4
Immigration's Effects on Jobs and Wages: First Principles

What impact does immigration have on the U.S. economy? Economic theory points to possible effects on the employment and wages of domestic workers, U.S. trade with other countries, the size and growth rate of the economy, and the prices that Americans pay for goods and services. Given the broad scope of these potential effects, this chapter and its companion, Chapter 5, necessarily cover wide ground. The evidence presented in Chapter 4 relies on theoretical insights of the likely effects, whereas Chapter 5 offers a comprehensive empirical documentation of the actual effects.

Our theoretical presentation in this chapter relies on a ''first principles" discussion of immigration's likely impact on domestic labor markets. This treatment highlights the main insights from economic analysis about the effects of immigration. While our primer is written principally for noneconomists, economists may also be interested in the panel's collective judgment (with which some may disagree) about what the essential truths are. In particular, the chapter is not offered as a technical contribution to an extremely rich and growing literature in economics, from which we have borrowed extensively. Moreover, the panel decided not to cover all possible theoretical scenarios, from which an array of alternative outcomes becomes possible. Rather, we believe our most important task is to provide its best professional judgment about which models accurately capture the most salient impacts of immigration on labor markets.

To highlight these essential insights, we start with a very simple model to demonstrate that, on balance, immigration benefits the American economy. This model is then extended to provide a richer set of implications about the distributional impacts of immigration on the wages of domestic workers, the adjustments in the location of domestic employment, and the effects on the prices of domesti-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 135
--> 4 Immigration's Effects on Jobs and Wages: First Principles What impact does immigration have on the U.S. economy? Economic theory points to possible effects on the employment and wages of domestic workers, U.S. trade with other countries, the size and growth rate of the economy, and the prices that Americans pay for goods and services. Given the broad scope of these potential effects, this chapter and its companion, Chapter 5, necessarily cover wide ground. The evidence presented in Chapter 4 relies on theoretical insights of the likely effects, whereas Chapter 5 offers a comprehensive empirical documentation of the actual effects. Our theoretical presentation in this chapter relies on a ''first principles" discussion of immigration's likely impact on domestic labor markets. This treatment highlights the main insights from economic analysis about the effects of immigration. While our primer is written principally for noneconomists, economists may also be interested in the panel's collective judgment (with which some may disagree) about what the essential truths are. In particular, the chapter is not offered as a technical contribution to an extremely rich and growing literature in economics, from which we have borrowed extensively. Moreover, the panel decided not to cover all possible theoretical scenarios, from which an array of alternative outcomes becomes possible. Rather, we believe our most important task is to provide its best professional judgment about which models accurately capture the most salient impacts of immigration on labor markets. To highlight these essential insights, we start with a very simple model to demonstrate that, on balance, immigration benefits the American economy. This model is then extended to provide a richer set of implications about the distributional impacts of immigration on the wages of domestic workers, the adjustments in the location of domestic employment, and the effects on the prices of domesti-

OCR for page 135
--> cally produced goods and services. This extension also illustrates how trade with other countries modifies the economic impacts of immigration. Our initial theoretical model is static—that is, it examines immigration's impact at a moment in time. This issue is addressed in the second half of this chapter, which discusses the channels through which immigration may affect economic growth. Chapter 5 focuses on the empirical evidence concerning the role of immigrants in the labor market. The issues covered range over the changing relative economic status of immigrants, their ability to assimilate economically, the effects of immigration on the wages and employment of native-born workers, and the impact of immigration on the prices of goods and services. First Principles: Labor Market Effects of Immigration Whether native-born Americans gain or lose, immigration most directly affects the welfare of immigrants themselves. Immigrants come with the expectation that they will gain from immigration. If they had not felt that they would gain, they were free not to immigrate. Economic betterment is only one of many reasons why immigrants come here. Some may come expecting economic loss, but this must be offset by higher perceived gains in other things they value, like political freedom or reunion with their families. These gains are all before the fact. It may turn out that some immigrants are disappointed with life in their new country and some who fail to realize a gain return to their country of origin.1 Beyond the immigrants themselves, their immigration may have implications for the economic well-being of those who remain in the sending countries. Because immigration reduces labor supply in those countries, the income of other workers rises and the income of other factors of production falls there. But such effects are beyond the scope of the panel's work and are not dealt with in this report. Baseline Analysis The easiest way to see how immigration affects economic outcomes is to take a very simple and highly aggregate view of the economy as a system that combines two inputs—one for which immigrants are good substitutes, and the other for which immigrants are complements—to produce a single final good, which we can think of as gross domestic product (GDP).2 In the very simple 1   Not all of those who leave should be seen as disappointed immigrants. Many who come and leave (the sojourners) may have fully foreseen—even planned—that they would eventually return home. 2   These inputs are also assumed to have fixed supplies (completely inelastic supply curves). In addition, we assume for the moment that the production technology is constant returns to scale (for example, a doubling of all inputs leads to a doubling of output). These assumptions are relaxed below.

OCR for page 135
--> model, there is no trade between nations, so America consumes what it produces. Even this highly stylized model illustrates fundamental points about the effects of immigration on the domestic economy. Later, we relax some of these simple assumptions to obtain an even richer set of theoretical implications about the effect of immigration on the domestic economy. This baseline analysis focuses first and foremost on immigration's direct effect on the economy through the addition of workers to the labor force. At the most basic level, immigration increases the supply of labor in the economy. More labor means more goods and services being produced, so that national output (GDP) rises. Immigration also affects the prices of the inputs that are used to produce these goods and services. Those inputs for which immigrant labor substitutes will suffer as the prices of their services fall. Simply put, "substitutes" means two things that are very similar to one another. As a homely example, red apples and green apples are almost perfect substitutes, so that an increase in the number of red apples would not only reduce the price of red apples, but also simultaneously lower the price of green apples by about the same amount. In the context of immigration, whereas we shall see many immigrants are unskilled laborers, the strong presumption is that immigrants are substitutes for domestic unskilled labor.3 Therefore, an increase in the number of immigrants will generally decrease the wages of domestic unskilled workers. Immigrants are not substitutes for all domestic workers. A disproportionate number of immigrants are low-skilled relative to native workers, and so tend to be poor substitutes for workers other than the low-skilled—that is, they do not do the same things at all. In the jargon of economics, two factors that are not substitutes are called "complements." For a simple example, think of supervisors and production workers. Suppose that, for every 50 production workers, we need one supervisor. If we increase the number of production workers, we will need more supervisors and their wages will rise. An increase in the number of immigrants, then, will raise the wages of those domestic workers who are their complements. The common presumption is that skilled domestic workers are complements for immigrants, so that an increase in the number of immigrants will raise the wages of domestic skilled labor. Capital may also be a complement to immigrant labor, although the evidence on the complementarity of unskilled labor and capital is more ambiguous than that of skilled and unskilled labor. In summary, an increase in immigration flows will lead to higher incomes for productive factors that are complementary with immigrants, but lower incomes for factors that compete with immigrants. This very simple model implies that an increase in immigration will lead to an increase in national output, an increase in the income of complementary fac- 3   We are considering the effects of the entry of new immigrants. Domestic labor includes both native-born workers and foreign-born workers who immigrated to the United States in the past.

OCR for page 135
--> Figure 4.1 Effects of an influx of immigrants on a simple model of the labor market. tors, and a reduction in the income of substitute factors. Since immigrants to the United States are disproportionately low-skilled workers, people usually think of the substitute input as unskilled labor and the complementary input as capital or skilled labor. But large influxes of immigration in some highly skilled workers, such as mathematicians and nurses, imply that some highly skilled natives will also see their wages or job opportunities worsen with immigration while capital or less skilled labor gains. To sum up, immigration makes national output go up, some domestic workers suffer, and other workers benefit. Although it clearly begets winners and losers, can we make an overall statement that the winners win more than the losers lose, so that in the aggregate native-born Americans are better off? With plausible assumptions about factor supply and returns to scale, we show that immigration produces net economic gains for the native-born. To understand this perhaps surprising result, we use a simple diagram (Figure 4.1). In the simple world portrayed by this diagram, we have two types of domestic workers: those who are perfect substitutes for immigrants (unskilled labor) and those who are complements (skilled labor).4 Only one good is produced (GDP), and the numbers of unskilled and skilled domestic workers are fixed. Figure 4.1 plots the demand curve (CF) for domestic unskilled workers. Before immigration, there are S domestic unskilled workers, who are all paid 4   In particular, in this simple world there is no capital, so that workers receive all the income produced by selling the single good.

OCR for page 135
--> a wage W0 (the wage that equates demand and supply), so that the total amount that domestic unskilled workers are paid is S times W0, or the area OBDG. Although we deal explicitly only with unskilled workers in this diagram, we can also determine how much skilled domestic workers are paid. To see this, note that the height at each point along the demand curve is the value of the extra national output produced by another unskilled worker.5 Therefore, total national output (GDP) is the area underneath the demand curve up to S unskilled workers (OCDG) so that the remainder (BCD) is the amount paid to domestic skilled workers. Now let new immigrants come into this country, increasing the supply of all unskilled labor in the work force to S + I. The new wage that equates the demand and supply of unskilled labor falls to W01; that is, the wage of substitute domestic unskilled workers falls to W01 . Unskilled domestic workers are clearly worse off. Since the total amount that all unskilled domestic workers are paid falls to OAKG, the domestic unskilled workers lose ABDK as a result of immigration. Unskilled immigrants are paid the same wage as domestic unskilled workers so, as a group, immigrants receive the area GKEH. What about skilled domestic workers? Before immigration, they received the area BCD, but what do they get now? Once again, we can calculate their incomes as a residual. With these new immigrants added to the workforce, total national output (GDP) will rise, so that it now equals the area under the demand curve up to the total number of unskilled workers, S + I. Instead of their preimmigration incomes of BCD, domestic skilled labor now receives the area ACE (everything that isn't paid to either unskilled domestic workers or immigrants). Total GDP is now the area OCEH, so that the value of domestic output has increased by the area GDEH. But new immigrants get only the rectangle GKEH, so that, on net, domestic workers must gain by the size of the triangle KDE. Immigration thus raises national output and national output per domestic worker. One way of seeing that the native-born must gain from immigration in this simple model is to recognize that new immigrants help produce new goods and services, but they are paid less than the total value of these new goods and services. The rest goes to domestic residents, who collectively are better off than before, by the triangle KDE. 6 Figure 4.1 also illustrates that, although the net gain is positive domestically, some workers may lose and others may gain. In fact, although domestic unskilled workers lose ABDK, domestic skilled workers gain ABDE. The area in common 5   Technically, the height is the marginal product of unskilled labor. 6   The fact that immigrants are not paid the total value of the new goods and services does not mean that immigrants are in any sense "exploited." Some of the extra output produced should be credited to native workers, who are now being worked more intensively. For example, if native workers were only supervisors and immigrants were only production workers, each native worker would be supervising more immigrants.

OCR for page 135
--> is the rectangle ABDK, which is simultaneously (and equally) a loss to domestic unskilled workers and a gain to skilled domestic workers. Therefore, although immigration yields a positive net gain to domestic workers, that gain is not spread equally: it harms workers who are substitutes for immigrants while benefiting workers who are complements to immigrants. Most economists believe that unskilled domestic workers are the substitutes, so their wages will fall, and skilled domestic workers are complements, so their wages will rise. The potential size of this redistribution among the native-born compared with the size of the net national gain is one reason why our national political debate stirs such strong emotions. As Figure 4.1 indicates, at least in this simple economy, the magnitude of the redistribution rectangle is proportional to the number of unskilled domestic workers. In addition to being a triangle (which alone would make it half as large as the redistribution rectangle), the size of the net national gain is proportional to the immigrant flow (generally a much smaller base). Consequently, the redistribution among domestic workers could be large relative to the positive gain to the nation as a whole. There is a good argument that specific policies should be based only on whether a national gain or a national loss exists, and that we should use other policies to achieve our desired income distribution. However, the magnitude of redistribution relative to the aggregate national gain is clearly one important element in the national political debate.7 From the diagram, it is also clear that the magnitude of the net gain to the domestic economy may depend on the magnitude of the loss to substitute domestic labor. The area of the triangle in Figure 4.1 that represents the net gain to the domestic economy is equal to one-half multiplied by the number of new immigrants times the fall in the wages of domestic unskilled workers. Since the number of immigrants is fixed, the bigger the drop in the wages of domestic unskilled labor, (W0 - W01), the larger is the net rise in the incomes of domestic workers. If the wage of domestic unskilled workers did not fall, no domestic worker (unskilled or skilled) would gain or lose, and there would be no net domestic gain from immigration. In this case, the value of all the new domestic output immigrants produced would go to the immigrants themselves. There is a direct correspondence between the fact that some domestic workers suffer wage reductions and the fact that we gain as a nation. This wage reduction is, in fact, the reason that the nation as a whole gains from immigration. This simple point is often lost in the voluminous empirical literature, which 7   It is the case that the gain from immigration could compensate the losers since the size of the total gain exceeds the size of the total loss. This is the concept of Pareto optimality. A policy is a Pareto improvement if it results in a net gain. Such a policy is an improvement in the sense that it would be possible to compensate those for whom the policy produces a loss by redistributing income from those who receive a gain, and hence all parties could potentially be made better off.

OCR for page 135
--> searches long and hard to find those native-born Americans who experience wage reductions that are then labeled the "cost" of immigration. Although these wage declines are real losses to these affected workers, they are also the source of a national "gain" from immigration. The gains and losses of immigration are therefore closely intertwined. As long as immigrants substitute for some natives, the larger the loss to those natives, the greater the benefits of immigration to the aggregate domestic economy. Similarly, if immigration has little effect on the aggregate economy, it must have little adverse effect on native workers for whom the immigrants are good substitutes. This analysis focuses attention on the extent to which natives and immigrants are similar and whether they are substitutes or complements at workplaces. If immigrants have skills that substitute for those of some native workers, those native workers will lose from immigration. If immigrants have skills that complement those of other native workers, those native workers will gain from immigration. In the extreme, if natives and immigrants had utterly different distributions of skills—if, say, all immigrants had less than a grade school education and all natives were high school graduates or better—immigrants might be complements to all native labor. Then immigration would improve the economic position of all natives and harm none. Immigrants might also take jobs that no native would or could take. This would enable the economy to produce goods or services that it would otherwise not produce. Finally, to close the logical possibilities, assume for a moment that the immigrants who arrive are exactly the same as the domestic workers who are already here. If immigrants have exactly the same skill distribution as domestic workers (the same fractions are unskilled and skilled workers) and if they have brought sufficient capital with them to maintain the U.S. capital/labor ratio, then natives will neither benefit nor lose from immigration. In this case, all inputs and national output will increase by the same amount and the wages of all workers will remain constant.8 It is only because immigrants and native workers differ from one another that immigration yields a net national gain. These differences between natives and immigrants, which may well be a legitimate source of concern about the ability of immigrants to assimilate socially and culturally, are the very reasons why the nation gains economically from immigration. This analysis directs attention at differences between the skill distributions of natives and immigrants in analyzing the economic effects of immigration. Table 4.1 gives the distributions of natives and immigrants by educational attainment in the 1990 Census of Population. Columns 1 and 2 show the distributions for the two groups for the United States as a whole, and Columns 3 and 4 the 8   This assertion depends on the assumption of constant returns to scale, which is relaxed below.

OCR for page 135
--> TABLE 4.1 Percentage Distribution of Immigrants and Natives by Educational Attainment, 1990, United States and California   United States California Years of Educational Attainment Native Immigrant Native Immigrant Less than 9 4.2 23.4 2.2 28.6 9-11 15.0 16.0 11.9 17.4 12 32.0 20.1 24.0 16.7 13-15 29.5 21.1 37.5 20.5 16 13.8 12.0 16.4 11.0 16+ 6.6 8.4 7.9 5.8   Source: Calculated from U.S. Bureau of the Census (1993). distributions for the two groups for California. The educational attainment of immigrants differs substantially from that of natives. Many among the immigrants have very low levels of education, particularly in California, and thus may be complementary with most American workers. Many among the immigrants have college degrees or better, and thus may compete with native college graduates and be complementary with less skilled Americans. Many also have 9 to 11 years of education and may compete with native high school dropouts and be complementary with more skilled Americans. But the salient finding in the table is the large proportion of immigrants with less than nine-years of schooling, 23.4 percent in the United States and 28.6 percent in California, who may be complementary with most native American workers. New immigrants lower the wages of groups for which they are close substitutes. The closest substitutes for newly arriving immigrants may well be prior waves of immigrants, rather than native workers. We show in Chapter 5 that, in terms of the jobs they do, newly arriving immigrants most closely resemble their immediate predecessors. Therefore, the bulk of the wage reduction induced by new immigrants may be concentrated on prior immigrant waves. In sum, the baseline analysis suggests that immigration raises national output and on net improves the economic well-being of the native-born. Immigration also redistributes income from workers who compete with immigrant labor to factors that complement immigrant labor. The analysis also implies that Americans benefit most from immigrants whose skills are very different from those of natives. Model Extensions This very simple model effectively highlights many of the fundamental implications about the economic effects of immigration, but other implications can be illustrated only by taking a broader perspective. These implications include

OCR for page 135
--> the effect of immigration on the prices of the goods we consume, the displacement of native workers (that is, the potential changes in where they are employed), and the relation between immigration and international trade. As before, we initially assume that the United States is a closed economy—that is, it does not trade with other countries, and that total supplies of domestic factors (skilled and unskilled labor) are fixed. The United States then still must consume only what it produces. The fundamental extension is that we now let the U.S. economy produce two goods instead of only one. Although this change may sound trivial, it allows immigration to alter the relative prices of goods and services, so that domestic consumers can now gain or lose depending on which prices change the most. It also allows domestic labor to choose the sector of the economy in which they will work, so that sectorial displacement of domestic workers is possible. Because the analytics of the model can quickly become quite complicated, we present only a brief summary of results in the text. Appendix 4.A presents a detailed model on the analytics of the gains from immigration in this two-good, two-factor case. In this simple economy, we now have two goods (call them X and Y) being produced efficiently with both unskilled and skilled labor. 9 Both goods need both types of labor, but they differ in the relative amounts of each factor they need. For convenience only, assume that good X needs a lot of unskilled labor relative to skilled labor, and good Y definitionally is the opposite. These factor proportions are crucial because they determine which sector immigration will affect more. Without immigration, and with no international trade, our hypothetical economy efficiently produces and consumes a certain amount of good X and a certain amount of good Y.10 What happens to this economy if there is an inflow of immigrants? Once again, assume that these immigrants are unskilled compared with domestic workers. This relative increase in the supply of unskilled labor will, as in the very simple economy, lower the wage of unskilled labor relative to the wage of skilled labor. As before, unskilled domestic labor suffers a loss and skilled domestic labor a gain from immigration. Because good X uses more unskilled labor relative to skilled labor, this increase in immigration will also lower the price of good X relative to the price of good Y. These changes in relative prices have a number of important consequences. Because the relative price of good X falls, domestic consumers will want to consume more X and less Y. Among native consumers, those who like good X the 9   As before, each good's production function is characterized by constant returns to scale. 10   This situation is depicted graphically in Appendix 4.A: Figure 4.A3, which plots the maximum amounts of X and Y that can be produced efficiently domestically. This situation is depicted as point A on this graph.

OCR for page 135
--> most—who consumed relatively more of it before immigration—will gain the most from its falling price, and those who like good Y the most will gain the least. The example is not fanciful. Immigrants are highly concentrated in certain sectors of the economy, so that relative price effects are possible. Many observers have argued that one of the primary effects of immigration is that it has reduced the cost of household service (cleaning house, caring for young children), a benefit that may be largely confined to the well-to-do. In Chapter 5, we go beyond mere speculation by providing solid empirical evidence on how widely concentrated or dispersed these price effects are across domestic consumers. With this expanded model, immigration has distributional effects: skilled domestic labor and domestic consumers of immigration-intensive products gain, and domestic unskilled labor and domestic consumers of goods not produced by immigrants lose or gain the least. But can we go beyond simply distributional gainers and losers to whether these cumulative domestic gains or losses are positive or negative in the aggregate? It turns out that we can, and, on net, the native-born will still gain from immigration.11 Domestic workers will shift their production to good Y, the good that they are better at, to take advantage of its higher relative price. One source of the native gain from immigration is that it allows the nation to use domestic labor more productively, specializing in producing goods in which we are relatively more efficient. In short, if immigrant labor can produce something for $3.00, what do we gain from producing it ourselves for $3.50? We can let immigrants produce it and pocket the 50 cents we save. Because of the shift in domestic production, some domestic workers, especially the less skilled ones, who had been working in industry X, may now have to move to industry Y. In addition to wage effects, immigration has "displacement" effects. Some domestic workers will be "displaced" by immigrants, in the sense that they will now have to work in a different industry. In the simple model, we are assuming that the process of displacement is costless, in that displaced workers will eventually find employment in the other sector. This is a good characterization of the long run, but in the short run adjustment does have costs. It may take time to find this new job, with all the anxiety associated with that search. Changing jobs may mean moving out of one's neighborhood, city, or even region, with a loss of family, friends, and familiar schools and churches. Many Americans who perceive themselves to be displaced by immigrants resent having to make this adjustment. "Perceive" is an important word in this sentence because an attribution problem emerges when it comes to 11   To see this visually, return to Appendix 4.A: Figure 4.A3. Before immigration, the economy was producing and consuming at point A. Immigration lowers the price of good X relative to good Y. Let this new relative price be depicted by the slope of the line L0L0. After compensating immigrants for their contributions, the domestic budget corresponds to all points on the line L0L0.

OCR for page 135
--> immigration. Some may associate their displacement with immigrants when the real causes lie elsewhere. Whatever its costs, more efficient domestic production is not the only gain from immigration. There is also the gain associated with specialization in consumption. Just as the presence of immigrants allows natives to specialize in production, it allows them to consume something different from what they can produce themselves.12 As a nation, we may be very good at producing good Y, but we really like good X. Immigration is one way we can have the best of both worlds; making what we are good at and also consuming what we like. The welfare gain to natives from immigration thus can be decomposed into two parts: the gain from shifting production toward more valuable activities that use the relatively more skilled native labor, and the gain in consumption toward commodities whose cost has fallen.13 In sum, the net welfare gains from immigration stem from two sources. By having immigrants specialize in the production of goods requiring a lot of low-skilled labor, it allows us to shift our domestic production toward those goods (Y) in which natives are relatively efficient (those that need a lot of skilled labor) and away from those that can be produced more cheaply by immigrants. The second component is the gain in consumption. Before immigration (and with no international trade), we could consume only that which we could produce domestically. Immigration breaks this rigid link between domestic consumption and domestic production, allowing us to produce goods of which we are relatively efficient producers and to consume those goods that conform to our tastes. The models thus far highlight the potential substitution of immigrant labor for some competing native workers and the possible redistribution among native workers. Such models may well exaggerate the harm and understate the economic benefits from immigration. Immigrants may not directly substitute for many domestic workers at all if immigrants produce goods and services that largely would not otherwise have been provided. For example, we might well not have any domestic textile industry (or any domestic workers, most of whom are unskilled, in it) if it were not for immigrants. In this case, the gains from immigration are not reflected in wages at all, but in the lower domestic prices of goods related to textiles. All domestic consumers gain in this case, including unskilled domestic labor. These purely domestic gains from immigration do not involve only substitutes for imports. As we shall see in the next chapter, immigrants are extremely 12   In Appendix 4.A: Figure 4.A3, domestic factors alone would produce at point C, but domestic consumers will consume at point E. 13   In Appendix 4.A: Figure 4.A3, the overall welfare gain to domestic workers is represented by the difference between U2 and U0. The movement from A to D represents the gain associated with shifting consumption, and the movement from D to E represents the gain associated with shifting production from A to C.

OCR for page 135
--> Immigration and Changes in Technology Observed patterns of migration underscore the importance of endogenous growth (that is, growth that can be manipulated). In particular, the simple framework with exogenous growth (that is, growth over which people have no control) implies that returns to skilled labor should be high when the level of skilled labor is low, since returns to all factors are decreasing. Yet we do not observe migration of skilled labor from rich nations with high proportions of this factor to poor nations with low proportions. This fact has been cited as one of the most telling conflicts between the exogenous growth framework and observed behavior (Lucas, 1988; Romer, 1994). In the case of the most highly developed nations, endogenous growth models suggest an avenue of influence of immigration on growth beyond the mix of factors in the simple framework. These models highlight the role of research and development in sustaining economic growth. Immigrants are important in many research and development sectors in the United States: they are disproportionately represented in many research laboratories, in the community of Nobel laureates recognized for work in the United States, in postgraduate programs in the sciences and engineering, and increasingly on the faculties of U.S. universities. The system of postgraduate education in the United States has long been central in this process. Widely recognized as the best in the world, this system brings in large numbers of graduate students from other countries, tries (usually successfully) to keep the best in the United States, and returns the others. Those who stay concentrate in research and development sectors important to sustaining economic growth, and those who return are among the most skilled in their countries of origin. Another mechanism through which immigration may affect growth involves immigrants who exhibit extraordinary entrepreneurial ability. Entrepreneurs, especially the successful ones, create wealth and jobs well beyond their numbers. They often supply the energy, ingenuity, ideas, and capital that are the sinews of economic growth. Although no direct systematic evidence exists on differences in entrepreneurship between natives and the foreign-born, some indirect evidence indicates the extent to which immigrants are self-employed. Table 4.3 lists rates of self-employment for the two groups in the 1990 census. It reveals essentially no difference in the probability of self-employment between immigrant and native-born workers. Self-employment varies considerably among source countries, running from a little less than 1 in 5 among European men to about 1 in 20 among Mexican women. With these data, however, it is difficult to argue that self-employment and any entrepreneurship associated with it mark a critical distinction between immigrants and the native-born. In addition, immigration is a very unlikely vehicle for increasing total national savings. Because immigrants still represent only about 8 percent of the population, savings differences between immigrants and the native-born would

OCR for page 135
--> TABLE 4.3 Self-Employment Rates by Region of Birth and Education Level   Percent Self-Employed Group Men Women Natives 13.1 7.2 All foreign-born 13.1 8.3 Foreign-born, by region Europe 18.4 10.2 Asia 15.9 9.8 Mexico 7.2 5.7 Other America 11.1 6.5 Other 11.2 7.1 Foreign-born, by education level 8th grade or less 9.6 6.8 Grades 9-12, no degree 11.8 8.5 High school degree or certificate 13.4 8.8 Some college 14.1 8.4 Bachelor's degree or greater 15.6 8.9   Source: Calculations based on reported class of worker at the time of the census, in sample of employed people ages 25-64 from the 1990 Census Public-Use Microdata Sample. have to be extremely large to alter aggregate savings.34 Although there exists little firm evidence about savings differences between immigrants and the native-born, it is more likely that immigrants actually save less than the native-born. Low-income families tend to save very little and the preponderance of immigrant families in the low-income deciles implies that their savings rate is likely to be low. The Impact of Immigration on Income, Wealth, and Growth The theory of economic growth identifies several features of the economy that are keys to determining the impact of immigration on income and wealth in the long run: the skill mix of the immigrant population, the ability to substitute 34   For example, assume that the native-born save 10 percent of their income and represent 92 percent of the population. If immigrants saved at twice the rate as the native-born (i.e., 20 percent, which is extremely unlikely), then national savings would rise to only 10.8 percent.

OCR for page 135
--> different labor skills and capital in production, the rate at which immigrants enter the economy, the rate of economic assimilation of immigrants, the response of investment in labor skills and physical capital to changes in factor returns, and the participation of immigrants in research and development. The evidence points to a substantial capacity of the economy to absorb immigrants. The rate of immigration can have important consequences for the absolute size of the economy (as documented elsewhere in this report), but it is not likely to have major long-run consequences for the distribution of income or the rate of capital formation. The very success of immigrants in assimilating economically across generations mitigates their potential to alter long-run growth rates. Conclusions To summarize the main findings of this chapter, as a nation, on net, we gain economically from immigration for several reasons. On the production side, immigration allows domestic labor to specialize in producing goods at which we are relatively more efficient. Specialization in consumption also yields a gain. Immigration thus breaks the rigid link between domestic consumption and domestic production. From this perspective, its effects are comparable to those of international trade. Because the two processes are so similar, when trade is relatively free, changes in the number of immigrants will affect the incomes of the native-born less than they would have without trade. Even when the economy as a whole gains, though, there can be losers as well as gainers among different groups of U.S. natives. The gainers are the owners of productive factors that are complementary with immigrants—native higher-skilled workers, whose incomes will rise. Those who buy goods and services produced by immigrant labor benefit, too. But there may also be losers: the less skilled native workers who compete with immigrants and whose wages will fall. To the extent that immigrants specialize in activities that otherwise would not have existed domestically, immigration can be beneficial for all the native-born. In this case, there is little substitution against native workers, and native consumers gain from the lower prices of these services. Overall, immigration is unlikely to have a huge effect on relative earnings or on gross domestic product per capita. Many other factors are far more critical to the U.S. economy than is immigration, including savings and investment and the human capital of U.S. workers. In the long-run, assuming constant returns to scale, immigrants can affect rates of economic growth only to the extent that they differ from the native-born—if, for example, they arrive with a different mix of skills from those possessed by the native-born. To affect growth rates, this difference between immigrants and natives must persist over each new generation. If the grandchildren of immigrants come to be just like the native-born, all that immigration does is

OCR for page 135
--> augment the population and the scale of the economy; it scarcely changes per capita income in the long run. References Baumol, W.J., S.A.B. Blackman, and E.J. Wolff 1989 Productivity and American Leadership: The Long View. Cambridge:MIT Press. Borjas, G.J. 1995 The economic benefits from immigration. Journal of Economic Perspectives 9:3-22. Borjas, G.J., R.B. Freeman, and L.F. Katz 1997 How much have immigration and trade affected the U.S. job market? Brookings Papers, forthcoming. Burnside, C.A., M. Eichenbaum, and F. Rebelo 1995 Capital utilization and returns to scale. NBER Macroeconomics Annual. Fukuyama, F. 1995 Trust: The Social Virtues and the Creation of Prosperity. New York: The Free Press. Grossman, G.M., and E. Helpman 1994 Endogenous innovation in the theory of growth. Journal of Economic Perspectives 8:2344. Hall, R.E. 1988 The relationship between price and marginal cost in U.S. industry. Journal of Political Economy 96(5):921-947. Hamermesh, D.S. 1986 The demand for labor in the long run. In O. Ashenfelter and R. Layard (editors), Handbook of Labor Economics. Amsterdam: Elsevier Science Publishers. 1993 Labor Demand. Princeton, NJ: Princeton University Press. Johnson, G.E. 1997 Estimation of the Impact of Immigration on the Distribution of Income Among Minorities and Others. University of Michigan, mimeo, February. Lucas, R.E., Jr. 1988 On the mechanics of economic development. Journal of Monetary Economics 22:3-42. Malthus, T.R. 1798 An Essay on the Principle of Population. First edition published 1798. New York: Random House (1960). Romer, P.M. 1994 The origins of endogenous growth. Journal of Economic Perspectives 8:3-22. Smith, A. 1776 An Inquiry into the Nature and Causes of the Wealth of Nations. Dublin: Whitestone. U.S. Bureau of the Census 1993 1990 Census of Population. Washington, D.C.: Department of Commerce. Varian, H.R. 1984 Microeconomic Analysis. New York: W.W. Norton and Company.

OCR for page 135
--> Appendix 4.A Immigration in the Two-Good, Two-Factor Model Consider a simple closed economy that produces two goods, x and y, using two factors of production, labor and capital. Let x and y be produced by constant-returns-to-scale production functions, F and G, respectively. We can write the following equations: x = F(LX, KX) = LX F(1, KX/LX) = LXf(kX) (1) y = G(LY, KY) = LY G(1, KY/LY) = Lyg(kY) (2) L = LX + LY (3) K = KX + KY (4) The Edgeworth box in Figure 4.A1 presents all feasible combinations of x and y that can be produced given this economy's endowments of labor and capital. Assume that good x is relatively intensive in its use of labor and that good y is intensive in its use of capital. The OXOY locus traces all the efficient combinations of production where the marginal rates of substitution between capital and labor are the same in the two sectors. Suppose that production takes place at point A. The capital-labor ratio in sector x is then given by the slope of R0. Similarly, one can connect the Oy with Figure 4.A1 Edgeworth box.

OCR for page 135
--> point A and take the slope of that radius to arrive at the capital-labor ratio in sector y. It is easy to see that, if we move production from A to, say, point C, the equilibrium capital-labor ratios will fall in both sectors. In fact, each point on the efficiency locus corresponds to different capital-labor ratios within each sector, with the range of possible capital-labor ratios being bounded by the available quantities of labor and capital. From equations 1 and 2 it follows that the marginal physical products are functions of the capital-labor ratio only: xL = f(kX) - kXf'(kX) and xK = f(kX) (5) yL = g(kY) - kYg'(kY) and yK = g'(kY) (6) By dividing the marginal physical product of labor by the marginal physical product of capital in both equations 5 and 6 and by taking the total differential, it follows that capital-labor ratios are strictly increasing in the marginal rate of substitution between capital and labor, independent of the endowments, d(kX)/d(w)= f'(kX)/(f''kX) + w) (7) and d(kY)/d(w)=g'(kY)/(g''kY) + w) (8) where ω = xL/xK = yL/yK Preferences determine where production will take place. The Y0X0 locus in Figure 4.A2 traces efficient production in y - x space. One can imagine a set of social indifference curves such that the highest possible level of welfare, given the budget as represented by the production frontier Y0X0, is obtained at point A. Both production and consumption take place at point A. What happens to this economy if we allow an inflow of immigrants? For simplicity, assume that the new immigrants bring no capital with them, so that they add only to the stock of labor in the economy. With x being relatively more labor intensive, immigration allows relatively more x to be produced, and the production frontier will expand to, say, Y1X1. If the price of x in terms of y, given by the tangency to the production frontier Y0X0 at point A, does not change, consumers (natives and immigrants) will have an excess demand for y and an excess supply of x. The relative price of x will have to fall to restore equilibrium. Preferences determine the size of that change. Thus, immigration has changed relative prices, lowering the price of the good that uses labor intensively in terms of the good that uses capital intensively. Suppose the new production point is given by point B. The question that

OCR for page 135
--> Figure 4.A2 Production possibilities with and without immigrants. arises is whether domestic workers are better off because of the immigration. To see that they are, first realize that the domestic budget is given by the tangency of the new price line at the old production frontier. That is, the difference between the two frontiers is made up of payments to the new immigrants. Given constant returns to scale, national income, Y, in terms of y must add up the payment to the factors of production. That is, Y = y + px = GL(Ly + IY) + GK K + p FL(LX + IX) + p FKKX (9) which, with p = GL/FL = GK/FK, simplifies to Y= GL(L + I) + GK K (10) where I represents new immigrant labor. Similarly, at the old production frontier national income at point C is given by Y*= G*L + G*K K (11) The difference between Y and Y* is equal to the compensation of new immigrant labor if and only if GL = G*L and GK = G*K. Ruling out factor-price-intensity reversals, relative prices at B and C are the same and imply equal marginal rates of substitution between capital and labor. Since the marginal rate of substitution uniquely defines capital labor ratios, which in turn define the marginal physical

OCR for page 135
--> Figure 4.A3 The gain from immigration in a model with two factors and two goods. products given in equations 5 and 6, it follows that the marginal physical products are equal at B and C and thus the difference between L0L0 and L1L1 gives the compensation to new immigrants. L0L0 thus corresponds to the budget for domestic workers after immigration has taken place. Domestic workers have shifted their production from A to C to take advantage of the higher relative price of good y. The presence of new immigrants allows domestic workers to specialize in production and to consume outside the old production frontier. Consumption and production are no longer constrained to be the same. At the new relative price, a higher social indifference curve can be attained and domestic workers thus experience an overall welfare gain. This welfare gain can be decomposed into two parts: the gain from shifting production toward more valuable activities, and the gain from shifting consumption toward commodities whose cost has fallen. In Figure 4.A3 the overall welfare gain is represented by the difference between U2 and U0. The movement from A to D represents the gain associated with shifting consumption, while the movement from D to the new consumption equilibrium at E represents the gain associated with shifting production from A to C.

OCR for page 135
--> Appendix 4.B Assessing the Labor Market Benefits and Costs of Immigration In standard models of immigration and trade, the income losses of natives who compete with immigrants are more than matched by the income gains of natives whose skills or capital complement those of immigrants. How large might these gains and costs be? To answer this question, economists use variants of the model shown in Figure 4.1. By aggregating groups of workers into a small number of complements or substitutes for immigrants; by assessing the extent to which these workers are complementary with or substitutes for capital or (what amounts to the same thing) of the responsiveness of factor prices to a change in immigrant labor supply; and by assessing the potential response of capital and labor supplies to immigration; one can put numbers to the labor market benefits and costs of alternative immigration flows. The basic framework for estimates is a production function Q= f(K, S, L), where Q is output, K is capital, S is skilled labor, and L is unskilled labor. Under constant returns the income received by residents is Q = FKK + FSS + FLL (1) where FK, FS, FL are the marginal products of capital, skilled labor and labor, respectively. Output is divided between capital, skilled labor, and unskilled labor according to their marginal products and magnitudes. If immigrants contribute more to the supply of unskilled than of skilled labor, as in the United States in the past three or so decades, then under virtually any plausible assessment of the effect of immigrants on the marginal productivity of skilled labor, unskilled labor, and capital, unskilled labor will lose and skilled labor and capital will gain. However, even the smaller number of immigrants who add to the nation's skilled labor can reduce the earnings of skilled labor as well, with the result that the sole beneficiary of immigration is capital. If increased labor supplies in the United States induce capital to move to the United States, however, U.S. capital will not gain at all: it will simply reallocate from other countries to the United States. Formally, let I be the increase in immigrant labor to the United States. Assume for simplicity that this increase consists entirely of unskilled labor (what matters is whether the immigrant supply consists disproportionately of skilled or unskilled labor). With a constant returns to scale production function, the change in the national income, Δ, accruing to current U.S. residents depends on the amount of their factor resources and the effect of the immigrant supply on factor prices:

OCR for page 135
--> ∆ = (∂fK/∂I)K + (∂fS/∂I)S + (∂fL/∂I)L (2) The effect of unskilled immigrants on the factor prices of capital and skilled workers will be positive, but it will be negative on the wages of unskilled labor. The economic analysis in Chapter 4 explained why (barring external diseconomies of scale) Δ will be positive. To turn this expression into a quantitative estimate of the effects of immigration on the income accruing to natives, we need estimates of the effects of immigrants on the factor prices, and/or on the marginal productivities of the factors. Economists have some knowledge of the interrelations among factors, but this knowledge is not hard and firm. Different analysts have made somewhat different assumptions about the parameters governing the interrelations, which yield somewhat different estimates of the magnitude of effects. Still, a general consensus has emerged that, overall, the net gains to the United States of immigration are rather modest measured relative to gross domestic product (GDP). Assuming no response of the supply of capital to immigrant-induced higher returns, Borjas (1995) estimates that the approximate 10 percent of the American workforce who are immigrants added roughly 0.1 percent (one-tenth of one percent) to the GDP accruing to other Americans—or about $7 billion in a $7 trillion economy. Under the same assumption, Johnson (1997) estimates that a hypothetical increase in unskilled immigrants by 10 million (roughly 8 percent of the U.S. workforce) would raise national output accruing to natives by just $2.5 billion-—or about 0.036 percent of GDP. Borjas et al. (1997) estimate a gain to natives of $9.1 billion, or about 0.13 percent of GDP. These are all small numbers, which would be hard to detect in any aggregate data, given the many factors that influence GDP. When capital responds to immigrant-induced higher returns, the gains accruing to U.S. natives are even lower. In this case, the gain to capital—the major complement to immigrant labor—draws in additional capital, so that the return to native capital is ultimately driven down. In terms of our equation, ∂fK/∂I = 0 Johnson (1997) estimates that the gain to natives in this case is a bare $1.1 billion, all of which accrues to skilled labor, given his assumption that immigrants were entirely unskilled. Borjas et al. (1997) estimate that the gain to natives is $3.5 billion when capital is completely elastic in supply. The gains accrue to skilled labor because immigrants are disproportionately unskilled. All of these assessments place the economic gains to the country from immigration in the same modest range of magnitude: from $1.1 billion to $9.5 billion in a $7 trillion economy. This consensus reflects what economists view as plausible values of the relevant elasticities of factor prices to changes in labor supplies, or of the economic parameters that govern an aggregate production function. To see how virtually all such calculations yield relatively modest estimates of the gains to immigration, consider a simpler but extreme model with two fac-

OCR for page 135
--> tors, capital and skilled labor taken as one input, and unskilled labor (high school dropout workers) taken as a second factor. Assume that unskilled labor makes up 20 percent of the U.S. workforce and that unskilled labor earns 10 percent of U.S. GDP—all reasonable order of magnitudes for high school dropout labor. Now let immigration increase the supply of unskilled labor by 33 percent. The gain to natives in terms of additional GDP from this flow of immigrants would be the triangle in Figure 4.1, which is one-half the change in labor supply times the change in the wage of unskilled labor. The change in the supply of labor that substitutes for immigrants is 0.33. If the elasticity of the wage of unskilled labor to this change was, say, 0.3, the wage loss to these workers would be about 10 percent. The triangle would be 1/2 (.33) (.10) for a group whose earnings represent 10 percent of GDP. Thus, relative to GDP, the gain to all natives would be about 0.2 percent or $14 billion in a $7 trillion economy. Immigration in this case has reduced the pay of a relatively modest number of U.S. workers by 10 percent while increasing the total GDP accruing to natives by 0.2 percent.