CHAPTER 17 CONCLUSIONS AND SUMMARY OF MAIN FINDINGS This book analyzes the economic effects upon natives of legal immigrants who have come to the U.S. in order to improve their economic lot. The findings are intended to cast additional light upon a national debate that has been going on for more than a century, and in which there is now high interest: How many immigrants, of what kinds, should the U.S. admit each year? My general conclusion, based on the various analyses and findings, is: More than at present, and chosen more for their economic characteristics and less on the basis of family connections. Though the general conclusion offered here runs contrary to much public opinion, it is mostly in consonance with the economic literature, and the overwhelming consensus of the most respected American economists agrees, as documented in Appendix C. MAIN FINDINGS The findings and conclusions are summarized here in this first chapter. The introduction contains an idealized analysis of the consequences, plus some reflections that may be relevant to those who may read with an eye to future research. Immigration Theory and Trade Theory. (Chapter 2) Contrary to intuition, the theory of the international trade of goods is quite inapplicable to the international movement of persons. There is no immediate large consumer benefit from the movement of persons that is analogous to the benefit from the international exchange of goods, because the structure of supply is not changed in the two countries as a whole, as there is when trade induces specialization in production. Goods that are traded internationally are, in the usual case, produced by workers who are paid much less in the producing country than they could earn in the country where the goods are sold and consumed. The inter- country difference in that labor cost -- which can be very great -- is shared among the consumer (who pays a much lower price for the traded good than if the good were produced in the consumer's own country) and the employer of the labor which produces the goods (who sells the good for a considerably higher price abroad than if it had to be sold only at home). But immigrants usually receive higher wages in the richer country to which they move. The immigrants benefit by the higher wages. But there is little or no immediate gain to the consumers in the new country analogous to the gains when goods are traded internationally. An example may make the point best: You can hire a driver for a day much cheaper in India than in New York, to your considerable benefit if the driver is competent; the difference in the labor price arises from the different overall relative price structures in poor and rich countries. But if the driver moves to New York and is competent to work there, you will have to pay him/her close to the usual going rate in New York; the driver gains from moving because he/she "owns" himself/herself and obtains the benefits from the use of her/his human capital (which would not be the case if you were to buy a slave). Crucial to the difference between the trade in goods and the movement of people is that the goods which are traded internationally -- tires, beer, and autos, for example -- can be shipped separately from the persons who produce them. But the services that individuals perform -- in driving and in dentistry, say -- cannot be moved unless the provider of the service moves (though modern communications does make more possible the rendering at a distance of services such as computer programming, which will reduce the difference in wages between computer programmers in rich and poor countries.) The Size and "Quality" of Immigration (Chapter 3). Contemporary immigration is not high by U. S. historical standards. There were 800,000 legal immigrants in peak-year 1980, whereas near the turn of the century, immigration substantially topped the million mark in many years. And the burden of absorbing immigrants was much greater in those earlier times relative to the population size. For instance, immigrants who arrived between 1901 and 1910 constituted 9.6 percent of the population; those who arrived between 1961 and 1970 constituted only 1.6 percent, and between 1971 and 1980 only 2.0 percent. Or consider this: In 1910, 14.6 percent of the population was foreign-born; in 1970 only 4.7 percent had been born abroad, and in 1980 only 6 percent, or approximately 1 person in 17, and this includes those who have been here many years. Amazingly, this "country of immigrants," as the politicians often put it, has a smaller share of foreign- born persons than countries thought to be far more "homogenous", including Great Britain, Switzerland, France, and even Sweden. As to the "quality" of immigrants, the central economic fact now--and also throughout U.S. history--is that, in contrast to the rapidly aging U.S. population, immigrants tend to arrive in their 20s and 30s, when they are physically and mentally vigorous and in the prime of their work lives. Immigrants have about as much education as do natives, on average, and this was true even at the turn of the century. Furthermore, immigrants are disproportionally professional and technical persons, a great benefit to the U.S. Behavioral Characteristics of Immigrants (Chapter 4). Immigrants tend to have especially desirable behavioral characteristics from the economic point of view. Compared to natives, their rate of participation in the labor force is higher, they tend to save more, they apply more effort during working hours, and they have a higher propensity to start new businesses and to be self-employed. They do not have a higher propensity to commit crime or to be unemployed, and (for better or for worse) their fertility rate is not higher. The Effects of Immigrants Through the Public Coffers-- Transfers and Taxes (Chapters 5 and 6). Analysis of a large Census Bureau survey shows that, contrary to common belief, immigrants do not use more transfer payments and public services than do natives; rather, they use much smaller amounts in total. The services that catch the public eye are welfare and Supplemental Security, unemployment compensation, Aid to Families with Dependent Children, and food stamps. Comparing immigrant and native families of similar education and age, there is almost no difference in usage levels. The costs of schooling are somewhat higher for immigrants after the first few years in the U. S., because their families are younger than native families, on average. But when we include public retirement programs-- Social Security, Medicare, and the like--immigrant families on average are seen to receive much less total welfare payments and public services than do average native families. Summing the categories we find that the average immigrant family received $1,404 (in 1975 dollars) in welfare services in years 1-5, $1,941 in years 6-10, $2,247 in years 11-15, and $2,279 in years 16-25. In comparison, natives' receipts averaged $2,279, considerably more than the immigrants got during their early years in the U.S. Furthermore, these early years are especially relevant because rational policy decisions weigh the distant future less heavily than the near future. The costs of Social Security dominate the entire system of transfers and taxes. Natives get a windfall from immigrants through the Social Security mechanism. By the time the immigrant couple retires and collects, the couple typically has raised children who are then contributing Social Security taxes and thereby balancing out the parents' receipts, just as is the case with typical native families. In this way there is a one-time benefit to natives because the immigrants normally do not arrive accompanied by a generation of elderly parents who might receive Social Security. Admitting additional immigrants may be the only painless way for the U.S. to ease the trade-off between Social Security benefits and taxes that now cramps the nation's economic policies. If immigrants paid relatively little in taxes, they might still burden natives by way of the public coffers, even with less use of welfare services by immigrants than by natives. We lack direct information on taxes paid, but from data on family earnings we can estimate taxes tolerably well. Within three to five years after entry, immigrant family earnings reach and pass those of the average native family, due mainly to the favorable age composition of immigrant families. The average native family paid $3,008 in taxes in 1975. In comparison, immigrant families in the U.S. 6 to 10 years paid $3,359, those in the U.S. 11 to 15 years paid $3,564, and those in the U.S. 16 to 25 years paid $3,592. These are substantial differences which benefit natives. Taken together, the data on services used and taxes paid show substantial differences in favor of natives. If we assume that 20% of taxes finance activities that are little affected by population size--for example, maintaining the armed forces and the Statue of Liberty--the difference is an average of $1,354 yearly for the years 1-5 in the U.S., and $1,329, $1,535 and $1,353 for years 6-10, 11-15 and 16-25 respectively, in 1975 dollars. These are the amounts that natives are enriched each year through the public coffers for each additional immigrant family. Evaluating the future stream of differences as one would when evaluating a prospective dam or harbor, the present value of a newly-arrived immigrant family discounted at 3% (inflation adjusted) was $20,600 in 1975 dollars, almost two years' average earnings of a native family; at 6% the present value was $15,800, and $12,400 at 9%. Illegals such as the Mexicans who cross into the U.S. get little in welfare services due to their status. Representative estimates of the proportions of these illegals using such services are: Free medical, 5%; unemployment insurance, 4%; food stamps, 1%; welfare payments, 1%; child schooling, 4%. Practically no illegals receive Social Security, the costliest service of all. But 77% of illegal workers paid Social Security taxes, and 73% had federal income tax withheld. Even more telling are the cost-benefit studies which show illegal immigrants paying five to ten times as much in taxes as the cost of the welfare services which they use. The Use of Existing Physical Capital by Immigrants (Chapter 7). New immigrants employ existing plant and equipment capital when they start work, and they use social "demographic" capital such as schools and hospitals previously paid for by natives. Additional capital is required to accommodate the additional people. We wish to know how this use of physical capital by immigrants affects the pockets of natives. The question is best broken down into three kinds of capital--privately-owned industrial plant and equipment, publicly-owned plant and equipment, and publicly-owned demographic capital. If the private sector of the economy were like the government sector--where a worker's pay may be assumed to equal the full value of what the workers produce, with nothing left over for the owner of the capital--then capital dilution would indeed lower average native income. But in the private sector, theory suggests that an additional worker implies higher earnings by the firm's owners about equal to the loss of earnings by other workers. This tradeoff leaves overall native per-person income in the private sector roughly unchanged (a slight gain to natives) by an additional immigrant. This tradeoff implies, however, that "workers" suffer as "capitalists" gain. That is, to the extent that the "classes" are separate, there is a transfer from workers' pockets to owners' pockets. But in fact much of our private capital is owned indirectly by "workers" through pension funds, and those same workers obtain part of the gross profit (called "return to capital") by way of the taxes paid on interest and dividends. Hence the loss to the "worker class" is not as great as it may seem to be at first glance. A capital-dilution effect of some importance occurs in the use of government-owned plant and equipment. In this case there are no benefits captured by the native owners of the capital as there are in the private sector, and hence workers (including immigrant workers) receive in their salaries all the returns to capital. About 8% of immigrants work for the government, which translates into a loss to natives equal to about 2% of all immigrants' income. Though not trivial, this sum is about the same size as the gain to natives by way of the returns to private capital discussed just above. (We might also note, however, that this 8% of immigrants working for the government is only about half the native rate, a difference which has various benefits to natives.) As to the public capital used by immigrants, we must be concerned with the additional capital outlays needed to equip immigrants--the extra schoolrooms, hospital beds, firehouses, and the like. Not relevant is use of public goods that does not affect natives' use or pocketbooks--looking at the Washington Monument, or riding on a lightly used interstate highway. Furthermore, to a considerable extent we are on a pay-as-you-go basis with respect to capital expenditures: The debt service on past public borrowings covers much of the outlays on new capital. Therefore, through their taxes, immigrants pay "rent" on public facilities. This is an important additional reason why Malthusian capital dilution is not a crucial problem. For perspective, the negative effect of immigrants upon natives' incomes through capital dilution, mostly for demographic capital, is about one-fourth the size of the positive effect through taxes-and-transfers. Effects of Immigrants Upon Natives' Human Capital Utilization: Technology and Productivity (Chapter 8). Though the direct effect upon industrial productivity is hard to nail down statistically, in the long run the beneficial impact upon industrial efficiency of additional immigrant workers and consumers is likely to dwarf all other effects. Some of the productivity increase comes from immigrants working in U.S. industries and laboratories that are at the forefront of world technique. American citizens benefit along with others from a contribution to world productivity in, say, genetic engineering that immigrants would not be able to accomplish in their home countries. Also, the presence of more immigrants means that there are more working persons who will think up productivity- enhancing ideas. Other increases in productivity due to a larger population-- and on these which we have solid evidence--come from increased production through learning-by-doing, together with other gains from larger industry scale. Also, increasing the number of customers and workers increases investment, which brings more new technology into use, due to immigrants swelling the population. There are also less direct effects through human capital changes. The technical level of the persons that a native--call her/him Alpha--works with may affect Alpha's productivity in several conflicting fashions. First, if immigrants have less skill on average than do natives in the same occupation with the same amount of education--and this is likely because people in poor countries almost necessarily have less skill than do persons in the same occupation in rich countries--then even if immigrants who come have average or even better-than-average education, enough of them might drag down the skill levels of the natives they work and interact with. In other words, if there is a huge flood of immigrants from Backwardia to Richonia, Richonia will become economically similar to Backwardia, with loss to Richonians and little gain to immigrants from Backwardia. The effect may therefore be expected to depend in non-linear fashion upon the proportion of immigrants in such fashion that a few immigrants from Backwardia may have a beneficial effect for other reasons, whereas many may have a deleterious effect for the reason at hand. So even if some immigrants are beneficial, a very large number coming from poorer countries (that is, numbers that make new immigrants a significant percentage of the work force, say 10%, 20%, or 30% having come within 5 or 10 years) may have the opposite effect. Second, and working in the opposite direction, is the stimulative cross-pollenization effect that arises from encounters among people who have different ways of doing things. Though I have not found any systematic evidence on the matter, there is much anecdotal evidence in contemporary society and throughout history that this hybridization effect is important. It should be noted, however, that as with the learning- externality effect, a given immigrant may be expected to make more of a contribution in this fashion when he or she is one of only a few rather than one of many immigrants, because the essence of this effect is that the immigrant be different than those persons he/she works with. Third, the technical "quality" of casual associations that occur away from the own workplace and throughout the society affect a persons's skill and productivity. If a large number of technically-deficient immigrants were to arrive in a town or country, the productivity of the natives there would be likely to be negatively affected by this influence. Effects Upon Natural Resources and the Environment (Chapter 9). Immigrants are frequently said to cause a natural resource squeeze for natives. Much of that proposition is demonstrably bunkum. The water and food ingested in the U.S. have been improving in past decades by every reasonable measure of quantity and purity, though the fact is too-little known. The air in the U.S. also has been getting less polluted. And over the long run, natural resources have been getting less scarce rather than more scarce, as indicated by the trends in the fundamental economic measure of cost. Additional people do increase resource demand and prices in the short run. But in the longer run, when the system has had a chance to find new sources and substitutes, the result is that resources are typically more available and cheaper than if the temporary shortages had never arisen. An Integrated Model of Aggregate Effects (Chapter 10). Using a simple macro-model and what seem to be reasonable parameters, I estimate the difference between natives' incomes with or without immigrants. When looked at by natives as an investment, similar to such social capital as dams and roads, an immigrant family is an excellent investment worth somewhere between $15,000 and $20,000 to natives, even calculated with relatively high rates for the social cost of capital. (This is in 1975 dollars, to be compared with mean yearly native family earnings of about $11,000 in that year.) Labor-Market Effects of Immigrants (Chapters 11 and 12). The most politically powerful argument against admitting immigrants always has been that they take jobs held by natives and thereby increase native unemployment. The logic is simple: If the number of jobs is fixed, and immigrants occupy some jobs, then there are fewer jobs available for natives. In the immediate present, the demand for any particular sort of worker is indeed inflexible. In theory, therefore, additional immigrants in a given occupation must have some negative impact on wages and/or employment among people in that occupation. There must even be some additional general unemployment while the economy adjusts to additional workers. But we want to know whether the effect will be huge or trivial. To guage the magnitude of the effect,a model based on queuing theory is constructed, using available estimates on such factors as the average length of time persons who lose jobs remain unemployed. This line of reasoning suggests that, all else equal, an additional immigrant is likely to cause well less than two months of additional unemployment of natives. Also discussed in the theoretical section is a model by Harrison suggesting that, to the extent that immigrants consume and purchase even before they go to work, they decrease native unemployment, even if immigrant unemployment is observed to be much higher than native unemployment. Several recent studies have recently tackled the matter of "displacement" empirically using a variety of approaches. No study has found across-the-board unemployment caused by immigrants, either in the U.S. as a whole or in particular areas of relatively high immigration. And effects on particular groups are surprisingly small or non-existent, even groups such as blacks and women in California seemingly at special risk from Mexican immigrants. In short, immigrants not only take jobs, they make jobs. They create new jobs indirectly with their spending. They also create new jobs directly with the businesses which they are more likely than natives to start. Concerning wages now: Additional persons outside of native Alpha's occupation--especially persons with different levels of education--improve Alpha's earning situation because they are complements to her; if, for example, Alpha is a person with much education and skill, she can therefore be better off if there are additional immigrant people of low skill available, as a trained surgical nurse may be more productive if there are additional less-skilled helpers available. Similarly, if Alpha has few skills, Alpha benefits from having additional immigrants of high skill enter the country. Additional persons in Alpha's own occupation, however, drive down Alpha's earnings due simply to additional competition. Evidence concerning both the competitive and complementary effects on wages suggests that the effects are small, at least in the U.S., but competitive effects are observed to drive down some natives wages--as immigrant physicians worsen the economic position of native physicians (though other persons benefit from lower prices). Effects Upon the Income Distribution (Chapter 13). There is no evidence that immigration widens the income distribution in the United States. Effects Upon Sending Countries and the Rest of the World (Chapter 14). This topic is touched upon briefly because it is outside the scope of the book's main interest. The effect of international migration upon the world as a whole is unambiguously positive because it comprises the movement of factors of production from lower-value to higher-value utilizations. The effects upon the sending countries are theoretically mainly positive, according to conventional static considerations, but such benefits do not seem likely to be realized. Illegals, and Guestworker Programs (Chapter 15). Illegal immigrants tend to have lower-than-average amounts of human capital, and therefore they increase the competition that native unskilled workers face. But the damage to the latter group is far less than is popularly imagined. And the overall effect of the illegals is positive in every manner of influence examined here. Particularly noteworthy is that the immigrants use very small amounts of public services, as mentioned above, both because of their favorable age distribution and because they are afraid of apprehension if they attempt to obtain services. At the same time they pay income and Social Security taxes many times the cost of the services that they use. Illegal immigrants are not otherwise discussed because they are necessarily quite different than legals in their economic behavior as well as in their background characteristics. Policy Recommendations (Chapter 16). A variety of admission-selection policies are evaluated with respect to their effect upon the standard of living of natives. Policies that discriminate on the basis of economic characteristics, and especially a system such as allocating admission by auction that self-selects immigrants according to productive economic characteristics, are especially recommended. The most general policy issue concerns the total number of immigrants. Taking in immigrants at a rate equal to, or even far above, our present admission rate improves our average standard of living, on balance. American citizens even do well while doing good when admitting refugees. Rather than being a matter of charity, we can expect our incomes to be higher rather than lower in future years if we take in more immigrants. Therefore, increasing the total immigration quota is recommended. Unlimited immigration is not discussed here as a policy alternative for the U.S. The reasons are as follows: First and most important, we have little understanding of how many people would choose to immigrate in the shorter and longer run if the door were completely open. At one pole of possibilities is the common fear that U.S. natives would be inundated by tens of millions immigrants per year, quantities of immigrants proportionately as great as, or even much greater than, the turn- of-the-century rates, which were about 6 times as great as in the 1970's decade. At the other pole is the relatively small number of people--not much more than one million, a large proportion of whom migrate for family rather than purely economic reasons--who are currently registered as applicants for permission to enter the U.S. Whereas fear on the part of natives pushes upwards the former estimate, the latter figure is pushed downwards by the expectation by many would-be immigrants that it would be a waste of time to apply because admission would not be obtained or would take so long to obtain as to not be worth considering. This is a wide range, and hence any estimate of immigration under an open- door policy must be very uncertain. And we have even less basis for predicting how the mix of immigrants by education and skill would change were immigration to be allowed without limit. A second reason for not discussing unlimited immigration is that we lack empirical evidence to help estimate at which point some of the economic effects that are positive in the broad neighborhood of present admission levels might become negative at immigration rates several multiples of present levels. One such possibility is negative human-capital externalities if immigrants were to lower the skill level. But there are no signs of these constraints operating at present admission levels. A policy of admitting people selectively by economic characteristics such as education would avoid the key theoretical constraint, and would have enormous positive impacts on native income, but whether this is ethically or politically acceptable is another question. Under such circumstances prudence would seem to suggest that any alterations in policy be stepwise and sufficiently close to present levels so that we can feel our way--say increments of a doubling in rate at any one time rather than a greater multiplication of it. Third, unlimited immigration is unlikely politically, and is a red flag that upsets and arouses many. Literary prudence therefore seems to dictate avoiding discussion of the matter. 86-80 Conclu17 1/4/88 FOOTNOTES 1It would be a grave error, however, to believe that non- economic issues are not of enormous importance in the U.S., too. Appendix B shows how powerful an issue this was in years past when people felt free to speak of it publicly. Nowadays the references are more oblique, usually phrased in terms of other people's thinking rather than the writer's, as, for example, this introduction to testimony before the Joint Economic Committee by Governor Richard D. Lamm of Colorado: [M]assive immigration involves serious and profound risks. Ethnic, racial and religious differences can become a wedge; they can grow and eventually splinter a society. (May 29, 1986, p. l.) 2More generally, the literature in this century on the economics of population is not well-developed. In previous centuries there was much greater interest by economists, however. It was the economics of population that led Petty, perhaps the first great economist, to the study of economics in general. As Hayek notes: [T]he science of economics may well be said to have begun when in 1681...Sir William Petty...[became] fascinated by the causes of the rapid growth of the city of London. (Mimeo, Chapter III, pp. 13, 13a) 86-80 Conclu17 1/4/88