Program In Comparative International Development Department of Sociology Johns Hopkins University Baltimore, MD. 21218 USA Working Paper #8 DIVERGENT DESTINIES: IMMIGRATION, POVERTY, AND ENTREPRENEURSHIP IN THE UNITED STATES Alejandro Portes and Min Zhou April 1992 Written for the project on Poverty, Inequality, and the Crisis of Social Policy, Joint Center for Political and Economic Studies, Washington DC. Contemporary immigration to the United States is seldom associated with the traumas of poverty in the specialized literature. This is not due to the absence of objective indicators of poverty, such as low incomes, among recent immigrants, but to the low incidence of those social pathologies commonly associated with the domestic "underclass" (Marks 1991). Instead, the areas where the foreign born concentrate are characterized by a different set of social traits including the widespread use of languages other than English, great diversity in terms of both national origins and socioeconomic backgrounds, the large and semiopen presence of unauthorized aliens, and the rise of ethnic business enclaves. The research literature on immigration has focused on these and related topics in contrast to those that most concern students of domestic poverty. Indeed when the two literatures come together, it is in the form of comparisons that point to the relative progress of immigrants relative to native-born minorities despite initial disadvantages, such as lack of English knowledge and unfamiliarity with the host culture. These comparisons are notable because immigrants must often endure conditions every bit as harsh, if not worse than the domestic poor. The difference is that immigrants' poverty is embedded in a social context that makes it appear less hopeless and more transitory than the kind afflicting the native born. We review some of the reasons for this contrast below following an overview of the magnitude and characteristics of contemporary immigration. I. Immigration Today In 1990, the United Stated admitted 1,536,483 foreigners for legal permanent residence. This is the single largest annual total in the twentieth century, exceeding by 240,000 the preceding record year of 1907 (INS 1991). The 1990 figure is an inflated indicator of new arrivals, however, since it is dominated by legalizations of formerly unauthorized aliens, as mandated by the 1986 Immigration Reform and control Act (IRCA). A total of 880,372 foreigners already residing in the country legalized their residence in 1990. Added to those who did the same in 1989, the total number of admissions of former illegals under the IRCA regular legalization program reached 1.35 million of which over 70 percent came from Mexico and another 10 percent from Central America. Even after discounting the new legalizations, immigration has continued to climb steadily reaching 656,111 in 1990 or a 15 percent increase over the figure five years earlier. For the decade, and discounting legalizations, legal immigration added almost 6 million newcomers to the U.S. population, a total second only to the 1901-10 decade (INS 1991). To this figure must be added the unauthorized immigrant population. Although the legalization program of IRCA reduced its numbers, the deterrents built into this law for preventing new illegal entries have proven ineffective, leading to a rebound in the unauthorized inflow. Apprehension of illegal aliens did decline sharply following passage of IRCA in 1986, but then turned upwards again reaching 1.2 million in 1990. Although apprehensions are only an imperfect proxy for new entries, they serve as a rough indicator that the unauthorized movement continues. Adding to the newly legalized, a conservative estimate of the remaining unauthorized population in the U.S. furnished by Census Bureau demographers (Woodrow et al. 1987), we would have a total of 3 to 5 million additional immigrants that arrived and settled clandestinely in the country. The immigration literature has repeatedly noted the geographic concentration of the present U.S.-bound inflow, both in terms of countries of origin and areas of destination. As seen in Table 1 only a handful of countries accounted for the bulk of legal, unauthorized, and refugee migrations during the last decade. Only 12 nations out of the more than 100 sending immigrants to the United States accounted for 62 percent of regular admissions for permanent residence; only 10 countries for 90 percent of applications for legalization under IRCA; only 5 countries for 96 percent of new apprehensions of unauthorized arrivals in 1990; and only 6 nations for 80 percent of all refugees during the decade. ------------------------------------------------------ Table 1 about here ------------------------------------------------------ A more careful perusal of the figures in Table 1 brings to light two additional findings. First, immigration and, in particular unauthorized immigration to the United States is, to a large extent a Mexican phenomenon. That country alone accounted for over one- fifth of legal immigrants, three-fourths of legalization applicants, and over 90 percent of apprehended aliens during the last decade. Second, every country sending large numbers of immigrants or refugees has been closely entangled with the United States at some point during its recent history, each having been subject to extensive North American political and economic influence during that period. Although this is not the place to develop this argument in detail, a brief review of the history of the top five countries sending legal immigrants (Mexico, the Philippines, China/Taiwan, South Korea, and Vietnam); the top two sending unauthorized immigrants (Mexico and El Salvador); and the four main sources of refugees (Vietnam, Laos, Cambodia, and Cuba) will support this conclusion. Just as the geographical origins of contemporary immigration are highly concentrated, so are its areas of destination. As during the turn of the century, immigrants today are overwhelmingly urban-bound, going to key metropolitan areas in a few states. As shown in Table 2, the states and cities of destination of present-day immigrants differ, however, from the predominantly Northeastern destinations of turn of the century arrivals. California is by far the preferred state and, within it, Los Angeles-Long Beach is overwhelmingly the most favored metropolis. In 1990, the 24 percent of legal immigrants intending to reside in Los Angeles exceeded the number going to any state, except California itself. Los Angeles also received more immigrants during that year than the next four most popular metropolitan destination combined. Although the predominance of Los Angeles is somewhat exaggerated in 1990 because of the residential preferences of the newly legalized, figures for this year are representative of a trend observed during the entire decade. What IRCA legalization did was to bring into the open the absolute dominance of Southern California and its displacement of New York as the prime site of destination of contemporary immigration, both legal and unauthorized. ------------------------------------------------------ Table 2 about here ------------------------------------------------------ As seen in Table 2, just six states, including California, accounted for three-fourths of immigrant destinations during the 1980s; just twelve cities, including Los Angeles, represented 60 percent. Past studies also show that individual immigrant nationalities tend to select different places of destination, giving to each receiving city a different ethnic mix. This is illustrated in Figure 1 that presents the composition of inflows going to the six major metropolitan destinations in 1987, the year before the IRCA legalization program started. Los Angeles-bound immigration is dominated by Mexican, Filipinos, and Koreans; New York receives primarily flows from the Caribbean and so does Miami where Cubans are dominant. Appropriately, given its mid- country location, Chicago is the most diverse of major destinations receiving large inflows from three continents. ------------------------------------------------------ Figure 1 about here ------------------------------------------------------ By 1990, the foreign born population of the United States stood at approximately 18 million or 7 percent of the total. This is the largest absolute number of immigrants living in the United States ever. The previous peak of 14 million was recorded in the 1930 census and was surpassed sometime in the 1980s. As a percentage of the total population, however, contemporary immigration does not even come close to the peak years of the 1890s when 15 percent of 1 in every 7 persons living in the country was foreign born (Fix and Passel 1991). The surge of immigration during the last decade has produced two consequences of note. First, it has led to a doubling of the Asian population to more than 7 million in 1990 and to a 53 percent increase in the number of Hispanics to 22.4 million (Rumbaut 1992). Neither group even existed as a statistical category before 1965. Second, immigration prevented serious declines in the population of some of the states where it has concentrated, while contributing to the extraordinary growth of others. An estimated one million immigrants to New York State helped compensate for its 1.7 loss of people during the decade. Similarly, Illinois' 1 million population loss was partially compensated by a net immigration of 400,000. At the other end, rapid growth in Florida and California was fueled, in part, by the 439,000 immigrants going to the first state and 2.33 million to the second (Fix and Passel 1991). Hence, while leaving most American states and cities largely untouched, immigration has had a major impact on those areas where it concentrates. This effect is not only demographic, but economic and political as well. In combination with the 1990 Voting Rights Act that mandated better representation of minorities in the electorate, the recent inflow will have, for example, a significant effect in the reapportionment of electoral districts. Already in Miami, where Cubans have been joined by significant Nicaraguan and South American inflows, Cuban- American state representatives are expected to increase to eleven from the present seven, or more than half of Dade County's delegation; Cuban-American state senators are likely to double to four and congressional representatives to two (Fiedler 1991). The economic impact will be as significant although its character will vary with the diverse origin and incorporation of immigrants, as explained below. II. Modes of Incorporation Figure 2 presents a typology of the patterns of entry used by contemporary immigrants to come to the United States and seek legal permanent residence. The diversity of these channels is correlated with the future economic performance of each national inflow and its insertion into American society. In general, regular immigrants tend to come from higher socioeconomic backgrounds. This is especially true of those nationalities that make heavy use of the occupational preferences of the 1965 Immigration Act. These groups come mostly from Asia, Africa, and South America. Refugees are also screened abroad prior to entry although their socioeconomic backgrounds vary greatly. The initial refugee waves from such countries as Cuba and Vietnam contained high proportions of well-educated, formerly well-off individuals from the elite and middle class of their respective countries. Subsequent waves came from increasingly more modest backgrounds (Portes and Bach 1985: Ch. 3; Gold 1992). Surreptitious entrants are predominantly former urban laborers and rural workers and hence tend to have lower educational levels. Those coming from Central America and Haiti are often escapees from political turmoil at home, although their requests for political asylum have seldom met a sympathetic ear by U.S. immigration authorities. Mexicans represent the overwhelming majority of surreptitious entrants. They seldom claim political asylum but, as seen above, have been the prime beneficiaries of the IRCA legalization program. Finally, visa overstayers is a socioeconomically diverse category that includes many immigrants from modest backgrounds, but also skilled workers, artisans, and even university-trained professionals. ------------------------------------------------------ Figure 2 about here ------------------------------------------------------ This diversity of origins underlies the wide range of educational credentials among contemporary immigrants. Education is commonly used as an indicator of the "human capital" brought by the foreign born to their new country. Table 3 compares the educational composition of the immigrant population of the United States and selected nationalities with the U.S. averages on two dimensions: percent completing high school and percent completing 4 years or more of college. The foreign-born population as a whole has 13 percent less high school graduates than the U.S. total population, but the number of college graduates is about the same. As shown in the table, a number of nationalities exceed the U.S. averages, often by a considerable margin. Notable in this regard, are the educational profiles of Iranians, Egyptians, Taiwanese, and Nigerians. ------------------------------------------------------ Table 3 about here ------------------------------------------------------ Most highly educated nationalities come from Asia and Africa and, with the exception of Iranians who often arrive as political refugees, they are regular immigrants. The same is true for all other groups above the U.S. educational averages. Noteworthy as well is the fact that some of the most highly educated immigrants are recent arrivals. This is the case of Iranians, 72 percent of whom arrived during the 5 years preceding the 1980 census and of Taiwanese and Nigerians over half of whom came during the same period. At the other extreme are groups formed by refugees and former refugees, such as Laotians. Low average education is also associated with groups whose mode of entry is often surreptitious such as Mexicans, Guatemalans, and Salvadoran. As seen above, some immigrants from these countries manage to regularize their status as asylees or permanent residents and, during the 1980s, were the primary beneficiaries of the IRCA legalization program. Given their modest backgrounds, formerly unauthorized immigrants tend to depress reported educational levels for their respective nationalities. The same is true of Dominicans many of whom came by crossing surreptitiously the Mona Passage into Puerto Rico or were visa overstayers. The remaining low-education groups are older immigrants, such as the Italians and the Portuguese, the majority of whom came in earlier migration periods. Diversity among contemporary immigrants is not limited to their education and modes of entry, but also includes the process through which they insert themselves in the host community and labor market -- that is, their modes of incorporation. This concept encompasses three different levels of reception: first there is the government stance toward different national inflows. In the United States some groups, such as refugees, are granted special resettlement assistance; others are granted legal entry and have access to the same general social programs available to the native born; still others are actively persecuted, their claims for asylum are routinely denied and, if legalized, do so under an inferior legal status that limits their entitlements for extensive periods. A second reception level is that involving civil society and public opinion. For a variety of historical reasons, a few immigrant minorities have been greeted with open arms in America; others have arrived with little fanfare and their presence has been mostly a matter of public indifference; still others have been highly impopular and their arrival actively resisted in their places of settlement. Haitian "boat people" and Mariel refugees arriving in South Florida during the early 1980s are prime examples of the latter situation (Portes and Stepick 1985). Refugees from the 1956 Russian invasion of Hungary and the first (1960-62) exiles from Castro's Revolution is Cuba are examples of the first. Barring such specific historical circumstances, white and English-speaking immigrants tend to experience the least amount of resistance from the American mainstream, while those who are phenotypically or culturally distinct endure much greater social distance and discrimination (Marger 1991: 126-129). This second dimension of public acceptance, indifference, or rejection is not necessarily dependent on the first, or governmental reception. For instance, the large number of unauthorized Irish immigrants in Boston have generally been accorded a favorable reception, despite their illegal status. By contrast, perfectly legal Iranian and Ethiopian refugees and Jamaican immigrants have endured much greater discrimination (Tumulty 1989; Cichon et al. 1986). The third reception level is that of the ethnic community itself. Some immigrants belong to nationalities too small to form distinct communities and hence find themselves dispersed among the native born. Others join communities composed primarily of immigrant workers. These communities offer some comfort against outside prejudice and the shock of acculturation commonly experienced by new arrivals. However, even if sizable, such communities provide few opportunities for upward mobility since the assistance that they can give to newcomers seldom goes further than finding them a manual, low-paid job. Still other immigrants are lucky enough to join communities where their conationals have managed to create a substantial entrepreneurial presence. These communities, referred in the literature as ethnic enclaves, are characterized by a large number of interlinked small and medium- sized firms. Enclaves tend to offer new arrivals an option to dead-end jobs on the low-paid or secondary labor market and better opportunities for economic ascent than those available to immigrants of comparable human capital who join working- class communities (Wilson and Martin 1982; Stepick 1989; Zhou 1992). The combination of these three levels is the mode of incorporation of a particular immigrant group to American society. Assuming, for illustration, that positive official reception is defined by active government assistance and that societal reception is determined by the phenotypical/cultural characteristics of each immigrant nationality, it is possible to cross-classify these dimensions in a typology of modes of incorporation. This is done in Figure 3. Ethnic community, in this figure, is classified according to the size of a particular nationality and their under or overrepresentation in small entrepreneurship, as indicated by average self-employment rates. Small groups lacking any visible concentration in the U.S. are defined as having no community; those that have sizable concentrations, but where business owners are underrepresented are classified as working class communities. ------------------------------------------------------ Figure 3 about here ------------------------------------------------------ Cell entries in the Figure illustrate the context of reception encountered by new immigrants of a particular nationality in 1980. The absolute number of persons from that group residing in the U.S. in 1980 are in parentheses, followed by median household incomes in 1979 and percentage of adults who completed four years or more of college. Illustrative of the significance of contexts of reception are Ethiopians and Ghanians -- both small nonwhite groups of very recent arrival. Despite their very high levels of education and some governmental assistance to Ethiopians as political refugees, their economic situation in 1980 was poor with incomes much below the national average. These cases can be fruitfully contrasted with those of Argentines, Cubans, and Jamaicans. Argentines are also a recently arrived group lacking any significant concentration anywhere in the country. Being white and of European extraction, Argentines mix easily, however, with the native white population, avoiding the discrimination meted on African immigrants. Neither Cubans nor Jamaicans are particularly notable for their high levels of education and most Jamaicans are black, yet both groups possess sizable and entrepreneurially oriented communities in their respective principal places of destination. Cubans, mostly in Miami, and Jamaicans in New York commonly manage to turn their human capital into business pursuits by taking advantage of the favorable context created by earlier arrivals (Portes and Bach 1985: Ch. 6; Marshall 1983). This advantage is reflected in their higher incomes, both in absolute terms and in relation to average human capital. Thus, while each percent increase in college graduates was associated with a $324 increase in 1979 median incomes for Ghanians and a still lower $267 for Ethiopians, the figure for Cubans was $1,014 and for Jamaicans, $1,390. III. Income and Poverty Table 4 presents individual and family incomes for the immigrant and native-born population of the U.S. in 1979, the last year for which a complete census enumeration is available. Figures in this table suggest reasons why "immigration" and "poverty" are not closely associated concepts in the research literature. Immigrant incomes are below those of the native born, but not by a very large margin. European-born men had greater earnings than American men in 1979 and Asian families topped the average American family, a result of the greater propensity of Asian immigrants to pool incomes. Only Latin American immigrants fell significantly below on both measures, a result to which we will return below. ------------------------------------------------------ Table 4 about here ------------------------------------------------------ Diversity in modes of entry and incorporation is well reflected in the wide disparity in average incomes among immigrant nationalities. Table 4 presents median household incomes for: 1) the twelve wealthiest foreign- born groups with a population of at least 20,000; 2) the twelve groups of at least this size with most similar incomes to the U.S. average; and 3) the eleven poorest nationalities of at least 20,000. Together the groups enumerated represent 66 percent of the total immigrant population. In 1979, Asian Indians were the wealthiest immigrant group in the United States, followed by Filipinos. Seventy percent of the highest income nationalities came from Asia and Africa. Only one Latin American (Argentines) and one European (Dutch) groups are represented in this category. With the exception of some Chinese who arrived as refugees, all these groups are formed by regular immigrants. They all have high levels of human capital and, in the case of the Chinese and the South Koreans, they have developed dynamic entrepreneurial communities in different parts of the country The same is true for Cubans, a group composed mostly of former refugees that tops the mid-income category. The remaining groups in this category are mostly regular immigrants from very diverse geographic backgrounds -- Europe, Canada, East Asia, and South America. Also included are Jamaicans, the only black group of any size to reach household incomes comparable to the U.S. average. As noted above, Jamaicans, are known for their entrepreneurial orientation in their principal places of destination. At the bottom of the income distribution are an array of nationalities topped by the largest immigrant group, Mexicans, followed by Vietnamese and Salvadorans. Irregular modes of entry distinguish these groups, including surreptitious entrants, visa overstayers, and would-be asylees. They also include, however, a number of regular immigrants and regular refugees. The European groups included are either formed by an elderly population, survivors of earlier migration waves (Swedes, Finns) and/or include a sizable group of recent refugees (Soviets). Recent refugees are also common among other impoverished nationalities such as Iranians, Vietnamese, Laotians, and Cambodians. The economic situation of Iranians and Nigerians is noteworthy since, as seen in Table 3, they figure among the most highly educated immigrants. Their low incomes may be attributed to their recency in the country, outside discrimination, and small numbers that, by 1980 at least, precluded the emergence of a protective ethnic community. Studies conducted during the 1980s reported a rapid development of small firms among Iranians in Los Angeles (Light 1991) and Dominicans in New York (Portes and Guarnizo 1991). Given the high concentration of these groups in the respective cities and apparently rapid business growth, it is possible to expect an improvement in their economic position during the decade. No similar development has been observed among Mexican and Salvadoran immigrants or among Cambodian and Laotian refugees. These four nationalities comprise today the most sizable foreign groups among whom poverty is common and, for this reason, merit additional comment. Lower than average incomes among the Latin-origin immigrant population are a consequence of the enormous weight of Mexicans, and to a much lesser extent Salvadorans, in the regionwide averages. As seen in Table 4, immigrants from Argentina, Cuba, and other Latin American countries have attained or surpassed U.S. average household incomes. Mexican poverty in the United States is the result of the continuous inflow across the border of a mass of unauthorized workers of modest educational and occupational backgrounds. At present, a substantial proportion of the Mexican-origin population of the United States is composed of formerly unauthorized immigrants and their descendants. The same is true of Salvadorans, although they are primarily first generation immigrants. The latest Census data on the Spanish-origin population of the United States reflects the modest backgrounds of the Mexican origin groups and its consequent failure to approach economic parity. These data are presented in Table 5. The immigration-driven character of the Mexican origin group is reflected in its average youth and in its high level of labor force participation. College graduates among Mexican origin adults are only one-fourth of the U.S. average, while they are heavily overrepresented among those with less than an elementary school education. As a result, Mexicans concentrate in the bottom rungs of the labor market, their mean family income lags by more than $13,000 the national average, and they are more than twice as likely to be living in poverty. ------------------------------------------------------ Table 5 about here ------------------------------------------------------ Among Spanish-origin groups, only Puerto Ricans, mostly a mainland-born minority, does worse. Although the educational profile of Puerto Ricans is somewhat better, their levels of labor force participation are significantly lower while the relative numbers of female- headed and poor families are much higher. Central and South Americans are a foreign-born category that includes Salvadorans. As seen above, this nationality has similar modes of entry and similar class backgrounds as Mexicans, thus depressing the average figures. Even then, Central and South Americans as a whole do much better economically than either Mexicans or Puerto Ricans, a result owed to the inclusion in this category of better educated and more affluent immigrants from Argentina, Colombia, Chile, and other countries. Among Spanish-origin groups, only Cubans have reached parity with the U.S. population. As seen in Table 5, Cubans are much older on the average than other Latin groups and even than the total U.S. population. This increases their average work experience. Their education is also higher than other Spanish-origin minorities and they have been more entrepreneurially oriented. Results of this last pattern will be examined in greater detail below. The relative economic success of Cubans has been attributed to the considerable resettlement assistance that they received as refugees from the U.S. Federal government (Pedraza-Bailey 1985). However, other refugee groups have received even more generous official aid without its being translated in significant economic advancement. This is the case of Southeast Asian refugees, particularly Laotians and Cambodians. Table 6 presents data from a random sample of Southeast Asian refugees residing in San Diego in 1983- 84. The data show the extremely low levels of education among Hmong (from Laos), Khmer (Cambodians), and Chinese Vietnamese and the rural origins of the first two groups. These are predominantly peasant minorities displaced by the various regional wars during the 1960s and 1970s and whose marketable skills and contact with Western culture were very limited prior to departure. Hence, official assistance has focused primarily on teaching these refugees language and basic work skills and gradually introducing them to the American labor market. As shown in Table 6, in 1984 the process still had a long way to go: over 75 percent of Chinese Vietnamese and Khmer families were living below the poverty level and the figure among the Hmong was an astonishing 88 percent; labor force participation among these groups barely reached one-third of eligible adults. ------------------------------------------------------ Table 6 about here ------------------------------------------------------ There is evidence, however, that, with the passage of time, significant improvements have taken place in labor force participation as well as declines in the poverty rate among Southeast Asians (Gold 1992; Rumbaut 1990). Among the Vietnamese, in particular, there has been notable improvements coinciding with the rapid growth of small firms and the regrouping of this minority in certain areas, particularly in California. In 1979, the Census already detected rapid economic improvement among the Vietnamese who had been in the country ten years or more. While family poverty among Vietnamese immigrants who arrived during 1975-79 was a whopping 36 percent, it declined to only 9 percent among those who came during 1970-74 and 6.8 percent among pre-1970 arrivals (Gardner et al. 1985: 34). Hence it is possible to expect significant improvements in the economic condition of Southeast Asians when 1990 Census figures are released, although it is not likely that groups such as the Hmong and Cambodians have escaped poverty altogether. Haitians have not been mentioned so far because, in 1979, the Haitian-born population of the United States had average annual household incomes above the threshold of $13,000. Unauthorized Haitian migration, most of it coming by sea, accelerated during the 1980s transforming the Haitian immigrant profile. A random sample of 500 post-1980 Haitian arrivals interviewed in South Florida in 1983 yielded extraordinarily high rates of unemployment and poverty. Thirty-six percent of Haitian males and 81.4 percent of females were jobless. Only 3 percent of the would-be asylees had held a steady job since arrival. Median family income was a paltry $7,200 per year and 59 percent of respondents lived below the poverty level (Stepick and Portes 1986: 339-41). A follow-up conducted three years later indicated significant improvements in labor force participation, but still income levels were very low. With modest levels of education averaging only 4.6 years and very limited knowledge of English, Haitians lacked the human capital necessary to compete in the South Florida labor market. To this must be added their tenuous legal status and widespread discrimination for reasons of color and culture. A weak and resourceless ethnic community did not compensate for these barriers leading to one of the most precarious situations faced by any immigrant minority during the last decade. Along with Mexicans, Salvadoran, Laotians, and Cambodians, Haitians comprise today one of the poorest foreign-born groups and one of those for whom the future looks least enviable. IV. Immigrant Entrepreneurship Since Ivan Light (1972) noted the fact for the first time, specialists in immigration have emphasized the greater propensity for self-employment among the foreign born. Following Light's initial formulation, immigrant entrepreneurship was originally attributed to the discrimination faced by newcomers in the American labor market, forcing many of them to seek marginal niches for survival. This view of immigrant small business as a simple alternative to destitution is still predominant among many social scientists, despite recent evidence showing that self-employment can also serve as a vehicle for economic progress. First, the association between immigration and self- employment remains as prominent today as when Light first noticed it. Not all foreign-born groups are inclined toward entrepreneurship, but among those who do the rate of self-employment frequently exceeds twice the figure for the domestic population and three times that for native-born minorities. In 1980, the rate of self- employment per thousand population in the U.S. was 48.9; Korean, Japanese, Chinese, Cuban, Greek, Lebanese, and Asian Indian immigrants surpassed that figure by 20 percent or more (Fratoe and Meeks 1985). Second and more important is the relationship between entrepreneurship and earnings. There is evidence of a positive association at both individual and aggregate levels. Table 7 presents preliminary data illustrating the relationship between family incomes and self-employment among major ethnic groups. The correlation between self-employment and median income for these groups is both positive and highly significant (.53). ------------------------------------------------------ Table 7 about here ------------------------------------------------------ A similar positive relationship is found at the individual level, yet many scholars continue to insist that ethnic small enterprise is a refuge from labor market discrimination and poverty (Sanders and Nee 1987; Bates and Dunham 1991; Gordon 1992). Alternatively, it is argued that the positive association between self- employment and earnings is entirely accounted for by the higher human capital of the entrepreneurs. The latter conclusion raises the question of why should the best- endowed immigrants choose small business if the latter is, as depicted in the same writings, a mere alternative to destitution. Be it as it may, the issue is significant enough to deserve additional attention since its resolution has significant theoretical and policy implications. If the entrepreneurial route among immigrant minorities is a mere consequence of labor market discrimination, then the obvious course is to try to reduce the latter so as to allow immigrants to reap full returns for their individual human capital. On the other hand, if entrepreneurship yields significant economic payoffs then the alternative suggests itself of supplementing antidiscrimination measures with programs in support of small business development among ethnic minorities, including the most downtrodden. To examine this question, we conducted an individual level analysis of four of the most entrepreneurially oriented foreign origin groups, based on data from the 1980 Census. These groups are Cubans, Chinese, Japanese, and Koreans. The analysis is limited to males, aged 25 to 64 who worked at least 160 hours and earned a minimum of $500 in 1979. Both foreign-born and foreign-origin individuals are included, although the former comprise between 75 and 95 percent of all groups, except the Japanese. Table 8 presents socioeconomic and demographic characteristics for each group, classified by employment status. Without exception, average earnings of the self-employed exceed by a significant margin those of salaried workers. The difference ranges from $10,800 per year among Cubans to $12,500 among Koreans. Differences in annual median household incomes are more attenuated, but run consistently in the same direction. ------------------------------------------------------ Table 8 about here ------------------------------------------------------ A notable result is that differences in years of education between the self-employed and employees are trivial in most of these groups, contradicting the conclusion that earnings differentials are owed to the higher human capital of entrepreneurs. Differences in knowledge of English tend to favor the latter, but they are not large. The principal characteristics differentiating workers and entrepreneurs are marital status and length of U.S. residence. In all groups, the self-employed are much more likely to be married and much less likely to be recent immigrants. A final important difference is hours worked: in agreement with the literature on entrepreneurship, the self-employed consistently work more hours both per week and in the course of a year (Green and Pryde 1990). These preliminary tabulations do not suffice, however, to clarify whether the observed positive relationship between self-employment and earnings is entirely accounted for by other factors. To examine this question, we regressed average earnings on the set of independent variables associated with the human capital function plus self-employment. Also included are variables uniquely relevant to immigrants, such as knowledge of English and length of U.S. residence. Definition of variables and measures of central tendency are presented in the Appendix. Results for each group are presented in three columns: 1) the gross effect of self-employment; 2) the net effect of self-employment controlling for human capital variables and marital status; 3) the net effect of self-employment controlling for all these variables plus annual hours worked. Control for the latter variable is problematic since a higher work effort is a well-known and integral part of the entrepreneurial role. Hence, controlling for it statistically eliminates one of the defining characteristics of entrepreneurship. We include this variable, nevertheless, for the sake of comparability with earlier results. ------------------------------------------------------ Table 9 about here ------------------------------------------------------ The set of human capital predictors plus immigration-related variables do a good job in predicting annual earnings, accounting for between one-fourth and two-fifths of total variance in the dependent variable among the four groups. Effects of education, occupation, and work experience are all highly reliable and in the predicted direction. Since the dependent variable is regressed without logarithmic transformation, coefficients express the annual dollar increment or decrement of a unit change in each predictor. Although the analysis contains several interesting results, for our purposes the central one is the effect of self- employment. As seen in Table 9, the gross effect of this variable is very strong, reproducing the finding from the earlier tabulations. This effect is not entirely accounted for by differences in other predictors. The introduction of human capital and immigration- related variables cuts down the self-employment effect by between 19 and 33 percent. As expected, the addition of hours worked further reduces this coefficient to between 57 and 70 percent of its original value. This last result reflects again the intimate association between entrepreneurship and a higher work effort. In no instance, however, is the self-employment effect reduced to insignificance. Even after statistically equalizing the samples on 13 different predictors, entrepreneurs continue to have a significant advantage in earnings. This finding contradicts earlier dismissals of self- employment as irrelevant to the economic mobility of minorities or as a spurious consequence of human capital differences. We interpret it instead, along with results presented above, as signalling the significance of ethnic enterprise as an avenue for gaining the upper hand in the American economy, both collectively and individually. V. Conclusion: The Decade of Immigration Passage of the 1990 Immigration Act and the fading deterrent capacity of the 1986 IRCA indicate that the last ten years of the century will probably surpass the first ten as the "decade of immigration." Despite restrictionist voices, heard with increasing frequency during the 1980s, the overall sense of the U.S. Congress and Administration is that the effects of immigration on society and the economy are benign. In signing the inclusionary 1990 immigration bill, President Bush declared that it would be "good for families, good for business, good for crime fighting, and good for America." Indeed the bill appeared to respond to recent objections to immigration raised by different constituencies by increasing the number of immigrants said to be missing from the pre-1990 inflow without reducing earlier categories. Overall, the ceiling on legal immigration is raised so that it can increase from its pre-IRCA (1987) level of 600,000 to 700,000 between 1992-94 and 675,000 thereafter. To satisfy criticism that the "quality" of immigrants was not as high as it could be in terms of education and professional experience, the bill sharply increases admissions under occupational preferences from 58,000 to 140,000 per year and reserves the vast majority for skilled and professional immigrants. In practice, the number of such workers that will arrive under this preference is much less, because it also includes their spouses and children. In the past, only about 40 percent of occupational preference immigrants have been the actual workers (Fix and Passel 1991). To assuage complaints that the current inflow has become dominated by Asians and Latin Americans, the 1990 bill reserves 55,000 visas for a new "diversity" category designed to encourage immigration from other parts of the world. In practice, a large proportion of these visas will be absorbed, at least during the first years, by Irish immigrants many of whom already reside in the country illegally. The decline of Western European immigration during the last two decades has not been due to the dearth of visas, but to the rapidly improving economies of those countries, especially relative to that of the U.S. The availability of the new "diversity" visas is not likely to reverse this trend, although it may encourage new inflows from Ireland, the Eastern European nations, and the former Soviet Union. The law recognizes for the first time the category of immigrant entrepreneurs, reserving 10,000 visas for persons who invest $1 million or more and employ in their firms at least 10 workers. This floor is lowered to $500,000 for foreign investors who establish businesses in depressed areas, for whom 3,000 visas are set aside. Given past trends in entrepreneurial immigration, it is likely that this category will be dominated by Asians. Given past practice, it is not certain that these entrepreneurs would prefer to hire the unemployed native born over their own conationals and other immigrants. These changes in immigration law are not accomplished at the cost of other admission categories which are either preserved or increased. Admission of immediate relatives of U.S. citizens remains unlimited. There is a new "cap" on family related admissions of 520,000 per year in 1992-94 and 480,000 thereafter. Although the latter figure represents an increase of 45,000 over previously authorized levels, it is likely to be soon pierced because the law sets a floor of 226,000 for other relative admissions to prevent immediate kin of citizens from crowding them out. In addition, the prior limit of 25,000 per country per year is waived for immediate relatives of U.S. permanent residents. Given that most such visas will go to recent immigrants from Asia and Latin America, its effect will be to increase, not reduce the preponderance of these regions in the legal inflow (Fix and Passel 1991). More remarkable still are provisions designed to ease family reunification for beneficiaries of the IRCA legalization program. The law reserves 55,000 visas per year for three years for immediate relatives of those legalizing under IRCA. In addition, it bars the deportation of all spouses and children of the IRCA legalizees who were already in the country on May 5, 1988. The exact number of beneficiaries of this "family fairness" provision can not be estimated exactly, but will likely reach the hundreds of thousands. Finally, and largely at the insistence of human rights advocates and the government of El Salvador, the 1990 Act grants "safe haven" to an estimated 500,000 Salvadorans already residing illegally in the United States for a minimum of 18 months. This period can be extended subsequently. Adding IRCA legalizees, their immediate relatives in and out of the country, Salvadorans and Irish currently unauthorized, the combined IRCA and 1990 acts will bring "above board" a number that can be estimated conservatively at 4 million. This is in addition to the newly expanded limits for legal immigration. Despite the efforts by Congress to regularize the vast unauthorized immigrant population and to deter new illegal entries, all evidence indicates that, after an initially successful period, the deterrent features built in IRCA are failing to have the desired effect (Papademetriou et al. 1991). This is suggested by the increasing number of apprehensions of surreptitious entrants by INS and the Border Patrol after an initial decline in 1987-88. A number of detailed field studies conducted in cities throughout the country point in the same direction (Bach and Brill 1991). Barring draconian changes in enforcement, the unauthorized flow is likely to continue and even increase during the coming years. Hence, it is not an exaggeration to say that the 1990s are likely to be the century's decade of immigration and that several cities are slated to repeat the experiences of the 1910s, when half or more of the urban population was foreign-born or of foreign parentage. This massive inflow will be diverse, but not along the national lines explicitly promoted by the "diversity" provision of the 1990 Act. Despite it, Mexico, Asia, and the Caribbean will continue to predominate among sources of immigration. Diversity will rather be in modes of entry and incorporation as outlined above. Hence,it is possible to expect a growing number of prosperous communities of immigrant entrepreneurs and professionals alongside a large number of poorly educated immigrants filling menial jobs and clustering in impoverished inner-city neighborhoods. New unauthorized immigrants from Mexico and Central America, as well as sizable groups of earlier refugees and would-be asylees will be in the latter category. It is too soon to tell just how this massive inflow will affect the American labor market and what consequences it would have for native-born minorities. There is a notable disparity, however, between the rationale in Congress for expanded immigration and the employment opportunities for the latter groups. Increased immigration was justified in the debates preceding the 1990 Act by a supposed "mismatch" between rapidly growing labor demand in this decade and the coming century and a shortage of domestic workers. This position, advocated by an influential Hudson Institute report (Johnston and Packer 1987) stands in stark contrast with high levels of unemployment and poverty among native-born blacks, Puerto Ricans, and other minorities. It would seem as if the idle labor potential of these groups was discounted or held to be unemployable by proponents of the "mismatch" thesis. It is not yet clear what the employment of a growing mass of low-wage immigrant workers would do to the economic situation of the native born. The preference for more motivated and more docile immigrant workers manifested by employers in a number of field studies may accelerate the displacement of domestic minorities and increase their already high unemployment levels. Alternatively, the economic dynamism promoted by the presence of the newcomers may open supervisory and other positions for English speaking personnel that can provide a mobility path for domestic minorities. The balance between these contradictory trends is uncertain at present. The propensity for business creation and the development of ethnic enclaves by certain foreign groups will have beneficial consequences for their members, for reasons explained above. However, it is not probable that these benefits will extend to domestic minorities given the well-known propensity of immigrant entrepreneurs to hire and promote their own. The emergence of vibrant ethnic entrepreneurial communities may have the effect, however, of energizing native born minorities into attempting to follow the same path. Already a number of black community leaders and academics are advocating the reinvigoration of black business communities that effectively disappeared following integration. For these authors, the abandonment of the autonomous economic base provided by ethnic enterprise in the rush to enter the social and economic mainstream was a tragic mistake that must now be corrected. We believe that such efforts are pointed in the right direction, but not that the conditions making for the success of immigrant enclaves may be difficult to reproduce among large native-born minorities. Elsewhere, we have analyzed the unique combinations of human and social capital underlying the emergence of successful immigrant entrepreneurship (Portes 1987; Portes and Sensenbrenner 1991). It is a task for future policy and research to examine when such conditions can be reproduced among other minorities, native and foreign born, and when this is the case to devise programs for supporting such efforts. Along with policies designed to improve individual education and skill acquisition among impoverished groups, the promotion of those collective networks on which viable ethnic enterprise is always grounded can offer an important path for economic mobility in America. Bach, Robert L. and Howard Brill. 1991. Impact of IRCA on the U.S. Labor Market and Economy. Report to the U.S. Department of Labor, Institute for Research on International Labor, State University of New York, Binghamton. Bates Timothy and Constance R. Dunham. 1991. "The Changing Nature of Business Ownership as a Route to Upward Mobility of Minorities." Paper presented at the conference on Urban Labor Markets and Labor Mobility sponsored by the Urban Institute, Arlie House, Virginia, March. Bendick, Marc and Mary Lou Egan. 1991. "Business Development in the Inner City: Enterprise with Community Links." Community Development Research Center, New School for Social Research, February. Borjas, George J. 1990. Friends or Strangers, the Impact of Immigrants on the U.S. Economy. New York: Basic Books. Bouvier, Leon F. and Robert W. Gardner. 1986. "Immigration to the U.S.: The Unfinished Story." Population Bulletin 41, p. 51. Butler, John S. 1991. Entrepreneurship and Self-Help among Black Americans. Albany, New York: State University of New York Press. Cichon, Donald J., Elzbieta M. Gozdziak, and Jane G. Grover. 1986. "The Economic and Social Adjustment of Non-Southeast Asian Refugees." Report to the Office of Refugee Resettlement. Department of Health and Human Services. Washington DC, mimeographed. Fiedler, Tom. 1991. "Politically, we'll hardly recognize the place at all in '92." The Miami Herald. Viewpoint Section, October 6, p. 5C. Fix, Michael and Jeffrey S. Passel. 1991. "The Door Remains Open: Recent Immigration to the United States and a Preliminary Analysis of the Immigration Act of 1990." Program for Research on Immigration Policy, The Urban Institute and the Rand Corporation, Working Paper. Fratoe, Frank A. and Ronald L. Meeks. 1985. "Business Participation Rates of the 50 largest U.S. Ancestry Groups: Preliminary Report." (June). Washington, DC: Minority Business Development Agency, U.S. Department of Commerce, mimeographed. Gardner, Robert W., Bryant Robey, and Peter C. Smith. 1985. "Asian Americans: Growth, Change, and Diversity." Population Bulletin 40 (October): 1-43. Gold, Steven J. 1992. Refugee Communities. Newsbury Park, CA: Sage Publications. Gordon, Ian. 1992. "The Impact of Economic Change on Minorities and Migrants in Western Europe." Paper presented at the conference on Poverty, Inequality, and the Crisis of Social Policy, Joint Center for Political and Economic Studies (September). Green, Shelly and Paul Pryde. 1990. Black Entrepreneurship in America. New Brunswick, NJ: Transactions. Immigration and Naturalization Service. 1987. 1987 Annual Report. Washington DC: U.S. Government Printing Office. -----. 1991a. 1990 Statistical Yearbook. Washington DC: U.S. Government Printing Office. -----. 1991b. 1990 Statistical Yearbook of the Immigration and Naturalization Service. Washington DC: U.S. Government Printing Office. Johnston, W.B. and A.E. Packer. 1987. Workforce 2000: Work and Workers in the 21st Century. Indianapolis: The Hudson Institute. Light, Ivan. 1972. Ethnic Enterprise in America: Business and Welfare Among Chinese, Japanese, and Blacks. Berkeley: University of California Press. -----. 1991. "The Iranian Ethnic Economy of Los Angeles." Department of Sociology, University of California-Los Angeles, unpublished manuscript. Marger, Martin N. 1991. Race and Ethnic Relations. 2nd Ed. Belmont, CA: Wadsworth. Marks, Carole. 1991. "The Urban Underclass." Annual Review of Sociology 17: 445-466. Marshall, Dawn I. 1983. "Toward an Understanding of Caribbean Migration." Pp. 133-53 in M.M. Kritz (ed.) U.S. Immigration and Refugee Policy. Lexington, MA: D.C. Heath. Papademetriou, Demetrios G., Robert L. Bach, Howard Brill, Deborah A. Cobb, Roger G. Kramer, Briant L. Lovell, Martina Shea, and Shirley J. Smith. 1991. Employer Sanctions and U.S. Labor Markets: First Report. Washington DC: Bureau of International Labor Affairs, U.S. Department of Labor. Pedraza-Bailey, Silvia. 1985. Political and Economic Immigrants in America: Cubans and Mexicans. Austin: University of Texas Press. Portes, Alejandro. 1987. "The Social Origins of the Cuban Enclave Economy of Miami." Sociological Perspectives 30 (October): 340-72. -----. 1990. "From South of the Border: Hispanic Minorities in the United States." Pp. 160-84 in V. Yans-McLaughlin (ed.) Immigration Reconsidered. New York: Oxford University Press. Portes, Alejandro and Robert L. Bach. 1985. Latin Journey: Cuban and Mexican Immigrants in the United States. Berkeley: University of California Press. Portes, Alejandro and Luis E. Guarnizo. 1991. "Tropical Capitalists: U.S.-Bound Immigration and Small Enterprise Development in the Dominican Republic." Pp. 101-131 in S. Diaz-Briquets and S. Weintraub (eds.) Migration, Remittances, and Small Business Development. Boulder, CO: Westview Press. Portes, Alejandro and Rub‚n Rumbaut. 1990. P. 91 in Immigrant America: A Portrait. Berkeley: University of California Press. Portes, Alejandro and Julia Sensenbrenner. 1991. "Embeddedness and Immigration: Notes on the Social Determinants of Economic Action." Paper presented at the thematic session on Immigration, meetings of the American Sociological Association, Cincinnati. Portes, Alejandro and Alex Stepick. 1985. "Unwelcome Immigrants: The Labor Market Experiences of 1980 (Mariel) Cuban and Haitian Refugees in South Florida." American Sociological Review 50 (August): 493-514. Portes, Alejandro and Min Zhou. 1991. "Gaining the Upper Hand: Economic Mobility among Immigrant and Domestic Minorities." Paper presented at the conference on Poverty, Inequality, and the Crisis of Social Policy, Joint Center for Political and Economic Studies, Washington DC. Rumbaut, Rub‚n G. 1990. "The Structure of Refuge: Southeast Asian Refugees in the United States, 1975-85." International Review of Comparative Social Research 1 (Winter): 95-127. -----. 1991. "The Agony of Exile: A Study of the Migration and Adaptation of Indochinese Refugee Adults and Children." Paper presented at the conference on Refugee Children, National Institute for Child Health and Human Development, Bethesda, MD. -----. 1992. "The Americans: Latin American and Caribbean Peoples in the United States." In A. Stepan (ed.) The Americas: New Interpretive Essays. New York: Oxford University Press (forthcoming). Sanders Jimy M. and Victor Nee. 1987. "Limits of Ethnic Solidarity in the Enclave Economy." American Sociological Review 52: 745-773. Stepick, Alex. 1989. "Miami's Two Informal Sectors." Pp. 111-31 in A. Portes, M. Castells and L. Benton (eds.) The Informal Economy: Studies in Advanced and Less Developed Countries. Baltimore: Johns Hopkins University Press. Stepick, Alex and Alejandro Portes. 1986. "Flight into Despair: A Profile of Recent Haitian Refugees in South Florida." International Migration Review 20 (Summer): 339-41. Tumulty, Karen. 1989. "When Irish Eyes Are Hiding..." Los Angeles Times. January 29. U.S. Bureau of the Census. 1983. 1980 Census of Population General Social and Economic Characteristics - U.S. Summary. Washington DC: U.S. Department of Commerce. -----. 1984. Socioeconomic Characteristics of the U.S. Foreign-Born Population Detailed in Census Bureau Tabulations. Washington DC: U.S. Department of Commerce. -----. 1991. The Hispanic Population in the United States: March 1990. Washington DC: U.S. Department of Commerce. Wilson, Kenneth and W. Allen Martin. 1982. "Ethnic Enclaves: A Comparison of the Cuban and Black Economies in Miami. American Journal of Sociology 88 (July): 135-160. Woodrow, K.A., J.S. Passel, and R. Warren. 1987. "Preliminary Estimates of Undocumented Immigration to the United States, 1980-1986: Analysis of the June 1986 Current Population Survey." Proceedings of the Social Statistics Section of the American Statistical Association. Zhou, Min. 1992. New York's Chinatown: The Socioeconomic Potential of an Urban Enclave. Philadelphia: Temple University Press.