How Immigration Harms Minorities

The Public Interest

Norman Matloff

footnote: Norman Matloff is a professor of computer science (formerly statistics) at the University of California, Davis. A speaker of Chinese, he has been active in the Chinese immigrant community for more than 20 years.

Summer, 1996

The popular press tends to portray proposals to reduce yearly immigration quotas as pitting whites against nonwhites. Yet they are overlooking the fact that the downsides of immigration often fall more heavily and more directly on nonwhites, i.e., on blacks, Asians, and Latino Americans. As one Mexican American put it, when immigration's negative impacts come, ``We get hit first."

It is thus no wonder that many minorities strongly support immigration reform, both in terms of legal and illegal immigration. An Empire State Survey taken in 1993 found that half of the immigrants in New York agreed with the statement, ``Immigration has made this city a worse place in which to live." In California's 1994 vote on Proposition 187, the rates of ``yes" votes amongblacks, Asian Americans, and immigrants were all near the overall statewide rate of 59 percent. A Hispanic USA survey found that up to 80 percent of MexicanAmericans agreed that ``there are too many immigrants."

And, though it has become politically popular to draw a sharp line between legal and illegal immigration, the fact is that minorities tend not to make this distinction, since the problems arising from the two kinds are typically the same--overcrowded schools and depressed wages. Moreover, the yearly volume of legal immigration is several times larger than that of the illegal kind. Thus, in this article, the term ``immigration" will refer both to legal and illegal immigration, with an implicit focus on the former.

Most analyses of the impact of immigration focus on economic issues. These are not unimportant, and I will describe the economic harms brought upon minorities by immigration. Yet the noneconomic problems are at least as important, and probably more so. The current high yearly immigration quotas are contributing to a pervasive (though largely unconsciously created and maintained) new American caste system among U.S. minorities, with Asian immigrants at the top, native blacks on the bottom, and Latino immigrants in between. This has the unfortunate effect of undermining America's commitment to improving the condition of blacks.

Not Enough Jobs

Even among those who claim that positive economic effects flow from immigration, or who downplay the negative effects, there is a general consensus that earlier-arriving immigrants are hurt by the addition of later-arriving immigrants. George Borjas of Harvard University, for instance, has found that a 10 percent increase in immigration populations reduces immigrant wages by 10 percent (a staggering statistic, in view of the fact that the 1990 Immigration Act increased yearly immigration quotas by 40 percent). One major factor is that poor English skills force immigrants who had been professionals in their native lands to seek low-skilled jobs in immigrant communities. This swells the number of workers seeking such jobs, depressing wages.

While economic theory would suggest that immigration would slow as wages fall, other forces are apparently at work: would-be immigrants often do not know of such wage trends; they may regard even weakened employment opportunities here as superior to those back home; and they may believe that opportunities are better here for their children.

The complex econometric models that statisticians use to understand such trends are helpful; they are, however, no substitute for direct observation. For example, Po Wong, director of the Chinese Newcomers Service Center in San Francisco, told National Public Radio that, of the 11,000 new arrivals who tried to find work through his agency, only 2 percent were successfully placed. More recently, he told journalist Sanford Ungar,

I don't think our community is equipped to welcome this large a number. It is especially difficult to find employment for those who speak only Chinese, who have very little education, or who have never acquired a skill to compete in this new market. It's very depressing to see so many people come here looking for work.

The pattern is the same in Latino immigrant communities. When asked why most Latino Americans wish to see reduced immigration, Antonia Hernandez, president of the Mexican American Legal Defense and Educational Fund, explained that

...migration, legal and undocumented, does have an impact on our economy...(in) competition within the Latino community...There is an issue of wage depression, as in the garment industry, which is predominantly immigrant, of keeping wages down because of the flow of traffic of people.

The competition for jobs among Latinos was illustrated in a rather dramatic manner in a recent article, ``Immigrants Split Over Job Scarcity: Legal Residents in Marin Tell INS About Illegals," in the San Francisco Chronicle. It begins, ``A shortage of jobs is provoking cutthroat rivalry among immigrant day laborers in San Rafael's Canal Area, where some (legal immigrants) are getting ahead byturning in their undocumented peers to the INS."

Those immigrants who bring some savings with them to the United States may seem on the surface to be better off than day laborers, as they can start businesses. Indeed, entrepreneurship is taken by many pro-immigration analysts to be a major selling point for high-immigration policies. Yet the entrepreneurs suffer the same problems as the workers do. Not being familiar with American culture and business, the would-be immigrant entrepreneur will typically start the same kind of business as his friends and relatives have--thus creating an oversupply of businesses of that type. Sociologist Peter Kwong of Hunter College has written that ``in the 1980s, business in (New York's) Chinatown reached the point of saturation: too many immigrants, too many new businesses, and exorbitant rents. Suicidal competition developed throughout the community." In 1995, Sung Soo Kim, president of the Korean-American Small Business Service Center in New York, noted that ``we are in the middle of a tragedy. Last year, we had 700 stores open but 900 close. Growth has completely stopped."

The current yearly rates of immigration are higher than the immigrant communities themselves can stand. Sadly, those who profess support for immigrants are often the least concerned about this point. Ethnic political organizations oppose reductions in yearly immigration quotas, in spite of the support for such reductions among their putative constituents. (Few, if any, of these organizations have large memberships among the groups that they claim to represent.)

Faulty Econometrics

Evidence suggests that the surplus of labor then spills out of the immigrant-enclave economies into the surrounding nonimmigrant economy. Since the immigrants tend to cluster in urban areas that contain large numbers of blacks, the latter population is particularly hurt. In regions with high immigration levels, low-skilled jobs in hotels, restaurants, airports, and so on--which could be held by African Americans (and often used to be)--are now typically held by immigrants. For example, the traveler at the San Francisco International Airport will see a workforce dominated by immigrants. By contrast, at the Baltimore and Philadelphia airports, areas in which there are many fewer immigrants, one sees many black employees, suggesting that the claim that blacks refuse to do such work is wrong.

In fact, a now-famous study by the General Accounting Office found that employers in Los Angeles had deliberately and systematically fired black janitors and replaced them with lower-paid immigrants. In most cases, however, the effects are less direct. When jobs become open through attrition or company expansion, the potential is produced for increases in black participation in a particular market--but that potential is not realized, because immigrants are often hired instead.

Econometric models may not detect this preference for immigrants over blacks. They may report only a small reduction in black employment in the market, missing the fact that such employment should have increased. Though economic studies performed entirely within immigrant labor markets are more clear-cut, econometric models are often powerless to detect effects outside the immigrant communities and, in many cases, will actually produce potentially misleading results. Data-oriented immigration analysts concede, indeed complain, that full data on immigration are unavailable.

One problem in such models is that the analysts do not know what data to use. Consider a pair of variables, X and Y, whose relationship we wish to study. One of the most frequent errors is to omit from the equation an additional explanatory variable, W, known as a covariate. That is, we predict Y from X alone when we should be expressing Y as a function of both X and W (or of X, W, and several other covariates). Model misspecification of this sort could result in wrong conclusions about relations. For example, some studies of multiple cities have found only very weak negative correlations between immigration (X) and employment (Y) rates. But William Frey of the University of Michigan has found that many natives are fleeing the high-immigration urban regions, often due to poor employment opportunities. If the very people whose job displacement we wish to detect have left town, failure to take into account this ``W" could lead to false optimism about immigration's impact.

Moreover, immigrants are naturally attracted to cities with growing economies. This phenomenon, holding everything else constant, would produce an apparent positive correlation between immigration and employment, but it would be wrong to conclude, merely from such a statistic, that the immigrants were causing the growth-when in reality the growth may have been causing the immigration. The only way to resolve such a causation question would again be to examine covariates, in this case variables that measure fundamental growth factors, like infrastructure quality and business policies.

Blue-Collar Competition

But why do employers apparently prefer to hire immigrants over African Americans? First, because of networked hiring, employing immigrants is more convenient. Employers have little incentive to seek out black applicants for job openings when their immigrant employees are happy to provide their bosses with job applications from the immigrants' numerous friends and relatives. As a result, according to Richard Rothstein, a labor organizer and columnist for the Spanish-language La Opinion, ``In the garment districts of Los Angeles, New York, or Miami, entire plants are staffed by immigrants from the same village in Mexico, El Salvador or China." Significantly, Rothstein adds, ``once such powerful networks are established, policy is impotent to break them."

Second, many employers think that immigrants are better and more reliable workers than blacks. Latino workers in Los Angeles, according to Peter Skerry of the Brookings Institution, characteristically use carpools to get to work, whereas blacks do not. Thus employers conclude that a black worker might not show up for work if his car breaks down. One might ask, ``Why blame the immigrants? Why can't blacks form networks and use carpools?" The answer is probably cultural; but, in any case, the availability of immigrant labor gives employers no reason to develop such work practices and skills among poor blacks. Much has been made of immigrant entrepreneurship, with the implicit claim that this provides jobs for natives. Immigrant entrepreneurs, however, tend to hire other immigrants, often members of their own families. Further, the replacement of native businesses by immigrant enterprises often produces an erosion of the tax base, according to sociologist Timothy Fong's analysis of Chinese immigration in southern California. Typically, a large, formerly non-immigrant establishment, such as an automobile dealership, is carved into several small low-profit immigrant shops whose collective sales taxes are much smaller than those produced by the original business.

It is claimed, however, that, in spite of problems of job displacement, immigration produces a net gain for natives when the job-creation power of immigrant consumerism is factored in. There is no doubt that immigration increases the absolute number of jobs, but it is easily demonstrated that the net change in jobs available to natives is negative. On a per-capita basis, immigrants are poorer than natives. Being poorer, the immigrants have lower levels of consumerism and thus create fewer jobs, relative to natives. Yet immigrants have the same or higher levels of workforce participation per capita as natives. In other words, immigrants do not generate as many new jobs as they hold--a net job loss for natives.

Such an argument is, of course, useful only as a first approximation. Yet more detailed inspection reveals the situation to be even worse. For example, let us break down immigrant job-creation effects according to skill level. Some high-skilled jobs might be created for the middle class, but whatever low-skilled jobs are created will for the most part be taken by the immigrants themselves. A glance at employment patterns in the low-skilled sectors of the labor markets in any city with a large immigrant population shows that there are virtually no low-skilled occupational categories in which immigrants have not become dominant.

White-Collar Blues

Though the focus of many analyses is on low-skilled jobs, it should be noted that there are some adverse effects in professional areas as well. Black teachers and social workers in Los Angeles are being squeezed out by immigrants, who are often not only bilingual but also identify culturally with their immigrant clientele.

Another example of adverse impacts on minority access to professional jobs involves the computer industry. Because employers hire many foreign nationals and sponsor them for immigration, many Asian-American university students who major in computer science face reduced employment opportunities, either just after graduation or later in their careers. Of the hundreds of computer companies in the San Francisco Bay Area, only two (Oracle and Sybase) recruit graduates from San Francisco State University, a predominantly Asian-American school. At UC Davis, where over one-half of the students in computer science are Asian American, the situation is better, but companies such as Microsoft only recently started recruiting graduates from there. Many new graduates must settle for semi-technical jobs, such as customer interface or system administrator.

It appears that foreign nationals are cheaper to hire. A statistical analysis I performed on the 1990 Census data revealed that average salaries for foreign-born computer professionals in Silicon Valley were nearly $7,000 lower than among natives of the same age and level of education.

footnote: A recent study by the pro-immigration Cato Institute found that foreign Ph.D.s were paid more than their native counterparts. However, it did not adjust for important covariates, such as geography and job type. There is a very substantial regional variation in wages in the computer field, and immigrant professionals tend to seek work in high-cost-of-living cities. Immigrant Ph.D.s tend to prefer work in the higher-paying industrial sector, rather than in academia. An analysis by UCLA Asian-American studies professor Paul Ong found that foreign-born engineers were paid up to one-third less than natives.

Finally, one of the most important items in the immigration debate is the question of whether immigration produces a net fiscal gain or loss. Numerous clever accounting schemes have been advanced, claiming to answer this question in one direction or the other. But the problem is that, due to methodological difficulties, many government expenditures are not included in such studies. This is not a problem for those studies that claim a fiscal loss, but clearly it renders suspect the studies that claim a fiscal gain.

A simple and direct analysis explains why. For the moment, consider only welfare services. Borjas recently found that an alarming 40 percent of all welfare dollars in California are going to immigrant households. The Urban Institute has found that the immigrants now entering the United States (not just California) use more welfare per capita than natives and have a lower average income than natives. The latter point presumably implies that immigrants are paying less in taxes per capita than natives. In other words, compared to natives, immigrants are paying less in taxes and using more in welfare. This indicates that there is a fiscal loss for natives. The two other main service categories in which there is a disparity in usage rates between immigrants and natives are education and Social Security/Medicare. Immigrants on average have more children than do natives; so on the educational front, we again see a fiscal loss for the natives. Though immigrants are on average younger than natives, and thus currently use less in Social Security/Medicare, they too will age and eventually use these programs. On the whole, therefore, the fact that immigrants are on average poorer than natives strongly suggests an overall net fiscal loss.

An American Caste System

Aside from the economic effects of immigration, there are political consequence too. In the past, African Americans could count on a certain degree of sympathetic attention from government and the media. They now must compete with the immigrant-dominant minorities (IDMs) for such attention. More often than not, the interests of IDMs and blacks conflict. In California, the largest immigrant-receiving state, African Americans are on the road to becoming what I term the Forgotten Minority.

Consider welfare, one area in particular where immigration harms African Americans. The public's backlash against welfare is owing to increased welfare costs driven in part by increases in IDMs who receive welfare. Elderly Chinese, Korean, and other Asian immigrants all receive welfare disproportionate to their numbers. This has, in part, provided motivation for congressional welfare-reform proposals in 1994 and 1995. After heavy pressure from Asian advocacy groups, the proposals were weakened to the point that they did virtually nothing to address the problem of immigrants on welfare. The revised bills (still under negotiation, as of March 1996

footnote: See attached update.
) would force a functionally illiterate African-American single parent off welfare while allowing an immigrant couple, with a combined income of well over $100,000, to keep their parents on welfare (providing they become naturalized citizens), in spite of the couple's earlier promise, made as a condition for their parents' immigration, to support them. This inequity is not deliberate, but it does exemplify the rise in power of the various immigrant advocacy groups, and the dwindling influence of black advocates.

Or consider their muscle flexing on the local level. From 1980 to 1993 in San Francisco, according to the San Francisco Examiner, the shares of civil-service jobs held by Asians and Latinos increased by 60 percent and 80 percent, respectively, while the share for blacks decreased 80 percent. And then from 1990 to 1993, the volume of city contracts to Asian businesses increased at a rate that was 2.9 times the rate for blacks.

Moreover, the IDM-black competition is becoming more heated, as one may see from an article by UC Berkeley Asian-American Studies professor Elaine Kim, a Korean-American [this was actually Prof. Kim's writeup of a narrative by Korean-American activist Bong Hwan Kim, not Prof. Kim herself--NM]:

Someone from MALDEF (Mexican-American Legal Defense and Educational Fund) pulled me aside during a meeting about redistricting to point out that due to demographic shifts, Asians and Latinos could work together because we never had very much to begin with. ``We have little to lose and a lot to gain by working together," he said ``while African-Americans stand to lose their hard-won civil rights gains from the 1960s, given their declining numbers."

Soon afterward, in nearby Oakland, Latinos and Asians successfully proposed a municipal redistricting plan that challenged the one proposed by blacks. Given the commonality of goals of the IDM political activists on some issues, such as bilingual education, one may reasonably project that alliances of IDMs against blacks will proliferate in the future. And though it is true that on some issues (notably affirmative action) Latino activists will side with blacks against Asians, it is clear that immigration is increasing political conflict between IDMs and blacks.

More Is Not Always Better

Handled in a sensible manner, immigration adds vigor to our society, broadens our world view, and simply makes America a more interesting place. But our current policy is an illogical hodgepodge in need of reform. Immigration is not the sole root of our problems, but it is certainly exacerbating them and, worse, making them increasingly difficult to solve. Like anything else, it is harmful to have too much of a good thing. Certainly policy changes are in order. Family-reunification laws, the heart of our immigration policy, typically do not serve the purpose implied by their noble-sounding name. ``Family reunification" often brings about chain migration, in which later links in the chain have never even met the earlier links, much less have a desire to ``reunify" with them. Bill Ong Hing, a Stanford University law professor and prominent immigration advocate, concedes that jobs, not family ties, are the greatest attraction to immigrants. This leads to much larger influxes of immigrants than can be handled by the communities they join, or by the neighboring black communities. Family-based immigration should be restricted to the nuclear family.

From the point of view of impacts on minorities it would be tempting to shift to a more skills-based immigration policy, but the fact is that, with corporate ``downsizing" and the reduction in the defense industry, there is no shortage of skilled labor in the United States today. One of the most powerful lobbyists for immigration, Sun Microsystems, a computer manufacturer, says that it needs to scour the world for the best and the brightest; but they actually seem more interested in hiring the cheapest. Sun boasted in 1993 to the Los Angeles Times that it had just hired 50 Russian programmers ``at bargain prices."

We should encourage and facilitate the immigration of exceptional talents, but they comprise only a small fraction of those sponsored for immigration by computer-industry employers. Of 56 awards given for industrial innovation by the Association for Computing Machinery since 1980, only one recipient was an immigrant. Of 115 U.S. recipients of computer-related awards given by the Institute of Electrical and Electronic Engineers since 1976, only nine have been to immigrants.

But the bottom line is that the significant problems associated with today's high yearly levels of immigration can only be solved by reducing those levels. In 1990, Congress, without seriously looking into the matter, and mainly in response to pressure groups, decided to increase yearly immigration quotas by 40 percent. Surely it would be reasonable to roll back current levels at least to their pre-1990 levels. As a nation that prides itself on its efforts to foster good race relations, the least we can do for native minorities is to formulate an immigration policy that follows the physician's cardinal rule: ``First, do no harm."

Update to ``How Immigration Harms Minorities''

Norman Matloff

October 4, 1996

Following are updates to some of the points made in the article.


The article had noted that the welfare-reform legislation, in the form pending March 1996, would do little to curb skyrocketing levels of welfare use by elderly immigrants. The provisions I was referring to at that time were essentially unchanged later, and thus remained in the legislation signed by President Clinton in August 1996.

The old welfare policy had essentially barred immigrants from welfare for five years. The new law, enacted in August 1996, imposes such a bar until the immigrant naturalizes. Since one can naturalize after five years, for most immigrants the new law merely replaces one five-year ban by another, thus no net change. This was a victory for the Asian advocacy groups, who had successfully fought for the withdrawal of a provision (in a separate bill) which would have extended an immigrant's sponsor's financial responsibility past the time the immigrant naturalizes.

(The press noted a big surge in applications for naturalization resulting from Clinton's signing of the welfare bill. However, that surge had begun in early 1994, in response to the bill's original introduction in late 1993.)

Economic Conditions for Immigrant-Dominant Minorities:

The article had noted that the immigrant communities themselves are among the biggest victims of the increasing influx of immigrants into those communities, causing wages to fall and making steady work harder to find. This trend has continued.

A report released in October 1996 by a coalition of 55 Asian American community agencies in San Francisco found that the proportion of Asian American children living in poverty rose from 20% in 1980 to 33% in 1990, a period of very high immigration growth. The report specifically cited as a cause of this increase the large number of immigrants whose low educational levels and lack of English proficiency limits them to (ever-lower) low-wage jobs in their ethnic communities.

In September 1996, during a hearing held by the San Francisco County Board of Supervisors, Chinese-American union leader Katy Quan testified that during the next few years there will be approximately twice as many immigrant seamstresses available for work as there are jobs in that field.

Political Conflict Between Native and Immigrant-Dominant Minorities:

In August of 1996, an Asian/Latino coalition filed a lawsuit against the city of Oakland, claiming city contracts are awarded to African-American firms to the detriment of the coalition's constituents. As noted in my article, two years ago Asian and Latino activists succeeded in pushing through an Oakland redistricting plan which increased their power and reduced that of the African-Americans.

The interested reader may wish to peruse my World Wide Web site on immigration issues, at

Norm Matloff
Tue Oct 15 22:01:34 PDT 1996