Productivity, Housing, Wages, and “The Cost of Living”

Posted on


One very typical line of dialogue in talking about politics goes like this:

LITERALIST: We need to raise taxes on rich people, the $100k and over crowd!
SUBJECTIVIST: $100k may be a lot of money in Kansas City, but try raising a family in New York or San Francisco on that and you sure won’t feel very rich.
LITERALIST: Look, one hundred thousand dollars is double the median household income in the United States of America. If you’re making twice as much money as the average household, you can’t go whining about hardship.
SUBJECTIVIST: Feelings, nothing more than feelings.
LITERALIST: After all, the high cost of houses in high-cost areas reflects the fact that an apartment in Manhattan is a luxury good.

Brad DeLong has a couple of recent posts that play out some of this dynamic. But after reading some Ed Glaeser more recently, I think this whole line of inquiry is mistaken.

The issue is that high-density areas are more productive:

Since productivity is higher in denser areas, wages are also higher. So it’s not at all clear that the guy earning $100k a year in New York could, in fact, take that salary to Kansas City with him. Meanwhile, in principle the price of housing ought to approximately equal the price of housing construction. The cost per square foot of building a 25-story apartment building is somewhat higher than the cost per square foot of building a single-family home, so the higher wages in New York ought to be partially offset by higher construction costs. In practice, however, the price of housing in New York is only very loosely related to construction costs. That’s because there are large non-trivial* regulatory barriers to construction. On the flipside, during the urban disinvestment period of the 1970s and early 1980s it was possible to purchase real estate in many parts of New York City** for less than the cost of construction. And some households are beneficiaries of rent control legislation and the like.

The upshot is that on average the higher wages in New York should, in fact, be more or less offset by congestion costs and the higher cost of housing.

But that’s only an average. In practice, the distributional consequences of supply constraints vary wildly. A person renting in New York City and earning $100,000 a year may in fact be “really” not that rich, despite his high income relative to the national median. The flipside of this is that he’s probably much less good at his job than he believes, and is earning a substantial wage premium based on his willingness to bear the high housing costs. But someone who bought an apartment in New York in 1982 and now earns $100,000 a year is actually much richer than his income would suggest. People like this (i.e., people like a lot of my parents’ peers) are really a kind of urban rentier class whose wage income may be pretty small relative to the financial value of the regulatory privileges they’re receiving.

And the real point to be made about this isn’t so much about how much taxes anyone should be paying as it is about how massively inefficient it is. If the restrictions on density were relaxed, we’d have more people living in high-wage, high-productivity areas and those places would be denser (meaning more productivity and higher wages) and housing costs would be lower relative to wages, meaning more disposable income. The point to recall is that the high per square foot cost of housing in New York and San Francisco isn’t really the cost of “living” it’s the cost of regulatory privilege.

(I thank Ryan Avent’s discussion of WMATA for suggesting this analysis)

* By which I mean it goes well beyond conventional rules like the ban on child labor, basic worker safety, etc. that apply everywhere.

** And other American cities, with the precise time period varying from place to place.