Originally posted on Streets.MN.
Everybody knows the Minneapolis-St. Paul (MSP) metropolitan area is building new housing. It’s been evident for years in the disappearing parking lots of the Mill District and the North Loop, the new windows on old Lowertown warehouses, and the profusion of mid-rises that increasingly characterize Uptown and Dinkytown.
But almost nobody has the free time and the interest to roam around and get a good grasp on the full scope of what’s happening in the city. Residential development in particular can mislead. Housing is an issue that we all experience as both participants and observers. New high-rises can convey an impression of rapid growth, renovations in existing buildings can obscure it. Some neighborhoods may boom, while others remain unchanged. An eye-popping rental listing might suggest that prices are skyrocketing, a deteriorating old duplex might suggest the opposite. Depending on what you see on your daily routine, you might come away with a completely different impression of what’s going on than someone else. This is why data is especially important to clarify the picture.
Over six years ago, before I moved away from the Twin Cities to study planning in Philadelphia, I began keeping track of the work of the Minneapolis Planning Commission. I wanted to get a more thorough sense of what was happening in terms of new development in the totality of the city.
At the same time, I’ve also begun collecting housing starts data from the United States Department of Housing and Urban Development. You can find this data through HUD’s charmless but functional data portal located here. This data tracks the amount of new housing units by the date that they break ground, and it’s easy to find annual data going back as far as 1980. I’ve compiled this data in a spreadsheet for the Minneapolis-St. Paul metropolitan area and its two core cities here.
With ten years of detailed data on what got approved in the urban core and over forty years of larger scale data on what actually got built, It’s as good a time as any to step back and look at what the numbers tell us about what’s actually going on. How has the landscape of housing in the Twin Cities shifted over the past decade, where might it be heading in the future, and what does that tell us about the overall health and direction of the region?
Today, I’ve focused on what the metro housing starts data can tell us. Tomorrow, in a second post, I’ll zoom in and share some results from the Minneapolis approvals data.
1: The MSP Metro is back to building (almost) pre-recession levels of housing.
Total housing starts in the MSP Metro (1980-2020).
The story of Minneapolis-St. Paul’s housing production in the last four decades is a story of three valleys and three peaks.
Starting in the early 1980’s, the economy recovered rapidly after a decade of inflation and recession. In the mid-80’s housing production peaked, then fell just as swiftly into a trough that persisted until a low-point in 1992. From that point, housing production grew in cycles, greased by cheap credit for homebuyers, finally culminating in the 2006-2009 crash.
The 2009 nadir of just 4,036 housing starts was just over 15% of the region’s 1986 peak of 25,518, and probably the region’s worst year for housing production since World War II.
The 2010’s were characterized by a halting recovery (with an interesting mini-bump in 2012-2013, more on this below), followed by a rapid rise to almost pre-recession levels. While it’s unclear what kind of market might follow the 2020-2021 COVID-19 pandemic, for now 2019 marks the peak of this era of housing construction. However, there’s some reason to think that unmet demand might push housing production higher in the post-COVID recovery.
During this forty-one year period, the region’s actual population growth has been fairly steady. Every decade in this period, the region added between 315,000 and 440,000 people. Despite the ups and downs of the business cycle, housing production over time had also been mostly steady, averaging between 140,000 and 180,000 housing starts per decade. However, the most recent decade, which was impacted by the long tail of the housing market crash, meaningfully underperformed. From 2010 to 2019, just under 110,000 total homes started construction, even as migration stayed relatively constant. This contributed to overall housing scarcity and helped to drive home prices and rents up.
2: The two core cities are building more housing than they have in a long, long, long time.
Total housing starts in Minneapolis and St. Paul combined (1980-2020).
Total housing starts for the MSP metropolitan region are shown in the background.
Within the overall housing numbers are shifts in where and what type of housing is being constructed. While the MSP metro as a whole has a series of peaks and valleys that over this four-decade stretch have sloped downward, the opposite is true for the two core cities of Minneapolis and St. Paul. When looking at combined housing starts in these two municipalities, the trend is of a series of peaks and valleys that over time have sloped upward.
While housing construction in the two core cities boomed along with the metro in the 1980’s, it did so in much lower proportions. When the market softened, developers abandoned Minneapolis and St. Paul almost entirely. In 1992, a year after what we now know to be the peak of America’s urban crime wave, just 131 homes started construction in the two cities combined. In 2020, twenty-two individual projects with higher unit counts were proposed in Minneapolis alone. The ten-year stretch from 1989 to 1998 saw just 3,967 homes break ground in both cities combined, the lowest of any period in the data.
As the housing market started to take off in the 2000’s, building returned to the core cities, and in 2004 new housing starts nearly equaled the total from that ten-year low. But the success wasn’t sustained, and housing starts slowed, then crashed again with the rest of the economy.
The true headline of this analysis is what came next. In the past decade, the two core cities have blown away their previous housing numbers, and the trend has only increased. 2019’s peak of 6,388 housing starts in Minneapolis and St. Paul combined is the highest number by far in this data, and probably the highest number even further back, at least since the cities’ population crested in the 1950’s.
Despite some predictions to the contrary, residential building has dipped only slightly during the global pandemic. The core Twin Cities have not seen the scale of disruption to the housing market that has been seen elsewhere, especially in the most expensive, supply-constrained coastal cities. Remarkably, in 2020, St. Paul (Population: 304,547) started more units of housing than San Francisco (Population: 874,961). In 2020, Minneapolis and St. Paul combined (Population: 724,871) started more housing than San Diego (Population: 1,423,852), Manhattan (Population: 1,628,706), or San Francisco, Oakland, and San Jose combined (Population: 2,321,853).
3: Starting in 2012, there was a major and sustained shift towards urban housing.
The percentage of housing starts in the MSP metro in multi-family buildings and in the core cities (1980-2020).
The dotted lines show the average up to 2011, and the average from 2012 onwards.
In the two graphs above, the year 2012 stands out. Housing production surged in that year and in 2013, before falling back slightly in 2014 and 2015. Student housing led this small boom. As the broader economy recovered slowly, developers pivoted towards the University area, constructing over 2,000 units of housing in the Marcy-Holmes and Stadium Village neighborhoods alone.
But a funny thing happened after 2012. Even as developers shifted more broadly into building housing for older adults, they stayed in the city and they stayed with multi-family, to a degree that has likely not been seen since the middle of the last century or earlier.
From 1985 to 2001, the percentage of housing units built in the MSP metro that were located in the two core cities never rose above single digits, and in 1992 it was just 0.81%. From 2001 to 2011, there was a uptick into the teens, with the two core cities averaging 11.3% in that period.
But almost overnight, starting from 2012, that percentage doubled. From 2012 to 2020, the two core cities were the location of 22.9% of the metro’s housing starts.
The trend of densification also meant more multi-family buildings, in both the cities and the suburbs. In the 1980’s the percentage of housing starts in the MSP metro that were in multi-family buildings hovered around 40%. It fell precipitously to a low of just under 12% in 1991, and stayed in the teens for much of the 1990’s before rising again. From 2001 to 2011, the percentage of housing units built in the MSP metro in multi-family buildings averaged 31.4%, an increase given recent history, but well below historic norms.
Then came the shift. From 2012 to 2020, that percentage jumped to 54.1%. In this eight-year period, over half of housing starts in the metro area came in multi-family buildings. In 2019 and 2020, the percentage was over 60%.
The Twin Cities metro region has now experienced almost a decade of housing development that has been denser and more centrally located than at any other period in at least the past forty years, and quite likely longer than that. These trends have held steady during this period, although there is considerable uncertainty about what will happen post-pandemic.
4: Most of the recently-built housing in two core cities has been located in Minneapolis—but St. Paul is, at long last, starting to make an impact.
Housing starts in both Minneapolis and St. Paul shown cumulatively (2001-2020). The total for Minneapolis builds on top of the total for St. Paul.
Focusing in on the most-recent twenty years of development in the two core cities, it’s easy to see the spike in housing construction that began in 2012 and has continued since. It’s also easy to see that Minneapolis has shouldered far more of the weight than St. Paul. Just once in any of the years for which I can find data, in 2004, did the state’s capital city out-build its larger neighbor.
The post-recession housing boom has kept to that trend, but finally in the last two years has St. Paul started to pull its weight. Housing starts rose above 1,500 in 2019, and landed at 2,077 in 2020, just shy of that earlier anomalous high mark.
Take a tour of St. Paul, as local newspaperman Fred Melo did recently, and you can find a number of projects underway. Portions of University Avenue are finally seeing investment, new construction continues to creep southwest on West 7th, the Snelling and Marshall corridors are getting the odd project, and downtown continues to slowly add housing density. Nobody will mistake St. Paul for a boomtown just yet, but there’s a lot of evidence of life.
There’s also probably the opportunity to see more in the future—especially if St. Paul were to take a look at its permitting and inspections process, which just about everyone seems to believe is exceptionally cumbersome. Eliminating parking minimums would also help significantly.
But development in the city is also still contentious, in a way that is not the case next door. A recent proposal to build homes for lower-middle class tenants on a vacant lot near the Green Line was denied by the city’s planning commission in a narrow vote, and its fate seems likely to be ultimately decided by the city council or a court. The next few years will be ones to watch in St. Paul.
5: Small multi-family buildings are becoming more prevalent, at least in Minneapolis, but not yet in meaningful numbers, and not at all in St. Paul.
Cumulative housing starts in 2-4 unit buildings (2001-2020). The total for Minneapolis builds on top of the total for St. Paul.
If asked to sum up the last few years of the housing conversation in the Twin Cities, many people might settle on “triplexes” or “fourplexes.” That’s of course a result of the stormy debates over the Minneapolis 2040 plan, which ultimately legalized the possibility of building the former on every lot in the city. But to paraphrase one-time planning commissioner Nick Magrino, as dramatic as this change might have been on paper, the reality was always going to be quite incremental.
There’s no getting away from the small amounts of housing being built in small multi-family buildings. In the three years of 2018, 2019, and 2020, nearly as many units were built in duplexes, triplexes, and fourplexes in both Minneapolis and St. Paul than in the previous ten years (2008-2017) combined. But the overall number of homes that we’re talking about (160 to 162) is regularly equaled or surpassed by individua; mid-rise apartment projects.
It is also clear that a construction boom of small multi-family buildings began before the 2040 Plan was passed in 2018. In 2020, the first year of citywide triplex zoning in Minneapolis, the number of newly-built units in small multi-family buildings increased, but only very slightly. The most ballyhooed piece of the city’s zoning overhaul has yet to make a discernable impact in the data.
It’s also notable that Minneapolis and (especially) St. Paul have yet to even match the amount of small multi-family building construction that both cities saw in the early aughts. The core cities’ 2003 peak of 115 homes built in small multi-family homes has been challenged in recent years.
It’s clear that there remain obstacles for small multi-family residential construction. Renters may prefer to live in larger buildings with more amenities and there are few development firms focusing on this market. Unfortunately, neither city appears poised to limit existing regulatory barriers beyond what they’ve already done. Minneapolis city councilmembers recently rejected proposed rules that would’ve made it much easier to build these types of small multi-family buildings (and smaller single-family homes as well), instead substituting a far-less ambitious alternative that does little to change the status quo.
***
Coming tomorrow: a deeper dive into every project approved in the past decade by the Minneapolis City Planning Commission.