Depending on where you live, you might be quick to call accessory dwelling units (ADUs) one of the biggest housing success stories of the past decade. Last fall, Nolan Gray from California YIMBY covered his state’s rapid expansion of ADU permitting — highlighting Los Angeles as a particularly exceptional story, where thousands of ADUs are being permitted every year. This March, the Seattle Times covered new data from Seattle showing an ADU boom of their own. An impressive headline led this coverage: Seattle is now building more ADUs than single-family homes.
But in this conversation, Minneapolis and St. Paul are nowhere to be found. As a few local sources have reported, the core cities have seen only minimal construction of ADUs in the past few years. To quantitatively contextualize this, consider three graphs below.
The first shows the absolute numbers of ADUs in Los Angeles, Seattle, and St. Paul. The second shows the number of ADUs permitted as a percentage of that year’s housing permits — the relative contribution of ADUs to housing supply growth. The third shows ADUs permitted per 10,000 residents, which gives a rough idea of how meaningfully ADUs are adding to citywide housing supply. These numbers show 1) how amazing Los Angeles’s ADU growth is, 2) that Seattle is nearly as impressive, and 3) that ADU development in St. Paul has been practically meaningless, no matter how you slice it.
(Two data notes:
- I requested data from the City of Minneapolis and wasn’t able to get clear annual data. But last December, the Star Tribune reported that 176 ADUs had been permitted in Minneapolis since 2014 — this puts them at the same magnitude as St. Paul.
- In these figures, I used ADU permits issued for all three cities, which does not always translate to ADUs ultimately built as projects are occasionally canceled. But it’s where the data are consistently available, and still a useful indicator.)
From looking at the Twin Cities alone, it’d be easy to conclude that ADUs are just not that promising of a model. But when looking at West-coast success stories, they seem like a miraculous housing tool.
Why is the difference so huge? There are surely some regulatory barriers and institutional adjustments holding the Twin Cities back from more ADU development. But the biggest issue is that the regional housing markets are completely different.
The Many Varieties of Local Housing Markets
The housing markets in Los Angeles and Seattle are fundamentally different than the housing markets in the Twin Cities. Research published by economists Liyi Liu, Doug McManus and Elias Yannopoulos in 2022 helps show how. The authors analyze how housing in different cities has “filtered” over the years — as a house gets older, how do the incomes of that house’s new buyers change? In a healthy housing market, we might expect buyers’ incomes to decline over time as older stock filters down the income distribution. Similarly, cars almost always filter down to purchasers at lower incomes as they age and are resold.
The graph below shows how the incomes of a home’s buyers change as the home gets older:
In the Twin Cities (this paper looks at metropolitan areas, hence “Minneapolis” in this graph actually refers to 15 counties in the metropolitan area), older homes are typically bought by slightly lower-income homebuyers, while in markets like Chicago and Detroit the average income of homebuyers decreases as the homes age. Cities like Los Angeles, however, have a problem: when homes get older, more high-earning people buy them — meaning that houses in those markets “filter up.” Other estimates from Liu et al.’s paper put Seattle in this same category.
What does this tell us about the difference between these cities? Even as homes age, they continue to be bought by people with higher incomes than the previous owners. This is a sign that these cities have demand for housing that is greatly outstripping supply (if you look at rent growth or housing price growth, similar differences appear).
As a result, when ADUs are suddenly allowed, cities like Los Angeles and Seattle immediately see strong pressure to build them. Similarly, when Santa Monica lost its zoning control after failing to comply with state housing law, developers filed for 4,797 permits within one week — and this was housing with affordability restrictions.
Alex Schieferdecker offered the same insight in a post on this site last year, when he looked to explain the lackluster growth in housing permitting after the Minneapolis 2040 Plan loosened a variety of zoning restrictions. He wrote that in pricey coastal cities “zoning restrictions act like a cork holding back a shaken bottle of champagne,” while Minneapolis “is more akin to a flat soda.”
The relative growth of ADU permitting across these cities is another reflection of that same difference in demand.
Fixing What We Can
Zoning codes are often dense and arcane, and development that they allow in theory may not always be allowed in practice. For example, at least part of our slow take-up of ADUs has to do with the owner-occupancy requirement, which Minneapolis only eliminated in early 2021 and St. Paul only eliminated in early 2022. Under this rule, an ADU could only be built if the owner of the lot lived in the ADU or the main property — a strict requirement that surely shrank the potential for ADU development.
As Bill Lindeke highlighted last year, other rules continue to be a problem with the development of ADUs. In Minneapolis, veteran-focused ADU developer YardHomes couldn’t build an ADU because it would’ve exceeded zoning maximums for floor area on the lot. Multiple neighbors testified in support of the development, but the existence of two garages on the lot meant that the ADU wasn’t allowed — another example of the arbitrary rules in our zoning code that undermine our stated housing goals.
Behind Los Angeles’s and Seattle’s ADU construction booms were smart policy changes. Like the Twin Cities, these cities used to have planning rules that technically allowed but practically prevented ADUs. Through continued analysis, these cities found out what wasn’t working and adjusted their rules to embrace ADU development. The Twin Cities should continue to embrace these kinds of iterative, responsive policy reforms.
Of course, a host of technical, practical problems are not regulatory in nature. Institutional change and new expertise are likely needed to support ADU development. In general, the local development ecosystem doesn’t have much experience building ADUs, while many homeowners may not even know that they can build these units. Similarly, interested ADU builders today may have trouble accessing the right financing to pay for construction costs upfront. As Mark Thieroff suggested in a story for Streets.mn, construction will hopefully become easier as more developers or homeowners gain familiarity with ADUs.
Across these various roadblocks, we should keep in mind what Daniel Herriges of Strong Towns calls the “limiting nutrient.” Borrowing a concept from biology, Herriges argues that complex processes, like urban development, are typically constrained by a specific binding limitation. As a result, changing other non-binding limitations won’t affect the system. That seems to be at work here — we’ve made ADUs easier to build in the Twin Cities, but there is still a binding “limiting nutrient” blocking their growth.
Low-Impact Good Housing Policy Is Still Good Housing Policy
Other cities’ successes with ADUs show that they can provide a powerful path for increasing local housing supply and choice. They also make the Twin Cities’ results thus far a disappointment.
We can’t change the economic conditions that are driving other coastal cities’ ADU booms. But remember — this is a good thing. Certainly, our housing is too expensive, and we should break down barriers that drive up costs while targeting investment to our underserved low-income population. However, unlike other cities, we don’t have deep housing shortages causing homes to filter up and rents to grow at a breakneck pace.
Even if better policies don’t lead to immediate ADU booms (or booms of other types of housing), they’re still the right direction for land use policy to take.
While we are unlikely to see enough ADUs to significantly increase market-wide supply and drive prices down, ADUs can also create a different kind of affordability. ADUs help share land costs within a lot, and because they are small they are typically operated at lower marginal costs. Even if we don’t get enough ADUs to drive down the overall price of housing, each ADU is a unit that can be rented more cheaply, allowing people to make tradeoffs between living space and other preferences.
Although the Twin Cities is not an extremely supply-constrained housing market, it’s worthwhile to preempt such a situation from ever happening. A few generations ago, Los Angeles had plenty of capacity to build new housing — but rapid downzoning in the 1960s and and 1970s laid the foundation for an extreme housing shortage. The Twin Cities have been wise to chip away at these regulatory barriers with things like ADU reform.
While the Twin Cities could still make meaningful progress on ADU development, they’re unlikely to ever replicate the sweeping successes of places like California. That’s OK. In other cities, extreme supply restrictions and surging housing demand have created an untenable status quo. And ultimately, a good first principle for housing markets is to seek never to become like California.