“Missing Middle Housing”: missing in action for 30 years

Housing completions by # of units in building

Missing Middle Housing” refers to a broad spectrum of smaller-scale multifamily building types that occupy the middle ground between single-family detached and large apartment buildings. At its edges, the term can cover rowhouses and 5+ unit apartments, but in my mind, the core of the “missing middle” are “plexes” — two, three, or four flats on a single city lot. Small apartment buildings like these line the side streets of most of America’s walkable urban neighborhoods: two- and three-flats in Chicago, Polish flats in Milwaukee, three-deckers in Boston, or brownstones in New York.

Census data show that 2-4 unit buildings have almost evaporated over the past generation. As recently as the early 1970s, over 1/4 of multifamily units built were in small buildings. Now, it’s only about 4%.

Despite the steep mid-1970s recession, over 10X as many units were built in small multifamily buildings as today. In boom years, 14X as many were built. It was during the Reagan era that these buildings broadly fell out of favor.

2-4 unit buildings have been a relatively small share of all multifamily construction in the postwar era. Since the 2008 recession, they’ve practically disappeared — with only a few thousand units built nationwide every year:

Housing completions by # of units in building

Next up: Wait, you mean today’s apartment boom isn’t actually that big?

Friday photo: Be careful what your zoning asks for

In the Penn Quarter neighborhood, as Mark Jenkins wrote in the Washington Post, “arts spaces are mandated by zoning, yet the arts scene is hard to find. Nearly a dozen exhibition spaces populate the area… most don’t have access to the street and aren’t clearly identifiable from outside. They are in lobbies or tucked away in office building interiors.” Here’s one, Terrell Place at the prime corner of 7th & F, that’s not entirely hidden — but also not exactly shouting its noble story to the public. Yet it’s somewhat striking in that it wastes what could be prime retail space across from the National Portrait Gallery and the Verizon Center arena.

The “arts” designation for the “Gallery Place” neighborhood was intended to foster a vibrant community of commercial art galleries alongside working artists, all in the shadow of the Portrait Gallery. Yet instead, there are scarcely any commercial galleries, which require considerable staff investment, and instead many pointlessly large office building lobbies which feature above-average displays of plop art. Perhaps some lawyers and lobbyists have their lives slightly enriched by walking past these paintings, but the general public now derives scant benefit.

The designation is so broadly written that at least two suspiciously spacious bars on E Street — Penn Social and Hill Country Barbecue — both owe their existence to the “arts” requirement. I’m glad that Penn Social has room to spare; it’s a reliable go-to for large events that might not otherwise fit into such a convenient downtown location. But was that really the intent of the zoning?

The situation illustrates the difficulty of trying to define great places, which depend upon a lot of “we know it when we see it” subjectivity, through legal means like zoning. Now that quality retail has emerged as the prime placemaking amenity — anywhere can have open space, but retail’s a lot more difficult — it’s also become something worth subsidizing for its placemaking value. Yet before you can subsidize something, you have to define it.

Doing so will be fiendishly difficult for public entities like municipalities, which will also run into arguments about whether retail is a public good or a private good. However, it should be easier for private entities seeking to cross-subsidize across revenue streams.

CNU conversations: Introducing the Place Corp

west block

Findlay Market in Cincinnati, where a nonprofit market manager and CDC are working alongside private developers to create a mixed-use destination based around small, local shops.

Our earlier lunchtime conversation in Dallas quickly morphed into some constructive criticism of the community development field. We’re doers, so we delved into some of the mechanics of how a community could control its destiny, and ensure affordability over longer time scales.

In theory, Community Development Corporations have tremendous capacity to manage community change, but even where good CDCs exist — as in Boston and Chicago — their efforts are often quickly overwhelmed by the monstrous quantities of capital involved in gentrification. CDFIs were started to help leverage greater sums of capital, but never quite seemed to move beyond a little niche.

The work that community development organizations do is so labor-intensive that it doesn’t end up being very scalable. Some of that is good — community organizing and small-scale development in disinvested neighborhoods is tough. Some of that is bad — there’s so much bureaucracy involved in securing grants or tax-credit financing (or securing property for a Community Land Trust), and in complying with restrictions on spending that money, that staff capacity gets stretched thin.

(As a side note, why don’t we at CNU talk more often with the community development sphere? Sure, there’s some buy-in when individuals from CDCs get invited to present, but otherwise there’s scarcely any overlap. Perhaps siloization is to blame: CDCs/CDFIs have their own conferences, their own language, their own processes.)

Yet surely there are other vehicles that can provide a middle ground of financing for community development. In a gardening analogy, LIHTC-funded, CDC-built subsidized housing is like a trickle of water from a hand-carried watering can, and Wall Street money is like turning a fire hose onto your flowerbeds. Surely there’s some happy medium: a way to use local capital within a neighborhood, to fund incremental, community-positive projects that make a solid but uneventful return, over the long run.

There’s recently been a lot of renewed buzz about low-profit corporations and B-corps. Richard Plunz’s magisterial history of New York City housing shows us that many of the most innovative efforts at community-building there stemmed from just such endeavors, many underwritten by philanthropists or unions who sought investments with steady returns.

We discussed some of the innovation going on with cooperatives — the most established of low-profit, broadly-owned corporations. I talked up some experiments in the Upper Midwest, like the River West Investment Cooperative in Milwaukee, the considerable excitement around the Northeast Investment Cooperative in Minneapolis, and efforts to minimize business displacement along the Central Corridor light rail. In both cities, local capital is fostering communities of cooperative businesses in “emerging” neighborhoods. And unlike the Mondragon cooperatives, these coops are focusing on serving neighborhood needs, rather than exporting. Nationwide, a few credit unions (like Self-Help in North Carolina) have built substantial commercial-lending businesses, even in real estate.

But as much as we liked the idea of cooperatives, not everywhere has the fertile legal or cultural setting of Minnesota or Wisconsin. While cooperatives are great for certain kinds of businesses, they’re not for everyone: They can raise modest sums of capital, but struggle with large sums; their structures aren’t always flexible enough to accommodate multiple capital classes; their plodding, consensus-based nature makes them resistant to entrepreneurs’ bold ideas.


The history of Minneapolis’ West Bank neighborhood is an instructive example of co-ops’ promise and limitations. Uniquely among the ’60s hippie havens, federal urban renewal was leveraged to place most of the neighborhood’s housing stock within limited equity co-ops or Section 8. Alas, idealism and local control proved unable to adapt to either obsolescence or succession [PDF]. Photo: Michael Hicks, via Flickr.

Co-ops are a proven success within steady markets that provide community necessities, like workforce housing or groceries or child care or bike repair or bakeries — but perhaps aren’t the best vehicle for more colorful businesses. In many ways, commercial real estate development is just such a business.

Which brought us to the big small-business finance innovation of recent years: crowdfunding. Numerous “community investing” vehicles have been floated, like direct/P2P lending, lending through community financial institutions, conventional Kickstarter-style rewards (perhaps with interest paid in-kind), equity investments, etc. Crowdfunding has become almost de rigeur in certain popular, high-growth businesses. It first swept thin markets for trinkets that would typically be distributed online — e.g., board games. As the dollar amounts have grown, and as regulators have gradually warmed to the idea, platforms specializing in more capital-intensive bricks and mortar projects have sprouted for sectors like breweries and restaurants. Business models that defy categorization, like the McMenamin’s brand of alcohol theme parks, might be particularly well-suited to crowdfunded equity.

Several new platforms have emerged within this space. I’m most familiar with Fundrise. It’s evolved its primary offering away from common equity investments and towards what it calls “project payment dependent notes,” a hybrid of mezzanine debt (it’s debt, with a promissory note) and preferred equity (even though there’s a pre-determined target return, dividends are only paid when there’s adequate cash flow). For small developers, crowdfunding — even in the currently half-baked, accredited-investors-only form — holds significant advantages, namely cheaper, slower money, plus broader community buy-in.

Interestingly for our purposes, at least one nonprofit community development institution has begun to use Fundrise as part of its capital stack, and in particular in an attempt to save a for-profit community institution. The San Francisco Community Land Trust used Fundrise to raise its 38% equity stake (a local nonprofit offered the other 62% as equity) in an attempt to buy a mixed-use building housing the nation’s oldest African American bookstore, in a neighborhood otherwise ravaged by urban renewal. (Alas, the attempt foundered when the landlord changed the offering price.) All that the city could do was to landmark the building, perhaps in an attempt to reduce its market value, and to offer the business inclusion on a “legacy business” roster.

Mezzanine debt seems to also be the sweet spot for small business-oriented lenders like ZipCap. Their business model mirrors the built-in customer base of a cooperative, “recruiting an ‘Inner Circle’ of customers who pledge to spend a set amount of money in a fixed period of time.” Instead of relying upon just that circle for funds [as a co-op would do with equity memberships, or Member Loans], ZipCap uses those pledges as collateral for outside loans.

Combining all of the above vehicles — not-for-profits, low-profit cooperatives, and crowd funded for-profits — into interrelated entities might solve the scalability problem by allowing each entity to contribute its own strengths to a capital stack:

investment co-op waterfall

We called this entity a Place Corp, which has the ability to construct a capital stack from various tranches of capital, while keeping control within the community. (Outside investors and developers can help to build pieces, with the Place Corp hopefully maintaining some control via equity, but are treated merely as means towards an end.) Investors can choose from a wide variety of risk/return combinations, and can invest either time or money, choosing negative returns (gifts, grants), zero return (co-op equity), or modest returns. The Place Corp is diversified across property types, but not in location — like the family firms or community banks of yore. The durable, long-term returns of placemaking will create financial rewards for some investors — but for most, the place itself will be the return.

Next in the series: Where? And what sort of services might a Place Corp provide beyond just financing?

Thanks to some of those who contributed to this series of conversations: Karja Hansen, Matt Lambert, Russell Preston, George Proakis, Padriac Steinschneider, Seth Zeren

Friday photo: DMU in Trenton

Hello DMU - NJ River Line

Look, an LRV with no catenary! The diesel powerplant in the middle (note the stained exhaust vents) produces its characteristic vibration and noise, and unlike a commuter train, the passenger compartments are all subjected to it. It’s also definitely slower; by one calculation, an electric LRV could do the same route 18% faster. It seemed out of place in the urban centers on either end, but more appropriate in the suburban villages in between.

These might be appropriate for other “interurban” lines with wide stop spacing and relatively low station-area densities. Given that most suburban areas are already arrayed along roads rather than rails, and that the train isn’t much faster than a bus, perhaps this isn’t the panacea for suburban transit.

CNU conversations: If CBAs are broken, should they be turned upside down?


Redevelopment of Alexandria public housing near Braddock Metro in 2013. The slightly taller buildings with gables in the back include replacement units.

A subsequent conversation at CNU bemoaned how “tollbooth zoning” (as Ben Ross calls it) has turned everyone into their worst nightmare of a money-grubbing Chicago machine pol, rasping “ubi est mea?” over a cigar.

In Chicago, it was simple — you paid, you played. You want to build something? Fine, pay up and we’ll talk. In cities today, it’s pretty much the same. We’re systematically under-zoning (and over-planning) everything to maximize the possible value granted through zoning relief, and demanding the difference back through legalized bribes we call community benefits agreements.

Yet because CBAs’ contents are up for negotiation at one point in time, communities end up with whatever’s convenient, not what’s actually needed. The results can be baffling. One local municipality has a surfeit of small, black-box theaters that don’t get used, since theaters can only be purchased in increments of one, and the cheapest performing-arts giveaway is a black-box theater. (Someday, I’ll pull together a tour of the laughable “arts spaces” that zoning’s required around Penn Quarter.)

The most obvious solution would be to buy upzones using cold, hard cash on a per-foot basis, but that’s not allowed — if it’s a tax (or impact fee), it requires a nexus. So instead, communities get whatever the developer feels generous enough to give. Not to mention that the entire process, since it’s all boils down to political power, favors those who already have political power — large, well-connected developers vs. squeaky-wheel communities. Small developers get shut out of any opportunity to build, and disempowered communities (the homeless, for instance) never get a chance.

Now, since we’ve established that the zoning is entirely arbitrary anyways, why not flip the equation and start with the community benefits? The quantity of stuff permitted on the site ultimately doesn’t matter as much as the quality of what goes there.

This “upside down pro forma” is already being done in several instances, notably in situations where municipalities are seeking to maximize affordable housing output. For instance, cities that have committed to 1:1 public housing replacement — as HUD’s Choice Neighborhoods program does — have long had to work backwards to find enough market-rate units to make the pro forma pencil out. Alexandria has operated under a 1:1 public housing replacement policy since 1972 (“Resolution 830,” PDF), and the results are very impressive — seamlessly integrated urban fabric, both socially and physically.

Canada never had HOPE VI as federal policy, and its municipalities have a firmer hand in land use control. (For instance, Ontario’s Section 37 permits cash payments within a negotiated CBA. Sounds filthy, but actually cash is nice in that it’s easily measured.) So, working backwards from the benefits to the proposal isn’t unusual.


The art inside the Woodward’s enclosed retail court shows a police riot that took place nearby.

In Vancouver, the mind-bogglingly complex Woodward’s redevelopment used a public RFP process to stack a vertical mixed-use community with just about everything onto an abandoned department-store block right on Skid Row.

In Toronto, zoning bonuses paid for a 68-unit artist live/work space. Regent Park’s redevelopment has created 1:1 replacement units, plus 15% so far. As part of the development agreement, space for social infrastructure was built early on; the Daniels Spectrum “includes several state-of-the-art performance spaces, a locally run café, a green roof, and two floors for various educational, arts and community groups that have long operated in and around Regent Park. Many relied on informal or rented space, and some had been uprooted when the demolition began. The need for this kind of social infrastructure remained.”

Regent Park redevelopment continues

Regent Park in Toronto, where a commitment to better than 1:1 replacement housing (and additional retail and social services) is resulting in much higher densities.

That’s certainly one approach for maximizing the community benefits, but it introduces a few huge risks. It still only works for huge projects, it still relies on political power, and it still subjects the CBA to the fads of the moment. And what happens if the approach fails? The developer might just decide there’s nothing in it and walk away. (Another argument for good phasing.)

One (faraway) city sees its authentic retail as “basic infrastructure,” even amidst gentrification

Singapore Hawker Center

A case study (quite literally) about the struggle to maintain authenticity through small-scale retail operations is profiled in this week’s Economist. In Singapore, the government somewhat inadvertently nurtured one of the world’s most exciting food cultures through its “hawker centers.” Anthony Bourdain explained their history in “The Layover“:

The hawker centers of Singapore were a shrewd strategy to incorporate and control what was once a chaotic but pervasive culture of street carts. To control the health, safety, and traffic aspects, vendors were brought inside these covered structures still open to the air, but in stalls with regulated running water, refrigeration, and strict rules of food handling.

In short, they’re food courts with one key twist — they’re government-owned, often located within HDB public housing estates. Today, rising rents and rising living standards both threaten the culinary treasures that are so integral to Singaporean culture:

The main problem is that Singaporeans have grown used to paying prices that the market can no longer bear… The government, says Dr [Leslie] Tay, “has committed to providing cheap food for the masses”. With tiny flats and cramped kitchens, and with the number of two-working-parent families steadily rising, plenty of Singaporeans count on hawker markets for their sustenance. But with the first generation of hawkers retiring and their replacements paying market rents, food prices will certainly rise.

And as the masses change, so will the food. Some Singaporeans lament that a recent influx of immigrants from northern China has made their traditional Teochew or Hokkien [ethnic groups from southeastern China] favourites harder to find.

The article arose from a blog post by Dr Tay where he recommended instituting non-market mechanisms for allocating space in the government-owned markets: “Wouldn’t it be great if the officers can just recognize his status as a Heritage Hawker and help him get a stall at pioneer hawker prices?” (Pioneering stalls were subsidized in an attempt to entice hawkers off the streets, and today those rents are 90% below market rate, according to the 2015 case study by Tan Shin Bin for NUS’ Lee School.)

Indeed, but would you necessarily trust a government to make good decisions about what constitutes “good food”? Perhaps not, but plenty of America’s most successful public markets (like those in Milwaukee and Seattle) are run by nonprofits. An early effort at a nonprofit hawker centre in Singapore failed, but the government has signed new management agreements with cooperatives and “social enterprises,” as recommended by a 2012 panel.

A prior experiment in renting out a hawker center to a for-profit operator failed, and won’t be repeated; Minister Vivian Balakrishnan, whose ministry runs most centers, maintains that “Hawker centres are social infrastructure – and not an opportunity for property speculation or rent seeking by commercial entities.” Instead, nonprofit managers will get the opportunity to reinvest surpluses into social development and entrepreneurship, and prioritize operators that provide affordable food and heritage cuisines.

“Social infrastructure,” as Minister Balakrishnan put it, provides a useful frame for talking about “authenticity.” Cities are living things that require supporting infrastructure: physical infrastructure, social infrastructure, green infrastructure. They also need room to grow and change.

How can cities simultaneously build physical infrastructure (transportation, housing, workplaces) while also supporting and fostering social infrastructure, and regenerating ecological infrastructure? The answer might seem simple, but unlike physical infrastructure, social infrastructure is itself also living — and a somewhat fickle, slow-growing, and slow-rooting species at that. These make them tremendously resilient, but also slower to adapt than other forms of infrastructure.

As Singapore’s experience indicates, the answer probably lies with the social enterprise sector. How could those be structured? One potential answer’s coming up.

Friday photo: a minimalist memorial

Richmond riverfront

Three Days in April 1865,” a surprisingly moving exhibit composed of a terse timeline (pictured here: “Richmond surrenders”), plus primary-source quotes from the fall of Richmond. Obviously, the surroundings help: swirling waters, road and rail traffic on bridges old and new, the city skyline and the woods.

Sometime soon, the walkway (built over a dam that fed an adjacent canal) will again cross the river.