How London developers responded to affordable housing requirements
A new study explored whether a stricter inclusionary zoning policy impacted the city’s housing supply. Here’s what they found—and what other cities can learn.
The housing shortage in high-demand cities needs many types of solutions, and one of the most popular policy tools in recent years has been the adoption of affordable housing requirements for developers, also known as “inclusionary zoning.” While the exact rules vary from city to city, inclusionary zoning generally calls for new development of a certain size to include a certain percentage of low-income units. The goal is to support inclusive neighborhoods that remain accessible to at least some households unable to pay market prices.
Two common criticisms of inclusionary zoning rules are, one, that they don’t produce a significant amount of affordable units, and two, that they discourage development in ways that dampen total housing supply. A well-designed new study, published this summer in the Journal of the American Planning Association, provides some evidence for the first concern but not for the second — at least for the housing market in London. The researchers Fei Li of Georgia State University and Zhan Guo of New York University conclude:
There was no net loss of new homes, though the strategic behavior of private developers could have dampened the affordable housing output … .
Let’s dive deeper into the work, and what we can learn from it.
What they did
For the study, planning scholars Li and Guo focused on a change in London’s inclusionary zoning law that occurred in 2008. That year, London expanded its rules to require affordable housing in new developments larger than 10 units. Prior to that time, only new development projects with 15 units of more were required to include low-income housing.
The 2008 turning point enabled Li and Guo to compare development patterns pre-2008 to those post-2008, in an effort to see whether the inclusionary policy had a measurable impact on the types of projects completed across the city. Using official city housing data, Li and Guo analyzed development patterns across the 33 London boroughs from 2004 to 2014, covering some 44,000 projects and 572,000 housing units.
This “quasi-experimental” study design — directly comparing pre- and post-policy activity — enabled Li and Guo to isolate inclusionary zoning as the main reason for developer decisions. In the absence of such a design, it would be hard to state with confidence that the policy change played a direct role in shifting the market, given all the complex factors that go into development patterns in big cities.
To further isolate the role of inclusionary zoning in any market changes, they controlled for factors such as population growth, density, and median housing price, and they excluded projects that weren’t subject to the policy, such as all-affordable developments.
Simply put, the researchers wanted to know how — if at all — housing projects changed in response to the 2008 rule. Did developers keep building 15-unit (or larger) projects despite the new affordable housing requirement? Did they build smaller projects that avoided the rule, leading to fewer affordable units? Or did they build fewer projects altogether, leading to less overall housing in the city?
What they found
The analysis was clear: developers changed their behavior in response to the affordable housing policy. Li and Guo report that, in the years following the 2008 policy change, London developers built substantially fewer projects subject to the new requirement (10- to 14-unit projects) and substantially more projects that fell outside the policy threshold (9 units or fewer), compared with pre-2008 market patterns.
On average, after 2008, London boroughs had 3.3 fewer 10- to 14-unit developments per year, against 5.6 more 5- to 9-unit developments. In particular, developers built roughly nine times more 9-unit projects (802) than 10-unit projects (88) after 2008. Other market segments, all the way up to 39-unit projects, experienced only “minor or insignificant” changes to development patterns, suggesting no broader market trends in play after 2008 — and lending further support for the direct impact of the inclusionary policy.
This main finding produced two clear insights about the impact of London’s inclusionary zoning policy:
First, London saw only a modest increase in affordable units. Li and Guo report a net increase of two affordable units per borough, per year, among the target developments of 10 to 14 units following the 2008 policy change. That means the policy worked as intended: projects of this size produced more affordable housing than they had in the past. But the volume of affordable units was very modest compared to the need and the goal, in large part because developers built more projects that fell outside the policy threshold.
Second, London saw “no net loss” of new housing as a result of the policy. The study found “no significant change” in the total number of housing units produced post-2008 in developments of 5 to 14 units, compared to pre-2008 trends. In other words, the higher volume of smaller projects (under 10 units) more or less offset the reduction in larger projects (over 10 units). That’s good news for the city’s housing supply as a whole, and a counterpoint to common fears that inclusionary policy stifles housing production.
What it means
To recap: London’s inclusionary zoning policy didn’t dampen housing supply but only produced a modest amount of affordable units. Several takeaways come to mind.
First, a reminder that it’s always hard to generalize policy results from one study or one city. The current study ended its analysis in 2014, and since that time the housing affordability crisis has only grown in high-demand cities like London. Long-term studies remain important complements to short-term research. Additionally, London’s inclusionary requirements tend to be more ambitious than those in American cities — targeted 30 or 50 percent affordability, against a 16 percent U.S. average — which makes a direct comparison tough.
Second, London’s regional policy approach led to geographic consistency. While some boroughs adopted the 2008 inclusionary policy more quickly than others, Li and Guo found no evidence that developers tried to complete larger projects elsewhere before the new requirement took effect, likely because they knew it was coming everywhere soon. In other words, the universal policy design discouraged developers from moving to areas that might otherwise have allowed a more lenient approach.
Third, policy design is about trade-offs. Developers did fundamentally change the type of projects they pursued in London post-2008, in ways that arguably hampered the production of affordable units. But if the policy had set even stricter size thresholds, it’s possible that total housing supply might have decreased, which would have been a harmful outcome, too. Li and Guo suggest that cities explore ways to “soften” the line separating the markets that fall inside and outside of an inclusionary policy threshold.
Finally, it bears repeating that inclusionary zoning is only one of many tools needed to address the housing crisis in growing cities. Low-income units are vital for neighborhood stability. But cities also need novel approaches to community wealth generation that help households keep up with rising prices over the longer term. They need construction innovation that makes it easier to produce more housing on faster timelines. They need technology that makes it more comfortable to live in smaller units. They need all of these ideas — and more.
Follow Sidewalk Labs with our weekly newsletter or subscribe to our podcast, “City of the Future.”
