The blog has been upgraded to http://www.urbanecononline.com ! Keep it in your bookmarks. Thanks for reading.

Essentially moving to an individual domain name allows posting dynamic contents such as interactive maps and graphs. And believe it or not, the copyright on contents posted on wordpress.com belongs to wordpress.

I’ll thus be upgrading previous posts such as the great description of Baltimore’s vacant houses during last year’s riots; and will be including better geographic contents in following posts.

The Great War of Realtors against Online Platform Zillow

While general interest media may be focused on the war between taxi drivers and the Ubers and Lyfts of the world, a similar war against the disrupter Zillow is being waged by the National Association of Realtors. Inman news reports that the National Association of Realtors is claiming more than $1b in damages for violation of trade secrets! Zillow, the Uber of real estate, just hired a top-level executive from Move, the internet company that manages realtors.com, the online service of the National Association of Realtors. The NAR claims that such top-level executive will reveal trade secrets to its competitor Zillow.

While the news are behind Inman news’ paywall, some of the comments there are revealing. A substantial chunk of the readers are professional real estate brokers. Example: “NAR and all local boards need to curb these third party websites such as Redfin, Zillow, and Trulia.” Sounds like some brokers are not preparing to be ready for disruption, but are trying to prevent disruption. We know that those kinds of battles don’t necessarily end well for the incumbents.

A few questions on the real estate business suggest the customer benefits/should benefit from the disruption of Zillow and the likes:

  • commission fees are a percentage of the transaction price, around 5-6% (see this FTC report), but more expensive homes are unlikely to be that much more expensive to sell/buy for the broker.
  • commission fees don’t vary much across the country (see Figure 2 of Hsieh et al)… when more competitive locations should command lower commission fees.
  • stories of collusion in the real estate business abound. The National Association of Realtors has been particularly successful in litigations.

Will Zillow win this latest stage in the online business vs realtors war? We’ll see what the superior court of the State of Washington says…

Reading suggestion: Hsieh, Chang‐Tai, and Enrico Moretti. “Can Free Entry Be Inefficient? Fixed Commissions and Social Waste in the Real Estate Industry.” Journal of Political Economy 111.5 (2003): 1076-1122.


Quandl and Zillow

Quick news that Quandl, through its R interface, is giving access to county-level price data from Zillow. The interface is OK and it’s cheap, so it seems to beat most of the competition — including buying county-level Case Shiller price data from Standard and Poors.


There are pluses and minuses.


  • The R interface is simple and easy to use. It takes five minutes to get the historical time series of one county’s median price.
  • Zillow provides median rents.
  • Zillow provides list prices as well as sales prices — this used to be available in the expensive CoreLogic data.
  • The number of foreclosures is also in the data. This is also available in the expensive RealtyTrac data (compiled from county records); but RealtyTrac still provides more information such as Real Estate Owned properties, Lis Pendens, Notice of Trustee Sale, i.e. the full timeline of the foreclosure process.


  • There are more than 3,000 counties in the U.S. Getting the full U.S. data thus requires more than 3,000 function calls. Not really the simplest thing on earth, especially when such data stored in CSV would take a few seconds to load on a statistical package.
  • It goes back to 1996, which may be good enough for you, but is not enough for some other applications.
  • Micro data is a must, and county-level data points are not sufficient any more. In that sense Quandl overstates the amount of information they deliver to their customers. They call one time series data set. Not really the standard way of defining a data set. (where is the “set” in data “set”?)

So we’re still going to rely on either Census values at the block group level, but they are self-reported and top coded; or on other data sets such as Data Quick, or FNC. I use FNC data in my latest paper.



Documenting Urban Segregation

During my visit to Princeton 8 years ago I was struck by how central questions of race and ethnicity were in U.S. social sciences. Segregation remains one of the major legacies of U.S. history. Research has shown that integrated and diverse cities fare better in a number of dimensions — yet the evidence presented below suggests that cities remain heavily segregated.

Today I am releasing a set of new interactive maps for the top 20 most populated metropolitan statistical areas of the United States. The remaining 330 metro areas will come online soon. These maps allow a web user to click on the name of a metropolitan area to see the racial make-up of neighborhoods.

What makes the data unique is that it is comprehensive, interactive and over the next few months data will go back in time (1970-2010 are in the works).

Census data does have quirks — for instance population in Central Park is non-zero — but it’s the best available data on the topic. I am sure you will find other quirks than Central Park population.

I start today with maps for the fraction of African Americans by census tract.

The maps may take a few seconds to load.

Top 20 Most Populated Metropolitan Areas

New York-Northern New Jersey-Long Island, NY-NJ-PA – Population 18,502,340 – link

Los Angeles-Long Beach-Santa Ana, CA – Population 12,617,920 – link

Chicago-Joliet-Naperville, IL-IN-WI – Population 9,298,407 – link

Dallas-Fort Worth-Arlington, TX – Population 6,294,195 – link

Houston-Sugar Land-Baytown, TX – Population 5,868,844 – link

Philadelphia-Camden-Wilmington, PA-NJ-DE-MD – Population 5,797,517 – link

Miami-Fort Lauderdale-Pompano Beach, FL – Population 5,487,714 – link

Washington-Arlington-Alexandria, DC-VA-MD-WV – Population 5,479,895 – link

Atlanta-Sandy Springs-Marietta, GA – Population 5,184,490 – link

Boston-Cambridge-Quincy, MA-NH – Population 4,392,939 – link

Detroit-Warren-Livonia, MI – Population 4,247,606 – link

San Francisco-Oakland-Fremont, CA – Population 4,245,474 – link

Riverside-San Bernardino-Ontario, CA – Population 4,148,968 – link

Phoenix-Mesa-Glendale, AZ – Population 4,113,465 – link

Seattle-Tacoma-Bellevue, WA – Population 3,374,336 – link

Minneapolis-St. Paul-Bloomington, MN-WI – Population 3,215,778 – link

San Diego-Carlsbad-San Marcos, CA – Population 2,993,347 – link

St. Louis, MO-IL – Population 2,756,664 – link

Tampa-St. Petersburg-Clearwater, FL – Population 2,733,998 – link

Baltimore-Towson, MD – Population 2,641,566 – link

Details and Acknowledgments



Positive results from the Moving to Opportunity Program (Finally)

The latest edition of the American Economic Review includes a paper with positive impacts of the Moving to Opportunity program (thereafter MTO) on children’s outcomes: The Effects of Exposure to Better Neighborhoods on Children: New Evidence from the Moving to Opportunity Experiment, by our Raj Chetty, Nathaniel Hendren and Lawrence F. Katz

The Moving to Opportunity (MTO) experiment offered randomly selected families housing vouchers to move from high-poverty housing projects to lower-poverty neighborhoods. We analyze MTO’s impacts on children’s long-term outcomes using tax data. We find that moving to a lower-poverty neighborhood when young (before age 13) increases college attendance and earnings and reduces single parenthood rates. Moving as an adolescent has slightly negative impacts, perhaps because of disruption effects. The decline in the gains from moving with the age when children move suggests that the duration of exposure to better environments during childhood is an important determinant of children’s long-term outcomes.

That is very interesting, because it shows positive impacts of the MTO program for children, when the impacts for adults were … positive but not particularly striking. From Long-Term Neighborhood Effects on Low-Income Families: Evidence from Moving to Opportunity, by Jens Ludwig, Greg J. Duncan, Lisa A. Gennetian, Lawrence F. Katz, Ronald C. Kessler, Jeffrey R. Kling, Lisa Sanbonmatsu (yes these randomized controlled trials take armies of researchers to complete):

We examine long-term neighborhood effects on low-income families using data from the Moving to Opportunity (MTO) randomized housing-mobility experiment, which offered some public-housing families but not others the chance to move to less-disadvantaged neighborhoods. We show that 10-15 years after baseline MTO improves adult physical and mental health; has no detectable effect on economic outcomes, youth schooling and youth physical health; and mixed results by gender on other youth outcomes, with girls doing better on some measures and boys doing worse. Despite the somewhat mixed pattern of impacts on traditional behavioral outcomes, MTO moves substantially improve adult subjective well-being.

That paper was in the working paper series of the National Bureau of Economic Research, the repository for some of the newest edgy research.

Now, what’s interesting about the AER paper (the first one) is that it is pretty consistent with the very substantial amount of research that shows that most of educational inequalities are shaped early in life. This is what Heckman says, this is what my rather brainy PhD mentor says.*

PS: back blogging after a pretty intense teaching period and a great talk in Manhattan, joint with New York University’s Furman Center for Real Estate, NYU Stern’s Urbanization Project, and the NYU Marron Institute for Urban Management.

For an introduction the Moving to Opportunity program, see the NBER’s description.

*  Although in the Maurin paper, children are older than the ones for whom there is a positive impact in the Chetty et al. AER paper.





Did “The Big Short” Get it Right? What Economics Says

The Big Short was released last week, and I took a small break from my research work to watch it with my wife. The movie theater was quite empty (looked more like a documentary about banking, mortgages, CDOs, Synthetic CDOs) but the movie has a nice IMDB rating (8.1/10).

The star-studded movie features Brad Pitt, Steve Carell, and many others, playing traders who figured out early on that Mortgage Backed Securities (MBSs) had AAA ratings that were too good to be true.

Some columnists (Forbes…) don’t like the movie: unhappy that regulators are not getting the blame for the mortgage crisis.

As with plane crashes, the 2008 crisis is not driven by one single incident that took down the economy. It’s a series of incidents happening at the same time:

  • Households did have inflated expectations of house price increases. That’s well documented in the usual Shiller references. Households were wrong on the pace, probably not on the fact that some MSA supply can’t expand (e.g. San Francisco). Irrational exuberance was everywhere, but likely more in places where land is abundant but prices keep increasing. Look at the scary Case Shiller house price index in Dallas: St Louis Dallas TX House price index
  • The securitization market has an asymmetric information problem: lemons are chasing the good products, a problem that was amplified by simple securitization rules (such as a minimum FICO requirement).
  • Securitized products can be complex, very complex. But that’s a pretty fixable problem: loan-level data does exist, and it’s fairly easy to dig into the characteristics of the mortgages to check that the MBS is sound.
  • Accounting rules such as the Basel agreements, give preferential capital treatment to securitized products. In plain English, they get a lower coefficient in the calculation of the capital ratios. That’s a substantial incentive to package products in CDOs.
  • Once mortgages are packaged into a single securitized product, the price of the MBS depends on an evaluation of the risk. Takes some additional homework to figure out whether the price is right.
  • Yes, pooling mortgages leads to a lower risk for the MBS as a whole, but that’s only true if the risks are roughly independent (that’s essentially an application of the central limit theorem in statistics). But that’s a wrong application of the central limit theorem if that leads people to think that packaging anything will lead to lower risk: if there are correlated shocks that affect all assets of the securitized product at the same time, packaging will do little to lower risk.
  • Catastrophe was brewing when Fannie and Freddie withdrew from the market and the so-called Private Label Securitizers took their market share. You can see this quite easily using data I got from Blackbox.
  • Adjustable rate mortgages (ARM) were OK as long as prices were rising. Households were not expecting to see the adjustable rate period kick in.

What about Fannie Mae and Freddie Mac’s housing goals?

Fannie and Freddie did have securitization goals (set by Congress) but these were for either underserved areas or low-income households. And careful estimation of the impact of these thresholds on mortgage supply suggests there isn’t much happening there. These housing goals don’t seem to be the reason the crisis. As mentioned earlier, it’s more that, when the GSEs withdrew, private label securitizers took the GSEs’ market share, with much less rigorous securitization practices.

What about monetary policy?

The fed started tightening around 2006 but no one can really say that a single interest rate can be the cause of this. The transmission channels of monetary policy are just too complex to be able to estimate the impact of nationwide interest rates on the housing market. Some markets didn’t heat up, after all.

So it’s got quite little to do with the GSEs Fannie and Freddie — it’s a story of capital ratios, PLS securitization markets,  and overinflated household expectations. Doesn’t seem like we’ve done much on these topics since the crisis.

My work on the topic is forthcoming in the Review of Economics and Statistics.

Rickshaw economics

I spent the last week in Sri Lanka, a good time to do some back-of-the-envelope urban/development economics, in a country mending itself after three decades of civil war. Sri Lankan growth has been quite impressive: while GDP per capita was about $100 higher than Indian GDP per capita in 1990, the country’s GDP per capita is now more than double the Indian one.

Walking in the streets of Colombo or along the roads of the inner countryside, it seems quite evident there are a few priorities that will enable Sri Lanka to leverage its growth further: improving urban planning, public and private transportation, and property rights.

I’ll start with the most typical mode of transportation there: the auto rickshaw.

Auto rickshaws make sense in cities with little or low-quality public transportation. They offer a low cost, flexible transportation method. On the downside, Sri Lankan rickshaws are using gasoline — not natural gas as Indian rickshaws do. It doesn’t take fancy measurement devices to feel dust and particles as a pedestrian. And what about turning Sri Lankan rickshaws into electric rickshaws? Such rickshaws could be inspired by the Renault Tweezy.

There are of course issues with such proposal:

(i) each rickshaw driver sets his price freely. This likely pushes rickshaw drivers to price close to their marginal cost, i.e. the cost of gasoline and the opportunity cost of time. Ride prices are bargained one-on-one with clients and are not recorded or observable. That’s bad for the rickshaw industry as a whole; and that’s not enough to save for future rickshaw upgrades and maintenance costs.*

(ii) the largest benefits of cleaner rickshaws accrue to people who are not the drivers themselves. Rickshaw drivers don’t have strong enough incentives to drive better and cleaner rickshaws.

(iii) there’s a tragedy of the commons in the sense that rickshaw drivers collectively may benefit from higher prices, better maintained rickshaws, and cleaner air, but they individually have very little incentive to invest.

All of those remarks point towards a rickshaw service with a minimum price, and frequent maintenance with a number schemes. That should be managed either by a few oligopolistic chartered companies such as an Uber or Lyft, or a municipal regulator.

* Rickshaw drivers also practice quite a bit of price discrimination, trying to assess where a customer is from and thus what is his willingness to pay for the ride. That can sometimes allow them to charge much more than their marginal cost of the ride. A regulated price wouldn’t allow them to price discriminate like this, but a regulated minimum price would. The problem is not that prices are too high, but that prices are too high. Seems like rickshaw drivers are hurting