Why Housing is So Expensive

Figure 7 from “Productivity and Potential Output Before, During, and After the Great Recession” by John Fernald, ultimately published in the 2014 NBER Macroeconomics Annual

Figure 7 from “Productivity and Potential Output Before, During, and After the Great Recession” by John Fernald, ultimately published in the 2014 NBER Macroeconomics Annual

There are two big reasons why housing is so expensive. The first is the obstacles put in the way of building new housing in many of the most desirable cities to live in. My detailed knowledge of these obstacles comes mostly from two Facebook groups: Market Urbanism and Market Urbanism Report. I highly recommend those Facebook groups. The political battle over regulatory obstacles to housing construction has heated up lately. For a long time there have been many “NIMBYs”: people who say “Not In My Back Yard.” But now there are also “YIMBYs” as a political force: people who say “Yes, In My Back Yard.”

The YIMBYs have had some successes lately. For example, California state law now says that local governments must allow garages that are not being used for vehicles to be converted into living space. Oregon state law now bans single-family-housing-only zoning. Minneapolis also now bans single-family-housing-only zoning. And Salim Furth has an article on a variety of pro-housing policies being pursued in different states.

The argument for state and federal involvement in policies affecting the quantity of housing that is built is a strong one because local governments are unlikely to fully factor in the benefits to people not now living in a city (and therefore not voting in that city) of being able to afford to move to that city. State governments are more likely to factor in that benefit for people moving from elsewhere in the state, while the federal government is more likely to factor in that benefit for people moving in from other states. (Note that the federal government does not need to mandate anything: it can simply dangle federal money in front of states that have a pro-residential-construction policy.)

One nice policy that targets total amount of housing without dictating specifics to local governments a policy in which the state establishes a goal for housing construction for each locality. If the local government meets the goal, it retains full control over the detailed zoning and other regulations. But if the local government doesn’t meet the goal, then a state agency can approve construction without any veto from the city. Most local governments would work strenuously toward achieving the goal in order to retain local control.

Having local governments get out of the way of residential construction reduces the regulatory cost of building housing and reduces the land cost of housing when taller buildings are allowed. But it still costs something to build the structure itself. Sadly, as John Fernald’s graph at the top of this post shows, there has been no real improvement in productivity in construction in the last 50 years. (“TFP” stands for “Total Factor Productivity.”) This accords with casual observation: the way houses are built now doesn’t seem that much different than when I was 10 years old in 1970. When one thinks about the dramatic transformation of many other industries in the last 50 years, the lack of transformation in construction is truly remarkable.

In particular, I marvel at how much greater the increase in productivity has been in manufactured goods made in factories than in construction. Is construction inherently so different from manufacturing? The mystery of low productivity growth in construction is one that more economists should try to answer.

I have a hypothesis for where policy went wrong, getting in the way of productivity improvements in construction. We could probably have had dramatic improvements in productivity in construction if we had moved more and more toward building pieces of houses in factories and then hooking those pieces together onsite. (Think Legos.) This would have required nationally standardized construction codes that could vary by soil type, earthquake or flood risk, average precipitation, etc., but would otherwise be the same all across the US. It is not to late to begin such a transformation.

I have heard the argument that housing is different because people want a customized house, but one could make the same argument about cars, and it is false. We manage to make high-end cars in factories that are customized enough for people to be happy with them. Moreover, the lower people’s income, the more willing they are to accept standardization if it means a less expensive house. The tragedy is that there aren’t standardized houses produced in pieces in factories by big companies that are a step above or three steps above a double-wide trailer house.

Bringing down the high cost of housing is something that is within our power. But it will take big changes from current policy.

Other posts on housing:

Links drawn from “Market Urbanism” and “Market Urbanism Report” and other authors: