“Urban doom loop.” “Office real estate apocalypse.” Today, anyone who reads business news has seen dire predictions for America’s downtown commercial towers, which emptied out when the coronavirus arrived and remain under-occupied three and a half years later. Most coverage centers on the most expensive big cities, such as New York.

But the focus on glittering superstar cities is misguided, because many more fragile downtowns—the likes of Dayton, Ohio; Birmingham, Alabama; and St. Louis—entered the pandemic with little margin for failure. Even Minneapolis, with a strong overall labor market, faced a high office-vacancy rate in 2019. Still more commercial space emptied out during the pandemic, and foot traffic downtown has waned. “It’s spooky,” one retail clerk told The Wall Street Journal.

To be sure, Manhattan office investors and their lenders certainly have plenty to lose, because participating in that market was so expensive to begin with. According to the 2023 outlook from the commercial-real-estate company Colliers International, asking rents for downtown Class A office space in Manhattan are $81 a square foot per year, down slightly from $85 the year before the pandemic. Current rents for comparable space in San Francisco ($79) and Boston ($72) also dwarf the rents typical in boomtowns such as Atlanta ($38), Denver ($39), and Dallas ($31). The rents in some of the priciest markets have started to come down—notably in San Francisco, where Class A rents, according to Colliers, hit $105 in 2019—but are still nowhere close to Sun Belt levels.

[Dror Poleg: The next crisis will start with empty office buildings]

Class A refers to a city’s most attractive buildings—typically recently constructed towers in desirable locations. If rents for such buildings in Manhattan must drop by half to return to normal occupancy, landlords will lose a lot of money. Some major real-estate investors in New York are halting debt payments for certain properties and giving up control to their lenders. The shift in the market could cost New York City 3 to 6 percent of its tax revenue, by some estimates. But the city will still be the world’s financial capital; a tech hub; the headquarters of a slew of major corporations; a home to major educational, medical, and cultural institutions—all of which generates demand for office space even in the remote-work era. New York, in other words, will be fine.  

By contrast, if office rents in the Rust Belt or the Mississippi River Valley drop by anything close to half, downtowns in those regions face abandonment—not only by white-collar businesses and the shops and restaurants that once served their employees but also by the owners of entire buildings. In a city such as Dayton—which, according to Colliers, has downtown Class A rents of $18 a square foot per month and had a vacancy rate of more than 25 percent even before the pandemic—rents can’t fall far while still yielding enough money to pay taxes and operating costs. Class A rents are comparably low in Memphis, Tennessee ($20); St. Louis ($20); Albuquerque, New Mexico ($23); Cleveland and Akron, Ohio ($23); and Birmingham ($23). St. Louis and Albuquerque also had pre-pandemic vacancy rates hovering around 20 percent or higher. Many cities, including Dayton, are working—with some success—to repurpose their downtown with new condos and apartments, restaurants, and entertainment venues. But how quickly struggling central business districts can replace what used to be their core economic activity is an open question. In the meantime, a lender who seizes a commercial building in so weak a market may turn around and surrender the property to the city rather than run up bills while awaiting a buyer.

That is what an actual public-policy crisis looks like: Think of Detroit, Buffalo, or Flint, Michigan—places where, over the past several decades, owners simply stopped paying property taxes and let the government take over. Many abandoned buildings were demolished for surface parking or left vacant altogether, in some cases prompting major publicly funded demolition campaigns that continue today.

When downtown commercial rents are high, it’s partly because the downtowns themselves are desirable places to work—and partly because the supply of office space is limited. New York, Boston, San Francisco, and other cities that are notorious for limiting housing construction also constrain the supply of commercial real estate. The high cost of building in some cities also helps explain high rents, but only up to a point. Indeed the New York Building Congress found that office-construction costs are 15 to 50 percent higher in New York than in most other major U.S. cities. This might justify rents that are persistently 15 to 50 percent higher, but the artificial scarcity is the primary explanation for why, before the pandemic, Class A rents in Manhattan were 74 percent higher than in Chicago and 82 percent higher than in Los Angeles.

For all the hand-wringing about New York, a major rent drop could end up being good for business. Brad Hargreaves, a New York–based entrepreneur, told me on Twitter (now X) earlier this year that his education start-up, General Assembly, rented a “beautiful” space for $29 per square foot in 2010. “In 2018–19 they were charging upwards of $75psf,” he wrote. “We never would’ve started GA if we had faced those rents on Day 1.” Acknowledging this threat to the city’s competitiveness, Mayors Michael Bloomberg and Bill de Blasio broke from anti-growth norms by rezoning areas such as Hudson Yards and East Midtown to permit more office space. Bloomberg’s “upzoning” of Hudson Yards alone legalized 28 million square feet of potential office construction—more than all of the office space of Portland, Oregon.

Not least because of that easing of regulations, Manhattan still has more than 11 million square feet of downtown office supply under construction, Colliers reported earlier this year. That’s about four Empire State Buildings. It’s nearly as much downtown office space under construction as in the entire South—which includes Atlanta, Houston, Austin, Charlotte, Dallas, and a dozen other cities. Even if the pandemic had never brought about an exodus from white-collar workplaces, the addition of so much new commercial space in New York would have forced the owners of existing office buildings to hold down or even cut rents. The new space, combined with the remote- and hybrid-work shocks to office demand, may foretell a Houston-like abundance of office space—which means that Manhattan office rents might conceivably fall to a Houston-like $40 to $50 a square foot.

If you own a dilapidated, highly leveraged building in Manhattan, you may lose it to the bank. But then the bank will auction it to a new owner, who might cut the rent by double digits or convert the property to another use to fill it back up. Nobody should even start to worry about a Dayton-style abandonment of Manhattan until its office rents fall below Houston’s or Atlanta’s. No foreclosing lender will simply abandon a tower that can still collect Sun Belt Class A rents.

The expensive superstar cities enjoy an advantage accidentally created by bad, anti-growth choices before the pandemic. Like nature, markets abhor a vacuum—and if office rents eventually fall far enough below residential rents, developers in cities starved for housing will find a way to take advantage.

Skinny buildings with lots of windows can easily be turned into apartments, particularly if their current zoning accommodates multifamily residential. But few office towers fit those criteria. In harder cases, extensive and expensive renovations, which in some cases may involve cutting huge lighting and ventilation areas dozens of stories deep, can produce high-end residential units. Cities could also change their building and zoning codes to allow dormitories and rooming houses with shared dorm-style kitchens and bathrooms that wouldn’t require threading in new vertical plumbing stacks for every unit.

[Tracy Hadden Loh: Downtown needs to change to survive]

Were New York and San Francisco farsighted in creating housing shortages and a “safety buffer” of priced-out people waiting to move in? Certainly not. Nevertheless, they do today in fact have hundreds of thousands of people ready to move in if they loosen their land-use regulations. The waitlist is shorter than in 2019, but NYC alone is still at least 300,000 homes short of demand.

If office rents really plunge, one last option comes into play: Desperate landlords will start renting out gray-market “artist studios” and not checking too carefully to find out whether people are staying overnight. Are unrenovated Class C office-building interiors ideal places to live? Not really. But neither are the large-floorplate 19th-century factories that have long supplied New York’s famous artist lofts. Impractical floor plans and bad plumbing didn’t stop artists from seeking big, cheap, gray-market factory loft studios when Manhattan began deindustrializing in the 1960s. It wasn’t just the 1960s, either; New York regularly updates the Loft Law to catch up with ongoing illegal factory and office loft conversions in the outer boroughs (and last did so in 2019). Although 1970s office buildings aren’t as pretty as lofts in 1870s factories, they’re also safer to live in.

New York and a few other cities have the easy option of changing the rules—or just looking the other way—as underused office buildings turn into apartments. But this alternative isn’t available to cities with more reasonable housing costs and fewer desperate tenants. Not many New Yorkers will shed tears for the incumbent landlords of Manhattan, whose supply-side comeuppance is long deserved, and an “office apocalypse” that lowers rents for start-ups and opens up space for artists could even make the city more vibrant. Instead, national policy makers and urbanists should be worrying about the already-cheap downtowns of cities that cannot survive any more rent cuts.

When Governor Josh Shapiro of Pennsylvania announced in September that the nation’s largest swing state would implement automatic voter registration, Donald Trump threw a conniption. “Pennsylvania is at it again!” the former president posted on Truth Social, his social-media platform. The switch, Trump said, would be “a disaster for the Election of Republicans, including your favorite President, ME!”

Trump’s panic is consistent with his (baseless) view that any reforms designed to increase voter turnout, such as expanding mail balloting and early voting, are part of a Democratic conspiracy to rig elections in their favor. But he may be wrong to fear automatic voter registration: Although Shapiro is a Democrat, if either party stands to gain from his move, it’s likely to be the GOP. In Pennsylvania, the reform “really has a potential to lean more Republican,” Seo-young Silvia Kim, an elections expert who has studied the system, told me. It’s “not great news for Democrats.”

First implemented in Oregon in 2016, automatic voter registration is now used in 23 states, including three—Alaska, Georgia, and West Virginia—that are governed by Republicans. Rather than requiring citizens to proactively register to vote, some states that use the system automatically enroll people who meet eligibility requirements and then give them the option to decline or opt out. The shift is subtler in Pennsylvania; the state has simply started prompting people to register to vote when they obtain a new or renewed driver’s license or state ID.

[David A. Graham: Actually good news about voting, for a change]

The seemingly minor change, which voting-rights advocates still place under the umbrella of “automatic” registration, is based on behavioral research showing that people are less likely to opt out of a choice than to opt in. By including voter registration as part of a commonly used process such as obtaining a driver’s license—and by presenting it as the default option rather than a form that citizens have to request—states have found that they can increase both registration and turnout in elections. “Even though the process isn’t that big of a shift, the effects are great,” Greta Bedekovics, the associate director of democracy policy at the left-leaning Center for American Progress, told me.

Democrats have led the move toward automatic voter registration, and their 2021 comprehensive voting-rights legislation known as the For the People Act included a requirement that state-elections chiefs implement the policy. (The bill died in the Senate.) But automatic registration does not inherently favor one party or the other, and it has appealed to Republicans in some states because it helps officials clean up voter rolls and safeguard elections. “I don’t know who it will help, and that’s kind of the point,” Sean Morales-Doyle, the director of the voting-rights program at NYU’s Brennan Center for Justice, told me.

A 2017 study by the Center for American Progress found that the voters who enrolled through Oregon’s automatic-registration system were more likely to be younger, more rural, lower income, and more ethnically diverse than the electorate as a whole—a demographic mix that suggests that Republicans might have benefited as much as Democrats.

Other research shows a more partisan advantage. While an assistant professor at American University in 2018, Kim, the elections expert, studied the effects of automatic registration in Orange County, California, the site of several hard-fought congressional races that year. She found that among residents who needed to update their registration because they had moved within the county, automatic registration resulted in no meaningful shift for Democrats. But it substantially boosted turnout among Republicans and independents—by 8.1 points and 7.4 points, respectively. “I was actually very surprised,” Kim said, adding that she’d expected that if any party gained, it would be Democrats. She suspects that Democrats may have been unaffected by the change because in 2018, they were already motivated to vote by Trump’s recent election.

The impact of automatic registration on any one election is likely to be marginal, but even small shifts could be significant in a state such as Pennsylvania, where less than one percentage point separated Trump from Hillary Clinton in 2016 and just more than one point separated Joe Biden from Trump four years later. Several factors suggest that the new system could benefit the GOP in Pennsylvania. Although Democrats have more registered voters in the state, Republicans have been closing the gap during the Trump era as more white working-class and rural voters who stopped voting for Democrats years ago have chosen to join the GOP. Democrats have countered that drift by capturing wealthier suburban voters, a group that helped Shapiro and first-term Democratic Senator John Fetterman win their races during last year’s midterm elections. Because this demographic already goes to the polls pretty reliably, though, automatic registration is more likely to boost turnout among the right-leaning rural working class.

An early-2020 study also suggested that the GOP stood to gain from higher voter turnout in Pennsylvania. The Knight Foundation surveyed 12,000 “chronic non-voters” nationwide before Democrats had settled on Biden as their nominee. Across the country, nonvoters said that if they cast a ballot, they would support the Democratic candidate over Trump by a slim margin, 33 percent to 30 percent. But in Pennsylvania, nonvoters went strongly in the other direction: By a 36–28 percent margin, they said they’d prefer Trump over the Democrat. The eight-point gap was the second largest (after Arizona) in favor of Trump in any of the 10 swing states that the organization polled.

[Jerusalem Demsas: Americans vote too much]

“Democrats sometimes have the mistaken opinion that anybody that doesn’t show up is going to vote Democrat,” Mike Mikus, a longtime Democratic strategist in Pennsylvania, told me. “It’s been one of the myths in Democratic circles for years. Quite frankly, given the changing of the respective party bases, it makes sense that [automatic registration] may somewhat benefit Republicans.” Other recent polls have suggested that the political realignment of the Trump era has made the GOP more reliant on infrequent voters.

The place where Democrats could most use stronger turnout—particularly among the party’s base of Black voters—is Philadelphia, which provided about one-sixth of Biden’s statewide vote in 2020. The city had higher turnout than Pennsylvania as a whole in both 2008 and 2012, when Barack Obama led the Democratic ticket, but it has lagged further and further behind in each election since. Last year, turnout in Philadelphia was just 43 percent, compared with 54 percent statewide.

Yet automatic voter registration might have less impact in Philadelphia than in other parts of the state. Studies have found that the switch drives higher turnout outside urban areas, where Democratic voters are most concentrated. That’s partly because automatic voter registration is operated through the state Department of Motor Vehicles—an agency with which people who rely on public transit are less likely to interact. For that reason, when New York implemented automatic registration in 2020, voting-rights advocates lobbied aggressively for the state to enroll voters through other agencies in addition to the DMV; as of 2018, a majority of the more than 3 million households in New York City did not own a car.

Pennsylvania has no plans to implement automatic voter registration beyond the state DMV. Democrats have been adamant that in enacting the new system, Shapiro was not trying to benefit his party but merely trying to reach the 1.6 million Keystone State residents who are eligible but not registered to vote. Although Republicans argued that the change should have gone through the state legislature, they have not formally challenged automatic registration in court. Few of them seemed to agree with Trump that the reform would doom the GOP. “Its impact will be somewhere between inconsequential and a nothingburger,” Christopher Nicholas, a Republican consultant in Pennsylvania, told me.

Democrats say it’s too early to assess the electoral impact of automatic voter registration, but they acknowledged that Republicans might gain more voters as a result. More than 13,500 Pennsylvanians registered to vote through the new system during its first six weeks of implementation, according to numbers provided by the Shapiro administration. Of that total, Republicans added about 100 more voters than Democrats. “Our former president is almost always wrong,” Joanna McClinton, who leads a narrow Democratic majority as the speaker of the Pennsylvania state House, told me. The fact that Trump is so opposed to the reform, she said, “reveals something we’ve always known, which is Republicans want to keep the electorate small, selective, and they don’t want to expand access to voting even if they could be the beneficiaries of it.”

Whether Trump regains the presidency next year could hinge on the tightest of margins in Pennsylvania. I asked McClinton if she worried that by implementing automatic voter registration, Shapiro had unintentionally bestowed an electoral gift on Republicans ahead of an enormously significant election. McClinton didn’t hesitate. “Not at all,” she replied quickly. “I look forward to seeing the full data, but I definitely am not looking at this from a political perspective but from a big-D democracy perspective.”