The Superstar Metro Argument... And Why It Fails

The Yes In My Back Yard (YIMBY) movement argues for less restrictive zoning based on the idea that by limiting the amount of housing construction, zoning reduces housing supply and thus makes American cities and suburbs more expensive. Anti-YIMBYs generally respond that new housing will not lower rents and/or that new housing creates negative externalities for its neighbors. 

One of the more scholarly defenses of the status quo is what I call the “superstar monopoly” argument. According to this theory, making housing cheaper in expensive cities will have negative side effects, because as Americans crowd into those cities, the rest of the U.S. will decline and be left behind. For example, one article by a group of legal scholars in Duke Law Journal points out that “deregulation of land-use controls in New York and California would shift people, jobs, and economic growth to superstar cities in those states and away from the Midwest, South, and Southwest” which, in turn, would mean that people who cannot or will not leave the latter regions would be stuck in “left-behind places.”

This argument seems to be based on a not-always-stated assumption: expensive superstar metros dominate the U.S. economy, and if we don’t price people out of those cities, they will dominate the economy even more, which means that the rest of the country will turn into one giant low-income, rotted-out city where only poor people live. To put it another way, housing poverty and homelessness in New York is just the price we have to pay for the alleged vibrancy of St. Louis or Gary. 

One obvious problem with this argument is that making big cities expensive hasn’t really helped “left-behind” cities: for example, St. Louis has lost over two thirds of its 1950 population, and Gary has lost almost as much. 

Why haven’t high New York rents saved cities like St. Louis? Because the “superstar monopoly” theory is based on a false dichotomy: Americans either live in big expensive superstar metros (which I will call “BSMs”) or in Rust Belt cities and decaying rural areas. The reason the dichotomy is false is that Americans have other options: they can live in stable suburbs of stagnant cities (such as Detroit’s wealthy northern suburbs) or in booming Sun Belt metros like Houston and Atlanta.

If the “superstar monopoly” theory reflected current reality, BSMs would be growing faster than those Sun Belt metros. The most infamously expensive metros with over 4 million people (New York , Los Angeles, Boston, San Diego and San Francisco) have experienced population growth ranging from 3 percent (New York) to 16 percent (San Diego) since 2000. By contrast, the four least expensive Sun Belt metros with over 4 million people (Dallas, Houston, Atlanta and Phoenix) have experienced growth ranging from 47 percent (Atlanta) to 60 percent (Houston). In other words, if New York and San Francisco lost all their jobs tomorrow, Rust Belt residents wouldn’t necessarily be stuck at home or in small metros — they could move to large Sun Belt metros (or as I will call them below, “LSMs”).

Data on job growth is a bit harder to find. However, the Brookings Institution's Metro Monitor includes a dashboard that enables us to compare metro area economic trends over the past decade or so. The dashboard shows that between 2013 and 2023, jobs in the five BSMs increased by about 15-20 percent: that is, the largest increase was 19 percent (San Diego) and the smallest was 15.6 percent (Los Angeles).  How do the four LSMs compare?  The smallest job growth number was Houston (28 percent increase) while jobs in Phoenix and Atlanta grew by over 40 percent. In other words, job growth in the Sun Belt is significantly higher than in the so-called superstar metros.

One obvious retort is: “Yes, people have been moving to the Sun Belt metros, but those are just poor people — all the 'good' jobs are in superstar metros, so those metros really dominate the economy after all.” If economic dominance means anything, it means growth in gross metropolitan product (GMP) — that is, the most dominant metros combine decent wages and lots of population, thus leading to a high overall level of wealth. So if the “superstar monopoly” theory made sense, GMP growth would be higher in the BSMs than in the LSMs. Is this the case? No. In four of the five BSMs, 2013-23 growth was stuck in the 22-29 percent range: the only exception was San Francisco (45 percent). How do the LSMs compare? In three of the four, GMP growth was in the 40-44 percent range: the only exception was Houston (22.6 percent growth). In other words, the general trend is that Sun Belt metros take up a larger share of the American economy than they did in 2013, and so-called superstar metros take up a smaller share — with Houston and San Francisco being exceptions.

A less implausible version of the superstar monopoly argument might be: obviously the Sun Belt is outcompeting New York and Boston now, but if the latter cities were cheaper, the Sun Belt would be as stagnant as metro St. Louis or Detroit. It seems to me that this argument is unpersuasive because even when housing was cheaper in the 20th century, Sun Belt metros grew faster than at least some of the BSMs. For example, the population of metro New York grew by about 45 percent (from 12.3 million in 1950 to 17.8 million in 2000) and the population of metro Boston grew by about 60 percent (from 2.5 million to 4 million). But the populations of each of the LSM metros quadrupled and then some: the slowest growth among the four was in Dallas (from 866,000 in 1950 to just over 4 million in 2000).* Having said that, I cannot find GMP data for individual metros, so more research would be useful here. 

In sum: the superstar monopoly argument is based on the assumption that BSMs are taking up an ever-growing share of American people, jobs and GMP. The reality is that less expensive Sun Belt metros are outcompeting these so-called superstars in all three areas.

*All of this data is from macrotrends.net.