What is the fiscal and economic impact of immigrants on the cities and states where they live? This question goes to the heart of the controversy over America's ability to assimilate immigrants into our economy and social infrastructure without imposing an undue burden on U.S. citizens. It is also a critical policy issue in the current debate over what is the "right number" of immigrants to admit to the United States each year.
The presumption inside the policy community is that immigrants impose a fiscal and economic burden on heavily affected states and cities. For example, Governor Lawton Chiles of Florida, one of the states so affected, complained that "federal immigration policy has created a nightmare for state and local governments."1
The nightmare is said to take many forms. One is an undesirable cultural transformation. Newsweek recently raised the alarm for Californians with the headline "Los Angeles 2010: A Latino Subcontinent."2 Within a generation, predicts the article, "California will be demographically, culturally, and economically distinct from the rest of America." Gordon J. McDonald, former chief of the U.S. Border Patrol, is even more blunt in his assessment of the urban impact of immigrants: "Major cities have already been turned into extensions of foreign countries," he warns. "Aliens threaten to seize political power within a few short years."3
Another problem allegedly created by large concentrations of immigrants in cities is unemployment. Brown immigrants are said to be pushing black Americans out of the job market.4 Still another impact of competition from immigrants in inner cities is white flight. "In the last half of the 1980s, for every 10 immigrants who arrived [in New York, Chicago, Los Angeles, and Houston], 9 residents left for elsewhere," claims a story in the New York Times Magazine.5 Immigration may be leading to the balkanization of America's once-mighty industrial cities.
What is beyond debate is that immigrants have a profound demographic, economic, and fiscal impact on America's largest central cities.6 More than half of all immigrants reside in just seven cities: Los Angeles, New York, Chicago, Miami, San Diego, Houston, and San Francisco. Most of the rest of the nation (at least directly) is negligibly affected by immigration.
There is also little question that many of America's largest cities--Detroit, Saint Louis, Buffalo, Newark, New York, Philadelphia, for example--have been suffering from a severe economic erosion during the past quarter century. The signs of urban sclerosis are evident in statistics on rising crime rates, increasing poverty levels, and losses of jobs, businesses, and the white middle class (though in the current economic recovery, the lot of central cities has improved).7 Yet not all America's central cities are deteriorating. In fact, many, such as Charlotte, Las Vegas, San Jose, San Diego, Austin, and Seattle, for example, are booming. Other cities, such as Dallas and Pittsburgh, are enjoying an economic renaissance that seemed improbable ten or fifteen years ago.
The critical policy issue is whether immigrants help create economic decline in central cities. Often anecdotal evidence is brought in to prove that immigrants are either a help or a hindrance to cities.8 There is no shortage of examples to prove or disprove that immigrants impose net costs on cities. In fact, however, the anecdotes prove little.
This essay attempts systematically to assess the impact of immigration on cities by contrasting the economic condition of high-immigration cities with cities with the lowest immigration over the period 1980–1990 (through 1994 when those data are available). The eighty-five most populous (200,000 or more) U.S. cities are examined.9
On eight of the nine fiscal and economic variables examined, high-immigration cities perform better, not worse, than low-immigration cities. The higher a city's percentage of foreign-born population, the higher its subsequent population growth, the higher its subsequent growth in employment, the higher its per capita income, the faster its increase in per capita income, the lower its poverty rate, the lower its subsequent increase in poverty, the lower its crime rate, and the lower its tax burden. The only economic variable where high-immigration cities fared worse than low-immigration cities was the unemployment rate, which was higher in high-immigration cities than in low-immigration cities.
When we build an index of cities' economic health based on these nine variables, we find that the median foreign-born population of the ten most prosperous cities in the 1980s was 9.0 percent (see figure 1). In the ten fastest-declining cities, the median foreign-born population in 1990 was just 3.8 percent. In fact, of the twelve fastest-declining large cities in the United States in the 1980s, only two, Miami and Houston, had higher-than-average immigration. And Houston's decline was clearly a result of the oil bust, not high immigration.
Figure 1. Percent Foreign-Born, 1990
|HIGHEST ECONOMIC GROWTH CITIES|
|LOWEST ECONOMIC GROWTH CITIES|
These findings do not prove that immigrants cause prosperity in states and cities. But they challenge the conventional belief that immigrants are a contributor to the economic decline of many American cities in that, with few exceptions, the cities with the most severe economic problems have almost no immigrants. Thus immigrants may be potential contributors to rebuilding and revitalizing America's inner cities.
The Impact of Immigrants on Cities: Two Views
Although most economists agree that immigration leads to positive economic benefits for the nation as a whole, the impact of immigrants on affected states and cities is a subject of considerable controversy.10
Immigrants may negatively affect the areas in which they locate in a number of ways.11 First, low-skilled immigrants may displace unskilled minority workers in central cities or depress wage rates in certain industries, such as fast food restaurants, janitorial services, construction, and apparel. If so, this suggests that the "victims" of U.S. immigration policy are low-income inner-city residents, most notably blacks. Because residents of large central cities, in particular, are significantly poorer than average-income U.S. households, it may be that job competition from immigration is widening the disparity between rich and poor.12
Second, immigrants may contribute to the decline of central cities by driving natives out. This is the premise of demographer William Frey of the University of Michigan, who finds that interstate and intercity migration of natives is heavily influenced by immigrant settlement patterns.13 Frey finds that from 1985 to 1990 whites tended to migrate away from states and metropolitan areas with large inflows of international immigrants. He also detects the flight of poor blacks from metropolitan areas with large international immigrant populations.14 Frey's research thus suggests that cities are have becoming balkanized by the influx of international immigrants in recent decades.15
Finally, international immigrants may impose a fiscal burden on cities. Several studies suggest that immigrants in high-migration cities, such as Los Angeles, Houston, and San Diego, use more state and local services--such as schools, hospitals, and roads--than they contribute in state and local taxes.16 Kermit Daniel, professor of public policy at the University of Pennsylvania, examined the budgetary impact of immigrants on twenty-eight major U.S. counties and found that immigration places a "disproportionate fiscal burden" on some of the nation's largest urban areas. He estimates that "immigrants cost localities as much as $8.4 billion a year" and argues for a federal transfer of funds to these areas to compensate city governments for the cost of providing services to immigrants.
But immigrants may also provide offsetting benefits to the cities in which they reside.17 There is evidence, for example, that immigrants have contributed substantially to the economic revitalization of urban areas, particularly gateway cities, such as Chicago, Dallas, Los Angeles, Miami, and San Diego.18 Immigration to some central cities has helped mitigate catastrophic declines in the out-migration of natives.19 Finally, immigrants may contribute to the expansion of the labor market in cities by their purchases of goods and services, by forming their own businesses, and by saving dying industries, such as the textile industry in New York.20
Let us now look at a number of studies in recent years that have highlighted the economic contributions of immigrants, including
- A landmark 1985 Urban Institute study on the impact of immigrants on California, which found that Los Angeles prospered in the 1970s in part because of large economic contributions by immigrants.21 Even with a huge inflow of Mexican manufacturing workers, black employment in Los Angeles for teenagers and adults increased at faster rates than the national average for the decade of the 1970s and through the early 1980s. The study found that, despite the 220,000 new Mexican immigrant households that entered Los Angeles during the 1970s, unemployment rates in the city fell relative to the nationwide rate and per capita income rose faster over the decade. Manufacturing jobs in Los Angeles grew at an astounding four times the rate of the nation as a whole, with immigrants filling between one-third and one-half of the new positions. An estimated one-quarter of the jobs filled by Mexicans would have disappeared and the apparel industry would have migrated across the Mexican border, according to the Urban Institute analysis, if immigrants had not been available. Finally, for California consumers, Mexican immigration meant lower prices for many goods and services and less inflation than the nation as a whole. "The bottom line consensus on the impact of Mexican immigrants on Los Angeles is that, on balance, they are an economic benefit," says coauthor Thomas Espenshade.22
- A separate analysis of Mexican immigrants in California cities, by Kevin McCarthy of the Rand Corporation in 1985, which reached similar conclusions,23 compared 1970 and 1980 U.S. Census Bureau data on demographic, employment, and earnings changes for California. The study found that "widespread concerns about Mexican immigrants are generally unfounded. Overall, the immigrants provided strong economic benefits for the state, with only minor dislocation effects, mostly among native-born Latinos." It further concluded that immigrants' "use of public services in general is not a problem," although it found that Los Angeles and other heavily affected cities had to bear a "disproportionate cost burden while receiving a less-than-proportionate share of the tax revenues."
- A 1986 study by the New York City Department of Immigration Services, which compared immigrants and U.S.-born New Yorkers on a whole range of social and economic statistics, found that the foreign born in New York have a higher labor force participation rate, a lower rate of using public assistance, a lower crime rate, and roughly an equal unemployment rate.24 The report concluded that "New York City seems well able to absorb immigrants at the rate at which they are now entering the city. There seems to be room for them in the job market, in the institutional structure, and in the infrastructure. . . . Immigrants do pose some problems to this city, but the balance is unquestionably favorable."
Clearly, the brief sample of studies cited above calls into question the prevailing opinion among Washington policymakers that immigrants impose net costs on cities and states.25
A Model for Measuring the Local Impact of Immigration
The Census Bureau compiles data on the immigrant population in the eighty-five largest U.S. cities. In the study I classified the cities as low-, average-, or high-immigration cities based on the percentage of their population that is foreign-born. I examine three immigration variables: the percentage of the resident population that was foreign-born in 1990, the percentage that was foreign- born in 1980, and the change in the foreign-born percentage from 1980 to 1990. Economic, social, and fiscal data by city are available from the Census Bureau as well. The Bureau of Labor Statistics gathers the necessary statistics on employment information. The nine economic variables examined here are (1) population growth, 1980–1994; (2) employment growth, 1980–1994; (3) unemployment rate change, 1980–1994; (4) per capita income, 1990; (5) change in per capita money income, 1980–1990; (6) poverty rate, 1990; (7) change in the poverty rate, 1980–1990; (8) crime rate, 1990; and (9) local tax burden, 1990.
Economic Conditions in Cities with High and Low Immigration
We define high-immigration cities in 1990 as those with a foreign-born population of 20 percent or more (see the seventeen cities in table 1). These seventeen cities are all located in six states: California, Florida, New York, Texas, New Jersey, and Massachusetts. I define low-immigration cities in 1990 as those with 3 percent foreign-born or fewer. Conveniently, seventeen cities also fall into this category (see table 1). Ten of those seventeen cities are in southern states.
Table 1. High- and Low-Immigration Cities, 1990
|New York City||N.Y.||28.4||Saint Louis||Mo.||2.5|
The twelve cities that had 15 percent or more foreign-born in 1980 we defined as high immigration. The thirteen cities that had foreign-born populations of 2.5 percent or less in 1980 we defined as low-immigration (see table 2). All twelve high-immigration cities in 1980 were on the larger list of high-immigration cities in 1990. Atlanta was a low-immigration city in 1980 but not in 1990.
Table 2. High- and Low-Immigration Cities, 1980
|New York City||N.Y.||23.6||Richmond||Va.||1.8|
Fifteen cities saw their foreign-born populations grow by more than 7 percentage points between 1980 and 1990.26 We define these as high-immigration increase cities. Conveniently, there were also fifteen cities that recorded a reduction in the percentage of foreign-born; we refer to these as immigration loss cities (see table 3). Twelve of the fifteen high-immigration increase cities from 1980 to 1990 were located in California, making that list very different from the 1980 and 1990 lists. Of the fifteen high-immigration increase cities, five--Fresno, San Bernardino, Houston, Riverside, and Oakland--were not high-immigration cities in 1980 or 1990. Ten of the fifteen cities that lost immigration from 1980 to 1990 were not low-immigration cities in 1980 or 1990.
Table 3. High- and Low- Immigration Growth Cities, 1980–1990
High-Immigration Growth Cities
Low-Immigration Growth Cities
|City||State||Percentage Change in Foreign-Born||City||State||Percentage Change in Foreign-Born|
|Santa Ana||Calif.||+20.4||Saint Louis||Mo.||0.1|
|Los Angeles||Calif.||+11.3||Colorado Springs||Colo.||0.3|
Tables 4, 5, and 6 show how these cities compare on various measures of economic performance. In the following sections we examine the results for each economic variable.
Table 4. Economic Growth in 1990 High- and Low-Immigration Cities
|City||State||Population Growth 1980–92 (in percent)||Employment Growth 1980–94
|Percentage Change in Unemployment Rate 1980–90||Per Capita Money Income 1989||Per Capita Money Income Growth 1979–89 (in percent)||Families in Poverty, 1989 (in percent)||Percentage Change in Poverty Rate 1979–89||Crime Rate 1991 (per 100 residents)||City Tax Revenue, as % of Money Income 1990|
|New York City||N.Y.||3||0||0.9||16,281||124||16.3||0.9||9.2||12.7|
|*Ranked in descending order by 1990 percent foreign-born.|
Table 5. Economic Growth in High- and Low-Immigration Cities, 1980
|City||State||Population Growth 1980–92 (in percent)||Employment Growth 1980–94 (in percent)||Percentage Change in Unemployment Rate 1980–90||Per Capita Money Income 1989||Per Capita Money Income Growth 1979–89 (in percent)||Families in Poverty, 1989 (in percent)||Percentage Change in Poverty Rate 1979–89||Crime Rate 1991 (per 100 residents)||City Tax Revenue, as % of Money Income 1990|
|New York City||N.Y.||3||0||0.9||16,281||124||16.3||0.9||9.2||12.7|
|*Ranked in descending order by 1980 percent foreign-born.|
Table 6. Economic Growth in High- and Low-Immigration-Growth Cities, 1980–1990
|City||State||Population Growth 1980–92 (in percent)||Employment Growth 1980–94 (in percent)||Percentage Change in Unemployment Rate 1980–90||Per Capita Money Income 1989||Per Capita Money Income Growth 1979–89 (in percent)||Families in Poverty, 1989 (in percent)||Percentage Change in Poverty Rate 1979–89||Crime Rate 1991 (per 100 residents)||City Tax Revenue, as % of Money Income 1990|
|*Ranked in descending order by 1980–1990 percent foreign-born.|
Migration patterns between areas may be the single-best measure of the livability of cities. People vote with their feet. They leave areas that are undesirable and enter areas that are more desirable. Hundreds of thousands of Germans fled from East Berlin to West Berlin before the fall of the Wall; few left West Germany for East Germany.
No single indicator captures the decline of many major U.S. cities more directly than population loss. Certain cities are being depopulated as Americans vote with their feet against these areas' destructive economic policies and social conditions. Since 1965 fifteen of the largest twenty-five U.S. cities have lost a combined three million people, although the total U.S. population has risen by 60 million. Cleveland, for example, had nearly one million residents in the 1950s but now has only about 500,000.27
The 1990 high-immigration cities, however, had a 20 percent population growth from 1980 to 1992. Only two of the seventeen high-immigration cities in 1990 lost population over this period: Yonkers and Boston. Although intuitively obvious, this result contradicts the hypothesis that there is a near one-to-one replacement of native-born with immigrants in cities.28 In five of the highest- immigration cities, the increase in the foreign-born accounted for more than 100 percent of the population growth, suggesting an out-migration of natives, as was the case in Miami, New York, Jersey City, Yonkers, and Boston. But the seventeen cities with the fewest immigrants in 1990 averaged a loss of 1.5 percent of their population. In fact, the six cities with the lowest percentage of immigration in 1990--Jackson, Shreveport, Birmingham, Memphis, Louisville, and Richmond--all lost population from 1980 to 1992. Only Lexington, of the seventeen low-immigration cities, experienced modest population growth.
Perhaps a better indication of how immigration affects the population growth of cities is to compare the cities with the most immigrants at the start of the period, 1980, with subsequent population growth. The average population growth of the twelve cities with the most immigrants in 1980 was 16 percent. The thirteen cities with the fewest immigrants in 1980 experienced subsequent population declines of 2 percent. Eight of the thirteen low-immigration cities lost people from 1980 to 1992. The population gain in the high-immigration cities in 1980 is at least in part explained by the settlement patterns of immigrants: they tend to be attracted to those areas where fellow countrymen also live, creating a chain migration effect.
Finally, we contrast the population changes in high- and low-immigration cities as defined by those that gained and lost the most immigrants from 1980 to 1990 (change in the foreign-born population as a share of population from 1980 to 1990). The fifteen high-immigration cities experienced a 30 percent increase in overall population. The increase in immigration did not account for all the population growth. In only two of these fifteen cities--Jersey City and Houston--did the increase in the foreign-born population account for more than 100 percent of the population increase. The fifteen high-immigration cities with the largest loss of foreign-born population as a share of income experienced a 4 percent decline in population. Only two low-immigration cities (1980–90) gained population: Colorado Springs and Omaha.
An indicator of an area's economic vitality, employment growth, is an indirect way of measuring new business creation, capital investment, and overall business climate in the area. Dying cities are typically characterized by dramatic job losses. Saint Louis, for example, has lost nearly two of every five jobs it held in 1960.29 Employment loss may be thought of as businesses voting with their feet.
When we examine the 1980 foreign-born data, we find that high-immigration cities had nearly twice as much subsequent job growth (19 percent) as low-immigration cities (10 percent). But the average-immigration cities in 1980 had the highest subsequent job growth, 29 percent.
The strongest relationship was between the change in foreign-born population from 1980 to 1990 and job creation from 1980 to 1994. The fifteen cities with the largest increase in immigration had 32 percent job growth, versus 27 percent job growth in average-immigration cities and just 7 percent job growth in the fifteen cities that lost immigration. All the high-immigration-growth cities gained jobs, and in only two of the fifteen--Glendale and Houston--did immigration growth outpace job growth. Conversely, seven of the fifteen cities that lost immigration in the 1980s also lost jobs. Although these findings do not indicate whether job creation expanded sufficiently to accommodate the inflow of all new immigrant and native-born workers, local job markets expand substantially as more immigrants enter the area.
These results indicate that job creation is more prevalent in high-immigration cities; they do not indicate whether immigrants are responsible for the job growth. Immigrants tend to be attracted to areas where economic opportunity is plentiful. We do not know, however, whether the jobs attract the immigrants, the immigrants create the jobs, or both cause the other.30 We do know, however, that the central cities that are losing jobs are areas with few immigrants.
Perhaps the issue of greatest concern to workers and Washington policymakers is whether immigrants displace U.S. citizens from jobs. Job displacement has been a worry as far back as the middle of the nineteenth century, when Frederick Douglas warned that "every hour sees the black man elbowed out of employment by some newly arriving immigrant."31 A huge volume of economic literature has investigated the issue; the surprising consensus that has emerged is that the job displacement impact of immigrants is either nonexistent or very small.32 There is little if any scientific support for the contention that immigrants cause unemployment in cities33 or in states.34
Still it is worth asking whether cities with high immigration have high or low unemployment. In the 1980s, California, Florida, Massachusetts, and Texas, for instance, each had high immigration but lower than average unemployment. States such as Iowa, West Virginia, and Wyoming had virtually no immigration but high and rising unemployment (relative to the U.S. average) in the 1980s.
We examined the change in the unemployment rate from 1980 to 1990. Cities with the most immigrants in 1980 had a subsequent 1.8 percentage point increase in their average unemployment rate from 1980 to 1990, versus a 1.0 percentage point increase in the unemployment rate for the cities with the fewest immigrants in 1980. The 1980–1990 unemployment rate fell in only one 1980 high-immigration city, San Diego. But it fell in five of 1980 low-immigration cities: Saint Louis, Lexington, Nashville, Indianapolis, and Louisville.
The average unemployment rate also rose faster in cities with the largest increase in immigration from 1980 to 1990 than it did in the cities that lost immigrants over that period. The rise in unemployment was 1.3 percentage points in the high-immigration cities (1980–90) versus 0.6 percentage points in the cities that lost immigrants.35
Although unemployment rates rose faster in high-immigration cities than in low-immigration cities, there is little statistical evidence that immigrants are a significant cause of that unemployment. Nonetheless, high-immigration cities experienced an average increase in unemployment of between 0.5 and 1.0 percentage points higher than in low-immigration cities in the 1980s. This result, combined with the result from the previous section, suggests that high-immigration cities generate substantial job growth but also substantial competition for jobs.
Per Capita Income
Per capita income is a reasonably good proxy for wage rates within an area. The average per capita income of the seventeen cities with the most immigrants in 1990 was $14,750. This was 8 percent and $1,000 higher than the $13,720 per capita income of the seventeen cities with the fewest immigrants in 1990. In fact, eleven of the seventeen high-immigration cities had a per capita income above $15,500. Yet not one of the thirty cities with the lowest percentage of foreign-born in their population in 1990 had a per capita income that high.
The twelve cities with the most immigrants in 1980 also had higher per capita incomes in 1990 than the cities with the fewest immigrants in 1980, though the differences were less dramatic. The per capita income for the high-immigration cities was $14,250, or roughly $500 higher than the $13,720 in the low-immigration cities.
Per capita incomes were higher but only by $300 per person in the cities with the largest increases in immigration from 1980 to 1990 than the cities with a net loss of immigrants. The cities with average immigration from 1980 to 1990 had slightly higher per capita incomes ($35 per person) in 1990 than the high-immigration cities.
Still the conclusion from each of these analyses is the same: in general, cities with high immigration are richer and have more dynamic economies that cities with low immigration. For the average urban household, the wealth differential can be as high as $2,000 more than in low-immigration cities. To the extent that per capita income is related to average wage rates, these figures suggest that average wages may be higher, not lower, in high-immigration cities, even if the immigrants compete with the native-born for jobs in selected industries. These figures are also consistent with those of Urban Institute economist Thomas Muller, who finds that personal income is higher in metropolitan areas with large immigrant populations than in those with small immigrant populations. Muller reports, for example, "a strong positive correlation between the percentage of foreign-born [in metropolitan areas] and black household incomes."36
Growth of Per Capita Income
Do high-immigration cities gain wealth at a faster pace than low-immigration cities? The answer appears to be yes. We examine whether the fifteen cities with the most immigrants in 1980 had higher subsequent per capita income growth from 1980 to 1990 than cities with the fewest immigrants. The per capita income growth was 95 percent in the high-immigration cities in 1980 but just 88 percent in the low-immigration cities in that year.
Those cities that had the largest percentage increase in immigration from 1980 to 1990 also fared better than cities that lost immigrants. The fifteen high-immigration cities (1980–90) had a per capita income growth of 84 percent, versus 76 percent for the fifteen cities with a net loss of immigrants. But for both high- and low-immigration cities, the gain in income was lower than the 92 percent income rise in cities with an average increase in immigration in the 1980s.
These results are particularly surprising in that when immigrants first arrive in the United States, they tend to have lower-than-average earnings and then go through a ten- to fifteen-year "catch-up" phase. The fact that residents of high-immigration cities are wealthier and gaining wealth faster than low-immigration cities suggests that the immigrants contribute to the earnings potential of the natives. At least one study provides evidence that this was the case in the 1980s.37
Poverty is a systemic problem in many inner cities today. Many cities have household poverty rates well over 20 percent. These high poverty rates contribute to the social decay of cities.
The cities with the most immigrants in 1990 had an average poverty rate of 13.3 percent. This was 20 percent below the 16.0 percent average poverty rate in the cities with the fewest immigrants. Two of the seventeen high-immigration cities had poverty rates of more than 18 percent: Miami, 25.7 percent, and El Paso, 21.2 percent. But nine of the seventeen low- immigration cities had poverty rates of more than 18 percent.
When we examine the high- and low-immigration cities in 1980, we discover the same inverse relationship between foreign-born population in 1980 and poverty in 1990. The poverty rate of the cities with the most immigration in 1980 was 14.9 percent. This rate was slightly above the eighty-five-city average of 14.3 percent but 2 percentage points below the 16.9 percent average in the cities with the fewest immigrants in 1980.
The most dramatic difference in poverty, however, appears between the cities with the largest increase in immigration from 1980 to 1990 and the cities with a net loss of immigrants over that period. The cities that gained the most immigrants had a poverty level of 13.3 percent, versus 17.7 percent in the cities that lost immigrants. With few notable exceptions, such as Omaha and Indianapolis, high poverty rates are characteristic of large U.S. cities with a low presence of immigrants.
Change in Poverty
Is the poverty problem becoming more severe in cities with high levels of immigration? One might expect this to be the case since newly arriving immigrants are themselves somewhat more likely to live in poverty than natives. Yet I find just the opposite. The poverty rate rose by only 1.4 percentage points in the 1980s in the cities with the most immigrants in 1980, compared with almost twice that increase (2.7 percentage points) in the cities with the fewest immigrants.
The cities adding the most immigrants from 1980 to 1990 had lower increases in poverty (2.5 percentage points) than the cities that lost immigration (4.2 percentage points). But the cities with average increases in immigration in the 1980s had the smallest increase in poverty, 1.5 percentage points.
One primary factor that leads to the breakdown of cities is high crime rates, which create an atmosphere of terror in many central cities and eventually contribute to the exodus of business enterprises and families. Anti-immigrant groups often allege that crime is exacerbated by large-scale immigration.
Whether immigrants are more or less likely to engage in criminal behavior is a matter of some dispute.38 But cities with large percentages of foreign-born in their population are typically low-crime, not high-crime cities. For example, the seventeen cities with the most immigrants in 1990 had a 1991 crime rate of 8.7 per 1,000 population. The cities with the fewest immigrants had a crime rate 17 percent higher, or 10.5 per 1,000 persons. Of the high-immigration cities, only Miami, which has the second-highest crime rate in the nation, had a crime rate of at least 12.0 per 1,000 or more. But of the seventeen low-immigration cities, six had crime rates that high: Jackson, Birmingham, Mobile, Saint Louis, Kansas City, and Baton Rouge.
The cities with the most immigrants in 1980 also had lower crime rates on average (9.2 per 1,000) than the low-immigration cities (11.1 per 1,000). A similar negative relationship exists for the crime rates and the increase in immigration between 1980 and 1990. In sum, crime is more rampant in cities with few immigrants than in cities with many immigrants.
Many state and local officials want the federal government to compensate them for the costs of financing services for immigrants. If immigrants make especially heavy use of municipal services--such as the schools, social services, hospitals, and infrastructure--and if this imposes a fiscal burden of billions of dollars on cities, as some researchers have estimated, then we might expect to see high-immigration cities with heavier tax burdens than low-immigration cities.
To investigate this issue, we examined the tax burden in 1990 for cities as measured by total city taxes as a share of personal income. We found that the high-immigration cities in 1990 had a tax burden (3.9 percent of personal income) that was 10 percent less than the 4.3 percent tax burden in low-immigration cities. Average-immigration cities in 1990 had the lowest tax burden, 3.4 percent of personal income.
The cities with the most immigrants in 1980 had a higher average tax burden in 1990 (4.5 percent) than cities with the fewest immigrants (4.2 percent). Again, the average-immigration cities had the lowest tax burden, 3.4 percent. The average tax burden in the high-immigration cities in both 1980 and 1990 is influenced disproportionately by New York City, a high-immigration city with a tax burden of 12.7 percent, at least twice the national average. If New York were excluded from the analysis and treated as an outlier, the average tax burden in high-immigration cities would fall substantially below the tax load in the low-immigration cities in both 1980 and 1990.
The strongest relationship between city taxes and immigration emerges when we examine the change in immigration from 1980 to 1990. The fifteen cities gaining the most immigrants in the 1980s had an average tax burden of 2.7 percent in 1990 compared with 4.0 percent in the fifteen cities that lost immigrants from 1980 to 1990. In fact, not one of the fifteen cities with the large gains in immigration in the 1980s had a tax burden exceeding the eighty-five-city average. Four of the five cities with the largest losses in immigration in the 1990s had tax burdens well above the eighty-five- city average of 3.7 percent: Detroit, 5.4 percent; Pittsburgh, 4.8 percent; Rochester, 5.7 percent; and Cleveland, 5.7 percent.
These findings seem to contradict the popular notion that immigrants impose a large fiscal burden on taxpayers. The case for a federal transfer of tax dollars to cities highly affected by immigration is especially unpersuasive given that the cities with the largest share of new immigrant arrivals in the 1980s had low tax burdens in 1990. Federal aid would require residents of cities with higher-than-average local tax burdens to pay more taxes to subsidize high-immigration cities with lower-than-average local tax burdens.
An Alternative Framework of Analysis
The analysis so far in this essay has contrasted the economic and fiscal performances of high- and low-immigration cities. An alternative framework is to examine the ten cities with the best and worst record on each quality of life measure. Then we can ask, Do the healthiest cities have more or fewer immigrants than the fastest-declining cities? We can also assess whether the healthiest cities in the 1980s gained immigrants or lost immigrants during that period?
The ten cities with the fastest population growth from 1980 to 1992 had a median foreign-born population of 8.1 percent in 1990, almost twice the 4.3 percent in the ten fastest population losers. The median 1980 foreign-born population was slightly lower (5.1 percent) in the ten highest population-gain cities than the top ten losers (5.7 percent).
The median foreign-born population in 1990 in the ten cities with the most rapid job growth was 6.6 percent, versus just 4.2 percent in the ten fastest job-losing cities. The foreign-born population in 1980 was slightly lower (4.9 percent) in the top ten job gainers than in the top ten job losers (5.4 percent).
The ten cities with the largest increase in unemployment from 1980 to 1990 had a median foreign-born percentage in 1990 of 4.8 percent. This was less than the median foreign-born population for all eighty-five large cities but 1 percentage point higher than the median foreign-born in the ten cities with the largest declines in unemployment. The two cities with the largest increases in unemployment were low-immigration cities: New Orleans and Shreveport. The city with the largest decline in unemployment was a high-immigration city: Sacramento.
Per Capita Income
The ten cities with the highest per capita income in 1990 had substantially higher median foreign-born populations in that year than the ten cities with the lowest per capita incomes (17.5 percent vs. 11.6 percent). The median foreign-born population in 1980 of the ten highest- and lowest- income cities were the same, but the change in the foreign-born from 1980 to 1990 was much higher in the ten highest-income cities in 1990.
Per Capita Income Growth
The ten cities with the highest growth in income had a median foreign-born population in 1980 that was higher than the median for the slowest per capita income growth cities (10.2 percent versus 7.8 percent). The cities with the greatest gains in income in the 1980s also gained more immigrants than the cities with the smallest gains in income.
The ten cities with the lowest poverty rate in 1990 had a median foreign-born population of 10.3 percent, which was twice the median foreign-born population of 4.3 percent in the ten cities with the highest poverty rates.
Poverty Rate Change
The median foreign-born population in 1980 was higher in the ten cities with the largest declines in poverty than in the median city among the ten with the largest poverty increases (5.3 percent versus 6.6 percent). The median change in foreign-born from 1980 to 1990 among the ten cities with the biggest decline in poverty was 2.1 percent, whereas the median in the largest increase in poverty cities was 0.1 percent.
Low-crime cities had roughly twice as large an immigrant presence as the high-crime cities in both 1980 and 1990.
For each of the three measures of immigration, the cities with the highest taxes had substantially lower shares of foreign-born in their populations than the ten cities with the lowest taxes. For example, the lowest-tax cities had a median foreign-born population of 9.8 percent in 1990. The median for the highest-tax cities was 5.0 percent.
Index of Economic Health
The nine economic variables examined above can be equally weighted and then converted into an urban economic health index. This index reveals that the ten highest economic growth cities in the 1980s had a mean foreign-born population in 1990 of 9.0 percent. The ten cities that declined most rapidly had a median foreign-born population of 3.8 percent (see table 7). Five of the ten fastest-growing cities were high-immigration cities: San Jose, San Diego, Fremont, Huntington Beach, and Las Vegas. Only two of the ten fastest-declining cities were high-immigration cities in 1990: Miami and Houston.
Table 7. Immigration in Cities with Highest and Lowest Economic Growth
|Highest Economic Growth Cities||State||
Foreign-Born Population (% of total)
|Percentage Change in Foreign-Born|
|Lowest Economic Growth Cities|
The ten highest-growth cities in the 1980s also started the period with larger immigrant populations (6.0 percent) than the cities with the most severe economic deterioration (3.5 percent). Birmingham, Saint Louis, Baton Rouge, and Shreveport all started the decade with virtually no immigrants and suffered severe declines in the 1980s.
Nor did these cities add immigrants in the 1980s. The median increase in immigration in the ten highest-growth cities was 3.0 percent, versus a median increase in the foreign-born population of zero in the fastest-declining cities. Of the ten cities that improved their economic and social conditions the most in the 1980s, only one, Colorado Springs, lost immigrants. But five of the slowest-growth cities--Milwaukee, Saint Louis, Shreveport, Cleveland, and Detroit--lost immigrants.
Do Immigrants Cause Urban Revival?
The correlations between city immigration measures and city economic performance variables are low in all cases. But in the majority of the cases the signs indicate that high immigration and good economic performance are weakly related.
The implications of the findings cited above are confounded by the fact that immigrants have historically been attracted to states and areas with strong economies and opportunities. Indeed, this is one of the beneficial characteristics of international immigrants: their mobility leads them to areas where their labor services are most in demand.
Are immigrants contributing to the economic well-being of cities or simply choosing areas that are already prosperous? This essay, unfortunately, cannot provide the answer. The best evidence we have that immigrants are contributing to the success of the cities where they live is that those cities with the largest numbers of immigrants arriving before 1980 generally had strong subsequent economic growth during the 1980s. Because the immigrants came before the growth, the subsequent growth could not have lured them.
The research reported in this essay does not prove that immigrants cause cities' economic prosperity. Rather, it challenges the conventional belief that immigration is a leading cause of urban decline. Most of the evidence presented here contradicts that argument. The presence of immigrants is unquestionably associated with urban growth and renewal, rather than decline.
Nearly two hundred years ago James Madison observed: "That part of America that has encouraged [foreigners] has advanced most rapidly in population, agriculture, and in the arts." This appears to have been precisely the case in the 1980s.
On a whole range of variables measuring the prosperity of cities--from population growth to poverty rates to income growth to crime to taxes--cities with large foreign-born populations fare better than cities with few immigrants. Cities that had the largest increase in international immigration in the 1980s also fared substantially better than the cities that lost immigrants over that period. High-immigration cities fared better than low-immigration cities on all the quality-of-life measures examined, with the exception of unemployment.
These results call into question the widely held belief that immigration is a major contributor to the demise of many of America's large central cities. Some researchers have pointed to a handful of high-immigration cities that are suffering from economic decline as evidence that immigrants cause urban sclerosis. There is a lively debate among urban scholars about the factors that caused the decline of so many of America's once mighty industrial cities. Immigration is probably not one of them.
Much of the research reported in this essay was financed by the Alexis de Tocqueville Institute in Arlington, Virginia.
1 Quoted in Catherine Yang, "Who Picks Up the Tab for Aliens?" BusinessWeek, March 28, 1994, p. 34.
2. Michael Meyer, "Los Angeles 2010: A Latino Subcontinent," Newsweek, November 9, 1992, p. 32.
3 Quoted in Gregory A. Fossedal, "Cry the Beloved Statue," Copley News Service, June 7, 1990.
4Jack Miles, "Blacks vs. Browns," Atlantic, March 1991.
5 William H. Frey and Jonathan Tilove, "Immigrants in, Native Whites Out," New York Times Magazine, August 20, 1995, pp. 44–45.
6 Richard B. Freeman and John M. Abowd, eds., Immigration, Trade, and the Labor Market (Chicago: University of Chicago Press, 1990).
7 Stephen Moore and Dean Stansel, "The Myth of America's Underfunded Cities," Policy Analysis (Cato Institute), 1993.
8 For example, the Federation for American Immigration Reform (FAIR) recently purported to show that immigrants have a negative impact on cities by selecting five high-immigration cities and five low-immigration cities. FAIR's conclusion was that "less immigration means a better family environment and a better quality of life for everyone." But the authors selectively chose the ten cities, out of a sample of more than three hundred metropolitan areas, that supported their conclusion. See FAIR, "A Tale of Ten Cities: Immigration's Effect on the Family Environment in American Cities," Washington, D.C., 1995. For a critique of the study, see John Berthoud, "FAIR's a Tale of Ten Cities: A Fair Analysis?" Alexis de Tocqueville Institution, Arlington, Va., 1995.
9 The data in this study cover the central cities, not the metropolitan areas.
10 For a review of the overall economic impact of immigration, see Julian L. Simon, The Economic Consequences of Immigration (Cambridge, Mass.: Basil Blackwell, 1990); Vernon Briggs and Stephen Moore, Still an Open Door? (N.p.: American University Press, 1994); and U.S. Department of Labor, The Effects of Immigration on the U.S. Economy and Labor Market (Washington, D.C.: Government Printing Office, 1990).
11 Randolph K. Filer, "The Impact of Immigrant Arrivals on Migration Patterns of Native Workers," paper presented at the U.S. Department of Labor conference on immigration, Washington, D.C., 1988.
12 Miles, "Blacks vs. Browns."
13 William Frey, "Immigration and Internal Migration for U.S. States: 1990 Census Findings by Poverty Status and Race," Population Studies Center, University of Michigan, September 1994; and William H. Frey, Kao-Lee Liaw, Yu Xie, and Marcia Carlson, "Interstate Migration and U.S. Poverty Population: Immigration Pushes and Welfare Magnet Pulls," Population Studies Center, University of Michigan, 1995; and William H. Frey, "Immigration and Internal Migration Flight from U.S. Metropolitan Areas: Toward a New Balkanization," Urban Studies 32, no. 4 (1995): 733–57.
14 William Frey, "White and Black Flight from High Immigration Metro Areas: Evidence from the 1990 Census," Population Studies Center, University of Michigan, 1995.
15 William Frey, "The Balkanization of America," New York Times Magazine, 1994.
16 California Senate Office of Research, "Californians Together: Defining the State's Role in Immigration," Sacramento, Calif., 1993.
17 The best and most comprehensive summary of the positive contributions of immigrants to cities is in Thomas Muller, Immigrants and the American City (New York: New York University Press, 1993).
18 For example, a 1993 New York Times report came to this benign conclusion on the impact of immigrants on the city: "Immigrants have become integral to maintaining the vitality of New York neighborhoods. According to city planners, the influx of newcomers has brought indirect urban renewal, reversing the blight that threatened New York in the 1970s and helping to avoid serious inner-city population loss that has plagued such cities as Philadelphia and Detroit."
19 See, for example, Michael Mandel, "The Immigrants: How They're Helping to Revitalize the U.S. Economy," BusinessWeek, July 13, 1992, pp. 144–21.
20 Briggs and Moore, Still An Open Door? pp. 123–33.
21 Thomas Muller and Thomas J. Espenshade, The Fourth Wave: California's Newest Immigrants (Washington, D.C.: Urban Institute, 1992).
23 Kevin McCarthy and R. Burcigo Valdez, Current and Future Effects of Mexican Immigration in California (Santa Monica, Calif.: Rand Corporation, 1985).
24 Elizabeth Bogen, Testimony before the Joint Economic Committee, U.S. Congress, Hearings on the Economic and Demographic Consequences of Immigration, May 29, 1986.
25 U.S. Commission on Immigration Policy (Jordan Commission) report presented to Congress, 1995.
26 This cutoff point was chosen because of a natural break of nearly 1 percentage point between the fifteenth city, Oakland, and the sixteenth city, Dallas.
27 Moore and Stansel, "Myth of America's Underfunded Cities."
28 This finding corroborates the conclusion of an analysis of Current Population Survey data in 1980 and 1985 by Princeton University economists Kristin Butcher and David Card, which finds that native population growth rates in cities and immigration levels into cities may be positively correlated. When Butcher and Card examined the population growth rates for twenty-four major U.S. cities, they discovered that immigration has neither a positive nor a negative relationship on native population growth. However, when they exclude the three most immigrant-intensive cities--Los Angeles, Miami, and New York--from the sample, they find that each additional immigrant led to an increase of 1.76 U.S.-born residents into the city. Americans are not leaving cities because of the presence of immigrants.
29 Moore and Stansel, "Myth of America's Underfunded Cities," p. 6
30 The fact that immigrants migrate to areas with the most jobs and opportunities is one of the economically beneficial traits of immigrants.
31 Quoted in Muller and Espenshade, The Fourth Wave, p. 92.
32. For a summary of these studies, see Briggs and Moore, Still an Open Door?
33 Julian L. Simon, Richard Sullivan, and Stephen Moore ("The Effect of Immigration upon Unemployment: An Across City Estimation," Journal of Labor Research, 1993) examined the effect of the immigration rate (new immigrant arrivals to a city as a percentage of the population of the city) in seventy-seven major immigrant-receiving cities on the change in the unemployment rate in those cities over the period 1960 to 1977. Using a pooled regression technique (which combines the cross-section and time series data), the study examined the impact of the immigration on the local unemployment rates for lag periods of one to five years. For example, what happened to the unemployment rate in a city in 1973, 1974, 1975, and 1976 in response to immigration into the city in 1972. The study found a small, statistically insignificant positive relationship between immigration to a city and unemployment two years later. An upper-bound estimate of the amount of unemployment created by the immigrant arrival to a local labor market is the one-time loss of 0.85 person-years of unemployment for a U.S. worker over the immigrant's working life. This is roughly the equivalent of saying that every thirty immigrant arrivals leads to one U.S. worker becoming permanently unemployed.
Economists Michael Greenwood and G. L. Hunt ("Migration and Interregional Employment Redistribution in the United States," American Economic Review, December 1984, pp. 957–69) found the employment effects of immigrants in fifty-seven metropolitan areas between 1958 and 1975. They found that each employed in-migrant leads to 1.26 jobs (the job for the immigrant plus 0.26 additional jobs); conversely, each out-migrant worker leads to the loss of 0.26 jobs over time. The researchers found a net displacement effect in only three of the fifty-seven cities: Baltimore, Grand Rapids, and Los Angeles. In sum, Greenwood and Hunt's analysis can be interpreted to mean not only that the job market stretches to accommodate an influx of immigrants but that employment expands to create additional job opportunities for the U.S.-born workers in the area.
Finally, a comprehensive study of the impact of immigrants on cities between 1970 and 1980 by economist Richard Freeman of Harvard finds evidence that job growth and immigration were positively related:
There is some evidence that in cities with more immigrants, employment grew more rapidly or declined more slowly in low wage industries where immigrants tended to find jobs and that less skilled natives moved into better jobs. The broad implication is that immigrants have been absorbed into the American labor market with little adverse effect on natives.
34 Richard Vedder, Lowell Galloway, and Stephen Moore ("Do Immigrants Increase Unemployment or Reduce Economic Growth?" Congressional Record, September 26, 1990) examined the U.S. experience with immigration from 1960 to 1990 to determine whether high-immigration states had higher or lower than average unemployment rates over this period. They found that, if anything, the relationship between immigration and state unemployment was negative. The ten highest-immigration states had an average unemployment rate of 5.8 percent, whereas the ten lowest-immigration states had an average unemployment rate of 6.6 percent.
35 We also found that the average unemployment rate in 1995 was substantially higher in the high-immigration cities in 1990 (7.9 percent) than the low-immigration cities (5.3 percent). None of the low-immigration cities in 1990 had an 1995 unemployment rate above 8 percent, whereas eight of the high-immigration cities did: Miami, Santa Ana, Glendale, Los Angeles, New York, Jersey City, El Paso, and Stockton. (The national average for unemployment was 5.6 percent.) But the data that come from the Bureau of Labor Statistics are less reliable, especially when conducting intercity comparisons, than the 1980 and 1990 Census Bureau unemployment data.
36 Muller and Espenshade, The Fourth Wave, pp. 174–75.
37 See Maria Enchaautegui, "Effects of Immigrants on the 1980–90 U.S. Wage Experience," Contemporary Economic Policy, July 1995, pp. 20–34.
38 See discussion in Muller and Espenshade, The Fourth Wave, pp. 213–17.