There is a coffee trailer parked at 5th and Madison most days; you’ll see it when you walk past City Hall, or when you step out of a meeting at the PacWest tower. Arrow Coffee & Desserts looks natural here – Portland is a food truck city, after all. What’s surprising is that Arrow Coffee & Desserts is parked in front of a vacant former Starbucks.

The coffee cart is popular, so you will have plenty of time waiting in line to ask yourself: why does the cart work where the shop failed? What has to be true about the economy for this to be the best solution? What makes this make sense?

If you’re lucky, you’ll get to meet Aimee, the founder of Arrow Coffee & Desserts, and her perspective will go a long way toward answering your questions.

“Tons of people who work in the building became customers of ours and kept telling us: ‘we are so happy you’re here. We can’t believe Starbucks left.’ Our customers keep suggesting that we open a storefront here… We’re parking in front of this beautiful, empty retail space that I have stared at for months. It is an incredible space and I can clearly see the vision.”

She hasn’t leased it. Although the vision is clear, a coffee shop in the former Starbucks space doesn’t work – at least not with the current variability in foot traffic, the standard ten-year lease terms, and the high fixed operating costs that come with a conventional brick and mortar café.

“I don’t have the right words to describe what I see,” Aimee said. “It’s backward.” She’s right – it’s backward because Arrow Coffee & Desserts gives the PacWest tower more than the tower gives back. The coffee trailer is a reason for tenants to come down in the morning, and a lively line of people that welcomes any prospective tenants touring the floors above.

The PacWest building is a barometer for retail across Portland. In areas with the weakest demand for upper-floor spaces, building owners no longer see the ground floor as a place to turn profit – they can’t find tenants to fill it. Instead, the ground floor is about creating vibrancy and demand. That’s why Arrow Coffee & Desserts operates through an arrangement that looks nothing like a retail lease.

Food trucks like this one thrive in 2026 downtowns because they’re built for volatility – a contrast made vivid in the comparison between the fortunes of Arrow Coffee & Desserts and Starbucks – so it’s tempting to focus on food trucks in lieu of storefronts. But a city filled with mobile operators is a city of failing buildings. The real project ahead is to bring the economics of food trucks indoors.

Building owners and policymakers have an opportunity to write a new operating system for ground floor retail that reflects the new fundamentals of demand – one that is closer to Arrow Coffee & Desserts than to Starbucks. What’s working and not working here at PacWest sheds light on models that are not only more effective (for property owners and vendors) but also more resilient to the urban economy of 2026 and beyond.

Two capital structures

A Starbucks lease is a high-value, long-term transaction. Building out the café costs between $470,000 and $860,000.1 The immense up-front cost, which is typically split between the building owner and operator through a tenant improvement allowance, can only be justified by amortizing it out over a ten-year lease. According to industry conventions, a café is viable when the cost of occupancy (rent plus triple nets) runs about 8% of sales. The rest of the revenue covers the costs of consumables, labor, advertising, amortized tenant improvements, etc. The average Starbucks needs 550 to 700 transactions each weekday to keep rent at a healthy share of sales, and the company negotiates its leases accordingly.2

A conventional ten-year Starbucks lease is structured, priced, and underwritten on the premise of five identical mornings and ten years of predictable demand. In a central business district – and especially the PacWest tower, across the street from City Hall and beneath 30 floors of professional services firms – that premise looked like a sure bet when the lease was signed in the mid-2010s.

Today, employees come to the office at the center of the week, if at all. Tuesday through Thursday carry about 70% of weekday office visits. Historically, more than half of a chain café’s revenue comes in before 11am. The morning window from 7 to 11 saw the greatest declines between 2019 and 2025, according to data from Placer.ai.

Even as the “curve” of a work week shrinks, flattens, and becomes more variable, Starbucks still has to pay for it high fixed costs. The share of sales allocated to rent has crept from 8% to 11% to 15% or more. A ten-year lease underwritten for five mornings can’t be repriced when two of those mornings disappear and the remaining ones falter. In 2025 Starbucks did the math and chose to close 627 stores in a single quarter, paying roughly $450 million to break its leases. A disproportionate number of the terminated locations were the seatless pickup format engineered for a now-nonexistent commuter rush.

Aimee is intimately familiar with the new, emergent workday patterns. “Nine to ten we have our morning push, then again during lunch from noon-one – and then it dies. After two o’clock, it is a ghost town.” It doesn’t take long for an operator like her to understand and begin predicting these patterns – she can open, close, or even move Arrow Coffee & Desserts around town to follow demand.

A truck can flex into the work weeks and work locations that people actually work in 2026 – and it can shift to any other location and opening hours overnight. Arrow Coffee & Desserts caters to events like the Portland Marathon. “There are typically 20,000 attendees,” she said, “and they’re going to want their coffee before their race.”

Even if Starbucks could close early when demand slows and save a bit on hourly staff wages, it would still have to carry the fixed cost of amortized tenant improvements and a ten year lease – that’s the difference between fixed and variable costs. Fixed costs are owed whether or not a customer turns up, variable ones rise and fall with revenue. A café makes more than a cart when it’s running at full capacity, but it also loses more at all times. A food truck, on the other hand, can close or move to follow the energy of its customers. When demand is variable, the mobile unit has a clear advantage over the storefront.

A new rent threshold

The conventional view of the food truck-storefront relationship was perfectly illustrated and legally codified in a landmark case at the Illinois Supreme Court. In 2012 the City of Chicago passed a comprehensive mobile food vendor ordinance that included a 200-foot exclusion zone around brick and mortar restaurants, and required food truck operators to install GPS trackers to prove their compliance.3 It took until May 2019 for LMP Services v. City of Chicago to make its way to the state’s highest court, which held that the City has a legitimate, rational interest in protecting traditional brick and mortar restaurants. The ruling reasoned that restaurants generate vital local tax revenues and economic benefits – casting the truck as a free-rider poaching foot traffic from fixed locations that pay rent, taxes, and tenant improvements to fixed spaces.

In a healthy downtown with predictable foot traffic, high office lease rates, a dynamic real estate market, and healthy municipal property tax revenue, this perspective makes sense. When an office building is operating below 70% occupancy and building amenities are a crucial factor in the prospects of leasing upper floor spaces, the ground floor takes on a new role. Brookfield and JBG Smith, among the largest office landlords in the country, schedule food trucks at more than 150 of their buildings. Platforms like Curbside Kitchen and Suburban Events – which programs PacWest tower – bring local food trucks to office towers.4

This is an opposing view of the truck: not a free-rider, but a vital contributor to the viability of the building, feeding tenants, animating the street, and driving demand for the office space by differentiating the building in a competitive market. Far from a free-rider to be excluded, the truck is a prize worth paying for.

A market-clearing rent happens when a tenant and landlord reach an agreement about value. It is hard to disentangle the direction of causality, but the reality today is that ground floor operators create value for building owners, and rent is falling.

Buildings’ balance sheets are starting to reflect the new conditions. Property owners around Portland are looking for tenants at any rent, in some cases as low as $0. This is a play from the classic suburban mall playbook. Anchor tenants like Macy’s never paid full freight – every other tenant subsidized their rent.5 Food and coffee trucks are the micro-anchors of Portland’s empty office towers.

When Starbucks gave up its lease, Arrow Coffee & Desserts effectively became the building’s café amenity, but it did not commit to a 10 year lease. Instead, the frontage is programmed by Suburban Events, a food truck booker that holds the permit from the Portland Bureau of Transportation. Suburban Events and PacWest share the permit costs, and Arrow Coffee & Desserts, along with a rotating cast of food trucks, pays Suburban Events a share of daily revenue. Most trucks that work with Suburban Events have a minimum guarantee – around $400 to $800 per service. If sales come in under that floor, the sponsoring building owner is responsible for paying the difference (plus a tip) to the truck operator.

The arrangement is novel for Portland. The Arrow Coffee & Desserts trailer occupies the parking lane – a public right-of-way where city code PCC 16.20.150 has, for decades, prohibited vehicle vending. In January 2024 the parking lane in front of PacWest opened up as part of PBOT’s Food Truck Pilot Program run exclusively with Suburban Events. The company works with dozens of locations and food trucks in the Portland metro area, bringing a rotating menu of vendors to buildings, neighborhood associations, office parks, and breweries full of customers.

Suburban Events’ website reads “we understand from data and experience what types of locations produce the highest yield of customers.” The subtext is that ground floor demand dynamic is not happening ubiquitously across the city, or even across downtown. The average price of market-clearing rent and aggregate occupancy rate can be thought of as a block’s “health.” Health determines the likelihood of a property owner fielding multiple offers from prospective brick and mortar retailers willing to pay rent – versus paying for a truck to turn up in front of a vacant retail space.

Here again Arrow Coffee & Desserts is a good example: Aimee has experience at both ends of the spectrum. On a handful of prime blocks, such as Vancouver’s new waterfront, demand holds, and a fixed retail space is viable for a brick and mortar café. Aimee is actively exploring this opportunity, but acknowledges that “to be at the ‘it’ spot, you’re going to pay the price.” PacWest is at the other end of the spectrum, and it’s not alone.

We have run this experiment before

Food carts are an indicator of the building economics behind them – a role they’ve played through decades of Portland real estate cycles, as land availability and value fluctuate. The history runs in three acts.

Act I: abundance. Food Cartology, released in 2008, documents the policy foundation behind a thriving food truck economy. A restaurant with wheels and a towing axle is legally deemed a vehicle – and therefore does not require complex permits. Parking lot owners were happy to lease to trucks while waiting for development opportunities. Surface lots were the best home for trucks, as City code prohibited vending from a vehicle parked at the curb, and sidewalks allowed only push-cart-scale permits. In the early 2000s, many surface lots were formalized into “pods” – a distinctively Portland format that brings together 3-10 food trucks on a site with basic shared infrastructure. At the time, spots in these pods ran $500 to $600 a month. Protagonists of the burgeoning food truck scene had ambition: more than half of the vendors interviewed dreamed of graduating from a truck to a storefront.

Act II: displacement. Food trucks became the crackling experimental edge that gave Portland a unique flavor – and national recognition. The truck was a low-risk, low-cost format that served as an on-ramp for entrepreneurs. In 2008, Food Cartology documented a single transition from cart to storefront, but it was eventually followed by Nong’s Khao Man Gai, Lardo, Salt & Straw, Matt’s BBQ, and Bing Mi, among many others – businesses that have defined Portland’s identity as a food capital. But, as Portland’s food truck scene blossomed, cheap surface lots became scarce. Land values recovered through the 2010s, and parking lots were incrementally developed.6 The tension reached a climax in June 2019 when the Alder Street pod, downtown’s largest, home to around fifty carts, and the one The Amazing Race had broadcast to the world, was cleared to make way for the Block 216 development.

Act III: courtship and confusion. Five years after clearing the Alder Street pod, the City allocated $269,000 to situate food trucks at Ankeny Square. PBOT’s two-year Food Truck Pilot opened parking lanes for vehicle vending in 2024 (paving the way for Arrow Coffee & Desserts). In 2025, obtaining a license to vend at 41 city parks became a same-day permit that costs $5 per day. While the city has prioritized food trucks, the County has threatened to raise health license fees 33%, and is now implementing a 6.5% increase (2026-27). Pod sites run $750–$2,400 a month. Oregon Health Authority rules issued in 2020 went into effect in Portland in 2023, and are reaching a compliance deadline this year. Those rules place significant infrastructure costs on property owners, and some cannot keep their pods open in compliance. Different levels of government are stepping on the accelerator and brake at the same time.

Portland is entering Act IV. Hundreds of storefronts are vacant. In some neighborhoods and buildings, ground floor retail rent is effectively zero: building owners will lease a space for only the cost of triple nets. But vendors no longer have much interest in moving from a truck to a brick and mortar space. Kim Jong Grillin’, a cart celebrated in local and national outlets, is perhaps the most telling. It went from a cart to brick and mortar and back to a cart in 2025, and now offers catering to diversify revenue. The markers of success, for some truck operators, have changed.

Four instruments

Food trucks remain a vital part of Portland’s food scene and broader identity, and they are well suited to the present economy. A cart can run laps around a brick and mortar, flexing with increasingly variable demand. Where downtown is suffering most, food trucks are an important tool not only for bolstering a building’s leasing prospects, but also improving the life of the street for all Portlanders.

It’s tempting to extrapolate the success of the food truck as a model, and make an argument for buildings to turn inside out. The food truck is a retail format fit for the economy of 2026!

But more food trucks isn’t necessarily the answer. Taken to the extreme, a city of food trucks is alive at eye level and insolvent above it. Trucks can animate sidewalks, but they can’t retire mortgage debt, maintain façades, or contribute to a building’s property taxes. In the long run, a city fails when its buildings fail, and food trucks can’t carry the City’s budget.

Today’s political focus on food trucks is encouraging, but myopic. There is an opportunity, instead, to learn from what is making trucks successful today, as a starting point for retooling Portland’s ground floor. Vendors, building owners, and policy makers must explore new models – and the advantages of the truck should be design specs for storefronts.

• Instrument one – regulatory parity downward. Launching a storefront should be as cheap and reversible as launching a truck: pre-approval for activation-ready space; by-right or fast-track temporary occupancy for low-risk uses (similar to the cart-as-vehicle hack, made intentional and moved indoors); fee waivers when carts graduate to storefronts. Most importantly, align the public sector: city, county, and state postures.

• Instrument two – the percentage lease. Building owners price a base rent low enough that a slow month doesn’t kill the operator and agree on a gross sales threshold above which the owner and vendor share the upside. Percentage rent is nothing new – it’s a standard tool for retail leases, especially in shopping malls. Here the old logic can be rebuilt for situations where the landlord needs the activation and the tenant faces significant uncertainty in future sales. Suburban Events, and their contract with Arrow Coffee & Desserts, is a working precedent. Today the revenue share is with the broker who holds the permit; a percentage lease would align incentives directly between tenant and landlord.

• Instrument three – the shared data. If incentives are aligned, the owner and tenant have a reason to share information. Each has data from which the other can benefit to reduce risk or more effectively capture opportunities: transactions, leasing (actual and prospective), foot traffic, events, and neighboring properties. Here again, Suburban Events is a precedent – it offers forecasting based on past data as a value proposition to food truck operators. With good data, they can plan and flex intelligently. A fixed location can get closer to “following demand” the way a cart does – driving inbound traffic and opening when demand is greatest. Shared data can also help surrounding building owners price their leases. A food truck’s receipts are a signal in a district with no recent building transactions. Owners and brokers should treat them as a stand-in for a missing comp set.

• Instrument four – the alternative use. Where retail cannot work at any rent, arguing about the lease is a mistake. In these situations, alternative ground floor uses may be more viable. Retail spaces can be used for small-batch makers / light production (think: Orox Leather), neighborhood services, micro-logistics, live-work, residential, or arts and culture. The blocker is convention, not viability. Building appraisals and financing often still assume ground floor = retail, but the viability of tomorrow’s programmed, non-retail ground floors should be evaluated with financial models designed to understand them.

Now what

There is a real risk of continued real estate market collapse downtown. But Portland is a food city, and trucks are a recovery instrument. Value is flowing between food trucks and buildings in new ways: Aimee is showing that a mobile trailer can work in Portland, just as she proves a storefront cannot – and by negotiating for a brick and mortar space across the river, she’s showing that the numbers aren’t permanently or ubiquitously “backwards.”

The city’s sidewalk vending experiment goes to council for a permanent decision this year. Property owners and policy makers have their pencils out – now is the moment to rewrite the regulatory and financial operating system, not just for the curb, but also for the ground floor. Regulatory parity, along with shared data and streamlined changes of use, will bring food truck dynamics indoors and foster a generation of creativity on the blocks that need it most. The entrepreneurs who will define Portland’s food, culture, maker and arts scenes are here, and they need low-barrier entry points to prove what they can do.

Subscribe now

To buy carpet is to be overwhelmed with options; loop pile, wool, nylon. It is not the most expensive flooring option (compared to, say, hardwood) but it’s easy to spend over $20 per square foot on nice material and installation in a commercial building.

Or, for about the same cost, you could just buy the building.

In December of 2025, Portland’s Commonwealth building sold for approximately $29 / square foot. The building isn’t a dump. Designed in 1948 by Pietro Belluschi (the same architect as Big Pink) and originally known as the New Equitable Building, this 14-story tower is clad in blue-green glass and aluminum. It is architecturally significant: it was the first modern curtain-wall building in America, the first to be sheathed in aluminum, double-glazed, fully sealed, and air-conditioned. The Commonwealth predates New York’s more famous Lever House, and it set the standard for what a corporate building should look like. It was the Armani suit of architecture; this building taught every office tower after it what to wear.

The Commonwealth most recently transacted for $6.5 million, 57% below its assessed value, and 9.4% of its previous transaction cost.1 The story of the Commonwealth building is uniquely ironic – how it tidily bookends the office era – but it is just one in a wave of similar transactions.

The Merchant Bank building sold in March of this year for $42 /sf. The Morgan Building, Park Square, and the Solomon Courthouse all sold in December of 2025, costing between $11 and $47 /sf. Montgomery Park went for 87% off its previous price in a courthouse auction. Jeff Swickard picked up Big Pink and Five Oak – a total of 1.38 million square feet in downtown Portland – for about $55 million total, all cash.

These are good buildings; many of them genuinely fine ones. They are historic, familiar to anyone who has spent time in downtown Portland. The wave of fire sale transactions is not a symptom of low building quality – it is a sign of the financial climate nationwide. Pre-COVID leases are expiring and buildings are emptying out, right as institutional debt matures into a stubbornly high interest rate environment, making refinance impossible (more on that here).

Owners who held onto their hopes for post-COVID recovery are finally giving up, or lenders are forcing a sale. “We’re six years from the shock of Covid,” Jim Costello, an executive director at MSCI – a company that analyzes commercial real estate data – said in a recent Wall Street Journal article. “That’s how long it takes someone to capitulate and give up such a highly valued asset.” Across the country, investors bought up 204 distressed office buildings in 2025, up from 133 the year before.

Who is Buying Portland?

We traced eleven downtown buildings through their last two sales, to understand the relationship between “Transaction B” (the most recent sale) and “Transaction A” (the previous sale).

Across all of the buildings, the average price per square foot of Transaction A was $263, while Transaction B averaged $47. There is a certain shock that comes from seeing buildings trade for less than a fifth of their previous value, but a more interesting trend emerged beneath the sticker price. Across the Transaction A cohort, only two of the buyers were local. Every other transaction went to a non-local investor. When our city was on its way up, from 2008-2019, Portland’s downtown buildings became line items in portfolios managed from Zurich, New York, Munich, Dallas, Houston, Newport Beach, Santiago.

On the way down, only one went to a non-local buyer (two were bought by an anonymous LLC with a Beaverton PO Box). These buyers are familiar to you – restaurateurs, entrepreneurs, a naturopathic medical university, family real estate firms with generations in the city. Portlanders are buying up Portland’s downtown.

Two Kinds of Value

Every investor cares about the numbers, but local and non-local entities tend to see buildings differently. A building represents two different and simultaneous kinds of value – its use value (a function of what it does for its occupants, the reason people pay money to lease it, the operating income it yields for its owner) and its exchange value (the cost to acquire it, and the price it will fetch on the real estate market). No one sees exclusively one or the other, but institutional real estate investors – those non-local firms that hold large, nation-wide portfolios – tend to prioritize exchange value.

The standard tool to estimate the intrinsic value of a building today is a discounted cashflow (DCF) model, which is based on the building’s projected future cashflow. The math starts with a simple assumption: a dollar tomorrow is worth less than a dollar today.2 How much less is set by the discount rate, which reflects risk. At a standard 15% discount rate, a dollar of income one year out is worth 87 cents now; five years out, 50 cents; ten years out, 25 cents.

Every prospective deal is compressed into a standard score, the internal rate of return (IRR), which models the investment’s projected yield. That metric is typically calculated over a five- to ten-year hold period, which aligns with standard commercial real estate investment business plans and exit strategies. If an investor needs to see a 16% return in 7 years, and they can estimate a building’s risk, they know exactly how much it is worth to them right now.

IRR is useful for apples-to-apples comparison of investment opportunities, and even to compare a building with any other financial asset. Real estate can be considered next to a treasury bond, for example, based on its risk and return profile and the time horizon of its yields. In practice, that means an office tower in Portland has to compete for capital against data centers and energy infrastructure, as well as equities and bonds. When the demand fundamentals crack or risk increases or competing opportunities skyrocket – as office demand cracked in 2020 and data centers skyrocketed in 2025 – national investment capital is simply allocated elsewhere.

IRR introduces a bias for quick returns. Because it is based on DCF, and DCF assumes the present value of future dollars decays, IRR rewards how fast money comes back, not necessarily how much money comes back over the long haul. In fact, there is no reason to hold a building for the long haul, because cash received early is counted at full value while the value of cash received later withers to zero. Transactions modeled on a building selling at year seven will beat those modeled on a building held to year fifteen – even if the second building ultimately earns far more cash after year ten. The bias toward a short-term exit is compounded by the fact that most real estate investment capital is managed in closed-end funds that owe money back on a fixed timeline. Exit is deliberately scheduled, and it is not optional.

Until 2019, institutional capital made big, skyline-shaping bets on Portland, bringing us projects like Block 216 and scooping up icons like the Commonwealth building. But the standard DCF/IRR modeling did not factor COVID and hybrid working into its risk profile when calculating the discount rate. Buildings emptied, and soon debt built on the premise of a short-term exit came due. Unable to cover costs or refinance, all but one of the buildings we analyzed went back to their lenders between 2020 and 2025. No bank wants an underperforming asset on its balance sheet for long, so they put their Portland buildings on the market, priced to move quickly. Hence the fire sale.

Institutional investors around the country have chalked up their Portland bets as failures, and they’ve moved on. PwC and ULI’s Emerging Trends in Real Estate, 2026 ranks Portland 80th of 81 U.S. markets, second-to-last for the second year running. The judgement of nearly two thousand investors, developers, brokers, and lenders across the country suggests that non-local funds have no interest in allocating their capital here any more. That’s certainly what we saw when we mapped Transaction B. But we also saw that local investors do have an interest in Portland.

Local money

This is not a jingoistic celebration of local ownership.3 Rather, it’s an argument for understanding the holding patterns and the risk and return expectations that define local investors’ rational behavior. The choices they make will shape our city’s next decades.

Local principals expect to own a building for its whole life, not just a fund cycle. As a result, they tend to prioritize operating income (associated with use value), not exit price (exchange value). For a lifetime owner, a building is a slow, revenue generating asset to be maintained, not a one-shot capital event that must yield returns before an exit at year seven. A building held this way is an option on every use the city might need in the future – and a fire-sale basis is the cheapest option premium ever offered.

All of the local buyers we observed own more than one building in the Portland area, and most own multiple buildings just in the downtown. A single-asset institutional owner hardly pays attention to the neighborhood, and if they do, they see it as an externality. An owner with a multi-building portfolio has already begun to internalize neighborhood factors, whether they’ve noticed or not. What happens with one building affects the other.

They care about the place. That sounds soft and qualitative – and in some ways it is – but that is not the only interpretation of care, and it does not diminish the influence of care in decision-making. A restaurateur bought the Old Town building his pizzeria has anchored for decades; a university bought the complex it means to occupy. These are owners whose businesses are intimately tied to the use value of the buildings they now own. If institutional investors held Portland real estate as a line item, the new generation of owners hold Portland buildings for their details and specificities.

For these three reasons – longer horizons, portfolio exposure, and personal stakes – local money is poised to rebuild downtown. It still might. But there has been little, if any activity in the buildings we studied. Why?

Recovery is a choice

Newly minted building owners have essentially four options.

1. Sell. The buildings have recently transacted for pennies on the dollar, and, with national capital wary of repeating its ‘Portland mistake,’ the total addressable market is tiny. Selling takes time, cost, and effort. → No

2. Upgrade. Putting more money into an office building only makes sense if it will command a premium that pays back the investment. In Portland that is unrealistic. The businesses that would fill it are struggling or moving away. There is roughly an equilibrium between businesses and spaces – no influx means no competition to win with a premium space. Upgrading buys an empty floor. → No

3. Hold. Keep the building as-is, lease what you can, and wait for demand to increase. Even a building that is less than half occupied can cover its operating costs, maybe even eke out a return on cost inside of fifteen years. Maintenance can be deferred and taxes barely increase, so waiting is cheap and has no deadline. → Yes

4. Convert. This option is tricky, for a few reasons. Most conversions run into issues with the size of the floor plate, MEP systems, facade treatments, seismic risk categories, and access. These are knowns, priceable because they are tangible. Regulatory issues are unknowns. Permitting is a wildcard, subject to delays and conflicting departmental demands; it is hard to predict approval timelines. The Design Commission may not exist (in any recognizable form) by the end of 2026. Finally, because the future converted use has an entirely new tenant base, its financial returns are difficult or impossible to predict (but they are likely to be poor, because no single building conversion can generate tenant demand on its own). The one certainty in conversion is that money is leaving the owner’s hands. For all of these reasons, converting any single building is irrational. → No

The final answer is to hold. Even well-capitalized new owners are leasing what they can lease and waiting for the market to improve. The smartest money in the market is using this strategy, and it is perfectly rational on its own terms. It has worked for decades; to date, the cycle has always turned.

That is to say, the hold option is based on the premise of cycles. We’ve argued before, however, that that premise is faulty; demand patterns have fundamentally changed. After six years of holding and hoping for a post-COVID recovery, no one seriously believes office work is coming back in full force. The case is closed.

The logic also starts to break down at the district scale. If every building owner waits for the next cycle, the next cycle will never come. A downtown full of half empty buildings is not an attractive place to be. Each individual owner may be making a rational decision, but the net result is all of them suffering.

So there is a dissonance. The only apparently viable option, holding, doesn’t make sense at a longer time frame and a larger district scale. In fact, it is the riskiest strategy available for the owners that now carry the keys to Portland’s downtown. Holding on to an office building as an office building is the only path that’s actually been tested, and it has proven not to work.

The only way to break the paradox is to change the input assumptions. Selling depends on outside capital, and upgrading depends on increased demand for premium offices from well-funded companies expanding in Portland. Both are far outside of our control. The impasse will only break when the cost and uncertainty of conversion – things we can control – diminish.

Running the “doom loop” in reverse

The “doom loop,” – a spiral of vacancy, falling rents, distress, tax erosion, and urban decline – cascades because one building’s failure increases the odds of its neighbors’ failure. Boarded up windows degrade the whole block. The post-COVID pattern proved that buildings in a district are enmeshed: each one’s viability is a function of what surrounds it.

That mechanism can run in reverse.

Of course, a full building doesn’t automatically upgrade its neighbors – its vibrancy is necessary but not sufficient. Direct benefit comes only when the building uses are complementary, because each de-risks the other. A cluster of uses (say, childcare, grocery, housing, restaurant, and a park) in a meaningful geography each have an easier time leasing up. When these kinds of clusters are programmed together with the residents and businesses who already know the place and who will actively support those new uses, risk dips even further. A conversion that can’t pencil against today’s tenant demand pencils just fine if lease-up is close to certain, and that happens when buildings complement each other.

Risk spreads across a portfolio. Instead of a single building owner making a risky bet to convert a single property, a group of building owners make a coordinated bet on a cluster of properties – or, a “portfolio.” Each use is viable because the others exist; each succeeds because they all do.

The importance of coordination

Just about every celebrated turnaround of this kind had a singular protagonist – in Detroit, it was Bedrock Development, in London, Related-Argent is reimagining King’s Cross. These entities have concentrated land, patient equity, and most importantly, they can single-handedly balance complementary uses.

Portland’s fire sale isn’t producing a Bedrock or Related-Argent; it is producing a dozen local owners with aligned incentives – but no coordinating institution. The project ahead of us is untried: catalytic development through coordination rather than consolidation.

The new generation of owners has the difficult task of cooperatively creating district-scale places. Imagine property owners deciding to convert three buildings, demolish one for a purpose-built replacement, and upgrade one as flexible office, but only after the first buildings have stabilized.

This kind of coordination could happen two ways. The first is through a light collaborative planning effort across the ordinary real estate market. Existing owners and new buyers align investments, share models of projected returns, and sequence complementary uses. This would require no new institutions and no legislation – only a coordinating entity that helps individual owners treat adjacency as information.

The other path, more similar to Bedrock or Related-Argent, would be a special purpose vehicle focused on district recovery. A corridor trust, for example, could syndicate a fund and acquire property across a defined geography. It would make decisions about building uses with the input – even the investment – of residents and local businesses.4 Each building’s capital stack would be broken into discrete time tranches, such that short-horizon money takes years one through five, mid-horizon equity takes years six through twelve and long-horizon stewards – families, endowments, the city itself – carry years twelve and beyond. These breaks are natural moments for another conversion if usage patterns shift significantly. This is not exotic; Albina Vision Trust is already testing cohesive design and community-centered stewardship at the district scale.

It also means reimagining what downtown is for. Buying at a low basis creates extraordinary flexibility. This is already starting to happen in cities around the country. In Washington D.C., a 940,000 sf federal building is becoming housing. In Chicago, a 485,000 square foot office building is becoming an urban vertical farm. “The buy-in at this distressed price allows us the opportunity to afford change,” real-estate developer Marc Calabria said to the Wall Street Journal. He recently bought the building for $4 million, after it sold for $68.1 million a decade ago.

Creative conversions like Calabria’s only work when the entry price is historically low – and that’s exactly what we are seeing in the Portland fire sale. It gives us a license to experiment.

The work remaining

Now is the moment to imagine new, creative, and coordinated uses of our downtown real estate. This kind of place-based transformation is harder than anything the industry has attempted in at least half a century.

For decades, commercial real estate has performed like a bond: steady and dependable. A new breed of real estate investors need to perform more like hedge funds, seeking out risk positions that others cannot manage and making active interventions to generate returns. Owners need to manage risk asymmetrically, which demands more imagination, agility, and local attunement. Suddenly use value matters a whole lot more than it did in a purely financialized approach. The next chapter of real estate will reward those owners who have a fundamental understanding of their buildings and the people who use them – a skill set that blends financial, physical, and human-centered capabilities.

There is no doubut that this transformation will benefit Portland and its residents in the long run. But this is not a charity case; no one is being asked to burn cash for pure public benefit. Coordinated, catalytic investment will yield significant returns for those who bought buildings at a historically low price. The ask is much simpler: “price in” the surrounding district, and make coordinated investments in creative conversions.

None of this makes conversion easy. There are a host of physical challenges, and the regulatory environment is fickle. Coordination will require a clear owner with new templates and broad buy-in.

Field States is actively engaged. We are building a rigorous, standardized method for measuring conversion cost, making options priceable. We are creating district-scale demand modeling, so complementary lease-up can be underwritten. And we are working directly with owners, kickstarting the coordination process inside the ordinary operations of the local market.

Our goal is to put Portland’s existing square footage to work. We are here to partner with the ambitious vanguard of new real estate investors to build value in our communities. Now is the moment to choose what the recovery looks like – and everyone making that choice is already here in Portland.

Subscribe now

A room full of architects, developers, technologists, and public officials gathered in a revitalized industrial warehouse in Detroit to ask why the sector that builds our futures is so bad at imagining its own.

The built environment is often pinned with hopes for a better future. I’m as guilty as anyone. As an optimist who has dedicated a career to this field, I often find myself feeling aspirational about what buildings and cities can do to improve communities – and I know they can do more.

The built environment is ripe with opportunity, in fact. New construction methods might cut the time and cost of delivering housing; new materials can enable “circular” end-of-life reclamation when a building reaches its expiry date; by prioritizing conversion, new investment models can push the building’s expiry date back by decades; new engagement methods can bring more voices into the design process seamlessly and joyfully.

These stand to deliver radically better outcomes – but are they really possible?

See, the problem isn’t as simple as creating new technologies, services, algorithms, or platforms. The architecture, engineering, construction, and operations sector (AECO) has an adoption problem. Think of a construction site. Think of a permitting official. Think of an title and escrow company. None of these is biased toward innovation, and as a whole, the machinery that produces buildings is slow to change.

The challenge of innovation in AECO isn’t just coming up with something new, but also addressing the systemic context around it. How might we prepare the next generation of built environment professionals to create and deliver new approaches?

Last month, Jonathan Massey, Bryan Boyer, and the team at Taubman College convened a room full of architects, developers, technologists, and public officials to ask tough questions about the future of AECO – and inform the design of a Masters of Urban Technology that will be launched at University of Michigan Center for Innovation (UMCI) in 2027. The degree will be the first-of-its-kind, building on UM’s Bachelor of Science in Urban Tech).

A few ideas that have stuck with me.

MEASURE TO LEARN. If we want new outputs, we need new inputs (and we need to measure outputs, too). Society needs us, the AECO sector, to do better. Buildings can be more adaptable, healthier, more equitable – but the information that goes into the design process is thin and static. Post-occupancy analysis is rare. It is nearly impossible to say, with any confidence, that a building delivered a year ago is successful, beyond lease-up numbers and energy data. That means there is no objective loop – there is no feedback mechanism, and a system without feedback cannot learn. In short, the value of architecture is justified on charisma and aesthetics alone.

CHAIN DRAIN. Knowledge leaks at every interface in the value chain, and the chain is long (this is attributed to Bryan Boyer, whose brilliant piece on Chain Drain you can read here). Hundreds of individuals are involved in the value chain of a single building, from initial site selection and project financing, through design, engineering, construction, and operations. Because each actor uses their own software, has unique expertise, and carries specific liabilities, knowledge escapes at each interface, or “handoff.” Decisions upstream constrain the option space downstream – and, because incentives of different actors are misaligned, there is seldom coordination. The result is a value chain optimized for cost and risk reduction, with each stage blind to the implications their choices carry for others. This jeopardizes building quality and long-term performance that owners and communities actually care about, and it short circuits recursive improvement loops.

“Handoffs between the firms involved in delivering a building or public space also chop up the feedback loop between design intent and actual lived outcomes… Think of this as chain drain: each link in the supply chain creates an opportunity for optimizing the cost but comes at the expense of learning.” — Bryan Boyer, “Beware of Chain Drain” at Urban Technology at University of Michigan on Substack

REVISE THE STANDARD, don’t dilute the project to meet outdated requirements. We often try to regulate or support new technologies with existing policies and regulations – even if they are an awkward fit (when dockless scooters arrived onto city sidewalks, Montreal used a food truck policy to regulate them, while Boston reached for a rule that was developed for group Segway tours decades ago). That means innovative ideas get sanded down to satisfy outdated rules. We need to continually ask what the policy is actually producing or precluding, and what its intent is. Then look at what is repeatedly falling on the cutting-room floor because it didn’t comply. Should the rule change? Could new constraints create pressure to innovate? Urban technology could have a design review process that learns from its own rejections and variances.

REIMAGINE BUILDINGS, in every way: from program to ownership to finance. The session was graciously hosted by Synecdoche Architects and the community at The Vander – a reimagined warehouse in Detroit’s Milwaukee Junction neighborhood. It is a surprising building, surrounded by vacant buildings, empty lots, and a construction site. I thought I was in the wrong place right up until I opened the door. Inside I was met with an exquisite gathering space, creative design studios, storefronts, and a woodshop where people were working in full swing. But none of these caused my surprise, really; it was only later, when I heard Lisa Sauvé describe the financing and operations model of the Vander – built on a rare mix of solidarity, distributed ownership, and sheer creativity – that I was truly blown away. She has fundamentally rethought the inputs and outputs of architecture. It is so successful that Sauvé and her firm Synecdoche are replicating it in nearby buildings, while fielding interest from other neighborhoods and other cities. The Vander proves that conventional debt, revenue, and return models aren’t the only way of making a building work.

The AECO industry cannot deliver adaptive places until it becomes adaptive itself – until it learns to sense, measure, and revise cooperatively between the public and private sectors. Buildings will be designed and operated in the image of the inputs we give them, and the outputs we gather should inform the next generation of projects. I’m excited to see how the new UMCI Masters of Urban Technology degree program will transform the sector.

Subscribe now

The Design overlay zone ensures that Portland is both a city designed for people and a city in harmony with nature. The Design overlay zone supports the city’s evolution within current and emerging centers of civic life. The overlay promotes design excellence in the built environment through the application of additional design standards and design guidelines that:

• Build on context by enhancing the distinctive physical, natural, historic and cultural qualities of the location while accommodating growth and change;

• Contribute to a public realm that encourages social interaction and fosters inclusivity in people’s daily experience; and

Promotes quality and long-term resilience in the face of changing demographics, climate and economy.

Last month, Portland’s leadership called the Design Commission into question. The Design Review Reform Report explores improvements – and potential suspension – of the Design Review process, with the broader objective of accelerating development.

Perspectives on the Design Review process are mixed – what it should entail, how it can improve, and if the Design Commission should exist at all. This is an active surface area for debate, and the outcome will shape the next generation of Portland’s building projects.

This week’s post is a transcript of a talk that Matthew Claudel gave to Portland’s Design Commission. It picks up on a provocative idea in the Design Commission’s mandate – supporting the city’s evolution – and advances a strong view on Design Review, grounded in Adaptive Urbanism. For a video of the full session (including a presentation by Will Smith and Randy Gragg, Q&A and discussion) see the city’s records here – and many thanks to Peter Andrews, Tim Heron, and the members of the Design Commission for your thought partnership.

Today I am going to talk about how the built environment adapts to what society and the economy need. Obviously, what society and the economy need is changing quickly, and what I’m going to share is a way that design commission can play a role in the process of the built environment adapting.

I’m going to make the argument that the work of the next decade is adaptive urbanism, and that design review can actually guide that process. And I’m going to be concrete because you’re dealing with very concrete issues.

The first part of my talk might sound a little pessimistic, but hang in there. You’re going to have a starring role, and it is, in fact, very optimistic.

I’m going to start off with what I call a tale of two towers. These are two very familiar and very iconic buildings in Portland. The first is Big Pink, the US Bancorp tower. It was designed by SOM with Pietro Belluschi, completed in nineteen eighty-three for around sixty million dollars.

For forty years, this building traded steadily upward. A $165 million in 2000, $286 in 2006, and $372 million in 2015. There are rumors that there would have been a transaction in 2019 that would have put this well over four hundred million.

The building was renovated in the mid-2010s and achieved LEED platinum status in 2019.

But in July 2025, Big Pink sold for $45. That’s less than its 1983 construction cost in nominal terms, and it’s a fraction of the capital invested over decades. It’s probably less than the land is worth.

This building was beautifully designed. It was beautifully maintained. It was owned successively by JP Morgan and JLL and UBS, and it was managed by Unico, and these are the best in class. But Big Pink was designed for one purpose. It was designed to be a corporate office tower, and when demand for that purpose disappeared, the building’s value disappeared too. The big idea here is that buildings designed for a single purpose are stranded when that purpose disappears.

The second tower I mentioned is Block 216, the Ritz Carlton building. This is arguably Portland’s most impressive tower since Big Pink. The design is top-notch. It has premium branding with Ritz-Carlton. It had sophisticated institutional debt behind it. The financing was ambitious, but Portland in 2017 was extremely hot, and the note was based on the best possible projections.

But 9 months after groundbreaking, the world locked down. The tower was delivered in 2023 into a market that the developers could never have imagined when they started in 2017. On opening day, the office space was 15% leased. By mid-2025 it was 23% percent leased.

The hotel was operating around 45% below Ritz-Carlton brand average, and only eleven of the condos had sold. As you all know, in 2025, the lender took possession, and the total exposure was around $560 million.

But, again, I want to stress that this was not ineptitude. These are smart people working from information that had been reliable for decades. My read is that it wasn’t a bad bet, but it failed because they put all of their chips on the table for a single future, and that future didn’t emerge. The building was essentially locked into a single path. Even when it became apparent that they needed to pivot, they were unable to change.

These two towers are part of a cascade that we’re about to see in Portland and across the country. If you look at 7-10 year leases that were signed before COVID, they’ll be expiring now to 2027 or 2028.

Tenants are going to have to decide: do we want to take on this lease even though the rates haven’t materially decreased, or do we want to exit entirely? What we’re seeing now in terms of vacancy rates is a leading indicator of something that’s going to continue and perhaps get worse because when someone leaves, it’s easier for the next person to leave.

At the same time, the debt on a lot of these buildings is maturing into a high-rate environment. Owners bought these buildings at pre-2019 valuations, and their debt was 3-4%. But now they’re going to have to refinance. Money is between 7% and 9%, and the assets are worth a lot less, 50-80% less.

And that’s why over 700 commercial property owners in Portland filed tax appeals last year. We can expect a wave of foreclosures – that’s just basic math and basic logic.

But it isn’t necessarily a bad thing. It doesn’t have to be a bad thing. The risk is if foreclosure is followed by a long hold.

This is when institutional distressed asset investors buy up the building and essentially wait out the cycle. The problem is that if everyone is waiting for the next cycle, the next cycle is never going to come. It’s essentially a tragedy of the commons.

Big Pink and Block 216, are very different buildings. They’re 40 years apart. One failed after decades of success, the other was dead on arrival. They represent spectacular losses – but neither was a result of incompetence. And they’re trying to teach us the same lesson, which is that they failed to adapt.

Big Pink could only be interpreted as an office building. The Ritz couldn’t get off the train tracks that it had set.

It’s tempting to write off COVID as a single exceptional moment that rocked the built environment, and it did. But thinking of it as a single, one-off event – an exception – is, I think, both naive and dangerous.

The pace of economic and social change is increasing. The magnitude of these changes is increasing, and these shocks are compounding. I read this as a paradigm shift in thinking about what the built environment is for.

I grabbed these headlines today. AI and autonomous vehicles. Data centers are now the most investable asset class that you can find in real estate. But quantum computing is about to go prime time, and when it does, all of those data centers are going to be obsolete. We need more preschools. “Downtowns as experience” is booming.

What I’m trying to demonstrate is that society and the economy are changing, and they’re going to keep changing, and they’re going to keep changing more significantly. What we had before, as an essentially cyclical process, is no longer going to hold. Society and the economy are shouting: we don’t need downtown offices.

Those low-priced office buildings are actually a remarkable asset, and the best way to think about them is adaptive urbanism.

The task of the next decade is not only to reimagine what our built environment is for, it’s also to build the capacity to continuously adapt in a faster future. Right now, our built environment – the buildings and the institutions that govern them – are not designed to run at a very fast pace.

You know this. You know that it takes years and years to examine a UGB expansion. You know that it takes years to go through permit review.

Adaptive urbanism is an approach to designing buildings and blocks and districts and the systems behind them to learn and evolve in response to changing economic, social, and environmental conditions.

Essentially, you can think about this pointing in two directions, forward and backward. Forward-facing is to design buildings fit for adaptation. Think about all the buildings that don’t get stranded, things like early 1900s warehouse buildings. These things have been industrial, they’ve been creative offices, they’ve been loft apartments, they’ve been retail on the ground floor.

I used to work at a building in Old Town called the Povey Building. It was originally a glassworks – they used to have high-heat furnaces to melt glass. Then it became a home for marketing and creative agencies. For a while, it was packaging and fulfillment for a cannabis reseller and it now has advanced manufacturing in the basement and design studios up above. Tomorrow it could be retail, or any number of things.

Old warehouse buildings can adapt because they have certain architectural features. The specs are really unglamorous: generous floor-to-floor heights, a structural grid that allows for subdivision, and structure that’s decoupled from the core.

We can design buildings this way today. A perfect example is LEVER Architecture’s building for The Schnitzer School at PSU. This is a thoroughly contemporary building and it’s designed to adapt. It has structure at the perimeter – there is no structural core. The partitions are all movable, so the building can essentially become what students need it to be in 2040 or 2050.

This building, interestingly, was even designed so that a floor could be added or removed late in the design process, because funding wasn’t certain until late in the game. Imagine if Block 216 had that kind of flexibility during design! They would have been able to modify what the building would become, as they received market signals.

LEVER’s building is ready for a faster future. I’d argue that it’s actually the lowest risk investment that you can make – flexibility is an incredible insurance policy. If I’m an investor, I want a building that holds options on many possible futures.

Adaptive urbanism is also about backward-facing adaptation. When you have a very low basis on a building, you can turn it into something else. You have some flexibility. A lot of the conversion conversation right now in Portland and in cities around the country is focused on office-to-housing, but housing is really only viable in a limited number of cases. There are many more endpoints to conversion.

I’ve already mentioned a few of them. An obsolete office building could become a perfect data center, an early childhood care facility, or a site for advanced manufacturing. It could be indoor agriculture or accessible self-storage. These are the kinds of things that society and the economy actually need right now, and all of them can become viable when you have the right low cash basis.

We’ve developed an analytical framework that we call a Conversion Friction Coefficient. This essentially quantifies the cost and the complexity of converting a specific building from a specific origin to a specific destination. It accounts for the structural system and services and envelope and regulatory dimensions, and it flattens all of that into a number between zero and one.

LEVER’s building at PSU, for example, is close to zero for most endpoints, because it can adapt. But a standard ’80s office building is close to one for housing because it’s very hard to get there. You will have structural issues, MEP issues, and so on. The conversion friction coefficient builds on Stewart Brand’s concept of pace layers. A building is composed of layers that cycle at different rates, and you want layer independence, so that something that wants to cycle fast doesn’t get constrained by something that has to cycle slowly.

A couple of specific examples. The one on your right is the Houston Post. This was a 1962 postal facility that has become a cultural and commercial hub. It was modified or reimagined by OMA. And the one on the left is Detroit’s Roosevelt Warehouse – an Albert Kahn building, actually – that is now home to Newlab, a tech startup accelerator with advanced manufacturing and creative offices.

Think of adaptive urbanism as the evolution of the built environment. Every experiment, like office to childcare or a new use for a parking garage, is like a mutation in the process of evolution. Designers are causing mutations.

You’re probably thinking: the built environment has always adapted. There is a steady vernacular process of evolution. That’s true. But what’s different today is that the pace is so much faster. What used to be a slow vernacular process that had inbuilt adjudication is now pressured to move extremely fast, and the mechanisms that typically govern the evolution – things like zoning or permitting or construction timelines or financing or even just our agreement of what makes a good building and a bad building – they weren’t built to run at the pace that they now need to run.

My view is: we should certainly speed those up. We need faster permitting. We can create new financing vehicles for conversions. We can use data to sense change as it’s happening, rather than quarterly reports from JLL. We actually have a book coming out that enumerates all those things.

But it does not say “get rid of design review.” This is the crucial question, and it’s really what I want to ask you, the Design Commission.

If the the pace of urban evolution is accelerating, what’s playing the role of natural selection? What are the fitness criteria that we’re using to decide which mutations are the right ones?

In biological evolution, natural selection makes those decisions, and it shapes the overall arc of a species as it evolves. And if we enable experiments with new building structures and uses, which, as I just described, is crucial to our city’s future, those experiments are going to create new value, and they’re going to introduce new paths in Portland’s ongoing evolution.

It’s the design commission’s job to decide which paths are most desirable. You have to ask: Value for who? What kind of city do we want? You play a vital role in the process of urban evolution.

Your role is more important at this moment, not less important, because the value at stake is enormous. The purpose of buildings is changing, right now. Big Pink just sold, and it will become something new. With that change, the character of downtown and the identity of downtown and the purpose of downtown is going to change.

If Design Commission leans into this role, it has to pivot from a focus on maintaining aesthetics to a focus on guiding adaptive urbanism.

Every day, designers and developers are responding to what they think society and the economy need, and you’re at the leading edge of deciding how those ideas become urban form. This is exactly what you’re already built to do, in five important ways.

The first is discretionary variance, the second is problem-solving, the third is as a public interface, the fourth is flexibility as a criterion (maybe that’s new. You aren’t already designed to do that). And then the fifth is recursive learning. I want to go through these each one by one.

Discretionary variance. I want to credit Nolan Lienhart for articulating this really well and changing how I think about it. He described how design guidelines specify goals, and then the Design Commission interprets how those goals are best satisfied.

The interpretation changes over time. His example was above ground structured parking. The design goal is to have a vibrant street front. Years ago, a sculptural screening facade was sufficient. But recently, commissioners began looking for programmed space at the facade with parking behind it.

So to quote Nolan, “The way that today’s Design Commission interprets proposals may be completely different from the interpretations of Design Commission twelve or twenty years ago. That’s the adaptation function. The rule didn’t change. Their interpretation of what created an active and vital public realm changed.”

This is the Design Commission doing its job. It’s adapting in real time without waiting for code to catch up. Discretion, in other words, is a tremendously important way for regulation to stay current – and you’re already doing that.

But it’s important to consider this role strategically. Architects and developers are coming to you with new ideas (or “mutations” in evolutionary terms), and you’re performing the fitness function. But as you do that, you can’t treat variances as one-off, single decisions. You need to think of every mutation as part of a longer evolutionary branch. That means each one of them has to test a hypothesis about the future of the public realm. How will it play a part in a longer arc of change?

The second role of design commission is as a navigator. Because these experiments are unconventional, by definition, every variation is going to run into complexities. The design commission has an opportunity to help projects work through these thorny problems that are going to arise. The built environment evolves, and so too must the internal systems of city government. As you work through conflicting code or issues raised by experiments, you can help the mechanics of city government evolve.

Let’s say someone wants to convert a parking garage into an autonomous vehicle charging depot (this is a very real question, by the way). Don’t just focus on the facade screen. Help restructure the relationships between BPS, PP&D, PGE, and PBOT. When you do that, you’re setting precedents, so that we see the right kind of infrastructure deployed in the future.

The third role is as a public interface, or a bridge. The design commission is essentially the connection point between residents and city government when it comes to the future of the built environment. That is to say, the fitness function isn’t only internal. Ideally you’re incorporating residents’ perspectives into it and continuously improving it over time. Design Commission is a public forum where residents can comment on development, and, over time, developers and architects get exposed to public values. This is where they learn what Portlanders really care about.

Crucially, this is a two-way street – you’re also sharing out. This is important because so many planning conflicts arise when trade-offs in development decisions are misunderstood, or the long-term strategies are not visible to the public. You can change that.

The fourth role is to ensure that flexibility is a criterion. So if you think back on the PSU building, this is a building that can adapt, and I believe all buildings in the future should be built to adapt. That has environmental repercussions. It means we don’t need to tear down buildings when their use is obsolete.

And it means designing for layer independence. So the space plan, for example, isn’t constrained by the structure. There are very specific architectural ways to do this. The conversion friction coefficient I described earlier gives us a shared and evidence-based language for evaluating the adaptive capacity at a building level.

Low conversion friction should become a design review criterion alongside the context and the public realm and the quality. In practice, this looks like asking a question of every project: What possible programs can this building accommodate in the future?

Finally, the most important role that I want to emphasize. If there’s one thing you take away from my presentation, it’s that the Design Commission has to enable learning.

For every discretionary variance that you allow, document why you made that choice. What hypothesis is it testing? What guardrails did you set? What were the outcomes over time, and how did the private sector respond?

The same is true of problem-solving. If you’re navigating issues one by one without incorporating solutions structurally, it’s a Sisyphean boulder. You have to solve the problems but then stabilize the solutions. It means continuously evolving public values as part of the fitness function.

If you are enabling experiments, you should be seeing a variety of outcomes. In other words, failures are inevitable – but they won’t be mistakes. If you’ve structured the experiment well, you’ve mitigated the potential risks, and you will learn.

The Design Commission’s role is to embed public value into the evolutionary process of urban adaptation over time.

Without the Design Commission, we might get faster development, we might not. We might just get institutional investors land banking downtown. We’re guaranteed to get a more chaotic process of urban evolution in the next five, ten years. The Design Commission is uniquely positioned to guide Portland’s adaptation to a new economic and social reality, and shape what the physical expression of what that is.

I do work all over the country, and I think Portland’s at the leading edge. The systemic dynamics are visible here early and clearly. And I think Portland is poised to demonstrate what post-COVID urban adaptation actually looks like. We can build and demonstrate a new model for design-driven reimagining of the built environment and the institutional framework around it.

Subscribe now

If you have been through PDX airport, or the adidas campus, or the newest branch of the Multnomah County Library system, you’ve fallen in love with timber as an architectural material. And few architects have had more of a role in the renaissance of timber in the Pacific Northwest than Thomas Robinson. He is the founder of LEVER Architecture, the Portland firm whose Albina Yard, completed nearly a decade ago, was the first modern mass timber building in the United States. In 2015 LEVER designed Framework, whose structural system now underpins most of the cost-effective mass timber office buildings built in the country since. LEVER’s recent work includes The Schnitzer School of Art + Art History + Design at Portland State University and the thousand-foot Timberlab production facility in Millersburg.

What follows is one of the clearest practical articulations we’ve encountered of what the Design Imperative of Adaptive Urbanism looks like at the building scale. If buildings are options rather than bonds, then we need to know what an actually-adaptable building looks like, how it’s structured, and what materials make it possible. Robinson walks us through how to do adaptive design concretely – how a building can be designed to expand or contract mid-design, why digital fabrication has become inseparable from physical craft, what mass timber lets you do that concrete and steel don’t, and what working backward from already-built warehouses at Red Fox Commons taught him about how to design net-new buildings moving forward. Our conversation started with a book sitting on my desk: Stewart Brand’s How Buildings Learn.

MATTHEW CLAUDEL: What appeals to you about Stewart Brand’s thinking? What are some of the ideas that stand out?

THOMAS ROBINSON: I think that he sees buildings as part of a continuum, not as a single point in time that you’re designing for. That’s what’s really valuable about the way he thinks – about everything. He was the original “do it yourself to save the world” guy. The Whole Earth Catalog provided access to technologies and ideas that improve people’s lives, but also improve our relationship with the planet.

Stewart Brand is one of the reasons the Bay Area is such a hotbed of innovation. He put tools out there for people to do innovative things, providing access to resources and sharing all these ideas about living off the grid. How do you make your own power? How do you make your own computer?

CLAUDEL: This is a bold statement: at the core of Silicon Valley innovation is the idea that you can do it yourself. Now that I think about it, it tracks. There is a myth of the Palo Alto garage as a place where you can just do it. You don’t need anything fancy.

Stewart Brand also talked about buildings changing over time, buildings having a life and adapting. One of the things that really appeals to me about that is that buildings reflect the character of their inhabitants and change with the needs of their inhabitants. That’s something we perhaps forget as architects. But your practice returns to the materiality, the maintenance, and the tectonics of things. Do you think that echoes Brand’s mentality?

ROBINSON: In the trajectory of my education and practice, it seems like I’ve alternated from the conceptual to the physical… I had to learn how to weld. I learned how to do woodworking. I learned how to do machining. All at a very basic level but enough to be dangerous. But what it did was to flip me from a conceptual way of thinking about design to actually building things in a way that tries to evoke a conceptual idea.

A lot of the time we take all these specific ideas and theories and we make incredibly complicated and useless spaces. In terms of their longevity, it’s a one-time thing. You’re designing the space to only work for one thing, and for these specific people who need this single thing. But when the building is complete, or even halfway through, well, everything’s changed. We need something more flexible. That’s why I think the [prototypical] shed or larger spaces that are designed to be flexible in the first place are the ones that have a long life.

A lot of those types of spaces are factories, right? [The original designers] knew that the processing equipment they would put in there was going to have a shelf life. They knew they would have to take it all out and redo it or reconfigure it because they had a new product. We don’t think that way enough about new buildings that are not about making. Absolutely not.

We’ve done a lot of spec buildings, compared to a lot of architecture firms our age. So you don’t know what’s going to go in there, but you need to have enough flexibility that many things could work.

CLAUDEL: Architects love to build a Ferrari: beautiful and optimized for a single thing. You and I agree that the Ferrari is not the building that makes most sense right now because the world is changing so quickly.

But the alternative presents a dichotomy. On the one hand, you have a space that is quite neutral – what you’re referring to as the shed, something in which many different things can happen. It’s very simple. On the other hand is a building that can change over time. It’s not neutral, but it’s flexible.

So I want to ground this in concrete specifics. Let’s talk about the PSU building that you are just in the middle of finishing. The Schnitzer School of Art, Art History and Design. This building is a marvel of engineering. I want to walk through it as a worked example. Is it correct to think of it as a building that has the structural capacity to change?

ROBINSON: While we were designing [PSU], we had to design in the ability to easily add a floor or take away a floor because there was uncertainty about how much money would be available. It wasn’t clear that the legislature would vote to fund it entirely. The system was designed to change at any point – early or very late in the design process.

To do that without blowing up the process or the schedule, each individual element had to have a logic that was connected to how it’s detailed and connected to how it’s made. We call it the framework system, because Framework designed in 2015 was the first project where we used this system.

The structural system for Framework was driven by the fact that we only had enough money to do one fully loaded fire test for one component of the building. And so we designed a frame system connected by a thin beam made up of the CLT panels, in collaboration with our structural engineer KPFF. It was out of that necessity. Now pretty much every efficient office timber building in the US uses that system if they want to do something that’s more cost effective.

So PSU has 10 years of learned behavior built into the system that, when you look at it, looks super simple. It is relatively simple. But when that system gets repeated, and rotated, it becomes a lot more complicated.

CLAUDEL: The framework system is essentially structural modules fabricated offsite. It sounds like you have shifted clash detection and tolerance and a lot of what usually happens on site upstream. Is it happening during fabrication? What are the software tools that you’re using to span from the design process to fabrication to the construction process? And how does that shift your role as a designer?

ROBINSON: At this point, we’re still in the early days. Obviously we’re using Revit, and that connects into a program called Navisworks, which is the standard that contractors use for clash detection. But what we’re fundamentally doing is using that Revit model as the base for a digital assembly of every single component of the building with all the people that will actually be eventually building it. You need to have this in order to do what we’re doing with timber – to prefabricate and precut thousands of penetrations into the slab before they arrive on site.

You have to build the building digitally first, down to an eighth inch or three millimeter tolerance for every single piece. The timber is not actually the complicated part. It’s all the other systems that have to be integrated into it. Our team spent months saying, “Well, this pipe’s going here and that’s going there.” And then we actually disassembled them digitally, removing those systems from the model so we could see what’s left, which looks like a timber Swiss cheese. Then finally that feeds back into the model for the timber and that’s what’s fabricated. Then we bring everything back and fill in the holes with the building systems on site during construction.

There are a series of operations and it just takes a lot of effort from a lot of different people. I’m sure that, like you say, buildings learn. I’m sure artificial intelligence can learn how we did it with the teams. And it’ll be interesting to see what happens with that, right?

CLAUDEL: I have no doubt. Feel free to tell me if this is crazy, but I’m thinking about your relationship with Timberlab, your relationship with some of the MEP engineers and so on – the architect sounds a lot more like a product designer than an architect.

ROBINSON: Yeah, I think the architect is [like a product designer]. Interesting architecture is a version of product design. I don’t see them as that different, at least from my perspective, because I get really interested in how things are made, and that making is related to how we design. The material characteristics of what you’re building can be the keys to innovative architecture – just as one detail can then lead to a whole building. To me, it’s like cellular DNA. The replication of the detail creates the building.

Some of the people that I admire the most, like Jasper Morrison [think this way]. The way he talks about how he designs is inspiring. In the end, a building is a product, right? It’s a thing in the world.

CLAUDEL: What you’re describing requires an intimate knowledge of structural systems or material systems. You know how glulam works, for example, and you know how attachments work. You’ve gone through it. And that’s why the hanging system for the Timberlab factory building can become the parti for a whole building.

ROBINSON: Those are always the most exciting projects for me: when the detail is the building. When there are very few degrees of separation between the detail and the experience and the culture that you’re working in. I don’t know why we are not more focused on that as architects.

CLAUDEL: When you have a separation between the engineer and the architect, for example, each one starts blaming the other for problems that arise. You’re certainly not going to get a building that has the integrity of the component-to-full-building fractal you are describing. You’re talking about a continuum from the conceit of the building to its fabrication, to its construction. But the building doesn’t end when you cut the ribbon, right?

The literature for the PSU building says that the interior partitions are designed to be reconfigured. In theory it’s modular, so it can change over time. In practice, what’s locked? What is mobile? What would a future occupant need to understand if they’re thinking about this building changing? For example, if they decide the small studio concept isn’t working for the student of 2035. If we need to rethink what education looks like. How might the building react to that future condition?

ROBINSON: At the most basic level, we try wherever we can to get rid of the core as a structural concept. A lot of times the core is the problem in terms of flexibility. If it’s structural, moving it is almost an impossibility.

The PSU building has no structural core – all of the structure is out at the perimeter. That’s also true in Spring Street, and it’s true in the project that we did in Kirkland, where we’re doing a timber exoskeleton. With the base move of how you handle the structure, you create the most amount of flexibility downstream. You can put up partitions or clear out a whole floor plate. The reality with PSU is that all of the walls can be removed at some future date. You could just come down to the basic shell with the braces and then begin again from there.

The other thing we did at PSU was separating systems so there’s not as much friction. We really tried never to put a wall and the structure right on top of each other – we always separate the two. The rhythm of the columns creates a series of micro galleries for students, but it also allows the wall to be removed with minimal interaction with the core timber structure.

We know the building systems will need to be replaced at some point. We know people are going to need smaller spaces, or they’re going to need bigger spaces. And the building’s completely set up to accommodate that. Obviously it will always cost some money but it’s not impossible. The other thing about PSU that’s interesting is that it’s basically a 40’x10’-foot module, repeated and rotated in a pinwheel throughout the entire building. All of these systems can potentially expand forever or they could only be a very nominal small chunk.

CLAUDEL: That implies extensibility in the future, even if what we’re building now is only phase one. If we decide it needs to be longer or shorter, it can.

ROBINSON: If you have a system that works, the building designs itself. We used the same method with the facades we did at the Arthouse project for the Pacific Northwest College of Art. We started with a panel we could afford – a Chevron-shaped metal panel. We’re like, let’s just design the whole facade around that as if it were a custom panel. The ready-made became something special because we used it and treated it like it is something special. Standardized products populate the world of architecture. If you really look at them hard enough and think about how to combine them in innovative ways, they can transcend their mundaneness.

CLAUDEL: Absolutely – let’s get into materials! With the PSU project, you’re describing a separation between the column and the wall, so that they’re not interacting with one another. That isn’t necessarily something that you can only do with mass timber. You could do that with many different structural systems. I’m curious – is there something you can do with mass timber that you can’t do with concrete or steel? Like cutting or reinforcing? In other words – is it a better material for adaptation? People make a carbon case for mass timber. Is there a modifiability case to be made?

ROBINSON: I would say so. Most houses in the U.S. are timber frames and they’re relatively easy to modify because of that. You don’t need to get a concrete saw in there to make an opening. Mass timber members are also a bit easier because the material itself is wood.

It’s not as processed as steel or concrete. Timber itself is something that actually was alive at one point. You don’t need tons of crazy tools to actually modify something made out of timber.

When we started learning more about it, we met this amazing guy, a Canadian timber installer. He and his crew were called “Chainsaw Divas” because they had these special chainsaws to get into tight spaces. When there was a problem with these early timber buildings, if they weren’t put together right – remember, it was just early days – the Chainsaw Divas would come out and actually modify huge panels or beams on site.

If you look at the PSU project, there’s an interface where 40-foot timber beams hang from the steel brace frame. To accomplish this we oversized the beams, then when they got them in the field, they just got a chainsaw and cut them on site to perfectly marry up with the steel.

When you see that this is possible, you can also see into the future that other modifications would be possible more easily.

CLAUDEL: So today we don’t have Chainsaw Divas – we have software like Navisworks.

ROBINSON: Right – we’re building and taking apart the building multiple times in the digital space, using a digital twin.

CLAUDEL: Presumably that twin will be a resource in the future, right? That’s an evergreen asset.

ROBINSON: Yes. Literally every single component of the building is in that model. It will be interesting when somebody feeds that model into Claude…

CLAUDEL: It’s probably already happened, Thomas.

ROBINSON: It’s probably already happened, yeah.

CLAUDEL: Over the years you’ve been at the forefront of mass timber architecture. But it’s been almost a decade since Albina Yard, and this is still fairly niche. What is holding everyone else back? What’s the actual barrier here? Is it code? Is it capital? Is it contractors? Is it design culture? Is it insurance? What’s going on?

ROBINSON: I would take almost the opposite view. I’m kind of amazed when I look at where we are today. There was nothing [mass timber] happening in the US 10 years ago. And now there’s like 3,000 projects and everyone, every developer and every person I talk to wants to build with timber. Editors at different publications are worried that they’re featuring too many timber projects. I think it’s really our own impatience, you know. 10 years is just a blip in the larger history of architecture and building, even in the United States.

We forget so easily that companies like Katerra invested billions and then went up in smoke. Google Sidewalk Labs – hundreds of millions. There have been a lot of very smart people who are like, “I am really smart and really good at writing software so I know how to be a contractor and build things.” And the reality is, they really don’t. I’ve been in architecture 30 years, and I still have so much to learn.

Companies that have been around for a hundred years like Swinerton are saying, okay, I’m gonna invest in this and figure out how to do it in a way that really works. We’re building the kind of culture that will allow timber to really take off in the next 10 years.

CLAUDEL: As we see more companies like Swinerton or LEVER, or maybe some folks you worked with 20 years ago who have now started their own firms – do you think that as more architects and developers start using this material, it will become less expensive? Can we anticipate economies of scale?

ROBINSON: When you have more manufacturers, more cultural knowledge, more assurance that if you decide to build out of this material, you can predict generally what the building will cost.

We’re moving from brewing beer in the garage to brewing beer in a converted warehouse to brewing at an industrial scale. As you go up in scale the quality has to be better, because the risk is higher. The controls have to be tighter, the systems have to be more efficient. I think we’re in the transition, at least in the United States, from a medium sized brewery to an industrial scale brewery. As we go through that, you’re going to see an increase in quality and also an increase in innovation – we will have to make it work.

CLAUDEL: There are two sides of the building culture that we haven’t addressed. Obviously architects care – they love seeing CLT and mass timber construction in magazines. Others are not really as enamored of it: permitting and workforce. Your Framework project has a permitting element to it, if I’m not mistaken. You’re publishing it openly – and presumably that makes it easier to go through permitting and makes mass timber construction more standardized for workforce training. Are you seeing that bear out?

ROBINSON: The research that was done for Framework in 2015 has been critical for high-rise timber projects elsewhere. They used the open source testing and detailing for Framework to get their projects permitted. That’s well documented. You also asked the question about workforce – I think that’s more complicated. D.R. Johnson was one of the first gluam fabricators in the Western U.S. to have a CLT press. There were other people doing industrial CLT for other purposes, such as temporary roads. But architectural CLT manufacturing is technically complex. You need to have engineers. You need to have quality control. You need to have people that are running the machines, and people that are running the shifts. And a lot of the people that do that work are engineers and require specialized training and they often want to live close to a city. It creates challenges.

CLAUDEL: It doesn’t sound like a bad thing. It just means that we’re solving for a new kind of economic condition.

ROBINSON: It’s pretty exciting. Have you been to Timberlab’s new facility? It’s awe-inspiring – the scale. If you had told me 15 years ago that I would be designing a thousand-foot-long, all-timber factory, I would be like, you’re nuts, right? It feels like something that [Pier Luigi] Nervi did during the ‘30s or ‘40s. I’m not saying that what we’re doing is equivalent. They were doing much more innovative structures out of concrete. But I think it’s something special about the culture and history of Oregon, and the timber industry. It’s a massive step.

CLAUDEL: It reminds me of the Metabolists in post-WWII Japan, also. Their ambition was to create infinitely extensible structures – linear cities. Buildings are a simple module that can repeat indefinitely. Metabolism also had a goal of buildings that change over time.

ROBINSON: What is different for me is the actual experience of [the architecture]. The modules [designed by Metabolist architects] are beautiful but also maybe a little inhuman. What’s interesting about timber is that even at a small scale, the material connects. That might seem like a minor piece of the equation, but I think it’s actually pretty major. When you amplify a material that feels connected to our humanity, I think it amplifies that connection we have to a space. When you take something that is highly processed, or has very little [human] connection – something blank, like a concrete wall – and you amplify and repeat it hundreds of times in a modular city, it amplifies the character of that material. I’m not necessarily proposing that we do massive cities of timber, but I think it’s a quality that architects should be thinking about.

CLAUDEL: You’re right. The Metabolist movement was enabled by innovations in concrete. At the time there was a sense that the plasticity of the material unlocked architects from a lot of the constraints that they had to worry about before. So they started to think grander and they started to think in terms of infinity – something an architect would never in their right mind have thought of before. Perhaps we can rescue some of the modularity that was really interesting about that movement, but blend it with the humanity of timber.

Shifting gears, I’m curious about your adaptive reuse projects – Red Fox Commons and Highland Hall are two examples. Maybe you have others. These started as industrial buildings, so they have a big structure that’s oversized for office or residential uses. What did that structure let you do? Where did it constrain you? And what is it like to work with an existing building rather than designing something from scratch?

ROBINSON: Redfox Commons and Highland Hall are different. At Redfox, there were two 30,000 square foot warehouses with 100 foot spans. The main building was built during or right before World War II, and used a timber truss system. We were talking to our engineer and they said it was designed right at the edge of what could work. There weren’t a lot of materials around, so they were trying to create the biggest possible span with the least amount of material. It was a huge metal shed with 100-foot span trusses and two rows of columns at each side.

Over time [the building] obviously evolved. So when we got involved, it was just a Pyrenesian structure with all these timber mezzanines with little dividers painted different colors, and people doing different things all throughout it. The first thing we did was to clear everything out because we had to seismically upgrade the building. And after we took it apart we found we had thousands of beams and columns and planks of various sizes.

We were in the middle of that demolition process, we thought, “well, what if we could use some of these? There were two warehouses. What if we use some of this timber to connect and create a bridge between the two?” That sounded good to the developer… we had all of this material that we already “owned” that we could reuse.

So the team cataloged a general typology of beams, and the ones we had the most of were the 4x12s that they’d used for flooring. There were a bunch of other ones, big old timber beams, the columns, that were worth a lot of money. So [the developer] sold a bunch of the timber, and we kept a certain amount. We had a timber grader [do an evaluation] – they come in and calculate the bearing capacity for each piece of timber.

We laminated two old 4x12s with one new piece to create a hybrid beam and column system. The contractor set up a production line in the existing building where they screw laminated all these pieces together. These guys were in there with screw guns, fabricating all the bridge pieces for weeks.

We saw an evolution – this building was filled with makeshift structures. We took it apart, cataloged it, and put it back together. And once we put it all back together, we blasted the whole structure with walnut shells to make it feel new. People really love that project. We won an AIA honor award for it. But it was really just about taking this material that happened to be there. It’s like if somebody handed you, I don’t know, like a hundred pounds of filet mignon, maybe even a tougher cut, and said “what are you going to do with it?”

CLAUDEL: It’s not crazy to imagine that the vacant buildings around the city are filled with similar – not the same, but similarly interesting – kinds of materials.

ROBINSON: Yeah. You’re seeing interesting research being done in Switzerland and Europe where they’re cataloging cities – all of the materials in all of the buildings that could maybe be reused. Creating a network of reuse.

Highland Hall is another evolution of that thinking. It was more intentional. [University of Oregon] had a grant to do a multi-story seismic test of a mass plywood panel structure at the Tallwood Design Institute at Oregon State University. There was a Master’s Architecture student who cataloged the components after [the tests] were complete. The budget was not large. So we thought, well, we need to try to re-use this material in the new School. It made that project really interesting. It’s an example of a more regenerative or synthetic approach where you’re holding two truths. [On one hand] I’m designing this building, [on the other hand] they’re going to be taking this thing apart. Can we land the plane at the same time? It took a relationship with a University of Oregon Architecture Professor and friend, Judith Scheine, who was one of the founders of the Tall Wood Design Institute.

CLAUDEL: If you back up a little bit, I think it’s interesting that the architecture school is in a building that has been a gymnasium and a theater.

ROBINSON: Right, but it could easily have been just a warehouse. It makes architecture more interesting when there are unexpected experiences and uses. At least I’m inspired when I go to a particular place and see these flavors or materials or ingredients that I’ve never experienced together.

For example you might have old growth timber – that feels contemporary when it is screwed together with state-of-the-art screws. Or you have an architecture school that’s actually reusing material that would have just gone into the landfill, but now has a refined platonic quality to it. To me that contrast is interesting.

CLAUDEL: Even if these materials you’re finding are humble, you manage to elevate them, and as you were saying, you can imbue them with new kinds of meaning over time. That’s typically happening in projects that work backwards – an existing building where you’re using a “harvesting” approach to find new materials and work them into the design of the renovation. But you’re also designing buildings that look forward, like PSU. Ones that can adapt in the future. How do you learn from old buildings? Are there lessons you’re learning from old buildings that you can embed into how you design buildings for the future?

ROBINSON: I think it comes from experience. I’ve learned what not to do. When I was a young architect, I worked for this architect named Joe Esherick, who’s a well-known Bay Area modernist. I was 23 and I was working on a brewery in Montana. He was in his 80s at the time, a World War II veteran. He just said, “Go do it.” So I did it. There wasn’t micromanagement in any way. It was 1995 and there was this high-tech aesthetic floating around. I thought – I’m going to put my life into this railing design!

When it was complete he took me aside and said: “You could have just done a simple picket.” It wasn’t mean. He wasn’t putting me down. But I always remember that. Sometimes you can just do something very simple, and it can be very good. It doesn’t necessarily need to take a massive amount of effort, but it does have to take into account the actual technologies, so that when you’re designing the thing, it is the thing. We need to always [remember] that what we are drawing will actually be a thing – as opposed to having a great idea that looks great in your head but you have no idea if it’s actually going to become something that’s buildable or real. You always have to be working with the thing itself.

CLAUDEL: That goes back to what you were saying right at the beginning – there’s a tectonic sensibility that architects used to have. You’re not just making drawings. Ultimately, you’re making a steel plate that holds a wood beam, for example. With a deep understanding of those different elements, you can actually start to play.

Shifting gears: we are facing increasing uncertainty and an increasing pace of change – what is the world asking of architectural practice that the discipline isn’t set up to deliver?

ROBINSON: One of the key things facing architects is the sense that what you’re creating is valued without having to justify or explain why it’s valuable. A lot of architects are very good at talking about their work, but the talk doesn’t always connect to the experience that people have in the space, or how they feel about its usefulness for what they have to do, or even just what the culture actually needs.

So I think architects need to demonstrate their value through innovation at every stage. When people say I’m going to hire an architect, that’s actually going to create a certain amount of value, both in terms of the experience [of the building], but also even in terms of the economics of a project. If architects can demonstrate that they create economic value and also societal and ecological value, then we’re going to have more ability to impact the future.

I also think, as we’re designing on a specific site, we really need to connect what we’re doing there to sites we may never visit, but are definitely connected to our project through the materials we build with and energy we use. The choices we make about the materials we use, how they were sourced, how they were transported and who puts them together can lead us toward a more regenerative approach to design. We should see our sites as something more – not just the program and the urban context, but also how materials matter in a larger context of the world.

CLAUDEL: That speaks to the value that you’ve been able to create with LEVER. Many of the systems you’re creating, whether it’s an attachment and structural system or whether it’s a piece of software, these are things that extend beyond the building itself. These are replicable systems. In a sense, you’re not just designing a building, you’re designing the first example in a larger system that will generate buildings in the future.

ROBINSON: Right. For me, architecture is more like a transformational tool rather than a single [object] on a site. We’re designing a set of tools that we can then deploy to solve different issues in different ways. And the more you apply and work with the tools, the more intelligent they get.

The connector for the Framework system is a good example. When we built it the first time, it was kind of clunky. The second variation utilized a cleat connection that is a manufactured product, and now it’s just timber to timber to timber connection using a mortise and tenon detail that is thousands of years old. It took a lot of evolution to arrive at something simpler than what we started with. There’s a beam and a post, right? But there is also 10 years of thinking about how they interact. Hopefully we keep evolving.

There’s a really interesting quote, I think it is attributed to Leonardo da Vinci. “Think well to the end, begin with the end in mind.” To me, that’s about thinking through, from the concept you have all the way through to how it will be made, and then you come back to the beginning.

A lot of times people think about the concept, concept, concept. And they never say, “Well, what would this really be if I just built it? What if I had to build it? What would it actually end up being?” That’s something we’ve done a lot. It’s definitely Stewart Brand’s ethos. We’re going to go out and do it ourselves: design a system that actually does what you’re asking for. We’re going to find the fabricator and we’re going to find the supplier and figure out how we would do it. We’re not going to do a mock-up that makes it seem realistic. We’re just going to put it all out on the table and make it real. I don’t know if a lot of people in the profession always value that. I do think there is space for more prosaic systems to be part of architecture and hopefully part of more people’s everyday experience of buildings.

CLAUDEL: That’s a beautiful place to end. Thank you Thomas.

Subscribe now

This is the eighth in our Adaptive Urbanism series. Previous essays have focused on foundational pillars – what we refer to as “imperatives” – for adaptive urbanism. The cognitive imperative focused on data to sense urban change as it is happening, the institutional imperative focused on governance and stewardship, and the economic imperative focused on finance and capitalization for the built environment, in a world of increasing volatility. This is the design imperative, and it is the most visible expression of Adaptive Urbanism.

The Houston Post, a hulking concrete structure at the center of the city, was designed for a single purpose: to process Houston’s mail. Construction was completed in 1962, and the building served its purpose well for half a century before it was decommissioned in 2014. By the mid-2010s, the infrastructure was aging, and the development option – sixteen acres of prime downtown-adjacent land – was more valuable than the building. The Houston Post was scheduled for demolition.

But a developer, Lovett Commercial, saw adaptive potential. The building’s heavy concrete frame is engineered for industrial loads. A regular column grid is sized for sorting operations. Floor-to-floor heights are much greater than contemporary office norms. By virtue of its initial use, The Houston Post had a generous structure that could, decades later, become a remarkable canvas for reinvention.

The developer hired OMA to realize the building’s adaptive potential. The team, led by Jason Long, carved vast atriums out of the sorting floor, built a rooftop lawn, added three monumental staircases – idiosyncratically shaped as an X, an O, and a Z – and brought light flooding into spaces that had been gloomy for fifty years. The building reopened in 2021 as a cultural and commercial hub with food halls, music venues, co-working, retail, and the largest rooftop park in Texas. POST Houston, as it is now called, is worth more (financially and socially) than it ever could have been as anonymous mixed-use development on the site of a demolished postal facility.

POST has an unmistakable life to it. Families gather in a buzzy dining area, office workers cross paths with concert-goers heading to the 713 Music Hall. Above, five acres of rooftop park spread across the building, paths winding between gardens, an organic farm, and an unobstructed view of the downtown skyline.

The building maintains its historic identity, but POST adapts the structure to fit the economic, social, and technological patterns of today. The building learned, playing host to a transformation its original designers could never have imagined – and adaptation was only possible because of its overbuilt structure. Some buildings have this capacity by accident, but every building should have it by design. In a world that changes rapidly, the least risky building is the one that holds options on multiple possible futures.

POST is a remarkable example of wholesale adaptation at a single point in time. Buildings more commonly evolve in fits and starts – clumsily, intermittently, often against their architects’ intentions – in a decades-long sequence of tenant improvements, structural retrofits, and code compliance upgrades.

In How Buildings Learn, Stewart Brand described that process of change: how it happens naturally, and how the physical elements of a building fight against change. To Brand, a building is a set of nested layers, and each wants to cycle at its own pace. He identified six, ordered from slowest to fastest:

• Site — the geographical setting and the plot itself. Stable for longer than human timescales.

• Structure — foundation and load-bearing elements. Thirty to three hundred years (but usually no more than 60)

• Skin — the exterior envelope. Cycles every fifteen to twenty-five years as the building weathers and building codes evolve.

• Services — the mechanical, electrical, and plumbing systems. Replaced every seven to fifteen years.

• Space Plan — the interior layout. Reconfigured every three to thirty years, and now, in commercial buildings, almost continuously.

• Stuff — the contents, like furniture and fixtures. Changes daily.

Brand observed that buildings fail because of layer entanglement. When a building feature that needs to change fast is constrained by one that is slow, the building cannot evolve to meet the demands of the moment. If services are cast into structural slabs, if the skin is inseparable from the frame, or if the space plan is locked to load-bearing interior walls, the building is doomed to obsolescence, locked into the technologies and assumptions of its original design.

Brand’s framework points to layer independence as a design principle. Each layer must be able to change at its own pace without forcing changes to the layers beneath it.

The Houston Post was designed for a specific program, and engineered for permanence, but it was amenable to radical change precisely because nothing fast was locked to anything slow. OMA’s redesign treated the layers of the building separately. The structure stayed, for example, but the services were entirely replaced, threaded through the existing frame using new risers. The Cold War era building will surely continue to evolve over the decades to come – and when it does, the space plan will shift more quickly than the services, the skin more quickly than the structure.

Layer independence is the foundation for an analytical protocol we call conversion friction. The number reflects the cost and complexity of changing a building’s use from a specific “origin” to a specific “destination.” The viability of pivoting any given building from office to residential, for example, or from retail to laboratory, is quantified as a “conversion friction coefficient,” between 0 and 1. A lower conversion friction coefficient means that it will be easier to exercise a particular conversion option; a higher conversion friction coefficient means that specific conversion will be challenging. Conversion friction has four dimensions, each with several contributing factors.

Structural friction encompasses floor-to-floor height, building depth, column grid, and load capacity. Structural friction carries the strongest weight in calculating the conversion coefficient because it tends to be very difficult or expensive to alter after initial construction. Take the example of an office building with 11ft slab-to-slab height. Laboratory spaces require 12 feet or more to accommodate fume hoods and specialized servicing. It may technically be possible to carve through a floor to offer greater height, but it is so challenging that the office to lab conversion option carries a very high conversion coefficient. The option is structurally foreclosed. In that same building, the floor to floor height may work well for residential conversion, but the depth of the floor plates result in a lack of interior daylight – another high-friction variable. The building is designed for a standard structural load capacity of 2.5-3.0 kN/m² – counterintuitively, less than the live load minimum for a parking garage, at 1.92 kN/m². Office to parking conversion is low friction, along with office to advanced manufacturing.

Services friction follows close behind structural. Plumbing constrains office to residential conversion, because drainage requires gravity flow, and routing waste pipes through existing slabs often means cutting through structure. A building with services embedded in the structure has extremely high conversion friction for office to residential, regardless of how generous its other dimensions are. Electrical capacity, vertical riser space, and HVAC compatibility are all part of services.

Envelope friction refers to the building’s skin. Sealed-facade office buildings, common since the 1970s, cannot be converted to housing regardless of what their structure or services would otherwise allow. Operable windows, daylight reach, and thermal performance under tightening building codes all sit in this layer.

Regulatory friction includes use class restrictions, building code differences across uses, historic designation, and accessibility standards that can foreclose options the building’s physical fabric permit. The building can be physically convertible and legally trapped.

Certain conversion origin and destination pairs will tend to have higher conversion coefficients – things like office to housing or warehouse to early education – while others, like office to self storage, will tend to be lower. Right now, those are mostly intuited. The concept of conversion friction gives a quantifiable cost basis.

Stewart Brand observed buildings naturally changing, and identified the time scales typical to each layer. Conversion friction preserves the sequential ordering of Brand’s shearing layers, but shifts the focus from the time scale of natural conversion to the cost and complexity of immediately exercising a conversion option.

Considered alongside market signals and capital availability, conversion friction determines the viability of an adaptive reuse project. A specific conversion may be high friction (expensive, complex) but still enjoy returns because the demand for the new use is so great. Most office to residential conversion is high-friction, but the cost of conversion will never be recouped through rental yields in a normal market. Many cities are experimenting with grants to reduce the cost. The concept of conversion friction helps us compare different conversion paths as apples to apples: office to housing, or self storage, or educational, or light industrial.

In that way, it helps surface an important policy question: will a grant program have a lasting, structural impact on the market, or will it temporarily and artificially subsidize a type of conversion that will never be viable in the real estate market without an injection of free capital?

The previous essay in this series presented a thought experiment: Building A, optimized for a single use; Building B, designed for adaptation, costing five to ten percent more at construction. Under stable conditions, A wins. Under conditions of uncertainty, B wins.

The specifications of a building that has many options – a building with low conversion friction to many endpoints – are unglamorous. It needs a floor-to-floor height of at least 12 feet (15 at the ground floor) to preserve residential, office, and laboratory options through the building’s life. It needs a 25-30 foot structural grid that allows subdivision. Load capacity above 5 kN/m² preserves laboratory and light-industrial options at a modest construction premium. Services organized in accessible central risers, with primary and secondary distribution separated, allow tenant-level changes without building-level disruption and allow major upgrades without demolishing finished spaces. It needs floor plates no deeper than 50 feet to glass, and direct street access. Ideally, the structure is separate from a central core.

For generations, architects have known how to design buildings for adaptation. But they have lacked a financial argument for why anyone should pay the premium. If the pacing problem is correct, and the future brings accelerating uncertainty, the financial value of reduced risk far outweighs the increase in upfront cost. The difference is the cheapest insurance an investor can buy.

Stewart Brand’s project was descriptive, showing what adaptation looks like in existing buildings. He did not, except in passing, ask how an architect would intentionally design a building that adapts well. Three decades on, the pacing problem brings that question to the foreground.

Field States worked with LEVER Architecture to answer it. We created the schematic design of collaborative hub building on an innovation campus (the specific client is anonymous while the project is in development). Because the future needs of the emerging industry are highly uncertain, the hub building was designed to maintain maximum optionality and extensibility. The architecture is based on a simple repeating structural unit that can be extended for any number of bays.

The design for the initial launch phase is based on our research and human-centered design process, and includes office, light manufacturing, labs, and an event and demonstration space. But the building is intended to flex and adapt to the observed patterns of use during years zero to five. If tenants spend more time prototyping than on Zoom calls, office areas can convert to light manufacturing bays. If the building is over subscribed, it can extend. If the industry grows more slowly than expected, the structure isn’t over-built, and it can become something else entirely. A light conversion will cost orders of magnitude less than a building that is stranded because it is optimized for a use that never materialized.

The success of the building depends on its capacity to adapt, and the direction of its adaptation depends on its capacity to sense the evolving demands of its users. What do people need, and how does that data inform the evolving design?

This is where the cognitive imperative informs design at the building scale. Quantitative and qualitative data – merging, for example, indoor positioning, tool utilization rates, event calendars, user interviews, and leasing patterns – is gathered in real time and analyzed longitudinally. Together these show what occupants need from the building.

The architects of the Houston Post were not designing for adaptation; they were designing for mail. The building was specified for a program that no longer exists, but its idiosyncrasies, like the floor to floor heights and column grid, are the reason the building still does. It happened to have adaptive capacity and learned in spite of itself.

POST belongs to a wider family of accidentally adaptive buildings: late 19th century warehouses, pre-war commercial blocks, post-war industrial sheds, and certain mid-century institutional structures. All of them have structural slack (born of caution, regulation, or the engineering conventions of their era) that can be taken up by multiple possible uses. These buildings have low conversion friction for many potential uses.

Field States’ project with LEVER embraces that ethos – but it is designed to learn from the start. An upfront investment minimizes the conversion friction for almost every future use, reducing long-term cost of ownership for a high-performing building. The owners – program directors, really – can exercise optionality throughout the building’s life.

The buildings that struggle most under conditions of uncertainty are the ones that have thoroughly internalized optimization, with sealed envelopes that hamper future ventilation needs, floor plates deeper than residential daylight thresholds, services embedded in structure to save inches of clearance, floor-to-floor heights specified for a single use case. These buildings are perfectly tailored and perfectly obsolete, the durable failure of a design culture that treated single ideal solutions as the highest objective.

Under accelerating uncertainty, a building tailored for a single predicted use is the riskiest and most expensive thing an architect can deliver. The cheapest insurance a building can carry is low conversion friction, and an embedded capacity to learn.

Subscribe now

Nolan Lienhart operates fluently across design, policy, and planning strategy. From his start working for Earl Blumenauer, to supporting organizations like Albina Vision Trust and 1000 Friends of Oregon, to civic engagement on the Oregon Capitol Planning Commission, Nolan sees the built environment through institutional and political lenses that most designers do not. But Nolan is also a designer, leading complex, district-scale development projects at ZGF and bringing a keen attention to the fine grain of urban fabric.

Portland’s Central City Roundtable launched this year, and is co-chaired by Nolan and Mayor Wilson with a mandate to “reimagine and reignite downtown for new patterns of work, culture, and community life.” We caught Nolan at a moment when he’s focused on translating ideas into action. Our conversation touches on Portland’s ground floor retail, the role of urban planning, and the challenge of balancing ambitious projects with quick wins as we create Portland’s next chapter.

MATTHEW CLAUDEL: You’re working on projects around the country. Is the way Portland is talking about its future different than other cities?

NOLAN LIENHART: The legacy of planning excellence in Portland is a blessing and a curse. There are some people for whom that legacy is a North Star. It’s a sense of identity and a belief in our ability to do great things together. But it is also a bit of a crutch: the belief, whether it’s explicit or implicit in the way we show up to the table, that planning excellence is in our DNA. High expectations can both serve as an inspiration and they can also serve as a real point of despair when you’re not meeting them.

When I look at other places, particularly places that weren’t as successful in the eighties and nineties – the kind of places that would have been here touring Portland when we were doing light Rail and Streetcar and the Pearl District – I think they are a little bit more appreciative of the base hits.

My old boss used to have this phrase: you can either swing for the fences or you can bunt to get on base. His point was there’s value in getting the progress that you can get.

You might look at Portland’s light rail system at that time, especially compared to other cities, and call it a swing for the fences, not a bunt to get on base. In some of the other places where we work, there is a greater appreciation for the small moves toward progress and celebrating those, whereas here, our expectations are high.

The other one would be the notion of creating great places that are attractive to tourists and new residents and business. In a lot of other places it’s not a given: they want to have a strong growing, thriving economy, and that means they work at being a place where businesses want to locate. In Portland, decades of growth convinced many that we don’t need to aggressively recruit and attract, compared with places that have a culture of hustling for growth.

Our job growth has really been driven by companies that are not in the city of Portland for the last 25 years. My sense is that a more diversified economy (across firms and sectors) is more stable than one driven by a few large employers. If you have fundamentals that serve as a flywheel, not only do you have a pretty resilient tax base, but the economic diversity and engine is much more resilient than ours here.

CLAUDEL: I’m curious about this characterization of bunting versus swinging for the fences. Firstly, do you think those two things are mutually exclusive? Or do you think a good economic development strategy has both going simultaneously?

LIENHART: I’m not a baseball person, but I know that good baseball teams win the World Series with both. You have to appreciate the contribution that each of them make.

I would reject a debate between big things and small things. We should focus on catalytic, transformative, impactful things. And sometimes those are going to be a series of small moves and sometimes those are going to be a big move. Many of the things that put Portland on the map in the seventies, eighties, and nineties were big moves. We’ve also done a great job attracting residents and visitors with a strong cultural identity that is made up of lots of small things – coffee shops, brew pubs, restaurants, bands...

CLAUDEL: When you talk about big things, do you mean in terms of development projects, or do you mean policy, or infrastructure?

LIENHART: All the above. Probably less on the policy side. I think some of the big moves we made on policy were probably more detrimental, including big tax changes that do not serve us well to this day. In terms of infrastructure, in terms of district planning and placemaking, a lot of those things we did are big moves.

I would also say that big moves are not always more complicated. When you talk about the policy, it’s about impact rather than scale: small changes can have a big impact when it comes to policy, good or bad.

CLAUDEL: Can you give an example?

LIENHART: Yes. On the face of it, proponents of inclusionary housing have a tendency to look at the impact in terms of how many affordable housing units the policy is producing.

I don’t think that the average proponent took seriously the idea that if you get the formula wrong, you could actually affect all of the market rate multifamily housing coming to a halt or being cut in half. 10% is the target for affordable units, and that feels small, right? But getting the formula wrong means that nobody builds any housing. Last year we had permit applications for 600 units. That’s a disaster. I’m not trying to simplify that particular policy question, but what I’ll say is unlike dollars and capital investment, where they tend to scale in parallel, getting a little policy wrong can actually have a really big economic impact, and a big impact to social and civic outcomes.

CLAUDEL: Is it fair to say that the ground floor retail requirement in zoning is another example?

LIENHART: Not in my opinion. I know some developers who really feel like if you’re building a five story building and you’re required to have the ground floor as commercial or active use, it’s a problem. They should call me and I’ll tell them how to address that in a more creative way.

CLAUDEL: They should!

LIENHART: I don’t mean to be arrogant, but I think if you look at the great walkable cities, the ground floor is always a fluctuating use.

It’s the shoemaker and then it’s the boutique and the cafe and then it’s somebody’s apartment, and then it goes back to being a cafe. Maybe we don’t have that level of construction, design, or use flexibility these days. But it is unlikely that there are a lot of buildings that would be going forward right now if not for the ground floor active use requirement.

The notion that we can’t do it is just as problematic as the notion that being forced to do it is a disaster. I would call for nuance and embrace the gray zone.

CLAUDEL: This idea is appealing. Let’s not be complaining about ground floor retail as a requirement, but instead think of it as a creative opportunity. What do we do with these ground floor spaces that could be retail one day and could be something else the next? On that note: you characterized a ground floor space as being quicker to change from one use to another, compared to upper floor spaces. Say more about that.

LIENHART: This is where I’ll make the distinction that you’re talking to a planner and not an architect. [A planner says] Oh, yeah, no problem. That’s simple.

There are some who would say no, a cafe needs a pretty specific set of design requirements. And these days, it probably does. There are probably a series of codes, and you would not want to make those investments if you weren’t going to do a restaurant or a cafe there. And there are certainly live-work units where nobody would ever put a store. Have we lost the ability to create big, huge, flexible spaces?

CLAUDEL: Creating a space that is amenable to many different uses – maybe slightly less optimized for any one of those uses, but which could be any of them – to me, that’s a less risky space to be owning and operating. I’d rather have that space than one that is perfectly designed and very pigeonholed. But it’s a planning concern too. This gets into zoning code and what we’re constraining or allowing in various districts.

LIENHART: There is a funny thing about the Pearl District. I’ve talked to people who are in real estate and have both developed and financed buildings. Some of them think the Pearl has too much retail and some think it has too little.

The Pearl is a great example of a place where there’s no main street. There are some people that love retail sprinkled all over the place. But from an efficiency of retail perspective and a pedestrian energy perspective, it might have been good to have one or two streets that people just know are the shopping streets. There are benefits from having one store after the next: you’re walking to get coffee and you pass Fjallraven and dip in there. I think we probably could have done that better. It is funny though that in the subsequent districts after Pearl, you still have the spread of retail everywhere, in the South Waterfront, and then Slabtown probably 10 years later.

I’m a big believer that if you have lots of people the ground floor will sort itself out – so long as you don’t preclude it through design.

By world standards, Portland is pretty low-density, and I think we’ve got enough room to adapt over the course of decades if we get it a little bit wrong. I’m very much not of the opinion that the planner has to be making sure we have enough of any given kind of space.

The planner’s responsibility is to make sure that we don’t have dead zones, blank walls that run 200 feet. Spaces that bring down the character and experience of being in an urban place or the perception of safety. It is much more about the experiential aspects of being in the city than the functional land use concern.

CLAUDEL: I agree that a planner shouldn’t be over-specifying in advance, and that a place should adapt over decades. If that’s true – shouldn’t the planner’s responsibility also be to enable the capacity to learn over those two decades?

LIENHART: Yes, in theory. We can all look back at a phase 10, 20, 30 years ago and recognize that it was a fad – every place has to have X. When we look back we realize it wasn’t necessary. You always have to be careful that you don’t decide something is so important right now, but find that it is totally useless or even counterproductive two decades from now.

If for some reason we planned a district and decided to focus all the retail on one street and then 10 years, 15 years from now, it became clear there is not enough retail in the rest of the district – we’re not developing so fast that we can’t correct that. The next developer is going to pick that up and realize there’s a real opportunity on the next block.

There was a debate about density on Twitter in the last two days. Somebody was saying that the right density is higher and another person was saying the right density is lower. And somebody else said maybe what we should do is have as few constraints as possible.

I said we shouldn’t impose a minimum density – and that’s a little controversial. Of course there’s a public interest argument that we probably shouldn’t put a one story building across from a major transit hub. But I do think if you have to fault, fault in favor of flexibility rather than faulting in favor of thinking you know what the prescription is.

CLAUDEL: I want to offer a third perspective, which is to say we can have rules based on expert-derived hypotheses about what the right solution is, but embed in those rules the capacity to learn based on what we’re eventually seeing. That’s not exactly flexibility. It’s adaptability. It’s a learning framework. Do you think that we have the capacity to do that today, at the institutional level?

LIENHART: Yes. My question would be, is it the policy and the framework of policies, guidelines, and regulations that’s adapting? Or is it the market that’s adapting?

I think a developer would tell you: if you give me a lot of flexibility, I will adapt. I will watch what works and what doesn’t on the street. The next project will respond to the last project.

On the other hand, if you say we want a flexible and adaptable policy framework then the question is: are you updating the policy as you see what works and doesn’t work? Or are you having some built-in discretionary function? Actually design guidelines and design review is a really good example of exactly what you’re describing. They specify a goal: we want active and vital streets, we want robust vegetation in the right of way, we want all transit modes to be accommodated. Developers, show us how you’re going to do that.

The way that today’s design commission interprets proposals may be completely different from the interpretations of the design commission 12 or 20 years ago. That’s the adaptation function. There’s a little bit of taste, a little bit of a reaction to the one we didn’t like or the one we loved that just got built.

A good example of this is above-ground structured parking. People used to think it was totally cool to just screen it with some architectural treatment or sculptural screening. Recently some commissioners have indicated a preference for programmed space at the facade with parking behind, so that when you look up, it looks like an active building, not a dark wall or open parking.

The rule didn’t change. Their interpretation of what created an active and vital public realm changed. You might say that’s the way we do adaptability today.

CLAUDEL: So you characterized generations of urban change. Now you see trends and workplace patterns and retail patterns – things like data centers suddenly being the most interesting target for development – these things are changing fast.

To me that says we don’t have the luxury of adapting our built environment on the generational scale, project to project. We need to be thinking about evolution within a project itself, having an in-built adaptation capacity.

Do you think that’s true? We can be concrete about it too. Portland, like many cities, has an extraordinary amount of commercial space. We have a lot of vacant retail space. Should it all be torn down? Can we find a way to do something with it?

LIENHART: My friend who made the point about the flexibility of ground floor spaces – I think he would say that buildings have been getting adapted for a long time. The buildings that we would most likely want to tear down – the ones that feel the least flexible, the historic building that’s not up to fire code, or has unreinforced masonry, or needs extremely expensive seismic retrofit – in many ways, those are the ones we could use most creatively if we don’t burden them with a whole series of requirements that they become a perfect 21st century building.

CLAUDEL: Are these the kinds of questions you’re contemplating in the Mayor’s Round Table? Where do you see that going?

LIENHART: So the Round Table is a relay team member for the Governor’s Downtown Task Force. The Task Force essentially established a value proposition for the central city: who cares about the central city, why is it important, and what are the things that we want to get out of a thriving central city?

That broke down three broad categories. One: it’s a place to gather arts, culture, and civic activities. The second was that it’s a place of innovation: work, education, innovation between people. The benefits of proximity. Third, it’s a great place to live because it has a concentration of services and amenities within walking distance. There has been huge investment in transportation that helps you get everywhere else you want to be more easily than any other place.

There was a handoff to the Mayor, and now it’s the mayor’s Round Table. He has framed it as an implementation table. Great work, Task Force – we agree on the value proposition, we agree that these are important strategies – now: what do we need to do to implement these strategies?

That will happen in concert with the other plans that the City has that address the central city. We have transportation plans, we have Advance Portland. There are a series of other plans, and in some cases it’s going to be about getting all the right resources to the table to do the things that we need to do.

In other cases it may be about political will to create policy change and make some of these things happen. The group that’s been assembled is smaller than the Governor’s task force. It is focused on people who are leading organizations that are deeply invested in the success of the central city.

We have an opportunity to lend our voices to all those initiatives. There may also be a role of coming up with the big idea, but I think we’re not going to be sitting at that table asking so what should we do? as much as we are going to be saying, how do we do it, or how do we get there? What does this catalytic idea need to be successful?

Portland, as a city, is not short on ideas. We are a city that needs alignment, needs investment. Pulling us all together on a regular basis is intended to help strengthen the alignment and accelerate investment.

Now on the big questions of how to adapt for the future – some of them may be topics that this table can grapple with. But I would also say the Mayor’s focus on key performance indicators is a reminder that he really wants the table to be action oriented. And accordingly there will probably be a little bit less focus on the things that take a long time to figure it out or even to identify the scope of.

CLAUDEL: Coming full circle: we need the quick wins. We need to bunt and get to first base.

LIENHART: We will be leaning into several different issues – but what “leaning into” looks like is going to be different on different topics. Somewhere it might be writing a letter. Somewhere it might be helping to go recruit a business or solicit the legislature for funding.

The intent is that we are dynamic and able to be responsive to anything that moves the needle on those key performance indicators.

CLAUDEL: Thank you, Nolan. I’m excited to see what comes of the Round Table.

LIENHART: I look forward to hearing the outcome of all your conversations. These are good questions.

Subscribe now

This is the seventh in our Adaptive Urbanism series. Each week, we go deeper into the challenges and the responses. Last time, we examined the institutional imperative — why adaptive urbanism requires a new kind of steward, and how the interests of public, societal, and private capital might be bound together by design. Today: why the capital markets that should be funding the adaptation of our cities are instead walking away from it.

Although the majority of meetings are now video calls, we have all had in-person meetings over the past two years. For me, one stands out, in particular. I parked in the building, walked into a vaulted lobby with a manicured green wall, and I was met with a crisp greeting from uniformed staff. I craned my neck to see the well-appointed fitness center before shooting up to the right floor. The floor was bustling with tenants. We used a conference room that was perfectly provisioned with technology that worked effortlessly. My host showed me the private, landscaped roof terrace on my way out, where tenant-only yoga was about to start.

Another meeting stands out, too. I circled the building a few times before parking several blocks away. There was a for-lease sign on the ground floor retail space, and the lobby was dingy, dusty, and desolate (the primary decor was a bright yellow “wet floor” A-frame sign). The elevator crept up to a floor where lights flickered over dated carpet, drop ceiling tiles crumbled, and my host – the only lessee on the floor – had to apologize for the stuffy air. “Something’s wrong with the HVAC.”

These two meetings happened in the same city, maybe even on the same block. Same urban fabric, same planning authority, but worlds apart.

The first building is filled with creditworthy tenants on long leases, and the landlord is backed by institutional capital. The second building is a perfectly good structure, well-located relative to transit, but facing vacancy and decline. This is a national trend, and the data is clear, but it takes boarded windows to make the reality visceral: while Class A offices absorbed 5.2 million square feet of new tenancy across the United States in 2025, the rest of the market lost occupied space in the same twelve months.1 The buildings filling up and the buildings emptying out are often on the same block.

This is not a market failure in the conventional sense. The capital markets are functioning exactly as designed. The problem is that they were designed for a world that no longer exists.

The commercial real estate market has quietly split into two domains. The distinction between them is the most important thing happening in real estate finance, and almost no one is talking about it clearly.

System A comprises the specialist assets where traditional finance continues to work: high-quality amenity-rich offices, data centers on twenty-year hyperscaler contracts, life sciences campuses, logistics facilities at critical supply chain nodes, prime towers leased to anchor tenants for whom the address is a business signal. Demand is definable, tenants are creditworthy, and leases are long. Discounted cash flow works well.

System B is everything else, comprising approximately 70% of commercial stock. Class B and C offices, aging retail, downtown mixed use, mid-rise buildings in outer neighborhoods, suburban strip malls. These buildings are not obsolete enough to demolish and not specialist enough for System A capital.

The crisis is in System B. These buildings could adapt, but adaptation is rare – it takes capital and creativity, and an appetite for risk, in situations where traditional investment models don’t work.

The numbers are stark. In the United States alone, more than 217 million square feet of office leases are scheduled to expire in 2024 and 2025, and over 800 million by 2028: the largest wave of tenant decisions in commercial real estate history (CRED IQ). Portland’s downtown office vacancy rate of 34% is among the highest of any major US city. Much of that space will not be re-leased on existing terms. Whether these buildings adapt or atrophy – and ultimately, whether Portland’s downtown survives or withers – is primarily a question of how investors model the return on capital.

Many System B buildings have adaptation potential; comfortable floor plates, tall ceilings, regular column grids, overbuilt structural engineering. The Five Oak building near Big Pink (a building we’ve looked at before) is a good example. The building could become an upgraded office, residential, ground floor retail, early childcare, self-storage, advanced manufacturing, or a mix of several uses, depending on decisions nobody has yet made.

Adapted System B assets can be profitable as these uses fulfill local needs. In a normal stable market capital flows toward profit – but that isn’t happening. Why? It’s a measurement problem. The instruments designed to value and finance these assets were built for System A, and applying them to System B produces not merely imprecise answers but systematically wrong ones.

A standard discounted cash flow (DCF) model assumes you can project rental income with confidence, calculate a terminal value based on a use that persists, and apply a discount rate that reflects quantifiable risk.

All three standard DCF assumptions break simultaneously. The analyst cannot agree on stabilized occupancy, lease-up period, or $/sf lease rates. The terminal value becomes speculative at best. The discount rate is chosen to produce a comfortable number rather than something derived from analyzable risk. The Swickard Group bought the building at a steep discount, based on intuition, grit, and a strong belief in Portland – not financial projections.

“Every great city goes through its hard seasons,” Swickard said. “We just have to decide if we want to lean in or not… In the case of what we’re doing, we’re leaning in and we’re not trying to time the market. We just have a long-term belief in Portland and the people of Portland and Portland’s identity,” (quoted in KOIN).

DCF, applied to adaptive assets, systematically undervalues them. A building capable of pivoting between uses appears, through the DCF lens, as a building with uncertain income; a risky building deserving a lower value. Its flexibility – a strategic advantage that makes it valuable – is coded as uncertainty that makes the building dangerous. From a traditional investor’s perspective, dangerous buildings don’t deserve additional capital; they should be held indefinitely, or sold. The result is that adaptation never happens.

To paraphrase Antony Slumbers: we are valuing buildings as bonds in a world where they are options. The bond model assumes a fixed income stream, known in advance. The options model applies methods used in financial services to assess the prospective future value, at a fixed point in time, of an asset under different conditions. The owner may have multiple options at multiple junctures; each will have a different cost and future value. Under genuine uncertainty, option value is not marginal. It is a building’s strategic advantage.

This inversion matters enormously. If uncertainty increases the value of adaptive capacity – and it does, by the same logic that makes financial options more valuable when volatility rises – then conventional valuation methodology points in entirely the wrong direction.

The mispricing creates an opportunity.

Some investors understand System B as a distinct asset class: not a failed version of System A but a different game with different rules. They see the value in buildings acquired at significant discounts to replacement cost, in locations with genuine underlying demand, capable of serving uses that System A frameworks cannot. They have the capital and creativity needed to play the optionality in a strategic way.

Jeff Swickard is not alone – we are seeing this strategy starting to emerge in Portland. Buildings are transacting for pennies on the dollar: The Merchant Bank Building (80% of prior sale), Downtown Portland Marriott (63% of prior sale), Gus J. Solomon U.S. Courthouse (sold for $1.8 million), and of course Big Pink (12% of prior sale).

The arbitrage potential is extraordinary. Returns are imminently achievable, as Portland’s downtown rebounds – and that will depend on these buildings coming to life, and a generation of others following suit.

The instruments for that transformation are missing, however. Four things need to change simultaneously.

Valuation must evolve. Real options theory offers the alternative to standard DCF models: they assess a building’s total value as the intrinsic value of its current use plus the option value of its potential futures. That option value is calculable – it is given by the range of plausible uses, the conversion cost for each, and the returns for each, with a volatility factor based on timing and depth of demand for each. Buildings with low conversion friction and broad demand diversity carry substantial option value that conventional appraisal simply cannot see. Unfortunately, this methodology is not mainstream amongst CRE professionals (Field States is solving that problem with a light software package that runs a heavy math engine for DCF + Real Options Modeling – to be formally announced soon).

Debt must accommodate transition. A standard commercial mortgage assumes the building will serve the same purpose for ten, fifteen, or twenty-five years. For System B assets, that assumption is the problem. Covenants built around debt-service coverage ratios calibrated to a single use trip into technical default precisely when an asset is doing what adaptive finance should enable: moving between uses. Adaptive debt structures accept that metrics fluctuate during transitions and treat planned adaptation as a strategic moment of operational transition, not distress. Again, these products are not mainstream, but Field States is helping to define adaptive capital stacks that blend the most appropriate instruments.

Equity must participate in the adaptation premium. When a district improves through strategic repositioning and active stewardship, values rise. Under conventional structures, that premium accrues to whoever holds the deed. Adaptive equity, whether patient capital with longer carry, shared appreciation arrangements, or the Perpetual Purpose Trust structures we discussed in the last post, is designed to participate in that uplift and recycle it into continued stewardship. The flywheel only turns if equity captures what it helped create.

Public capital has a role that is neither subsidy nor charity. Government should act as risk-bearer of first resort, through first-loss positions, guarantee structures, or development finance that accepts non-standard returns, creating the conditions under which private capital can enter System B at the scale the problem requires. This is co-investment in genuine uncertainty, structured to share both risk and reward.

Blending these non-standard approaches into a System B-specific ‘adaptive capital stack’ is challenging. Investment in this asset class requires building new financial muscles. Those investors who can master it will have significant competitive advantage as cities lift out of doom loop spirals.

The four instruments outlined above constitute what we call Adaptive Finance: new valuation methods that see option value, new debt instruments that accommodate transition, new equity structures that participate in the adaptation premium, and public capital that de-risks the entry point without crowding out the market. Together, these can create a capital stack designed for System B on System B’s terms, rather than System A instruments applied to buildings that have already been left behind.

For the past five years, pundits across the country have framed Portland’s downtown as a cautionary tale, the poster child of the doom loop. We believe it can become a proving ground for a new kind of real estate finance. We’re ready for adaptation.

In the next installment, we’ll turn to the buildings themselves and the need for an adaptive design imperative.

Subscribe to Field Discovery to follow the series.

Subscribe now

In our last essay we established the cognitive imperative: cities must learn to see clearly and act before decline becomes irreversible. Here, we define the institutional imperative: why our institutions of urban governance struggle to match the pace of change – and what should replace them.

Adaptive Urbanism begins with the Pacing Problem: the shear between the pace of social, economic, and technological change and the pace of urban action. The Pacing Problem is most visible in high-profile development projects that are dead on arrival. But it is also structurally encoded in the institutions we rely on to govern district-scale development.

Consider who creates value and who captures it in a major urban transformation project. Residents, small businesses, local organizations, and civic institutions generate value over decades, through economic activity, social networks, identity, and the slow accumulation of livability that makes a neighborhood worth investing in. Governments capture value at two extremes: political wins in the short term, property tax revenue and urban vitality in the long term. Developers capture value in the five-to-seven year window defined by their IRR projections – beyond that horizon there is no value worth observing. Each actor operates on a different clock. Each bears a different kind of risk, exercises a different form of responsibility, and seeks a different reward.

A common institutional mechanism for aligning these interests is the Community Benefits Agreement (CBA). A CBA is negotiated at the very beginning of a development process that will unfold over five, seven, ten or more years. It attempts to specify, in advance, the benefits a community will receive in exchange for endorsing (or at a minimum, agreeing not to oppose) a development project.

CBAs have generated positive outcomes in communities across the country, but the mechanism is also structurally flawed. It is adversarial by design: each stakeholder advocates for their own interests, and the resulting agreement reflects negotiated concessions, not shared purpose. Once signed, the community is functionally outside the project as anything but a legal claimant. If the developer fails to deliver, enforcement falls primarily to the coalition, and legal action it must fund itself. Worse still, CBAs are anticipatory, written before anyone can know what the project will actually require or produce at each phase.

This is anticipatory and compartmentalized governance applied to a process that demands continuity and adaptation. The CBA asks the right question – how can community interests shape a large-scale development project that still pencils? – but answers it with an instrument designed for a world where the future can be specified in advance. It cannot account for what only becomes knowable as a project unfolds: how risk and reward shift as conditions change.

The institutional imperative is a call to build governance models with structures that align stakeholders’ inputs and project outputs across the full lifecycle of urban transformation. Each contributes what they are best positioned to contribute, each is accountable to the others, and value is defined and created phase by phase rather than negotiated once and enforced from outside.

In 2015, twenty-five organizations gathered in Amsterdam to sign something unusual: a manifesto.

The Circular Buiksloterham Manifesto tied together the City of Amsterdam, housing corporations, utility companies, developers, and community groups to a shared framework for the redevelopment of a post-industrial district on the far bank of the river IJ. It specified not outcomes but principles: circularity in material flows, self-sufficiency in energy, shared infrastructure, community governance, transparent monitoring across a twenty-year horizon. The signatories had no legal obligation to each other. What they had was a shared metabolic understanding of the district – the product of a comprehensive mapping of its energy, capital, water, and people flows – and a commitment to act within principles rather than toward a predetermined end.

The Manifesto created something planning rarely achieves: cadence. A shared calendar around which diverse actors could align. When review points came around, participants prepared. When commitments came due, the accountability was social before it was contractual. The district coalesced around a constitution rather than a masterplan – establishing the conditions within which multiple futures could be navigated, rather than optimizing for a single predicted one.

Eleven years on, Amsterdam Noord looks nothing like a masterplanned community. De Ceuvel – a contaminated shipyard – became a creative workspace through phytoremediation and near-zero energy design. Schoonschip grew into a floating neighborhood of forty-six households that own their shared grid, their heat exchange, their own governance. NDSM Wharf blossomed with cultural life alongside the Buiksloterham area. None of these outcomes were predicted and managed as part of an optimized masterplan; all of them developed from a emergent understanding of what the place actually was. Together they created some of the most valuable property in the city. Accountability here runs between co-signatories, not upward to a single owner, and that is precisely why it flourished.

On the other side of the North Sea, a different kind of institutional innovation has been taking shape; quieter, less theorized, but pointing toward the same conclusion from a different direction.

In 2012, Vastint acquired twenty-six acres of derelict industrial land in Stratford, east London, and made a commitment that most property businesses structurally cannot: to develop it across a fifteen-to-twenty year horizon, retaining ownership throughout, with no forced sale point and no investor redemption pressure. Vastint – a property company within the Interogo group, the holding entity connected to IKEA’s founding family – is not a developer in the conventional sense. It does not raise external funds with defined return periods. It does not answer to shareholders with quarterly expectations. Its capital is patient by design.

What that patience makes possible is visible in Sugar House Island today. Over six hundred residents in a development that remains a work in progress, but is shaping to become a vibrant community, including a two-form entry primary school and the UK’s first national Talent House for urban dance and music. The latter is co-funded with Arts Council England, the Greater London Authority, and the London Borough of Newham – a piece of cultural infrastructure whose costs and credits are shared across a genuine coalition. Nine different architects were commissioned to deliberately prevent the homogeneity that single-developer schemes typically produce. Retail is filled with a curated selection of independent businesses rather than chains, because the long-term economics of a thriving district beat the short-term economics of a higher-rent chain occupier.

This is not corporate philanthropy from a socially democratic benefactor. It is a commercial investment thesis, held with unusual consistency: that patient capital committed to genuine placemaking compounds more value over twenty years than capital optimized for three-to-five year returns. The yields are real, and, although the returns are slower, they stand to be substantially larger. What Vastint demonstrates is that private capital, structured with the right time horizon and freed from the extraction pressure of short-cycle finance, can behave like a steward – curating, investing, absorbing setbacks – and still deliver commercial returns that vindicate the model.

Neither case, on its own, fully resolves the institutional problem.

Buiksloterham distributes accountability across a genuine coalition of interests. Its “manifesto” established governance that is constitutional rather than prescriptive: setting principles within which multiple futures could unfold rather than optimizing for a single predicted outcome. But the Manifesto is voluntary. Its twenty-year ambition rests on continued alignment without a formal mechanism for capturing and recycling the value that district improvement generates. And its terms were set at the outset: an act of collective vision, but still an anticipatory one.

Vastint has patient capital and a genuine long-horizon commitment. Its financial structure eliminates the forced exit that makes many conventional developers structurally incapable of stewardship. But the community’s interests are served at the owner’s discretion, not by governance design. Residents are tenants, not stakeholders, and the value the district generates flows to the owner. The benefits are a product of good intent rather than institutional obligation.

Both cases point toward the same conclusion from different directions: long-horizon urban transformation requires a structure in which public institutions, community stakeholders, and patient private capital each contribute what they are best positioned to contribute – regulatory stability and public investment from government, legitimacy and local knowledge from community, financial commitment from capital – and remain accountable to one another across the full lifecycle of the project. Accountability is not externally enforced through litigation, but built into governance: peer-to-peer, ongoing, and, yes, still structurally enforceable.

The Oregon Museum of Science & Industry brought in Field States to run a community listening process that surfaced ideas and aspirations to shape the district’s future. What the process uncovered was a more fundamental question: whether the CBA model was the right vehicle for translating community aspirations into the governance of a development project at all.

Together with OMSI’s leadership, we designed an alternative we call the Civic Value Framework. Where a CBA is confrontational, the CVF is purposeful, grounded in values shared by the developer, landowners, residents, and civic organizations. Where a CBA is anticipatory, attempting to specify benefits at the very beginning of a decade-long process, the CVF is emergent, designed to discover, prioritize, and enact opportunities for civic value at every phase of development: planning, construction, activation, and ongoing operations. Where a CBA is zero-sum, premised on negotiated concessions, the CVF is oriented toward creating shared value. And where a CBA leaves the community outside the project after signing, the CVF keeps stakeholders inside governance throughout, with a seat at the table in the ongoing definition of what civic value means and how it will be created.

Practically, this works through phase-specific committees of project partners, organizations, and community members, each articulating what civic value looks like for their phase, all contributing to the overarching project vision. Phase-specific negotiations are smaller, lower-stakes, and grounded in emergent reality rather than speculative projection. The project partners commit capital and in-kind investment to create civic value, and are held accountable by the committee and the public. In 2023, Portland’s Design Commission approved the OMSI District Master Plan with the Civic Value Framework as a component.

The CVF addresses governance: how stakeholders align, define civic value, and remain accountable across the lifecycle of a project. There is an important complementary question: how capital itself is structured to serve that governance over decades rather than extract from it. The institutional imperative requires legal vehicles that bind patient capital to long-horizon stewardship. At Field States we are innovating in the application of a specific tool – the Perpetual Purpose Trust – to real estate.

A PPT is a legal entity organized around a defined purpose, maintained by a stewardship committee, and policed by a trust enforcer. It is capable of owning and operating real estate with the flexibility of an LLC but the long-term accountability of a mission-driven institution. PPTs gained global visibility as part of Patagonia’s 2022 ownership restructuring; closer to home, Oregon’s Organically Grown Company has demonstrated the model’s versatility. The Kensington Corridor Trust in Philadelphia pioneered the application to real estate. At Field States, we are building on this work, developing PPT structures where patient capital and effective governance are bound together by design.

The institutional imperative, then, is a call for better structures. The built environment demands governance that is adaptive, not anticipatory, keeping stakeholders inside the project; capital that is patient by design, not by the disposition of a single benevolent owner; accountability that runs horizontally among co-invested parties, not vertically to a single authority, and risk distributed with intention. We are building these structures, and working to prove, project by project, that they work.

In the next post in this series, we will turn to the economic imperative: how capital must be structured to flow toward adaptive assets rather than away from them, and why the current financial architecture systematically penalizes exactly the flexibility that cities most need.

Subscribe now

Markets as information systems

The meteoric rise of betting platforms like Kalshi and Polymarket has prompted an important public conversation about the role – and the legality – of markets as information systems. These high-profile companies use crowd intelligence, via bets, to predict the outcomes of any (truly any) decisive future state.

But crowdsourced prediction is not the only way that markets can convey information. A classical definition frames a market as a tool for aligning supply with demand, using price as the basic “unit” of information (a unit that accounts for material and labor costs, scarcity, and time). “The so called ‘market mechanisms’ of capitalist economic theory can be regarded in ideal form as self-regulatory cybernetic feedback systems implementing the collective aspects of the preferences of individual decision makers,” according to Campbell (1999 p.19).

Markets are effective, but they aren’t the only self-regulatory cybernetic feedback system for that kind of information. In the mid-twentieth century, a diverse field of political economists, technologists, management theorists, and revolutionaries were all asking variations of the question: is it possible to design – and actually construct – a non-capitalist solution to the complex information problem of supply and demand?

This week’s post is animated by a charismatic protagonist, unfolds in post-revolutionary Chile, and rises to a climactic scene set in a room that can only be described as “Bond villain vibes”. The story is fun, of course, but it is also instructive. Cybersyn, as it was called, has lessons for large-scale information management problems, even (and especially) within a market dynamic like real estate or urban data analysis. This is a historical perspective on the Cognitive Imperative of Adaptive Urbanism.

A cybernetic experiment

“[We are] now in a position from which it is possible to implement on a national scale – at which cybernetic thinking becomes a necessity – scientific views on management and organization,” stated a 1971 letter from Fernando Flores, technical director of Chile’s new economic development agency CORFO. The letter was addressed to Stafford Beer, a larger-than-life management consultant from the United Kingdom, and it included an invitation to Chile. It probably would not have taken a national-scale implementation opportunity to get Beer, onto a first class flight to Santiago, but an opportunity of this magnitude was impossible for him to ignore.

Stafford Beer was a unique kind of management consultant. After studying philosophy, he served in the UK military, where he gathered observations of efficient command under conditions of uncertainty. In the career that followed, Beer merged scientific operations research with practices for organizational learning, and embraced the concept of cybernetics from computation. “Cybernetics” refers to feedback loops and control mechanisms that enable a system to act, learn, and adapt in complex, changing environments.

By the time he received the letter from Chile, Beer had become known for his work consulting with companies, books, and lectures. But Flores was describing a once-in-a-lifetime opportunity: to save a turbulent socialist economy with technology.

On November 4, 1971, Stafford Beer left behind a stagnant UK – gray weather, conservative politics – for the heat of a Chilean summer, still sparking with revolutionary energy. With the imprimatur of Chile’s government, Beer was tasked with designing and implementing a technical system for managing Chile’s industrial ecosystem without a capitalist market.

The task was not theoretical. When Salvador Allende was elected Socialist president of Chile in 1970, his agenda centered on nationalizing hundreds of companies and incorporating worker participation in their operational management. As foreign managers fled and workers seized factories, there was a sudden lack of managerial expertise. Allende’s government faced the challenge of maintaining industrial output, without price signals or competitive feedback.

Chile was not the only Socialist country grappling with the challenge. At the time, “Cybernetic Socialism” was an area of study, but real solutions had never been attempted at scale with the cutting edge technologies available. For Beer, it was an unprecedented opportunity to advance the theory and practice of cybernetics.

Project Cybersyn

In under a year, Beer and his collaborators designed a technical nervous system for the Chilean economy and named it Project Cybersyn. It had four components. Cybernet linked factories to a central communications hub via a repurposed network of telex machines. Cyberstride was statistical software for processing factory data and modeling policy impacts. Cyberfolk was a system for continuously aggregating citizen sentiment, using a simple happiness dial in every home (it looked like a thermostat, and enabled citizens to share their satisfaction with the government in real time). The Opsroom was a futuristic hexagonal chamber with swivel chairs, toggles, and wall-mounted displays, offering total immersion in economic data.

Practical success never measured up to Stafford Beer’s conceptual ambition. The Cybersyn suite was something between a prototype and film prop – Cyberfolk was never implemented, and the Opsroom was an intricate configuration of false walls and manual data entry.

Cybernet offered some practical value during a CIA-backed factory strike, where the telex network allowed the administration to monitor scarcity in real time and deploy support from loyalists strategically. They moved faster than the operatives trying to destabilize the industrial ecosystem, and the strike was quelled. Cybernet was helpful as an information network, but it was far from performing as a cybernetic economic nervous system for connected parts to share information and adapt.

Pinochet’s US-backed coup came on September 11, 1973. With the presidential palace in flames, Allende dead, and tanks surrounding CORFO, team members fled. When soldiers finally reached the Opsroom, they stabbed each screen with a knife.

A political theory of information flows

Beer’s experience in Chile transformed him from a management consultant to somewhat of a political theorist, convinced that cybernetics was more interesting as a theory of continuous institutional adaptation than as an efficient management technology.

He went on to develop ideas about sociotechnical systems operating at the intersection of bureaucracy and technology, asking questions like: How does technology deliver relevant information about supply and demand? Who makes decisions? Is the system structured for stasis or for continuous testing, learning, and adaptation?

These questions are no less relevant today, half a century later, and in a very different political economy. Cybersyn aimed to replace the market with a technical information system. What if a technical information system could enable a market to learn and adapt toward positive civic outcomes?

Urban planning and development systems worldwide are built around stocks: zoning allocations based on projected housing need, transport models fed by decennial census data, economic strategies keyed to employment statistics that arrive months after the conditions they describe. Data on stocks, gathered in periodic snapshots, describes what the city was, not what it is becoming. To measure continuous flows is to see the city clearly, sensing change as it happens and revealing emerging conditions before they cascade irreversibly.

Cybersyn aimed to capture flows of information and use them for real-time management of the industrial ecosystem. In socialist Chile, control was centralized – but real-time information on flows would be no less valuable in a market context.

Local real estate is a good example. In the United States, more than 217 million square feet of office leases are scheduled to expire by 2028, the largest wave of tenant decisions in commercial real estate history. Yet the data describing this shift remains fragmented: municipalities hold property records, developers hold pipeline data, a handful of brokers hold transaction intelligence, platforms like CoStar and Loopnet have listings, and data does not flow freely between any of them. Vacancy figures are self-reported, methodologically inconsistent, and often months out of date. Transactions are (sometimes) recorded when deals close. There is no shared infrastructure for tracking occupied and unoccupied space in real time – and this is exactly the information that property owners need to make critical decisions about investing in building conversion or upgrades, selling, or exploring meanwhile uses.

The goal of a contemporary cybernetic information system should not be to eliminate real estate markets but to make them function adaptively. What might be possible with a comprehensive data infrastructure that treats real estate information as a shared civic resource? This kind of information would sharpen competition among private firms, shifting competitive advantage from relationship capital to analytical capability. It would enable experimentation with nimble supply-side models that meet emergent demand – models that would be rewarded with activity and revenue. Governments would gain the visibility they need to make proactive decisions about land use, incentives, and housing, rather than reacting to crises already underway. There are no meaningful technological barriers to this kind of data. The primary challenge is public-private collaboration and data stewardship.

The city that can see itself clearly – the city that has built recursive feedback between what its buildings are doing and what is needed of them – is one capable of generating real civic value. The legacy of Project Cybersyn may not be in replacing markets, but in laying a sociotechnical foundation for understanding what is possible when actors have the information they need to act well within a market.

“I said before that there are solutions, but I have also shown that they concern organizational modes. They concern engineering with the variety of dynamic systems. By continuing to treat our societary institutions as entities, by thinking of their organizations as static trees, by treating their failures as aberrations – in these clouded perceptions of the unfolding facts we rob ourselves of the only solutions.”

— Stafford Beer, “The Real Threat to ‘All We Hold Most Dear,’”

CBC Radio Broadcast, 1973

Subscribe now

For the rest of March, you’ll see something curious if you walk past the PacWest tower on a weekday morning. There is a temporary coffee kiosk on the sidewalk in front of a vacant retail space. It is impossible to miss the irony: the vacant space was previously a coffee shop.

For years, the space has been listed for lease. The ground floor retail on the other side of the building was previously a restaurant – it’s empty too. A block away, the storefront at 1407 SW 4th Ave has been listed on the market for 1114 days. Some of these spaces, like the retail suites at 1515 Market Square, are fully built out. Others are all raw concrete and gravel floor.

Without a comprehensive understanding of what is going on at street level – block-by-block market dynamics, attractors and detractors, demand patterns that make a kiosk viable and a retail space unleasable – cities will be entirely unable to respond effectively.

Of course this is a data problem, but there is a deeper issue at stake. Before cities can act adaptively, they have to see and understand adaptively, and most of them don’t. Building this capacity is the Cognitive Imperative of Adaptive Urbanism.

The majority of urban datasets we use are lagging indicators. In a slower world, they worked. Census data is refreshed every decade. Economic statistics aggregate away the district-level patterns that matter. Vacancy figures, where they exist at all, are compiled from broker records, landlord disclosures, and tax assessments, each held by a different actor, each lagging real conditions by months. Governing a city this way is like trying to drive a car using only the rear-view mirror: the picture is accurate, it is just describing somewhere you have already been.

These instruments are too slow, too fragmented, and too static for the kind of change cities are now trying to govern.

Urban governance has been built around stocks: how much office space exists, how many housing units, what the vacancy rate is, what the assessed value is. These measures tell us what is there. They say almost nothing about what is happening.

That distinction matters more than it might seem.

A district at 20% vacancy that was at 10% a year ago is in an entirely different condition from a district at 20% vacancy that was at 30%. The stock is identical. The trajectory is not. One is tipping. The other is stabilizing. If you only look at the snapshot, you miss the motion. In urban systems, the motion is almost always the story.

Adaptive urbanism starts with a shift from measuring stocks to measuring stocks and flows.

Not just how much space exists, but the velocity of take-up and release. Not just the number of empty units, but how long they have been empty, in what pattern, and with what knock-on effects. Not just whether a district is underperforming, but whether it is deteriorating, stabilizing, or beginning to recover. In many cases, the difference between these conditions is visible well before the official data registers a trend: early enough to act – if anyone is looking for it.

This is where most cities struggle. No single entity holds the full picture. Broker data captures listings, not occupancy behavior. Landlord information is private. Public records lag. Utility usage, footfall, business formation and closure, lease roll – each actor holds a slice. The city, meanwhile, is expected to infer causality from a patchwork of incomplete signals assembled after the fact. That is a poor basis for policy.

The task is not simply to collect more data and hope for enlightenment. That way lies dashboard theatre (no offense to smart city tech UX designers out there). The task is to build a genuine capacity to: sense, interpret, and respond.

Sense means detecting change earlier – at the level where the trajectory is still visible, before the threshold is crossed. A street with 15% vacancy and rising is a different intervention target from one with 15% vacancy and falling. The signal is in the direction, not the number.

Interpret means distinguishing signal from noise. A spike in vacancy could indicate structural distress, or it could be lease rollover, planned retrofit, or a data artifact. Better intelligence means better judgement: not automation for its own sake, but the capacity to ask harder questions about what the evidence actually shows.

Respond means connecting perception to action within the timescale of the problem. This is where the failure usually lives. A city may understand, in broad terms, that a district is weakening. But by the time the data confirms it, the conversation has already hardened into crisis management. The city is no longer steering; it is chasing.

And when urban decline begins to spiral, that lag is expensive. Vacant units reduce foot traffic. Reduced foot traffic weakens neighboring businesses. Falling revenue reduces investment. Reduced investment undermines confidence. Once that loop takes hold, better data cannot solve the problem.

The Cognitive Imperative is not just a call for better data, more data, or a smarter dashboard. It is a call for a different kind of urban intelligence.

That means combining public and private signals rather than treating them as separate worlds. It means working at the district level, where thresholds become visible, rather than relying on citywide averages that smooth away trouble until it has compounded. It means asking not only “what is the current state?” but “what is changing, how fast, and in which direction?”

Cities that build this capacity can see a district tipping before decline hardens into structure. They can distinguish a passing fluctuation from a structural shift while the distinction still has practical consequences, and can allocate intervention where trajectories, not just conditions, warrant it.

None of this eliminates politics. Better cognition doesn’t tell a city what it should value, or how to allocate scarce resources, or which neighborhoods deserve priority. It does something both more modest and more important: it helps a city see early enough to choose. Most cities are ready to act. The first failure of urban adaptation is that they do not know enough, early enough, to act well.

In Portland, as in many cities, the debate about downtown oscillates between denial and fatalism. Either recovery is just around the corner, or the old model is gone and nothing can be done. Both positions are too tidy. The harder truth is that cities in transition cannot navigate structural change using yesterday’s categories. They have to learn their way forward – and learning begins with seeing and understanding.

A city may see understand what is happening and still be unable to respond. Data does not change zoning, and capital still needs to flow. Insight does not create institutional authority. Someone with mandate, flexibility, and staying power sufficient to matter across the timescale of the problem still has to act. The next imperative of Adaptive Urbanism: institutional.

Subscribe to Field Discovery to follow the series. In the background, we are working on a more detailed ‘Practitioner’s Guide’. We will be making it available to subscribers.

Subscribe now

As the Titanic was sinking, Engelhardt Ostby made sure his only daughter, Helene, got a place on a lifeboat, and heroically went down with the ship. Helene survived clutching tightly to a briefcase that contained precious stones and original molds for Tiffany & Co.’s diamond solitaire ring.

Upon her return, Helene became a co-owner of her father’s firm, Ostby & Barton, a Providence jewelry company. Today, Tiffany still uses the molds—a continuity that spans a shipwreck and 100 years of craftsmanship.

This week, we explore manufacturing identities beyond Detroit and Rochester, diving into Grand Rapids and Providence. These cities were shaped by networks of concentrated production, iconic brands, and distinctive architecture.

Providence

Whether it was the burgeoning Hollywood film industry, the expanding middle class, the ability to inexpensively mass produce replicas of famous heirloom pieces, or just Coco Chanel’s influence, 20th century America had a tremendous appetite for costume jewelry—and Providence satiated it.

Providence was the undisputed capital of American jewelry production for most of the twentieth century. At its peak in 1978, over 32,500 workers produced 80% of America’s costume jewelry, shipping 1 million pounds of costume jewelry a week.

The Coro Building was the largest costume jewelry factory in the world, and Coro just was one company among dozens. Unlike cities with a small number of dominant companies (like Detroit and Rochester), the majority of Providence jewelry businesses were small independent operations with 25 to 100 employees. They were served by a number of specialized suppliers, from jewelry-box makers to precious and semi-precious material suppliers. Firms collaborated and competed ferociously within the small jewelry district. Workers moved between firms, carrying techniques and trade secrets. Artisans and companies made innovations such as the first rolled plated gold process, “seamless-filled wire” for chain making, and electroplating—and within months, an innovation could spread from one company to the entire district.

The Providence jewelry industry was concentrated in a 19-acre industrial district of small brick warehouses—buildings that were designed for flexibility. The A.T. Wall Building, completed in 1910, pioneered mushroom-column flat-slab construction, with open floor plates that could accommodate whatever manufacturing process a tenant required.

Foreign competition undercut American companies, but decline came unevenly. Some firms failed while others adapted. Independent designers still work in the district today—using the same plating shops and specialized manufacturers that served the major brands decades ago.

Providence’s buildings were built for adaptation, and they continue to adapt. Today, the old jewelry factories house design studios, technology firms, and medical offices. The network logic that made the district function—many small actors, tightly interconnected, sharing infrastructure and knowledge—is re-emerging as a crucial factor.

Grand Rapids

Since the late nineteenth century, Grand Rapids, Michigan has called itself “Furniture City.” When Grand Rapids earned its name, furniture was primarily made of wood, hand-carved in a decorative style. The city was home to a deep talent pool of Scandinavian immigrant woodworkers who had settled along the Grand River and spent decades refining the craft of cabinetry. By the end of the 1800s, the city led U.S. furniture production; 47 furniture factories employed one-third of all Grand Rapids workers.

It was such a powerhouse, and its name was so synonymous with quality, that in 1902 the local furniture guild began using “Grand Rapids” as a kind of trademark—any local tradesperson could imprint the moniker and logo on their furniture. Its credibility was even enforced through multiple lawsuits, and industry-specific trade publications like the Grand Rapids Furniture Record exposed violators as frauds.

At its 1925 zenith, the Grand Rapids Furniture Market operated 10 exhibition halls with 560 dealers. This is where America came to find its furniture—particularly ornate Victorian residential pieces that were popular at the time.

But style is fickle. When changes in consumer demand compounded with the Great Depression, Grand Rapids’ furniture industry was decimated.

The shock was an economic crisis for the region, but it was also an opportunity for reinvention, in terms of aesthetic, materials, and manufacturing processes. By the 1950s, Grand Rapids had once again earned its title as “Furniture City” with a sharp pivot to office furniture.

The seeds of success had been planted decades earlier. Herman Miller was founded in 1905 just outside Grand Rapids, and Steelcase (then called the Metal Office Furniture Company) followed soon after. Both rose to prominence in the years following World War II. In 1954, the Metal Office Furniture Company renamed itself Steelcase. Herman Miller hired George Nelson, who invented the modern workstation, a systematic approach to office furniture that treated the workspace as a design problem. Charles and Ray Eames followed, creating the lounge chair that entered MoMA’s permanent collection in 1956. These were not woodworkers continuing a craft tradition; they were industrial designers inventing a new category of product.

By the early 1960s, Grand Rapids once again produced a significant share of American office furniture. The current industry is focused on ergonomics, high-profile designers, and office systems, all manufactured from steel and aluminum, engineered polymers and textiles, wood and leather. Today, the region is home to over 225 furniture-related companies, including suppliers of wood, specialty hardware, textiles, specialized machinery, and furniture-dedicated transportation companies. The cluster effect is self-reinforcing, and most of its members proudly advertise their Michigan roots.

Industrial identity can survive complete material transformation. The name Furniture City became a self-fulfilling prophecy—a regional brand that outlasted the products that earned it.

In Providence, flexible architecture and a dense district allowed for continuous adaptation. The historic product—in this case, jewelry—matters less than the urban form it created.

Grand Rapids showed quite the opposite. The myth of the product has persisted for almost two centuries, from Scandinavian woodworkers to workshops producing Victorian replicas to post-war office furniture giants.

There is no single recipe for a resilient urban economy, but the cities we have examined in these Production Potential posts offer good ingredients: a culture of ingenuity, an ecosystem of vendors and suppliers, flexible architecture in dense districts, and the persistence of a shared narrative. These are still the right ingredients, even today.

In recent years, cities have been buffeted by an economy that changes rapidly and unpredictably. Demographic changes, the rise of hybrid work, the whiplash of tariffs and isolationist trade policies, and the ever-accelerating changes in consumer demand patterns—these destabilize what was, for a long time, a stable environment. In the next posts we will look at global economic patterns in greater detail, and uncover what they mean for cities.

Michael Wilkerson works at the intersection of public policy and private real estate economics, and he is arguably the most visible economist working on Portland’s urban condition. He has authored the analyses that validate and in some cases, challenge the Adaptive Urbanism thesis.

Our conversation starts with the diagnosis. We get into the question of whether Portland’s situation is cyclical or structural, and talk about the transition that the built environment can undergo in the years to come. And finally, we talked about the response. What might we do as we step forward in the years to come?

MATTHEW CLAUDEL: You’ve presented the State of the Economy report at the Portland Metro Chamber for four years now. When you delivered this year’s presentation, and as you walked off the stage, how did it feel different compared to previous years?

MICHAEL WILKERSON: I think it felt different in the sense that we’re now building off of previous presentations with an expectation that the public is now absorbing and digesting what we have been saying.

That allows us to expand the conversation beyond just looking at leading and lagging indicators. In previous years if I said we’re in a recession, it was so novel that people needed to grapple with it. Now this is intuitively obvious. It doesn’t need 30 minutes of content around it because the data are so consistent and overwhelming. Instead of spending the bulk of the conversation there, now I can just say recession and move on without having to justify why.

CLAUDEL: I also noticed that in years past, you’ve used recovery as a benchmark, and this year you didn’t. Am I reading too much into that? Are you indicating a structural shift?

WILKERSON: That’s the point. Thinking about recovery is flawed in the sense that it implies we can go back to where we were. Even if you believed that you could do that, the time to recovery is unfathomably long. You can’t actually work towards that goal because it’s a decade or more away.

Rather than trying to say, can we get back to where we were? I’d like to see theories emerging around what the future might be, assuming that it is going to be different than the past. We don’t know exactly what it’s going to be, but we cannot assume that we’re going back to where we were and that the playbook is known and that we can redeploy known policies that have worked in a previous structure. In the last year, that clarity emerged. Now it’s near irrefutable.

CLAUDEL: What gave you that sense of clarity?

WILKERSON: We saw certain indicators showing what seems to be a recovery, but those reversed in a way that suggests it wasn’t actually happening. There hasn’t been a reversion to the mean—it was noise. We always get the question are we at the bottom? Are things about to get better? But there is no bottom. The idea of a bottom indicates that we are going back to where we were.

CLAUDEL: If I understand what you’re saying correctly, it’s that we need to actually have a structural shift in how we think about the supply that meets a new form of demand.

WILKERSON: That’s right. We need to be talking about demand because the nature of demand has changed. Instead of prioritizing how do we align the supply with the demand, we need to focus on what is the demand now, and what is it going to be? It’s a demand issue primarily, or certainly a demand issue that needs to be looked at in combination with supply. It’s not a market failure solving for some cyclical nature, that just needs a temporary fix because the structure is still aligned in the future.

CLAUDEL: I want to come back to this and the difference between a cycle and a cliff and how we might actually put numbers behind that difference. But first I want to call out one of the most dramatic numbers in the report. One that you and I have both used in presentations. It’s very effective.

Portland is 80 out of 81 in ULI’s development attractiveness ranking. Not seven years ago it was near the top. We’ve had a precipitous fall in development attractiveness. But in a sense it’s an indicator of the creativity of the development market because development attractiveness is related to factors like rent growth expectation and cap rate compression potential, population growth, job growth – the traditional indicators.

In a world where the previous market fundamentals held, Portland is an abject failure, but if those aren’t the relevant fundamentals going forward, maybe there is a disconnect in a market opportunity. Of course Portland is not going to compete in the old game. But can we compete in a new one?

WILKERSON: That’s the realization that we’re having now. If we thought we were going to be one of the top 10 fastest growing metros from a population and jobs standpoint… no, we are not. Certainly not in the near term. From that perspective, yes, the ULI ranking is actually valid. But is it measuring the wrong things?

CLAUDEL: Exactly. That’s the question.

WILKERSON: Historically, it has been a leading indicator. When the structure is not changing, the ULI ranking is actually a very good leading indicator because it measures sentiment and sentiment reflects future development decisions.

Developers in the aggregate are not risk takers. On an individualized basis they very much are, but they’re not going to move together into a place of risk. There’s going to be a person who moves into a market individually and that person typically gets outsized returns and eventually that becomes the market norm. That’s where we are in Portland. We need to have one or two catalytic projects, and then the process starts. And at that point, we will jump massively in the rankings.

There is no silver bullet. There isn’t even a list of 10 things that you could do today in Portland to demonstrate that this is a new norm and that Portland is adapting. I used the analogy of Groundhog Day. Until the catalytic project, there is nothing that is going to change that perception of Portland through the lens of the old structure. We deserve our ranking.

CLAUDEL: Right, because the ranking measures essentially the bulk of the bandwagon. Portland does have a massive inventory of distressed assets. There is a huge untapped potential for upside from repositioning buildings. So it could be that we’re scoring poorly on conventional metrics, but we have opportunity in terms of alternative investment strategies.

The question is: how much contrarian capital is out there? What you’re saying is that we need a couple lighthouse projects. Those will only happen through contrarian capital; an alternative investment thesis. Those lighthouse projects will cause the bandwagon to change its direction, and then we will see a jump in the rankings. We need to re-rack the market.

WILKERSON: Yes, and I think one of the ways it gets re-racked is vacancy rates and capitalization rates for any given building.

CLAUDEL: Broadly this report projects that vacancy is going to peak and then gradually decline. But we’ve also argued that there is a structural threshold, at 30% or 40% vacancy, we will see an accelerating decline. You and I have also talked about the reality that leases written before the pandemic are going to be expiring in the next year or the one after. The reason that’s important, beyond the vacancy rate itself, is that a lot of debt that was structured pre pandemic, in a very different interest rate environment, is now going to be maturing into a very high interest rate environment, and that’s happening right as the vacancy cascade hits buildings. It’s a recipe for foreclosure.

Help me understand where you see this going. Are we going to hit a bottom on vacancy then stabilize, or do you think this is a cascade has yet to emerge in its fullest expression?

WILKERSON: I think what you’re suggesting is very plausible. A year ago I would have said it was plausible that we were going to hit a peak and have a stagnation to very slow recovery.

Now, if you look at the leasing volume in this last year in the central city, it’s around half of what it was one year ago, and it’s lower than it was in 2020. The data suggests a real possibility that we are not seeing a peak vacancy rate, and that we may not hit an actual peak ever. Or if we do hit a peak, that it won’t inevitably be followed by a recovery.

There was a period of time when Portland was a tech satellite. That recruitment strategy, paired with domestic natural growth, got us to that point where we were the darling of the ULI rankings.

Why would a company choose to locate here today? The answer is they aren’t. We are no longer a tech satellite. The data could not be more clear: it’s one way traffic. There are no large leases of any kind coming into the city. More problematically, the companies who are here aren’t growing organically or investing to grow.

Understanding that we cannot rely on that specific kind of recruitment is going to be very challenging, and we just don’t see examples of organic growth happening in this market because the comparative nature of operating a business here is such that you can’t extract a competitive advantage.

CLAUDEL: What you’re implying is that in the past we also had an X factor, which was the quality of life and the agglomeration effects of the streetscape. We had a comparative advantage. That is why tech companies put satellite hubs here, because people wanted to be here.

WILKERSON: Companies could operate more profitably, having remote workers at a time when remote wasn’t a thing. Now that remote is a thing everywhere, you would think we should be absolutely dominating, but it’s the opposite because we have eroded our comparative advantage.

CLAUDEL: Competitive and comparative advantage points back to something interesting I noticed in the 2026 State of the Economy report. There is mathematic barrier to trading on our X factor. You found that even a full return to office would only get us about halfway to pre-pandemic downtown foot traffic. This was shocking to me.

You’re essentially making the case that the old single use downtown model can’t work.

WILKERSON: Correct. Even if we get back to full occupancy, the old model of downtown cannot work. This is what we have been saying. This is not where we should be spending our energy. It’s not that full occupancy wouldn’t help, and some businesses might even opt into it as an individual business decision.

If the hope is that full occupancy is the strategy that’s going to “get us to recovery,” it’s a mathematical impossibility. We should understand that reality for what it is, and then talk about alternative productive strategies.

CLAUDEL: So what is downtown? What do we want it to be?

WILKERSON: Yes, we have to think about what downtown’s value proposition is in the context of our broader central city. The problem for downtown is that we have what I would describe as a new mixed use character in many of our other central city neighborhoods.

When you look at the streetscape, at the block level, our concentration of problem areas is immense, while there are pockets of strength around them.

What are the big targeted interventions that can provide enough of an anchor that you could transform downtown in a strategic way? A lot of people are saying it just needs to be more housing. But is there a plausible value proposition for a resident to move in as the first tenant of a transforming downtown?

Maybe, but if it has to be massively subsidized, we need to be clear about the implications. And maybe the answer is that there is no value proposition. If that’s the case, then Portland’s downtown needs to transform itself into a different type of thing, contextualized by the rest of the central city.

CLAUDEL: So if I could simplify what you’re saying into a dichotomy, the neighborhoods around the central city are, in a sense, becoming more like historic downtowns, because they are now places where people work. What if downtown becomes more like a historic neighborhood? That hypothesis is why we’re now having a conversation about office to housing conversion.

Now, I think it’s worth looking back at the housing conversion study that you worked on. You found that, and correct me if I’m wrong, only about 5% of the building stock is amenable to housing conversion. There are a lot of headwinds, right? We have architectural headwinds. We have policy and land use and building code headwinds. Cost is obviously a headwind.

Conversion is just very challenging. You just brought up an interesting one, which is the cultural headwind. Here we are, sitting next to Nordstrom, right off of Pioneer Square. This works as a business district, but it doesn’t really feel like the kind of place where I want to have my kids running around.

WILKERSON: It’s not an after five o’clock neighborhood.

CLAUDEL: It is not an after five o’clock neighborhood. I’d call this a cultural headwind. So I’m curious, how should we think about office to residential conversion? Are we seeing that the decline in the actual acquisition cost of any given piece of real estate now is making the financial piece maybe more viable? Does that change the balance of headwinds and tail winds?

WILKERSON: No, unfortunately. In that report we were projecting into the future, to a time when acquisitions could happen at $50 a square foot. They were not happening at that point then, but they are now. That is to say, we were already looking ahead to a future state of distressed acquisition. So no, there is no possibility that the value of the buildings could drop so much so that it would alleviate the need for other interventions.

For office to housing conversion to be viable, you still need other interventions. That was the “aha” moment in the conversion report. There is a need for structural changes to fire and life safety codes. Until you do that, nothing else matters. On top of code amendments, there are a handful of necessary but insufficient conditions, like SDC waivers, tax exemptions, and per-unit subsidies.

The question is, structurally, can that even happen today, or is it just a market impossibility? I’m not saying definitively yes or no. I’m saying we need to explore that “if” statement. If this is an impossibility, then trying to build units of housing will be a failed policy.

I’ll use SDC waivers as an example. Three years ago, SDC waivers would have been immensely valuable. We weren’t ranked 80th out of 81. At that time, the market was close enough to delivering housing, and SDC waivers would have been a signal.

The thought was that SDC waivers would generate thousands of units. Now we can see that it has done effectively zero. So there are still tons of headwinds. You’re not going to see the market respond to the opportunity for conversion without just moving mountains.

CLAUDEL: We are pushing an immense boulder up a hill.

WILKERSON: Exactly. The intention is not wrong, and 2,500 units of housing might be the ultimate outcome.

The bigger question is really around sequence and timing. If this isn’t a neighborhood, a place for families, what do we need to do to make it so? Do we need to build parks? Do we need to build a children’s museum? Do we need to build more arts and entertainment and cultural things?

Maybe the long-term strategy is changing the value proposition of downtown versus trying to subsidize office to residential conversion in an unscalable way. It might not be an optimal strategy now, at scale. But if the intention is to make downtown a neighborhood, there need to be many targeted interventions between now and then, and maybe it can happen. I would rather not wait for five years to then realize, oh, we weren’t ready to do that yet.

CLAUDEL: What if we take a step sideways? Residential is not the only target for conversion. Suddenly you’ve eliminated a lot of those headwinds, specifically the structural and fire and life safety issues, which bring huge costs, and obviously the associated zoning questions. I wonder if there might be a near term strategy for putting some of this square footage to work before we see the fundamentals of residential start looking different.

WILKERSON: That’s the hypothesis that makes more sense in this moment. And you’ve done a good job of starting to outline why that is the case.

CLAUDEL: What sorts of targets could you imagine? What do you think would be the most interesting?

WILKERSON: I think this is part of the challenge for Portland. Today we’re hearing we need an economic development strategy. We’ve evolved the conversation, but when seven different economic development strategies are happening at the same time, they’re very unlikely to be aligned.

We’re trying to eek out some competitive advantage – what will allow us to do that? The answer, to me, is that we probably need to do the thing that is politically challenging, which is to say we don’t know, and we’re going to take a varied approach, some of which will work and some of which will fail. And here is our plan to mitigate the impact of failures.

Trying a one-off only approach and thinking that you are going to succeed – that is unlikely to happen.

CLAUDEL: One of the most important, if not the most important role of an economic development strategy is to say what we are not going to do.

WILKERSON: Correct. I would argue it’s the only thing you’re really trying to do with an economic development strategy. It’s hard to know what you are going to do. But if we can be clear about what we are not doing, that is way more helpful. People can react to execute around that.

CLAUDEL: Exactly. You need to say, these are the outcomes that we’re working toward, and then we’re going to allow for the creativity and the exuberance of the market, with specific regulations and incentives built in.

WILKERSON: Build the sandbox for us and then we’ll figure it out. Don’t build the sandbox so small and prescriptively that it won’t work well.

CLAUDEL: And also make sure that there aren’t seven different sandboxes. There is a huge risk right now in Portland and in Oregon more broadly that everyone comes up with their own economic development strategy.

WILKERSON: Finally we all agree that we need something like this to progress, but maybe we need to pause a little bit.

CLAUDEL: You’ve said before that these reports are data for the conversation, not a list of solutions. But if we could take a moment, Mike, I’d like to hear your own speculations about what we might need to move forward today, especially as you’ve presented this report now four or five years running. How has your sense of what we need to be doing sharpened and what are the near term and longer term levers that we should be paying attention to?

WILKERSON: I get asked this question all the time. What would you do? And my answer is we have to understand what we’re optimizing around first, before we can talk about what we would do.

What is clear is that we have a demand problem. That is foundational to everything. And our historic policy interventions are ill-equipped or not equipped at all to deal with that. That’s a challenge.

Our tax structure is so flawed and limited that it severely impacts demand. If our tax structure in Oregon is taken as a given and a status quo, we can have a conversation about what actions we should take.

Alternatively, rather than trying to come up with an ill-equipped solution within the status quo, to me, a logical first step is structural reform. At that point, we could do very different things that would be much more likely to succeed. The challenge is that structural reform is, at minimum, a three year process.

Your question is what would I do? My answer is we should work towards that three year process. That’s a tough political decision, to say there’s a pathway here, and we are willing to take the time and energy to do the hard thing. In the meantime, are we going to implement a solution that we know is entirely insufficient because it’s better than doing nothing?

CLAUDEL: You’re very genuinely asking what the parameters of our action are. You also implied that there’s an issue of knowing, based on data, when the right action needs to be taken. You’re a data guy…

WILKERSON: I look at data occasionally. It’s true. Or you can say I live in it, a little bit.

CLAUDEL: We’ve been advocating for the idea that cities need to shift from a prediction paradigm to a learning paradigm, because no matter how much data we have, we can’t predict what the next five years, let alone ten are going to look like. I’m thinking this also relates to a comment you made a minute ago about creating the sandbox and then allowing ourselves to play toward the optimal outcome.

WILKERSON: Optionality is critical in this moment we are in. You and I both agree with that. But that is very different from the traditional paradigm in commercial real estate, and it is not what people traditionally want to hear.

CLAUDEL: Optionality has a cost. Why would we do that? Why would we build in optionality? Well, if you consider real options modeling or expected utility or risk adjusted returns, this is very clearly the right path today.

You can price risk and you can think carefully about what sorts of actions you should take at any given threshold once you see it. But you shouldn’t try to predict exactly when you’re going to see it, or which threshold you will reach. So let’s advocate for better data, but let’s also equip ourselves to be learning as best we can, as we’re taking action. So, with what data should we be furnishing ourselves as we step into this uncertain future?

WILKERSON: That’s a great question. We have to understand the totality of whatever we’re trying to capture. So if that’s people, jobs, income, GDP, what is the entirety of that? And then we can look at what our share of that is, what our trend in that is. Maybe, for the purposes of comparison, we can look at what others are doing, so we can understand if we are actually treading water? Because we have to abstract it up to a big enough observable thing that we can actually understand it.

And then I think we’re seeing emerging indicators that are more real time that we can start to get at. Foot traffic as an example of this, and pairing consumer spending with foot traffic with other things starts to give us a more real time understanding of what’s happening. There is a whole slew of proxies that we can start to build to get at some more cultural questions.

CLAUDEL: I was here on Saturday at Pioneer Square to see the winter lights, and it was packed. People were loving the winter lights.

WILKERSON: Absolutely. That’s a huge example of a successful targeted intervention that’s clearly delivering benefit.

CLAUDEL: So having a more attuned understanding of where culture is moving as people think about the value of downtown and the reasons for coming down here. We need data to understand what overcomes the magnetic pull of Netflix and the safety of a neighborhood. We need to understand what the cultural drivers are at any given moment.

So final question. You’re an economist. You teach at PSU in the real estate program. You live in Portland. How are you going to approach the State of the Economy report differently next year? And what are you going to be keeping your eyes open for, over the next year as you work toward writing that report?

WILKERSON: It’s an interesting report because I think for some people it’s a once a year opportunity to get a checkup of what’s going on.

In my mind, we’re doing a state of the economy report on a monthly basis. It’s the sum of all the work that we’re doing through a variety of clients. It always evolves, and as we get into the Fall we already generally have a sense of what’s going on.

In this particular instance, we have so many ongoing or soon to commence economic development efforts, strategies, et cetera, that are gonna play out largely through the Fall. I think that will be very indicative of what the need is at that moment in time. Are we going to say we were unexpectedly very coordinated here, and this actually is now something that we can begin to prioritize and execute on? Or will we have the seven different sandboxes problem? And is seven sandboxes better than zero? Probably. But we need to get more coordinated around it.

With near certainty I can say that the lagging indicators are going to be exactly the same. Nothing we’re doing now (and this is the hard conversation) is about changing the next 12 months, in a meaningful way. Hopefully we can sit down and have this conversation again in a year and we’ll see how prescient we were with our thoughts.

Does a city’s resilience have anything to do with its narrative? Last week we made the case that it does. Cities with a stronger sense of self—a clear story about who they are, grounded in what they make—have better tools for reinvention when they are hit by economic and social shocks. Those cities know what they’re building on and they have inherited resources. Rochester was a strong example, and this week we go deeper, bringing Detroit into the mix.

Cars and cameras don’t have much in common. But the cities that built them do. Both Detroit and Rochester were shaped by dominant, campus-bound companies with charismatic products that the whole world knew. Both cities concentrated their industrial identity on strong brands—Ford, GM, and Chrysler on one side; Kodak on the other. Both built sprawling, self-sufficient campuses that operated like cities within cities, with their own power plants, roads, and workforces in the tens of thousands. And both cities watched as their business titans contracted, leaving behind vast footprints and uncertain futures for the urban economies they once bolstered.

What follows is an overview of two centralized production models—and what happened when each confronted a set of disruptions: digital technology, offshoring, and the slow unwinding of twentieth-century manufacturing. Neither city escaped unscathed. But both found, in different ways, something worth building on. The outcomes, as we’ll see, were not the same. And the reasons expose something structural about how industrial identity is built, lost, and, in some cases, recovered. Identity is not the only determinant of resilience—the structure of the regional production economy is also a defining factor.

Detroit

Henry Ford is known for the automobile, but his true innovation was the assembly line—an efficient method of making a complex product quickly and consistently. Ford’s River Rouge Complex, ninety-three buildings spread across 2,000 acres, was itself an assembly line. Iron ore and raw materials arrived at one end; finished automobiles departed from the other days later. Between those two points blast furnaces, coke ovens, a steel mill, a glass plant, a tire factory, a power station and assembly facilities worked in tandem. By 1940, over 100,000 workers clocked in each day. By 1950, Detroit was the fifth-largest city in the United States, and River Rouge set new benchmarks for the size and efficiency of an integrated industrial complex.

Diego Rivera spent three months sketching the interior in 1932, captivated by the molten metal, the human-machine choreography of the assembly line. Charlie Chaplin visited too, studying the Rouge for Modern Times. Both artists were enchanted by industry as a total system, self-sufficient and self-contained.

Chrysler and General Motors followed Ford’s lead, creating vertically integrated facilities of their own. Across each of the “Big Three” the goal was to vertically integrate, eliminating the need for local suppliers and creating advantages in terms of efficiency and control. Vertical integration created disadvantages, however, for the Detroit region.

A combination of pressures—foreign competition, rising labor costs, and a failure to anticipate fuel-efficiency demands—took their toll on The Big Three. When Ford, GM, and Chrysler contracted, there was no ecosystem of smaller firms to catch displaced workers and no network of specialized manufacturers that might pivot to new products. Detroit had large automobile companies, it did not have a regional industrial ecosystem for creating automobiles—a web of interdependent firms sharing knowledge, competing for talent, and spinning out new ventures.

The Detroit Three automakers went from controlling over 90% of the U.S. market share in the 1940’s, to only 17% of the U.S. total vehicle volume today, and in 2013 Detroit became famous for largest municipal bankruptcy filing in U.S. history.

But the story doesn’t end in bankruptcy. The River Rouge facility still operates. Ford builds F-150 Lightnings there now, on 600 acres—less than a third of the original footprint. The blast furnaces, glass plant, and tire factory are gone. What remains is an assembly operation, its components arriving from suppliers scattered across the US and far-flung continents.

The sense of opportunity that followed Detroit’s bankruptcy drew a new kind of energy to the city. Land was cheap, and a generation of inventive projects emerged. Little Village is an emerging cultural corridor that encompasses a skate park, a sculpture garden dedicated to the late artist Charles McGee, a boutique bed-and breakfast, a bar and two restaurants, as well as office space for nonprofits. West riverfront park would almost certainly have become a private mixed use development if the City had not experienced a decline and re-emergence. Shinola brought watchmaking back to the United States with its Detroit workshop—and now includes an expansive manufacturing operation and a boutique hotel. After decades in the shadow of the Big Three, a regional ecosystem is flourishing.

Michigan Central is a punctuation mark to this story. The 30-acre project is focused on mobility innovation, societal impact, and culture, and it is centered on the long-derelict Michigan Central Station, an icon of pre-automobile transportation. This is the most visible signal of Detroit’s attempt to rewrite its industrial identity—and Ford was heavily involved. Today, the campus hosts startups and research groups working on autonomous vehicles, micro mobility, and electric propulsion, through a partnership with NewLab. Field States contributed analysis and strategic design work for Michigan Central, envisioning a new workplace fit for a today’s innovation processes. Michigan Central is a billion dollar bet that Detroit’s manufacturing depth—its engineers, its supply chain knowledge, its production culture, and its self-narrative—is the right substrate for the next generation of mobility technology.

Rochester

Throughout the four decades from 1950 to 1990, a full wall of New York’s Grand Central Station displayed an advertisement for the American experience. Well, an ad for Kodak, but the camera company was an American icon. This was true in more ways than one—as much as Kodak represented travel, leisure, and family values, the company was a symbol of American manufacturing.

A short train ride away from Grand Central Station, Rochester was the global center of camera invention and production. At its peak, Kodak employed 62,000 people in Rochester, roughly half of the city’s working population. Kodak’s campus, the Eastman Business Park, had 154 buildings, including laboratory, manufacturing, and office space, as well as its own power plant, railroad, fire department, and thirty miles of internal roads.

Unlike Ford, Kodak did not vertically integrate. An ecosystem of suppliers sprang up around it, including Bausch & Lomb, Wollensak Optical Company, and Graflex Inc. Together, these companies manufactured every component in a complete photographic supply chain, including film, optical lenses, camera shutters, photo paper and chemicals.

When digital photography emerged, Kodak moved slowly and the business model collapsed. A long decline ended in Kodak’s bankruptcy filing in 2012. Fifty buildings have been demolished since; controlled explosions drew crowds of former workers to watch an icon reduced to rubble.

But the campus is alive with something Kodak never anticipated. A cross-sectoral ecosystem for optics innovation—including the American Center for Optics Manufacturing, Optimax Systems Inc, and AIM Photonics—has grown from the knowledge Kodak spent decades accumulating. Rochester has contributed to the Hubble Space Telescope, scientific advances in the coherence theory, high-precision military applications, advanced lithography for semiconductor manufacturing, and now quantum computing. The campus has also become a home for entirely new industries—companies as diverse as cannabis and advanced textile production have adapted vacant industrial space.

Six thousand people work at the Eastman Business Park today, a fraction of Kodak’s peak workforce, but a telling representation of contemporary business models and production methods; it is currently home to 114 companies and is leasing over 200 acres.

Before George Eastman died in 1932, he asked that his ashes be interred at Kodak Park. He loved this place, and his work had an extraordinary influence on what Rochester became. Eastman’s legacy is alive in the campus that is thriving once again; a multitude of innovative companies in place of the American monolith that Eastman built. Even outside of the campus, Rochester’s economy is laced with the history of Kodak.

Resilience depends on narrative and manufacturing networks

Rivera’s murals captured a moment when industrial control was strength. The decades since have demonstrated otherwise, and as behemoth firms stumbled, the cities they bolstered rapidly declined. When production concentrates in a single organization—vertically integrated, campus-bound, self-sufficient—the organization’s decline leaves little behind. Those that weathered deindustrialization best had diversified manufacturing bases—dense networks of small and medium firms that could adapt when any single company failed. When a city’s production capacity is distributed across an ecosystem of firms sharing knowledge, competing for talent, and spinning off new ventures, disruption has a softer impact on the system. Rochester proved this too, imperfectly and at real cost, but durably.

Detroit and Rochester offer the same structural lesson from different angles. Where production concentrates without an ecosystem, narrative can sustain potential but cannot substitute for network. Detroit is rebuilding, thanks to the ecosystem that emerged post-bankruptcy. Where narrative and network reinforce each other, the city is more resilient. That distinction—between a story a city tells about itself and the actual web of firms and people that carry industrial knowledge and narrative forward—is what determines the rate and character of revival.

The next Production Potential post turns to Grand Rapids, Providence, and Portland—cities that didn’t concentrate production in a single campus or brand.

Subscribe now

You know a district has crossed a threshold before the data confirms it.

The coffee shop that opened with optimism eighteen months ago has paper over its windows. The ground-floor retail unit that was “available” is now just dark. Someone is sleeping in the doorway of what used to be a sandwich place. The flowers in the planter boxes have died and no one has replaced them.

The office workers who kept these businesses alive aren’t coming back five days a week. Some aren’t coming back at all. The lunch crowd that kept a small business humming (and paying rent) is now a Tuesday-Thursday trickle. The café owner did the math, and the math doesn’t work.

By the time you notice, the damage is done. The storefronts are a symptom. The underlying problem is in the floors above.

Last week, we looked at two Portland towers—Big Pink and Block 216—and asked how sophisticated actors making rational decisions ended up with stranded assets. We asked if the current trends are part of a cycle, or a larger restructuring of urban demand patterns. If it’s a cycle, the right move is patience; if it’s a transition the right move is swift, decisive action.

This week we explore how cities have passed a cliff that prevents a cyclical upswing, and why waiting for recovery is the wrong bet. It is only by understanding these dynamics that we can design the response our cities need—a topic we will focus on in weeks to come.

The mechanics are familiar by now. Vacancy rises. The landlord offers ballooning concessions to attract tenants—eighteen months free, generous TI allowances, break clauses at year three. Each sweetener is a bet that the cycle will turn before the giveaways consume the economics. But the headline rents are still high; a signal of the building’s underlying financial distress. Tenants stay away. When income falls far enough, the building is in breach of its debt covenants, and it enters special servicing. There is a forced sale at a low price, which is reflected in the assessed value the following year. Tax revenue contracts precisely when the city needs resources to respond. Services thin. The public realm degrades. Tenants who remain find it harder to justify staying.

This cascade has been well documented. There are local variations across cities like Portland, San Francisco, Seattle, and Denver but essentially the pattern is repeating. Big Pink lost $98 million in assessed value in a single year, a 68% reduction in property tax revenue from one building.

But the doom loop itself is not the story. There is a far more important question that no one is asking: are we in a cycle or a transition?

If real estate and regional economies are in a cyclical pattern, we’re waiting for the bottom before an upswing kicks in, and patience is the right strategy. If it is not a cyclical pattern but a structural transition, there is no bottom—just a new equilibrium at a permanently lower level of demand for specific kinds of uses. That question changes everything. And answering it requires looking at specific indicators.

Markets soften gradually, at first, then they cross specific thresholds and the rules change. We can identify those thresholds with enough precision to know which side of the threshold any given city is on.

Threshold 1: Building Breakeven

A standard leveraged office asset needs 70-80% occupancy to cover operating costs and debt service. Above that line, owners have options and time. Below it, they’re injecting capital, renegotiating with lenders, or heading toward default. The concession packages that preserved headline rents have already consumed the cushion.

This threshold is building-specific, but the pattern is systemic. When a critical mass of buildings in a district falls below breakeven simultaneously, the workout process itself becomes a drag on recovery—distressed sales resetting comps, lender caution freezing refinancing, capital allocating elsewhere.

Threshold 2: Retail Clustering

Retail depends on critical mass. Customers come because multiple options exist; they expect a sense of discovery and surprise. The audience pulled by one shop augments the revenue of the shop next door. Research on the viability of shopping streets suggests the tipping point falls between 15-20% vacancy. Below that, a street is functional. Above it, the clustering effect inverts: foot traffic thins, remaining tenants lose the spillover that justified their rent, visible distress repels the foot traffic needed for recovery.

This threshold is observable: walk down the street and count the dark storefronts. If one in six units is empty, the street is likely to recover. If one in five units is empty, the street is in danger. If it’s one in four, the street may already be past the point where organic recovery is possible.

Threshold 3: District Signal

When office vacancy crosses 30-40%, the market receives a very specific signal. High vacancy stops reading as “Opportunity! Space available, terms negotiable” and starts reading as “Risk! Something is going wrong here.” Capital doesn’t flow toward that signal. It flees to districts that still appear stable, concentrating investment in fewer locations and accelerating divergence.

San Francisco’s SOMA submarket hit 47.6% vacancy in late 2024. That is not a soft market. That is a market where nearly half of all space sits empty, where the critical mass required for street life has been lost, and where the reasonable expectation for a prospective tenant is continued decline.

Economists call this property hysteresis: systems that retain the effects of shocks even after the initial cause is removed.1 The path down is not a path back up. Recovery, if it comes, requires overcoming accumulated disadvantages—not simply waiting for conditions to normalize.

In real estate, the practitioner’s instinct is to wait. This instinct has served developers and investors well through previous downturns, when markets have self-corrected.

In past cycles, three mechanisms restored equilibrium. Rents fell until demand returned. Construction stopped until vacancy absorbed. Policy reflated values through rate cuts and stimulus. Today, none of these mechanisms is operating.

Construction has collapsed 88% from 2019 levels—and vacancy still climbs. Demand is not absorbing excess supply. Portfolio real estate holders are shedding capacity that is no longer in demand. Meanwhile, policy is adding pressure, not relieving it: elevated interest rates, tightening insurance markets, energy performance requirements that will render a significant share of existing stock non-viable by the decade’s end.

Are we seeing a pause in demand? If tenants are waiting for conditions to improve before returning to the same spatial patterns, it is a cycle. At this point, more than half a decade since lockdowns began, it is safe to say that normal work patterns are not just paused; a new normal has emerged. Companies have rewritten their HR policies, fully remote teams have developed collaboration patterns, and firms compete for the best talent in a global applicant pool by offering hybrid and remote jobs. Mandated ‘return-to-office’ may work when the labour market is weak, but when it becomes a seller’s market, employees will vote with their feet. Hybrid and remote work is a stable configuration with physical spaces—home offices, third places, on-demand co-working—that didn’t exist at scale five years ago. Companies just don’t need as much office space as they did before. The evidence shows that activities that generated demand are reorganizing into different forms entirely—lower rents cannot summon demand.

Historical analogy offers little comfort. The 1990s crisis was supply-side: too much construction, eventually absorbed as demand grew into the excess. The 2008 crisis was financial: over-leveraged assets written down, recapitalized, repriced. Both resolved because underlying demand remained intact. Today’s crisis is demand-side. The activities that filled these buildings have permanently reorganized. There is no historical playbook for waiting that out.

The buildings falling vacant aren’t waiting for tenants to return. They’re watching tenants build lives that no longer need office buildings.

So the change we are seeing is structural, not cyclical. We believe that the response must also be structural. Not better marketing for buildings, or deeper concessions, and certainly not patience.

Buildings designed for one purpose need to become something else. Districts optimized for a single use need to diversify. Institutions built for predictable growth may need to learn how to adapt—not a single period of adaptation followed by stability, but a deep capacity for continuous adaptation as conditions keep shifting.

The pacing problem we identified in our first essay—the mismatch between how fast the world changes and how fast cities can respond—doesn’t have a pacing solution. Speeding up approvals won’t help if you’re approving the wrong things faster. The problem requires an adaptive solution: building capacity to adjust as conditions reveal themselves, rather than betting the house on forecasts that no longer hold.

Diagnosing the underlying problem is not about despair—it is a necessary step to design effective action. Once you stop waiting for the cycle to turn, you can start asking better questions. What do these buildings need to become? What does this district need to thrive? What would it take to build a city that learns?

Our next posts will begin answering those questions.

Subscribe to Field Discovery to follow the series. New essays every Tuesday, with supplementary posts on Thursdays moving from theory toward action.

About Field Discovery

Field Discovery is the research arm of Field States, surfacing the structural shifts affecting cities and building the evidentiary foundation for Adaptive Urbanism. We work with cities, developers, investors, and institutions to create places that thrive in conditions of change.

Matthew Claudel & Simon Giles

Subscribe now

For decades, Rochester was America’s undisputed capital of cameras and film. Why is the Kodak sign important in a city that no longer makes photography equipment? Because identity matters. Cities with clear self-narratives consistently recover from adversity—they draw on their past to navigate their future. Now, with office vacancy crossing 20% nationally and climbing to 40% in some downtowns, cities face a choice of chasing the next mobile industry or building on who they are. This essay explores why the knowledge economy proved brittle and why identity matters for urban resilience in key American cities.

He had driven past the Eastman Business Park hundreds of times before, on his way to and from downtown Rochester. On an otherwise unremarkable late summer day, he pulled over sharply. Workers were dismantling an iconic red sign, and he needed answers. “Why are you messing with the big sign?!”

The construction team reassured him that they weren’t removing anything—just rewiring the sign so the new tenants didn’t have to foot the electric bill. “My heart skipped a beat,” he later wrote. “I know it’s just a sign. But it’s also part of our collective story.” Almost every American of a certain age, and certainly every Rochester resident would recognize it: Kodak.

The company made extraordinary breakthroughs in photography, and became almost synonymous with the American dream. As it grew, it brought burgeoning energy to Rochester. Almost as quickly, Kodak declined, as digital photography eclipsed film and foreign companies dominated the new market.

Since Kodak’s bankruptcy, Eastman Business Park has transformed from a company campus into a cross-sector innovation hub housing 114 companies—photonics firms, biopharma manufacturers, and even cannabis cultivation. The Kodak sign stays, proof that Rochester has earned its place in history. The memory of Kodak matters to the workers who made film in these buildings, for their children who work here now, and for the Rochester resident walking by.

What allowed Rochester to reinvent itself? This essay explores how a city’s legacy can animate its future.

Why Identity Matters

Research shows that individuals with clear self-narratives recover faster when adversity hits. They draw on past experiences, find purpose, pivot with confidence. A strong identity adapts without breaking.

The same is true for a city. Historically, a city’s physical form, economy, and identity develop hand in hand over time. And often, one industry becomes so foundational that the city is the industry.

As companies succeed, buildings are constructed to serve their distinct needs, with features like tall ceilings, large windows, loading bays, or basement storage. Distinctive building types emerge, forming whole neighborhoods that are laid out for production. Urban form calcifies in the image of the industry and its products. Jewelry studios line Providence’s downtown manufacturing district. Detroit’s “big three” automakers each developed a campus.

Industries shape the social fabric as well, as innovators and employees build networks. Knowledge is created and stolen and adapted through a competitive process. The city acquires a self-narrative that is intimately connected to the industry, and there is a pride in place.

Industries bust almost as much as they boom. When international trade prospects change or raw material supplies run out or technological advances make an industry irrelevant, urban decline follows. When whole regions are struck, names like the rust belt stick. The population diminishes, wealth shifts out of the city, and buildings are vacated. In addition to economic decline, the result is often a collective existential malaise, because a city is its industry.

Resilience Depends on Narrative

Kodak emerged into its heyday in the late 1970s, spinning out scientific advances and animating a cluster of optics innovation and production in Rochester, New York. Then Kodak tumbled, falling quickly at first, then slowly, excruciatingly, over decades. By the 1940s, The Detroit Three automakers controlled over 90% of the U.S. market share. Today, the Detroit Region produces only around 17% of the U.S. total vehicle volume.

But some cities prove to be resilient. The ones that reinvent themselves share something: a strong narrative.

Grand Rapids emerged as a furniture capital in the late 1800s and the city’s name became a mark of quality by 1902. Decades later, the city’s furniture show was eclipsed by a larger fair in Chicago, as its artisans were caught unaware of—or perhaps were unwilling to accept—the American public’s changing tastes. When changes in consumer demand compounded with the Great Depression, the industry was decimated.

After a period of decline, brands like Steelcase and Herman Miller rose to prominence in the years following World War II. They became internationally successful for a new kind of design: modern metal office furniture. By the early 1960s, Grand Rapids once again produced a significant share of American furniture. What allowed the pivot? It was not craft knowledge (prior expertise was in wood carving) or the existing buildings stock (artisan woodshops replaced by industrial metal fabrication). What mattered was the identity. Grand Rapids believed it made furniture, and when Victorian reproductions died, the city didn’t ask what else it could make. It asked what furniture could become.

For a century, this pattern repeated: boom, decline, reinvention. But the 1980s broke it. Globalization and offshoring struck every manufacturing sector at once. Knowledge work took its place—first in downtown office towers, then in converted warehouses as cities competed to attract tech companies and startups.

Knowledge work was a strong economic base, but firms were highly mobile. Workers moved quickly between companies, and companies moved quickly between cities. Soon cities competed to attract tech company headquarters, rather than cultivating industrial craft or a stable production base. Post-industrial buildings became staging grounds for this competition—warehouses converted to startup offices were trendy. They traded on urban character without creating lasting ties.

The tech economy proved brittle. A software developer can work from a warehouse in Portland or a glass tower in London or a beach in Bali. The rise of remote work in 2020 only intensified this placelessness. Without roots or narrative, downtowns hollowed out.

Manufacturing is intimately connected to place, in a way that knowledge work isn’t. When Kansas City rebranded as the Silicon Prairie, did anyone feel inspired? Did the identity run deep enough to weather a downturn?

Rebuilding on a Narrative Foundation

Cities must rebuild their economies. The next chapter of American urbanism cannot be written in the language of knowledge work alone. As in any biological system, monocultures fail under stress.

Cities adapt successfully when they build on their sense of identity; pride in place becomes the foundation for reinvention. For decades industry globalized and municipal leaders wooed the technology industry, homogenizing their cities and their urban economies. Years of post-pandemic malaise have stalled development. With the shift to remote and hybrid work, companies are shedding their real estate portfolios. In 2024, the US office vacancy rate crossed the 20% threshold for the first time, and in some cities these rates are much higher: 36-40% of office buildings. Now is the time for reinvention. A place-specific narrative may be dusty (or rusty), but it is still there.

Rochester kept its sign. It survived by building on what it was—and other cities can do the same. In the coming weeks, we’ll trace these narrative through-lines in five American cities: Detroit, Rochester, Providence, Grand Rapids, and Portland. Each tells a story of identity, loss, and reinvention.

Why these cities? They share crucial traits: each has a strong industrial identity (cars, film, jewelry, furniture, shoes). Each faces high vacancy rates. None is located in a state that has prioritized and funded manufacturing in the recent push for reshoring—these cities must define their next chapter endogenously, that is, based on their own intrinsic strengths, characteristics, and narratives.

Subscribe now

At sunset, the U.S. Bancorp Tower earns its nickname: Big Pink. Pietro Belluschi designed the building to catch light, with pink Spanish granite and windows glazed with copper and silver that shift from reflective to translucent as the day turns. For forty years, Portlanders have watched Big Pink glow at the end of the working day, a landmark visible for miles along the city’s long east-west streets.

The building still glows, but only around 40% of the offices behind those windows are occupied. The tower’s eponymous anchor tenant, U.S. Bancorp, has gone. The light is fading on an era.

Big Pink had many good years. Completed in 1983 for a construction cost of $60 million, it traded steadily upward through successive ownership cycles, thanks to increasing office lease rates and decreasing capitalization rates: $165 million in 2000, $286 million in 2006, $372.5 million in 2015. Even the 2008 financial crisis barely registered. Between 2013 and 2015, the owners invested $15 million in renovations, and in 2019 it achieved LEED Platinum certification. The asset was maintained, upgraded, and well positioned for continued institutional ownership. Rumors of a transaction in late 2019 and early 2020 put the price in the high $400 millions.

Then the game changed. By July 2025, Big Pink sold for $45 million, less than its 1983 construction cost in nominal terms, a fraction of the capital invested over four decades, a sum estimated to be less than the land value.

A ten-minute walk away, Block 216 tells a different story.

In 2017, Portland ranked near the top of the Urban Land Institute’s development attractiveness index; few developers weren’t interested in Portland. A new project at Block 216 began forming, and quickly got the attention of the Ritz Carlton group.

The tower, Portland’s largest since Big Pink, was designed with office space on floors one through seven, a luxury hotel on floors eight through eighteen, condominiums on floors twenty-one through thirty-five, and a food hall on the ground floor. The plan was an exercise in prudent diversification across revenue streams, with the imprimatur of premium branding. The Ritz Building was grounded in sophisticated risk management.

Nine months after the construction groundbreaking, the world locked down for COVID-19. Work proceeded, and in 2023 the tower was delivered. The doors opened to a very different world than the one Block 216’s developers had envisioned when they started the project in 2017.

On opening day, there were only two office tenants, representing 15% of the available space. By mid-2025, all three components of the building were underperforming. The office space was 23% leased. The hotel was running 45% below the Ritz-Carlton brand average, and 11 of the 132 condominium units had sold. In July 2025, the lender took possession of the building, with a total exposure of $565 million.

On paper, the project followed standard risk logic. In practice, a bet worth hundreds of millions did not pay off. Office leasing, luxury hotels night, and premium downtown residential all depend on a well-established pattern of business travel, urban vitality, and professional norms. When those patterns shifted, diversification did not protect the project. It wasn’t too big to fail, it was too big to succeed.

These two buildings are five blocks and forty years apart, and today, both are stranded.

What makes these cases instructive is not the losses, spectacular as they are. It is the competence of the actors involved.

Big Pink was designed by Skidmore, Owings & Merrill with Pietro Belluschi as consulting architect. It was anchored by U.S. Bancorp for four decades and owned successively by JPMorgan, LaSalle, and UBS. Unico Properties managed the asset. These are not amateurs. At every stage, sophisticated professionals made rational decisions on the basis of information that had been reliable for decades.

Block 216 was developed by BPM Real Estate Group. The project went through multiple design iterations, secured major brand partnerships, achieved complex entitlements. The debt was structured by institutional lenders applying rigorous institutional standards.

Big Pink and Block 216 are not stories of incompetence; both failed because the rules changed while the game was in play.

Cities operate on two clocks. One measures how fast cities can change: development cycles, planning approvals, construction timelines, lease terms. The other measures how fast conditions actually shift—things like hybrid work adoption, retail format cycles, technology diffusion, and the churn of corporate tenants.

For most of the post-war period, these clocks ran roughly in sync. A developer who noticed an emerging trend still had time to design a response, secure approvals, obtain financing, build, and deliver into a market that wanted what was built.

Since 2020, that relationship has inverted. Clock one has slowed, as permitting, legal process, and construction each swelled and all compounded in a cascading effect. And Clock two has accelerated: hybrid work went from a quirky anomaly to mass adoption in roughly eighteen months. By the time a building like Block 216 could have adjusted—redesign, re-entitlement, refinancing, and construction—the market has already reset. Changes of a similar magnitude are coming from AI, climate change, advanced manufacturing, autonomous vehicles, and fast-shifting international trade relations.

That is to say: Block 216 and Big Pink were not simply victims of COVID. To blame their failure on the pandemic is too easy.

Five years later, hybrid work has institutionalized and we are no longer waiting for the pandemic to end. Yet development remains slow, or frozen. Over the course of 2025, the U.S. saw a 44% drop in new office construction starts, measured year over year. Why? Because developers who can see the mismatch between Clock One and Clock Two are increasingly unwilling to take expensive bets on what demand patterns will emerge five to seven years in the future. No one can anticipate the next tectonic shift, but we all know it will happen.

These buildings were both victims of the pacing problem, the growing gap between the speed of change and the capacity to respond, and it matters beyond Portland.

San Francisco’s office vacancy has climbed above 37%. Seattle is above 25%. Denver, which attracted significant pandemic-era investment, has seen vacancy climb from single digits to approximately 20%. The pattern repeats, with local variations, across virtually every major American office market. We are starting to see similar patterns emerge in international markets as well.

Across the U.S., more than $1.5 trillion in commercial real estate debt is scheduled to mature by 2027, much of it originated during years when interest rates were low and optimistic building tenancy assumptions still held. This is happening right as tenants are considering renewal of leases they signed pre-COVID, an opportunity many will use to reimagine their hybrid and remote work policies. Buildings that were well-performing assets three years ago are entering, or will soon enter, special servicing as owners confront debt they cannot repay and cannot roll over.

Real estate portfolio owners are staring down the barrel of a balance sheet crisis as leases come up for renewal and valuations are forced to adjust. But this is more important as a problem for governments that depend of property taxes to fund their operations. Cities across the U.S. earn 12-45% of their general fund dollars from property tax.

Is this a cycle that will self-correct? In the past, when vacancy rises, rents fall enough to attract tenants back, and distressed assets trade to new owners with a lower basis. Cities have, during some historical periods, been cyclical.

But they have also undergone structural changes. If the demand patterns that generated the prior equilibrium—patterns of work and travel and urban vitality—are not merely getting back on course, but are instead permanently reorganizing, then what we are seeing today is a structural shift.

If this is a cycle, patience is the appropriate strategy. If it is a structural change, patience is an existential danger to cities.

Subscribe to Field Discovery to follow the series. New essays every Tuesday.

Subscribe now

Welcome to the Povey Building

If you haven’t experienced it before, Thermoplastic Polyurethane 3D printing is surprising. It leaves behind a fine white powder as it turns a mostly-recycled stock material into a finished, environmentally sustainable, and aesthetically remarkable product. It’s even more surprising to see that powder dusting a historic brick masonry wall, collecting at the base of old-growth Oregon timber columns, inside a century-old building in the center of Portland’s Old Town Chinatown neighborhood.

This is the scene you’ll find when you walk into the production lab at HILOS. Engineers are designing and manufacturing shoes in the same room, something the industry abandoned more than a century ago. Using advanced manufacturing technologies, HILOS moves from a concept to a fully functional sample product in about one day. The industry standard is weeks or months, and the total production cycle from design to shelves is 18 months.

Co-working desks at WeWork—the late-2010s startup office trope—wouldn’t be suitable for the kind of work that HILOS does. During the height of the COVID-19 pandemic, the fledgling startup moved into Portland’s historic Povey Glassworks building. Like most of the buildings around it, the Povey building, originally built in 1905, had been converted to offices for knowledge workers during the “creative class” boom of the early 2000s and 2010s. And in 2020, almost overnight, it emptied out. The Povey building, alongside millions of square feet throughout the city, became stranded assets. As an office, the building had no real utility—but that didn’t mean it was useless. When HILOS found it, the space was vacant, the price was right, and it was a remarkable raw material for architectural reinvention.

HILOS leveraged that opportunity to create a new kind of work space. The former creative office, the generic kind of real estate that remains impossible to fill in the aftermath of the pandemic, became a hotspot for both designing and producing the next generation of sustainable footwear. Its architectural characteristics—tall ceilings, regular column grid, loading bay, natural light—embodied what Adaptive Urbanism calls the design imperative: structure that outlasts any single use. Because of its manufacturing history, the building required no building or zoning code updates. It was primed to adapt, and it did. Today, the Povey Building is thriving. It is productive, in all senses of the word.

Three Generational Trends

HILOS signals the convergence of three generational trends: urban cores are hollowing out, supply chains are in flux, and breakthroughs in manufacturing technology are bringing the next generation of products into sharper focus. A new urban industrial economy is bringing production to the core of American cities.

Is it simply reshoring? Not in any conventional sense. To this day, most reshoring attempts have failed. They will continue to fail if they attempt to replicate outdated manufacturing models that were built in a very different context. Conventional mass production is synonymous with astonishingly low wages, inhumane work conditions, and environmental degradation. Continuing those practices—let alone replicating them in the United States—is neither ethically defensible nor economically viable.

Today’s market realities are distinctive. Consumers expect new products monthly, not yearly—and they reward companies that disclose sustainable and humane practices. Trade policies swing unpredictably between administrations, but they are trending worldwide toward protectionism. These shifts demand manufacturing that can adapt: batch-on-demand production, just-in-time delivery, supply chains that absorb volatility rather than shatter under it.

The infrastructure for this transformation already exists. Advanced manufacturing has moved from peripheral novelty to mainstream capability. And cities are burdened with millions of square feet of vacant space, especially the office buildings that once defined their downtowns.

Domestic production is viable only if the fundamentals of manufacturing radically transform. That means rethinking the product itself and how it is made.

Through the late twentieth century, companies optimized their supply chains by squeezing suppliers and chasing the lowest possible wages. They created an intricate, efficient, and shockingly fragile global production system, and the COVID-19 pandemic proved just how brittle it was. In its aftermath, resilience has replaced efficiency as the primary operational objective for companies that manufacture. That means more (and often redundant) suppliers in more countries, especially on-shore and near-shore providers. Flexibility as risk management has become the operating principle, and it is the economic imperative of Adaptive Urbanism.

For decades, 3D printing remained peripheral to manufacturing: useful for prototyping, embraced by hobbyists, but too slow and too limited for production at scale. Recent advances in materials, equipment, and automation have changed the calculus. Advanced manufacturing now operates at production scale, and it stands to reshape the relationship between companies, their customers, and their employees—and the geography of production.

Every day, HILOS shows that it is possible to make finished shoes in downtown America, and they are not the only one. Gildform designs, 3D prints, and manufactures custom jewelry on demand from its facility in Detroit, and seamlessly transitions creators into a vetted global manufacturing network for production at scale.

Honda has operated an advanced design facility in downtown Los Angeles since 2012. In 2021, the company merged “all auto development, planning, purchasing, strategy and manufacturing into one company, thus strengthening what Honda calls a one-floor approach to the product development process,” (Honda News). The production R&D facility creates full-size models and prototypes that are moved into scale production at a facility in Ohio.

The Next Downtown

For decades, downtowns have served office workers, and today they need a new purpose. The simple intuitiveness of office to residential conversion makes it popular, but architectural analyses have repeatedly proven that it is only feasible in a narrow subset of the building stock. Even feasible conversions require heavy public subsidy or yield market-rate units, often both. Office-to-housing is neither scalable nor a solution to housing, especially affordable housing.

Office to production is largely absent from the architectural, political, and public conversation. The value proposition is clear, and it should be a call to action for civic leaders, economic development professionals, real estate developers, and architects: this is a once-in-a-generation opportunity to (re)define urban production. Achieving this goal will take a cross-sectoral coalition, because the economics of manufacturing and consumer demand are changing faster than zoning, development, and real estate finance can respond. This an example of the pacing problem, and overcoming it requires coordinated action.

When design and manufacturing share the same space, each accelerates the other—ideas move from concept to prototype to production in days, not years. A new method of making will reshape what products look like, how urban economies function, and what vacant buildings become. This future is neither speculative nor theoretical. A new economics of production is emerging at a time when the financial infrastructure of American cities is strained to the edge of collapse. The fundamentals have changed; there is no return to a pre-pandemic equilibrium.

The next urban economy will emerge when entrepreneurs, developers, planners, and communities recognize what they already have: an abundant supply of vacant buildings waiting for a new purpose. This is production potential.

This post is part of a series about Production Potential, a long-term project that began as an installation for the 2025 Venice Biennale of Architecture, a collaboration between Versostudio, an Arsenalia company, and Field States. If you like it, subscribe to Field Discovery and receive posts about Production Potential, Adaptive Urbanism, and the Methods & Tools we use as an adaptive design practice.

Subscribe now

You send your prompt, the AI starts thinking, and you casually check your email. It’s something about HR. You delete it, and barely have time to think what am I doing here, really? before the AI delivers a thoroughly researched report with clearly presented strategic actions.

For decades, large consultancies have offered high-credibility knowledge at a scale and speed that justifies premium fees. At the most basic level, they sell time: everything comes down to the billable hour. That was the right building block when information was scarce and expertise was hard to transfer.

Research, analysis, and first drafts are suddenly abundant. When knowledge at a vast scale and near-instant speed is available to everyone at a negligible fixed cost, selling time no longer works. The traditional consulting model is dying.

Smart, well-resourced consultancies are still doing good work, but the basic economics behind them no longer add up. This problem is widely recognized, and it has been debated ad nauseam. We are interested in a far more interesting question: what comes after the consultancy? Or, more specifically, what is the urban design and real estate strategy business fit for an era of AI?

The LegacyCo Trap

Usman Sheikh, writing at Framebreak, has been tracking this shift across professional services. He frames it as the move from “LegacyCo” to “NewCo.”

The old model (LegacyCo) assumes that competitive advantage comes from scale: huge teams of young analysts converting messy reality into tidy outputs, overseen by senior experts whose judgment is scarce and expensive. This is often called the “pyramid” model.

It worked when information was hard to access and hard to process. It fails when the unit economics of analysis shrink to near-zero, and the firm is composed of only senior partners with enterprise AI subscriptions. Consulting, as a practice, is still valuable, but the economics and the deliverables (billing for the cost of producing an analysis, then walking away) no longer work.

The predictable response for consultancies is to use any tool possible to gain advantage: copilots, chatbots, and AI-powered software. These are all single-point solutions that bolt on to the same basic machine. They may yield marginal productivity gains, but once everyone has the same tools, no one has a meaningful advantage. The consultancies quickest to implement new tools will see a comparatively slower delay in their decline, but without a fundamental shift, decline is inevitable.

The shift to NewCo isn’t about adopting AI, it’s about redesigning the practice around what AI changes:

→ Speed is no longer scarce, content is no longer scarce, PDFs don’t matter.

→ Judgment, accountability, and real-world proof are mission critical.

Differentiation moves to whoever engages with the underlying dynamics of change by actually delivering projects. The answer is to create learning systems: frameworks that are tested in the world, updated quickly when they break, and encoded into tools and workflows that compound. This is a fundamentally ‘Human + Machine’ paradigm.

What Replaces the Billable Hour

1) Proprietary frameworks, published openly

As the world floods with competent but generic output, the premium shifts to original thinking: frameworks that depict reality in a way that helps people make difficult decisions, and design strategies that reflect place and community. This isn’t a veneer of marketing pablum—it is the foundation of shared language and trust.

There was a time when firms held tightly to their intellectual property, fearing that others would steal their ideas. Publishing openly builds momentum around transformational ideas that create new value for everyone associated. It demonstrates the kind of credibility that earns you the right to be in the room when decisions are made.

2) Delivery as proof

Today ideas are cheap; it’s proof that matters. In the built environment, you don’t get to hide from outcomes. Capital gets allocated, buildings get built, and communities absorb the second-order effects. Ideas have consequences, and if you’re wrong, the world tells you. Differentiation comes through delivery and experience.

The next model of advisory work requires staying engaged through execution, being accountable for outcomes, and using real-world feedback to improve the frameworks.

3) Interdisciplinary capability from the ground up

Adaptive Urbanism doesn’t live in one field. The transformation our cities need—driven by innovations in flexible architecture, productive districts, flexible finance, alternative governance models and more—doesn’t fit inside a single discipline. It sits at the interface of architecture, real estate, policy, finance, and technology.

Forward-facing clients are ready to tilt at systems problems; they’re ready for adaptive urbanism. But most of their vendor firms are organized around a single lens, like architecture or development finance. Clients are left to manage complex projects by maintaining the strategic vision while doing the day-to-day work of integrating firms across silos. This happens slowly, in fits and starts, and at great expense. In a word: badly.

A firm designed for Adaptive Urbanism has to be interdisciplinary by default. It goes beyond partnering and coordinating; it is structurally built to cross boundaries without losing time to handoffs. It is a system in itself.

Field States is Adaptive-by-Design

Last week we argued that cities face a pacing problem: the rate of change now exceeds the response capacity of the institutions designed to manage them. Development cycles run five to seven years. Policy cycles longer still. Meanwhile, work patterns, capital markets, and technology shift in months.

Adaptive Urbanism is a response to the pacing problem that is gripping cities, and a provocation that cities must become capable of learning, not just optimizing. Developers, policy-makers, and real estate investors need to make decisions at the pace of change, and implement strategic programs and deploy capital that positions them to continue succeeding under volatile conditions. They need a strategic partner that can support them in learning faster than conditions change, and proving it through delivery.

Developing the Adaptive Urbanism framework gave us an excuse to ask a hard question: if we believe cities must become more adaptive, why wouldn’t our own practice need to be adaptive-by-design? And if our clients need interdisciplinary integration, why would we organize ourselves as a set of separate disciplines stitched together by meetings?

So we’ve structured Field States as a loop; an integrated learning system where research, delivery, tools, and proof points reinforce each other. We’re early. This essay is by no means a victory lap (it’s an example of Proprietary Frameworks, Published Openly). It’s deliberate design.

At a high level:

• Field Discovery generates original research and frameworks that we publish openly—building shared language and a body of work that compounds over time.

• Field States turns that work into delivery: projects where our frameworks meet reality. Every project produces evidence that flows back into the learning system.

• Field Tech codifies our methods into repeatable workflows and standalone tools that increase speed, defensibility, and throughput, so we can deliver more value without reverting to the billable-hour trap.

• Field Ventures are the projects that grow to stand on their own feet. The system spins out real assets that bring long-term stability to the business.

Under conditions of rapid change, you can’t afford to separate research from practice, or practice from tools, or tools from proof. The loop is the product.

What Comes Next

The firms that survive this transition won’t be smaller versions of AECOM or Arup or Gehl. They will be NewCo architectures: technology-enabled, outcome-oriented, differentiated to topics and places, and built around proprietary knowledge systems that improve through use.

Over the coming weeks, we’ll take practical deep dives into how we make this real: showing the work, not just the theory.

Over time, as we build our creative differentiation, we will continue to make our knowledge available while building a library of proprietary data, insight and knowledge.

We’re starting with Production Potential: a collaboration between Field States and Versostudio (an Arsenalia company) for the Venice Biennale that explores how adaptive reuse, industrial capability, and new forms of value creation can be designed as a coherent system.

The challenge of implementing Adaptive Urbanism is immense; the conventional consultancy is not equal to the task. Field States is designed from the ground up to help cities succeed in a faster future.

Field States is building a design practice for cities that learn. Subscribe to Field Discovery for research on adaptive reuse, productive cities, and the future of urban practice—and follow along as we test these ideas in the world.

Cities, and the buildings that compose them, are poised to thrive in the decades to come. That may sound like a naive or counterintuitive claim, as the prospect of a cascading failure in property markets and institutions looms large. Cities face a very real danger. But a clear-eyed understanding of the causes that are driving change in the built environment reveals the potential for a fundamental paradigm shift.

Adaptive Urbanism is a move from a prediction-based approach that assumes a knowable future to a learning-based system that has the capacity to adapt as unexpected futures emerge. This essay outlines Adaptive Urbanism as an architectural, economic, institutional, and technological practice—and a generational opportunity for investment in urban revitalization.

The building without a purpose

In San Francisco’s South of Market district, a nearly-new office building sits largely empty. It is functional, well-designed, professionally financed, and approved by every committee, signed off by every expert. By every conventional measure, nothing is wrong with it. And yet it has become a liability to its owners and investors, worth less than the debt used to build it, generating less revenue than its operating costs, contributing less to the city’s tax base than the surface parking lot it replaced.

This building is not an anomaly. In fact, it is an example of a whole generation of buildings that have lost their purpose.

And it is not just an American problem. From Canary Wharf to La Défense, from Frankfurt’s banking district to Toronto’s Financial District, cities that embraced modernist planning and built their economies around knowledge work are discovering the same uncomfortable truth: the buildings they constructed for the late twentieth century do not fit the early twenty-first.

Across the United States alone, $1.5 trillion in commercial real estate debt is maturing into a market that no longer wants what it financed. Office towers designed for a world of five-day commutes now operate at 50% occupancy—totaling about 1 billion square feet of vacancy across the country. Hollowed-out high streets, vacant strip malls, and zombie mega-malls mark the landscape of retail transformation. Suburban office parks, the backbone of post-war commercial development, are discovering that their highest and best use may no longer exist.

The vacancy we are witnessing today is only the first tremor of an approaching earthquake. Many companies remain locked into leases they signed before the pandemic, before hybrid work, before the current wave of AI-driven productivity gains. They are now operating with a reduced staff of fully remote workers, and they are choosing to let their spaces sit vacant while they wait for their pre-COVID leases to expire—leases that will not be renewed. Vacancy rates will continue to rise.

It is tempting to blame the occupancy crisis on COVID, but the mismatch between what cities have built and what cities need is not a temporary disruption. Something more fundamental is happening, a structural shift, and our urban institutions were not designed to handle the new urban condition.

The Pacing Problem

Cities face a pacing problem: the rate of urban change now exceeds the response capacity of the institutions designed to manage it.

Think of it as two clocks running at different speeds.

Clock One measures how fast cities can respond to any new condition. Development cycles run five to seven years from building conception to completion. Zoning amendments take two to five years to write and adopt. Comprehensive plans project fifteen to twenty years into the future. These timelines made sense when they were designed. They allow for deliberation, community input, and careful allocation of capital.

Clock Two measures how fast conditions actually change. Hybrid work restructured where knowledge work happens in eighteen months. Retail formats now cycle every three to seven years. The average lifespan of a company in the S&P 500 has compressed from sixty-one years in the late 1950s to under eighteen today; and it continues to fall. Climate pressures are forcing physical transformation on timelines that planning processes cannot accommodate.

Clock one is the pace of supply, clock two is the pace of demand. In the post-WWII era, both clocks ran at comparable speeds. From offices to housing to retail, it was possible to observe emerging demand, forecast future needs, build to meet them, and operate profitably before conditions shifted again. Prediction worked because the future resembled the present long enough to act on it.

When Clock Two runs faster than Clock One, prediction fails. By the time you finish building, the world has moved on. The building that made perfect sense when it was conceived no longer has a purpose by the time its doors swing open for the first time. The housing crisis that grips cities like Portland and San Francisco is a stark example of the pacing problem.

The gap between Clock One and Clock Two is widening. Four factors are pushing them apart: new technologies such as AI and robotics are restructuring work faster than any previous technology; climate change is forcing physical transformation across entire regions; bureaucratic burden is slowing down the pace of development; and demographic shifts are fragmenting demand patterns that have held stable for decades. Each of these alone would strain our political, financial, and social institutions. Together, they are overwhelming the urban status quo.

The Cascade

The pacing problem does not produce gradual adjustment. Under these conditions—when structural mismatches align with financial vulnerabilities and institutional rigidity—it can trigger cascading failure.

The stranded building in San Francisco is not failing because of bad decisions. It is failing because it was designed for a world that no longer exists, and the systems that created it cannot adapt fast enough to respond—a structural mismatch.

You might expect markets to self-correct: vacancy rises, rents fall, new tenants arrive, equilibrium restores. That is how healthy markets work. But when the mismatch is structural rather than cyclical, decline can feed on itself in ways that defy simple correction.

When cities cannot adapt to the circumstances they face, the consequences extend far beyond individual buildings. The pattern is grimly predictable. Vacancy rises, rents fall, and property values collapse. At that point loans default and municipalities without a tax base are forced to cut services. The quality of the downtown fabric declines. Communities fragment as investment flees.

Tax bases erode as commercial values collapse. In most American cities property tax is the single largest revenue source, contributing between one third (Boston, San Francisco) and close to one half (New York, Portland) of the general fund operating budget. Revenues enable cities to function; when they cannot respond to the conditions they face, their legitimacy erodes.

If commercial real estate debt defaults accumulate at scale, the ripple effects could destabilize financial markets that have long treated property as a bedrock of stability. Pension funds, insurance companies, and banks have built their balance sheets on the assumption of predictable real estate returns. That assumption is now in question. As the property analyst Antony Slumbers has observed, commercial real estate no longer operates like a bond. The stable, predictable income streams that underpinned traditional valuation models have given way to something far more volatile and uncertain—more like a stock.

These forces are converging to push cities beyond the threshold where a market correction is possible. Each stage of the cascade makes the next more likely. Economists call this a “doom loop.” Once it starts, breaking free requires intervention that our institutions are poorly equipped to provide. As the paradigm shifts, urban property markets and the institutions that underpin them are on the brink of collapse.

The Forgotten Capacity

There is reason for optimism. Embracing the paradigm shift—especially before a structural collapse—is a remarkable opportunity for reinvention. The simple solution to the pacing problem is in the very fabric of cities.

Before the twentieth century, cities evolved through continuous, distributed adaptation. Buildings changed use as conditions shifted. A house became a shop. The shop became a warehouse. The warehouse became apartments. Districts transformed as economies restructured. Knowledge accumulated in the physical fabric of the built environment through countless small adjustments by countless small actors.

This vernacular, adaptive process of urban development was messy and slow. It was often unjust in its resource allocations. But it worked. Cities learned and evolved.

Consider any historic district that has survived centuries of transformation: Barcelona’s Eixample or London’s Soho. As global and local economies and demographics shifted, the buildings changed, the uses changed, the people changed—and through it all, the districts remained vital. Something in their structure enables continuous adaptation.

What happened? In the twentieth century, we made what might be called an urban Faustian pact. We traded slow organic learning for fast formal planning with Euclidian zoning that separated urban uses horizontally rather than vertically. Governments centralized decision-making. Developers standardized building types. Investors financialized real estate into tradeable assets. The vernacular capacity for adaptation was systematically suppressed in favor of prediction, control, and scale.

Under stable conditions, the pact made sense. Post-war development needed to move fast, and it did. Cities grew at unprecedented rates. Standards of living rose. The modern built environment took shape. Phrases like suburb and central business district replaced neighborhood identities as the primary marker of urban character.

But the pact had a cost. We optimized for a world of predictable change and locked ourselves into systems that cannot handle anything else. Now that the stable regime conditions have shifted, we are discovering that cities have been stripped of the capacity that made them resilient for millennia.

This capacity can be recovered. We can once again design buildings and cities that learn, not by returning to pre-modern vernacular practice, but by combining vernacular logic with contemporary tools that our predecessors never had.

Adaptive Urbanism

Adaptive Urbanism treats cities as learning systems rather than static fixtures. The core insight is simple. If we cannot predict the future, we should stop pretending we can. Instead of optimizing for a single forecast, we should build capacity to adapt as futures emerge. Buildings should be designed to evolve. Instead of governing through rigid rules, we should create flexible, outcomes-oriented frameworks that enable continuous adjustment.

From an architectural perspective, Adaptive Urbanism has two faces, one looking backward, one looking forward.

Retrospective adaptation addresses the buildings we already have: the stranded offices, the empty retail stores, the vacant strip malls, the underperforming assets that fill our hollowed-out cities. These buildings are not worthless; they are a remarkable raw material for reinvention. With the right approach, they can be converted, repositioned, and reactivated for uses that match current and future demand. The trillions of dollars in “stranded” assets represent an immense, once-in-a-generation urban reinvestment opportunity.

Prospective adaptation addresses the buildings we will create: designing new construction with embedded capacity for future transformation. This means flexible architecture that can quickly and sustainably shift between uses without structural intervention. It means financial structures that value optionality rather than punish it. It means governance frameworks that enable adjustment rather than prevent it.

Both faces share a common principle: under conditions of uncertainty, flexibility is not a luxury; it is the lowest-risk strategy available.

The Operating System

Making Adaptive Urbanism real requires an operating system: a coordinated set of capabilities that work together. There are four fundamental imperatives, and all of them must function in concert.

The Cognitive Imperative. Cities cannot adapt to what they cannot see. Governance fails not primarily through bad policy but through failed perception. Cities are run with lagging indicators that were designed for stability, indicators like census data that arrives years late, or economic statistics that aggregate away the patterns that matter. Adaptive cities also need real-time sensing, pattern recognition, and feedback loops that reveal emerging conditions before cascades become irreversible.

The Institutional Imperative. Information is useless without the capacity to act on it. Adaptation requires sustained effort over time horizons that neither markets nor government electoral cycles can maintain. New entities, such as Civic Corporations, District Trusts, Cooperatives or even revamped Tax Increment Finance Districts must be designed with the capacity to steward adaptation across decades. These entities own common assets, manage coordination problems, capture value uplift, and reinvest in the next cycle of improvement.

The Economic Imperative. Traditional real estate finance assumes every parcel has one optimal use: find it, build for it, operate forever. That doctrine is failing. We need valuation frameworks that move beyond static highest and best use analysis to recognize the value of real options, the ability to pivot as conditions change. Adaptive assets hold options on multiple futures. Capital must recognize that adaptation is the lowest-risk strategy under conditions of volatility. A building that can transform is worth more than one locked-in to a single use. At the portfolio level, investors can manage uncertainty through diversification across adaptive asset types, reducing exposure to any single demand scenario.

The Design Imperative. Buildings and districts must be physically structured for change. As Stewart Brand articulated in his iconic book How Buildings Learn, this means understanding architecture as layered systems where different elements change at different rates: structure lasts centuries, services last decades, fit-outs last years. It means designing buildings that can accommodate multiple future scenarios rather than optimizing for one. It means creating districts with the fine grain, connected streets, and shared infrastructure that enable adaptation at every scale.

These imperatives are not a menu to choose from, they are an operating system to install. Sensing without stewardship generates data no one acts on. Stewardship without finance lacks resources to invest. Finance without design has nothing flexible to invest in. The system works together, or it does not work at all.

What We Are Building

Field States is making Adaptive Urbanism real.

We are not waiting for conditions to improve, or writing reports that sit on shelves. We are building an integrated system: research that surfaces patterns, design and advisory and development management that creates actual places, technology that supports continuous learning, and ventures that demonstrate what is possible.

Our work is grounded in a simple conviction: the public and private sectors can align. Adaptive Urbanism is not a zero-sum game. When cities succeed, communities succeed. When projects succeed, investors succeed. The loop can be generative rather than extractive, if we design it that way.

We believe in learning in the open: we will publish our research, share our methods, and build tools that others can use. Our ambition is not to monopolize Adaptive Urbanism but to make it the norm. The stakes are too high for any other approach.

What Comes Next

This essay is the first in a series. Over the coming months, we will explore each element of Adaptive Urbanism in depth.

We will examine the pacing problem, for example: how the two clocks fell out of sync, why conventional responses fail, and what the cascade looks like when it takes hold. We will recover the forgotten history of urban adaptation, showing how cities learned before planning and prediction replaced learning. We will work through each imperative—cognitive, institutional, economic, and design—with the specificity that practitioners need to act.

We will look through the lens of different stakeholders. What does Adaptive Urbanism mean for developers navigating uncertain demand? For financiers underwriting assets in volatile markets? For mayors trying to maintain tax bases and community vitality? For institutions anchoring districts in transition? As we work across these stakeholders, we will find opportunities to align incentives, mitigate real and perceived risks, and collect shared metrics and indicators.

We will look at real places: districts that are adapting successfully, buildings that have transformed, governance models that work. We will also look at failures, the projects that stalled, the approaches that misfired, the lessons we are still learning.

We will explain why we have structured Field States as an integrated learning loop, reflecting our conviction that research, practice, and technology must work together if cities are to learn. We are committed to sharing the tools we create: frameworks, indices, and methods that help cities and investors navigate conditions of uncertainty.

Our goal is to build a shared language across a community. The challenges facing cities are too large and too urgent for any single organization to address. We need a field of practice: researchers, practitioners, investors, and policymakers working from common premises toward common goals.

Cities that learn will outperform those that do not. Can we build the institutions, tools, and practices to make urban learning possible? Can we do it fast enough to matter?

We believe it is possible. Join us and find out.

Matthew Claudel & Simon Giles

About Field States

Field States creates places that thrive in conditions of change. We work together with forward-facing cities and developers to discover the potential in buildings, public spaces, and districts—and turn that potential into lasting cultural and economic value.

Subscribe to Field Discovery to follow the exploration. New ideas every two weeks.