The Impact of Large-Scale Immigration on American Workers
Testimony Prepared for Senate Committee on the Judiciary Subcommittee on Immigration and the National Interest
March 16, 2016
by Steven A. Camarota, Ph.D.
Director of Research
Center for Immigration Studies
1629 K Street NW, Suite 600
Washington, DC 20006
(202) 466 8185
fax, (202) 466 8076
Economic Theory. There is no question that by adding workers immigration makes the U.S. economy (GDP) larger by perhaps $1.7 to $2 trillion a year. However, by itself a larger economy is not a benefit to native-born Americans. Though the immigrants themselves benefit, there is no body of research indicating that immigration substantially increases the per-capita GDP or income of natives. Basic economic theory predicts that immigration should create a small net gain for natives, but to do so it must redistribute income from workers in competition with immigrants to business owners, and those workers not in competition with immigrants. Economic theory also predicts that the size of the redistribution will be much larger (about $531 billion) than the net gain (about $54 billion). Less-educated Americans who are already the poorest workers bear the brunt of immigrant competition and wage losses. Moreover, the National Academy of Sciences' comprehensive look at the economic and fiscal impact of immigrants (taxes paid minus services used) found that the net fiscal burden immigrants create (taxes paid minus services used) is actually larger than the immigrant surplus.
Recent Employment Trends. Putting aside economic theory, the last 15 years have witnessed an extraordinary situation in the U.S. labor market — much of employment gains have gone to immigrant workers, even though natives have accounted for about two-thirds of the growth in the working-age (16-65) population. Even before the Great Recession, a disproportionate share of employment gains went to immigrants. Despite improvements in job growth in the last two years, in the fourth quarter of 2015 only about two-thirds of working-age native-born Americans actually had a job; as recently as 2000 about three-fourths were working. America does not have a shortage of workers, it has a shortage of jobs.
Immigration and Aging. Whenever the impact of immigration on the labor market is discussed the argument is often made that immigration can fix the problems associated with our aging society, in particular the decline in the share of the population who are workers. However, this is not the case. For example, if we remove the 17.3 million immigrants (legal and illegal) who arrived in 2000-2014 and their 3.9 million U.S.-born children from 2014 Census Bureau data, 66 percent of the U.S. population would be of working age (16 to 65); if they are included, 66.2 percent are of working-age — a miniscule difference. Whether we look at the impact of current immigrants or project the impact in the future, it is clear that immigration has only a tiny impact on slowing the aging of American society. Equally important, the problem in the United States is not that there are too few people of working age, rather the problem is the enormous number of working-age people who are not working.
In this testimony, I will focus on four areas. First I will summarize the theoretic impact of immigration on the U.S. labor market. Second, I will discuss what the research shows about the actual impact of immigration on the wages and employment of natives. Third, I will describe what has been happening in the U.S. labor in the last 15 years. Finally, I will discuss the impact of immigration on the aging of American society and the need for workers. The overarching theme of my testimony will be that immigration does not confer significant economic benefits on the native-born. Instead, immigration tends to redistribute income from some American workers, often from the poorest and least educated, to business owners and other American workers. Moreover, the chief justification for the current high level of immigration that there is need for workers is not supported by the available evidence.
Theoretical Impact of Immigration on the Labor Market
A Larger Economy. Immigration adds significantly to the size of the U.S. population; and there is no question that a larger workforce means a larger economy. In the fourth quarter of 2015, immigrants accounted for 16.9 percent of those in the labor force. If we make the standard assumption that labor accounts for 70 percent of GDP and the economy totaled $18 trillion in 2015, then immigration may add $2.1 trillion to the U.S. economy.
This calculation assumes that immigrants receive a share of all labor incomes in proportion to their share of the labor force, but we know that immigrants on average earn about 80 percent what natives earn. Therefore, the actual increase in the size of the nation's GDP is probably more likely to be roughly $1.7 trillion, still a very large figure. It must be emphasized that this "contribution" to the U.S. economy does not measure the gain or benefit to the native born.
As the nation's top immigration economist, George Borjas, points out the, "contribution to the aggregate economy, however, does not measure the net benefit to the native-born population." This is because roughly 98 percent of the increase in GDP goes to the immigrants themselves in the form of wages and benefits. They are the ones providing the labor, so it is entirely fair that they reap the benefits of their efforts. Immigration does make the nation's economy larger, but that fact does not make natives significantly richer. What is the gain to natives?
The Gain to Natives. There is a standard way of calculating the benefit from immigration, also referred to the as the "immigrant surplus", that goes to the existing population of natives. A 1997 study by the National Research Counsel (NRC), authored by many of the top economists in the field, summarizes the formula for calculating the benefit.1 The NAS study updates an earlier study by the nation's top immigration economist, George Borjas of Harvard.2 In 2007, the president's Council of Economic Advisers (CEA) also used the same formula to estimate the benefit of immigration to Americans.3 At the Center for Immigration Studies, we published an updated version of these calculations by Dr. Borjas in 2013.4
The next gain from immigration can be estimated using the following formula:
Net gain from immigration as a share of GDP = -.5 * labor's share of income * wage elasticity * immigrant share of labor force squared.
"Labor share" refers to the percentage of GDP that goes to workers, which is usually thought to be 70 percent, the rest being capital. The immigrant share of the labor force was 16.9 percent in December of 2015. "Wage elasticity" refers to the percentage change in wages from immigration increasing the size of the labor force by 1 percent. The size of the elasticity is a contentious issue. The NAS study assumed an elasticity of .3, and so will I in the calculation below. This means that each 1 percent increase in supply of labor caused by immigration reduces wages by 0.3 percent. Put a different way, if immigration increased the supply of workers by 10 percent, it would reduce the wages of American workers by 3 percent. Putting the values for 2015 into the formula produces the following estimate:
0.3% =-.50 * .70 * -0.3 * (.169*.169)
Thus the net gain from immigration is 0.3 percent of GDP. (Expressed as a decimal it is .0029.) If GDP is $18 trillion, then the net benefit would be about $54 billion. Three important points emerge from this analysis. First, the net effect of immigration on the existing population is positive overall, though not for all workers. Second, the benefits are trivial relative to the size of the economy, less than one-third of 1 percent of GDP. Third, the benefit is dependent on the size of the wage losses suffered by the existing population of workers. Or put a different way, the bigger the wage loss, the bigger the net benefit. Those who contend that immigration has no impact on the wages of natives are also arguing, sometimes without realizing it, that there is no economic benefit from immigration.
Two important caveats about these calculations. First, the model does not tell us how the benefit is distributed among business owners, consumers, and workers not in competition with immigrants. But the fact that business owners lobby so hard to keep immigration levels high is an indication that much of the lost wages are likely retained by them. Moreover, there is general agreement that gains in productivity in recent decades have almost entirely accrued to owners of capital, not labor.5 This is certainly consistent with the possibility that by increasing the supply of workers, immigration has shifted bargaining power to employers, allowing them to capture these gains. The second caveat should be obvious, the U.S. economy is enormous and its complexity is not captured by a simple model of this kind.
Wage Losses. The same model can be used to estimate the wage losses suffered by American workers. Wage loss as a fraction of GDP = - "labor's share of income" * "wage elasticity" * "immigrant share of labor force" * "native-born share of labor force.
Putting the numbers into the equation you get the following:
3% = -0.7 * -0.3 * 0.169 * 0.831
This is 3.01 percent of GDP, or $531 billion in wage losses suffered by American workers because of immigration. This is not trivial. There is nothing particularly controversial about this estimate and it stems from the same basic economic formula as the one above. Think of it this way: Labor is 70 percent of our $18 trillion economy, or $12.6 trillion. If the elasticity is .3 and immigrants are 16.9 percent of the labor force, then wages will decline by 5.1 percent (16.9 * .3). Thus the total wage loss must run into the hundreds of billions of dollars. If we are to accept the benefits that the model implies from immigration, then we must also accept the wage losses that the model implies. In short, the winners from immigration gain $594 billion and the losers lose $531 billion, for a net gain for $63 billion. The central and unavoidable reality is that immigration creates winners and losers.
Immigrants tend to be concentrated at the bottom end of the labor market in occupations that require modest levels of education. For example, half of maids and one-third of construction workers and meat packers are immigrants, compared to 7 percent of lawyers. Immigrants are not evenly distributed across the economy and this fact has important implications for those who win and lose from immigration.6
Although not the focus of this analysis, the fiscal impact of immigrants (taxes paid minus services used) may be relevant to the discussion of the immigrant surplus. In their 1996 study of the economic and fiscal impacts of immigration, the National Academy of Sciences estimated that the net fiscal drain from immigrant households was larger than the estimate immigrant surplus.7
The empirical literature tends to show that the wage and employment impacts of immigration are consistent with the theoretical effects outlined above. However, the results vary depending on the methodological approach.
Cross-City Comparisons. Studies done in the 1980s and early 1990s, which compared cities with different proportions of immigrants, are now widely criticized because they are based on the assumption that the labor market effects of immigration are confined only to those cities where immigrants reside. The interconnected nature of the nation's economy makes cross-city comparisons difficult to interpret, as the movement of people, goods, and services tends to diffuse the impact of immigration. Moreover, immigrants generally choose to settle in areas of high wage and employment growth, creating a correlation-causation problem when assessing the link between immigration and economic activity.8
Even quasi-experimental studies across cities are problematic. For example, the famous "Mariel boatlift" — a large and unexpected wave of Cuban immigrants to Miami in the summer of 1980 — has sometimes been cited as proof that immigration has no effect on wages.9 Last fall, however, economist George Borjas re-analyzed the data and found that the boatlift did have a large negative impact on native high-school dropouts in Miami. Borjas's study was quickly followed by attempted rebuttals. Regardless of how the debate is resolved, however, it is clear that the boatlift is not an ideal test case. Sample sizes are small, the comparison cities are somewhat subjective, and the timing of expected wage impacts is unknown.10
The National Perspective. In order to overcome the problems of cross-city comparisons, researchers over the last two decades have sorted workers into skill groups defined by education and age, and then compared the impact of immigration across these groupings. In his updating of the standard skill-group model for the Center for Immigration Studies, George Borjas shows that a 10 percent increase in the size of a skill group due to the entry of immigrants (both legal and illegal) reduces the wages of native-born men in that group by 3.7 percent and the wages of all native-born workers by 2.5 percent.11 Further support for the findings using the skill-group approach has been shown in recent studies in Canada, Mexico, Germany, and Norway.12
As with all economic modeling, the wage effects on skill groups can vary considerably depending on the assumptions that are built in. Nevertheless, immigrants are seen to lower the wages of competing workers in most scenarios and across a wide range of assumptions. For example, perhaps the most cited paper claiming that immigrants do not lower the wage of low-skill natives, written by Gianmarco Ottaviano and Giovanni Peri, requires the assumptions that the capital stock has fully adjusted, that immigrants and natives are not perfect substitutes even within skill groups, and that there are no meaningful skill differences between high school dropouts and high school graduates. Even under those particular conditions, immigrants still lower the wages of prior immigrants, just not the wages of natives.13
One of the newest studies in the skill-group literature shows one of the largest negative wage effects. In a working paper updated in January of this year, Joan Llull noted that external events such as overseas wars and natural disasters influence the level of immigration to the United States. Because those events are largely unrelated to demand for immigrants here in the United States, Llull was able to separate the effect of immigration on wages from other economic factors. The result is that a 10 percent increase in immigrations leads to a remarkable 12 percent reduction in wages.14 Whether effects of this magnitude are sustained in future studies remains to be seen.
Employment. Although economists have focused more on wages, a handful of studies have attempted to measure the impact of immigration on the employment patterns of natives. In an extensive study of California, the RAND Corporation estimated that between 128,000 and 195,000 natives in California were either unemployed or withdrew from the labor force because of immigration from 1970 to 1990.15 Two more recent studies concluded that immigration not only reduces the employment of less-educated black men, it also increases crime and incarceration among that population.16
Research by Christopher Smith, an economist at the Federal Reserve, has found that immigration has played a significant role in reducing employment for teenagers,17 and our own research at CIS tends to support these findings.18 However, the issue of how immigration impacts the employment opportunities available to natives remains underexplored.
Recent Labor Market Trends
Explosion of Non-Work. The vast majority (95 percent) of workers in the United States fall into the 16 to 65-year-old age group, so focusing on this population makes sense when considering the population of potential workers. Looking at the fourth quarter of 2015, the most recent quarterly data available, and comparing it with the same quarter in 2007, just as the Great Recession began, shows that despite significant job growth in the last two years there were still 1.3 million fewer native-born working-age Americans working at the end of 2015 than in the same quarter of 2007. The table below reports these figures. In contrast, the number of immigrants (16 to 65) working was 1.8 million higher in 2015 than 2007. Thus, over the whole time period all of the net gain in employment 2007 to 2015 went to immigrants. This is the case even though natives accounted for 61 percent of overall population growth among the working-age. Put simply, natives accounted for 61 percent of the increase in the number of potential workers 2007 to 2015, but none of the net growth in actual workers in this age group. Employment figures for the first quarter of each year for the working age can be found in the table.
To be sure, things have looked better for natives in the last two years. However, though the number of natives working has gone up significantly, it has still not gotten back to the level it was in 2007. In contrast, the number of immigrants working has entirely recovered. Perhaps most important, despite improvements in job growth in the last two years, in the fourth quarter of 2015 only 68.2 percent of working-age native-born Americans actually had a job; this compares to 71.8 percent in the fourth quarter of 2006, a year before the Great Recession started, and 74 percent in the fourth quarter of 2000. Of course, the employment rate (share of working-age people working) fluctuates with the economy, but the long-term trend over the last 15 years is clearly downward for the native-born. At the same time, the number of working-age immigrants holding a job increased 32 percent, from 18.5 million in 2000 to 24.3 million in 2015.
A Dearth of Jobs. One way to think about what has been happening in the labor market is that employment or job growth has not come close to matching new immigration and natural population increase. As the figure shows, between 2000 and 2015 the number of working-age people (both immigrant and native) in the country increased by 25.9 million. But the number working increased by only 8.4 million. The difference of 17.5 million represents the increase in the last 15 years in the number of people ages 16 to 65 who are not working. (It should be noted the figures do not include those in institutions such as prisons and nursing homes.)
Of course, not everyone wants to work. Parents staying home with young children, the disabled and those with disabilities, and others often cannot work or do not wish to do so. But this has always been true. Looking at the entire population ages 16 to 65, if the same percentage were working in the fourth quarter of 2015 as in the fourth quarter of 2000, there would be 152.9 million immigrants and natives employed, rather than the 142.4 million who were actually working — a 10.5 million difference. It is reasonable, therefore, to argue that there is currently a jobs deficient of 11 million. To be clear, the big growth has been among those entirely out of the labor market. That is, they are not working nor are they looking for work.
At the heart of the immigration debate is the idea that there are not Americans available for work. The data collected by the government shows this is not the case. Rather, there has been a dramatic decline in work, particularly among the young and less educated.19 The employment situation in the United States is so bad that it is absurd to suggest that there is a general shortage of workers.
The Impact of Immigration on Aging
When the issue of immigration and workers is discussed, its impact on slowing the aging of the U.S. population often comes up. While it may seem plausible that arrivals of relatively young immigrants who have higher fertility rates than natives would rejuvenate the age structure of the U.S. population, in fact immigration's impact on aging is quite modest. We can see this is the case using the most comprehensive dataset the government collects on the U.S. population, called the American Community Survey (ACS). The most recent ACS available is from 2014 and it shows that the median age of an immigrant was 43 years compared to a median of 35 years for natives. The median overall age in the United States was 37. The same survey also shows that 13 percent of both immigrants and natives are over age 65. Immigrants typically arrive in their late twenties, but the higher median age for the overall immigrant population and the share over 65 is a reminder that immigrants age over time like everyone else.
The main concern with an aging society is that there will not be enough people of working-age to pay for government or support the economy. In 2014, 66.2 percent of the total population was 16 to 65. If all 17.3 million immigrants in 2014 who indicated that they arrived in 2000 or later are removed from the data, 65.1 percent of the population would be of working age. If we remove post-2000 immigrants plus the 3.9 million native-born children born post-2000 immigrants, 66 percent of the U.S. population would be of working age. Again this compares to 66.2 percent when these immigrants and their children are included. Clearly the impact of immigration on the share of the population that is of working-age is quite small even though it added 21.2 million people to the population. Immigration adds to the working-age population, but it also adds to the population too old or too young to work.
We can also use the ACS to estimate immigrant and native fertility. Looking at fertility provides insight into the long-term impact of immigration on the aging of America. The total fertility rate (TFR) of immigrant women is 2.2 children, compared to 1.78 for natives. The TFR is a measure of fertility used by demographers to measure the number of children a woman can be expected to have in her lifetime given current patterns.20 The total fertility rate in the United States (immigrant and native) is 1.85. Without immigrants the rate would be the TFR for natives of 1.78. Thus the presences of immigrants raises the TFR of the country by .08 — 4 percent.21 While immigrants do tend to arrive relatively young, and have somewhat higher fertility rates than natives, immigrants age just like everyone else, and the differences with natives are not large enough to fundamentally alter the nation's age structure. Demographers, the people who study human populations, have long known this is the case.
In an important 1992 article in Demography, the leading academic journal in the field, economist Carl Schmertmann, explained that, mathematically, "constant inflows of immigrants, even at relatively young ages, do not necessarily rejuvenate low-fertility populations. In fact, immigration may even contribute to population aging."22 The Census Bureau also concluded in projections done in 2000 that immigration is a "highly inefficient" means for increasing the percentage of the population that is of working age in the long run.23 In a paper presented at the annual meeting of the Population Association of American in 2012 by myself and several co-authors we also showed that immigration has only a small impact on aging, but a large impact on the size of the U.S. population.24 There is a clear consensus among demographers that immigration has a positive but small impact on the aging of society like ours.
It is difficult to find a government policy that impacts American society more than immigration. There are currently 61 million immigrants and their young children living in the United States, accounting for nearly one in five U.S. residents. In my testimony I have tried to make clear that immigration makes our economy larger, but the gains to natives seem to be extremely small. However, economic theory and a significant body of research indicates that immigration does adversely impact the wages and employment of the native-born. And the impact can be significant for some workers, often the poorest and least educated. However, the size of the impact is debated among economists.
What is not in dispute is that the employment situation for Americans looks dismal when we expand our analysis beyond just those who are officially unemployed. The number and share of native-born Americans ages 16 to 65 who are not in the labor market (not working or looking for work) is at or near a record level, with no meaningful improvement in the last two years. Any suggestion that the nation needs immigration because there is a shortage of labor is not supported by the available data.
1 Barry Edmonston and James Smith, Eds., The New Americans: Economic, Demographic, and Fiscal Effects of Immigration, Washington D.C.: National Academy Press, 1997.
2 George Borjas, "The Economic Benefits of Immigration", Journal of Economic Perspectives, Vol. 9, No. 2, Spring 1995.
3 "Immigration's Economic Impact", white paper, June 20, 2007.
4 George Borjas "Immigration and the American Worker: A Review of the Academic Literature", Center for Immigration Studies Backgrounder, 2013.
5 Ian Dew-Becker and Robert J. Gordon, "Where Did the Productivity Growth Go? Inflation Dynamics and the Distribution of Income", NBER Working Paper No. 11842, 2005. Lawrence Mishel, "The wedges between productivity and median compensation growth", Economic Policy Institute, 2012.
6 Steven Camarota 2016, "The Demographic, Economic, and Fiscal Impact of Immigration", presentation, 2016.
7 Barry Edmonston and James Smith, Eds., The New Americans: Economic, Demographic, and Fiscal Effects of Immigration, Washington D.C.: National Academy Press, 1997.
8 For a more detailed discussion of this problem, please see Steven A. Camarota, "Each Skilled Immigrant Creates 2.5 Jobs for Natives?", Center for Immigration Studies Backgrounder, August 18, 2015.
9 David Card, "The Impact of the Mariel Boatlift on the Miami Labor Market", Industrial and Labor Relations Review, Vol. 43, No. 2 (January 1990), pp. 245-257.
10 Jason Richwine, "Immigration and Wages: The New Debate over the Mariel Boatlift", Center for Immigration Studies Backgrounder, January 2016.
11 George Borjas, "Immigration and the American Worker", Center for Immigration Studies Backgrounder, April 2013.
12 Borjas, "Immigration and the American Worker", Table 2.
13 Gianmarco I. P. Ottaviano and Giovanni Peri, "Rethinking the Effect of Immigration on Wages", Journal of the European Economic Association, Vol. 10, No. 1 (February 2012), pp. 152-197.
14 Joan Llull, 2016 "The Effect of Immigration on Wages: Exploiting Exogenous Variation at the National Level", Working Paper, January 2016.
15 Kevin McCarthy and George Vernez, "Immigration in a Changing Economy: California's Experience", Rand Corporation, 1997.
16 George J. Borjas, Jeffrey Grogger, and Gordon H. Hanson, "Immigration and the Economic Status of African-American Men", Economica, Vol. 77, No. 306 (April 2010), pp. 255-282.
Edward S. Shihadeh and Raymond E. Barranco, "Latino Employment and Black Violence: The Unintended Consequence of U.S. Immigration Policy", Social Forces, Vol. 88, No. 3 (March 2010), pp. 1393-1420.
17 Christopher L. Smith, "The Impact of Low-Skilled Immigration on the Youth Labor Market", Journal of Labor Economics, Vol. 30, No. 1 (January 2012), pp. 55-89.
18 Steven Camarota and Karen Zeigler, "A Drought of Summer Jobs: Immigration and the Long-Term Decline in Employment Among U.S.-Born Teenagers", Center for Immigration Studies Backgrounder, May 2010.
Andrew Sum, Paul Harrington, and Ishwar Khatiwada, "The Impact of New Immigrants on Young Native-Born Workers, 2000-2005", Center for Immigration Studies Backgrounder, September 2006.
19 Steven Camarota and Karen Zeigler, "The Employment Situation of Immigrants and Natives in the Fourth Quarter of 2015", Center for Immigration Studies Backgrounder, 2016.
20 For a more technical definition of the total fertility rate see the United Nations Statistics Division definition of TFR.
21 The TFR for natives is actually 1.778 and the overall TFR is actually 1.854, so the difference is .08.
22 "Immigrants' Ages and the Structure of Stationary Populations with Below-Replacement Fertility", Carl P. Schmertmann, Demography, Vol. 29, No. 4, November 1992.
23 The 2000 Census Bureau population projections mentioned above can be found here.
24 Stephen Tordella, Steven Camarota, Tom Godfrey, and Nancy Wemmerus Rosene, "Evaluating the Role of Immigration in U.S. Population Projections", presented at the annual meeting of the Population Association of America, May 2012. Using the Census Bureau's projections as a baseline, the paper shows that immigration between 2010 and 2060 would add roughly 140 million residents to the U.S. population. However, immigration would only increase the share of the population in 2060 that was of working-age (16 to 65) from 58.5 percent of the population (without immigration) to 59.9 percent. (See Figures 4 and 5 in that report.)