Essential infrastructure, such as the aqueduct that brings water to Los Angeles, unlocks economic potential.
Mark Hanauer/CorbisEssential infrastructure, such as the aqueduct that brings water to Los Angeles, unlocks economic potential.

Scholarly journals have published hundreds of articles about urban economies. The questions are always the same: Why do some cities grow faster than others? Why do some generate more wealth? Why do some decline? No simple answers exist, and much remains open to speculation. However, the accumulated wisdom of more than 50 years of research does allow us to state certain principles about the economies of cities. I propose five.

Before getting to them, a word on the nature of cities is in order. Cities are first and foremost places—agglomerations of people—rather than economic and political units. That fact complicates the study of urban economies. For starters, delineating urban areas can be done in a variety of ways. Exactly where does New York begin, and where does it end? We might opt to study the entire New York metropolitan area, figuring (correctly) that such a definition is more economically significant than merely the area of New York City proper. But even then, we have to remember that a metropolitan area’s borders can expand or shrink over time. The boundaries of the New York metro area aren’t the same today as they were 50 years ago, so one can easily draw mistaken conclusions from the statement that it has grown from 12 million people in 1960 to 20 million today.

Also, cities’ power to make economic policy is limited. (City-states like Singapore are an exception.) The policies that most significantly affect urban economies usually come from higher levels of government. This doesn’t mean that local policies don’t matter, but it does mean that their ability to affect broad economic and geographic trends is sharply circumscribed. It also explains why the scholarly literature focuses on determinants that urban policy can do little about, such as location and “agglomeration economies” (essentially, the beneficial effects of urban size and diversity).

Finally, that cities aren’t economic and political units in the way countries or even states are means that they face particularly fierce competition for mobile resources, especially for talent and brains. After all, it’s much easier to move your residence or your business to a nearby city than to move it to another country or another state. This reality is of fundamental importance in the knowledge economy, whose primary scarce resources are brains, skills, and entrepreneurial spirit. Much of the scholarly literature about cities therefore focuses on “human capital”; for example, the pioneering work of Harvard’s Edward Glaeser, a City Journal contributing editor, has shown that cities initially endowed with highly educated and skilled populations will be more successful down the road. Yet I have consciously chosen not to discuss human capital, talent, skills, the creative class, or whatever term one wishes to use. This is partly because we aren’t at all sure how to attract the young, educated, and ambitious (see “Urban-Development Legends,” Autumn 2011). Further, the presence of a skilled and educated population is as much a result as a cause of success. Talent will flock to successful cities and flee unsuccessful ones. And both success and failure usually have deep historical roots.

Those roots are the subject of my first principle of urban economics: cities’ size and location are key determinants of wealth. One of the more surprising findings of the research on city economies is that urban hierarchies are remarkably stable over time. Only rarely do top cities find themselves dislodged by newcomers. The first city to arrive at the top acquires a fixed advantage, which translates into higher levels of wealth. In every European nation, the biggest city a century ago remains the biggest one today. The population rankings of cities are also surprisingly stable. Lyon has long been France’s second-largest city. Today, it is one-seventh the size of Paris, a proportion that has barely changed over the last 200 years.

The advantages of size and location are the outcome of decades, even centuries, of investments in infrastructure and in institutions. Once in place, these accumulated investments define a good location and cannot easily be undone. Just look at a map of the rail and highway systems of Britain or France: hub-and-spoke networks centered on the largest cities, reinforcing their initial location advantages. Each new city that is connected to London or to Paris further increases the capitals’ market potential.

The United States and other New World nations are somewhat different, in that their settlement patterns remain in flux and their populations are historically more mobile. But even in the U.S., the relative standing of cities has remained surprisingly stable over time, once we put aside the country’s long westward migration. The three largest metropolitan areas east of the Mississippi in 1900—New York, Chicago, and Philadelphia—are still the three largest today, and in the same order, despite the draw of the Sunbelt since the 1960s. The population of the Chicago metropolitan area in 1900 was just under half of New York’s, a proportion that has barely changed since.

My first principle, then, teaches us that the possibilities of dramatically altering a city’s economic standing are necessarily constrained by the city’s location and relative size. Cities aren’t like nations, which can leap from rags to riches within a generation, as South Korea has done. A city’s initial size and location will largely determine which classes of economic activity are likely to succeed there and which are likely to fail. Philadelphia, lying in New York’s shadow, will probably never be a global financial or entertainment center, though it has been highly successful in other sectors, such as health. The constraints of size and location weigh even more heavily on smaller metropolitan areas, such as Syracuse, New York, and Scranton, Pennsylvania.

My first principle of urban economics doesn’t mean that every city’s fate is preordained. That brings us to my second: when cities do experience dramatic changes in their growth paths, the reason is almost always outside events or technological change. European postwar borders are an example of the way political conditions can shape growth. After the Iron Curtain was drawn in 1947, cities in West Germany had access to the growing European Economic Community, while cities in East Germany didn’t. One of the East German cities was Leipzig, which was Germany’s fourth-largest city before World War II but has fallen to 13th place today.

Technological change, too, can shift cities’ growth paths. For instance, as the technology of steel production grew less reliant on economies of scale, and as other metals and alloys entered the market and reduced the demand for steel, growth declined in steel towns, from Pittsburgh to Essen (in Germany) and Birmingham (in the United Kingdom). New transportation technology or infrastructure is an especially powerful agent of change, since it can alter a city’s location advantage, turning a good location into a bad one or vice versa. The emblematic example is the construction of the Erie Canal in the 1840s, which gave New York City access to western markets and solidified its position as America’s biggest city. Similarly, Buffalo’s stagnation since the 1950s can be traced in part to the construction of the Saint Lawrence Seaway, which allowed western goods headed east to bypass Buffalo. And the arrival of air travel has meant that the absence of a port is no longer a handicap for aspiring corporate and financial centers, such as Denver and Atlanta.

In the United States, the most momentous recent shift in urban fortunes has been the rise of Sunbelt cities. We owe that rise partly to new technologies. Miami would still be a fever-ridden swamp if not for drugs and improvements in sanitary conditions that eradicated malaria and yellow fever. Blistering-hot Phoenix, Las Vegas, and Houston probably wouldn’t have attained their current size without air conditioning. The Sunbelt shift also depended on changing demographics—namely, greater life spans, which produced a growing population of retirees in search of warmer temperatures. Technology facilitated that development as well, with air travel enabling the retirees to migrate back and forth easily to keep in touch with friends and relatives back home.

The essential lesson to draw from this second principle of city economies is that no location advantage is eternal, no matter how seemingly indestructible. A new technology can undermine a city’s economy overnight. Two parallel lessons follow. First, the more highly specialized an urban economy is, the more vulnerable it is, no matter how hip or high-tech the city’s star industry (see “Wall Street Isn’t Enough,” Spring 2012). Second, changes in transportation generally aren’t geographically neutral, since they imbue some cities with new location advantages and undercut those of others.

Also related to transportation is my third principle: accessible, well-connected cities exhibit higher growth. Studies of European and North American cities have repeatedly shown the relation between accessibility and growth, most commonly defining accessibility by measuring the number of destinations (often weighted by income) that one can readily reach from a given city, taking transportation costs and time into account.

Market access and connectivity may be even more crucial than human capital. The city that succeeds in positioning itself as the meeting place and market center for a wider region has won a tremendously important battle, since transportation and travel hubs have historically emerged as dominant finance and business centers, attracting talent, money, and brains. Chicago became the market center of the Midwest thanks partly to the city’s canal links with the Mississippi River system, which were promoted by the local business community. The construction of the Erie Canal, again an excellent example, shows that pivotal infrastructure need not be located within the city itself: the link that established Gotham’s access to the continent’s interior was constructed between the headwaters of the Hudson River, several hundred miles upriver from the city, and Lake Erie.

Twenty-first-century wars for connectivity are waged with highways, high-speed rail, and especially airports, which have become accelerators of the conference- and seminar-crazy knowledge economy. Few self-respecting, globally connected executives would seriously consider a job offer in a city lacking daily flights to New York and London. The rise of Atlanta owes much to Hartsfield-Jackson Atlanta International, now the world’s busiest airport. Or consider Frankfurt. Though considerably smaller than Munich, Hamburg, and Berlin, Frankfurt is where Lufthansa, Europe’s largest airline, chose to locate its hub and base of operations. Frankfurt’s arrival as continental Europe’s leading financial center is no coincidence.

Some cities start with a natural connectivity advantage, such as a central location or proximity to major markets. Monterrey, Mexico’s rising industrial metropolis, is near the Texas border on the principal highway linking Mexico City to core markets in the United States. In China, growth is concentrated in the two principal points of contact with world markets: Shanghai and the Guangzhou / Hong Kong area. In Europe, proximity to the continent’s economic heartland—a boomerang-shaped corridor, dubbed the Blue Banana by geographers, stretching from London to Milan—is a clear asset. That’s why Italy’s wealthiest cities are in the north and also why Barcelona and Bilbao, the Spanish cities most accessible to the Blue Banana, are growing industrial centers. Here again, transportation infrastructure has boosted the growth potential of well-connected cities; think of the Chunnel linking London and Paris.

In the United States, the growth-inducing effects of accessibility have produced urban economic corridors, such as the 60-million-strong northeastern megalopolis stretching from Boston to Washington, D.C. The megalopolis has continued to expand south along Interstates I-95 and I-85, and one might say that it now includes a string of cities as far south as Atlanta, including the cities of North Carolina’s Research Triangle.

A useful qualification to the rule comes from France, where research on high-speed trains suggests that it has identifiable growth effects. Whether the trains accelerate overall growth is open to debate, but they can clearly steer growth toward the cities located on high-speed rail lines. The problem is that there can’t be too many stops on such a rail line, since that would slow down the trains and defeat the whole purpose of high-speed rail. So high-speed rail differs from roads in at least one significant respect: it may actually reduce the accessibility of cities left out of the bonanza.

Principle Number Three teaches us that city fathers must be attentive to the opportunities of new transportation links. It also warns us that a loss in accessibility can seriously harm growth. Even a new airport doesn’t automatically improve accessibility, as my own city, Montreal, learned to its sorrow. Back in 1974, the Canadian government, with the enthusiastic support of Montreal’s business community, endowed the city with a shiny new airport named Mirabel (after a nearby village). Because Montreal’s market wasn’t big enough to support two full-service airports, Mirabel would receive intercontinental flights, while Dorval, the older airport, would continue to serve North America. But separating the two streams extinguished the hub function that an airport should serve. No Londoner flying to Cleveland would want to change planes at Montreal, since it would mean driving from Mirabel to Dorval.

So passenger traffic through Montreal plummeted during the 1970s. Not coincidentally, the same period saw a transfer of financial institutions and head offices to Toronto, which (like similarly sized Atlanta) understood that maximizing accessibility in the airplane age required a single, efficient, full-service airport. Toronto soon emerged not only as Canada’s principal air hub but also as the country’s financial center. Mirabel has since been shuttered, an embarrassing and costly white elephant, and all flights now arrive at Dorval. But the damage is done.

My fourth principle of urban economics is that every industry leaves its imprint on a city—and it isn’t always a good one. In North America and Europe these days, the best illustration of this principle is that cities with a legacy of heavy industry and large assembly plants generally exhibit slower growth. The first cities to industrialize, not long ago models of economic progress, are often among the most troubled today. Many have found it tough to move to the knowledge economy.

Their story tells us much about the dangers of industry specialization, or “clusters,” to use the currently fashionable term. When a single industry comes to dominate the local economy, the long-term results can be devastating. The obvious example in the United States is Detroit, the Silicon Valley of the early twentieth century, which was one of the country’s fastest-growing cities until the 1970s but now seems stuck in irreversible decline. Detroit is by no means unique: the great industrial cities of the Midwest, the English Midlands, and the German Ruhr have all registered below-average growth over the last few decades.

Common to these once-great industrial cities is the presence of large plants or other large-scale operations, such as railheads and dockyards. Such installations, while they remain operational, typically pay comparatively high wages for low-skill jobs. Large plants with high sunk costs also give companies a disincentive to move elsewhere, resulting in seemingly secure employment for workers. The outcome is a high-cost business environment that may benefit workers for a time but isn’t conducive to the start-ups that an urban economy needs over the longer haul. If you can get good wages with little schooling, why go to college? If your job is secure at the local plant, why start a business? And if you do decide to start that business, why not move to the city next door, where labor costs are lower?

An even deeper problem is that a city’s ingrained mind-set is hard to undo. The residents of a city with big, unionized factories will naturally come to expect good wages and job security, and their expectations will endure long after the last plant has closed its doors. I know of no example of a painless transition from heavy industry to the knowledge economy—and unfortunately, no standard-issue tool kit exists for helping cities make that transition. The tools may include worker retraining, counseling, small-business support, school reform, downtown revitalization, and industrial-land decontamination, but often the most important step is grudgingly accepting the fact that local wages (and local costs in general) must fall if the community is to regain its competitive edge.

That grudging acceptance has a corollary: city fathers must avoid the temptation to keep plants alive with public funds. Such advice is easier to give than to implement. Faced with sudden job losses, few elected officials will have the courage to do nothing. In an ideal world, every owner would give enough notice before closing a factory, or else downsize slowly enough, to enable the community to absorb the shock painlessly. That isn’t the way of markets, but at the very least, political leaders should comprehend that many established industries will eventually close and start pointing their cities in a different direction. This, too, is easier said than done. The turnaround seldom begins until the shock of closures drives home the need for change.

My fifth principle of urban economics: though much remains unexplained, good and bad policies do matter. Despite the best efforts of scholars, econometric models rarely succeed in explaining more than half of cities’ variations in growth over time. Many factors are impossible to quantify, such as the ability of a dynamic individual, such as a mayor or an entrepreneur, to make a difference. Local business and political culture surely plays a part in a city’s growth, but we know little about how such cultures form.

In the end, we probably understand which policies cause failure better than we understand which cause success. Poorly governed cities with a reputation for corruption, violence, or deficient institutions will pay a price. Recall the openness of city economies: people and firms can leave at will. In open societies, good governance is not only a matter of virtue but also a competitive necessity. New Orleans’s decline began long before Hurricane Katrina; its reputation for endemic corruption, cronyism, and parochial politics was part of the reason. Another was Louisiana’s long and unfortunate tradition of underinvestment in education, which resulted in one of the least educated populations in the United States. In France, the city of Marseille has an equally poor reputation for governance and has also consistently exhibited below-average growth. Both cities enjoy warm locations, but that advantage hasn’t been enough to keep them growing quickly.

At the other end of the spectrum, not only climatically, is the Minneapolis–St. Paul area, which has done well, despite chilly temperatures and a peripheral location. The area has a reputation for being well governed, and Minnesota boasts a tradition of investment in higher education. The Scandinavian roots of its population are undoubtedly a factor.

Crafting good urban policy isn’t limited to local government; higher levels of government also have to get it right. Montreal’s airport blunder was largely of federal doing. And policymakers should always remember that a single project, no matter how grand, will seldom be enough to turn an ailing city economy around. Two years after the airport mistake, Montreal embarked on the construction of a colossal stadium for the 1976 Olympics, this time with mostly municipal money. The project almost bankrupted the city and did nothing to bolster the long-term growth of the local economy.

But let me end on a positive note. Many policy initiatives have succeeded in sparking growth. Los Angeles’s phenomenal progress after World War I owed much to the foresight of its business community, which—defying nature and geography—built an artificial harbor for the city, along with vast systems of aqueducts to bring water from the distant Sierra Nevada. We can trace Houston’s similarly meteoric growth since the 1950s not only to nearby oil reserves (a natural advantage) and to the invention of air conditioning (a technological change) but also to such wise policies as the construction of navigable channels, as well as a strong municipal government. True, sustained growth is rarely the outcome of one action, and there is no easy road to success. But policymakers can nevertheless help their cities thrive.


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next