When Amazon decided to locate its second headquarters in New York, it cited the supposed advantages of the city’s talent base. Now that progressive politicians have chased Amazon out of town, the tech booster chorus has been working overtime to prove that Gotham, and other big, dense, expensive cities, are destined to become “tech towns” anyway, because of their young, motivated labor pools. That argument may sound great to New York Times readers or on local talk shows, but it is increasingly untrue.

In fact, as a new Brookings study shows, millennials are not moving en masse to metros with dense big cities, but away from them. According to demographer Bill Frey, the 2013 to 2017 American Community Survey shows that New York now suffers the largest net annual out-migration of postcollege millennials (ages 25 to 34) of any metro area — some 38,000 annually — followed by Los Angeles, Chicago, and San Diego. New York’s losses are 75 percent higher than during the previous five-year period.

By contrast, the biggest winner is Houston, a metro area that many planners and urban theorists regard with contempt. The Bayou City gained nearly 15,000 millennials net last year, while other big gainers included Dallas — Fort Worth and Austin, Texas, which gained 12,700 and 9,000, respectively. Last year, according to a Texas Realtors report, a net 22,000 Californians moved to the Lone Star State.

The other top metros for millennials were Phoenix; Nashville, Tenn.; and Charlotte, N.C., as well as four relatively expensive areas: Seattle, Denver, Portland, Ore., and Riverside — San Bernardino, Calif. The top 20 magnets include Midwest locales, such as Minneapolis — St. Paul, Columbus, and Kansas City, all areas where average house prices, adjusted for incomes, are half or less than those in California, and at least one-third less than in New York.

Perhaps even more significant has been the geographic shift within metro areas. The media frequently has exaggerated millennial growth in the urban cores. In reality, nearly 80 percent of millennial population growth since 2010 has been in the suburbs. Even in the Bay Area, the tech industry’s global epicenter, suburban Silicon Valley has continued to grow its STEM base rapidly, while San Francisco recently has seen rapid slowdown in tech jobs. Perhaps density, massive homelessness, and filthy and disorderly streets, not to mention unaffordable living costs, lose their appeal as couples contemplate childbearing.

Dense, high-priced cities attract young people straight from college, but they have trouble keeping them there. Some of the “best and brightest” still migrate to “superstar” cities, but many don’t stay long. The average resident in the downtown areas so popular with postcollege millennials has lived in the same house for about 2.4 years, compared with seven or more years in the suburbs and exurbs.

Urbanists’ faith in the inevitable appeal of major cities to high-end businesses, notably in tech, may be questionable. New York, for example, is not remotely the tech center that outlets like the Times suggest; according to estimates from the economic consultancy ESRI, its share of computer-related jobs barely equals the national average, well below even such unheralded spots as Dallas — Fort Worth or Columbus, Ohio. Tech workers, and the capital to finance them, may come to Manhattan or San Francisco, but the growth of these industries is slower than in places like Austin; Texas; Nashville, Tenn.; or Orlando, Fla.

Like all geographies, core cities must compete for businesses. The prosperity of urban cores demands favorable business climates. Core cities, such as New York, Los Angeles, Chicago, and San Francisco, may be exciting for recent college graduates or tourists, but they still need affordable standards of living and strong economic and educational environments to attract and retain the young families critical to their long-term growth. A middle class cannot be sustained by only elite workers. Politically, a city with a shrinking middle class, as we see in many superstar cities, will exhibit evermore radical politics as the young single population and poor dominate the electorate. The political problems evident in New York have surfaced in other pricey regions. Even in world-class tech center Seattle, a city council-imposed homeless tax on large corporations had to be repealed when Amazon responded by stopping construction on a downtown office tower.

The imperative is not to increase subsidies for favored companies, as New York tried to do with Amazon, but to address the basic conditions — taxes, public safety, schools, housing — that ultimately determine long-term economic growth. No amount of subsidy can make up for these failings. The road to enhanced growth lies in addressing the very issues most urban politicians like to avoid.

Joel Kotkin is the presidential fellow in urban futures at Chapman University and executive director of the Center for Opportunity Urbanism. His latest book is The Human City: Urbanism for the Rest of Us. Wendell Cox is the principal of Demographia, a public-policy consultancy, and a senior fellow at the Center for Opportunity Urbanism. This piece originally appeared in City Journal.