Office to Residential Conversions
Redeploying Real Estate to Address the Urban Doom Loop
Previously, I’ve discussed my research on the urban exodus in the last few years, as remote work allowed knowledge workers to move out to suburban areas, flattening the urban rent gradient. The cost of these shifts in work were born by urban office buildings, whose value has fallen substantially. Older office buildings in particular are are risk of becoming “stranded assets” in the wake of these technological disruptions, and cities broadly face the risk of negative spirals resulting from the loss of anchoring white collar work: the urban doom loop.
The Conversion Option
Our latest project — joint with Stijn Van Nieuwerburgh and Candy Martinez — is now out through the Hamilton Project evaluates the prospect of converting this now obsolete office space to residential uses. Despite the physical, regulatory, and financial challenges of these conversions (which we consider in detail); there are sound economic reasons that these conversions will likely happen anyway, as well as some key factors which mean that policymakers may want to nudge them along:
The loss in foot traffic in centers of cities is having a big impact on local retail consumption, the viability of mass transit, potentially crime as a result of fewer “eyes on the street” and ultimately losses in local government revenues.
Many older office buildings in question are also stranded assets from a climate perspective, producing large amounts of carbon emissions. Some cities like New York are starting to put carbon taxes on these buildings — you can see a local map of emissions and fines here.
Housing affordability remains a challenge in many cities, even those that have seen substantial remote work, like New York City.
Our project conducts a feasibility analysis and finds about 5-10% of “brown” or carbon-emitting office real estate can be plausibly converted into “green” residential buildings, typically in older buildings with narrow floor plates. While that’s not anywhere close to the majority of office space — reflecting the huge challenges that go into these conversion projects — it is a meaningful enough chunk to put a dent into office vacancies, thereby helping to stabilize that market. Converting these units would produce 367,750 rental units, equivalent to about 1.5 years of typical multifamily supply, hopefully moderating residential rent growth in the housing market, while adding new residents in the center of cities who will have short commutes (and so are more likely to go into the office). And finally it would take a large chunk of carbon emissions offline, addressing the broader decarbonization agenda.
These conversions will happen anyway as a result of the activities by private developers. But there are critical roles here for local and federal policymakers, especially considering the externalities involved. (emissions and urban vitality).
Cities typically need to make regulatory changes to zoning and building codes to facilitate these conversions. For example, in New York City, residential buildings are capped in density (FAR of 12) below the level of office buildings, and so many office buildings cannot be converted legally as of right to apartments.
Local governments may need to provide property tax relief to help conversion activity pencil out. You can think of such actions as a form of land value tax, in simply removing the disincentive to invest in real estate.
And finally, we identify tax provisions in the Inflation Reduction Act (IRA) which can go towards conversion activity. The White House has since clarified the use of IRA money towards these purposes, and is rolling out additional incentives — including subsidized interest rates — to facilitate such conversions.
To think carefully about these options, we have produced a public calculator which conducts a pro forma analysis of the conversion option. You provide inputs for a possible office conversion candidate, and the calculator returns whether or not the conversion is financially feasible — and, if not, the subsidy amount would make the resulting conversion pencil out. The model is also transparent that conversions to affordable housing units — a common desire among local policymakers — typically require additional subsidy in order to pencil out.
How Viable are Conversions Really?
I think it’s fair to ask whether conversions will really play out as we are anticipating, what are some other alternatives, and what are really the relevant headwinds.
First, it’s important to realize that NYC has a history of conversions and subsidies, with the 421-g program providing incentives for the conversion of about 13% of the office space in the Financial District. And of course New York has seen large scale redeployment of real estate for all sorts of reasons: the mass conversion of industrial space in SoHo into expensive lofts, for instance.
We’ve also taken a look at the Dutch experience: the Netherlands has actually produced 100,000 units through conversion over the last ten years — 15% of total housing construction. They have done so through the proactive development of state capacity: setting up centralized “transformation teams” which approach local developers and policymakers to provide information about concentrated areas of high vacancy suitable for conversions, providing logistical and technical support, and streamlining the regulatory bottlenecks around conversions.
Of course, redeployment of real estate can happen through teardowns as well as conversions. It’s just less ideal from the standpoint of “embedded carbon” — all of the carbon that went into the steel and the concrete which would have to get rebuilt; as well as in light of escalating building and labor costs. Most office real estate will stay put as is — either upgrading in quality to attract the top tenants, or reducing in rent to fill vacancies. And there will be some alternative uses as well — Universities, hospitals, lab space, and so forth — which will likely fill the gap. Another factor that helps is the concentrated nature of vacancies so far: JLL estimates that just 1% of buildings make up 13% of NYC office vacancies, while 10% make up 55% of overall NYC vacancy. Obviously, these buildings are prime candidates for conversion or redeployment in some other way.
These conversion targets probably work best in cities that have a strong enough housing market alongside a glut of office space. Surprisingly, as this CBRE report suggests, that results in a lot of conversions in some Rust Belt markets like Cleveland and Cincinnati. Alex Armlovich has a nice piece in The Atlantic focusing on office markets in these cities, highlighting how declines in rents there can make offices completely unviable given operating expenses, whereas superstar cities can manage with somewhat lower office rents. The task is a little more challenging in cities like San Francisco — which are suffering declines in both residential and office markets.
Since we started working on the proposal, we’ve also seen wild shifts in interest rates. Conor Sen is skeptical that conversions will really pencil out in these tighter financing environments. If so, that would suggest that we might be seeing a substantial amount of crowding out: higher government spending on manufacturing and fiscal deficits in general may be pushing up interest rates and then crowding out some activity on the conversion side. I think one silver lining is that interest rates for multifamily loans are still much lower than for residential mortgages — prepayment risk is so high these days that mortgages are averaging close to a 300 basis point spread over the relevant risk free rate; spreads for apartment loans are lower than that (more like ~185 basis points). But it’s a fair concern as well, one that might lessen in the coming months, and we might see additional government financing subsidies here too.
Finally, I think it’s worth keeping on the agenda even larger changes in building codes to help conversions pencil out. The key constraint in most buildings for conversion is the floor plate — it’s very wide in modern office blocks, and so to generate bedrooms with windows, you either need to cut a core in the center or side of the building, or else deal with a lot of dead space in the center. An alternative here is to allow more bedrooms without windows — as Charlie Munger likes in the dorms he designs, famously with UCSB and more under the radar with an existing dorm in Michigan. It’s not a policy without tradeoffs, but illustrates in a vivid way the challenges and choices cities face to remain vibrant and exciting places.