’s wealth is highly correlated with parents’ wealth. African Americans, who as a group were stripped of wealth and who were over generations systematically denied access to credit, have less wealth than non-Hispanic whites even after controlling for income and education. The absence of wealth among older generations means that it is more difficult for younger generations to accumulate downpayments and establish excellent credit scores. This puts generation after generation of minorities at a disadvantage when it comes to owning a home.

The combination of diminishing numbers of married couples, the fallout from the recession, and access to credit issues have pushed rental demand and therefore rents as well. While there are many methods for measuring rental affordability, perhaps the most telling is that in the vast majority of American metropolitan areas, median-income renter households must spend more than 30 percent of their gross income on the median rental unit. Economists like to talk about “choice,” suggesting that people “choose” to live in expensive housing. But both within and across our cities, affordable rental housing is not a choice that is available to the median-income renter.

Tuesday, October 13, 2015

A book that changed my life

"We believe that the confounding of the aggregate with the individual is as dangerous as it is pervasive...."  Page 81.

Sunday, August 16, 2015

Apartments, Energy and Bad Incentives

I am spending this academic year working in Washington working at HUD, so I am renting an apartment here.  When I signed the lease, I understood that I would pay utilities; what I didn't understand, because I did not read the lease carefully enough, is how I would be charged for utilities.

Even though I live in a professionally managed building that has something like 400 units, and even though each unit has its own circuit breaker the units are not metered individually.  Instead, utility costs are allocated on a pro rate basis based on unit square footage and number of residents in the unit.

Since moving into this place, I have been very conscientious about setting the air conditioner at 80 degrees F. before leaving the apartment for the day.  I will do my best to continue to do this, but the fact is, my principal motivation for doing so was thinking that I could reduce the very expensive cost of cooling an apartment in the hot DC summer.

Of course, as one of 400 units, my influence on electricity usage for the complex is small.   There is essentially no financial reason to avoid blasting the AC all day long.  It must be the case that how one consumes air conditioning has a large impact on how much one spends on air conditioning.  In fact, people who like Bikram Yoga could drive their air conditioning costs to nothing.

A Google search on the cost of individual metering implies that the cost of installing meters for electricity would be about $300-$500 per unit.  Summer electricity costs in my unit are about $100 per month, so if price incentives lead to a 10 percent saving, each unit could save about $60 per year on electricity (I am applying the savings to six months).  Beyond this, of course, energy use produces negative externalities, so the social benefit of metering would be greater than the private benefit. This implies the benefits of metering exceed the costs. [If I am completely wrong about either the cost of metering or the savings arising from it, I would be happy to hear about it].

So why don't landlords do this?  The answer might be that they can't get enough extra net rent from tenants to justify paying for metering.  The only private cost they bear is not being able to charge as much rent as they otherwise might.  It might be worth doing serious analysis to determine the social benefits of metering in the context of individual apartments, and whether such metering should consequently be subsidized.


Sunday, April 26, 2015

Is Free Trade Good for Everyone?

Greg Mankiw implies that it is, and that all economists agree that it is.  But it actually isn't.  Who says so?  Economists.

In particular, the workhorse theory of International Trade, the Hecksher-Ohlin Theorem, leads to the Stolper-Samuleson Theorem, which shows that when countries start trading with each other, the relatively abundant factor of production in each country becomes better off, while the relatively scarce factor becomes worse off.   In the US context, this implies that opening up trade will leave capital better off relative to labor, and skilled labor better off relative to unskilled labor.

Does trade increase the total size of economies?  Yes--this is something that economists do agree on. But in the absence of redistribution--something that seems to be anathema to we Americans--more open trade will make low skilled laborers worse off.

In my ideal world, we would pass the Trans-Pacific Partnership (TPP), a potential [quasi]-trade agreement among the US and 11 other countries of the Pacific Rim, and redistribute its bounty such that everyone would be better off.  There is no evidence that our political system would allow this to happen.

Despite all this, I do and will continue to support trade agreements such as the TPP because that there is some evidence that they prevent wars.  Of course, as someone who has had nothing but good fortune in life, it is easy for me to think that the abstract prevention of war is more important than the tangible reduction in other people's already low wages.  

Thursday, March 12, 2015

LA has zoned itself out of the ability to house its residents (h/t Matthew Glesne)

Once upon a time, the zoning in Los Angeles would have allowed for 10 million residents to live within its municipal boundaries.  Greg Morrow, in his UCLA dissertation, "Homeowner Revolution: Democracy, Land Use and the Los Angeles Slow Growth Movement 1965-1992," documents how this was eroded over time:

So LA really did create a moat around itself and pulled up the drawbridge.  For those of us who think the blessings of cities should be shared widely, this is a shame.

Thursday, February 26, 2015

It is hard to feel urban form sometimes.

I have spent a fair amount of time in Sao Paulo over the past 3-4 years, and always thought it sprawled more than LA, because it takes forever to get from one side of the place to the other.  So was I surprised when I went to Google Earth and looked at both of them from the same elevation.

Here is LA:

Now here is SP:

It is far more compact.  Metro LA has about 18 million people; SP has about 20 million. But it takes about 2 hours to get from Santa Clarita in the west to San Bernardino in the east--the distance between the two is 85 miles; it can take four hours to go just 30 kilometers in SP.  Sao Paulo feels much larger to me.