I frequently reference my post summarizing the interesting research at last year’s Urban Economics Association conference. This year was as insight rich as ever. And the Montreal setting was truly special – it’s a city that’s as European as any in North America. (But it’s still North American. One attendee from France told me he kept addressing locals in English because the setting felt so North American to him.)

Greenbelts

Several researchers presented papers about greenbelts or urban growth boundaries, but they were scattered across sessions.

Toronto greenbelts (Hempel)

Alexander Hempel studied Toronto’s greenbelt policy, which was implemented in the 2000s. Because it was relatively loose at the time of implementation, the price effects were muted: just a 3% price increase, less than a twentieth of the total price appreciation during that period. The price effect within the greenbelt was about twice as large, which puts a low ceiling on the amenity value of living in the greenbelt. The technical challenge for Hempel is to distinguish the effects of the greenbelt from those of the accompanying urban upzonings.

Daniel Bigelow and Justin Gallagher are studying Oregon’s urban growth boundary (old slides). Their work is mainly descriptive. They find that the UGB does what it says on the tin: density is much lower outside the boundary than inside. But that, of course, only emerged over time. Oregon passed its UGB law in 1973 and mandated implementation by 1980. It was only in the mid-1990s, though, that they see evidence that UGBs were binding. This is a caution for Toronto and other recent adopters: a UGB that seems generous at implementation will eventually tighten.

I didn’t see Lukáš Makovský and Capucine Riom‘s presentation on London’s Greenbelt. But their poster from another conference shows that a 1% increase in wages in a typical UK local authority results in 0.8% in-migration. But in the Greenbelt, wage increases have no population effect.

Delays and designs

Job market candidate Danny Gold analyzed Seattle policies that have been a reform priority for Cascadian YIMBYs: curbing the uncertainty and delay of design review. He uses the city’s rules through 2018, which required design review for buildings of more than 8 homes. As expected, there is bunching right at 8 units and larger buildings take longer to approve than small ones. Design review with a unit-count threshold caused developers to build fewer, slightly larger apartments.

But some 9-17 unit buildings are built, and those have relatively quick approvals. He interprets this as evidence of self-selection: builders go ahead with projects just above the threshold only if they expect easy approval.

Using the universe of building permits, he estimates that design review adds 4 or 5 months to the approval process (median: 14 months). Going through design review adds $14,000 to the cost of the median building, but with a large right tail.

Thanks to state-level reforms limiting design review, Seattle suspended design review a week before the conference.

Gabriel Ahlfeldt presented a summary of the Economics of Architecture, work undertaken with E. Pietrostefani and Ailin Zhang.

The internal premium for distinctive design is +15%.
The external premium is +9%, leaving room – at least in theory – for government intervention.
In a simulation, they find that the optimal policy is to subsidize distinctive design by 10% of construction cost.
Having “design districts” can be OK.
But restricting supply and giving a density bonus for distinctive design is worse than doing nothing.

Seattle’s policy – restricting supply and giving a density bonus for design review that doesn’t even achieve visibly distinct results – is even worse than the worst idea Ahlfeldt considered.

Christian Hilber presented work with Michael Ball, Paul Cheshire, and Xiaolun Yu. Their data – from an insurer – covered 140,000 England sites from 1996-2015. Construction is slow in the UK: 597 days, versus 371 in the US, despite the former having smaller build sites.

Planning restrictiveness (Ball et al)

Delays are endogenous. With higher demand growth, developers speed up projects to try to reach the market while it’s hot. But that motive is driven by competition. With less competition, they have less motive for speed. In UK locales with less available land, more restrictive regulation, or higher market power, demand shocks have less impact on delay.

Generalizing, they estimate that reducing regulatory restrictions by a standard deviation would decrease construction duration by 11%. Increasing land availability or decreasing market power by one S.D. would decrease construction time by 8%.

Market power

In the same session as Hilber et al, a few papers studied the market power of large landlords, who may have the ability to set prices monopolistically.

Fern Ramoutar used Chicago’s uniquely (?) partitioned rental market to study landlords’ market power. As evidence that rental markets are segmented, she showed that most moves take place within small geographic areas and (more surprisingly) among buildings of similar sizes. In Chicagoland, larger buildings (50+ units, typically high-rises) are the most expensive, and 2-4 flats the cheapest.

Once the market is sliced into neighborhood x building size segments, one landlord can plausibly dominate the segment. Using a staggered diff-in-diff, she finds that rents rise 5% following an acquisition of a competing property by a “dominant” landlord, relative to acquisition by a non-dominant landlord. (I’m not in IO, but I was confused that “dominant” was defined as any landlord with at least 1% of market share within a segment. Is that a plausible cutoff?)

Duke student Joe Fish studied Boston and Houston to get a broader sense of the importance of large landlords.

The share of large landlords has declined slightly over the past few decades.
Research in this area has been limited because most buildings are owned as unique LLCs, so it’s hard to know who owns what.
Larger landlords buy cheaper properties (mostly because they concentrate in cheaper neighborhoods). (But is this just a function of the fact that if you invest $20m in a cheap neighborhood, you’re more likely to pass Fish’s size cutoff?)
Acquisition by a medium or large corporate landlord is associated with a 4% increase in rent over 4 years. There’s no rent change from acquisition by a small corporate landlord or non-corporate. But the effect is mostly driven by landlords who pull a renovation permit.

There’s some tension between Fish’s work and Ramoutar (which excludes renovation cases).

Cautions for reformers

As a group, urban economists are natural YIMBYs. They’re mostly young and live in high-cost cities,...

Market Urbanism

Message History

Add new keyword

Delivery option(s)