The sense of crisis around housing affordability in many of America's cities is at the highest level it's been in a long time. Responses to the crisis are often politically contentious, but the existence of a crisis is no longer so. And policy makers have been marshaling quantitative studies to make the case for their preferred reforms, leading to a barrage of headlines such as:
California needs to build 3.5 million new housing units by 2025, according to Governor Gavin Newsom, who promises to make it happen.
Portland, Oregon needs to add 3,500 - 4,500 housing units each year for the 30-year timeframe to 2035, according to the city's Housing Needs Analysis, whereas from 1997–2017 it averaged under 3,000 per year.
Numbers are a good rallying cry. Numbers are a measurable goal that can inform concrete action. They're useful.
But numbers also simplify. They create the temptation to become fixated on the number itself and lose sight of context. They cause us to fixate on outcomes and forget process.
The Process Matters
At Strong Towns, we're all about process. In the tradition of urban thinkers such as Jane Jacobs, we think that how our cities are planned, financed, and built is more important to get right than knowing the type of city that will ultimately emerge. And yet that's hard to reconcile with the urgency with which people, justifiably, talk about housing affordability, which leads to proposals like these:
"Ban zoning statewide."
"Abolish all height limits. Allow only 50-plus-story towers near transit stations."
"Just build it all as social housing."
"Use all vacant land owned by local government for high-rise affordable housing."
These are all things I've heard advocated in earnest by well-meaning housing advocates. In other words, just get the units built. Who cares how they're distributed or who builds them? Just make it happen! We need them!
The world looks simple when you optimize for one variable. And yet, anyone familiar with the history of North American cities (or, almost certainly, even just their own city) can cite examples of grave harm done by this sort of tunnel vision.
Optimize for the goal of removing visible blight? You get the bulldozers of urban renewal.
Optimize for the goal of providing public housing as cheaply and politically-uncontroversially as possible? You get horror stories like Pruitt-Igoe.
Optimize for the goal of moving traffic at high speed, despite the risks? You get statistically inevitable tragedy.
Our decisions have fallout. Optimize housing policy simply to increase the number of homes, and there's no question that we'll make decisions with the potential to radically transform a place's physical and social fabric in a short amount of time. And that they will have fallout too.
And yet, those numbers.
A common form of pushback we get at Strong Towns when we talk about incremental development as a process is that it's not up to the scope of the challenges our cities face. That, by asserting that cities are complex systems and we should tread lightly in disrupting them through top-down action, we're not taking the affordability crisis seriously. (Or climate change seriously. Or car dependence seriously. Et cetera.)
We've praised Minneapolis's plan to allow the next increment of development—duplexes and triplexes—in all single-family areas citywide. We've spoken positively of similar moves in Vancouver and urged strongly that Austin, TX follow those cities' leads. This kind of policy is a huge step toward re-legalizing the incremental, organic evolution of the city, and addressing housing affordability and ecological challenges as an additional plus.
But there's a segment of our audience that's unsatisfied: It's not enough. We can't stop at triplexes: we need larger buildings too. We need X number of units to deal with our affordability problem. We need X amount of density to make transit viable, or to make places financially productive. This triplex thing is small potatoes: it's good news, sure, but it's not up to the task of transforming our cities in the way that's necessary.
There are extreme versions of this argument out there, but I think an eminently fair and reasonable version is that made by Joe Cortright of City Observatory (a Strong Towns member and longtime friend of ours) in his essay "You're Going to Need a Bigger Boat." The whole thing is worth reading, but here are some representative excerpts:
Legalizing accessory dwelling units, duplexes, triplexes and fourplexes does hold the promise of adding to affordability while injecting some very “gentle density” into single family neighborhoods. But it is far from up to meeting the scale of our housing affordability problems. For that, we need to build larger multi-family buildings, including apartments.
The “Russell” [ed: a 68-unit apartment building in Portland that Cortright uses as a case study] avoids about 20 demolitions of single family homes. The “Russell” has a bigger impact on adjacent properties than the a single row house project, but you’d need 20 more row house projects, somewhere, to provide as much housing as the Russell....
There’s a good argument to be made that we’re underbuilding density in many locations where it makes sense. Neighborhoods with great transit, a mix of commercial uses, and high levels of walkability may justify more than just a duplex or triplex on a particular site. We ought to be think about the long term, what the demand for the neighborhood is likely to be in 2050 or even 2075, rather than in terms of the 20 year time horizon of most housing affordability analyses.
For the sake of this post, I don't want to get into the weeds of quantifying "need." Let's say Cortright is correct (and I agree that he is) that a place with the kind of pent-up demand and affordability issues that Portland faces needs a lot more new housing units than a policy of allowing four-plexes and row houses will produce on its own. So we're not really trying to argue with the numbers or the need, but rather to make observations about process.
Let me propose three rules of thumb for helping Strong Towns advocates think about addressing the housing crisis at scale, but also in ways that are as antifragile as possible.
1. Don't pick winners and losers. Let productive places reveal themselves.
Many cities have a "trickle or fire hose" problem, where development is intensely concentrated in just a few hot areas. In the county I live in, more than half of all new construction in 2016 occurred in just three ZIP codes, out of 27 total. In the rapidly-growing Seattle metropolitan area, I've documented that most residents live in neighborhoods that are virtually unchanging, while a small minority of areas undergo dramatic change (such as the Central District, which has gone from 70% African-American to only 20% in the face of rising rents and displacement).
This problem is exacerbated by public policy in two ways. One, many neighborhoods are virtually prohibited by their zoning from developing enough new homes to actually meet demand. Allowing the development of triplexes and ADUs is one way to loosen that straightjacket, but Cortright is almost certainly correct that there are places where it isn't enough to stabilize prices and let the neighborhood grow at the rate that it organically would in the absence of regulation.
Another way our local governments pick winners gets less attention from urbanist quarters, though, and that is policies that seek to concentrate high-intensity development in a small handful of areas, often near existing or planned transit service. Rail transit in particular gets used as a speculative development tool rather than as a means of connecting already-thriving places.
The intent is good: reduce car-dependence and associated fossil fuel emissions, and create more of a type of development (transit-oriented, mixed-use) for which there's demonstrated high demand. But these policies of targeting transit station areas for intense redevelopment are likely to be destructive in much the same way as building freeways to brand-new suburbs was in the 1950s and 1960s: they sapped the wealth out of existing inner-city neighborhoods by sucking up demand and investment to a handful of hot locations where large-scale developers could make a killing.
And upzoning a small minority of areas while leaving the single-family zoning straightjacket in place in most of your city has another consequence, which leads us to Rule 2:
2. Avoid policies that encourage land speculation.
Strong Towns founder Chuck Marohn has written extensively about the speculation problem that results from narrowly targeted upzoning. Think of suddenly allowing much more height and density in a small geographic area—say, along a major bus line, but not on the interior neighborhood blocks nearby. The incentives produced by this policy tend to bring incremental redevelopment to a grinding halt. If you own a vacant lot or small home near a shiny new ten-story building, it makes more sense to hold out for a developer to buy your property for a windfall than to do something more modest with it.
I suspect that broad upzones are far less likely to introduce this kind of speculation than narrowly targeted ones. In other words, if the city announces that a few blocks along a major street will be rezoned to allow 6-story apartment buildings, speculators will flood the market and seek to cash in. If the city announces, on the other hand, that several whole neighborhoods will allow 6-story apartment buildings, the developers of 6-story apartments will be selective, picking and choosing the best opportunities to build, without a huge effect on home prices where demand is lacking. This is as it should be. Let productive places reveal themselves.
We can think of this like the concept of a limiting factor in plant growth. If a plant needs a 3:1 ratio of nitrogen to phosphorus in the soil, and what exists is a 2:1 ratio, nitrogen is limiting. Adding a nitrogen-based fertilizer will likely accelerate the plant's growth.
On the other hand, if there is already a 4:1 ratio, adding more nitrogen will do nothing. Nitrogen isn't the limiting factor. Excess zoned capacity for homes, beyond the actual rate at which people want to move into a city or neighborhood, is likely to act like the nitrogen in this analogy.
3. Think about the "next increment of development" at a neighborhood level.
A narrow interpretation of "next increment"—a phrase we use a lot on this site—is that every individual building should only be allowed to be replaced with a building slightly bigger and more intense. Numerous critics have pointed out that, from the developer's standpoint, this is usually financially infeasible—it takes a bigger leap in intensity to overcome the cost of demolishing a previous structure. Cortright argues that it's also ahistorical:
It’s helpful to remember that most of the neighborhoods that exhibit a mix of housing types, including duplexes, triplexes and fourplexes were originally built that way, because early zoning codes (especially prior to World War II) didn’t prescribe only single family buildings. It’s rare to see a neighborhood completely retrofitted with just slightly higher density.
What we've actually suggested at Strong Towns in the past is a floating height limit based on the average heights of nearby buildings. This was a food-for-thought proposal, not a policy with all the kinks worked out. I would suggest defining "nearby" to span at least a few blocks. The goal should be that the neighborhood can incrementally fill in and add intensity (assuming it's a place people want to move to), even if that means significant changes to individual pieces of property when they're redeveloped.
A major goal of this kind of incrementalism is to discourage monocultures. Neighborhood should be in flux, always evolving, never finished. When a neighborhood is built to a finished state, the effects of future decline can be cataclysmic. For example, a cluster of apartments in Marietta, Georgia, built all around the same time for young professionals in the 1960s, became a concentration of poverty that so bothered city officials that Marietta spent $65 million in public money to buy it and demolish it.
If you can't imagine a concentration of homes built in a top-down, orderly-but-dumb way today to address San Francisco or Seattle or Austin or D.C.'s housing crisis becoming a similar headline 40 years from now, I submit that you lack imagination.
Build more. A lot more where necessary. But do it by letting millions of decision makers call the shots.
Incremental development does not mean slow development. It means decentralized, organic, emergent developent. It means the number of relevant decision makers is in the thousands or millions, not the tens or hundreds. Chicago grew incrementally from a city of 30,000 to one of 1 million in just four decades, 1850-1890. That looks like a huge leap if you only look at the numbers, but if you look at the process by which it was achieved, you get a different picture.
We're glad single-family zoning is receiving a kind of scrutiny that it hasn't in decades. For something that has shaped our cities so fundamentally, its role is almost invisible to many urban residents. "Of course my neighborhood doesn't have apartments in it. What kind of radical social experiment is this, telling me we have to allow apartments everywhere?!"
Places used to evolve and experience ongoing redevelopment. BuildZoom demonstrates this with their series of maps on the growth of "no-build zones" over time. The concept of a neighborhood which looks fundamentally the same in 2019 as it did in 1999 as it did in 1979, in a city that's growing, is a historical anomaly.
We need to get back to cities that are free to evolve and breathe. But we need to do it with attention to local context, and with a great deal of humility about the unintended consequences of well-intended policies, not by obsessing over the number of homes we need to build by any means possible.