COVID Didn't Change Real Estate — It Pulled Five Years of Change Into Eighteen Months

The story I had in my head about COVID and real estate was that the pandemic broke things in ways nobody saw coming.

The more I worked through what actually happened to office demand, retail patterns, warehouse absorption, and labor markets between 2020 and 2022, the more I realized that frame was wrong. Almost every "COVID change" was something already happening before the pandemic. What COVID did was compress the timeline.

Remote work was a slow trend. COVID made it the default in three weeks. E-commerce was steadily eating retail share. COVID jumped that share several points in a year. Automation was creeping into low-wage service jobs. COVID accelerated the rollout when "no human contact" became a feature instead of a downside. None of it was new. The pace was new.

COVID didn't change real estate's trajectory. It compressed five years of change into eighteen months — and the developers, lenders, and tenants who hadn't been quietly preparing for those trends found themselves caught flat-footed when the timeline collapsed. The full implications won't be clear for years, but the pattern of what got accelerated is already visible: less demand for traditional office, more demand for warehouse and distribution, faster automation displacing service jobs, a wave of public infrastructure spending colliding with a labor shortage, and a population reshuffling that the industry is still trying to interpret.

That compression is what I want to walk through here.

Remote work changed what office space is for

The single biggest acceleration was remote work.

Before 2020, many large companies had been experimenting cautiously with flexible arrangements — a day or two from home for senior staff, full remote for a small slice of roles. The infrastructure was uneven, the cultural acceptance was inconsistent, and most office leases were still being signed on the assumption that everyone would be in the building most days.

Then in March 2020, every white-collar office in the country sent its workers home overnight. Companies that had no VPN, no laptop allocation policy, no remote collaboration playbook — most of them — built one in two weeks. By the time office buildings reopened, the working pattern had permanently shifted.

The numbers vary by firm and region, but the broad pattern is consistent. Many professional service firms run at roughly 40% in-office attendance now. Tech companies often run lower. Some industries — finance, law — pushed harder for return-to-office and run higher, but still meaningfully below pre-pandemic levels.

For office real estate, the implications are still working through the system. The total amount of leased space hasn't dropped as fast as attendance has, because leases are 5–10 years long and tenants can't just walk away from existing commitments. But as those leases expire, many tenants are downsizing — taking 60–70% of their previous footprint, often in newer Class A buildings with better amenities.

That's the flight to quality in action. Tenants are using less space but paying for nicer space, which means older, lower-class office buildings face two simultaneous problems: they're losing tenants to nicer alternatives, and they're losing total square footage to remote work. Some of those buildings will get repositioned. Some will get redeveloped into apartments or mixed-use. Some will sit empty until somebody figures out a use that works for the bones of the building.

What was a slow grind before COVID became a rapid restructuring after it.

E-commerce demanded warehouses faster than the industry could build them

The other side of the same coin is what happened to industrial and warehouse space.

E-commerce was already growing steadily — roughly doubling every five to six years through the 2010s — and the warehouse infrastructure to support it was already expanding. Then COVID accelerated the e-commerce share of retail meaningfully in a single year, and at the same time, supply chain disruptions taught everyone that "just-in-time" inventory was much riskier than anyone had assumed.

The shift from just-in-time to just-in-case inventory is the more durable of the two changes. Companies that ran out of semiconductors, cleaning supplies, or raw materials during the pandemic decided they'd rather hold extra inventory than be exposed to another supply shock. That extra inventory has to live somewhere, and "somewhere" means warehouse space.

For industrial real estate developers, this turned into one of the strongest demand environments in decades. Vacancy rates dropped, rents climbed, and the kind of distribution centers that had been considered overbuilt in 2018 suddenly looked under-supplied. Amazon, in particular, leased and built at a pace that single-handedly reshaped industrial markets in many metros — though they later pulled back on some of that footprint when their growth normalized.

The longer-term question is whether the warehouse boom is permanent or whether some of it overshoots and corrects. The just-in-case inventory shift is probably durable. The pace of e-commerce growth has moderated since the 2020 spike. The mix of those two forces is what the industrial market is still trying to find equilibrium on.

Automation jumped a fence the labor debate couldn't

For roughly a decade before COVID, there was a slow-burning debate about minimum wage increases, automation, and which low-wage jobs would survive. The "$15 an hour" movement argued for higher wages; opponents argued that higher wages would just accelerate automation and eliminate the jobs entirely.

COVID resolved the debate by accelerating the automation independently of the wage discussion.

Walk into a restaurant in 2026 and you scan a QR code, order from your phone, and a runner brings the food out. The five front-of-house positions that existed in 2019 — host, two servers, a bartender, a busser — got compressed into one or two. The change wasn't strictly because of wages. It was because COVID forced restaurants to reduce contact, and the reduced-contact workflow turned out to be cheaper and easier to operate.

The same pattern showed up in grocery stores (self-checkout went from a fraction of registers to often half), in retail (mobile pay and self-checkout in big-box stores), in hospitality (keyless mobile check-in), and in healthcare (telehealth and self-service kiosks).

For real estate, the implications are subtle but real. Buildings that were designed around heavy front-of-house staffing don't need that staffing anymore. Restaurants that used to need a 200-square-foot greeting area now need a smaller pickup counter. Retail stores that needed a long checkout line need a more compact self-checkout zone. The internal layout of these tenant spaces is changing, and developers building or renovating retail and hospitality property have to design around the new staffing model rather than the old one.

The harder question is what happens to the workers displaced by this shift. That's not a real estate question, but it's a downstream consequence — fewer service jobs in shopping centers means fewer commuters, less foot traffic, and a different kind of demand profile for the centers those jobs used to be in.

Infrastructure investment collided with a labor shortage

The other COVID-era development that will reshape construction for years is the wave of federal infrastructure spending that finally passed.

For decades, US infrastructure spending had been declining as a percentage of GDP. Roads, bridges, dams, power grids, airports, and ports were aging and under-maintained, and the political will to fund replacement at scale wasn't there. The 2021 federal infrastructure package put hundreds of billions of dollars into that gap — the largest single commitment in a generation.

The problem is that the construction industry was already running short of skilled labor before the spending wave hit. Welders, electricians, plumbers, carpenters, and equipment operators were retiring faster than new workers were entering the trades. The infrastructure boom multiplied demand for exactly the workforce that was already in short supply.

The consequence for private real estate development is straightforward and uncomfortable: project costs went up, schedules stretched out, and developers found themselves competing with public projects for the same labor pool. Bids on construction work that used to come in three at a time started coming in single-bid or no-bid. A project that used to take 14 months to build started taking 18 or 20.

In San Diego specifically, the labor question is intertwined with cross-border workforce. A meaningful share of the region's skilled construction trades commute north from Tijuana every day — and when the borders closed during COVID, the labor crunch became acute almost immediately. The region's construction industry has been quietly arguing for immigration reform on labor-supply grounds ever since, regardless of which way the broader political wind is blowing.

For developers, this is now a real underwriting variable. A pro forma that doesn't account for skilled labor scarcity is a pro forma that's going to look optimistic when the bids come in.

Demographic shifts that the industry is still trying to read

The fifth major COVID acceleration was the redistribution of where people live.

Three patterns showed up at once. Some people fled high-cost coastal cities for lower-cost interior metros — Boise, Phoenix, Austin, Nashville, Charlotte. Some people fled urban cores for suburbs or exurbs within the same metro. And immigration into the US slowed sharply during the border closures, then began to recover unevenly.

For real estate, this matters most in the metros that were either gaining or losing the most. Growing metros saw rapid rent increases, multifamily absorption that outpaced delivery, and a wave of new development trying to catch up. Shrinking metros saw the opposite — rent softening, slower absorption, and developers pulling back on new starts.

California is the clearest example of a state where the trend went the wrong way. The 2020 Census showed California losing a congressional seat for the first time in its history, and from 2010 to 2020, roughly 1.3 million more people left California for other states than moved into California from other states or other countries. The high cost of living is the main driver, and until that changes, the pattern is likely to continue.

The immigration piece is the wild card. International immigration has historically been a major source of US population growth, and when it slows, the demographic ceiling on housing demand starts to bind sooner than it otherwise would. The longer immigration policy stays uncertain, the harder it gets to forecast long-term demand in metros where new arrivals had been a meaningful share of growth.

For developers and lenders, the practical takeaway is that the population assumptions in pro formas need to be much more locally-specific than they used to be. "Population grows 1% a year nationally" doesn't mean your specific submarket grows 1% a year — it might be growing 4% or shrinking 1%, and the answer changes the project.

What I keep coming back to from the COVID period

The shift in framing that stuck with me is that COVID was much more an accelerant than a creator. None of the major shifts — remote work, e-commerce, automation, demographic redistribution — was new. All of them were trends that the patient observers had been tracking for years.

What COVID did was force everyone, including the people who weren't paying attention, to deal with all of those trends at once. Companies that had no remote infrastructure built one in two weeks. Cities that had been studying outdoor dining for a decade approved it permanently in three months. Industries that had been resisting automation moved hard toward it when contactless service became a feature.

The firms that did well in this period weren't the ones who reinvented themselves on the fly. They were the ones who had been quietly preparing for the trends — keeping office leases short, investing in warehouse capacity, building remote infrastructure ahead of when they needed it, hedging their tenant mix away from categories that were already declining. They didn't need to scramble because they had already done the work.

That's the lesson I keep coming back to. Most of what looks like a sudden change in retrospect was actually a long, slow change that finally became impossible to ignore. The reward for paying attention to slow trends isn't avoiding the change — it's not being the one panicking when the change finally arrives.

A few things I'm taking away

• COVID's lasting impact on real estate was much more compression than reinvention — almost every "new" trend was an acceleration of something already happening
• Office space demand is still adjusting to a hybrid work pattern that landed somewhere between fully remote and fully in-person, with the flight to quality favoring newer Class A buildings
• The shift from just-in-time to just-in-case inventory is the more durable of the warehouse demand drivers — the e-commerce growth pace has moderated, but the inventory cushion change is structural
• Automation in low-wage service jobs jumped forward years during COVID because contactless workflows became a feature instead of a problem, and most of that automation will stick
• The federal infrastructure spending wave is colliding with a long-running construction labor shortage, which means project costs and timelines are now structurally tighter than they were pre-COVID
• In labor-constrained markets like San Diego, immigration policy is now a real real estate variable — when the border tightens, the construction labor pool tightens with it
• Population redistribution accelerated during COVID, with high-cost coastal metros losing residents to lower-cost interior cities, and California specifically losing more residents to other states than it gained for the first time at scale
• Pro forma assumptions about local population growth need to be much more locally specific than they used to be — national averages no longer represent any particular submarket well

The thing that stayed with me most is how unsurprising the COVID acceleration would have been to anyone who was already paying attention to slow trends. Remote work was creeping up. E-commerce was eating retail share. Automation was nibbling at low-wage jobs. Population was shifting away from the coasts. None of it was hidden. The reward for noticing wasn't avoiding the change — it was being ready when the change arrived faster than anyone expected. The next acceleration will probably look the same way in retrospect: obvious in hindsight, dismissed in advance, and gentlest on the people who were already getting ready.

This post is part of my ongoing series on real estate development — Real Estate Development. The previous post is about why real estate evolves slowly. The next post is about the long-term forces — demographics, climate, affordability, and smart cities — that will reshape what gets built over the next two decades.