The most profitable and productive elements of our society are lodged in our cities. From the fantastically dense agglomeration of financiers who make up Wall Street to the cluster of artists and film studios in Hollywood, the concentration of resources in urban settings seems essential to creating world-class centers of commerce and industry. Cities facilitate trade, provide markets for specialized producers, and, perhaps most important, speed the flow of ideas. Because of these advantages, big-city workers earn more than their nonurban counterparts -- 28 percent more, controlling for education, age, race, occupation, and gender. Certainly there are cities in decline, especially those without a well-educated work force or those with too heavy a commitment to manufacturing. But the overall connection between urbanization and economic growth is such an empirical truth that one can hardly find a wealthy, modern country that is not also urbanized.

So it is disturbing to find geographic concentrations of impoverished ethnic groups in the midst of these productive environments. These districts, commonly called "ghettos," function culturally, intellectually, and economically apart from the busy downtown. The distance from Wall Street to the South Bronx, along these dimensions, is greater than that between New York and London or Tokyo. Cities throughout history have contained distinct ethnic districts. But rarely have they been so isolated and impoverished as the African-American districts found in U.S. cities today.

All major immigrant groups coming into the United States established their own residential areas. Irish and Eastern European immigrants in the early twentieth century actually were more segregated than blacks of that era; they lived almost as segregated as blacks do today. These immigrants clustered together in part because they were restricted from living in Yankee areas, but also in part voluntarily. They found it much easier to settle where they could speak the language and get foods that were at least somewhat familiar. As sociologist Herbert Gans described, Boston's Italian West End was a halfway station between the old country and new. Where outsiders often viewed residents as locked in a squalid, archaic society, Gans saw a healthy community that preserved a culture useful for making one's way in America.

Today, advertisers use only Spanish signs in many urban neighborhoods. Polish is the first language in parts of Chicago and South Boston keeps a decidedly Irish flair. Boston's Italian North End is a cherished urban asset, a nearby piece of Italy prized by residents and visitors alike. From the creation of Yiddish theater, to the influence of Irish politicians, to the restaurants of Chinatown, there are many indications that ethnic districts serve valuable social and economic functions.

Nevertheless, the isolation of African-American ghettos from the mainstream city can be quite harmful. Ghettos create artificial barriers that impede critical opportunities for trade and the exchange of ideas, and this deprives residents of the key advantage of living in an urban setting. In addition, segregation impedes the rest of the city from developing advantageous financial, employment, business, and cultural contacts with the ghettoized group.

HISTORY OF AFRICAN-AMERICAN GHETTOS

The African-American ghetto is a creation of the twentieth century. The golden age of Northern black-white relations lies in the period before 1900, write Allan Spear and Kenneth Kusmer, historians of the Midwestern ghettos. Blacks at the time were not generally restricted from using public facilities, and they lived in much more integrated communities than their descendants do today.

Informal practices did limit integration in the North. But only in response to the large-scale black migration north, in the early twentieth century, did these restrictions harden. W.E.B. DuBois, the Harvard-educated black scholar, raised in Great Barrington, Massachusetts, was shocked at the deteriorating conditions he found in the nascent, turn-of-the-century Philadelphia ghetto inhabited by recent migrants from the South "Murder sat on our doorstep, police were our government, and philanthropy dropped in with periodic advice." The apparatus of legal segregation arrived soon thereafter -- zoning by race, restrictive covenants, and a myriad of other devices. The U.S. Supreme Court banned explicit zoning by race in 1917, and restrictive covenants were banned in 1948. But these legal restrictions had served as a mighty handmaiden of segregation; by 1920, the color line in Northern cities had fully hardened.

This reinforcement of ethnic barriers was hardly limited to antiblack initiatives in Northern U.S. cities. The South created its vast array of Jim Crow laws at the end of the nineteenth century. In the West, whites used restrictive covenants against Asians. In Boston, with a long history of attempts to bar Irish immigrants from Yankee institutions, these barriers, and anti-Semitic restrictions as well, were formalized in the early twentieth century.

Domestic tranquility was marred not just by conflicts between native Protestants and both blacks and immigrants, but by tensions between blacks and immigrants, and among different immigrant groups. In 1910, blacks were more segregated from the foreign-born than they were from native whites. Spear's history of the Chicago ghetto describes how immigrants were the fiercest opponents of blacks in that city, and how blacks moved into native white areas rather than face the more violent resistance of the newer Americans.

Segregation increased most in those cities with the greatest black in-migration. Whites felt more threatened by larger influxes of blacks, and their racism grew. Black migrants from the South also found in urban ghettos in the North many of the "attractions" seen in other urban immigrant communities. Most were arriving from an inhospitable, impoverished region that still relied on lynching as a tool of discipline, and many valued the comfort of their own community.

African-American ghettos also started out well, economically. In the Midwest, ghettos were built on high wages from manufacturing jobs. In New York City, the housing was superb. Developers in Harlem had built state-of-the-art apartment buildings around the new subway extension for upwardly mobile whites, writes historian Gilbert Osofsky. But they overbuilt, and entrepreneurial real estate agents, of both races, quickly filled vacant units with blacks. By the end of the 1920s, Harlem was home to the nation's largest concentration of African-Americans. Migrants from the South, to use Nicholas Lemann's phrase, generally had come to see Northern ghettos as "the promised land."

The segregation of the foreign-born also rose, for similar reasons, during their period of great in-migration, 1890 to 1920. But once America ended its open-door immigration policy in the mid-1920s, the segregation of the foreign-born began to decline.

African-American segregation continued to rise however, until it reached its peak in the 1960s. It rose in every decade and in cities of all sizes, and in all regions of the country. While the great growth came before World War II, segregation increased after the war as well. It continued to rise perhaps because the black migration north, stimulated by the cutoff of foreign immigration, extended over a much longer period than the influx of other immigrant groups. And white flight to the suburbs led to an increasingly isolated black inner-city population.

The segregation of blacks in Northern U.S. cities began to level off in the 1960s. The U.S. "segregation index" -- the number of blacks who would need to move to distribute the races evenly across metropolitan areas -- had reached an all-time high of 74 percent. The index thereafter declined quite rapidly to its current 56 percent level, and to 74 percent for twenty-four large Northern cities. Blacks nevertheless still live far more segregated lives than any other U.S. urban group. The segregation index for Hispanics, for example, is 38 percent. And the average urban black lives in a census tract that is 60 percent black; the comparable number for Asians is 19 percent.

The decline in racial segregation from its peak in the 1960s might stem from the end of the legal barriers needed to keep areas all white. Thirty years ago, ghettos existed primarily because legal restrictions made it impossible for blacks to leave. The barriers today are more subtle, and economic. David Cutler, Jacob Vigdor, and I, examining the price of otherwise similar housing, find that ghettos now exist primarily because whites will pay more to live in areas with few, if any, blacks. Middle-class blacks can buy their way out of the ghetto, but those at the bottom of the income ladder are unable to leave. The black segregation index declined primarily because areas that used to be all white now have a small number of blacks. The African-American ghettos have not become any less black. They just house a smaller share of the nation's urban black population.

Maps Black and poor in Boston

IMMOBILITY

Economic conditions in African-American ghettos have deteriorated quite sharply over the past three and a half decades. The inner city, which once might have looked like a promised land, doesn't much resemble one today. This is partly a statistical phenomenon. The ability of more affluent blacks to leave has lowered the average income of those who remain. The poverty of inner-city blacks also reflects the declining economic position of Americans of all races at the bottom of the income ladder. But a growing body of research shows that the segregation of American blacks in inner-city ghettos further damages their economic chances.

The oldest and the most easily understandable evidence on ghettos compares blacks who grew up in segregated neighborhoods with those raised in integrated neighborhoods. The literature began with a 1968 study, by economist John Kain, in which Kain documented that blacks who lived in ghettos had worse labor-market outcomes than those who did not. Kain's explanation was "spatial mismatch" -- that ghetto residents lived far from where the urban jobs were located. According to Kain, the key economic advantage of living in a city -- the opportunities urban environments create for trade and exchange -- thus lay beyond the reach of ghetto residents. Subsequent research has generally corroborated Kain's results. Extremely black neighborhoods are generally located far from job opportunities, and residents do worse, economically, than blacks from more integrated areas.

There is a methodological problem with this type of study, however. A connection between living in a ghetto and being poor need not imply that ghettos create poverty. Poverty could also create ghettos -- it could be that poor people can't afford to live elsewhere.

Katherine O'Regan and John Quigley published a particularly fine study that addressed this issue in the May/June 1996 issue of the New England Economic Review. O'Regan and Quigley's study examined young blacks and Hispanics who still live at home. Since their parents chose the neighborhood, the labor-market outcomes of these young people should have little effect on where they live. So in any correlation between neighborhood and labor-market outcomes, causation should run from neighborhood to outcomes.

O'Regan and Quigley found, in the neighborhoods around Newark, New Jersey, that blacks and Hispanics who live in ghettos are far more likely to be idle -- to be neither in school nor working -- than those from more integrated communities. Their results suggest that the chance the average black or Hispanic youth would be employed or in school would rise a dramatic 10 percentage points if he or she moved to the neighborhood where the average white youth lives.

Why is this so? In addition to spatial mismatch, poor whites may do better because their neighborhoods are economically more heterogeneous. A critical problem with ghettos today is that almost everyone who lives there is poor. Ghettos lack the variety of incomes and skills found in other urban neighborhoods, so opportunities for trade and the exchange of ideas -- again, the key economic advantages of living in cities -- are again unavailable to ghetto residents.

NO CROSSING THE RIVER

Another way to gauge the effects of ghettos is to compare black economic outcomes across different metropolitan areas. Cutler and I divided the metropolitan areas of the United States in half -- into more and less segregated communities -- and examined various outcomes. We found that blacks between ages twenty and twenty-four in the more segregated metro areas are far more likely to be idle 22 percent are neither at work nor in school, compared to 15 percent in the more integrated areas. Segregated blacks are also more likely to have dropped out of high school 26 percent versus 21.5 percent. And segregated black women ages twenty-five to thirty are more likely to have become single mothers -- 45 percent versus 40 percent. These effects are big and statistically significant. They also hold up under alternative methods of estimation and after controlling for region, city size, and the racial composition of the metro area.

(Our study, coincidentally, found no effects of segregation on whites. Whites in segregated areas may seem to monopolize the economy's better-paying positions or otherwise "gain" from segregation. But their incomes, single motherhood, and schooling outcomes are essentially identical to those of whites in more integrated communities.)

It is possible, of course, that black poverty at the metro level causes segregation, not the other way around. (This issue of identifying causation is equivalent to the problem, in the intra-city studies, of determining whether ghettos create poverty or poverty creates ghettos.) Cutler and I examined this issue using a variable created by economist Caroline Minter Hoxby, based on her notion that topographical barriers often serve as neighborhood boundaries. We found that metro areas with more natural boundaries -- like Cleveland with the Cuyahoga River running through it -- are more segregated and have worse black outcomes. The chain of causation here must run from rivers to segregation to poverty. (Rivers presumably do not cause poverty directly; and neither segregation nor poverty causes rivers.) We thus conclude that segregation -- whether created by natural or man-made factors -- results in poor black outcomes.

AMERICAN DREAMS

The African-American ghettos of the mid-twentieth century appear to have been much less harmful than those of today. In the most segregated cities, such as Chicago, Cleveland, and Detroit, African-Americans prospered as workers in America's industrial centers. The fortunes of the ghettos changed, in part, as a result of downturns in manufacturing in postwar America. But the declining vigor of African-American ghettos also resulted from a pervasive feature of all immigrant ghettos. David Cutler, Jacob Vigdor, and I found that immigrant ghettos are generally beneficial, or at least not harmful, for the first generation of residents. Today, first-generation Asians, who often do not speak English, seem to be helped by living in segregated Asian communities. But when we look at later generations still living in the earlier generation's ghetto, we see deleterious effects. This was true of Irish immigrants still living in ghettos in 1910, long after the major Irish immigration waves, or of Eastern European immigrants still living in their ghettos in 1940.

This overall pattern helps us understand why ghettos form and why they can be harmful to residents. The first generation of migrants benefits from the social networks, the cultural comforts, and the protection against native hostility. But ghettos deprive their children of contacts with the broader world and with the informational connections that make cities so strong. The negative effects of ghetto isolation are exacerbated as many of the ghetto's most able children then leave for more integrated communities, or for more prosperous segregated communities. So thirty years after the immigrant ghetto was a vibrant community, it typically becomes an island distant from the city, whose inhabitants rarely experience the best features of U.S. urban society.

RESPONSIBILTY

The empirical evidence clearly indicates that ghettos hurt blacks a great deal. Ghetto walls separate residents from mainstream society, from mainstream jobs, and from contact with successful whites and blacks. The suffering is real, as is the resulting crime, disorder, and social distress. The magnitude of these problems, moreover, is sufficiently large to merit significant government intervention.

While the evidence justifies action, policymakers have little idea about what should be done. In the past, many well-intentioned interventions caused more harm than good.

Perhaps the most egregious example is the large-scale housing projects of the 1950s. This generally well-intentioned policy squeezed as many minorities into as small an area as possible, increased segregation, and worsened ghetto conditions. Forced school integration, or busing, as Charles Clotfelter documents, led to a substantial outflow of white children into private schools, not to increased integration. And enterprise zones, which are currently in vogue, might slow what has been, for other ethnic groups, the process of neighborhood exodus and evolution.

It does seem crucial to lessen discrimination in the housing market. Racism in individual consumer tastes seems to be the primary problem, and government cannot legislate racism away. But government can combat discrimination in real estate marketing and finance.

Policies that generate choice and use incentives instead of controls also hold promise. Housing vouchers and magnet schools, for example, attract individual blacks and whites most willing, or eager, to live and go to school with one another. The nation can also hope that evidence showing a decline in racism over the past twenty-five years is correct, and that the trend will continue.

The damage caused by African-American ghettos reinforces the importance of the idea of the "informational city." Ghetto residents live in cities and face most of the costs --monetary and otherwise -- of urban residence. But the ghetto cuts them off from the informational connections and job markets that make city living worthwhile for so many people.

The city is an enormously positive social institution. It should be able to answer the problems of its own inner core. Breaking down ghetto walls is no small task. But it will be a great achievement to connect inner-city residents to the informational advantages of downtown America.

Genesis

Ghettos are formed in three ways:

• As ports of entry where minorities, and especially immigrant minorities, voluntarily choose to live with their own kind.
• When the majority uses compulsion -- typically violence, hostility, or legal barriers -- to force minorities into particular areas.
• When the majority is willing and able to pay more than the minority to live with its own kind.

All three causes are typically present in the formation of any particular ghetto. But compulsion played an unusually large role in forming the African-American ghettos. We would expect these ghettos to be much more harmful than immigrant ghettos, where immigrants clustered more voluntarily.

It is often alleged that ghettos and the separation of the races create more racism and that racism -- not segregation -- explains why black outcomes are so much worse in segregated cities. This argument, however, relies on the claim that white racism is more extreme in segregated communities.

To examine the link between segregation and racism, David Cutler, Jacob Vigdor, and I examined evidence collected by the National Opinion Research Center. For the past twenty years, the Center has asked respondents whether whites and blacks should be allowed to marry, their assessment of how violent blacks are, and a myriad of other questions designed to display discriminatory attitudes.

Cutler, Vigdor, and I found that whites living in more segregated communities are indeed more likely to have discriminatory attitudes regarding housing. Compared to whites who live in completely integrated areas, those in completely segregated areas are 20 percentage points more likely to believe they have a right to segregated housing; they are 36 percentage points more likely to say they would not live in a neighborhood that was 50 percent black.

But we found no connection between segregation and discrimination on questions not directly connected with housing. Whites in segregated areas actually had a more favorable assessment of blacks on some issues, such as perceiving blacks as violent. For most questions, however, there was just no connection between and segregation and discriminatory attitudes.

White discrimination in housing decisions would seem to be at least partly responsible for residential segregation. But the lack of strong connections between segregation and other racist attitudes suggests that segregation may not lead to more hatred between the races. The ghetto walls themselves, not any increase in racism they may engender, thus seem primarily responsible for the poor black outcomes associated with increased segregation.

so-called, were areas in European cities where Jews were forced to reside. The term, according Jakob Lestchinsky's article in the Encyclopedia of the Social Sciences, originated with fifteenth-century Venice and comes from the Italian getto, or iron foundry, a nearby landmark. Today, the term applies primarily to concentrations of blacks in Northern U.S. cities.so-called, were areas in European cities where Jews were forced to reside. The term, according Jakob Lestchinsky's article in the Encyclopedia of the Social Sciences, originated with fifteenth-century Venice and comes from the Italian getto, or iron foundry, a nearby landmark. Today, the term applies primarily to concentrations of blacks in Northern U.S. cities.