scorecardresearch Skip to main content
Opinion | Mike Ross

Build more in the middle

Triple deckers on Telegraph Street. John Tlumacki/Globe Staff/Globe Staff

For years, Boston’s housing construction has resembled a doughnut: On one side, there’s a tranche of single-family structures. On the other, large (and expensive) multi-unit buildings. But in the middle, there’s nothing except a hole.

Yet — and stick with my baked goods analogy for one more second here — what Boston really needs its housing policy to resemble is the dark fudge portion of a half moon cookie. That is, mid- to large-sized apartment structures without the one-, two-, and three-unit buildings.

Since 2000, Boston’s housing growth has been slow — and slower still than other cities, including Washington, D.C., San Francisco, Seattle, and Portland, Ore.

Part of the reason for the slower rate of growth is what Boston is building. A review of the type of structures that have been built reveals a stubborn persistence to build housing we no longer need — I have complained about this in the past. Our housing stock of one-, two-, and three-unit buildings doesn’t fit the city we’ve become. What is needed are larger multi-unit apartment buildings. Nonetheless, since 2000, Boston’s focus continues to be on building one-family homes.

Other cities have figured this out already. In San Francisco, according to available census figures, only 7 percent of the new housing stock is one-family construction. Boston, however, has devoted over 18 percent of all new homes built to the creation of one-family units.


We’ve built more one-family structures than all five-unit to 19-unit apartment buildings combined. The only category of housing we build more of are apartment buildings that are 50-units or greater. While this is good for housing creation, these new luxury buildings do not typically serve the middle market — and most are snapped up by investors.

In other words, we do a really good job of building big and building small, but we do a poor job of building in the middle.

One reason for the lack of mid-sized apartment buildings is the high cost of land and construction. With the exception of San Francisco, housing construction in Boston is more costly than in nearly all comparably sized cities. It therefore is much easier to spread the cost of construction over more units than fewer. For infill housing (typically, one-, two-, and three-unit buildings), the economics are different. As a result, housing creation avoids the middle.

To create more units, Boston will have to reduce the drag on the system.

One idea is to take a second look at Boston’s inclusionary zoning policy, the requirement where buildings of 10 or more units must include a percentage of affordable housing. Again, for larger buildings, it’s easy to spread the cost of the subsidy across the other units. But as the units decrease, each individual unit’s cost goes up.

Predictably, builders are building less to avoid tripping the requirement. Why build 10 when you can build nine?

There have been 46 buildings (or 414 units) built in nine-unit apartments — where none are required to be affordable — since the policy began in 2000, according to data provided by the Department of Neighborhood Development. The number of 10-unit buildings built in this same period? Sixteen (or 160 units).


Other cities, such as Washington, D.C., and San Francisco, also have inclusionary zoning policies. However those cities have more flexible buyout provisions. Boston should consider loosening the restrictions for buildings with 10 to 20 units. In the end, robust housing creation will contain the cost of housing.

Recently, Brian Golden, the director of the Boston Redevelopment Authority, has been telling people that we are in the biggest building boom in Boston’s history — bigger than when the Back Bay was filled in the late 19th century and bigger than when Boston’s high spine was built in the 1960s and 1970s. He estimates that each year since 2013 there has been more than 13 million square feet built, 80 percent of which is residential.

Indeed, last year’s creation of 4,681 units of housing (with approximately 20 percent affordable) was a record.

Yet that alone won’t fix the supply-demand deficit. It will also require housing policies that facilitate more, not less, housing creation.

Or, said another way, it will require an end to the doughnut and an embrace of the half moon cookie.

Mike Ross writes regularly for the Globe. Follow him on Twitter @mikeforboston.