Early in the COVID-19 pandemic, it became clear that people were abandoning pricey cities and taking refuge in more affordable neighborhoods. Why pay lofty downtown rents when all of the urban amenities like trendy restaurants, concert venues, theater districts were shuttered? Why live close to the office when you are no longer required to go there every day?
Newspapers last spring were filled with stories of empty downtowns as firms converted overnight to remote working, freeing office workers from the grind of their daily commutes. Not needing to report to the workplace, countless office workers, especially in tech and finance, moved to smaller communities where initial COVID infection rates were lower and housing more affordable.
The most expensive and densely populated cities got the most press attention, most famously New York and San Francisco. A Google news search of “New York pandemic exodus” returns 220,000 articles. “San Francisco pandemic exodus” finds another 50,000 hits. Some articles layered on the claim that people were fleeing not just expensive housing markets but also high taxes and onerous regulations found in coastal “blue” states and moving to “red” Sunbelt states with lower taxes and more business-friendly climates.
Commercial real estate market data seems to provide further confirmation of these trends. Apartment rents plunged at record paces in San Francisco and New York as vacancies soared, while rents were actually climbing in many small and medium-sized cities. Similarly, office rents fell sharply in many leading gateway markets as leasing tumbled and sublease space mounted. Vacancy rates soared by double-digits in San Francisco and New York during the pandemic while rising less than five percentage points in Atlanta and Dallas. (To be fair, however, vacancies in these sunbelt markets were, and remain, much higher than in the gateway markets.)
What is being reported isn’t always the reality. So how do we learn what is really happening to our cities’ populations? Did masses of people really abandon San Francisco, New York, and other leading metros in favor of greener and less expensive areas? Well, like any good story, it’s complicated.
First, a quick look at the common data sources being used to follow these trends. The gold standard for measuring migration is the annual survey from the U.S. Census Bureau. The survey is very detailed in terms of both geography (where people are moving to and from) and demographics (who is moving). However, this data is collected only once a year. Unfortunately, for assessing the impact of the COVID-19 pandemic on migration levels and patterns, people are surveyed in March of each year, and then the results are published nine months later in December. That means that in 2020, the data was collected only during the first month of the pandemic, and we didn’t see those results until last December. So, we won’t have the Census pandemic migration data until the end of this year. Thus, we need to turn to other sources for more timely data.
Resourceful researchers have relied on a variety of real-time data sources like cell-phone tracking, moving companies, and move request forms submitted to the postal service to analyze these pandemic migration patterns, but each suffers from limitations. U-Haul tracks where their trucks and trailers are picked up and dropped off. There is similar data from moving companies like United Van Lines and North American Van Lines. We can observe some interesting trend data here, but an obvious drawback is the relatively small sample size for each source. There’s also no corresponding demographic data.
Several companies track anonymized cell phone usage, supplemented by various approaches to profiling consumers. These are excellent sources for tracking where people are situated at any point in time and are especially useful for monitoring shopping patterns. For example, during the worst of the pandemic lockdowns, these sources were able to clearly document the spooky quiet of downtown business districts and the relative jump in activity in more suburban neighborhoods. But they can’t directly track migration patterns, only where population concentrations are now compared to where they used to be, with no sense of how long-lasting these changes might be. Did these people move to a new location, or are they just visiting? The data can’t tell us.
Mortgage data can also provide insights into migration patterns. CoreLogic tapped its database of home-purchase loan applications to provide interesting comparisons of in- and out-migration by metro, as well as the share of mortgage applicants who are changing metros. However, their analysis excludes renters, all-cash home buyers, and anyone moving into another household, and thus is not fully representative of all movers. Renters tend to be more mobile than homeowners, so focusing only on mortgage applicants may distort the observed migration trends.
The most comprehensive and reliable alternative to Census migration data is the Postal Service database compiled from change of address request forms. A key benefit is that this database distinguishes ‘temporary’ (less than six months) from ‘permanent’ moves, which customers indicate on the request form. The data comes out monthly, so we can see how patterns change over time. Importantly, demographic patterns can be inferred by comparing the population characteristics (average income, ethnicity, occupational distribution) of the origin and destination zip codes.
Finally, survey data can fill in some of the missing pieces, especially concerning the characteristics and motivations of movers. The Pew Research Center provided some of the best survey migration data, while a variety of firms like Gensler and PwC surveyed remote workers to understand their motivations and plans. Though these sources inevitably contact far fewer people than does the Census, they’re nonetheless helpful in understanding the broad trends in recent migrations.
American on the move?
The gusher of headlines during the pandemic made it seem like there was a mass migration out of the expensive gateway cities. The data says otherwise. First, as shown in the first graph, Americans are moving much less than they used to. About 11% of Americans moved last year, according to a Harris Poll survey commissioned by Zillow. If confirmed by Census data later this year, that rate is up marginally from the 9% that moved during 2019. But migration rates have been falling pretty consistently for over five decades. The share of Americans moving each year is barely half that in the 1960s. The pandemic did not induce a dramatic increase in household moves.
And we know that when Americans do move, they tend to stay local. About two-thirds of all moves are within the same county (indicated by the blue bars in the next chart) a figure that has remained remarkably steady since the Census Bureau started tracking migration patterns over 70 years ago. Inter-county moves have been rising modestly since 2010, reaching 39 percent of moves in 2019, as shown by the sum of red bars (different county/same state) and green bars (different county and state).
This makes sense if you think about it. With much of the country in various forms of lockdown for extended periods, it was hard for people to move very far. Plus, despite the idea that Americans are willing to pick up and leave for more opportunity, almost three-fourths of Americans live close to the city where they were born. When the San Francisco Chronicle analyzed household and business move data covering the five central Bay Area counties during the first eight months of the pandemic. They found that less than 4 percent left the state, that adds up to a total of only 4,300 in a region with nearly five million residents and half a million businesses. By contrast, almost three-quarters moved elsewhere in the Bay Area, and one-fifth moved to another part of California. Similar trends were seen in New York.
More broadly, 84 percent of households who moved during the pandemic stayed in the same metro, according to a Bloomberg analysis of Postal Service data. Another 7.5 percent moved to another metro within the same state. Similarly, the CoreLogic analysis of mortgage loan applications found that just 17 percent of mortgage applicants were changing metros last year, up marginally from 16 percent in 2019.
The moves out of central cities tend to be even more local, typically to an adjoining or nearby county within the metro. For example, the National Association of Realtors calculated that 84 percent of residents leaving the City of San Francisco moved to one of four neighboring counties, while 88 percent of people departing Manhattan left for one of nine counties in the Tri-State metro (that includes New Jersey and Connecticut).
That’s not to deny the clear migration patterns favoring less expensive markets. When Zillow crunched moving data from North American Van Lines, they found that Sunbelt metros like “Phoenix, Charlotte, N.C., and Austin, Texas, saw the highest net inbound moves in the first 11 months of 2020” while “some of the country’s largest and most expensive housing markets saw the highest net outbound moves, including New York, Los Angeles, San Francisco, and Chicago.” Nonetheless, intraregional moves far outnumbered the better-publicized intercity and interregional moves.
How many and for how long?
Since most households did not move very far during the pandemic, it will be relatively easy for most movers to return to their former neighborhoods or at least move closer once they’re sufficiently motivated. That is, once the urban amenities reopen and especially when their firms call them back into their offices and other workplaces (or when mom and dad kick them out of their old bedroom).
Relatively few workers, even in the tech sector, can work remotely 100 percent of the time: the so-called ‘untethered class.’ The vast majority of office workers will go into the office at least once a week (and most more than that). That means they might be able to reside further from their base office than they used to pre-COVID, but they cannot move out of the region altogether. Thus, many people who moved away from their city apartment during the pandemic ultimately will move back once they need to start regularly commuting to their office, even if it’s less frequently than before.
The critical question, then, is how many households will remain in their new far-flung communities and for how long? It’s still too early to know for sure. Though many leading companies have announced plans to call their workers back this summer, most offices still have not yet reopened for most workers. Data compiled by building security firm Kastle Systems suggests that fewer than a third of office workers in ten leading markets are back in their office. Even in Texas, which was quick to reopen, less than half of workers are back in the office, while in New York and San Francisco, only one in five workers is back. [However, note that these figures likely overstate the share of office workers still working from home because Kastle’s “back to work” estimates are based on security card swipes, an amenity typically found only in the relatively expensive office buildings leased by larger firms that are more likely to have formal work-from-home programs.]
Also unknown is what share of workers ultimately will return to their workplaces once they reopen. Many workers have gotten used to working from home and now prefer it, at least for some of the week, according to a JLL survey. Moreover, many workers say they’d quit if forced to come into the office full-time. A recent Morning Consult survey reported by Bloomberg found that 39 percent of all office workers (and 49 percent of younger workers) would quit if their employer did not allow any remote working. This tracks with a previous Robert Half survey in which a third of professionals currently working from home would look for a new job if required to be in the office full time.
Finally, there’s the question of where the office will be located when it does reopen. Many firms say they will try following their workers to less dense and less expensive office space outside the central business districts (CBDs). Some are experimenting with a “hub and spoke” locational strategy with several smaller satellite offices to replace the traditional CBD headquarters. Thus, some workers may be able to commute to a new suburban office and therefore not need to move back to the central city.
With all these factors in play, we won’t know for a while how many former city residents will remain in their more suburban locations. But there seems little doubt that a meaningful share of the pandemic outmigration from some gateway markets will endure for a while. Central cities were already starting to experience a modest population exodus before COVID as aging millennials wanting to start families were moving to suburban markets in search of larger homes and more open space. Plus, with a much greater share of employers working remotely at least part of the week, they’ll need larger homes than they had in the city to accommodate their home office.
With the pandemic reinforcing prior demographic trends, suburban markets are sure to outperform central business districts coming out of the pandemic recession. This would mark a sharp reversal from trends during the last two property market cycles when CBDs generally experienced more moderate downturns during the recessions and grew faster early in the expansions.
This time, it seems it will be the suburban markets that recover quicker, while vacancies linger longer in the CBDs. These demographic trends have significant consequences for property markets. Some apartment occupancies will revert from suburban to urban markets as former city dwellers move back to their old neighborhoods. But equilibrium occupancies and rents in suburban markets should be higher going forward than they were pre-pandemic, while urban apartment occupancies and rents in many markets could be lower. Office markets should follow a similar pattern, as will support services and retail.
Not all the impacts on urban markets will be negative. Some of the most expensive markets could benefit, even if landlords take a temporary hit. One unfortunate consequence of the extreme prosperity of some gateway markets over the last decade was that too many people and businesses were priced out of the market, robbing these cities of the diversity that made them interesting in the first place. Artists and musicians, cooks and florists, cops and teachers, all could no longer afford to live there. That may finally be changing (for now). With residential and commercial rents inching down, more people and businesses should again be able to afford the rents to diversify the urban landscape. In this way, urban properties will find new life at a new price point, even if lower than their landlords grew to expect.