CHAPTER THREE THE DEMOGRAPHIC DIMENSIONS OF IMMIGRATION INTO THE UNITED STATES The aim of this chapter is to supply demographic data relevant to discussion of the consequences of immigration into the United States. Unfortunately there is not now available a reference volume presenting immigration statistics. Therefore, subject to constraints of space the chapter presents as wide a variety of data as I could find, covering as long time spans as possible. Closely-related subjects dealing with immigrants' behavior (such as fertility, and labor-force participation) rather than with immigrants' demographic characteristics at entry, are treated in Chapter 4. And data on the quantity and characteristics of illegal immigrants are presented in Chapter 15. THE VOLUME OF IMMIGRATION Gross annual immigration since 1820--the total numbers of immigrants arriving each year--is shown in Figure 3-1. The proportions of emigrants to immigrants, a magnitude which has received little attention, are shown in Figure 3-2.0 Net immigration--the difference between arrivals and departures, shown in Figure 3-3a--produces a different impression than does gross immigration. Sometimes, as during the 1930's, emigration has exceeded immigration. This has been the case even more dramatically in other countries of immigration such as Canada and Australia. (See Figure 3-3b for Canada). Emigration data make clear that gross immigration data are misleading when used to estimate population growth and labor force growth.1 -------------------------------- Figures 3-1, 3-2, 3-3a, and 3-3b ________________________________ The data on immigrant flows refute widely-held ideas. For example, the anti-immigration organization FAIR says, "Immigration to the United States is at record levels" (fund- raising letter signed by Roger Conner, no date, p.2). And a recent Commissioner of the U.S. Immigration and Naturalization Service (INS) wrote: "The United States is probably undergoing now the greatest surge of immigration in its history" (Chapman, 1976, p.2). The INS was including illegal immigrants, who are excluded from the data underlying the figures shown above. But the data on legal immigration are such that no plausible assumption about the rate of illegal immigration could justify such a statement, as we shall see below. The data shown in Figures 3-1 to 3-3 demonstrate that immigration into the U.S. is not an inexorable force, but rather depends upon people perceiving that circumstances in the U. S. are relatively favorable for new immigrants and for residents. The large scientific literature analysing the determinants of international and domestic migration mentioned in Chapter 1 shows that the volume of immigration is heavily affected by conditions in the receiving country, and less heavily by conditions in the sending country. (The number of countrymen in the potential receiving country also has a large effect on whether immigrants come [Bartel, 1982], and if they come, where they settle. Distance and the cost of transportation also have important effects, of course.) To be meaningful in discussion of immigration policy, a statement about the volume of immigration must be a comparison of some sort; inappropriate comparisons cause misleading and frightening conclusions. For example, one frequently comes across statements about the proportion that immigration bears to total population growth. That proportion may seem to be rising, even rapidly, especially if one uses very high estimates of the rate of illegal permanent immigration. For example, The Environmental Fund (TEF) says: "In 1980, the proportion of U. S. population growth due to natural increase (births minus deaths) was 59%, while population growth due to net immigration was 41%. By the year 2000, if current levels continue, natural increase would account for only 29% of U. S. population growth and immigration would account for 71%." (THE DATA, May, 1982, No. 1, p 1). And historian of immigration Graham (a FAIR director) writes: "Immigrants now supply half the nation's population growth, the proportion is steadily rising, and at some point, between the year 2000 and 2035, all population growth in the United States will derive from immigration." (The Center Magazine, 1978, quoted in an advertisement of The Center for the Study of Democratic Institutions). Again, unless one assumes there to be a volume of illegal immigration far beyond anything warranted by the scientific literature (see discussion below), these statements were quite wrong in 1978 when Graham wrote; births that year were 3.3 million, deaths were 1.9 million, legal immigrants were .6 million (U. S. Statistical Abstract, 1980, pp. 61 and 94), so legal immigration was perhaps one-fifth of total growth excluding emigration, and perhaps one-sixth including immigration, and perhaps a maximum of one-fourth including illegal immigration. Nor could the TEF 1982 statement have been anywhere near close to true even for the peak recent immigration year, 1980, when legal immigration was .8 million.2 By contrast immigration accounted for about 39% of natural increase during the decades 1881-1890 and 1901-1910 (Easterlin et al, 1982, p.3); even earlier, from 1830 to 1885, it was "almost a quarter" (Kuznets, 1977, p. 4). Even if the quoted assertions about the proportion of total U.S. increase represented by immigration were correct, such a measure does not tell anything meaningful either demographically or economically, because the proportion depends heavily upon native fertility. For example, if the native population were growing by exactly one human being per decade -- that is, at one person above zero growth -- and exactly one immigrant came, immigration would then account for fully half of the two-person population growth. Two immigrants would then account for two- thirds of growth, and so on. As fertility has fallen in the U.S., immigration inevitably constitutes a larger proportion of natural increase. This measure is not useful for any demographic-economic purpose I can think of except perhaps to indicate just the opposite of what those who have constructed this comparison have in mind: It shows how much scope there is for immigration to make up a shortfall in case one worries about low fertility leading to population decline. The measure is meaningful, however, if one cares about maintaining racial homogeneity (which we must note is no more illegitimate logically than wanting your children to look like you, even if some of the rest of us do not share your concern; morality is another matter). If immigration bears a high proportion to total population growth, there will be a relatively rapid rate of change of population composition by country of origin, assuming that the present composition of immigration differs from that of the native population. Immigration should also be seen in the context of native fertility. Births are projected to be around four million in the foreseeable future; immigration will be closer to 500,000 than to a million if present policies continue--something more than only a proportion of one to eight compared to fertility. A more meaningful ratio--demographically, economically, and sociologically--is the flow of immigration as a proportion of the yearly resident population, shown in Figure 3-4a. Immigrants who arrived in the decade 1901-1910 constituted 9.6% of the population at the end of that decade; by comparison, those who came in 1961-1970 and 1971-1980 constituted 1.6% and 2.0% of the population, respectively. That is, relative to resident population the rate of gross immigration -- the latter being more relevant from the point of view of absorption problems than is net immigration -- was five or six times greater in the earlier period than recently. And in the recent years 1980-1983, the ratio of immigrants to population was three or four times as large in Australia as in the U.S. (Ozdowski, 1985, p.3.) ------------ Figure 3-4 ------------ Another relevant comparison concerns stocks of people rather than flows: In 1910, 14.6% of the U.S. population was foreign- born, but in 1970 only 4.7% of the population was foreign-born, and in 1980 only 6.2%. That is, the recent proportion of foreign- born is less than half of what it was in peak periods in earlier decades. (The data from 1900 to 1980 are shown in Figure 3-5.) In 1850, in the "large" cities outside the South foreign-born whites were 39% of the population, and by 1920 were 66% (cities over 500,000; Kuznets, 1977, p. 3). And in Canada in 1981, 16.1% of the population was foreign born, compared to the 6% in the U.S. (Samuel and Woloski, 1985, no page number). ------------ Figure 3-5 ------------ Woodrow, Passell, and Warren show that the relative stock was rising rapidly around 1980, to 6.2%. And the rise continued through the Current Population Survey in June, 1986, at which time the proportion was perhaps 6.4% (1987, Table 3). The proportion was rising less rapidly during the period from 1983 to 1986 than from 1980 to 1983, however (1987, Table 4). In some flourishing countries there are very high proportions of foreign workers to total workers. (See Figure 3-6.) (U.S. data are not easily available, and in any case, official data would understate the proportion due to the presence of illegal workers). The data for Germany are particularly startling when one adds in (the numbers of) displaced persons, refugees, and immigrants from East Germany into West Germany after World War II. The proportion of the West German labor force that was either foreign or in one of these other categories was 31.5% in 1970, and must have been even higher in earlier years (Neal, 1983, pp. 7, 9). In the U.S. in 1860, 26% of white adult males were foreign born (Soltow quoted by Gallman, 1977). ------------ Figure 3-6 ____________ Though the key issue is whether an additional person, immigrant or native, is a benefit or a burden, stabilization of the population size has apparently become an implicit goal of U. S. policy, expressed even (or especially) in the context of immigration decisions. For example, the Staff Report of the Select Commission said that if certain immigration flows would occur, then immigration would reach "the figure for new entrants which would bring the United States to a stable population of 274 million...by the year 2050...The Commission's recommendations...would allow negative population growth within 80 years" (p. 423, 424). The Commission did indicate some doubt about this goal, when it continued the discussion quoted above with "...without the alleged deleterious effects of continuous growth", especially with respect to natural resources and the environment. But the Commission certainly treated population stabilization as a serious goal in its deliberations. I do not consider the relationship of flows of immigrants to projected stabilization to be a relevant comparison. Stabilization has no demographic or economic meaning in itself. As someone put it in the 1960's, the desire for zero population growth is simply a non-rational preference for round numbers. In brief, comparisons of proportions of immigrants to natives, as a flow during any given period and as a stock at any given time, appropriately measure the relative burden upon the native community as a whole, and upon individuals, of absorbing the cohorts of immigrants arriving at particular times. In contrast, absolute numbers of immigrants are of little economic interest. Nor is immigration as a proportion of total population growth of economic importance, though it may be of interest to those concerned about perpetuation of national and ethnic character. The ratios that are relevant indicate that the burden of refugee absorption has been relatively low in recent decades. AGE AND SEX COMPOSITION From the point of view of United States residents, the most important characteristic of immigrants is their age composition. As has been true of migrants in all times and places, contemporary immigrants typically are young adults just beginning their work lives. Their age composition is the key to their economic contribution to the native community, especially because immigrants lighten the burden of supporting the aged through Social Security, as will be discussed in more detail later. The reasons for the concentration of immigrants in the young-adult ages are easy to discern. Compared to an older person, a young person has before him or her a larger number of future earning years during which to recoup the costs of the investment in transportation, and of learning the ways of a new environment. And an older person has already made various investments in the home country, such as developing a network of personal contacts and developing various other capacities necessary to make a living, that cause the additional costs of earning a living in any future year in the home country to be lower for the older person than for the younger person; which for psychological reasons these sunk costs often exert more influence than sensible present-value calculations would suggest they should, causing the person to abjure change in personal circumstances. Let us consider the age-composition data for some representative countries. United States. The age distribution of legal immigrants is heavily concentrated in the prime labor-force years, much more so than the distribution of the resident population, as Figure 3-7 shows. A key element is the proportion that persons of ages 25- 34 bear to total population: 16.3% for natives, and 26.0% for immigrants. For Asians who entered between 1970 and 1980, the proportion was 31.0%. (Census of Population, PC 80-1-DI-A, U.S. Summary, Section A: United States, p. 1-7). And for those admitted under the occupational preference category, the proportion age 25-34 entering 1969-1976 was 43.4%; this proportion is a reasonable expectation of what would happen with a system of economic criteria, rather than the existing family- connections system. The story has been much the same throughout U. S. history, as Figure 3-8 indicates. In earlier years, there was a heavy predominance of males (Figure 3-9), among whom the labor-force participation rate is higher than among females. And if we generalize from the available data for Mexican undocumented illegals, are especially concentrated in the prime working ages of 16-44; additionally, they are disproportionately male and many are single; married men usually do not take their wives and children with them to the U. S. (See Chapter 15). The importance of the heavy concentration of immigrants in the younger working ages is especially important in light of the expected U.S. increase in the proportion aged 65 or over amony natives (see Figure 3-8), who will need to be supported by working persons. ------------------------- Figures 3-7, 3-8, and 3-9 ------------------------- Data on the recent distribution of immigrants by sex and by country are peculiar. For example, 72.4% of the immigrants from Finland in the fiscal years 1972-1979 were female, higher than among immigrants from any other country. Unlike the data for some other countries, these cannot be explained by the presence of U.S. military forces. It would seem that there is an "importation" -- or perhaps "exportation" from Finland -- of blond mates for American men. Economics assumes that there is a welfare gain to American men thereby. But what about the effect on American women? All kinds of interesting questions are raised thereby, that will not be explored herein. Experience in other developed countries confirms the U. S. pattern. (One reason why the following examples come only for developed countries is that one of the goods which richer countries buy is social science research on the nature of their societies.) Canada. Parai (1974, p.43) found that about 75% of immigrants were of labor-force age (15 plus) compared to about 60% for Canadian-born residents. Among immigrants who came between 1947 and 1972, 52.9% were classified as "workers", versus the 33% of Canadian-born residents classified as "workers" in the 1961 census (p.92). Higher labor-force participation among immigrant women than among native women aged 25-44 accounted for part of the difference (p. 44), but the main cause of the higher proportion of "workers" among immigrants is the difference in age distributions. Great Britain. Among the New Commonwealth immigrants whom Jones and Smith studied, the proportion of persons in the prime working ages from 25 to 44 was considerably higher than in the population at large: 46% versus 26% in 1961, and 47% versus 25% in 1966. (The 44% proportion for all immigrants in 1961 is similar to that for the New Commonwealth immigrants, confirming that the New Commonwealth immigrants are a fair sample of the larger group.) The proportion of immigrants who were 15-24 was also somewhat higher among immigrants than among natives, while the proportions 0-14, 45-64, and 65 and over were lower among immigrants (1970, p.19). Israel. Among the Soviet Jews (Israel's largest recent immigration wave, and the group that has been studied most intensively) who came to Israel in 1973, 68% were between the ages of 15 and 64, in comparison with 60% in the Israeli Jewish population. For the age group 15-44, the percentages favored Soviet Jewish immigrants 47% to 42%. The differences arise chiefly from the smaller proportion of children among the Soviet Jews. The Soviet Jewish immigration contained a slightly larger proportion of persons over 65 than found among the Israeli Jewish community, 10.5% versus 7.3%.2.1 This unusual characteristic for a wave of immigration was due to the Soviet Union's policy of allowing many elderly persons to leave in order to save on their pensions. Australia. Goddard, Sparkes, and Haydon (circa 1984) have provided a detailed comparison of the age compositions of immigrants and natives. In the top left diagram in Figure 3-10, the bars representing immigrants are wider in the prime labor- force years and narrower in the late middle-aged years than are the bars representing natives. -------------- Figure 3-10 -------------- Goddard et al. also divide total immigration into the groups that arrive in various admission categories--categories which correspond closely to U. S. preference categories {except for the "special eligibility" category, which is a small hodge-podge including children of Australian citizens born abroad; this category may be left out of the discussion}. The middle top diagram in Figure 3-10 shows that migrants who arrive in order to reconstitute families do not have a favorable age distribution relative to natives. And to the extent that female immigrants have a lower rate of participation in the labor force than do males (see Chapter 4), immigrants in this category also do not have a favorable sex distribution, though of course the relatively high proportion of females in the nubile ages cannot be considered separately from the age and sex distribution in the other categories. The top right diagram in Figure 3-10 shows the age distribution of immigrants who are admitted on the grounds that their occupations, including entrepreneurs, are in high demand. Such persons are concentrated in the 30-45 age brackets. Much the same is the case with the "independent" migration (bottom left of Figure 3-10) people "who have outstanding achievements in an occupation which is not necessarily in demand"; it takes some years of adulthood to attain such status, which explains their being older than other economic immigrants. The "labor-shortage" and "independent" groups are young but not as young as immigrants who come from the places that are entitled to enter Australia as through an open door with "no preference restriction." This may be seen in the bottom right diagram of New Zealand citizens, who may enter Australia freely; they are concentrated heavily in the 20-25 age bracket. The New Zealand-to-Australia pattern presumably is as close as one can get to a modern immigration totally free of special requirements, though it probably is somewhat more heavily weighted in the older age categories than would be a totally free immigration from poorer countries, because the age pyramid in New Zealand is much less heavily skewed toward the young ages than is the pyramid in poorer countries. (Migration among the Scandinavian countries also is very free. It would be interesting to study the patterns of behavior found in that context.) The age distribution of refugees (middle bottom, Figure 3- 10) is particularly interesting because it contradicts the common idea that refugees are less desirable immigrants than are "economic" immigrants. To the extent that economic desirability depends upon age, the refugees admitted to Australia are seen to be relatively desirable, though not so much as the wholly-free immigration from New Zealand. The pattern of immigrant labor-force participation within age and sex categories reinforces the effect of age composition of immigrants. Data on labor force participation are dealt with separately in Chapter 4 because they describe behavioral characteristics of immigrants rather than the demographic characteristics which are the subject of this chapter. EDUCATION AND OCCUPATIONAL SKILL The distribution of immigrants by skill, as measured by education and occupation--commonly but infelicitously referred to as the "quality" of the immigrants--strongly influences immigrants' effect upon natives through their earnings as well as through a variety of other channels. Immigrants' earnings largely determine their tax contributions, and also have a very strong effect on the amounts of transfer payments that they receive, as we shall see in Chapters 5 and 6. Other effects of skill and knowledge--such as the amount of innovation immigrants cause to happen, the extent to which co-workers are able to learn from them, and the way they function as citizens--are harder to document but probably are very important, also. A common belief is that immigrants are poor in the sense of not having much all-important human capital in the form of education. For example, W. Fogel writes: "Hundreds of thousands of people -- most of them poor -- are now immigrating to the United States" (flyer of the Center for the Study of Democratic Institutions, 1978). This is wrong. Contemporary immigrants have a bi-modal educational and occupational distribution relative to natives. As Table 3-0 shows, there is a disproportion of natives in the lowest education category. But there is also a disproportion -- much more important both numerically and economically than the low-education disproportion -- in the highest educational categories. In 1980, 16.1% of employed natives were professional and technical workers; the corresponding figure for immigrants who entered between 1971 and 1979 was 26%; in 1970 the figure for natives was 14.2%, whereas it was 23% for immigrants between 1961 and 1970. That is, a much larger proportion of immigrants than natives are in professional * and technical occupations (U. S. Statistical Abstract, pp.89 and 401, 1981). Figure 3-11 shows the comparative pattern for the year 1975. This paragraph refers to all permanent immigrants, legal and illegal. Given the relative size of net illegal immigration--perhaps one-third or less of legal immigration, and perhaps even zero--their educational and occupational distribution apparently is not such as to render incorrect the statements about immigration overall that are made in this paragraph. ------------------------- Table 3-0 and Figure 3-11 ------------------------- We may also compare the average educational level of the immigrants arriving each year with that of the stock of native workers, not holding any factors constant. Such a comparison tells us whether immigrants on balance raise or lower the education level of the U. S.labor force and its average productive capacity. Data come from the 1976 Survey of Income and Education (see Appendix to Chapter 5 for information on this survey.) Adult immigrants who arrived between 1965 and 1974 had 3/4 of a year less education on average than the average of natives in the labor force. Before 1965 the amounts of education were probably about equal for immigrants and the stock of natives, but there were probably somewhat larger proportions of immigrants in the highest education categories, in accord with the data on recent occupational distribution presented just above. A similar pattern may be seen for Canada in Table 3-1. Among immigrants who arrived between 1946 and 1961, the proportions in professional and technical and skilled work were far greater than the overall proportion of immigrants in the population. And in Table 3-2 we see that the proportions of both Asian and U. K. immigrants who arrived after 1960 with more than secondary schooling are vastly larger than the proportion of the stock of the Canadian-born. The high proportion of immigrants from the U. K. with advanced training other than university training is of particular interest, since it is likely to indicate technical skills. And the decline in the incidence of university training among the Asians from the earlier to the later period -- though it was still high in the later period -- probably was a result of many more persons in the later group arriving via family reconstitution rather than via the point system (and perhaps there was some influence of Vietnamese refugees, too). ------------------ Tables 3-1 and 3-2 ------------------ A similar pattern may also be seen for Australia in Table 3- 3. The data collected by Miller (1983, Appendix, in Harrison, 1983, Table 7) show this pattern for South Australia even when the data are adjusted for age. Those data also show similar average amounts of education for natives and for immigrants in South Australia. Not only does the stock of immigrants contain a share of educated persons larger than among natives, but the most recent cohort also contains a much larger share than the share among natives. (See also Baker, 1986, p. 18) And as in the United States, the Australian distribution is somewhat bi-modal, the proportion of immigrants apparently being greater--or at least not smaller--at the lowest educational levels than among natives. (See also Boyd, 1980, p. 2, re Canada.) --------- Table 3-3 --------- The offspring of immigrants are a consequence of immigration. Hence their educational attainments in Canada and Australia are seen in Table 3-3.1. In both countries, having one foreign- born parent leads to more education than having both parents natives. Having both parents born abroad leads to more education in Canada (ex Quebec) but less in Australia. ----------- Table 3-3.1 ----------- Another interesting piece of Australian evidence: A survey of General Eligibility Migrants -- those who are not refugees or relatives of persons already in Australia -- showed the following proportions in the various occupational groups: Professional and technical, 29%; skilled metal and electrical, 28%; skilled mining, building and construction, and other, 15%; clerical, commercial, 14%; semi-skilled, 8%. This is a remarkably desirable occupational profile, the result of the economic criteria used in selecting this group. (Northage & Associates, 1983) Table 3-3.2 shows the New Zealand distributions of immigrants, emigrants, and natives among various occupational- skill groups. Both migrating groups are more heavily concentrated in the high-skill groups than are natives as a whole. ----------- Table 3-3.2 ----------- Particularly interesting is that even throughout the early history of the United States, immigrants did not arrive with less education than natives had -- contrary to popular belief and contrary to the famous poem by Emma Lazarus at the base of the Statue of Liberty. Give me your tired, your poor, Your huddled masses yearning to breathe free, The wretched refuse of your teeming shore. Send these, the homeless, tempest tossed, to me: I lift my lamp beside the golden door. And of course the same "refuse" belief led to very different recommendations when the belief was held by another sort of person: Wide open stand our gates And through them passes a wild motley throng... O Liberty, white Goddess! Is it well To leave the gates unguarded? (Thomas Bailey Aldrich, quoted by Morrison, 1980, p. 353) Hill (1970, p.31) calculated a measure of the "labor force quality" of immigrants relative to that of natives, roughly equivalent to a percentage. The estimates are: 1870 - .97 (e.g. in that year immigrants had 97% the labor force "value" of natives); 1880 - .99; 1890 - .95; 1900 - .97; 1910 - .95; 1920 - .93. And according to Hill's (1970; 1975) analysis of the wages and occupations in censuses and other data sources, covering the period 1840-1920 but with special emphasis on the decade just before the turn of the century ...almost all the empirical evidence leads one to a conclusion in direct opposition to that reached in most of the historical literature...immigrants, instead of being an underpaid, exploited group, generally held an economic position that compared very favorably to that of the native born members of the society (1975, p48). And Bailyn's research on the Colonial-period Registry of Emigrants from Great Britain reveals that way back then, too, immigrants had excellent economic characteristics, being composed mainly of young "useful artisans" and farmers. Based on his data, he characterized as follows two groups of migrants, one from the "metropolitan" population center of Britain in the Thames Valley, and the other from the provinces: The metropolitan pattern, which characterizes the central migration from the Thames valley, is typified by a young man, in his early twenties, acting individually. He is not, characteristically, drawn from among London's most desperate, destitute slum dwellers; nor is he from the more stable or substantial segments of the population. He is, rather, an impecunious young artisan or craftsman who has served all or some part of his apprenticeship, or in a less formal way, learned something of a trade, found employment irregular or nonexistent, and, still unmarried and without family encumbrances, has decided to head out to the colonies alone. At the other extreme, there is a different pattern altogether. In this provincial pattern the characteristic unit is not an isolated male worker in his early twenties, a bondsman for several years of unlimited servitude. It is not even a person. It is, rather, a family, and a family that contains not only mature women but also small children, including a remarkable number of young girls. Few of these family members were indentured servants; in one way or another--often by liquidating all their possessions, real and personal--they had raised enough money, often just enough, to retain their freedom. (1986, p.31) Another way to measure the economic "quality" of immigrants is according to earnings level, which is an indication of productive capacity as judged by the market. Within a few years in the U. S.,the average contemporary immigrant male head of a family comes to earn more than does the average native male head of a family (e.g., Chiswick, 1979; Simon and Sullivan, no date). (Of course much of this superior earnings performance, and the relatively high education level, is due to the simple fact that immigrants migrate when they are young, and education levels have been rising all over the world. As I repeat many times, it is hard to overestimate the importance of the age distribution among immigrants, which is the source of so many of their economic effects.) In Canada, Table III-3.2 shows that within ten years or so immigrants earn as much on average as do natives. And natives with both parents born in Canada earn less than do natives of immigrant parentage, and less than even recent immigrants.3 ------------- Table III-3.2 ------------- NUMBER OF ALIENS ILLEGALLY IN THE UNITED STATES The number of aliens illegally residing and working in the United States enters into the discussion of immigration in many ways, and therefore must be discussed even though the measurement of illegal alien flows and stocks in necessarily fraught with uncertainty, and even though estimates are more specific and transient than the rest of the scope of the book is intended to be; the topic is discussed in Chapter 15, which is devoted to the entire subject of illegal immigration. COUNTRIES OF ORIGIN The national origins of immigrants have little or no economic implications for natives once immigrants' education and occupation are taken into account. Differences in economic performance related to ethnic origin are relatively small by any measure that I would consider meaningful. There may be some difference in the first generation between, say, Japanese and Haitians, but such differences tend to disappear after the first or second generation. A likely explanation for some of the differences is the distance that one must travel to reach the U. S., together with the political possibility of being able to return. These factors may explain the only reasonably large difference in performance which is observed, that between Mexicans and others. Despite the lack of economic importance of the matter, readers are likely to be interested in the ethnic composition of immigrants. The continents from which they have come are shown in Figure 3-12 (and also in Figure 3-1). The change in the immigration law in 1965 led to a major shift in pattern. ----------- Figure 3-12 ----------- THE DEMAND FOR IMMIGRATION Immigration policies make implicit assumptions about the number and types of people who would like to immigrate to the United States. Often the assumption is that the supply of potential immigrants is almost infinitely large. Perhaps typical is this letter in The Wall Street Journal: "We cannot make it easy for the millions of potential immigrants to come to this country. We would be overwhelmed". (Herbert N. Woodward, September 1, 1983, p. 25). Or another letter in the same newspaper: "We are faced with the fact that some large percentage -- and it could be a very large percentage -- of the earth's 4 billion non-U. S. inhabitatants would opt, if they could, to leave where they now live and come to this country. If that happened, Lifeboat U. S. would quickly sink" (Roger Lewis, Nov. 30, l981, p.23). The Staff Report of the Select Commission spoke of "the huge demand to immigrate to the United States" (p. 374). And the then-commissioner of the INS asserted that "a survey taken by the government of Haiti...revealed that 40 percent of the persons interviewed wished to migrate to the United States. Their motivation is economic" (Chapman, 1976, p. 11). There does not seem to be solid evidence for so large a supply of potential immigrants. One reason given for belief in a large supply is, as the first letter writer quoted above puts it, the assumption that immigration now is at an "incredibly high rate," but the early part of Chapter 3 refutes this assumption. There seems to be very little relevant evidence, a few scrappy pieces of which will now be mentioned. The most important conclusion implied by the review to follow is that systematic research into the topic would be exceedingly welcome. 1. The number of persons on the official lists waiting for immigration visas is astonishingly low, little more than one million as of 1980 and 1981 (Select Commission Staff Report, p. 377). Of course, this set of lists does not constitute a fair estimate of the numbers who would come from the various countries if they were allowed to immigrate freely; many persons are undoubtedly discouraged from applying because of their knowledge that the wait for admission is at least many years, and might be forever. 2. A very large number of Turks registered for positions as temporary workers in West Germany. There were 1.25 million persons in the queue as of 1971, at which time the Turkish government placed restrictions on applicants (such as that unskilled workers could not be over 25 years of age when applying, or skilled workers over 40). But the queue rose again to 1.2 million by 1973, representing a ten-year wait (Paine, 1974, p. 67). It should be noted, however, that the travel expenses were paid by employers in most cases, and jobs were provided in advance; both of these factors certainly increaed the demand for migration. Also, this program was not intended to lead to permanent family migration, the latter being a much more drastic step than is temporary migration by a single family member; this tends to lengthen the Turkish waiting list. 3. Most Puerto Ricans remain in Puerto Rico despite the absence of administrative barriers and the low cost of travel to mainland United States, though average earnings in Puerto Rico are considerably lower than in the mainland. 4. A relatively large number of Europeans immigrated to the U. S. in the years before World War II even though the passage was relatively more costly and difficult than now. Some of this large inflow probably was occasioned by worry that the U.S. would soon pass restrictive immigation laws, however. 5. Most Blacks remained in rural areas in the South until World War II opened up many job opportunities in Northern cities. This suggests that there will not be a large influx of ill- educated immigrants unless there are available attractive jobs that they think that they can obtain. Only in this fashion can one explain the observed low rates of internal migration in places such as Colombia, where income on the coast is much lower than in prosperous inland areas. 6. Most Mexicans do not enter the United States, despite the ease of entering illegally and the relatively low cost of the trip. And most who do enter illegally later return home after a relatively short time; this phenomenon is detailed in Chapter 15, and it is also evidenced by the extremely low rate of naturalization of Mexicans (and also of Canadians), as may be seen in Table 3-4. This point should be kept in mind even though the statistics concerning the number who do enter illegally arouse so much interest. If the United States were as universally attractive as so many Americans think it is, one would expect to find more illegal immigrants than there are now. --------- Table 3-4 --------- 7. During the 1930's, when labor market opportunities were poor, emigration exceeded immigration, as seen earlier in this chapter. 8. A body of studies, and especially the studies of trans- Atlantic migration mentioned earlier, generally find that the rate of immigration is most heavily influenced by the "pull" of employment, and only secondarily by the "push" of unsatisfactory labor market conditions in the country of origin. Recognition of the pull factor makes explicable the very large immigration to the U. S. in the decades before World War I, and it also implies that that episode is not necessarily a good predictor of future immigration rates if entry were free. The U. S. represented an extraordinary opportunity for many Europeans betweeen 1850 and 1914. There were great quantities of land to be acquired easily, and jobs seemed easy to come by. Credentials were not needed. To potential immigrants it seemed as if the streets were "paved with gold", even for those with relatively little skill. This also explains why so many fewer immigrants went to Canada and Australia and to some other places in the world where entry was also quite free. Under the present system, the effective demand for immigration is necessarily effected by the number of relatives of new immigrants. These numbers, categorized by elegibility and likelihood of immigrating, are discussed in Chapter XVI. 86-81 Demo3 1/16/87 FOOTNOTES 0Perhaps this is because it is never flattering to feel that people wish to depart your country and prefer to live elsewhere. It may cushion the blow to remember that some emigration occurs for financial reasons, such as taxes and the cost of living, especially after the age of retirement, and some other emigration is due to failure to adjust to the U.S. 1Contrary to my (and probably others') impression, emigration is not concentrated among older persons. Rather, among adults, the 20-29 group has a higher rate than any older group. The data for 1960-1970 are as follows, calculated as a per cent of the 1960 census count (from Warren and Peck, 1980, p. 78). Per Cent 1960 Emigrated Ages 1960-1970 _______ ________ 20 - 24 7.8 25 - 29 8.7 30 - 34 4.1 35 - 39 5.8 40 - 44 4.7 45 - 49 4.6 50 - 54 3.1 55 - 59 2.6 60 - 64 3.2 65+ 2.6 2For many years, a large proportion of Australia's natural population increase has come from immigration. From 1946 to 1975, 39% of the total population increase derived from immigration. That is, immigration was almost two thirds as large as natural increase. (Australian Immigration, Consolidated Statistics No. 8, 1976, p. 8). Yet this phenomenon has not had any noticeable economic ill effects. 2.1Israeli data used in this and other sections of this chapter are mainly derived from a follow-up study of a sample of 1,360 immigrants who arrived in 1969-1970, and similar samples up through 1972, done by the Central Bureau of Statistics (Israel) and the Israel Institute of Applied Social Research. The survey data are supplemented by Israeli census and national-income data. This report describes the immigrants from Russia, though data on other Eastern European immigrants are used to enlarge the sample where necessary. 3A large body of research based on comparisons of the earnings of immigrants who arrived at different dates has suggested that immigrants rapidly catch up with the earnings of comparable natives, and then pass them. Recently, however, Borjas (1985) has presented evidence suggesting that the observed recent results are due to a decline in "quality" of immigrants during recent decades, rather than due to assimilation. It will likely take some years of thrashing out the matter to arrive at a satisfactory understanding, and to know whether the currently- accepted generalization, based on data in other eras and places, is faulty. 86-81 Demo3 12/10/87 ------------- Table III-3.2 ------------- MEDIAN EARNINGS* OF CURRENT EXPERIENCED LABOUR FORCE, 15 YEARS AND OVER, IN 1970, BY SEX, CONTROLLING FOR BIRTHPLACE OF PARENTS OF THE CANADIAN BORN AND PERIOD OF IMMIGRATION OF THE FOREIGN BORN, REST OF CANADA (Excluding Quebec). (Standardized with Age Composition of Canadian Population.) Birthplace of Parents/ Period of Immigration Males Females Birthplace of Parents of Native-born Both Born in Canada $ 6635 $ 3218 One Born in Canada 7391 3444 Both Born Outside Canada 7369 3258 Period of Immigration of Immigrants Before 1946 7251 3316 1946-1960 7212 3466 1961-1965 6701 3333 1966-1971 (first five months only 5923 3061 of 1971) * Excluding negative and zero earnings Source: Richmond and Zubrzycki, 1984, pp. 105, 107, originally Statistics Canada, Special Tabulations 86-81 Demo3 12/10/87