CHAPTER 4 BEHAVIORAL CHARACTERISTICS OF IMMIGRANTS Tom [Thomas Corcoran, adviser to Franklin D. Roosevelt] was the grandson of an Irish immigrant. His mother used to point to the immigrants who were still pouring into America from Ireland when he was a boy. "The cowards never started", she would tell her son. "The weak died on the way." (Michael Straight, After Long Silence, New York: Norton, 1983, p. 136) This chapter reviews evidence on a range of immigrant behavior related to their economic contributions. These characteristics include the rate of labor-force participation, the rate of unemployment, the propensity to start new businesses, the propensity to save, the propensity to exert work effort and vitality, geographical mobility, fertility, and the propensity to commit crime. Because the data are derived from a wide assortment of countries, the samples differ from country to country by age, education, and category of entrance (family reconstitution, needed skills, etc.). If these background characteristics were held constant, comparisons would be improved, and it is hoped that research along these lines will be done. The best we can do now is to keep these problems in mind as we draw conclusions from the available data. SIMILARITY OF BEHAVIOR IN DIFFERENT COUNTRIES Figure 4-0.1a and 4-0.1b show that the adjustment of immigrants to their new societies is strikingly similar in Canada and Australia despite that these data are unadjusted for background and the composition of immigrants into the two far- distant countries is quite different. This is strong evidence about the value of data for one developed (and perhaps English- speaking) country for understanding the effects of immigration in another country, especially when sufficient data for the country of interest are not available. Other data for Australia or Canada in this and the preceding chapters strengthen this point. ------------- Figure 4-0.1 ------------- ------------- Figure 4-0.2 ------------- LABOR FORCE PARTICIPATION OF IMMIGRANTS AND NATIVES The differences in age composition discussed in Chapter 3 are important mainly because of the effect that they have upon labor-force participation. But even within particular age-and- sex categories there often are differences between immigrants and natives; this is especially true among women, where there is more slack for group differences to appear. Among men in the prime years, the overwhelming proportions are in the labor force for natives as well as for immigrants, and hence there is little possibility of group differences. In the youngest and oldest male cohorts, however, where there is possibility of variation, immigrant participation tends to be higher than among natives. All the data that are readily available will be presented, even those that lump together the age composition and the participation rates, because all help illuminate the general issue at hand. These data should be understood as reinforcing the data on age composition alone in showing relatively high proportions of working persons among the immigrant populations. United States. Historically, larger proportions of women and young persons at each age have been in the labor force among immigrants than among natives. This may be seen in Hill's Table 4-1a for 1890. Additional evidence comes from Table 4-1b which shows that this effect, together with age-compositional effects, led to much higher overall participation rates among immigrants than among natives. ---------------------- Tables 4-1a and 4-1b ---------------------- Contemporary data for the U.S. do not show the same pattern, however. Rather, the labor-force proportions among the foreign born, natives of native parentage, and natives with at least one foreign-born parent, are similar (See Table 4-2). --------- Table 4-2 --------- Canada. Many more data are available for Canada than for the U.S. Since World War II, markedly higher labor-force participation is seen for immigrants than for natives in every age-sex category (Tables 4-3 and 4-4). Overall age-adjusted participation rates for the Asian and U. K. immigrant groups that entered during various periods (See Table 4-5) reveal little difference among males between those groups and the Canadian- born; among females, participation is higher among the immigrants, even the most recent immigrants. ------------------------ Tables 4-3, 4-4, and 4-5 ------------------------ The children of a foreign-born parent also have a higher rate of labor-force participation than do the children of Canadian-born parents; age-standardized data are shown in Table 4-6. Among males, having just one Canadian-born parent is correlated with the highest labor-force participation rate of all; this observation is verified by the similarity of patterns for males and females in each area shown. ----------- Table 4-6 ----------- Israel. Whereas in the early 1970's some 53% of the civilian Israeli population aged 18 and over was in the labor force, about 65% of the immigrant Soviet Jews in that age group were in the labor force two years after their arrival. (During the period of adjustment up to that two-year point, fewer of the Soviet Jews had yet entered the labor force.) That is, a fifth more of the Soviet Jewish immigrant adults were in the labor force than among the existing Israeli Jewish population--which is a huge difference in economic terms. Great Britain. Partly because the immigrants included a disproportionate number of males, Jones and Smith found that "72% of coloured adults were economically active compared with only 62% of the (adult) British population" (1970, p. 30). Australia. The pattern for Australia, seen in Tables 4- 3.1and 4-7, is similar to the patterns for other countries seen above. Women immigrants age 25-54 participate in the labor force considerably more than do native women of the same ages. And older immigrant men -- those 55-59 and especially those 60-64 -- participate more than do native men. --------- Table 4-7 --------- In a paper summarizing much useful data about immigrants and labor markets in Australia, Baker (1984) shows that the ratios of immigrants' to natives' participation rates in most sex and age categories fell from 1972 to 1983. (See his Figures 4.2 to 4.10) The overall Australian age-adjusted participation rates for native and immigrant males in 1972, 1976, and 1981 show much the same patterns. But for women the age-adjusted participation rates are significantly higher for the overseas-born, with married females being the cause of most of the difference; as Australian-born women moved more heavily into the labor force between 1972 and 1981, however, the difference between natives and immigrants narrowed somewhat (Baker, 1984, Table A4.2). Table 4-7 contains more recent data categorized by both age and length of time in Australia. It shows much lower labor-force participation rates for immigrants in Australia less than two years than for those longer in Australia, but after the two-year mark the rates for female and older male immigrants are generally higher than for natives. (Please note the small sample sizes for the immigrant groups which make particular cell estimates unreliable, especially cells close in the table to other cells having less than 25 observations and therefore without estimates.) Among males in Australia less than two years and not in the labor force as of June, 1981, 60% were attending an educational institution, indicating that low participation was not necessarily due to inability to find a job. And of that group, 58% had been in residence less than one year, an indication of increase in participation within the two-year period (Inglis and Stromback, p. 14). The fact that 40% of this group had poor English is further explanation of the observed low rate of participation; it should be noted that normal economic immmigrants tend to arrive with good English, unlike the refugees who constituted a considerable part of this sample. The data for multiple job-holding in Australia (Table 4-8) show a mixed pattern by arrival data, with generally lower rates for immigrants than for natives. This runs against the general pattern of the other data presented here. --------- Table 4-8 --------- Summary: The similarity in patterns of labor-force participation among these countries of immigration is striking. This consistency provides confidence that the experiences are sufficiently alike so that we probably will not be far wrong if we generalize from data in one or more of these countries to the U. S., when data for the U. S. are not available. Taken as a whole, the data tell us that labor-force participation is substantially greater among immigrant women and older men than among natives in those categories, the differences being quite sizable from an economic point of view. The reasons for these differences are still a matter of speculation, and are in any case outside the scope of this chapter. UNEMPLOYMENT Great Britain. In Great Britain, the unemployment rates for the New Commonwealth immigrant labor force in 1961 and 1966 were difficult for Jones and Smith (1970) to estimate because of spotty coverage of the relevant categories in the census. The 1961 unemployment level in the "conurbations" (large, densely populated, urban areas) was considerably higher (4.3%) for the immigrants compared to 1.6% for the total native labor force (p. 39), but Jones and Smith did not consider this a representative comparison. They had more faith in the 1966 all-Britain figures of 1.63% for immigrants and 1.37% for the total labor force (p. 40; yes, rates were once that low!). This difference between immigrants and natives is apparently due to the much higher unemployment rates for New Commonwealth women (1.52%) than the rate for all women (.76%), which they interpret as being the result of immigrant women having a greater propensity to be in the labor force than native women. Canada. Data on the length of time it takes immigrants to find a job are interesting even without comparison to natives. In Canada in the early 1970's, the average period between arrival in Canada and starting work was 4.2 weeks (p.19). Unemployment after six months was 10.1%, but then dropped to 4.8% by the end of their third year (1972)--the final level being higher than the Canadian control group's 3.7% but below the average of 6.8% for the male Canadian labour force. At the same time, the duration of unemployment diminished until by the end of the third year after arrival the immigrant sample as a whole, at 2.8 weeks on average, was within striking distance of the control group on this measure. Altogether, considering that average levels of unemployment in Canada remained high during this period, it can be concluded that immigrants were very successful in finding employment in Canada. (Manpower and Immigration, 1974, p. 8) United States. Chiswick's analysis of the 1970 Census of Population, and of the 1976 SIE survey, produced these results: In 1969, a year of tight labor markets and very low unemployment, adult male immigrants in the U.S. for less than five years worked three fewer weeks during the year than native-born workers with the same demographic characteristics. This implies that the immigrant unemployment rate was higher or the labor force participation rate was lower by a combination of 6 percent of the labor force. The work-year of those who had been in the United States for five to nine years was only one week shorter than that of the native-born, implying that the immigrant unemployment rate was higher or the labor force participation rate was lower by a combined 2 percent. There was no difference in the number of weeks worked between those who had been in the United States for more than ten years and the native-born. A similar pattern occurred in 1975, a year of very high unemployment, although the intensity of unemployment for those in the U.S. less than 5 years was greater in the recession. ... Shorter work years are not necessarily proof that the immigrants were looking for work, but a direct examination of unemployment rates shows similar results. In the 1970 Census reference week, the last week in March 1970, immigrants in the United States for up to five years had a significantly higher unemployment rate than that for the native-born; for immigrants in the country a longer period of time, there was no difference in unemployment rates. It is important to note that both shorter work years and higher unemployment rates are short-term characteristics of immigrants that virtually disappear after the first five years in the country (Cafferty, et.al., 1983, pp. 17, 18.) Among Vietnamese refugees, the employment ratio for males aged 16 and over rose from 67% in August-September, 1975 (it is not clear how long they had been in the country at that time) to 86% in July-August, 1976. This is a rapid rise to a high rate of employment, compared to native-borns, especially when one considers that some of the males over 16 are in school or retired. DeFreitas (1985) found in a micro-sample from the 1980 census that among "Anglos" who work in industry, the 1975-1980 immigrants work slightly more hours per week. Those who had immigrated between 1975 and 1980 suffered slightly more unemployment per year than the native-born, though working more hours. With respect to Hispanics, the 1975-1980 immigrants do worse than the native-born in all respects (and have a far lower proportion with twelve years and more education -- 29% compared to 62%). Recent female immigrants do less well than the native- born in all categories. Israel. Within two months of arrival, 66% of the Soviet immigrants who came to Israel in 1972-1973 and were in the labor pool had found jobs. Within one year of arrival, 90% of those in the labor force who came in 1971-1972 were employed. For males, the available figures are even more impressive: 95% of the men in the labor pool who came in 1971-1972 were employed in less than a year. Australia. Among persons in the labor force, the unemployment rate in 1977 was slightly higher for immigrants who had been in Australia 16 years or less, and much higher for the first year or so until the immigrants become adjusted (Table 4-9), compared to natives. After 16 years in Australia, unemployment is lower amnong immigrants than among natives. Similar data for 1972 through 1975 show the same pattern (ABS, Migrants in the Labour Force, 1972-1976). And data for 1983, when economic conditions and general unemployment were somewhat worse than earlier, show much the same pattern but reveal considerable diversity in unemployment rates (and in participation rates) among groups from different countries of origin. (The educational-occupational mix also varies greatly among such groups). The authors of the latter study could find no satisfactory explanation for the differences among countries of origin, though they noted differences in proportions of elderly persons within the various ethnic community; that compositional difference could contribute to th apparent overall differences in unemployment rates. (Victorian Ethnic Affairs Commission, 1983) Harrison (1983, Tables 8 and 7) shows considerable differences between persons from English- speaking and non-English-speaking countries. --------- Table 4-9 --------- Table 4-10 shows more detail on the relationship of age and period of residence to unemployment in Australia, as of June, 1981 when native unemployment was relatively high and immigrant unemployment was even higher. (For an explanation, see Chapter ll. And see Figure 4-1. Also, Baker's data [1984] for the younger age-and-sex categories in various years from 1972 to 1983 show much the same pattern as does the aggregated Figure 4-1). Unlike participation, which tends to level off within a few years, unemployment continues downward over the entire observed period for married men, who are the most important segment of the labor force; this latter phenomenon is less marked for the other three groups for the crucial 25-44 age groups. It also is much less marked among those from the U. K. and Eire, and somewhat less marked among those from New Zealand, the most "economic" of all immigrants; additional evidence for the latter is seen in Table 4-11 (Inglis and Stromback, Table 3.2). -------------------------------------- Table 4-10, Figure 4-1, and Table 4-11 -------------------------------------- A more aggregated analysis sheds additional light from a policy point of view. In Table 4-12 we see that for the decade from 1973 to 1982, persons in Australia 17 months and more (the category which contains the bulk of immigrants) have unemployment rates not substantially different than persons born in Australia. This suggests that the rate of unemployment among immigrants is not a critical factor except insofar as the high rate among new immigrants indicates that help in finding work for them might be appropriate. ---------- Table 4-12 ---------- Among General Eligibility Migrants -- those who are not refugees or relatives of families in Australia -- "In the first six months only 10% had not found employment", a rather remarkable record which speaks once more to the desirability of allocating admission by economic criteria. (Northage & Assoc., 1983) The association between the general unemployment level and unemployment among immigrants is strongly seen in Figure 4-1. It is also shown by the fact that when the average duration of unemployment increases for the Australian-born, the duration increases even more for immigrants who are not from the main English-speaking countries, as may be seen in Table 4-13. ---------- Table 4-13 ---------- Summary. Immigrants do not remain long on the unemployment rolls drawing unemployment compensation. Rather, they find jobs, stay employed, and have unemployment rates that compare favorably with the rates for natives. PROPENSITY TO START NEW BUSINESSES United States A large proportion of new jobs in the United States arise in small businesses. Several studies show that firms with fewer than 20 employees, which employ about 33% of all working persons, create somewhere between 51 and 80 percent of the net new jobs. The data for firms with fewer than 100 employees, and those with fewer than 500 employees, show the same high propensity to create new jobs. (U.S. Small Business Administration, November 2, 1984, p. 3). And immigrants apparently have a high propensity to start their own businesses; this seems obvious to the casual observer. For example, the proportion of taxi drivers (often self-employed) who have foreign accents seems high from New York to Melbourne to Malmo, Sweden. And it is most noticeable where entry is easiest, e. g., Washington, D. C., where only small fees are required to put a cab on the road. There does not seem to exist a reliable study of immigrants' business-opening for the United States (though it is badly needed). In a 5% sample from the 1980 census, T. Sullivan found that immigrants "were slightly more likely than the U.S.-born to be employed by their own corporation." (Correspondence, May 8, 1985). But these data are not adjusted for age and other characteristics. Borjas studied the extent of self-employment (a more general concept than small business) in a variety of immigrant cohorts, and among natives, in the 1970 and 1980 censuses. He finds that "Immigrants are more likely to be self-employed than similarly skilled native-born workers" (1985, abstract). This result probably would be even stronger if physicians and dentists and similar professions were deleted from the sample. Borjas's analyses suggest to him that a "major reason for this differential is that geographical enclaves of immigrants increase self-employment opportunities, particularly for immigrants who share the same national background as the residents of the enclave" (abstract). Perhaps more important than the analyses with background held constant are the unconditional probabilities of self- employment found by Borjas: .165 for immigrant whites versus .117 for native-born whites, and .053 versus .037 for immigrant and native-born blacks. Immigrant Asians and Cubans also have higher propensities compared to native-born whites as well as compared to native-born Asians and Cubans, respectively. Hispanics, both immigrant and native-born, have lower probabilities than do native-born whites. Overall, and most important, immigrants clearly have a higher propensity for self- employment than do the native-born. Unfortunately, however, Borjas does not provide unconditional data for immigrants who arrived in various periods so that they might be compared against the native-born stock. In the 1890, 1900, and 1910 census data, Higgs found the number of male retail merchants per 1000 employed males in large cities to be greater for foreign-born whites than for natives (see Table 4-13.1). ------------ Table 4-13.1 ------------ Light (1972, p. 13) found a similar pattern for 6 large cities in 1920, with additional interesting results for white natives with foreign-born parents, and for orientals (See Table 4-69). Fine-grained studies of Japanese and other Orientals in Los Angeles and California reveal startling large amounts of entrepreneurial behavior in both retail business and agriculture (Bonacich and Modell, 1980; Modell, 1977). The same is true of Cubans in Miami and Koreans in Los Angeles and Washington (Light, 1984, p. 203). ------------ Table 4-13.2 ------------ A similar pattern is found in New Zealand: The leading occupations of Chinese and Indians for nearly half a century have been market gardening and greengrocery in which they have served the national economy with an every growing efficiency. Today, they own about 15 per cent of New Zealand's vegetable production. The large Asian gardens of the truck farming areas also provide both employment and accommodation for many Maori labourers and their families. In fruit and vegetable retailing also the Chinese and Indians have had an important role. In 1964 they owned 529 greengrocery shops, which comprised 56 per cent of the country's total. Since World War II the Chinese and Indian greengrocery shops have spread throughout the length and breadth of the country and they are now common in almost every town (Taher, 1970, p. 62). Canada The Canadian Manpower and Immigration department conducted a panel study of the first three years in Canada of a cohort of immigrants entitled Three Years in Canada. Almost five percent-- 91 of the 1,746 males plus 291 single females in the panel sample on which there is data (Manpower and Immigration, p. 142)--had started their own businesses within the first three years in Canada. Not only did they employ themselves, but they also employed others, "creating" a total of 606 jobs. Expressed as a proportion of the 2,037 total immigrants, roughly 30% as many jobs were created from scratch as total jobs were held by immigrants. Furthermore, these numbers surely were rising rapidly after the three-year study period; after one year there were 71 self-employed immigrants creating 264 jobs, compared with the 91 persons and 606 jobs after three years.1 From the 1971 Canadian census, Richmond determined the proportions of various immigrant and native groups that are employers of others, as seen in Table 4-14 and Table 4-14.1.1.1 Among both male and female immigrants in Canada more than eleven years, a considerably higher proportion are employers compared to persons both of whose parents were born in Canada. And native- born persons with one or two parents born abroad have a considerably higher propensity to be employers than persons with both parents born in Canada: two immigrant parents are associated with an even higher propensity than one immigrant parent. The second-generation effect could be due, at least in part, to children of immigrant parents having parents who turn businesses over to them, but if so, it simply confirms that the first generation has a substantially higher propensity to start businesses than do natives. This indirect evidence of the first- generation propensity is particularly neat in that it avoids potentially-confounding differences in date of immigration with respect to the census data, because the parents of the second generation probably have been in Canada a considerable length of time as of the census date. (These data are standardized by Richmond for age distribution; the effect appears much more strongly in the unstandardized data in the complete original table.) The fact that the propensity to start new businesses is higher even among immigrants only 15-24 years old than among comparable-age natives suggests that the experience of being an immigrant, as well as self-selection, may account for this propensity, because a considerable proportion of immigrants ages 15-24 must have come with their parents rather than having chosen to migrate of their own accord. --------------------- Table 4-14 and 4-14.1 --------------------- Some data on the capital that immigrants have available to start new businesses may be found in Chapters 7 and 12. Table 4-15 shows data for Asian and U. K. immigrants with various periods of residence, as well as for the Canadian-born. The high propensity of Asian immigrants, both male and female, to open businesses after ten years of residence is impressive. The low propensity for U. K. immigrants is also interesting. Together the two sets of data suggest that entrepeneurship is very much a function of "need"; immigrants who can easily find their way to a good living without starting a business are less likely to start one than are immigrants who face greater barriers to a well-salaried job, it would seem. ---------- Table 4-15 ---------- In Australia the picture is more mixed though the outlines are the same as for Canada. Harrison's Table 4-16 shows that even after five years, immigrants from English-speaking countries are less likely to be employers or self-employed than are Australian natives. On the other hand, immigrants from non- English-speaking countries have higher propensities to start new businesses than do native Australians, starting sometime after five years in the country. (Unfortunately Harrison does not present more detailed data for longer residence in Australia). These data are additional evidence that the more difficult the adjustment of the group to the wage economy -- as seen in the data on unemployment rates among non-English-speaking persons above -- the higher the propensity to start one's own business. ---------- Table 4-16 ---------- FERTILITY The fertility of immigrants is of interest for several reasons. First, though parents pay most of the costs of raising their children until the children grow up and become productive and creative members of society, children are something of a burden upon the society, mostly by way of public expenditures on education. Second, to the extent that having a large number of adults is either a burden or a boon for other persons in the society (and elsewhere I argue at length that, in a society such as the United States in the late 20th Century, the benefits of more adults vastly outweigh the costs; Simon, 1977; 1981) more children now represent more adults in the future. Third, and perhaps most important, nowadays having many children is viewed by some (though not by this writer) as a symptom of personal sloth and "primitivism." Fourth, a birthrate higher than that of natives leads some to worry about the effect on the U.S. racial composition and cultural homogeneity. United States. Table 4-17 shows that for women born in the twentieth century, fertility among immigrants has been about the same as among natives, except that as of 1970 fertility was considerably lower among immigrant than native women age 25-44. And it is interesting to note that the fertility of women of foreign or mixed parentage -- who may be thought of as a result of recent immigration -- show much lower fertility than do native women, except among women age 25-44 in 1970. ---------- Table 4-17 ---------- The oldest cohort of immigrant women shown in Table 4-17 bore more children than did native women of the same age. Higher fertility for immigrants than for natives was apparently the rule in the more "advanced" societies in the past (Isaacs, p. 185) which fits with the pattern that the countries from which the immigrants came had lower incomes and higher mortality on average than did the United States. The more recent pattern reflects various crosscurrents in fertility which are affected by several variables and which are not yet fully understood. (For more discussion, see Simon, 1977, Part II, or Simon, 1974). The bottom panel of Table 4-17 shows data from the 1980 census. Little fertility difference between native-born and foreign-born is seen. The data in Table 4-17 showing Cuban fertility much below all-U.S. fertility, while all Spanish-origin (presumably including Cuban) fertility is much above all-U.S. fertility, is a bracing corrective to the notion that fertility of immigrants is strongly a function of one or another culture. Canada. Table 4-18 shows lower fertility for immigrants than for natives in almost every age group. And Figure 4-2 portrays fertility rising with the number of generations in Canada up to at least the third generation, at all levels of education. ------------------------- Table 4-18 and Figure 4-2 ------------------------- Australia. Table 4-19 shows that, except for a few small groups, such as women born in Malta and the Netherlands, and Asian women ages 40-49, immigrant fertility is lower than native fertility, no matter which comparisons one makes. It is a bit surprising that the data show only slightly higher fertility for the newest- arrived cohorts, though in the younger cohorts the observations may be biased by differences in age distribution within the brackets. ---------- Table 4-19 ---------- Hugo and Wood (1983) find that Australian immigrant fertility in 1971 and 1976 (measured by the standardized fertility ratio) was 7% and 6% higher respectively among immigrants than among native women. But this difference is mostly accounted for by the higher propensity of immigrant women to be married; among married immigrant women, fertility is only 1% or 2% higher than among comparable natives. It also is interesting that fertility is lower among unmarried immigrant women than among unmarried native women, by 15% or 20%. Among immigrant groups there is considerable diversity in these respects, however, as Table 4-20 shows. ---------- Table 4-20 ---------- MORTALITY In Australia, mortality among immigrants is considerably lower than among natives. (See Table 4-21). This effect is most marked, and most interesting, in the 0-4 year-old category. There is some evidence that people in the countries of origin also have lower mortality rates than do people in Australia -- which is a most unusual pattern in itself, given the usual relationships among income and migration, and income and mortality --which might be a partial explanation of the observed phenomenon (Wood and Hugo, 1983b). Another possible contributory explanation, particularly interesting from an economic point of view, is that immigrants are self-selected with respect to hygiene and general self-preservation. ---------- Table 4-21 ---------- For the United States I could find no census data on mortality of immigrants. Schultz (1980) reports that immigrant children have lower rates of reported health problems than native children, though after the immigrants have been in the country for some time, the two rates converge. It seems reasonable to assume that the health of childen is correlated with the health and mortality rates of the their parents, suggesting that mortality among immigrants is at least as low, or lower, than that among natives. SAVING RATES Saving obviously is important to any society. Hence it is worthwhile for us to learn what we can about the extent of saving by persons who might be added to the society as immigrants. In the past, financial saving which was transformed by investment into physical capital was preeminently important: nowadays, however, saving in the form of investment in human capital rivals physical capital in importance. It seems reasonable that saving for the two kinds of investment should be correlated (though they may be substitutes in some cases). If they are positively related, evidence about one activity provides information about both. Two sorts of decisions affect saving: the allocation of the family's given income into current expenditure and savings, and the allocation of the family's time into leisure and the production of income. And there are several possible influences upon these decisions, and hence upon the amount of saving, including: 1) Age distribution. Immigrants tend to be younger than natives, and therefore they are likely to save more, according to the life-cycle theory of saving of Neisser (1944), Modigliani (1966) and Samuelson (1975). People save during the earlier work years so as to be able to consume when retired, as well as saving for special expenses during the later work years such as children going to college. 2) Wealth. It is reasonably obvious intuitively, and it can be shown simply with an adaptation of Patinkin's excess-demand- curve analysis (1965, Chapter 1), or on more complex assumptions using Becker's time-allocation analysis (1965 in 1983?; see also Simon, l987, Chapter 3), that greater wealth leads to the purchase of more leisure, ceteris paribus. It may be noted that this analysis directly contradicts Keynes's "psychological law" that the rich spend a smaller proportion of their income than do the poor. And the evidence of Feldstein and others on the effects of Social Security entitlement upon saving supports this analysis and contradicts Keynes' position. Greater wealth also implies less saving out of current income by way of the life cycle saving-and-consumption mechanism.3 (Some data on wealth brought by immigrants may be found in Chapters 7 and ll.) 3) Human Capital. Immigrants may differ from natives in amounts of education and in other background characteristics that affect saving. The topic of saving as related to immigration deserves careful and detailed analysis in order to lay bare the partial influences as well as the overall effects. Regrettably, this research has not yet been done, and therefore I will be able to do no more here than present some scraps of unrefined evidence culled from other countries' experiences. Data on Saving Behavior Home ownership data should be a meaningful indicator of saving behavior, though several apparently-reasonable assumptions are necessary to link the two phenomena. In Canada "the foreign-born were more likely to be home- owners except in the most affluent group" (Richmond, p. 408). Table 4-22 shows the evidence for Richmond's conclusion. Table 4-22 also shows that though cohorts in Canada ten years or less had lower rates of home ownership than the Canadian-born, those in Canada somewhat longer than that had much higher rates. The age of the respondents is a complicating factor, of course, but it seems unlikely that it explains all the observed differences. Furthermore, the fact that persons whose parents were born outside of Canada had much higher rates of home ownership than natives (interesting in itself) would seem to confirm the conclusion about immigrants, while being safe from the confounding influence of age (though it may be confounded by higher incomes of parents and higher education and incomes of immigrant respondents.) (The underlying data, together with data on mortgages, could be productively studied in more detail to provide more data on this topic. The SIE survey, discussed in the Afterword to Chapter 5, contains some similar data which might be used for this purpose.) ---------- Table 4-22 ---------- HOW MUCH DO IMMIGRANTS EARN COMPARED TO NATIVES? Immigrants work and thereby produce goods and services. They themselves receive much of the benefit of this production through their own earnings, which they then utilize to pay for immediate consumption or for later consumption by way of savings. But through a variety of channels the rest of society also benefits from the immigrants' work. Chapter 5 will trace the effect of these earnings on natives through the taxes that immigrants pay. Chapter 7 will show how natives benefit through the return to the private capital that immigrants work with but that they do not own. Chapter 8 will discuss how productivity is improved as a by-product of immigrants' purchases out of their earnings, and as a result of the ideas that immigrants contribute to the production process. The higher the individual immigrant's earnings, the greater the likely contribution through all of those channels, as well as through other channels such as the immigrant's savings. High earnings are also likely to indicate that the immigrant has a store of skills that he or she may communicate to native workers with whom he or she works. For all these reasons, the amount an immigrant earns is a measure of the immigrant's contribution to natives' economic welfare. The absolute amount largely determines the extent of tax contributions. Hence the level of average earnings by an immigrant cohort, relative to average native earnings, is the ultimate test, and Chapter 5 summarizes the cost-benefit situation in that regard. An immigrant's skills and productivity relative to those in the same occupation are of interest, also, because individuals' human capital surely influences the productivity and the human capital of those with whom they work. Therefore, an immigrant's earnings relative to others of the same level of education and age (as proxies for occupation) are of interest to us. The trend in immigrants' earnings is also a matter of interest. If a person's income is increasing rapidly relative to others with similar background characteristics, it suggests desirable personal economic characteristics. Chiswick did the first systematic work along this line, studying immigrants to the United States (1978); he has also studied other countries' immigrants (1982). This has blossomed into a large literature which will only be sampled here. Chiswick found that, holding relevant variables constant, earnings "cross over"--that is, they catch up with and pass-- natives' earnings after l0 to l5 years in the U.S. The work of North (1979) and of Greeley (1982) points in the same general direction. Simon and Sullivan (forthcoming), using discrete rather than continuous functions for length of time in the U.S., find that the cross-over may be earlier than Chiswick suggested, depending on how you think about it. We also find that English- speaking ability has an important effect. Blau (1980) applied Chiswick's technique to groups of immigrants, classified by industry and ethnicity, who were in the United States as of the Immigration Commission's survey in 1909. While the data with which she worked were cruder than Chiswick's, the analogy is sufficiently close to make the comparison interesting. Blau found much the same as did Chiswick: Upon arrival, immigrants earned less than did native workers by somewhere between 12%-18% for men, and between 10%-14% for women. The time required for men to catch up with and then pass natives was 11-17 years for males, and 8-12 years for women. And native whites with foreign parents earned more than native whites with native parents, suggesting that there is a continuing positive effect of immigration by way of psychological-cultural factors. There is also a fairly substantial body of research (e. g. Chiswick, 1982; Hirschman and Wong, 1980) comparing the earnings patterns of natives to particular ethnic groups, especially Asians and Hispanics: the Asians mostly do better, and the Hispanics mostly do worse. One could hypothesize that the longer the distance that immigrants have to come, and therefore the greater the difficulty of returning to the home country, the larger the likely element of self-selection on such characteristics as independence and confidence in one's capacities to adapt to a foreign land.1 One might also speculate that immigrants who cannot return home for political reasons should have a stronger motivation to invest effort in adapting to the new land, compared to those who can return home. But there are enough exceptions from the pattern, such as Filipinos (relatively less successful) and Caribbean blacks (relatively more successful), to call these speculations into question. More generally, there seems to be little solid scientific understanding of the underlying forces influencing the success or failure of particular racial or national or religious groups, as Greeley (1982) has argued persuasively. The social disorganization theory of W. I. Thomas is hardly more convincing than the genetic theories which it largely (but by no means completely) displaced; both are falsified by the speed of the rise by apparently-handicapped immigrant groups toward average native economic attainment. (See Greeley's data as well as Chiswick's. And the economic successes of the Vietnamese in the few short years since they arrived, in a condition that was described by casual observers as socially shattered and demoralized, constitute dramatic though unsystematic additional evidence; see evidence below). Greeley provides some empirical evidence that different immigrant groups follow different paths in adjusting themselves to the United States, and in approaching or surpassing the mean level of native earnings in their economic accomplishments. But (as he is the first to assert) this evidence is not enough to support a comprehensive or compelling theory. For this reason, and also because it seems out of the question for U. S. immigration policy to discriminate on the basis of national origin, I shall not further discuss this body of literature here. The generality of the findings reported by Chiswick in his 1978 paper is enhanced not only by the historical evidence but also by evidence in other countries, as follows. Canada. Tandon (1978) finds patterns in Canada quite similar to those in the U. S., which makes sense given that immigrants to the two countries come from somewhat similar backgrounds, and that the two countries' immigration policies have considerable similarity. And Basavarajappa and Verma (forthcoming) find that in 1980, immigrants (adjusted for age and educational attainment, though probably not adjusted in as detailed a fashion as in the work by Chiswick and by Tandon) from the United Kingdom earned more than the Canadian-born, but Asian immigrants earned less than the Canadian-born. The recent Asian immigrants do much worse than those who came earlier, though this is not so for those from the U. K. It is not possible to know from these data the extent to which this is due to the time necessary for absorption, and how much it is due to lower economic capacity of immigrants in later periods (due to the increasing influence of family reconstitution immigration, and of the Vietnamese immigrants who arrived in substantial numbers in the later period) relative to those who came without relatives purely on the basis of their economic characteristics as measured by the point system. Great Britain. Chiswick (1982) finds that the pattern for Great Britain is more complex in some respects. But the problem is not inconsistent with the findings for the U. S., including the finding that an additional year of schooling abroad leads to less income in the new country than does an additional year of schooling by natives. Australia. Stromback (1984) employed a model similar to that of Chiswick to study the earnings of immigrants from English- speaking and from non-English-speaking countries relative to the Australian-born. I rephrase his conclusions as follows: After adjustment for amount of education and for experience (closely related to age), migrants from English-speaking countries had earnings similar to those of the Australian-born: that is, the returns to their education and experience are similar to the situation for Austrailians. Among immigrants from non-English- speaking countries with more than average education, there are lower returns to education, both for those with no experience and for those with experience, and the returns to experience for these persons are relatively flat compared to Australians. The explanations for these patterns remain something of a puzzle. Israel. The pattern for Israel seems different than the other countries, perhaps because immigration there was to a considerable extent affected by ideology and flows of refugees from Europe and Asia-Africa, which are outside the scope of interest here. Chiswick, however, argues that nothing in the Israeli results conflicts in principle with the results for the U. S. and Canada, and I find his argument convincing. Borjas (1985) has argued that Chiswick's results for the U.S. are due to changes in cohort "quality." The issue is still a matter of controversy. But the corroboration of the contemporary U.S. pattern by turn-of-the-century data, and by the data for other countries, suggests to me that Chiswick's basic findings will hold up. MOBILITY Kuznets (citation unknown) remarked that it would be difficult to overestimate the economic importance of domestic mobility and its enhancement of economic flexibility. It is logical that new immigrants should be particularly mobile, because (unlike natives) they do not already have a stock of knowledge of persons and institutions which make it cheaper to remain in a given place than to begin anew elsewhere. And indeed, in Australia it is found that "recent migrants are particularly mobile, although migrants who have been in Australia for many years tend to be no more mobile that the Australian-born population..." (Norman and Meikle, 1983, italics in original.) For example, among immigrants from the U.K. and Ireland (representative in this regard of all immigrants, of those in Australia for various numbers of years, the proportions who moved their residence between 1976 and 1981 were as follows: 0 - 4 years: 9l%; 5 - 9 years: 59%; 10 - 14 years: 48%; 15 - 21 years: 45%; 22+ years: 31%; (Norman and Meikle, 1985, p. 80). Data from the 1970 U. S. Census tell a more complex story. Table 4-23 shows that for those persons who resided in the United States during the entire previous five years, the proportions who moved from one county (or state) to another are astonishingly similar among the foreign born, natives of native parentage, and natives of at least one foreign-born parent. But if we include persons who were abroad five years earlier, a much higher proportion of immigrants than of natives moved into the county in which they are now living during that five year period. On the one hand, this latter observation suggests that more immigrants had moved to a new area, and therefore there was more new matching of job applicants to job openings (though this observation is diminished in importance by other data that show immigrants tend to move where there are countrymen rather than where unemployment is low; see Chapter 12). On the other hand, the similar proportions moving of those within the country suggests similar propensities to move in the future. In other words, the mobility of immigrants may be a one-time thing, beneficial at that time but not indicating a continuing behavioral characteristic. ------------ Table 4-23 ------------ It is reasonable to expect more mobility among new immigrants than among the existing labor force simply because new immigrants are relatively young, and younger workers are more likely to change jobs than are older workers. Evidence for this is found in Figure 4-3a, which shows the geographical mobility of various age groups in the U.S., and in Figure 4-3b, which shows the job mobility of various age groups in Australia. More direct evidence on the mobility of immigrants is shown in Table 4-24 which shows that persons who arrived within the last two years changed jobs much more frequently than the average of the Australian-born, and much more frequently than persons who had been in Australia a longer time. This mobility is also attested to by the greater willingness of those who arrived within the last two years, compared to the average Australian-born person, to move from the state of residence in order to find work -- 51% to 38% (Cooper and Constantides, p. 77). It is also interesting to note the greater propensity of immigrants from the main English-speaking countries of origin to change jobs than for other immigrants; this probably reflects their greater ease in communications within the job market. But there is only slightly greater willingness of persons in that group currently looking for work to move from the state than for other immigrants, or even for the Australian-born (who have more roots) -- 40% compared to 35% and 38% (Cooper and Constantides, p. 77). ------------------------- Figure 4-3 and Table 4-24 ------------------------- catalogue no. 6209.0, December, 1983] Kindleberger (1967, p. 197) cites much higher labor turnover in Germany in earlier years after World War II for foreign than for domestic workers -- 23% compared to less than 11%, and even this much domestic turnover was thought to be abornmally high because "substantial proportions of the expellees and refugees [from East Germany and elsewhere] have not decided on a permanent residence". Kindleberger views this high foreign-worker turnover as a business cost because of the expense of hiring and training. It may indeed be costly from the point of view of the firm, but the benefits to the economy as a whole should not be lost sight of. Kindleberger also suggests that the mobility of foreign workers may enable the domestic work force to avoid the necessity of moving, thereby reducing the flexibility of the native work force (p. 204), but this does not seem a weighty matter to me. VITALITY AND THE PROPENSITY TO EXERT WORK EFFORT All of us recognize that people vary in the amounts of effort that they exert in various tasks. Yet effort is not a variable in received economic theory. (On this general topic see Simon, 1987.) And there is very little evidence that bears directly upon the matter. Therefore, it is necessary to digress a bit to establish the reality of the phenomenon, as well as its importance. Anyone who has exhorted herself or himself to "try harder" physically or mentally in a sporting match or in an examination, or who has "racked the brain" in trying to solve a problem, implicitly believes that the results are variable in response to inputs other than time and money. The same is true for cirumstances in which you have "driven yourself" -- a telling term -- to continue expending effort despite unpleasant sensations of pain from fatigue or injury or sickness. The same is true for the experience of exerting self-discipline, the essence of which seems to be a capacity to bear pain for a purpose. Studies of athletics show wide variation among persons in willingness to exert this sort of effort. Attention to the task at hand, and to the production of new ideas, seems to be related to this concept of effort. Change of any kind, including adjustment of one's ideas and institutions, seems to require effort. And given that migrating requires adjustment, it seems reasonable that immigrants are capable of exerting the unusual amounts of effort required to bring about change. Not only do we observe that effort varies with the circumstances, but as consumers and employers we act upon that observation. We offer extra pay to a taxi driver to "try" to make the airport in time. We give bonuses for superior performance in business and in professional athletics. Sellers will literally pay prospective buyers to "try" new products. There are several reasons to expect that immigrants will exert more effort than natives. One reason is that immigrants on average arrive with less material assets than natives possess, which follows from the fact that the immigrants are coming from countries where the level of income is lower on average (though see data on capital transfers in Chapter 7). We may model the effort-exertion function as depending upon wealth (and opportunity) as follows: (Wealth ex post accepting the opportunity - ) Effort = f ( wealth ex ante) accepting the opportunity) (____________________________________________) ( wealth ex ante ) This formulation is abbreviated and therefore cryptic: full explanation requires an entire book (Simon, 1987); hopefully, however, even this short statement is suggestive of the idea lying behind the notion that immigrants' lack of wealth is a spur to their achievement. Another reason to expect greater effort from immigrants than from natives is that immigrants are likely to be self-selected on the basis of economically positive characteristics. It makes sense that those persons who believe that they have characteristics that will enable them to succeed in a new country will be the most likely persons to migrate: they are likely to earn a greater return on their investment of money and time in moving to and adjusting into a new country than are persons less likely to succeed in the new country. And the propensity to exert effort certainly is a characteristic that predisposes one to success in a new country. I have found little direct systematic evidence on the extent of effort exerted by immigrants compared to natives. Therefore, we are thrust back on two sorts of evidence, anecdotal accounts from newspapers and elsewhere, and indirect evidence from various proxies for effort. "Four owners of full-service gasoline stations in the Aspen Hill-Wheaton area have gone to court in Montgomery County to see if they can contain sales at "pump-your-own" stations that they say are driving them out of business. The plaintiffs, who already enjoy more protection from self- service operations than their colleagues in most other states, are seeking monetary damages and a judge's order forbidding two owners of self-service stations from selling gasoline below a 7 percent markup that the plaintiffs say state law ordains... The defendants Boo Chung, a 42-year-old Korean immigrant with a station in Wheaton, and Abdhol Hussein Ejtemai, 35, an immigrant from Iran who runs a station in Aspen Hill, agree their prices are low. Ejtemai's charge of $l.ll9 for a gallon for self-service regular is in fact the lowest-priced gasoline in the metropolitan area, according to American Automobile Association figures. The defendants say they are simply passing to their customers the savings of their no-frills operations and not selling gas for less than they paid for it. Returning to the observation that immigrants have a very high propensity to be taxi drivers, there would seem to be no other likely explanation than that immigrants are more willing to work hard, and more willing to work for earnings which depend upon their output, than are natives. It also seems that immigrants have a tendency to seek out jobs as domestics, which requires a certain amount of enterprise. For example, of the first thirteen newspaper advertisements our family called in the Washington area, twelve were placed by immigrants. The data which indicate that immigrants have a greater propensity to save, to take second jobs, and to earn more than immigrants after some time on the job, and after starting at lower salaries, are all consistent with the idea that immigrants exert more effort. The data showing that new immigrants work fewer weeks per year than do natives in the U. S., ceteris paribus, point in the other direction, but this latter result is more likely to be the result of the time it takes to enter the labor market than a shortfall of effort. Shift work in France provides some direct evidence. Shifts other than daytime must require special motivation, and in fact the other shifts are usually paid somewhat more and are assigned to people with low seniority. A study of 54 French industrial firms that employed a high proportion of foreign workers found that while 54% of the first shift was composed of foreigners, the second shift had 64 percent foreigners, and the night shift 80 percent foreigners. (Rerat et. al., 1974, reported in a publication whose identification has been lost and whose original publication I could not obtain.) Other direct piece of evidence comes from an Australian survey of employers. "Our surveys confirm that Australian employers find migrant workers generally more dedicated, hard- working...than the Australian born. There is a clear positive preference for migrant labor." (Norman, 1984, p. 2). This observation is confirmed by the fact that decade after decade, in country after country, the greatest opponents of immigration--the labor unions--pay immigrants the backhanded compliment of saying that native workers "cannot compete" with immigrants because the immigrants are willing to work harder than natives in dirty jobs that natives will not do. Embarrassing one would think, and therefore persuasive. As FAIR put it, "American workers are hurt by being forced into competition with illegal immigrants who work hard and scared and sometimes `off the books,' who don't complain about unsafe conditions or low pay" (Conner, 1980). If those who are biased against immigrants, and who supposedly have first-hand knowledge of the workplaces, say immigrants work harder, should not the rest of us believe it true? (Additional discussion of this issue may be found in Chapter 15.) A factor influencing effort may be the perceived absoluteness of the break with the country of origin and hence the commitment to the country of destination. And indeed, Cubans have dome much better in the U.S. than have Mexicans. "We had one major advantage," Mr. Suarez, the television cameraman, said. "We knew from the beginning that there was no way back to Cuba, and even more to pre-Castro Cuba. Those who came before us, for a long time thought it is possible, and some of them still think so, living in an old historical capsule. New York Times, "Cuban Refugees Adapt to Life in U.S.," September 29, 1985. Running against this theory, however, is that people from other Caribbean nations, who can return if they choose, do well in the U.S. But many of them do not have a language problem. Intellectual vigor shows in the school results of immigrant children. Item: An astonishing number of new oriental immigrants have been winners of the high school Westinghouse Science Talent Search awards -- 12 of 40 finalists in 1983, 9 of 40 in 1984, and 7 of 40 in 1985 (Bell, 1985, p. 26.) Item: The Maryland Function Test results for Montgomery County show that Asian students, while they do considerably poorer (92% pass) than whites (99% pass) on Grade 9 reading, do considerably better in Grade 9 math -- 85% pass versus 82% in 1983, and 88% versus 82% in 1984 (The Washington Post, Jan 9, 1985, p. B4). Item: In a study of 1400 Southeast Asian refugee households, Caplan, Whitmore, Bui, and Trautmann ... investigated the academic achievement of approximately 350 school-aged children from the sample of refugee families. All of the children had arrived in the U.S. after October 1978; most of them spoke no English upon arrival here. Yet, after an average of just three years in this country, these children were doing extremely well in school. On national standardized tests of academic achievement, 27 percent of the refugee children scored in the 90th percentile on math achievement -- almost three times better than the national average. And although they scored somewhat lower than the national average in English language proficiency, they outperformed their school-aged peers on general grade- point average, with 27 percent earning A or A-minus. (1985, pp.5,7) Item: "The Berkeley student body is now 22 percent Asian- American, UCLA's is 21 percent, and MIT's 19 percent. The Julliard School of Music in New York is currently 30 percent Asian and Asian-American" (Bell, 1985, p. 26). Item: "-- 35 of the 150 American Nobel laureates in the past 79 years were naturalized citizens of the United States" (McCafferty, et.al., 1983, p. 7). How much of the vigor of immigrants is due to their being "hungry" rather than being settled and affluent, how much due to their being self-selected for vigor among the populations they come from, and how much due to the fructifying effects of living in the tension of two cultures is an open question; the success of Vietnamese refugee students in the American schools does not answer the question because those refugees were hardly a random sample of the Vietnamese population. An answer to the question would be interesting, but it is not crucial in a policy context, however; what matters is that the vigor exists. (A more difficult question, and one that probably has more implications, concerns the relatively poor performance of the Hispanics and blacks, as, for example, on that Maryland Functional Test in both reading and mathematics. Hypotheses there are aplenty, but a solid explanation is sorely lacking.) THE CHILDREN OF IMMIGRANTS The children they produce are an effect of immigrants. Earlier, this chapter presented evidence for higher labor-force participation for the children of immigrants than for other natives, and a higher propensity to start new businesses. And Chiswick (1984) finds that in the 1970 Census data "men with a foreign-born parent have 15 percent higher earnings" than do natives (p.29). A third of this difference is from higher earnings with the same background characteristics, whereas most of the rest of the difference is due to geographic concentration of the immigrant children in the North and urban areas. The children of immigrants do not, however, have a greater propensity to have substantially more years of education than other natives (Schultz, 1980). Richmond and Zubrzycki (1984) show in Table 4-25 that the children of one or two immigrants earn much more than the children of two natives. These unconditional data (though age- adjusted) are the proper basis for policymaking, rather than data adjusted for education and other background characteristics; they properly show the effect of immigration (in the next generation). ---------- Table 4-25 ---------- The fact that having one rather than two immigrant parents generally leads to higher earnings is fascinating, but speculation about its causes is outside the scope of this volume. CRIME Do immigrants have a greater propensity to commit crimes than do natives? I could find no study on the rates of crime among contemporary immigrants. But it seems reasonable that the records of crime among immigrants in other places and in earlier years should throw light on the matter. Kindleberger (1967, p.207) reviewed the evidence on guestworker crime in Europe since World War II. He concluded that the scare stories in the newspapers were mainly the product of overactive imaginations, and he found that there was no statistical evidence to suggest a higher age-and sex-adjusted rate of crime among immigrants than among natives; German data actually show less crime among immigrants. Many decades ago, Commons arrived at the same conclusion for the United States (1924, pp.168-169, reference from Kindleberger). Steinberg studied the rates of crime among U.S. immigrants up to the first quarter of the twentieth century for the Select Commission (Appendix A). He notes that "Immigrants have been considered a source of crime and disorder in America since long before anyone ever thought of the United States. Some of the claims about immigration which would evolve into the twentieth century were first heard early in the colonial period" (p.474). Steinberg found no evidence that criminality has been prevalent among immigrants. He concluded that, by any measure, the rate of serious crime has been less among immigrants than among natives, though the rate of petty crime (vagrancy, disorderly conduct, breach of the peace, drunkenness) has sometimes been grater among immigrants. When age and sex are controlled for, the rate of all crime has been less among immigrants than among natives. More specifically, two chief findings are: 1. "During both periods [1820-1880 and 1890-1924], immigrants were more likely to commit minor offenses [vagrancy, disorderly conduct, breach of the peace, drunkenness] than were the native born. Natives tended to commit more property crimes and more crimes of personal violence than did immigrants." (p.464) 2. "When immigrants and natives are compared using controls for age and sex, immigrants are, at worst, equally prone to crime and, usually, less prone than natives." (p.464) Briefly discussed in Chapter 15 is the finding by Muller (1984) that crime by Mexican illegals in California, standardized by age and sex, is lower than for natives (check). DISCUSSION Various hypotheses have been advanced to explain the superior economic performance of immigrants. One such hypothesis has been that the least successful immigrants later depart the U.S.; if a sample of persons arrive with average characteristics, such a phenomenon would leave behind a better-than-average sample. However, by comparing the time pattern of earnings of various ethnic groups that have greater and lesser ability to emigrate, Chiswick (1984) is able to arrive at a persuasive conclusion that selective emigration is not in general a crucial factor. (However, in a comparison of schooling distributions by age in the 1960 and 1970 U. S. censuses, he finds evidence that among Mexicans with less than 5 years of education, there was a greater propensity to return to Mexico than among Mexicans with more education, though this finding is not mirrored in the behavior of other ethnic groups [p. 24]. This evidence suggests by default that, at least to some extent, economic behavioral characteristics are a cultural phenomenon rather than being determined entirely by the economic conditions that the individual faces. This may seem perfectly obvious to most people, but in fact the empirical evidence concerning cultural factors is very limited, and is hard to develop rigorously because of the difficulty (or impossibility) of controlling all the relevant variables. The tendency to assume that immigrants are different than natives in their general styles of life may often be off the mark, however. Morrison (1980) found that for immigrants living in New York City in 1907, holding income constant, "ethnicity is a poor predictor of spending habits", including the propensity to save. SUMMARY Almost without exception the behavioral characteristics of immigrants are conducive to economic advancement for the community as well as for the immigrants themselves. Compared to natives of the same sex and age, immigrants work harder, save more, have a higher propensity to start new business, and are more likely to innovate. Two frequent negative allegations -- that immigrants are more disposed to crime, and that they have large numbers of children which are a burden upon the native community -- have no basis in fact. With the demographic dimensions and behavior characteristics of immigration described in Chapters 3 and 4 in hand, we may now proceed to the analysis of the various effects of immigration. 86-81 Other10 12/10/87 FOOTNOTES 1There is little reason to believe that these results are due to a single isolated event. Aside from the single individual who employed more than 100 persons, and the 14 who did not answer about the number of employees, there were two employers with 20- 44 employees, three with 11-19 employees, five with 6-10 employees, 15 with 3-5 employees, eight with 2 employees, 19 with one employee, and 24 with no employees. 1.1Both Tables 4-14 and 4-14 are included because they do not completely overlap. They cannot be combined because the samples are somewhat different geographically. 3Consider a two-period work-and-retirement model. Initial endowment is $5,000, income is $10,000. Consumption in each period will be $7,500, with $2,500 being saved from current income. Now assume initial endowment is $0, and income is $10,000. Consumption in each period will be $5,000, with $5,000 being saved. In general, for given income, higher wealth implies lower saving out of current income. 86-81 Other10 1/14/87 --------- Table 3-0 --------- High and Low Education, 1980 Census NATIVE IMMIGRATED 1970-1980 TOTAL ASIAN IMMIGRANTS IMMIGRANTS _____ ___________________________ Percent with "College - 8.7% 9.7% 17.6% 4 or more years" (Persons 25 years and older) Percent with "College - 7.5% 12.5% 19.8% 5 or more years" (Persons 25 years and older) Percent with "Less than 2.9% 12.8% 7.4% 5 years of Elementary School" (Persons 25 years and older) Source: U.S. Department of Commerce, Bureau of the Census. 1980 Census of Population. Detailed Population Characteristics, March 1984. ------------ Table 4-13.1 ------------ 1890 1900 1910 _________________ _________________ _________________ Na- Na- For- Na- Na- For- Na- Na - For- tive- tive- eign- tive- tive- eign- tive- tive- eign- born born born born born born born born born Whites Blacks Whites Whites Blacks Whites Whites Blacks Whites _________________ _________________ _________________ 13 cities 71 16 83 48 10 66 44 14 83 6 Northern 70 16 80 46 11 64 41 13 80 7 Southern 78 16 120 56 9 101 58 15 140 (Robert Higgs, Competition and Coercion, p. 92.)