Housing: An irresistible force meets an immovable object

In Part 1 of this two-part series, I wrote that rising household sizes, declining homeownership, tighter lending standards, and a sell-off of single-family houses by the nation’s fastest growing demographic — senior citizens — will keep the for-sale building industry depressed for the next 10 years. Yet demographers expect the US to add 33 million residents by 2020 — so where will these people live and what part of the housing industry will be strong?

The answer, in a word, is rental.

The rental share of the housing market could surpass 41 percent by 2020, says Arthur C. Nelson, professor of city and regional planning at the University of Utah. Nelson is a numbers cruncher with one of the best big-picture housing forecast track records in recent years. He presented new data at the Forum on Land and the Built Environment in Cambridge, Massachusetts, jointly sponsored by the Lincoln Institute of Land Policy, the Harvard Graduate School of Design, and the Nieman Foundation for Journalism at Harvard University.

The US now contains about 39 million rental units. That figure will increase by 9 to 12 million by 2020, according to Nelson, meaning there will be an average rise in rental housing of more than a million a year throughout the decade. That’s a tremendous number, indicating that the rental market, the construction of units for lease, and the conversion of owner-occupied dwellings to rental will be very strong.

This dynamic is already well under way. Even as new home sales are the lowest they’ve been since records began being kept half a century ago, apartment absorption is stronger than it has been in a decade. This trend will only gain momentum, because it is the mirror image of declining demand for new for-sale housing. According to Nelson, the Urban Land Institute, and other experts, declining homeownership is the inevitable outcome of social, demographic, and economic trends that are pushing in concert.

A flood of new rental units could come in many forms — new apartment buildings; condo buildings converted to rental; accessory units attached to single-family houses; and existing owner-occupied houses that are flipped to rental — but the most popular locations will be mixed-use, transit-friendly neighborhoods.

“Forty-seven percent of households want urbane living; that’s a big change from 10 or 20 years ago,” says Nelson, referring to a recent National Association of Realtors (NAR) finding on that percentage of households that prefer to live downtown or in mixed-use city or suburban neighborhoods. “Back in ‘70s or ‘80s, people wanted drivable suburbs. Now 70 percent want to walk to discernable destinations, from transit to grocery stores. This wasn’t the case until recently.”

A growing imbalance

Three reputable studies — by NAR, the Robert Charles Lesser & Co. (RCLCo), and Nelson — all found a nearly identical, massive imbalance in US housing supply and demand. The 2009 American Housing Survey found that 28 percent of houses are attached, 29 percent are detached on small lots, and 43 percent are detached on large lots. The three studies found that only 24 to 25 percent of Americans would prefer to live in large-lot single-family houses (see graph "Housing preference versus supply"). Consequently, there’s an oversupply of approximately 28 million units in the drivable suburbs.

Attached housing and small-lot housing, on the other hand, are undersupplied — by about 12 million and 13.5 million units, respectively.

This imbalance is likely to grow in the years to come, reports Nelson. The generation that is currently moving into the housing market — Millennials — is the most urban-oriented cohort since at least before World War II. A whopping 88 percent of this generation want to live in an urban setting, according to a survey by RCLCo. Factor in rising gas prices — which make drivable suburbs more expensive — rising suburban poverty and the difficulty of selling houses on once-exclusive cul-de-sacs, and suburban appeal erodes further.

Transit will be key

Nelson believes that much of the new housing supply will be provided in the form of transit-oriented development (TOD), which is highly appealing now. Only about 12 million housing units are within an easy walk to a rail station, Nelson says. In 2004 NAR found that 47 percent of Americans want transit access. Furthermore, TOD generally provides attached and/or small-lot housing and walking access to grocery stores and other destinations — all desired by a growing majority of Americans.

The demand for TOD will grow throughout the next three decades to about 68 million units, Nelson calculates (see graph "TOD supply and demand"). He forecasts that 40 million TOD units will be added in the next three decades — well over a million a year — and even after that, the US will still have a shortage of about 16 million units relative to demand.

It’s worth noting that TOD units can be provided in two ways. They can be built from scratch around transit stations, or the transit can come to already existing housing units. Washington, DC, has started to build a 37-mile-long streetcar system that would place 50 percent of residences within walking distance of rail transit — up from 16 percent now.

A good deal of TOD will be provided in cities — both through new units and new transit systems — but even more of it will be developed in suburbs, Nelson believes. Why? Similar to Willie Sutton’s reason for robbing banks (“That’s where the money is”), “The development will take place in the suburbs because that’s where the land is — but it won’t be sprawl,” Nelson says.

Land in suburbs is often spread out, underutilized, and ripe for increasing intensity. Close-in suburbs are near to jobs and therefore are good candidates for TOD. Major TODs have moved forward in many suburban locations recently. Examples include Wyandanch, on Long Island, NY; Wesmont Station, in Wood-Ridge, NJ; Ho’opili near Honolulu, Hawaii; and New Carrollton, Maryland.

A focus on TOD and walkable neighborhoods would have positive effects — among them, improved household finances. The average American family spends 32 percent of its income on housing and 19 percent on transportation, leaving 49 percent for all other expenditures. Those who live in auto-dependent suburbs spend 25 percent of their income on transportation, leaving only 43 percent for all other expenses. Those who live in transit-rich neighborhoods spend only 9 percent on transportation, Nelson shows (see graph "Location efficiency"), leaving far more money for discretionary expenses.

A side benefit of the shift toward rental housing will be increased mobility of the American workforce, Nelson says. Incomes could rise for households that shift from owner-occupied, suburban houses to flexible, urban housing where they can pick up and move if better opportunities arise.

Can builders and governments adapt?

A big question is how the major builders — Pulte, KB Home, Lennar, Toll Brothers, DR Horton among them — will adapt to the abrupt and significant changes in the coming decade. Most are either losing money or barely profitable as conditions now stand.

For the housing industry to meet the nation’s needs in the next 10 years, we are depending on builders that have two generations of experience mostly of providing the opposite of what the market is now calling for. I think the national builders could adapt — or other developers would take their place — if public officials establish an infrastructure and regulatory framework that promotes walkable, transit-friendly development.

But that indicates the bigger problem: We have zoning codes that mostly prevent the needed response from happening. And we have transportation systems that are geared more to the 1950s than to current needs. The result: We now depend on the nimbleness and vision of public officials — in a political climate characterized by short-term thinking and gridlock.

This sounds to me like an irresistible force meeting an immovable object.

Editor's note: This article was revised to include all three sponsors of the Cambridge symposium.