False Assumptions in CAP Study Yield False Results

Executive Summary

A study put out by the Center for American Progress (CAP) in March 2013 made the fantastical claim that adoption of amnesty legislation could benefit the U.S. GDP over a ten-year period by $1.1 trillion, could create new 159,000 new jobs, and would result in the collection of an additional $144 billion in taxes.¹ Of course, that claim is dismissed later on by the authors as a “most politically unlikely scenario.” Still, the study asserts that somewhat lesser benefits would be likely. The report is based on flawed research and ignores the fiscal costs associated with illegal immigration that would escalate with access to welfare benefits.

The study cites various sources of earlier research findings for its conclusions, but the one report cited throughout, and upon which the authors model their scenarios of amnesty benefits, is a study commissioned by the U.S. Department of Labor (DOL) published in 1996. ² That study was based on survey results among beneficiaries of the 1986 amnesty five years later. The authors either misunderstand or misrepresent the results of that DOL study. They fail to recognize that the sample of questionnaire respondents was self-selecting and, therefore, more likely to reflect those with positive experiences than negative ones, and the respondents were also not representative of all amnesty recipients. The survey excluded the two-fifths of amnesty beneficiaries who gained amnesty through work in agriculture – a factor of major importance in the education, skills set, and English competency that were shown to have a major role in influencing the economic effects of the amnesty on the beneficiaries. 

The survey data in the DOL study found that the amnesty beneficiaries who gained the most were visa overstayers, mostly those who came as students from Europe and Asia. Most, if not all, of those visa overstayers were likely to have been among the three-fifths of the amnestied illegal aliens sampled in the survey rather than among the two-fifths working in agriculture who were not in the survey.

Even setting aside the unrepresentativeness of the sample data relied on by CAP’s authors, they also seem to misread the findings with regard to earnings benefits from the amnesty. Rather than there being some magical across-the-board economic benefit from amnesty, the DOL report’s authors found that overall the amnesty recipients did not gain any ground in earnings relative to workers who were already legal workers. This failure of the entire cohort to realize wage gains compared to other workers, while some, i.e., the overstayers, were gaining some advantage, means that overall the illegal border crosser population lost ground compared to other workers. The illegal border crosser population constituted a majority of the amnesty beneficiaries and, presumably, virtually all of the two-fifths of the amnesty recipients who were working in agriculture. Therefore, drawing on the experience of the 1986 amnesty to project future economic results from a new amnesty should reflect economic hardship rather than an economic panacea.

The CAP study inflates benefits and ignores costs

Unlike fiscal cost studies, such as FAIR’s estimate of the annual fiscal costs of illegal immigration at the national, state, and local level at $100 billion,³ and the Heritage Foundation projection of $6.3 trillion in expenditures over the lifetime of the amnesty recipients,⁴ both of which were net costs after estimating offsetting tax collections, the CAP study ignores the costs of amnesty. The CAP authors’ rosy amnesty depiction projects a jump in earnings by the recipients of 15.1 percent as a result of gaining legal status and an additional 10 percent earnings bonus as a result of becoming U.S. citizens. Those increased earnings are then projected to lead to increased GDP, job creation and tax collection.

The 15 percent wage gain applied used by the authors is derived from the DOL survey research conducted among the amnestied 1986 population. In trying to promote amnesty for illegal aliens, the CAP authors chose to ignore the critical fact that the DOL report also documented the same 15 percent wage gain for all “non-supervisory U.S. workers.” The legalized workers at the time of gaining amnesty were earning on average 84 percent of average wages for the comparable U.S. workers and, five years later, they were still earning on average 84 percent of the comparable U.S. worker wages. They had not benefitted relative to other workers from the amnesty. 

The study reported that, “Singer [one of the study’s authors] found that the wages of legalized men increased at about the same pace before and after legalization.” It continued, “However, real losses were sustained by those in the United States longest and those completely unable to speak English. Respondents in families whose 1991 total incomes fell below $12,000 were trapped on the wrong side of the widening wage gap. They had experienced more than a 10 percent drop in average real hourly earnings since they arrived in this country over a decade earlier.”

Further, the DOL study of the legalized population found that five years after amnesty the unemployment (“not in labor force”) rate for the legalized population had increased from 15 percent to 19 percent.

The CAP authors also do not acknowledge that during the 1987-1992 period over which the 15 percent wage gain was reported the federal hourly minimum wage was raised from $3.35 to $4.25 – an increase of 26.9 percent.⁵ Many of the illegal alien workers may be assumed to have benefitted from that minimum wage increase rather than as a result of gaining legal status. The ripple effect from the rise in the minimum wage will also have generated wage increases for hourly wage workers who had been earning above the old minimum wage.

The CAP study argues that one of the factors that would produce wage gains for the amnestied population would be gains in English competency. They correctly note that lack of English competency is a limiting factor for job mobility and advancement. They cite the DOL amnesty follow-up study as evidence that most of the amnesty beneficiaries spent some time studying English following the amnesty. This assertion and resulting assumption misses two important facts. The first is that the 1986 amnesty required applicants to study English (except for the agricultural worker beneficiaries) in order to gain legal permanent residence. Second, the authors appear to be unaware that the DOL study of the 1986 amnesty beneficiaries found that, “Despite the satisfaction of the English requirement and the fact that nearly a third (31 percent) completed additional English training, gains in proficiency between 1989 and 1992 were negligible. …Upon direct questioning in 1992, after more than a decade of continuous U.S. residence, 22 percent of legalized adults said they still spoke no English whatsoever.” 

Factors that mitigate economic progress for legalized aliens

The CAP study’s authors don’t just use an unfounded and unrealistic earnings gain estimate to project out over ten years, they also ignore factors that would make that projection misleading even if the underlying data were correct. 

One of those factors is the fact that many of the illegal alien workers in the United States are sending money back to their home country. These remittances represent a drag on the U.S. economy because when the earnings do not circulate locally they cost the economy in sales, tax collections and jobs.⁶ As the amount of remittances tends to be sensitive to economic changes, it is reasonable to expect that if earning were to rise from adoption of an amnesty, that the flow of remittances out of the country would also rise. That would limit any benefit to the economy and would exacerbate the nation’s negative balance of payments.

The CAP authors also seem mindless of the fact that illegal alien workers are often preferred over legal workers precisely because they do not have legal status. The lack of legal status means that they are more exploitable by the employer. The illegal workers are generally paid less than legal workers, are less likely to protest illegal work conditions such as the failure to pay overtime, and are less likely to risk losing employment by supporting unionization of the worksite. These are all factors that could influence an employer to dismiss a former illegal worker who gains legal status. Because S.744 requires amnesty recipients to avoid periods of unemployment to maintain their legal status and gain ‘green cards,’ the legalized workers are unlikely to gain leverage with their employer.

This tendency is most easily observed in seasonal crop agriculture where the wages that must be paid to legal agricultural guestworkers are deliberately set at rates designed to protect U.S. workers. That higher wage and requirements to pay transportation and to provide shelter for legal guestworkers have resulted in a widespread practice of hiring workers with no documents of fake documents over legal workers. 

The existence of such practices in agriculture and among other employers put in doubt the CAP authors’ assumption about wage gains. Employers who are taking advantage of the lower wages they can pay to illegal workers are unlikely to voluntarily pay higher wages simply because the legalized workers have received legal status as long as there remain illegal workers available to work for less. 

Impact on U.S. Workers

The CAP report’s projections of gains from amnesty speculate that the benefit will be based on the legalized aliens’ ability to compete for more jobs. Even greater economic potential and job opportunities would result from acquiring U.S. citizenship, claim the authors. The report ignores the economic impact of amnesty on other workers in our economy. The law of supply and demand suggests that creating a greater supply of job applicants will offer an advantage to employers in holding down wages. That, in turn, puts in question the assumption in the study that wages will rise unrestrained by competition for employment. The study does not include an assessment of the possibility of decreasing wages for both the amnesty beneficiaries and U.S. citizens as a result of the amnesty. There is also the political question of whether it is fair to U.S. citizens to artificially create greater job competition, and therefore job insecurity – especially at a time with unacceptably high unemployment.

Other flaws in the CAP study

The authors apparently also relied on a 2002 study by Kossoudji and Cobb-Clark (“Coming out of the Shadows: Learning about Legal Status and Wages from the Legalized Population”) that in turn relied on the DOL’s follow-up study based on interviews with a selection of the legalization beneficiaries after five years of legal status. Like the original DOL study, the authors of this 2002 study make no allowance for the minimum wage increase that was legislated following the amnesty but before the follow-up survey. Kossoudji and Cobb-Clark did note, however, that the minimum wage had been raised in California by state law, and that event could have affected the calculations of a wage gain by the legalized population because so many of them resided in California.

The authors of the CAP study also rely on Census Bureau data that show U.S. citizens have, on average, higher incomes than non-citizen immigrants. They infer from this fact that if the non-citizen immigrants became citizens they would gain higher incomes. This assumption ignores the issues of work skills, education and English competency that they later acknowledge as critical factors in upward job mobility. In the process, they also ignore that the non-citizen immigrant population on whom the Census Bureau collects its data include both legal immigrants who have not become naturalized U.S. citizens as well as illegal aliens. These groups have very different characteristics which represent very different limiting factors for upward mobility.

Similarly they cite research that shows earnings for Mexicans legally residing in the United States are much higher than for those residing illegally. But this does not prove that legal status results in greater income. Another, more plausible explanation of the earnings differential is that Mexicans residing legally in the United States include professionals who have entered with NAFTA visas and those who enter in the agricultural guestworker H-2A visa program, both of which would explain higher salaries than would be available to those residing illegally in the country. That would not change simply because of granting legal status.

Other research cited by the CAP study’s authors to justify assumptions about higher earnings from legalization found that naturalized immigrants earn more than legal noncitizens. To suggest that the higher earnings are a result of gaining U.S. citizenship ignores the fact that the English-language civics exams that applicants for naturalization must pass discourage poorly-educated and non-English-speaking immigrants from becoming citizens. Higher earnings for the naturalized immigrants are likely to result from those personal characteristics rather than from the different naturalization status.

Finally, a careful read of the CAP study reveals that the authors recognize that they have cut some corners to arrive at their projected benefits from adoption of an amnesty. They acknowledge, for example that they did not have data on English language competency to plug into their regression analysis. And they acknowledge that “…since our regression analysis mixes already-documented legal noncitizens with undocumented noncitizens, it does not measure the effect of a policy change aimed only at the undocumented.” These caveats also challenge the reports rosy projections.

Conclusion

The CAP study depicts a vastly inflated boon from adoption of an amnesty for the illegal alien population while ignoring the fiscal costs associated with providing the beneficiaries with access to the nation’s broad social safety net. Even if the projected economic benefits were realistic – which they are not – the offsetting costs would outweigh the benefits.

The fundamental flaw with the CAP report is that it based on a misinterpretation of research of the effects of the 1986 amnesty on its beneficiaries. Rather than great economic advances by that population, the data show that those beneficiaries did not gain relative to other workers and a majority lost ground economically.