The $1.2 Trillion Question: What Immigration Actually Costs and Creates in the American Economy

The United States has approximately 47.8 million foreign-born residents — about 14.3% of a total population of 334.9 million [1]. That share has roughly doubled since 1990. Over the same period, the U.S. economy grew from $5.9 trillion to over $31.4 trillion in nominal GDP, nonfarm employment expanded from 109 million to over 158 million jobs, and total crime fell by roughly 60% [2][3][4].

Whether immigration caused, coincided with, or occurred despite those trends is the subject of one of the most data-rich and least resolved debates in economics. The answer depends on which immigrants you are talking about, which time horizon you examine, and which costs and benefits you choose to count.

This report treats legal permanent residents, temporary visa holders, refugees and Temporary Protected Status (TPS) recipients, and undocumented immigrants as distinct populations with different economic profiles — because they are. Lumping them together produces statistics that obscure more than they reveal.

Source: U.S. Census Bureau / American Community Survey

Data as of Mar 21, 2026CSV

The Fiscal Ledger: Federal Gains, Local Costs

The Congressional Budget Office estimated in 2024 that the surge in immigration beginning in 2021 will add $1.2 trillion in federal revenues over the 2024–2034 period, primarily through individual income taxes and payroll taxes. The same immigration surge is projected to lower federal deficits by $0.9 trillion over the same decade, with annual revenue increases reaching $167 billion — or 2.2% of total revenues — by 2034 [5].

That is the federal picture. The state and local picture is different.

The CBO found that the immigration surge imposed measurable fiscal burdens on state and local governments, particularly in 2023 when the influx peaked. The largest cost increases were in primary and secondary education, shelter and related services, and border security operations. State and local tax revenues from immigrants also grew, but spending grew faster [6].

The 2017 National Academy of Sciences report, updated in 2024, found that over the course of their entire lives, immigrants tend to contribute slightly more in taxes than they receive in federal benefits [7]. But that lifetime average masks enormous variation. High-skilled immigrants on H-1B visas or with advanced degrees are large net fiscal contributors from the day they arrive. Refugees and undocumented immigrants with low education levels are net fiscal costs for their first 15 to 20 years, with the fiscal balance turning positive only in the second generation — their U.S.-born children who attain higher education and earnings [7].

The Manhattan Institute's 2025 update on the fiscal impact of immigration emphasized that the fiscal calculus depends heavily on the level of government examined. At the federal level, immigrants — including undocumented immigrants — are net contributors. At the state and local level, particularly in jurisdictions with high concentrations of immigrant families with children, the costs of education, healthcare, and social services can exceed the tax revenue immigrants generate locally [8].

The Welfare Debate: Cato vs. CIS

Two organizations have produced dueling analyses of immigrant welfare usage, both using the Census Bureau's 2024 Survey of Income and Program Participation (SIPP), and reaching opposite conclusions.

The Cato Institute's January 2026 analysis found that immigrants consume less welfare per capita than native-born Americans at every income level. Cato measured individual recipients and attributed benefits received by U.S.-born children to the U.S.-born category, not to their immigrant parents [9]. The finding: immigrant adults use fewer means-tested benefits and entitlement programs than comparable native-born adults. Cato's broader fiscal analysis concluded that immigrants reduced federal deficits by $14.5 trillion since 1994 [10].

The Center for Immigration Studies reached a different conclusion using the same survey data. CIS found that 53% of all households headed by immigrants use one or more welfare programs, compared with 37% of U.S.-born households [11]. The difference in methodology is the crux: CIS counts at the household level and attributes all welfare consumed within an immigrant-headed household — including benefits received by U.S.-citizen children — to the immigrant household.

Both approaches are methodologically defensible. The question is what you are trying to measure. If you want to know whether immigrant adults personally consume more welfare than native-born adults, Cato's individual-level analysis answers that: they do not. If you want to know the total fiscal burden associated with immigrant-headed households, including their U.S.-citizen dependents, CIS's household-level analysis answers that: it is higher than for native-born households [12].

Roughly one-fourth of all children in poverty in the United States are U.S.-born with an immigrant parent. As American citizens, these children have full welfare eligibility. CIS argues that this fiscal cost is a direct consequence of immigration policy and should be counted as such. Cato argues that attributing the costs of American citizens to their parents' country of birth distorts the picture [12][11].

Wages: The Borjas-Card Divide

No question in immigration economics has generated more academic dispute than the wage effect. The debate is anchored by two economists whose work reaches fundamentally different conclusions.

The Case That Immigration Depresses Wages

George Borjas of Harvard has argued for decades that immigration reduces wages for competing native-born workers. His framework holds capital fixed in the short run and estimates that a 10% increase in the labor supply from immigration reduces wages for competing workers by approximately 3% [13]. Applied to the period 1980–2000, when the immigrant share of the U.S. labor force roughly doubled, Borjas estimated that immigration reduced the wages of native-born high school dropouts by 7.4% [13].

Borjas's 2017 reappraisal of the Mariel Boatlift — the 1980 arrival of 125,000 mostly low-skilled Cuban immigrants to Miami — found that wages for high school dropouts in Miami dropped substantially in the years following the boatlift, contradicting David Card's original finding of no wage effect [14].

Steven Camarota of the Center for Immigration Studies has extended this argument, asserting that the evidence shows immigration reduces wages "significantly" for low-skilled native workers, particularly Black Americans without college degrees who compete most directly with low-skilled immigrants in construction, food service, and manual labor [15].

The Case That Immigration Has Minimal or Positive Wage Effects

David Card of UC Berkeley — whose 1990 study of the Mariel Boatlift is the most cited natural experiment in immigration economics — found no discernible negative effect on wages or unemployment for Miami's low-skilled workers after the boatlift increased the city's labor force by 7% [16]. Card's framework allows for capital adjustment: if employers invest in response to a larger labor supply, the capital-labor ratio stabilizes and average wages remain roughly unchanged.

Giovanni Peri and Vasil Yasenov re-examined the Mariel evidence using a synthetic control method and confirmed Card's original finding: no significant wage departure between Miami and its control cities from 1980 to 1983 [17]. Peri's broader research suggests immigration can have a slightly positive effect on native wages because immigrants and native workers are not perfect substitutes — they tend to specialize in different tasks, with immigrants concentrating in manual labor and native-born workers shifting toward communication-intensive roles [17].

Recent NBER research examining the period 2000–2022 found that immigration had a positive effect of +1.7% to +2.6% on wages of less-educated native-born workers, a finding that directly contradicts the Borjas framework [18].

Why They Disagree

The core technical disagreement concerns the elasticity of substitution between immigrant and native workers, and the speed of capital adjustment. If immigrant workers are close substitutes for native workers and capital is slow to adjust, Borjas is right: wages fall. If immigrants and natives specialize in complementary tasks and capital adjusts within a few years, Card and Peri are right: wages hold steady or rise.

The honest answer is that both mechanisms operate, and the net effect depends on the specific labor market, time horizon, and skill composition of the immigrants in question. A surge of low-skilled workers into a local market with fixed capital likely does depress wages in the short run. Over a longer period, with capital adjustment and task specialization, the effect diminishes or reverses. Neither camp has definitively refuted the other.

Source: FRED / Bureau of Economic Analysis

Data as of Mar 21, 2026CSV

Sector-by-Sector: Where Immigrants Work and What Happens Without Them

Foreign-born workers are not distributed evenly across the economy. They are concentrated in agriculture (where they represent 68% of the workforce, with approximately 42% being unauthorized), construction (roughly 30%), meatpacking (over 50%), home healthcare (25%), and technology (where H-1B holders fill specialized roles) [19].

Agriculture

The American Enterprise Institute's 2025 report on immigration enforcement and agriculture documented that agricultural employment dropped by 155,000 workers between March and July 2025 — a period that would normally see seasonal hiring increases. The comparable period in 2024 saw a 2.2% employment increase [19]. The consequences were tangible: significant shares of edible produce went unharvested, increasing food waste, lowering yields, and raising production costs. States with the highest concentration of undocumented farmworkers — California, Florida, Texas, Washington — experienced the most acute disruptions [19].

Construction

The 10 states with the highest concentration of undocumented immigrants in the construction industry saw a 0.1% drop in construction employment during a period when other states saw a 1.9% increase [20]. Construction companies shed 10,000 jobs since May 2025. The National Association of Home Builders has repeatedly identified workforce shortages as the leading cause of construction project delays, estimating that the industry needs 500,000 additional workers annually to meet housing demand [20].

The Broader Labor Market in 2025

Between January and August 2025, the number of foreign-born workers in the U.S. declined by more than one million [20]. Agriculture, construction, food processing, transportation, hospitality, caregiving, and certain manufacturing segments all reported tighter labor conditions, reduced staffing, and increased difficulty meeting production needs [20].

CNBC reported in August 2025 that economists were observing measurable labor force contraction attributable to immigration enforcement policies, with downstream effects on output in labor-intensive industries [21].

Source: Bureau of Labor Statistics

Data as of Mar 21, 2026CSV

The Price You Pay: Wages vs. Consumer Costs

If immigration restrictionists are correct that reducing immigration would raise wages for native-born workers, what would that wage increase actually look like — and what would it cost consumers?

The National Academies estimated that each 1% increase in the supply of immigrant workers reduces wages by approximately 0.3% [22]. Inverting this: removing immigrants from a sector would raise wages for remaining workers, but by a modest amount relative to the price increases consumers would face.

Construction and extraction workers' compensation accounts for 2.2% of GDP. Cleaning and maintenance is 1%. Food preparation and serving is 1.1%. Healthcare support is 0.9% [23]. These are small GDP shares, which means wage increases in these sectors would produce relatively small aggregate price effects — but those price effects would be concentrated in housing, food, and services, the categories that consume the largest share of low- and middle-income household budgets.

The Peterson Institute estimated that deporting 1.3 to 8.3 million undocumented immigrants would reduce U.S. real GDP by as much as 7% by 2028 while increasing inflation [24]. The National Foundation for American Policy projected in October 2025 that deportations could cut annual economic growth by almost one-third by 2035 [24].

CIS's Steven Camarota has pushed back on the inflation argument, contending that immigration cannot significantly reduce inflation because the labor-cost share of immigrant-heavy sectors is too small relative to total economic output. In this framing, the wage suppression effect on native workers is real but the consumer price benefit is marginal — the worst of both worlds for low-wage American workers [23].

The distributional question is uncomfortable for both sides. Employers and consumers benefit from cheaper labor. Low-wage native-born workers — disproportionately Black and Hispanic Americans without college degrees — bear the competitive pressure. The moral arguments (compassion for immigrants, rule of law) should not be confused with the economic data, which shows a real tension between aggregate economic gains and concentrated distributional costs.

Immigrant Entrepreneurs: The Job Creation Record

The American Immigration Council's 2025 analysis of the Fortune 500 found that 46.2% of America's largest companies — 231 out of 500 — were founded by immigrants or their children, the highest share since tracking began in 2011. These companies generated $8.6 trillion in revenue in fiscal year 2024 and employed over 15.4 million people worldwide [25].

At the small business level, Gusto's 2024 survey found that 17% of new business owners in 2023 were born outside the United States. One in three founders was either an immigrant or the child of an immigrant [26]. Immigrant-owned businesses were more likely to have employees (91% vs. 84% of all new businesses) and more likely to plan additional hiring [26].

Harvard Business School's 2024 working paper on immigrant entrepreneurship found that immigrants start businesses at higher rates than the native-born across nearly every demographic category, with the gap widest for immigrants who have been in the country between 5 and 15 years [27]. The OECD reported that between 2011 and 2021, migrant self-employment created 3.9 million jobs across 25 member countries, representing 15% of total employment growth [28].

These numbers are concentrated at both ends of the spectrum. Immigrant entrepreneurship includes both the Indian-born engineer founding a tech startup in Silicon Valley and the Salvadoran-born tradesman starting a landscaping company in Houston. The economic profile — capital intensity, job quality, tax revenue generated — varies enormously.

Social Security and Medicare: The Undocumented Subsidy

Undocumented immigrants paid $25.7 billion in Social Security taxes, $6.4 billion in Medicare taxes, and $1.8 billion in unemployment insurance taxes in 2022 [29]. They are not eligible for Social Security retirement benefits or Medicare health insurance, even after decades of contributions [29].

The Institute on Taxation and Economic Policy (ITEP) estimated in 2024 that undocumented immigrants pay an effective tax rate comparable to that of middle-income Americans when federal, state, and local taxes are combined [30]. Many file tax returns using Individual Taxpayer Identification Numbers (ITINs), a system the IRS created specifically for people who are required to file taxes but are ineligible for Social Security Numbers [29].

The Social Security Administration's actuaries have estimated that undocumented workers contribute a net surplus to the Social Security trust fund of approximately $12 billion per year — money paid in that will never be claimed in benefits [31]. If this revenue stream disappeared, the projected insolvency date for the Social Security trust fund would move closer by an estimated 1 to 2 years, depending on the assumptions used [31].

The Bipartisan Policy Center noted that this creates a paradox in the immigration debate: undocumented workers are simultaneously characterized as fiscal burdens and as people who subsidize a retirement system they cannot access [32].

Crime: What the Data Shows and What Drives Perception

The Aggregate Evidence

Ran Abramitzky of Stanford and colleagues published research in the December 2024 issue of the American Economic Review: Insights showing that immigrants have been less likely to be incarcerated than the U.S.-born for the last 150 years. Today, immigrants are 60% less likely to be incarcerated than all U.S.-born men and 30% less likely than white U.S.-born men [33].

A study published in the Proceedings of the National Academy of Sciences (PNAS) using Texas Department of Public Safety data found that undocumented immigrants were roughly half as likely to be arrested for violent and property crimes as native-born U.S. citizens [34]. Legal immigrants had even lower rates.

Between 1980 and 2022, the immigrant share of the population doubled while total crime dropped approximately 60% [4]. Correlation is not causation, but the data flatly contradicts the claim that more immigration produces more crime at the aggregate level.

The Cases That Drive Public Perception

Aggregate statistics do not erase specific incidents that generate legitimate community concern.

Springfield, Ohio: An estimated 12,000 to 20,000 Haitian immigrants, most with Temporary Protected Status, resettled in Springfield — a city of about 58,000 — beginning around 2013–2014. They came to fill open jobs in manufacturing and logistics. The influx reversed the city's population decline and produced higher wage growth than Ohio's average [35]. Haitian entrepreneurs opened at least 10 new businesses including restaurants, grocery stores, and a food truck [35].

But the rapid population increase strained housing, schools, and infrastructure. Residents reported landlords doubling rents, overcrowding in shared housing, and pressure on local services not designed for that pace of growth [36]. When TPS uncertainty intensified in 2025–2026, the economic contributions began to reverse: food insecurity rose, immigrant-owned businesses saw declining revenue, and money circulation slowed [36]. Ohio Governor Mike DeWine expressed concern about the economic impact of Haitians potentially losing their legal status [37].

Aurora, Colorado: Claims that the Venezuelan gang Tren de Aragua had "taken over" apartment complexes in Aurora became a national political flashpoint in 2024. The reality was more complicated. TDA is a real transnational criminal organization with a small presence in the Denver metro area. Aurora's mayor called the gang takeover claims "grossly exaggerated." The interim police chief stated that "gang members have not taken over this complex" [38]. Tenants interviewed by the Colorado Sun said they were more afraid of white supremacists and anti-immigrant vigilantes than of any Venezuelan gang [38]. But some residents did report intimidation, and the political attention produced real consequences for the broader immigrant community in Aurora [38].

Sanctuary cities: Research has consistently found that sanctuary jurisdictions — which limit cooperation between local law enforcement and federal immigration agents — have equal or lower crime rates than comparable non-sanctuary jurisdictions [4]. The mechanism proposed by researchers: when immigrant communities trust that contacting police will not result in deportation, they are more likely to report crimes and cooperate with investigations, making everyone safer.

Fraud: Scale and Context

The Deporting Fraudsters Act, introduced by Representative Dave Taylor in March 2025 and passed by the House 231–186, would make fraud involving SNAP, Social Security, or other federally funded programs grounds for deportation and bar foreign nationals who commit fraud from entering the U.S. [39].

The legislation was prompted partly by the Feeding Our Future scandal in Minnesota, in which an organized ring stole nearly $250 million in federal child nutrition funds. Several defendants were Somali nationals [39]. A separate childcare funding fraud scheme, also based in Minnesota, involved hundreds of millions in stolen federal funds [39].

These cases are real and serious. They are also not representative of the immigrant population at scale. Benefits fraud and identity theft occur across all demographic groups. The Government Accountability Office has documented billions in improper payments across federal programs annually, with the vast majority involving U.S. citizens [40]. Worksite identity theft — where undocumented workers use borrowed or fabricated Social Security Numbers — is a genuine problem that affects the credit histories and tax obligations of the identity theft victims, who are often other low-income Americans [40].

The question is whether the legislative response — making fraud a deportable offense specifically for immigrants — addresses a problem unique to immigrant communities or creates a parallel punishment system for conduct that is already illegal for everyone.

Source: Congressional Budget Office

Data as of Mar 21, 2026CSV

High-Immigration Metro Areas: The Real-World Laboratory

The Brookings Institution's Metro Monitor 2026 examined the relationship between immigration and regional economic performance across U.S. metropolitan areas over the past decade.

The findings: metro areas with larger increases in the foreign-born share of their working-age population saw stronger growth in gross metropolitan product and employment between 2014 and 2024. They also recorded higher productivity growth and stronger wage growth for both native-born and foreign-born workers [41].

The typical native-born worker in metro areas with larger foreign-born population growth earned approximately $3,000 more than a comparable worker in metro areas with lower foreign-born growth, after adjusting for cost of living [41].

Miami, where immigrants make up 37% of the population and generate 38% of economic output, has seen sustained economic expansion despite — or because of — its status as one of America's most immigrant-dense cities [42]. Houston, where one in four residents is foreign-born, has experienced consistent employment growth and diversification of its economic base beyond energy [42]. Los Angeles, despite well-documented affordability challenges, has maintained lower unemployment than the national average for most of the past decade [42].

These are observational comparisons, not controlled experiments. High-immigration metro areas tend to be economically dynamic for reasons beyond immigration — they have ports, universities, industry clusters. But the data does not support the claim that high immigration correlates with worse economic outcomes for native-born workers at the metro level. If anything, the correlation runs the other direction.

Conservative economists have offered an alternative interpretation: high-immigration cities attract immigrants because they are already economically dynamic, not the other way around. Immigration follows economic opportunity rather than creating it. This selection effect makes it difficult to isolate immigration's independent contribution to metro-level prosperity [15].

The H-1B Question: Who Benefits From Skilled Immigration?

The H-1B visa program has become its own contested terrain. In September 2025, the Trump administration implemented a $100,000 fee per new H-1B petition — up from the previous $2,000 to $5,000 range [43].

The fee's impact is asymmetric. Amazon, the largest sponsor of H-1B workers in fiscal year 2025 with over 10,000 visa holders, can absorb the cost. So can Microsoft, Meta, Apple, and Google [43]. For a startup, a research lab, a rural hospital, or a public university, a $100,000 fee per foreign hire is prohibitive [44].

Northeastern University researchers reported in September 2025 that the fee hike would hit startups and mid-sized companies hardest, potentially pushing them to abandon foreign hiring entirely [44]. Large technology companies, meanwhile, have fallback options: expanding R&D centers abroad, using intracompany L-1 transfer visas, or simply paying the fee as a cost of doing business [44].

The H-1B and L-1 Visa Reform Act of 2025, supported by the administration, raises the wage floor to the median for skill level 2, shortens typical H-1B duration to three years, and tightens compliance requirements [43]. Proponents argue the reforms protect American workers from wage undercutting. Critics argue they consolidate the program's benefits among the largest corporations while eliminating access for the small and mid-sized employers where H-1B workers often have the greatest marginal impact.

The DACA and Natural Experiment Evidence

Beyond the Mariel Boatlift, several other natural experiments and policy changes have produced evidence on immigration's economic effects.

Studies of the H-1B lottery — which randomly assigns visas among qualified applicants — have found that firms receiving H-1B workers show increased patent filings and no measurable reduction in native employment [18]. The random assignment addresses the selection problem that plagues most immigration research.

Research on DACA (Deferred Action for Childhood Arrivals), implemented in 2012, found that recipients experienced increased employment, higher wages, and greater likelihood of starting businesses [18]. The program's limited scope — approximately 600,000 active recipients — makes it a relatively clean test of what happens when unauthorized immigrants receive work authorization without a path to permanent residence.

State-level policy variations provide additional evidence. States that adopted restrictive immigration enforcement measures (Arizona's SB 1070, Alabama's HB 56) experienced agricultural labor shortages and increased food prices without measurable improvements in native-born employment or wages in the affected sectors [19].

The strongest evidence from natural experiments suggests that moderate, well-managed immigration flows produce small positive or neutral effects on native wages and employment. Large, sudden, and concentrated influxes can produce short-term competitive pressure in specific labor markets. The policy-relevant question — how much immigration, of what kind, at what pace — does not have a single empirical answer.

The Chilling Effect: Costs of Enforcement

Immigration enforcement produces its own economic costs, some intentional and some not.

The American Immigration Council documented in 2025 that families in immigrant communities were avoiding healthcare facilities, resulting in children missing vaccinations and preventable conditions going untreated [20]. Public health researchers have identified a measurable "chilling effect" in which legal immigrants and U.S.-citizen children of immigrants reduce their use of public benefits — including programs they are legally entitled to — out of fear that participation will jeopardize a family member's immigration status [20].

Schools in high-enforcement areas have reported increased absenteeism among children of immigrant families. Hospitals have reported declines in emergency room visits that do not correspond to improvements in community health [20].

These costs are difficult to quantify but they are real. A child who misses vaccinations today becomes a public health risk tomorrow. A worker who avoids seeking treatment for an injury becomes a more expensive patient when the condition worsens. The enforcement savings must be weighed against these downstream costs, and the honest accounting on both sides of this ledger remains incomplete.

What Remains Unresolved

The economic evidence on immigration is extensive, high-quality, and genuinely contested among credible researchers. Several tensions remain unresolved:

The distributional problem. Immigration produces aggregate economic gains — larger GDP, more business formation, lower consumer prices in labor-intensive sectors. But those gains are not distributed evenly. Employers and consumers capture most of the surplus. Low-skilled native-born workers, particularly those without high school diplomas, bear a disproportionate share of the competitive pressure. The size of that burden is disputed (Borjas says it is large; Card and Peri say it is small) but its existence is not.

The fiscal level problem. Immigrants are net fiscal contributors at the federal level and net fiscal costs at the state and local level, particularly in the first generation. Federal policy determines how many immigrants enter; state and local governments pay for their schools and emergency services. This mismatch between who decides and who pays is a structural feature of the system, not a bug that better data will fix.

The time horizon problem. Over a lifetime, the average immigrant is a slight net fiscal positive. Over 10 years, low-skilled immigrants are net costs. Over 30 years, their children are large net contributors. The "right" time horizon for policy analysis is a political question, not an empirical one.

The counterfactual problem. Estimating what would happen with significantly less immigration requires assumptions about whether native-born workers would fill the vacated jobs, at what wage, with what productivity, and what would happen to the industries that could not find workers at any domestically available wage. The 2025 evidence from agriculture and construction suggests that many of these jobs simply go unfilled, with output declining rather than wages rising [19][20].

The economic case for immigration is strong in the aggregate and weaker at the margins where specific communities and workers absorb concentrated costs. The economic case against immigration is strongest at those same margins and weaker in the aggregate. Both cases deserve serious engagement, and the data supports neither the claim that immigration is an unqualified economic boon nor the claim that it is a net economic drain. The tension is real, the tradeoffs are measurable, and the resolution is political rather than empirical.