Thinking at the European Central Bank: House Prices and Small Business Lending

The Center for European Policy Research organized a labor market workshop at the European Central Bank recently. The conference happened in the ECB’s new building, on the banks of the river Main, a deconstructivist structure designed by the architectural firm Coop Himmelb(l)au. The firm has designed buildings in Lyon (Musée des Confluences), in South Korea (the Busan Cinema Center). The ECB building has quickly become a Frankfurt landmark.

The focus of the conference was not urban economics — although one can argue that unemployment has major implications for mayors. I discussed a paper that aims at showing the impact of the housing market on job creation, in a way similar to Adelino, Schoar, and Severino (2015). The purported causality runs as follows: a change in house prices affects the value of small entrepreneurs’ houses; this affects the value of their equity, and thus the value of the collateral in case entrepreneurs use this equity to borrow and fund their ventures. The authors argue that an increase in house prices will lead to more job creation by small and medium entreprises, and symmetrically a decline in house prices will lead to fewer job destructions. This is the collateral channel.

Small businesses in areas with greater increases in house prices experienced stronger growth in employment than large firms in the same areas and industries. Adelino, Schoar, and Severino (2015)

Why small and medium enterprises? Well some authors have argued that small and medium-sized firms are the main drivers of job creations. But that remains debated (see Davis, Haltiwanger, Schuh 1996, and here): yes, small businesses create more jobs, but they also exit at a much higher rate. In gross flows, small businesses indeed create most jobs. In net though (accounting for the exit rate), their contribution is substantially smaller than larger firms.

Of course, arguing that house prices affect job creation through the collateral channel is hard to demonstrate: it is no surprise that house prices and job creation correlate strongly; for instance, jobs in construction and real estate services vary in an easy-to-understand way with house prices. That is the real estate sector channel. House prices partly reflect demand for construction and real estate services, and that’s a much simpler Econ 101 explanation of the correlation than the collateral channel.

Another reason why it’s hard to make the point that house prices cause job creation through the collateral channel is because local banks suffer heavy losses when house prices decline. And thus extend less credit in tough housing market times. Houses are a big share of banks’ balance sheet… and thus the decline in house prices doesn’t cause a limited ability to borrow because of the entrepreneur’s own house value, but because the bank holds mortgages that default with a higher probability. This is the mortgage channel.

I love papers that show that the housing market “causes” the labor market. But it’s really tough to make a bullet-proof case for the collateral channel.

Adelino, Manuel, Antoinette Schoar, and Felipe Severino. “House prices, collateral, and self-employment.Journal of Financial Economics (2015).

Davis, Steven J., John Haltiwanger, and Scott Schuh. “Small business and job creation: Dissecting the myth and reassessing the facts.” Small business economics 8.4 (1996): 297-315.

Federal Reserve of St Louis “Are Small Businesses the Biggest Producers of Jobs?” by Kliesen and Maues.

Portland’s Land Use Regulations


I am in Portland for the Urban Economics Association meeting, an exciting meeting of like-minded economists studying the dynamics of cities, the relationship between the demand/supply of housing and prices, and many other topics.

Being here is also a great opportunity to learn a bit about what everybody’s talking about when it comes to Portland’s urban dynamics: Portland has stringent land use regulations that push prices up and hinder the metropolitan area’s growth. This is what the economist says about this.

Some evidence supports the hypothesis that land use regulations push prices up. In particular, house prices (unlike GDP) are back on an upward trend that means that today’s prices are back to their 2006 level, at the peak of the housing boom.

Are Portland’s land use regulations that stringent? A quick look at the best available data, from Joseph Gyourko’s land use regulation survey data, reveals the following:

  • Portland’s regulation index iPortland WRULIs around the median (0.11), which suggests that the claim that Portland is one of the most regulated cities in the US isn’t supported by the data. The regulation index summarizes cities’ answer to about 11-12 questions.
  • It doesn’t seem that the length of the zoning process, the length of the permit process, school crowding, council or citizen opposition are major factors that constrain the supply of housing units.
  • The single most important factor that stands out from Joseph Gyourko’s survey is simply that the demand for single family housing far outstrips the supply of zoned land for such single family housing.

Baltimore’s Vacant Housing Crisis

The recent Baltimore riots highlighted the plight of inner city — mostly African American — residents of Baltimore. A good way to understand the issues at hand is to map the spatial distribution of housing, and in particular vacancies. In the glorious days of post-war Baltimore (population peaked there at  around 950,000 residents in 1950), demand for housing in the city center was outstripping the actual supply of housing units. But the city’s population has shrunk by about 37% in 60 years, and there are vast amounts of unoccupied housing in the city center.

An interactive map of neighborhoods (a.k.a as census tracts in geographers’ parlance) can be found on my website here. I present in this map the fraction of vacant housing units in each neighborhood, together with the number of housing units in each neighborhood.

The map presents the distribution of vacancies for the entire metropolitan area. The rectangular boundaries of the city — much smaller than the metro area, which includes Baltimore’s suburbs — clearly delineate urban blight. For instance, Rosedale, Maryland is a small community right to the East of Baltimore city. Interestingly, Rosedale is a more than 70% white and middle-class suburb, which sits right next to inner-city neighborhoods where a fifth to a quarter of all houses are vacant.

The World is Not Flat: Distance Matters

The early 1990s were full of excitement about the then-new idea of `telework’. A study titled ‘Telework: A New Way of Working and Living‘ told us that:

Telework is ten years old. In its short time its capacity for redrawing the geographical and organisational boundaries of the traditional, centralised entreprise has been amply demonstrated. The positive consequences: […] new employment opportunities for various categories of workers, potentially without geographical limits.

The dream was that we could work from anywhere: congested roads during the morning commute would be a thing of the past as we could simply stay at home; or move to an inexpensive countryside house with a pool and a garden — all we need is an internet connection.

We would avoid the long flight across continents to meet face-to-face or discuss contracts. Conference calls and visioconference would wipe out jetlag fatigue and impersonal hotel stays.

Of course that vision didn’t come through. Technology has made face-to-face communication more important, not less. Online communication is much more likely with somebody living in the same metro area than with somebody living in another urban cluster. Wannabe IT entrepreneurs still need to meet their Venture Capitalists face-to-face: would Marc Andreessen endow a new entrepreneur that he hasn’t met in person? The whole premise of the Silicon Valley is that people want to live close to other people with the same interest.

In a recent paper titled ‘Does Distance Matter in Banking?’, Kenneth Brevoort and John Wolken show that most banking relationships remain local, with a median distance between lender and SME of 5 miles. More significantly perhaps, such distance has actually decreased in the early 2000s.

Distance matters. The world’s cities are as important as ever as families are ready to pay large sums of rental money to live in crowded apartments rather than in less crowded rural or semi-rural houses. Close to 55% of the world’s population now lives in urban areas. The world is not flat.

U.S. Housing Market

A quick look at the house price index (either the Case Shiller index or the U.S. Federal Housing Agency index) might worry us. After all this is what the graph looks like from 2000 to now:

Sounds like house prices are back to their 2005 level, right? Except that much of the increase is driven by nominal, not real increases. Deflating house prices by the Consumer Price Index reveals a slightly different graph.
In real (rather than nominal) house prices have only slightly climbed above their 2002 levels. That’s a much smaller increase, but still reveals that there is an imbalance in the demand and supply of housing units. Housing starts are more than 35% below the 1500 level.
There is a tight connection between population and housing demand. The construction of new homes is not keeping up with the demand pressure. What’s new in 2015 is that the supply of new homes is not keeping up with demand. As the graph below suggests, since 2007 we seem to have entered an area of low construction.
One interesting thing: there is a large increase in the construction of new multi-unit structures. A look at construction starts for 5-Unit Structures or More reveals a much different pattern; the U.S. construction industry is back at building dense, urban dwellings.

Thursday Readings – London Housing Crisis, the Green Belt, and the height restrictions

  • The Economist reports on the growing housing crisis in London. And on the urban “regeneration” projects that push low- or middle-income tenants out of central London. Link here.
  • The Global Urbanist argues in favor of lifting the London Green Belt’s building restrictions.
  • But rather than building on the Green Belt, what about building taller in London? The Guardian argues that taller buildings (e.g. in Canary Wharf) have not prevented rising transaction prices… but these are commercial buildings, not residential developments.

Shanghai Tower opens this year

NPR reports on the planned opening of Shanghai Tower later this year. They report that the only building in the world that is taller is Burj Khalifa in Dubai. The project is also unlikely to break even for any private investor. Somehow a number of these large scale Empire State Building like projects don’t make sense from the point of view of a private investor. But do they make sense from the point of view of the city or country that is building them? Is it the economics or is it something else that is motivating such investment? A few ideas:

  • High rise buildings lead to higher densities within metropolitan areas. Higher densities are generally leading to more innovation, creativity and exchange of ideas. See this paper for instance.
  • Higher densities also lead to higher demand for local retail, generating a greater diversity of offerings.
  • Higher densities lead to lower transportation times, length, and costs. Higher buildings lead to less urban sprawl and thus to fewer negative externalities for other residents of the metropolitan area. Urban sprawl is bad.

In particular, it is interesting that a city state like Singapore or the city of Abu Dhabi have chosen to allocate limited land for urban expansion, preferring to build vertically rather than expanding horizontally. The city of Singapore still has plenty of greenery, while Abu Dhabi is fairly vertical despite a potentially unlimited supply of land.

In many cases it is rather more efficient to limit the use of land, by precomitting to using only a limited amount of the physical space rather than expanding.

As the incentives of private investors are not to build vertically, it is good for regulators and institutional investors to put forward bylaws and land use regulations, and promote developments that lead to higher densities.

The NPR story here. The top 10 tallest skyscrapers.

Nechyba, Thomas J., and Randall P. Walsh. “Urban sprawl.” Journal of economic perspectives (2004): 177-200.

Carlino, Gerald A., Satyajit Chatterjee, and Robert M. Hunt. “Urban density and the rate of invention.” Journal of Urban Economics 61.3 (2007): 389-419.