Misrepresenting Wage Gains for 1986 Amnesty Recipients: The Center for American Progress Distorts Data to Fit Their Amnesty Agenda

The claim that the 1986 amnesty led to a 15 percent increase in wages for amnesty recipients is based upon a 2010 study authored by Raul Hinojosa-Ojeda, an associate professor of Chicano Studies at UCLA, and published by the American Immigration Council (CAP).  Hinojosa wrote:

Surveys conducted by Westat, Inc. for the U.S. Department of Labor found that the real hourly wages of immigrants who acquired legal status under IRCA's general legalization program had increased an average of 15.1 percent by 1992 — four to five years after legalization in 1987 or 1988 (p. 7).

This 15 percent figure has been cited by Senator Schumer and Grover Norquist, and the Cato Institute published a condensed version of Hinojosa's CAP report in the Winter 2012 The Cato Journal.  The presentation of the Department of Labor (DOL) survey is based on a misreading of the data because Hinojosa and others are making the argument that amnesty will raise the "wage floor" for the average illegal alien.  What the DOL survey actually revealed was that the earnings of those 1986 amnesty recipients who were younger, better educated, spoke English well, were earning better than average wages before amnesty – indicating higher levels of job skills – and were not from Mexico, continued to outpace the earnings of those who did not fit this profile.1 

The survey revealed that "...the wages of Mexican men were the lowest, and those of male LPRs from non-Latin American countries the highest" (DOL, p. 44).  The non-Latin American cohort amnestied in 1986 was relatively small and was not representative of the general amnestied population (see Table 1).  Of those surveyed by the DOL, 70 percent were from Mexico.  Non-Latin Americans who received amnesty were much more likely to be visa overstayers (DOL, p. 79). Women also saw gains much higher on average than men, which was in line with a trend from the 1970s through the 1990s that saw wages for female workers rising disproportionately compared to those for male workers.

Table 1
1986 Amnestied Aliens: Region of Origin2
Mexico 70%
Central America 13%
Asia 5%
South America 4%
Caribbean 3%
Africa 2%
Europe 2%
All Other 1%

Table 2
Wages 1991 Median
Household Income 3
  At Application 1992 % Increase  
U.S. Workers $8.98 $10.33 15.033 $35,939
Amnestied Workers $7.57 $8.71 15.059 $20,147

The 15 percent gain for amnestied workers that Hinojosa attributes to amnesty was the same for all comparable U.S. workers (non-amnestied) and should be considered the baseline measure of earnings increase between 1986 and 1992 for all production or non-supervisory, non-farm workers.  If wages for a sub-group of these workers did not increase by 15 percent, those workers would have lost economic standing in the economy, and thereby made a smaller or negative economic contribution.  Increases below 15 percent are indicated in Table 3 in red.

Table 3
Wages of Workers of Amnestied in 1986 4
Variable Worker Characteristics Wages
    At Applicationi 1992 % Increase
Sex Male 8.33 9.43 13.205
Female 6.15 7.41 20.488
Region of Origin Mexico 7.14 8.11 13.585
Central America 7.30 8.42 15.342
Other 9.64 11.46 18.880
Period of Arrival Before 1975 8.37 9.06 8.244
1975-1979 7.83 8.82 12.644
1980 or later 6.88 8.42 22.384
Age at Arrival 17 or less 7.10 8.34 17.465
18 or more 7.76 7.88 1.546
Education Level 6 years or less 6.91 7.56 9.407
7 to 11 years 7.29 8.09 10.974
12 years or more 8.93 10.92 22.284
Ability to Speak English Not at all 6.08 6.65 9.375
Not very well 7.22 8.11 12.327
Well/Very well 8.87 10.33 16.460
Place of Residence California 7.67 9.02 17.601
Texas 6.34 7.17 13.091
New York 8.38 9.13 8.950
Other 7.75 8.68 12.000
Age at Application 20-29 6.50 8.34 28.308
30-39 7.83 8.91 13.793
40 or more 7.85 8.68 10.573
Family Income Previous Year $11,999 or less 6.04 6.10 0.993
$12,000-$19,999 6.80 7.42 9.118
$20,000-$29,999 7.58 8.67 14.380
$30,000 or more 9.80 11.93 21.735

i   Wages for U.S. workers is the annual average wage for U.S. production or non-supervisory workers on private non-farm payrolls in 1987.  Wages for amnestied workers is the workers' wage the week prior to application in 1987 or 1988.

Hinojosa ignores all other economic factors and attributes the increase for all workers to amnesty, even though this should have resulted in a disproportionate increase for amnestied workers in relation to non-amnestied U.S. workers.  That was not the case.  In many instances, amnestied workers lost ground compared to non-amnestied workers.  Their wages were on average 16 percent less than comparable U.S. workers both before and after amnesty, while the average difference in household income in 1991 was 44 percent.   Hinojosa misrepresents the survey data as a positive portrait of economic gain by the average amnesty recipient.  This directly contradicts the conclusions of the researchers whose work he cited.  

Findings by DOL researchers:

  • "...by 1992, the likelihood of unemployment was higher for legalized than for other U.S. men – a reversal of the pattern seen prior to legalization" (DOL, p. xv).
  • "...after five years of legal U.S. residence, a disproportionate share of legalization families were still below the poverty threshold" (DOL, p. 80 ).
  • "...it is clear...that at the time they applied for legalization, the majority of pre-1982 applicants possessed relatively little human capital" (DOL, p. 78).
  • "Although upward mobility was limited, during the week prior to legalization, English-proficient, skilled applicants were noticeably better positioned within the labor market than were the unskilled and those unable to speak English" (DOL, p. 79).

The DOL survey does not include those who received amnesty under the Special Agricultural Worker (SAW) provision, so the data does not account for two-fifths of those who received amnesty in 1986.  SAW amnesty recipients would have been the least-educated, least-skilled workers, with very few being able to speak English well, if at all.  This cohort would have performed poorly in the economy after legalization, as indicated by comparable workers who participated in the DOL survey.  Not including SAW workers paints a more favorable picture of post-amnesty performance than was actually the case.

In a separate analysis, The Immigration Policy Center found that 61 percent of Mexican men who had received amnesty in 1986 had no increase in their earnings five years later (IPC, p.  5).  Nancy Rytina, a researcher for INS, wrote in 2002 that fifteen years after amnesty:

Some upward mobility is suggested by the movement of 7%-19% of IRCA LPRs, including 9% of SAWs, into professional, managerial and technical occupations by the time of naturalization. On the other hand, some degree of lateral movement or downward mobility is implied by the movement of nearly 30% of IRCA LPRs in precision production and farming to laborer occupations at the time of naturalization. Overall, the net direction of movement is not obvious from these data (Rytina, p. 4).

Despite one of the greatest and most prolonged periods of economic growth in U.S. history during the late 1980s and the internet boom of the 1990s, those who received amnesty in 1986 on whole showed no visible signs of upward mobility between 1986 and 2001.  This is truly a dynamic analysis of the economic benefits of amnesty.

Minimum Wage

Hinojosa also failed to point out that the federal minimum wage increased 27 percent between 1986 and 1992.   A rise in the minimum wage this substantial had two direct effects on the wage survey results.  First, a 27 percent increase in the minimum wage would have raised the average wage for all workers.  The fifteen percent increase in wages for all workers would have been in part due to this increase in the minimum wage.  Second, the minimum wage is used as the floor for hourly workers, and an increase in the minimum wage generally leads to higher hourly wage rates for workers earning above the minimum wage, as well as causing some states to adjust their minimum wage rates upwards.5  The rise in the minimum wage was not related to amnesty and the failure of Hinojosa to adjust for this increase is another factor that renders his economic model defective.

Just as in 1986, the current average hourly wage for most illegal aliens is already above minimum wage, so if amnesty is passed today, the small cohort of illegal aliens who are making below minimum wage would see an increase in their pay.  That increase will have little positive effect on the economy, and some of these workers may lose their jobs upon legalization, which was the trend following the 1986 amnesty.  If the minimum wage were to rise comparable to the increase after the 1986 amnesty, the federal minimum wage would go up from $7.25 to $9.20 an hour.  This would boost the average earnings for hourly workers, but it would likely lead to a slow-down in hiring.

Compounding the Error

The Center for American Progress recently released a report authored by Robert Lynch and Patrick Oakford on the purported gains in the GDP that would occur in states with high illegal alien populations if amnesty were passed.  This publication is based upon the methodology used in an earlier report that simply assumed a 25.1 percent increase in wages that illegal alien workers would enjoy if amnesty were passed, and that this increase in earnings would "ripple through the economy." (Lynch and Oakford, p. 2).

This rosy scenario ignores the fact that amnesty will not change the education or skill levels of illegal aliens, which are the reason these workers earn less than others.  The authors also misrepresented the 1992 DOL survey, which actually undermines their argument that amnesty would result in higher earnings.  In the current economic climate, it is highly unlikely that wages for the average amnestied worker would increase, let alone rise 25 percent.  The authors also ignored that "comprehensive immigration reform," the road to amnesty for illegal aliens residing in the United States, is coupled with increases in legal immigration and expanded guest worker programs.    Perversely, the Gang of Eight bill would legalize millions of illegal alien workers, while at the same time bring in millions of low-skilled guest workers, creating an even greater oversupply of labor and resulting in wage depression and increased unemployment.   

The GDP Chimera

Hinojosa also claimed that based on the economic gains brought about by IRCA, amnesty today would increase the Gross Domestic Product (GDP) $1.5 trillion over ten years, which is $150 billion a year.  U.S. GDP in 2012 was $15.66 trillion.  That means amnesty would raise the GDP less than one percent a year.  Hinojosa calculated it to be 0.84 percent.  Even if one accepted all of his assumptions, the increase would still be minimal considering the size and scale of a mass amnesty.

Lynch and Oakford also contended that amnesty would increase GDP between 0.5 and 0.9 percent.  Amnesty advocates are hanging their hats on the shaky premise that granting amnesty to more than ten million illegal aliens will grow the GDP by less than one percent. The Congressional Research Service  (CRS, p. 2-3), and more recently Harvard economist George Borjas, have plainly stated that growth in the GDP attributable to immigration accrues to immigrants and employers of immigrants.  Even if GDP increases by adding more workers to the economy, it is unlikely that any of these gains will benefit the average American worker. 

Growing the economy is not the same thing as increasing economic prosperity.  The increase in per capita GDP has lagged behind overall GDP gains over the last 30 years.  This is ignored by CAP authors.  They also ignore the difference between economic and fiscal considerations.  Importing a low-wage foreign workforce benefits employers and it does add to the GDP, but it leads to fiscal costs that are incurred by American taxpayers.  The recent Heritage Foundation estimate of the cost of amnesty emphasized these socialized, taxpayer-funded costs ignored by pro-amnesty proponents. 


The claims made by the Center for American Progress that amnesty will benefit the U.S. economy have been cited in support of the immigration legislation currently pending in the Senate.  But these claims are being made by researchers who failed to recognize or were unable to grasp basic economic facts; or who deliberately constructed flawed models in order to arrive at preconceived conclusions.  Neither Hinojosa's nor Lynch and Oakford's conclusions hold up under even the most basic scrutiny.  Hinojosa misrepresented his source data, and any claims based on his work are not valid.  Lynch and Oakford make fantastical assumptions that no scrupulous economist would endorse.  

Recently, there was a peer-reviewed article published in the Journal of Regional Science in which the authors stated the following:

We find that improvements in employment outcomes from a new legalization program are limited, and for many possibly zero — at least in the short run. Specifically, the employment outcomes of immigrants who crossed the border without documentation do improve over time, but none of these improvements are attributable to gaining legal status. These immigrants are typically low-skilled workers with little education or proficiency in the English language. On the other hand, those immigrants who gained legalization after violating the terms of a temporary visa are likely to demonstrate some occupational mobility that may be related to acquiring legal status. On average, these workers are more highly skilled than those who crossed the border illegally. The differences we observe in occupational earnings growth between these two groups of unauthorized immigrants are specifically attributable to their differences in skill: we find that highly skilled immigrants in both groups exhibit occupational improvements after gaining legal status.

There is already a glut of low-skilled labor in the United States.  Granting amnesty to millions of illegal aliens and bringing in millions more low-skilled guest workers will only exacerbate the problem.  A review of the 1986 amnesty and an honest appraisal of the current illegal alien population makes it clear that amnesty today would not broadly benefit the U.S. economy.  

May 2013

  1. Shirley J. Smith, Roger G. Kramer, and Audrey Singer, "Characteristics and Labor Market Behavior of the Legalized Population Five Years Following Legalization," U.S. Department of Labor, May 1996.
  2. Market Behavior of the Legalized Population Five Years Following Legalization," U.S. Department of Labor, International Labor Affairs, May 1996.
  3. Smith, et. al, "Characteristics and Labor Market Behavior of the Legalized Population," p. 18.
  4. Ibid, p. 102.
  5. Ibid.
  6. Ralph E. Smith and Bruce Vavrichek, "The minimum wage: its relation to incomes and poverty," Monthly Labor Review, Bureau of Labor Statistics (June 1987): pp. 24-30.