Why America’s cities are leaving the rest of the country behind

By Jonathan Gruber and Simon Johnson

America is often called “the land of opportunity,” but in recent decades, the opportunities have been shrinking—in a geographical sense—as we increasingly rely on a small set of superstar cities located on the coasts to drive our innovation economy. Meanwhile, much of the rest of the country has missed out, as people are unable to move to these economic epicenters because of limited high-priced housing. This lack of mobility has rein- forced our nation’s divided politics, specifically the splits across regions and also between large cities and smaller towns or rural areas.

America does not have to rely on a small set of superstar cities for growth. There are opportunities for rapid growth around the country that are just waiting to be tapped. Cities outside the superstar centers on the coasts have excellent educational institutions, talented populations, and a high quality of life. They are ready to take their place as growth engines driving the technology-based economy of the future.

Unfortunately, the path to widespread economic opportunity is blocked by the allocation of the public and private infrastructure of research and development.The continued targeting of both early-stage investment dollars from the private sector and of basic research from the federal government to the superstar cities simply reinforce the on- going regional inequity. States are trying to resist these trends through the kinds of tax breaks that Amazon was offered in its recent competition to determine the location of its second headquarters, but that just leads to a race to the bottom, where corporations win and US taxpayers lose.

There is a better way. The federal government has a long history of undertaking place-based policies that target areas around the nation, beginning with support provided to various locations as the frontier moved westward during the nineteenth century. Other contributions include the land-grant colleges that created the modern US educational system, the Tennessee Valley Authority that modernized a large area in the Southern United States, and the distribution of military bases around the nation, which helped share the benefits of the postwar boom.

It’s time to update this thinking and apply it to an expansion of research and development that will keep the United States at the forefront of the modern world economy.

Superstar Cities Pulling Away

While overall rates of growth have slowed in the United States, a set of superstar cities is thriving—and pulling away from the rest of the country.

To see this, we can look at the numbers on average earnings per worker across metropolitan statistical areas (MSAs) in the 48 contiguous US states in 1980 and 2016. These MSAs pick up what is happening in both cities and in their surrounding suburban areas. Using public survey data, we have created a list for both 1980 and 2016 of superstar cities, which we define as the top ten mainland urban areas with highest earnings per worker.

In 1980, five of the top ten MSAs were in Michigan, and a sixth was Casper, Wyoming. These were MSAs whose income derived primarily from manufacturing and natural resource extraction. By 2016, the top ten included Boston, New York, San Francisco, San Jose, and Seattle, which were outside the top ten in 1980; Boston and New York weren’t even in the top twenty in 1980. The good jobs today are in information technology, biotechnology, and financial services. None of those previously star Michigan areas make the top ten list today—indeed, they don’t even make the top twenty.

There is also an evident shift in prosperity toward the coasts. In 1980, three of the top ten MSAs were on the East or West Coast; by 2016, nine of the top ten were on the coasts.

In addition, there is a widening disparity of incomes between the most prosperous areas and the rest. In 1980, the top ten cities had 30 percent higher earnings per worker than cities elsewhere in the country. By 2016, the top ten MSAs had earnings that were 57 percent higher than elsewhere.

Disparities are also opening up between the very highest-earning cities and those merely doing very well. Average earnings in the top three cities in 1980 were 8 percent higher than average earnings in the rest of the top ten in that year. Average earnings in the top three cities in 2016 were 25 percent higher than the rest of the top ten in that year.

It is the shift to a knowledge-based economy that appears to be driving income divergence across geographic areas. This process creates greater agglomeration, meaning the crowding together of similar or related activities. For example, there are over two thousand tech companies in Silicon Valley, which is the densest concentration in the world. The Economist identified ninety-nine listed technology companies with market values of over $1 billion in Silicon Valley, which are together worth $2.8 trillion and account for around 6 percent of all corporate America’s profits. As another example, the Boston/Cambridge area is home to about one thousand biotech-related businesses, and Kendall Square has the highest concentration of biotech companies in the world.

Agglomeration in the technology economy arises when more talented workers in an area raise the economic returns to other talented workers who are located nearby. For example, recent research shows that inventors are more productive when there is a higher local density of other inventors in their area. Thereby, places that are able to initially attract talent are able to reward additional talented workers at a higher level, leading to a concentration of talent in certain places—while others fall behind.

Agglomeration arises from a number of economic forces. Firms and individuals can share resources ranging from better public transportation (airports and roads) to higher-quality schools to sports arenas. Having more people living in an area with similar skills makes it easier to find workers whose skills match what firms need—and can also help workers find the kind of opportunity they want (in the job market or even in the dating market). Finally, more interactions with a wider set of skilled individuals means more opportunity for learning.

Surprisingly, the concentration of economic activity in superstar cities doesn’t seem to be slowing down. The high cost of living in successful cities, along with the reduction in communication barriers brought about by the internet, should—you might think—lead individuals to move out of those cities and into new locations with lower costs of liv- ing. Bogged down by their high costs, existing technology hubs might give way to new growth centers.

In fact, the opposite is happening. The success of technology companies depends critically on the entire economic ecosystem around the company so that it is hard to break away from an existing hub. The geographic divergence we have seen recently in the United States reflects the fact that some areas have succeeded in creating ecosystems that foster growth, attracting more skilled workers, and continuing to grow faster.

Well-performing areas look set to do even better. The cities that have the most college graduates are the ones adding college graduates the fastest. The cities with the highest-earning college graduates are the ones where college graduate earnings have been growing the fastest. The cities with high life expectancy are the cities where life expectancy is growing the fastest. Every day, the American landscape splits further and further between the haves and have-nots.

Recall that this trend has been reinforced in an important way by the government; the locations that have been favored by public research spending are by and large the same superstar areas of the country. Seven of the ten highest-earning MSAs are in states that are in the top ten in terms of per capita public R&D. None of the top ten—or even the top twenty—highest-paying MSAs are in states that are in the bottom half of spending on public R&D per capita.

The Rent Is Too Damn High

If some cities are doing so much better than others, then why don’t more people move there? The United States is a nation with historically high mobility and a history of strong-willed pioneers striking out for new areas in search of riches. There is no secret regarding where the best opportunities are today. Why isn’t everyone in their covered wagon heading to those cities?

After all, US history is replete with boomtowns. In 1850, there were thirty thousand people in Chicago; by 1910, there were more than two million residents. However, today, economic booms no longer necessarily lead to burgeoning urban populations. The metro areas offering the highest pay in 2000, such as San Francisco, Seattle, and Boston, have some of the slowest population growth rates in the country, while individuals have flocked to lower pay and less productive metropolitan areas. New York, San Francisco, and San Jose are today smaller, relative to other cities, than they used to be.

The problem with living in booming cities such as San Francisco is quite obvious—they have the highest cost of living. The list of cities where the cost of living is the highest reads very much like the list of cities that have most benefited from the knowledge economy: San Jose, San Francisco, Boston, New York, Washington, DC, Los Angeles, and so on. This high cost of living forms a barrier deterring entry into the city and into its workforce.

The high cost of living in these places is not mostly about the cost of food or other services. A quarter pounder with cheese costs only twice as much in Boston, Massachusetts, as it does in the lowest-wage parts of the country. Rather, the most significant difference is in the cost of housing. Housing expenditures are the single largest element of family budgets, accounting for 40 percent of spending on average. The cost of housing varies a great deal across cities in the United States.

Indeed, the same dispersion that we noted earlier in wages is also present in what folks have to pay for a house. In 1980, families living in the ten highest-earning cities paid $188,880 on average for their homes; in all other cities, house prices averaged $151,050. By 2016, those in the top ten highest-earning cities paid $607,530 for their homes, compared to $222,020 nationally. That is, housing prices in the top cities in 1980 were 25 percent higher than other cities in 1980; by 2016, they were more almost three times as high.

It might not seem surprising that housing is more expensive when everyone wants to move to a city. This simply reflects the basic law of supply and demand. However, housing is much more expensive than it has to be in prosperous cities because of a mechanism interfering with the unfettered operation of the housing market: zoning regulations. Throughout the United States, local regulations restrict the use of land in ways that make it impossible to provide the supply of housing that is required to meet its demand. This means that as areas grow, more and more workers are bidding for a limited set of places to live reasonably close to the cities where the good jobs are located.

For much of US history, local economic booms were matched by local building booms, but this era has ended. For example, in Manhattan, there were thirteen thousand new housing units permitted in 1960 alone, which is nearly two-thirds of the total permitted during the entire decade of the 1990s. The cost of constructing housing does vary some- what across areas; for example, it is cheaper to build in the flat southwest than the hilly northeast, but this variation is very small relative to the enormous variation in house prices. Local restrictions on constructing housing are the primary driver of house price variations.

The United States is relatively unique in having land use under local control, as opposed to national land-planning agencies in places like the UK and France. The problem this raises is that existing owners have an incentive to restrict housing supply, both to increase the local amenities (such as by having height restrictions on buildings to make the skyline more appealing) and to keep house prices high.

The set of restrictions imposed varies widely, from minimum lot sizes (most Boston suburbs have a minimum lot size of over one acre for the majority of their homes), to maximum heights, to multistage and costly review processes for environmental and other restrictions. While the set of restrictions is complicated, the net effect is to significantly raise home prices and rents. Researchers estimate that in one-sixth of the metropolitan areas in the United States (including most of the highest-earning ones discussed earlier), the prices are at least 25 percent higher than construction costs, and in markets such as Los Angeles and San Francisco, they are twice as high.

Consider in particular the example of Palo Alto, California, the epicenter of Silicon Valley. The technology industry in and around the city has grown spectacularly, but its housing stock has not. Low-slung, single-family housing dominates, even with thousands of students and young employees who might prefer smaller apartments. According to Zillow, single-family homes trade at a median price of $2.6 million, and median rents have climbed from $3,800 in 2011 to $6,000 in 2017.23 Despite this, Palo Alto leadership has focused on restraining the growth rate of jobs rather than building more housing. In 2017, the city planning commission approved an extension of the city’s recent restriction of new office development to no more than fifty thousand square feet per year in an effort to slow development.

Another good example is the suburbs around Boston. The Route 128 corridor that circumscribes Boston’s core has long been the site of a cluster of prominent companies, initially focused on computer services and more recently on financial services, management consulting, and biotechnology. In response to the growing industrial cluster, cities and towns in this area aggressively used zoning to maintain the predominance of single-family housing. According to the Fair Housing Center of Greater Boston, the type of development that occurred around Route 128 served to intensify racial segregation in the metropolitan area by excluding lower-income people and people of color from living near good job opportunities. Specifically, zoning tactics like setting minimum lot sizes served “to control density, protect open space and artificially inflate housing prices.” Today, towns around Route 128 are among the most expensive in Massachusetts.

Income now flows to a smaller set of cities, but individuals can’t afford to move to those cities to take advantage of the opportunities—and therefore get left behind in the places that are not benefiting from the agglomeration economies. Indeed, one recent study estimates that restrictive housing policies cause millions of workers to be “missing” from the most productive cities in the economy. If barriers to housing supply were removed, this study estimates that many more workers could live in these more productive cities; if this reallocation had occurred over the past several decades, the United States as a nation could have grown 50 percent faster in that time.

Moreover, mobility, like everything else in America, has become polarized. Almost half of college graduates move out of their birth states by age thirty, while only 17 percent of high school dropouts do so. This at least partly reflects the fact that higher education leads to higher wages, which leads to a better ability to afford to move to thriving cities, which means even faster wage growth in the future. And as a result, we get a continued divergence between the most and least educated in our society.

At the same time, people living in and around the highest-income cities are spending much longer commuting. In the top ten cities of 1980, 5.4 percent of workers had to spend an hour or more commuting, compared to 3.8 percent nationally. In 2016, 14 percent of commuters in the top ten cities spent more than an hour commuting, while nationally it was less than 6 percent. Meanwhile, in San Francisco, a proposal to build high-density housing close to mass transit is meeting opposition from, among others, environmentalists—who would in other forums likely argue for lower carbon emissions.

Baby Einsteins Without Access To Capital

One contributor to this ongoing dispersion in economic opportunities is the method of financing innovation. Despite fluid national and international capital markets, VCs like to invest where they are located. This follows naturally from their concern about moral hazard—early-stage investors are more eager to invest when they know the entrepreneurs and can monitor them closely.

Once again, this is a profit-maximizing strategy for investors, but it has the consequence that the early success of VCs in existing superstar cities has led to a concentration of start-up capital in those cities. And this concentration is striking. Twenty-five percent of all VC financing in the United States is concentrated in the San Francisco area, while another 15 percent is in nearby San Jose. Another 10 percent is in the New York area, 10 percent is in Boston, and 5 percent is in LA. This means that two-thirds of all VC financing is focused in five places in our entire nation—and these are the places that are already the epicenters of the knowledge economy.

The result is that new discovery is not being financed in those parts of the country that do not already have successful investors, and this can have long-run implications for polarization.

A vivid illustration of this point is recent research that contrasts the place of birth of individuals who do and do not eventually create patented technologies in the United States. Future inventors grew up in exactly the places where existing inventors are concentrated. In 1980, the rate of patenting in the top twenty cities was about 2.5 times as high as other large cities in the United States. By 2010, it was 6 times as high—that is, the highest-earning cities are also creating new knowledge at a faster rate than other cities in the United States.

This is particularly disturbing because it suggests a strong mechanism for perpetuating the polarized nature of the US labor market. There are undoubtedly many “future Einsteins” born in parts of the country where inventing is not happening—but productive opportunities are lost when there is no mechanism for the discovery of ideas from these parts of the country, nor the financing for the development of these ideas into valuable goods that can create jobs and grow the economy.

Jonathan Gruber and Simon Johnson are the authors of “Jump-Starting America: How Breakthrough Science Can Revive Economic Growth and the American Dream,” from which this article is excerpted.