One central tenet of the Strong Towns approach is incrementalism. Specifically, we’ve suggested the following as one of six key features of our approach to local development and infrastructure:

“Relies on small, incremental investments (little bets) instead of large, transformative projects.”

The Oxford English Dictionary defines incrementalism as "belief in or advocacy of change by degrees; gradualism.” Merriam-Webster’s definition is similar: "a policy or advocacy of a policy of political or social change by degrees."

The word "incremental" is troublesome because, while we've found it's often the best term for the thing we are trying to say, it also has some connotations that aren't really what we're trying to say. Specifically, I've learned over time that some people see the word "incremental" and read into it things like slow, timid or change-averse. For example, in a city with a dire housing shortage, many will tell you we must come to grips with the fact that a large building or development project can provide many times more new homes than a small one (and the status quo is that those large buildings do account for a very high percentage of new housing built). In a city with inadequate public transit, critics will observe that a large, complex rail project can meaningfully expand transit’s appeal to those who don’t already ride it in a way that tweaks to bus schedules and routes won’t.

But these critics of incrementalism often misunderstand it. They are comparing the impact of one big project to that of one or a few small projects. The correct comparison is something more like one big project versus one thousand small projects. And if your objection is that we don’t have the coordinating capacity, time or resources to make those thousand projects happen, then that should get you thinking about what would successfully unleash those thousand small bets.

The answer is a rethinking of our whole approach to planning and development, and a relinquishing of control by authorities. Your job can no longer be to orchestrate change, but to create the conditions for it to flourish organically. When we get that right, the incremental approach is often the only one that actually is up to the challenge of a rapidly evolving problem.

It's the Size of the Step, Not the Size of the Change

A process can be incremental, in the sense of proceeding by many small steps, whether it is rapid or slow. A process can also be non-incremental, in the sense of proceeding by a smaller number of large, consequential decisions, whether it is rapid or slow.

For the mathematically minded, here's what that looks like, graphed, using the example of the rate at which new homes are built in a city:

What does the Rapid Incremental Growth pattern look like in terms of housing production? It looks like an army of small-scale developers doing relatively small (for their context) projects on relatively small (for their context) pieces of land, all over the place. It looks like things such as 20 percent of all housing permits in Los Angeles in 2018 being for backyard accessory dwelling units (ADUs). That's a huge number of people independently making the same small decision, enough to add up to a significant result.

Decentralized and incremental housing production happens in ways that are less obvious, too, such as the under-the-table subdivision of homes where official channels don't allow this or make it difficult. For example, former New York City planner (and frequent Strong Towns contributor) Nolan Gray observes that 73% of Queens housing production in the 1990s was from this kind of subdivision of existing homes—far more than the highly visible large-scale new construction going on in places like Long Island City.

What does the Rapid but Non-Incremental Growth pattern look like in terms of housing production? It looks like the latest Lennar or Del Webb master-planned community. It looks like The Villages. Nearly every fast-growing Southern metro area in the U.S. in recent decades fits this pattern: Orlando, Las Vegas, Austin, Houston, Charlotte—in all of these, most of the new housing has come in the form of large subdivisions erected by industrial-scale national homebuilders. (There are relatively few examples in North America of high-rise urban housing driving comparably rapid population growth, but Seattle in recent years (specifically centered on the South Lake Union neighborhood) might be one, and this is a common enough pattern outside North America.) This model works as long as the business model of those builders is favorable, but conditions can change for them very rapidly and growth can screech to a halt.

Incremental Growth Hasn’t Meant Slow, Historically

In theory there is no reason growth by small increments can't be rapid—you just need a lot of small increments, quickly. But surely in practice, the barriers to this kind of growth are too big?

Chicago grew from 30,000 to over 1 million inhabitants during a four-decade period of dazzlingly fast expansion. More to the point, this expansion was carried out mostly by a vast number of small builders and developers working at (individually) small scales.

This hasn't always been true: Even if you constrain your lens to America, you need only look at the spectacular growth rates of 19th century cities, which happened almost entirely through small-scale development—just a huge volume of it. In 1850, Chicago was a prairie settlement of 30,000. By 1890, it was a metropolis of over 1 million inhabitants. No 20th- or 21st-century population boom, not even those of the fastest-growing cities today like Austin, can match this rate, or the similar more-than-doubling per decade of Cincinnati (1810–1840), St. Louis (1840–1870), or Los Angeles (1890–1930).

There are lots of structural and basic demographic reasons that kind of explosive growth of cities doesn’t happen today, and lots of reasons we might not want it to—I’m not saying rapid growth itself is inherently good. My only point is that if your argument is that incremental development cannot power a building boom as rapid as any happening today, history says otherwise.

The explosive growth that gave us the row-house neighborhoods and courtyard apartments of 19th and early 20th century America happened in the absence of a lot of restrictive bureaucracy and formal planning. It was comparatively easy back then to buy a small plot of land and build what you wanted on it; today, you might contend with expensive permit requirements, delays of months or years, and opposition from neighbors, and all of these things drive up the cost and uncertainty.

That's not a nail in the coffin of small-scale development. That's an argument for actually attacking the things that create untenable fixed costs for small-scale development, and creating a regulatory environment where small projects that represent modest change can sail through with minimal hassle, while large projects with high stakes for the community receive an appropriate level of scrutiny.

This is important to pursue, because the development regime we have—one with a lot of bureaucratic inertia and complicated rules and procedures—benefits the biggest developers the most, while still not producing enough homes. We’re talking about a system in which the developer of a large apartment building has to bribe a city with millions of dollars in extra public amenities, only to still face denial of their project. This is supposed to scale up to meet the challenge of the housing crisis in a place like Northern California?

Or take public transit. Multibillion dollar light-rail projects twenty years in the planning are supposed to be something we can do over and over again, enough times to get a large share of current U.S. drivers riding transit instead?

To those who say the decentralized, incremental approach is not up to the scale of our problems, I say there’s no evidence that the coordinated, master-planned approach is either. There’s plenty of evidence that it’s inadequate: not nimble or scalable or farsighted enough.

Incrementalism as a Bias Toward Action

The defining feature of incrementalism as a planning or policy approach is not slowness or timidity: quite the opposite. Incrementalism entails a bias toward quick action over exhaustive planning: you take the next, easiest action to address the immediate situation you’re facing, and you take it right now. You don't wait to have the whole road map to your policy goal laid out for you.

But the piece that comes along with this is that it means a bias toward constant, repeated small actions. You don't take one small step and call it a day.

In infrastructure, this means do things like improve the experience for transit users—speed, frequency, timeliness, conditions at the stops—but do it rapidly, do it with minimum fuss, do it in lots of places, and track the results. Then make another round of improvements and another, each informed by what you’ve learned.
In housing, it means make the smallest bar to entry low (something like converting a home to a duplex, as of right), so that it happens in lots of places without the need for coordination, permission, or valuable time spent on the part of planners and city officials.
In economic development, it means taking more seriously things like micro-lending and economic gardening, rather than spending years working on one big incentive deal to lure one big company to town.

Incrementalism in this sense would have city-builders—planners, civil engineers, developers, and so forth—thinking more like software engineers. In fact, incrementalism is a dominant paradigm in Silicon Valley for good reason.

The principle of the “minimum viable product” is an example of incrementalism as an important mindset in the tech industry.

Tech people will often talk about the "minimum viable product." We don't really know how our customers will react, what problems they'll encounter that we didn't anticipate, or how the world will have changed in the time spent developing something. You can spend years and millions of dollars perfecting a product according to your initial vision, only to have it flop when it goes on the market. So the best thing to do is to get something into their hands and see how they respond to it. And let that guide the changes you make for the next release, and the one after that. Free up your coders, too, to solve problems they are competent to solve without going through many layers of approval.

Software developers know that trying to predict the future is doomed to failure, because they work in a realm in which the world is changing incredibly quickly, and the only constant is that the future is unpredictable.

Policy makers, however, aren't any more prescient than software developers. In fact they are almost certainly less so, because their work deals with the full spectrum of human experience, not just the user experience with a product which occupies a particular niche in your life. The city is something everyone “uses” every day in dozens of different ways. So why not get the ability to make small, rapid, responsive changes into the hands of the “users” of a place?

Incrementalism is the preferred policy approach of those who aren't prescient and don't think they are. And it doesn’t have to be slow or timid at all: it can be viral, experimental, even dizzyingly fast and full of possibilities.

Creating Housing Opportunities in a Strong Town

Want to put the power of incrementalism to work where you live? Want to develop a plan of action you can implement right away? Our brand new housing course will show you how.

See the Course Description