Chapter 6: The Economic Impact of Immigration David Coates

Skill levels (or at least education levels) vary between native-born Americans and Latino immigrants, legal or otherwise. Mexican immigrants arrive here with higher levels of education than the Mexican population in general, but they come with fewer years of education behind them than those possessed by the native-born workers they join. They also come with English as, at best, their second language. Illegal immigrants are particularly disadvantaged in both these regards. As we saw in Chapter 1, researchers at the Pew Hispanic Center recently compared high-school drop-out rates for unauthorized migrants, legal immigrants and native-born Americans. They reported 49% of unauthorized immigrants dropping out of high school, as against 21% of legal immigrants and just 11% of native-born Americans.³ Arriving in the United States under-schooled is particularly burdensome now because the educational credentials of the native-born labor force are currently rising. High school diplomas may be, at best, only an imperfect guide to educational attainment, let alone to adult skills; but such diplomas are not easy things to acquire, and their possession does tell us something of importance about the ability of their recipients to go on acquiring skills. In 1970 half the native-born labor force in the US economy lacked a high-school diploma. By 2004 only 6.6% of native-born workers were so ill-equipped. In 1980 only 10.9% of high school drop-outs were foreign born: by 2000 that figure had risen to 40.9%.⁴ There is, therefore, a genuine educational gap between Latino migrants as a group and the indigenous population as a whole, a gap larger for unauthorized immigrants than for those legally here.⁵ Significantly, however, for the jobs they seek, that educational gap is smallest with African-Americans.

Numbers of that scale and kind affect the total supply of labor available to the American economy, and they affect the skill mix of that labor supply. Both would be a problem if indigenous population growth already exceeded the anticipated trajectory of job numbers, or if the skill mix they created was out of line with known skill needs. But neither of those conditions currently applies.

On population growth we know several significant things. One is that immigration has long been the primary cause of US population growth. Without it, the population here in the United States in 1800 would likely have grown to less than 100 million by the year 2000. In fact the population that year stood at 270 million.⁶ We also know that currently birth rates among native-born Americans of all ethnic groups are low and falling. Natural population growth is presently running at 0.6% per year. In the 1960s it was running at 1.4%. Latino immigration, like immigration before it, is likely to be the great corrective here, though by itself it will still not be enough to restore US birth rates in the next century to the normal rate in the last. Immigration – legal or otherwise – is not spurring a population explosion as so many critics imply: rather, as Dan Griswold argues more fully in Chapter 13, it is helping to save “the United States from a population implosion.”⁷ The Pew researchers estimate a 48% increase in total population from 2005 to 2050, as against a 64% increase from 1960 to 2005. They also estimate that fully 82% of the population increase we can expect from now to 2050 will be immigrant-based – immigrants and their children born here – as against an equivalent figure of 51% for the years 1960-2005; and that the bulk of that population increase will be Latino in origin.⁸

Such immigrant labor is clearly needed, both immediately and over the long-term.⁹ It is needed over the long-term if only to hold at bay a looming crisis in pension-provision created by the retirement of the baby boomers.¹⁰ Social Security is a ‘pay-as-you-go’ pension scheme. Funds paid in now pay pensions now. They do not accumulate to pay the pensions of the workers who are actually contributing. The scheme worked well when the ratio of workers to pensioners favored workers. That ratio was 16:1 in 1935 as the scheme began. It is now 3:1 and falling, brought low by the prolonged longevity and diminished fertility of native-born Americans. By contrast, 84% of all unauthorized immigrants of working age (18-64) are under 45. The equivalent figure for native born and legal immigrants is a mere 60%.¹¹ In the United States, as elsewhere in the advanced industrial world, immigration is helping to hold the impending pension crisis at bay, as the Social Security Trustees openly acknowledge. Their 2007 Report argued that “an additional 100,000 new immigrants per year would increase the long-range actuarial balance by about 0.7% of taxable payroll.”¹²

The labor of undocumented immigrants in particular is currently also sustaining a significant number of key US industries. In 2005 undocumented immigrants made up 24% of all workers employed in farming, 17% in cleaning, 14% in construction, 13% in hotels, 12% in food preparation, and 10% in textiles. For specific occupations the percentages were even larger: 36% for insulation workers, 29% for roofers, 27% for butchers and food processing workers, and 26% for landscapers.¹³ The labor performed there by undocumented immigrants then sustained more senior positions, positions disproportionately occupied by native-born Americans. As we will see in more detail later, in general jobs in such industries have not been lost by native-born workers because of immigration. All that native-born workers have lost (or, more accurately, left behind) are the basic “grunt jobs” that earned the lowest wages; and there is nothing particularly new in that. Traditionally in the American immigration story, the last group to arrive took the lowest jobs, monopolizing them in spite of their best efforts until the next wave of immigrants arrived to take their place. American immigration has long been less a story of melting pots than of escalators, and it is proving to be so again.

The reliance of those particular industries on immigrant – and often undocumented – labor is itself a reflection of the changing pattern of demand and supply for unskilled labor in the US economy. The demand for such labor remains high and is projected to stay that way. The Department of Labor projects a continued and expanding demand for unskilled labor: between now and 2016 some 4.1 million more occupational slots that require, at most, only short-term on-the-job training.¹⁴ Indeed fourteen of the thirty US occupations with the largest projected job growth are in that short-term, on-the-job training list. When retirement patterns and labor turnover are factored in, the net replacement number required to staff these job categories may be twice that: approaching 8 million.¹⁵ Yet, at the same time, the supply of such unskilled labor from native-born households is fading fast. “In absolute numbers, the number of high-school dropouts in the workforce declined by 4.6 million between 1996 and 2004.”¹⁶ There is, therefore, a serious and impending skills gap at the bottom of the US labor market. Immigrant labor across the US-Mexican border clearly helps to bridge that gap; and it arrives illegally in the main because so few visas are annually made available for the entry of workers with low level skills. The Department of Labor anticipates the creation of 400,000 new low skill jobs annually over the next decade: yet Congress awards currently only five thousand visas for workers of that kind, and the current cap on H2-B visas (non-agricultural guest workers) is a mere sixty-six thousand.
Wages
Anyone trained in the laws of supply and demand should expect that a significant increase in the supply of any commodity, including labor, should directly reduce its price: that the effect of immigration on the wages of native-born workers should be negative. Of late, the most widely cited economist making that case has been George J. Borjas. Down the years his estimate of the impact of immigration – legal or otherwise – on the wage levels of native-born workers has varied. In 1996 he argued that “immigration may account for perhaps a third of the recent decline in the relative wages of less-educated native workers.”¹⁷ In 1999 he calculated the negative impact of immigration between 1979 and 1995 on the relative wages of high school drop-outs at about 5 percentage points, writing that “immigration reduced the income of the average black native by about $300 a year.”¹⁸ More recently, his best estimate of immigration’s impact is a reduction of 4.0% in the level of real wages of all native-born men between 1980 and 2000, with a significantly larger fall (7.4%) in the real wages of native-born high school drop-outs.¹⁹

But Borjas’s calculations have long been challenged by economists using different methodologies and data, and “many studies continue to find no effect or only weak negative effects of immigration on low-skilled workers or workers in general”.²⁰ David Card’s early work, for example, on the Mariel Boatlift – the unexpected influx of 125,000 Cubans into southern Florida in 1980 – found no adverse impact on Miami wages, even among earlier unskilled Cuban arrivals. His later research confirmed his general thesis: “overall, evidence that immigrants have harmed the opportunities of less educated natives is scant”.²¹ And Gianmarco Ottaviano and Giovanni Peri, in a recent and widely-cited research paper, have the impact of immigration on American wages as actually positive. Counter-intuitive as it may seem, they calculated that in the 1990s the average wage of native-born workers increased by 2-2.5% because of immigration, that the inflow lowered the real wage of native workers without a high school [diploma] by 1%, but increased the real wage of native workers with at least a high-school [diploma] by as much as 3-4% overall. ²² Not surprisingly perhaps, Borjas has recently bounced back with a counter-paper, querying Ottaviano and Peri’s methodology, and insisting that we cannot yet “reject the hypothesis that comparably skilled immigrant and native workers are perfect substitutes”.²³

This variation in findings is not surprising, because contradictory forces as well as contradictory methodologies are in play. What has to be determined in each case is the degree to which immigrant labor and native-born labor substitute for each other or act as mutual reinforcements. “In a closed economy model, immigrants will lower the price of factors with which they are perfect substitutes, have an ambiguous effect on the price of factors with which they are imperfect substitutes and raise the price of factors with which they are complements.”²⁴ To the degree that immigrant labor increases the labor supply and competes with native-born labor, it seems sensible to expect a negative impact on the trajectory that native-born wages would otherwise have taken. But to the degree that immigrant labor also increases total production, and pushes native-born labor up organizational ladders made more commercially-sound by the growth of GDP, this incoming labor is best conceptualized as complementary, and a force for general wage growth among native-born Americans. In fact, it is possible to conceive of both impacts occurring simultaneously: with immigrant labor adversely affecting those native-born workers most similarly skilled and closest in location to the immigrants, while (via its general impact on output and demand) reinforcing, even increasing, the wages of native-born workers more “distant” from the immigrants in the production process. In practice, both those effects seem present in the available wage data.

The large-scale immigration of unskilled labor does impact the wages of the native-born unskilled. Since the labor pool of the unskilled is heavily structured by race, the main “loser” would appear to be unskilled African-American male labor; though unskilled white labor is equally vulnerable, as are “the contingent of previous immigrants, who compete for much more similar jobs and occupations with new immigrants.”²⁵ How big a loss is hard to quantify, because the general demand for such labor remains high and growing over time, and because the available supply of African-American labor has been significantly reduced of late by unprecedented levels of black incarceration.²⁶ But without immigration, the supply of unskilled labor would undoubtedly be more limited, and demand would pull wages up. That wage loss is more likely to be experienced by native-born African-Americans and native-born Hispanic workers because they are “67% and 37% more likely respectively to be employed in the negatively affected occupations than are native-born whites.” Moreover, since those workers are low waged anyway, a wage loss here is likely to “represent a more significant reduction in the material prosperity of those groups”.²⁷

Borjas, Grogger and Hanson have recently estimated that “a 10-percent immigration-induced increase in the supply of a particular skill group is associated with a reduction in the black wage of 4%...and…among white men, reduces the wage by 4.1%.”²⁸ Their reference to “skill groups” implies that the immigration of skilled labor is likely to have a similar effect on wages higher up the salary ladder. Again, precision is difficult given the even greater shortfall in native-born supply in certain key skills, the associated heavy demand for them, and the willingness of US employers to pay premium salaries to attract foreign talent. George Borjas has recently estimated that “a 10 percentage immigration-induced increase in the supply of doctorates lowers the wages of competing workers by about 3%,”²⁹ but this is entirely out of line with Batalova’s findings that “for the overwhelming majority of native workers, the higher presence of immigrants in skilled jobs is not associated with a decline in earnings.” Only where skill groups contain over 35% of foreign-born workers – she estimates 5-7% of natives work in those jobs – is wage deflation likely.³⁰

Even more difficult to isolate is the impact of illegal immigration on wage levels, particularly among the unskilled. When flows of illegal immigration thicken, one possible response by both native-born workers and earlier immigrants is to relocate. To the degree that relocation occurs, an increase in immigration – legal or otherwise – might have no significant wage effect locally, while still possibly pulling wages down across regional or national labor markets as a whole. George Borjas has argued for the importance of this more complex impact in a number of recent publications,³¹ but even this complexity is contested. Card and his colleagues have found little evidence either of labor flight, or of a change in the wage gap nationally between native-born high school drop-outs and unskilled immigrant labor.³² Likewise, the highly prestigious report for the Fondazione Rodolfo Debenetti in 2002 recorded “the near uniform finding…that immigration has, at most, a small negative effect on wages…that a 10% increase in the fraction of immigrants in a region lowers native wages by less than 1% (often an amount not statistically different from zero).”³³ Even economists who publish with George Borjas – like Gordon Hanson – are on record as arguing that “there is little evidence that legal immigration is economically preferable to illegal immigration.”³⁴ This is a territory of strong claim and counter-claim, which is why Ron Haskins’ advice needs to be borne in mind: that “when economists who are greatly respected by their colleagues disagree sharply over an issue like the impact of immigration on employment and wages, it seems wisest for outsiders to resist forming a strong conclusion and simply say, instead, that the jury is still out.”³⁵