← The New Commonwealth

Paper 02: Wages and immigration

Signed 1 October 2026.

Plank: Pause large-scale low-wage immigration and enforce it against the employers who hire illegally, not the workers they hire.

1. What we want

We want a labor market where the wage rises when a worker is scarce, because that is the most reliable mechanism for pushing working-class pay up. For a long stretch, cheap and often illegal labor has let employers meet demand without raising pay, and the country has let them. We want to close that option with mandatory employment verification for every employer at the federal level. The penalty for hiring an unauthorized worker falls on the business, up to federal debarment, meaning a bar on receiving federal contracts, loans, and grants. What happens to the worker is a separate question, addressed in its own paper. At the state level, we want verification tied to the license a business already needs to operate. At the local level, we want verification of every contractor and subcontractor working on a public project paid with public money. We want new low-wage guest worker visas brought down to near zero, because a guest worker program is a legal channel for exactly the wage suppression we are trying to end. None of this is a permanent wall; it is a pause on new low-wage arrivals, long enough for wages to catch up. The pause also gives the country time to settle what happens with the roughly 14 million unauthorized immigrants already here before it decides what the next chapter of immigration policy looks like. That pause reflects our own judgment about how much change a country can absorb at once. It is not a conclusion from the studies in the next section; the wage evidence and the pause argument are two separate claims, and we are not going to blur them together.

2. The evidence

Economists disagree about how immigration affects native wages, so any paper that treats the question as settled is not being honest with the reader.

George Borjas, the Harvard economist most associated with the negative-wage-effect view, models immigration as adding workers who substitute closely for natives at similar skill and experience levels. Using the 1990–2010 immigrant influx, he estimates wages of native workers without a high school diploma fell 6.2 percent in the short run, before capital adjusts, and 3.1 percent in the long run once it does. In an earlier model using 1980–2000 data he found a larger short-run effect, 8.2 percent, falling to 4.8 percent long run (Borjas testimony to the Senate Judiciary Committee, 2016).

Giovanni Peri and Gianmarco Ottaviano take the opposing view. Their central objection to Borjas is that immigrants and natives with the same schooling are not perfect substitutes (they tend to specialize in different tasks even within the same education bracket), and once that is built into the model, the picture changes. On their estimates, the 1990–2006 immigration wave lowered wages for native high school dropouts by 0.7 percent in the short run and then raised them 0.3 percent in the long run (Ottaviano and Peri, NBER Working Paper 12497). A broader survey of the empirical literature finds two-thirds of published wage-effect estimates cluster between minus 0.1 and plus 0.1 percent for every one percent rise in the immigrant share of the workforce, a real but small effect in either direction (Peri, IZA World of Labor).

The single most-argued-over natural experiment is the 1980 Mariel boatlift, when roughly 125,000 Cubans arrived in Miami within months, expanding the city's low-skilled workforce by about 7 percent almost overnight. David Card's original 1990 study found essentially no effect on the wages or unemployment of Miami's existing low-skilled workers, including its Cuban population, and this became the empirical anchor for the view that immigration does not depress wages. Borjas re-examined the same data in 2017 and reported a very different result: wages for Miami high school dropouts fell 10 to 30 percent relative to comparison cities, concentrated among the workers most likely to compete directly with the new arrivals. Michael Clemens and Jennifer Hunt then argued that Borjas's result is a statistical artifact: the survey that supplied the wage data changed how it sampled low-skilled Black workers at almost exactly the same moment as the boatlift, and that alone can produce the appearance of a wage drop. A later reanalysis using synthetic control methods, comparing Miami to a weighted blend of similar cities, found no significant difference from Card's original null result. Borjas maintains the timing does not support the sampling-artifact explanation. The dispute has not been resolved by the profession, and we are not going to pretend it has.

The bracero program brought Mexican farm labor into the United States from 1942 to 1964 under a series of bilateral agreements. It ended December 31, 1964, when Congress declined to renew the underlying statute after twenty-two years of biennial extensions. Economists Michael Clemens, Ethan Lewis, and Hannah Postel studied what happened to farm wages in the states most exposed to the program's end. They found no measurable wage or employment gain for domestic farmworkers (American Economic Review, 2018), which cuts against the simple version of our own argument. Farms did not raise pay to attract replacement labor; they mechanized faster and shifted which crops they grew. Cutting off a low-wage labor supply does not automatically raise the wage of the job that supply used to fill; it can just as easily eliminate the job. That is a real possibility for some of the work now done by unauthorized labor, and we say so rather than promising a wage increase that history does not guarantee.

3. Why it happened

Congress did pass real sanctions in 1986, so it would be wrong to say sanctions were never attempted. The legal structure that let this persist began with that law, which was built to fail.

The Immigration Reform and Control Act of 1986 was the first federal law to punish employers for knowingly hiring unauthorized workers. It required every new hire to complete an I-9 form backed by one of a long list of acceptable documents, but it did not give employers or the government a reliable way to check whether those documents were genuine. Fraudulent Social Security cards and green cards became widely available almost immediately, and because prosecutors had to prove an employer "knowingly" hired someone unauthorized, a facially valid fake document was close to a legal shield. Worksite enforcement was also never funded or staffed as a serious federal priority in the years after the law passed. The result: the law that was supposed to end illegal hiring coincided with the unauthorized population more than tripling over the following decades (Migration Policy Institute, "IRCA in Retrospect").

A real verification tool did eventually exist. The 1996 Illegal Immigration Reform and Immigrant Responsibility Act created the Basic Pilot Program, an electronic system that checked a new hire's information against Social Security and immigration records instead of relying on the documents themselves. It was renamed E-Verify in 2007. But it was written as voluntary, and it has stayed voluntary for essentially every private employer in the country for three decades. As of mid-2025, only about 14 percent of U.S. employers were enrolled (American Immigration Council). The one place the federal government made it mandatory was its own contractors: an executive order first issued in 1996 and amended in 2008 was implemented through a Federal Acquisition Regulation clause that took effect September 8, 2009, requiring companies with new federal contracts of 120 days or more to use E-Verify on their employees. Where the federal government had leverage over who it was already paying, it used mandatory verification. It has never applied the same rule to employers it does not pay.

States that acted on their own show what a real mandate does. Arizona's Legal Arizona Workers Act, effective January 1, 2008, required every employer in the state to use E-Verify and threatened a business's license for a first violation and permanent revocation for a second. Enrollment went from about 300 Arizona employers in March 2007, the month the law passed, to roughly 15,000 by mid-2008 and 36,000 by January 2011. Researchers at the Public Policy Institute of California found the state's unauthorized working-age population fell by about 17 percent, or roughly 92,000 people, in the first year the law was in effect. The mechanism that failed at the federal level in 1986 for lack of a real verification tool and enforcement worked at the state level in 2008 once both were in place and the penalty was a business's ability to operate.

4. What to change

Federal: make employment verification mandatory for every employer in the country, not just federal contractors, using the E-Verify infrastructure that already exists. The penalty for hiring an unauthorized worker is federal debarment, a bar on the employer receiving federal contracts, loans, and grants. This paper is about where the penalty for illegal hiring falls; what happens to the worker is a separate question, addressed in its own paper. Cut new low-wage guest worker visas, the H-2A agricultural program (315,328 visas issued in fiscal year 2024) and the H-2B non-agricultural program (roughly 131,000 issued over the same period), toward zero, phased down on a fixed schedule rather than ended overnight, since farms and seasonal employers built current operations around this labor and need time to adjust or mechanize.

State: tie business licensure to E-Verify enrollment and compliance, on the Arizona model. As of 2024, 21 states already condition some contracts or business licenses on E-Verify use; the remaining states should adopt the same structure, at the scale of "every employer," not just the largest ones or those with public contracts. That expansion is not friction-free: in early 2026, several Republican-led states weighing mandatory E-Verify ran into pushback from employers worried about labor shortages and compliance costs, a real cost this paper does not pretend away [16].

Local: require every contractor and subcontractor on a public works project to run E-Verify on their crews as a condition of the contract, with the same license-level penalty for violation, enforced by whichever agency issues the contract.

5. What a bill would do

The model bill makes E-Verify mandatory for every employer above zero employees, replacing the current patchwork of state size thresholds. Two of its clauses do the real work. The penalty clause triggers federal debarment, a bar on the employer receiving federal contracts, loans, or grants, for a fixed term on the first confirmed violation and permanently on the second; states that want to add the Arizona model's license-suspension penalty on top can do so through the parallel state bill described above. The phase-down clause steps H-2A and H-2B allocations down on a published multi-year schedule toward a de minimis cap, rather than a single cliff, so agricultural and seasonal employers have a fixed and known runway to adjust staffing or invest in mechanization. A third, lighter provision requires an independent report to Congress every two years on wage trends in the occupations most exposed to unauthorized labor, required to report increases, decreases, or no measurable change alike.

6. What would prove us wrong

Several years after employer verification is fully enforced and guest worker visas are down sharply, if wages in the occupations that relied most on unauthorized or guest labor (construction, meatpacking, agriculture, hospitality) have not risen relative to their pre-enforcement trend, adjusted for the overall labor market, the wage argument for this policy is wrong, in the same way it was wrong for bracero-era farm wages. If those employers simply automate or cut output instead of raising pay, as happened after 1964, that is a measurable, specific outcome we would have to report honestly and not explain away. The federal mandate closes the gap when people move to a non-enforcing state rather than leave unauthorized work altogether, so if mandatory-verification enforcement itself turns out to depress the unauthorized population by less than the Arizona precedent suggests it should, once corrected for national mobility between states, its failure to close that gap would be a second, separate way this policy could fail to deliver what we say it will.

Sources

  1. Michael A. Clemens, Ethan G. Lewis, and Hannah M. Postel, "Immigration Restrictions as Active Labor Market Policy: Evidence from the Mexican Bracero Exclusion," American Economic Review 108, no. 6 (2018): 1468–1487. https://www.aeaweb.org/articles?id=10.1257%2Faer.20170765

  2. George J. Borjas, "Immigration and the Labor Market," testimony to the U.S. Senate Committee on the Judiciary, March 16, 2016. https://www.judiciary.senate.gov/imo/media/doc/03-16-16%20Borjas%20Testimony.pdf

  3. Gianmarco I.P. Ottaviano and Giovanni Peri, "Rethinking the Effects of Immigration on Wages," NBER Working Paper 12497, 2006. https://www.nber.org/system/files/working_papers/w12497/w12497.pdf

  4. Giovanni Peri, "Do Immigrant Workers Depress the Wages of Native Workers?," IZA World of Labor, 2014. https://wol.iza.org/uploads/articles/42/pdfs/do-immigrant-workers-depress-the-wages-of-native-workers.pdf

  5. David Card, "The Impact of the Mariel Boatlift on the Miami Labor Market," Industrial and Labor Relations Review 43, no. 2 (1990). https://davidcard.berkeley.edu/papers/mariel-impact.pdf

  6. George J. Borjas, "The Wage Impact of the Marielitos: A Reappraisal," Industrial and Labor Relations Review 70, no. 5 (2017). https://journals.sagepub.com/doi/10.1177/0019793917692945

  7. Michael A. Clemens and Jennifer Hunt, "The Labor Market Effects of Refugee Waves: Reconciling Conflicting Results," Industrial and Labor Relations Review 72, no. 4 (2019). https://www.researchgate.net/publication/330749774_The_Labor_Market_Effects_of_Refugee_Waves_Reconciling_Conflicting_Results

  8. Giovanni Peri and Vasil Yasenov, "The Labor Market Effects of a Refugee Wave: Synthetic Control Method Meets the Mariel Boatlift," Journal of Human Resources 54, no. 2 (2019). https://www.jointdatacenter.org/literature_review/the-labor-market-effects-of-a-refugee-wave-synthetic-control-method-meets-the-mariel-boatlift/

  9. Pew Research Center, "Record 14 Million Unauthorized Immigrants Lived in the US in 2023," August 21, 2025. https://www.pewresearch.org/race-and-ethnicity/2025/08/21/u-s-unauthorized-immigrant-population-reached-a-record-14-million-in-2023/

  10. Migration Policy Institute, "IRCA in Retrospect: Guideposts for Today's Immigration Reform," 2011. https://www.migrationpolicy.org/research/irca-retrospect-guideposts-today-s-immigration-reform

  11. American Immigration Council, "E-Verify: Widening Use, Lingering Challenges," 2025. https://www.americanimmigrationcouncil.org/blog/everify-states-use-lingering-challenges/

  12. E-Verify.gov, "History and Milestones." https://www.e-verify.gov/about-e-verify/history-and-milestones

  13. E-Verify.gov, "Federal Contractors" (Executive Order 12989, as amended, and FAR Subpart 22.18, effective September 8, 2009). https://www.e-verify.gov/employers/federal-contractors

  14. Sarah Bohn, Magnus Lofstrom, and Steven Raphael, "Lessons from the 2007 Legal Arizona Workers Act," Public Policy Institute of California / Review of Economics and Statistics. https://www.ppic.org/publication/lessons-from-the-2007-legal-arizona-workers-act/

  15. U.S. Citizenship and Immigration Services / E-Verify.gov, "The Arizona Mandatory E-Verify Experience: Evaluation Findings," December 2010. https://www.e-verify.gov/sites/default/files/everify/data/EVerifyArizonaMandExperience2012.pdf

  16. Stateline, "E-Verify Requirements Draw Business Pushback in Some Republican States," February 26, 2026. https://stateline.org/2026/02/26/e-verify-requirements-draw-business-pushback-in-republican-states/

  17. U.S. Department of State, Bureau of Consular Affairs, "Nonimmigrant Visa Statistics," Report of the Visa Office 2024 (315,328 H-2A visas issued, fiscal year 2024). https://travel.state.gov/content/travel/en/legal/visa-law0/visa-statistics.html

  18. U.S. Departments of State and Labor, "H-2B Temporary Non-Agricultural Visa Program: Nonimmigrant Visa Issuance," Fiscal Year 2024 factsheet (131,442 H-2B visas issued). https://www.dol.gov/sites/dolgov/files/ETA/oflc/pdfs/DOS-DOL_H-2B_Selected_Statistical_Factsheet_FY2024.pdf

  19. Texas State Historical Association / Bracero History Archive, on Public Law 78 (1951) and the program's expiration December 31, 1964. https://www.tshaonline.org/handbook/entries/bracero-program