Population Forecasts and Housing Allocations Don't Work; We Should Replace Them With Vacancy Targets

Growth in national vehicle-miles traveled started leveling off around 2004, but it took at least 6 years for the Federal Highway Administration to adjust to the new reality and begin reducing its projections for future growth (which were still too optimistic). Source: Frontier Group.

Population forecasts, like virtually all forecasts, are bunk. There are just too many variables and unknowns to predict something like population growth reliably. As with projections of other future trends (such as vehicle-miles traveled; see above), they're made up, fantasy-land numbers that usually just assume that what's happened in the past will continue to happen in the future, indefinitely. As a result, they're often very wrong.

In LA's General Plan Framework Element (adopted 20 years ago), for example, the projected city population was 4.3 million people  by 2010. We actually just hit 4 million in the last year, and were at barely 3.8 million in 2010, so they overshot it by about 500,000 people. At the same time, they assumed that just 5 percent of that growth would occur in Central LA, which includes the area roughly from downtown to Westlake. In fact, since 1999, downtown alone accounted for 20 percent of the city's growth.

This wouldn't be a big deal except that forecasts really matter for how the city is planned. They play a major role in how community plans and specific plans are developed, and how housing targets are allocated throughout the city and the region. We forecast, imperfectly, and then we update zoning for corridors and neighborhoods to accommodate and guide that imagined future, creating strict standards and regulations for how new development should look, feel, and function.

And then the future actually happens, and its nothing like we thought. Suddenly downtowns are hip, everyone has an iPhone, the Great Recession happened, Uber is here and driverless cars are on their way, and, understandably, neighborhoods are completely unprepared to respond flexibly to the new demands being placed upon them. Since our plans don't match up with the reality we expected, we end up with general plan amendments, new height districts, and zoning changes. And then people freak out.

The inability to accurately forecast the time, place, and nature of growth is not a failure on the part of planners; changes like the success of the Adaptive Reuse Ordinance, or the resurgence of downtowns across the country, simply defy reliable prediction. The real failure is in believing that we can—or should—try to predict exactly how our city will grow and change, creating plans that fail to accommodate any other possible future. Those forecast-based plans stand in the way of elegant, responsive growth, and they've played a meaningful role in the city's ongoing housing affordability crisis.

So we should do away with housing and population forecasts and replace them with vacancy targets.

As I've written in the past, our affordability crisis comes down to a shortage of housing, and that shortage is most clearly evident in LA's historically-low residential vacancy rate of about 3 percent. When vacancies are low (under about 5 percent), landlords wield incredible leverage over their tenants, and rents climb as a result. We need to take vacancy rates as seriously as unemployment rates and push for immediate action when they veer into unwelcome territory.

To allow that, the only significant planning guidance on housing production should be a target vacancy rate—something that would vary between about 4 and 7 percent depending on the neighborhood. Downtown would probably target a vacancy rate in the upper end of that range; somewhere like Silver Lake would probably target something at the lower end (e.g.). Free of the micromanagement of individual parcels, planners could then focus on planning for the public space that we all share

Beyond setting a general framework for new development—such as encouraging development near transit and other local amenities—the city could leave the how of achieving that vacancy rate up to the community. You want to concentrate growth around a few major transit hubs and leave most of your neighborhood untouched? Feel free. Prefer to spread it around a bit? No problem, so long as there's enough capacity to meet your target. If vacancies are falling short, you'll need to do more; if they're staying within range, you can rest easy. 

With vacancy rates kept at equilibrium levels, rents would tend to stabilize in neighborhoods throughout the city, and with appropriate incentives those with new development would create plenty of new rent-stabilized and affordable housing to replace what was lost. (Neighborhood-driven zoning changes should of course include value capture mechanisms that create affordable housing rather than simply enriching existing property owners.) 

Community members opposed to any change in their neighborhood might not like this system any better than the current one, but at least they would have more of a voice in the shape of that change, if not the overall trajectory. Vacancy rates are comprehensible; it's readily apparent to the reasonable observer why it's better to have 2 people competing for every household than two dozen. Planning a neighborhood with a vacancy target in mind makes sense, whereas population and housing unit targets seem arbitrary—because they are.

With all the existing rules surrounding the updating of our General Plan and Community Plans, this would no doubt be a dramatic change, most likely requiring a massive overhaul of the planning process. But I think it's a change worth fighting for in the long run. The king is dead; long live the king!