A Reminder for Planners: "Every Projection is Wrong"

Brent Toderian has a bone to pick with planners. And generally, when this is the case, planners sit up and listen, because Toderian is one of the profession's leading lights—he's the former planning director for the city of Vancouver, founding president of the Council for Canadian Urbanism (a north-of-the-border analogue to CNU), and one of Planetizen's 100 Most Influential Urbanists.

Toderian recently tweeted an excellent thread about the profession's overreliance on forecasts and mathematical models, and the dangers this poses. Here's the text of it:

We put too much weight on projections in city planning — population projections, economic projections, traffic projections. I’ve analyzed & created projections my whole career, & I know one thing — EVERY projection is wrong. We just don’t know by how much & in what direction. 

Projections & models too often use the past to predict the future, & that’s increasingly questionable. They ALL are based on assumptions, which are often wrong. And even if projections or models ARE “right," it's often a self-fulfilling prophesy [sic] resulting from the model itself.

Your traffic model predicts more driving. You build new roads. More people drive. But would they have driven more without new roads? Your school model predicts few students downtown. You don’t build schools/daycare. Families avoid living downtown. Self-fulfilling prophesies.

The best transportation plan is a great land-use plan. But too many cities still let flawed traffic modelling actually determine or limit land-use, resulting in what I call the “sweet spot of failure” — enough density to cause local traffic problems, but not enough to solve them.

"The traffic model is the slave, not the master." -Steven Burgess. Models are frequently full of faulty assumptions, and rarely understand actual human behavior. Never let a traffic model tell you what to do. Decide what you WANT to happen, & Design how to get there. #multimodal

Traffic models & population projections need the same rethink — a shift from a "predict & provide" approach where we guess what WILL happen, to a "debate & decide" approach where we CHOOSE what we WANT to happen, & assess our choice against many future scenarios for resiliency.

These are wise words. Projections can severely limit our thinking about what the future could look like, by imposing a powerful status-quo bias. In fact, this is often their explicit purpose: by telling us we will need to do more in the future of what we've done in the past, and creating a sense of urgency and inevitability around this, they serve those who benefit from the status quo.

There's no better example of this than the propaganda that emanates from the Infrastructure Cult of organizations like ASCE each year. Congestion is always bad and getting worse. We're never spending enough on infrastructure, and year after year "experts" offer up eye-popping dollar amounts that we're told we the people "will need" to spend to set things right. The math and assumptions underlying these kinds of reports are mostly smoke and mirrors, as we've observed and discussed many different times.

Can We Choose Our Own Destiny?

I would go further than Toderian goes in his thread, in one crucial way, though. I would question "our" ability to, in fact, "choose what we want to happen." I'm inferring that by "us" here, Toderian means the public acting through their elected representatives in (particularly local) government. 

Projections are not necessary when things are built incrementally with ongoing feedback driving adaptation.
— Charles Marohn

I don't think for a second that Toderian himself believes this, but I encounter many people who believe that the purpose of planning is to script the future and bring it about in orderly, comfortable ways. That once We the People decide, say, how much population growth we wish our community to experience, or where its new homes and jobs should be located, or how people should travel within the community, then it's just a matter of waving a magic policy wand and making it so.

Yet if projections of what will happen if we don't alter the status quo are fatally flawed, then it stands to reason that projections of what will happen after we do alter the status quo are at least as flawed, and probably more so.

Cities are complex systems. They will always defy our best-laid schemes, our Vision 2040 or Blueprint for 2050. Everything we do in them launches a domino cascade of unanticipated consequences. This is at the core of why projections are so problematic to begin with.

A central goal of planning ought to be to make our human habitat more antifragile, so that we don't need to be right about the future. This is about having resilient institutions with a lot of redundancy and many sources of bottom-up feedback, not about having the right inputs and coefficients in your Excel model.

In Chapter 4 of Strong Towns: A Bottom-Up Revolution to Rebuild American Prosperity, Strong Towns founder and president Charles Marohn writes, "Projections are not necessary... when things are built incrementally with ongoing feedback driving adaptation.” And so a better description of what "we" get to do is that, while we can’t dictate our preferred outcomes, we do get to decide our values and priorities. And then we should implement them, with humility, in this simple but infinitely repeatably set of incremental steps:

 
 

Cover photo by Isaac Smith on Unsplash