The $1.7 Trillion Question: What Immigration Really Costs and Creates in the American Economy

Few topics in American public life generate more heat and less light than immigration's economic impact. Advocates claim immigrants are an unalloyed blessing; restrictionists call them a fiscal drain. The actual research — from the Congressional Budget Office, the National Academies of Sciences, Harvard's George Borjas, UC Davis's Giovanni Peri, and dozens of peer-reviewed studies — paints a picture that should make partisans on both sides uncomfortable.

This report treats legal immigrants, undocumented immigrants, refugees and TPS holders, and H-1B workers as distinct populations with different economic profiles. Because they are.

The Fiscal Ledger: Who Pays and Who Receives

The most contested number in immigration economics is the net fiscal impact — the gap between what immigrants pay in taxes and what they consume in public services. The answer depends entirely on which immigrants you're counting, at which level of government, and over what time horizon.

At the federal level, the Congressional Budget Office projected in 2024 that the recent immigration surge would generate $788 billion in income and payroll taxes over the 2024–2034 period, while adding roughly $109 billion in costs through refundable tax credits and premium subsidies [1]. The Manhattan Institute's 2025 update found that a college-educated immigrant arriving at age 30 generates a net fiscal gain of $1.7 million over 30 years, while an immigrant without a high school diploma costs a net $130,000 over the same period when descendants are included [2].

At the state and local level, the picture is worse. The CBO estimated that the immigration surge created a potential net cost to state and local governments of $9.8 billion in 2023 alone — $28.6 billion in spending against $18.8 billion in revenue [3]. Schools, emergency rooms, and county jails bear costs that federal tax receipts never reimburse.

This federal-versus-local split is critical: immigrants can be a net positive for the U.S. Treasury while simultaneously straining the budgets of specific communities that absorb them. Both things are true.

The Welfare Debate: Cato vs. CIS

The Cato Institute and the Center for Immigration Studies (CIS) have produced dueling analyses of immigrant welfare consumption that reach opposite conclusions — and both are methodologically defensible, depending on what you measure.

Cato's finding: Immigrants consume 24% fewer welfare benefits per capita than native-born Americans. Noncitizens, including undocumented immigrants, consume 53% less. All immigrants were 14.3% of the population but consumed just 11.9% of estimated benefits in 2022 [4].

CIS's finding: 53% of all immigrant-headed households use at least one welfare program, compared with 37% of native-born households. Among undocumented immigrant-headed households, the figure is 59% [5].

The disagreement is not about the underlying data — it's about the unit of analysis. Cato counts benefits received by U.S.-born children of immigrants as going to U.S.-born citizens. CIS counts those same benefits as part of the immigrant household's consumption. A U.S.-citizen child receiving Medicaid because their undocumented parent earns poverty-level wages is, depending on your framework, either a native-born American receiving a benefit or an immigrant household accessing the welfare state through a citizen child as conduit. The reader should understand both framings before choosing one.

Source: FRED / Bureau of Economic Analysis

Data as of Mar 20, 2026CSV

Wages: The Borjas-Card Fault Line

No question in immigration economics is more contested than whether immigrants depress native-born wages. The debate largely maps onto two economists: Harvard's George Borjas and UC Berkeley's David Card.

Borjas's finding: Between 1980 and 2000, immigration reduced the average wage of native workers with similar education and experience by 3–4%, and by as much as 9% for the least-skilled workers (high school dropouts competing directly with immigrant labor) [6]. His framework treats immigrant and native workers with similar education as close substitutes.

Card and Peri's finding: When you relax Borjas's assumption of perfect substitutability — recognizing that immigrants and natives often do different jobs even within the same education bracket — the wage effect shrinks to near zero or turns slightly positive. Peri and colleagues found in a 2024 NBER paper that immigrants and natives are more complementary than competitive, with immigrants pushing natives into communication-intensive roles that pay better [7].

The honest answer: both are partially right. Borjas is likely correct that in specific local labor markets and narrow skill categories, immigration creates short-term wage pressure on the native-born workers most similar to new arrivals. Card and Peri are likely correct that the broader, longer-term effects include productivity gains and occupational upgrading that offset or exceed those losses. The distributional question remains: employers and consumers benefit from cheaper labor; low-wage native workers bear the displacement costs. This is an uncomfortable truth that neither side likes to state plainly.

The Sectors That Cannot Function Without Immigrants

The 2025 immigration enforcement surge has provided a real-time natural experiment in what happens when immigrant labor is rapidly removed from the economy.

Between January and August 2025, the foreign-born labor force declined by more than one million workers — the first such decline in decades. The effects were immediate and sector-specific [8]:

• Agriculture: Employment dropped 6.5% nationwide between March and July 2025, compared to a 2.2% increase in the same period of 2024. Foreign-born workers comprise roughly 70% of the farm workforce [8].
• Hospitality: Labor force growth collapsed to 0.2% in June 2025, compared with 1.5% in June 2024. About 145,000 fewer immigrants were working in restaurants through August 2025 [8].
• Construction: The 10 states with the highest concentration of undocumented construction workers saw a 0.1% drop in employment while other states saw a 1.9% increase. Nearly one in three contractors reported being impacted by enhanced ICE enforcement [9].
• Meatpacking: Undocumented immigrants fill an estimated 23% of jobs. In roofing, drywall, and concrete, unauthorized workers make up over 30% of the workforce [10].

Fresh vegetable prices rose 2% from April to July 2025. Restaurant food costs increased 3% between December 2024 and September 2025 [8]. These are not hypothetical models — they are measured consequences.

Source: FRED / Bureau of Labor Statistics

Data as of Mar 20, 2026CSV

The Replacement Question

A critical test of the restrictionist case is whether native-born Americans will fill jobs vacated by deported or deterred immigrants. The evidence so far is not encouraging for that thesis. Over the past two decades, the U.S.-born population with a high school diploma or less has significantly decreased, particularly among workers aged 25–44 [10]. The jobs these workers historically filled — meatpacking, harvesting, roofing, elder care — have seen minimal native-born entry even as wages have risen 10–20% in affected sectors [9].

Labor elasticity research suggests that to fill current agricultural vacancies with native-born workers alone, wages would need to roughly double, which modeling exercises project would push food price inflation toward 9% under large-scale deportation scenarios [10]. The question is not whether higher wages would attract some native workers — they would — but whether the price increases consumers would face are a trade-off the public is willing to accept.

Innovation and Entrepreneurship: The Top of the Ladder

The economic contribution of immigrants is not limited to manual labor. In 2025, 231 of the Fortune 500 companies — 46% — were founded by immigrants or their children, the highest share ever recorded. These companies generated $8.6 trillion in revenue and employed over 15.4 million people [11].

The breakdown: 108 companies had immigrant founders; 123 had founders who were children of immigrants. They dominated professional services (80% of Fortune 500 firms in that category), manufacturing (65.6%), and information (57.5%) [11].

In STEM, immigrants file a disproportionate share of U.S. patents, particularly in computing, electronics, medical devices, and pharmaceuticals. Research shows immigrant inventors also increase patenting rates among their U.S.-born colleagues through knowledge spillovers [12].

The CIS critique of the Fortune 500 statistic deserves engagement: many of these "immigrant-founded" companies were started generations ago by immigrants from very different waves (European, early 20th century) under very different conditions. Attributing their success to current immigration policy is a stretch. The counterpoint: recent immigrant founders — from Sergey Brin (Google) to Jensen Huang (NVIDIA) — demonstrate the pattern continues [13].

Social Security: The Demographic Time Bomb

The Social Security trust fund is projected to be depleted by 2033–2034, at which point benefits would face an automatic 21% cut [14]. The fundamental problem is demographic: the worker-to-beneficiary ratio has fallen from 5:1 in 1960 to 3:1 in 2023, and is projected to drop below 2.5:1 by mid-century [14].

Immigration is the only near-term lever that can shift this ratio. 77% of immigrants entering the country are working age (18–64) [15]. By 2040, immigration is expected to be the sole driver of U.S. population growth [15].

The Bipartisan Policy Center estimated that the 2013 immigration reform bill would have solved roughly 8% of the 75-year Social Security shortfall [15]. That is meaningful but modest — immigration alone cannot fix Social Security. However, the CBO's 2024 demographic outlook shows that a 50% reduction in immigration would accelerate the trust fund's depletion, while current or increased levels extend solvency by several years [16].

The arithmetic is straightforward: young workers paying into the system support older retirees drawing from it. Reducing immigration while the native birth rate sits at historic lows makes the math worse, not better.

Source: Bureau of Labor Statistics

Data as of Mar 20, 2026CSV

Crime: What the Data Shows vs. What People Feel

The peer-reviewed evidence on immigrant crime rates is among the most consistent findings in the field. A landmark PNAS study using Texas criminal records found that undocumented immigrants were 47% less likely to be convicted of a crime than native-born Americans. Legal immigrants had the lowest rates of all [17]. A Cato Institute analysis of national data found undocumented immigrants were 33% less likely to be incarcerated than the U.S.-born [18].

Between 1980 and 2022, the immigrant share of the U.S. population doubled while total crime dropped 60% [17]. This does not prove immigration caused the crime drop — but it decisively refutes the claim that more immigration means more crime at the aggregate level.

But aggregate statistics are cold comfort to specific communities. Springfield, Ohio — a city of 60,000 — absorbed 12,000–15,000 Haitian refugees under Temporary Protected Status in a matter of years. Local officials confirmed that refugees filled manufacturing and healthcare jobs and opened businesses, but also acknowledged severe strain on housing, schools, healthcare, and public services [19]. The problem was not that refugees were criminal — they largely were not — but that the pace and volume of resettlement exceeded the community's absorptive capacity.

In Aurora, Colorado, police linked 10 individuals to the Venezuelan gang Tren de Aragua across several apartment complexes. The city's mayor called broader gang-takeover claims "grossly exaggerated" in a city of 400,000, but the real incidents — however limited — fueled national narratives [20]. The lesson: dismissing community concerns because the aggregate data is favorable is a political failure, even if it's statistically accurate.

Sanctuary cities: A Journal of Economic Behavior & Organization study found that sanctuary counties have, on average, 35.5 fewer crimes per 10,000 people than non-sanctuary counties [21]. A PNAS study found sanctuary policies reduce deportations without increasing crime [21]. This contradicts the enforcement narrative but does not address the specific mechanism concerns that drive sanctuary opposition.

Fraud: Real but Contextualized

Benefits fraud involving immigrant communities is real and should not be dismissed. House Republicans spotlighted large-scale childcare funding fraud schemes in Minnesota and California. The Deporting Fraudsters Act, which passed the House 231–186 in early 2026, makes public benefits fraud a deportable and inadmissible offense [22].

Estimates place total fraud losses across all government programs at $233–$521 billion annually [22] — the vast majority committed by U.S. citizens, given that citizens are the vast majority of benefits recipients. Immigrant-linked fraud is a subset of a broader problem, not its primary driver. That said, organized fraud schemes exploiting immigration status — identity theft at worksites, fraudulent benefit claims through citizen children — are documented, prosecutable, and should be prosecuted without being used to characterize entire populations.

The H-1B Divide: Big Tech vs. Small Business

The H-1B visa program illustrates how immigration policy can simultaneously serve and undermine different segments of the economy. In September 2025, the Trump administration imposed a $100,000 fee on new H-1B petitions — up from $2,000–$5,000 previously [23].

Amazon, the largest H-1B sponsor with over 10,000 visa holders, can absorb this cost. A 15-person startup cannot. The fee, combined with a new wage-weighted lottery system, effectively prices small employers out of the high-skilled immigrant talent pool while large corporations can either pay up or route hires through foreign subsidiaries using L-1 intracompany transfer visas [23].

The U.S. faces a projected $480 billion annual loss from a shrinking STEM talent pool [12]. The H-1B cap has remained at 65,000 new visas since 1990 (plus 20,000 for advanced degrees), while demand has ballooned — over 470,000 petitions were filed in a recent cycle [9]. The program needs reform, but reform that makes it accessible only to the wealthiest employers is reform that entrenches market concentration.

Why Business Lobbies Want More Immigration

If restrictionists are correct that immigration suppresses wages and strains services, the consistent pro-immigration stance of business lobbies — including the U.S. Chamber of Commerce, the Associated General Contractors, the American Farm Bureau, and the National Restaurant Association — requires explanation.

The answer is not complicated: businesses benefit from a larger labor supply that holds down wage costs. The Chamber of Commerce explicitly states that the U.S. "needs more workers of all skill levels" [9]. Internal industry analyses show that in agriculture, dairy, construction, and hospitality, the available native-born workforce is structurally insufficient — not because Americans can't do these jobs, but because the working-age native population in relevant skill categories is shrinking, and the wage levels required to attract sufficient native workers would fundamentally alter cost structures.

This is the distributional tension at the heart of the debate: cheaper immigrant labor benefits employers (through lower costs), consumers (through lower prices), and the immigrants themselves (through higher wages than in their home countries) — while imposing real costs on native-born workers in competing occupations and local communities absorbing rapid population growth. Acknowledging all four groups' interests simultaneously is the prerequisite for honest policy discussion.

What the Consensus Actually Shows

A 2024 NBER working paper by Borjas and Breznau documented ideological bias in immigration economics research itself — researchers' political leanings predicted their findings, raising questions about the objectivity of the entire field [24]. This is a sobering finding that both sides should take seriously.

With that caveat, here is what the weight of evidence supports:

1. Immigration is a net fiscal positive at the federal level and a net fiscal negative at the state/local level, with the balance depending heavily on immigrants' education levels and time in the country [1][2][3].
2. Wage effects on native workers are real but small — concentrated among workers without college degrees competing directly with new arrivals, and partially offset by complementarity effects over time [6][7].
3. Immigrants commit crimes at lower rates than native-born Americans, including undocumented immigrants, though rapid resettlement can strain specific communities [17][18].
4. The U.S. economy structurally depends on immigrant labor in agriculture, construction, hospitality, and elder care, with no realistic near-term substitute [8][10].
5. Immigration extends Social Security solvency modestly but meaningfully, and reducing it accelerates the demographic crisis [14][15].
6. High-skilled immigration drives disproportionate innovation, but the H-1B system increasingly favors large corporations over small employers [11][12][23].
7. Benefits fraud exists but is a small fraction of overall fraud and should be prosecuted on its merits rather than used as a characterization of immigrant populations [22].
8. Both pro- and anti-immigration researchers exhibit ideological bias, and the honest answer on many contested questions is that the evidence is genuinely mixed [24].

The American economy does not need open borders, and it does not need a wall. It needs a system that distinguishes between high-skilled and low-skilled immigration, accounts for local absorptive capacity, enforces the law without destroying industries, and stops pretending that a question this complex has a bumper-sticker answer.