Laura Bliss is a staff writer at CityLab, covering transportation and the environment. She also authors MapLab, a biweekly newsletter about maps (subscribe here). Her work has appeared in the New York Times, The Atlantic, Los Angeles magazine, and beyond.
New Census numbers show big population gains in America's big cities. According to our analysis, new housing in some cities may be finally making a dent in years of pent-up demand.
America’s fastest-growing cities are concentrated in the Sunbelt, according to the latest Census numbers, thanks to their (relative) abundance of affordable housing and jobs. The Texas tea boomtowns of Monroe, Frisco, and McKinney all make the top 10 list, as do Greenville, South Carolina and Bonita Springs, Florida. A dusting of business-oriented, non-Southern locales, including Bend, Oregon, also rank.
But many of the largest absolute increases in population hit bigger cities, with bigger problems. While housing-abundant Phoenix topped the list for the largest numeric gain in new bodies, Los Angeles, Seattle, Austin, and New York City weren’t far behind. These are towns with severe housing shortages, and many of their mayors have made pledges to increase stock.
But have the number of units increased much? It appears that housing gains are, in fact, picking up pace in many big cities. But developers are only beginning to respond to years of pent-up demand. Don’t count on finding an affordable place in these desirable villes tomorrow.
The “dwellability” crisis might be softening
I set out to compare population increases over the past six years in the ten top-gaining U.S. cities to how many people new housing units could accommodate in these places. This second metric is an original invention; I call it “new dwellability.” It is an estimate that could probably use some finessing. (Housing-wonk readers! I’m all ears.)
To calculate it, I took the number of added housing units in each city (not the entire metropolitan area!) for every year going back to 2011. I didn’t account for demolitions, reconstructions, or multi-family units versus single-family units; I simply measured the difference between total housing stock from one year to the next. Then I multiplied that number by the average household size for each city, which ranges between 2 and 3 persons per unit. (Seattle is surprisingly roomy, with an average 2.06 individuals per unit. Los Angeles is one of the most dense, at 2.8.)
Why not just show the number of units? Simply saying that San Antonio gained 29,000 people but only 3,600 housing units in 2014 certainly gives an impression of a severe shortage. But it’s more useful to think about how many bodies those new units are likely to accommodate (in San Antonio’s case, about 8,000), because for every new housing unit, at least a couple of people get a place to live.
How to read these charts
In the charts below, this “new dwellability” number is shown in red bars. Population gains are shown in blue. As you look at these graphics, bear a few things in mind:
- The blue bars show annual population gains, not total city population.
- The red bars show annual change in housing capacity, not total housing capacity.
- The Census currently has good data on housing numbers for the years 2010 to 2015, so that’s what I looked at here. An even better analysis would include data back to 2000, pre-Recession. Though I included the newly available 2016 population gains, exact housing numbers are not available for 2016 yet. So in this analysis, last year’s dwellability gains are an open question.
- Most of all, the charts do not show supply and demand! They show the relationship between dwellability and population gains, from 2011 to 2016. Demand is piling up every year a city gains new bodies. So when you think about how housing is responding in these markets, remember that it needs to respond to that built-up need, plus whoever’s moving in next year.
Below, check out how the ten U.S. cities with the largest numeric increases in population from 2015 to 2016 (in order of gains) are faring with respect to housing.
The biggest numeric gain in population hit Arizona’s sprawling capital, with 32,113 individuals setting roots inside city limits. Phoenix’s housing situation is not exactly pressed; its role in the pre-Recession housing boom has left it with plenty of excess stock. In fact, it may have lost units in 2013 and 2014, and added just a scant 60 units in 2015, which barely show up on this chart as 167 dwellable spots. (Those numbers are well within the margin of error, though— could just be noisy data.)
2. Los Angeles
My hometown has become an unlikely poster child for the urban affordability crisis sweeping the country. Tens of thousands of new residents are flooding the city every year, and housing additions have been uneven. Dwellability has increased every year since 2013, though—if that trend line continues, there could be enough new units at least for the new folks coming in every year. That doesn’t mean the end of the housing shortage, though: There are still years of latent demand to account for.
3. San Antonio
Though it slowed considerably in 2016, San Antonio’s growth has been explosive in recent years, and it’s begun to grapple with an affordability crisis. From the looks of it, housing has not not begun to catch up.
4. New York City
New construction in the Big Apple boomed in 2014, thanks to a wave of new luxury developments. That settled down in 2015, but building permits also surged that year. The numbers aren’t in the Census for 2016, but other sources show that the city was actually able to build or preserve an impressive 22,000 affordable apartments that year. Meanwhile, New York’s population growth has slowed way, way down—it’s pretty alarming to see it charted so clearly here. Rents will never be cheap in NYC, but they, too, are falling faster than most of the other primo markets in the U.S. For once, America’s biggest city resembles something of a tenant’s market.
Measuring the rate of growth for the cities on this list, Seattle was the second fastest growing big city after Phoenix. Unlike its sunnier counterpart, Seattle is grappling with a major housing crisis as it gains tons of new residents. There are positive signs, though: Seattle increased its dwellability considerably in 2015 through a massive rental boom. Reports suggest that the trend line will continue in 2016 and 2017. Even if many of these new homes are priced for luxury tenants initially, that stock should eventually “filter down” over time to meet the needs of lower-income renters. It’ll take years, but there could be hope for Seattle-ites.
Below, check out the next five top-gaining U.S. cities are doing.
Dallas’ population growth has slowed down, but dwellability ramped up in 2015. Home prices are on a steady upwards trajectory.
7. Fort Worth
Fort Worth’s population gains track closely with Dallas’, and its dwellability trend line registers a steady increase since 2013.
Houston saw a big dip in new residents in 2016, but a huge spike in dwellability. Is it enough to abate the city’s quiet affordability crisis? Maybe if construction continues at this pace.
Texas’ most expensive housing market saw new dwellability ratchet up in 2013, but it’s declined since. Austin is struggling with a severe shortage of affordable housing, but at least these numbers show a closer-than-average match between new residents and new capacity.
10. San Diego
Poor San Diego: when it comes to creating housing, America’s finest city it ain’t. The city seems to have shed units in 2012 (though that loss is well within the margin of error, so this could just be noise), and though it’s been adding unevenly since then, it’s still losing more affordable homes than it’s gaining. Meanwhile, new San Diegans keep pouring in.
So what’s our forecast? Many of America’s big expensive cities are seeing a rent slowdown, and if some of these reassuring gains in housing capacity continue, perhaps that will continue. The CityLab dwellability index is an ongoing experiment in comparing housing growth to population growth, and we’ll keep refining it. Check out this spreadsheet to see how I got these numbers; and please chime in on how this experiment could work better for the next batch of population numbers.