Playing Catch-up: Putting the Recent Home Building "Boom" in Context

• Since the Great Recession, there is a cumulative shortage of 1.35 million new single-family homes permitted in the 35 largest U.S. housing markets. That is 2.7 years' worth of permits at today's homebuilding rate.

• Home builders are doing what they can to meet unrelenting housing demand during the pandemic, but there remains a large deficit to fill to catch up from years of underbuilding after the Great Recession.

• The biggest new-construction shortfalls relative to population growth since 2008 have been in Dallas, Miami, Phoenix and Seattle.

The number of homes listed for sale in America ticked down in October, marking the end of 2021's modest inventory rebound that began in May, and leaving current levels down by more than a third from pre-pandemic norms. But how did we get here? Why are there so few homes available to buy, even as homebuilders seem to currently be firing on all cylinders to develop and sell more homes?

More than 1.5 million residential building permits were issued between February 2020 and February 2021 — a benchmark level of housing construction surpassed in every building boom since the 1970s. The problem is that after permit activity bottomed out in 2009 at the depths of the housing crisis, it took more than 11 years to get back to that threshold. The new home construction market today very much remains in catch-up mode, recovering from a decline that was both much steeper and took longer to come out of than in any previous construction cycle.

Total residential permit activity has continued to grow dramatically throughout the past 18 months — hitting a new post-Great Recession peak of 1,687,300 permits issued in the 12 months from August 2020 through August 2021 — leading some to wonder if builders are actually creating a new glut of housing. But this year's pace of homebuilding represents only the first full year of above-trend construction in 11-plus years, and in truth builders are only just beginning to meet the housing deficit and feed housing demand that has only been growing since the end of the Great Recession.

Historic Context

In the 40 years from 1960 to 1999, there were about 1,371,000 housing permits issued on average in a given 12-month period. Building followed a clear cyclical pattern, rising and falling around that average, but the duration of cycles stretched out to longer periods as time went on (similar to the duration of economic recessions and expansions in recent years). Prior to the U.S. onset of the pandemic in earnest in early 2020, the nation had just come out of the longest permitting "bust" period on record — on the heels of the longest boom on record, each almost 12 years in length.

The expansion of home construction in the early 2000s was unprecedented in its duration, running above the pre-2000 average construction volume for 141 months: April 1996 to December 2007. And the bust followed the boom: 142 consecutive months (January 2008 to October 2019) in which construction activity lagged the 1960-1999 average. But while almost identical in duration, the boom and bust were not equal in magnitude: After 2007, the shortfall below average (about 4,467,000 permits) exceeded the net above-trend activity before 2007 (4,059,000) by about 408,000 permits.

In the 23 months since overall permitting activity first broke above pre-2000 averages, about 355,200 total residential permits have been issued above that baseline pace. In other words, even after almost two years of robust activity, home builders still are working through a net permitting shortfall (355,200 permits issued above trend the past few years, versus a shortfall of 408,000 from the past boom/bust cycle). In some ways, the market has not yet even begun a new boom cycle, so much as only just getting back to par following the last bust.

But isn't population growth slowing down?

Admittedly, it's not clear if that 12-month average number of permits in the 40 years from 1960-1999 (1,371,000) should represent the ideal annual flow of new construction — it's only a useful benchmark for comparisons' sake. If we assume most new housing construction is geared toward meeting new housing demand from a growing population, it might be more useful to compare the pace of annual permitting to the pace of annual population growth. The United States is much more populous today and has a much bigger housing stock than in the 1960s or 1970s. Yes, population growth is currently slower in percentage terms than it was back then, but a larger population base will counteract that — in raw numbers of people, population growth has not changed dramatically.

Year-over-year population growth over this time frame essentially peaked in the early 1990s, according to Census data, before falling in aggregate through the late 2010s and hitting a recent bottom after 2020. We should also note that data from the early-period pandemic months were affected by some (hopefully temporary) shocks to all components of population growth: birth rates, death rates and net immigration.

Comparing homebuilding with population growth, measured as permits per extra person computed over the trailing 12 months, we see a similar boom-bust cycle in the high-frequency data. This measure reveals that, relative to population growth, the post-1996 boom-bust cycle happened at an even lower magnitude than the cycles of the 1960s and 1970s.

Because houses generally stand much longer than 20 years, and people do not need an entire extra home the day they are born, it also makes sense to measure these changes over a longer timeframe. Comparing the 20-year total volume of homebuilding with the 20-year growth of population shows a much smoother trend and suggests that the 20 years ending around 1990 represented a period of relative housing abundance compared to population growth. But even after the early 2000s boom, the 20 years ending around 2010 constituted a phase of relative housing scarcity. Probably not coincidentally, the pace of annual home price appreciation following 1990 was relatively slow, while for most of the 2010s it was quite rapid.

Total housing construction over the past 20 years, spanning the most-recent boom, bust and restarting period, is approaching the cumulative home building per capita achieved in the twenty years ending in 1979. But the average household size has shrunk, from 2.78 people in 1979 to 2.53 today, meaning all else equal the same number of people will occupy more homes — any given population size now will fill about 10% more homes than in 1979.[1]

Location, Location

We can quantify the single-family[2] permit shortfall at the local level by analyzing how many single-family permits each metro area would have issued between 2008 and 2020 if its local permitting pace relative to local population growth matched the 1985-2000 national average.[3] By this measure, the largest current shortfalls are in the Dallas, Miami, Phoenix, Seattle, Washington, D.C., and San Francisco metros. Most of the metros with "excess" single-family permits are the ones with modest population growth in this timeframe, including Chicago and Cincinnati. Even a region without population growth would be expected to issue some building permits as people replace or upgrade older homes, but, to be conservative in estimating shortfalls, this analysis does not account for that baseline “maintenance” level of construction.

By this measure, the cumulative housing permit shortfall across the top 35 metropolitan areas totals 1.35 million missing single-family home permits, even accounting for the building "surplus" in places like Chicago. That represents 2.7 years' worth of permits, even at the elevated building rate achieve in the 12 months through August 2021.

The implications of this shortfall are coming home to roost now, as home prices surged across the country throughout the pandemic. The demographic makeup of the U.S. and the ability for many to work remotely have boosted housing demand. The shortage of homes in the most heavily populated markets has left home shoppers competing for a limited supply, pushing up prices. Since the Great Recession, single-family residences have appreciated much more (up 47.9% since January 2008) than other types of homes, such as condos and co-ops (up 29.4% over the same period). The depth of the construction shortfall means we are still, at minimum, years away from a balanced housing market, much less one with abundant housing.

[1] Census Bureau, Historical Households Table 4, accessed November 12, 2021 at https://www.census.gov/data/tables/time-series/demo/families/households.html.

[2] While total residential building permit data — both single-family and multifamily homes — was used for the national calculations in this analysis, metro-level data only includes single-family permits. Historical single-family permit data at the metro level is available for longer timeframes than total permits data.

[3] An alternative measure would be to compare each metro area's post-recession permitting to its own local permits per population growth pre-2000, and then estimating the post-Great Recession shortfall of building permits by comparing the actual number issued with the number that would have been issued (given the observed local population growth from 2008-2020) if permits had matched that local pre-2000 permitting rate. This approach, similar to earlier Zillow research on the topic, ultimately delivers a similar conclusion: There is a notable shortfall of issued housing permits relative to population growth in a majority of the nation's large metro areas. Ultimately, we decided not to go with this alternate approach because it could have seemed to unfairly "punish" metro areas — Indianapolis, for example — that achieved high local permitting rates in 1985-2000 but then fell behind that pace in the post-Great Recession period. At the other extreme, it would "reward" localities that issued very few permits before 2000.

The post Playing Catch-up: Putting the Recent Home Building "Boom" in Context appeared first on Zillow Research.