Key Points
Policies: We examine two distinct policy scenarios regarding the deportation of unauthorized immigrants: (i) A 4-year policy where 10 percent of unauthorized immigrants are removed annually over four years (2025 – 2028), with unauthorized immigration returning to baseline levels in 2029; (ii) A 10-year policy where 10 percent of unauthorized immigrants are removed each year for a period of ten years, ultimately removing all unauthorized immigrants and preventing the entry of new unauthorized immigrants thereafter. In both scenarios, immigration laws concerning authorized workers remain unchanged.
Budget: The 4-year policy would increase primary deficits by about $270 billion before economic feedback effects and $350 billion with economic feedback effects. The 10-year policy would cost about $862 billion before economic feedback effects and $987 billion with economic feedback effects. Besides lost revenue, these primary deficits include new required outlays that are in addition to the monies (about $170 billion) already allocated in the 2025 OBBBA for border security, interior enforcement, and deportation.
Economy: Not surprisingly, most aggregate economic variables fall. By 2034, the 4-year policy reduces GDP by 1.0 percent; the 10-year policy reduces it by 3.3 percent. By 2054, the 4-year policy reduces real GDP by 1.0 percent; the 10-year policy reduces it by 4.9 percent. But per-capita economic variables also mostly decline, including a fall in the average wage of 0.5 percent (4-year policy) and 1.7 percent (10-year policy). Output per capita also falls between 0.5 percent and 1.1 percent.
Distributional analysis: However, the impact on wages varies by skill class. High-skilled workers see wages fall over time, between 0.5 percent (4-year policy) and 2.8 percent (10-year policy). But wages for authorized (including US-born) low-skilled workers can increase. Under the 4-year policy, their wages increase by 1.1 percent by 2034 but eventually fall 0.6 percent by 2054. Under the 10-year policy, their wages increase by 5.0 percent by 2034 and mostly persist thereafter. As a result, for the median authorized low-skilled workers born today, these changes translate into lifetime losses of around $8,600 (4-year policy) to lifetime gains of $63,600 (10-year policy). Overall, while a permanent deportation policy can benefit low-income workers, reversals can leave low-income workers worse off over time as higher debt decreases wages.
Mass Deportation of Unauthorized Immigrants: Fiscal and Economic Effects
The United States currently has over 11 million unauthorized immigrants, equal to about 3.4 percent of the total population, who disproportionately occupy low-skilled jobs. Jobs held by unauthorized immigrants are largely complementary to jobs of authorized high-skilled workers. Policies that deport unauthorized immigrants, therefore, generally harm high-skilled workers. However, jobs held by unauthorized immigrants are imperfect substitutes for low-skilled workers. Still, overall, policies that deport unauthorized immigrants generally help low-skilled workers at reasonable values of elasticity of substitution.
In this brief, we analyze two deportation policy scenarios:
i) The 4-year policy: 10 percent of unauthorized immigrants are removed annually for four years (2025 – 2028) and unauthorized immigration is halted over this period with enhanced spending on border security. The annual net flow of unauthorized immigration returns to baseline levels in 2029.
ii) The 10-year policy: 10 percent of unauthorized immigrants are removed each year for a period of ten years, ultimately removing all unauthorized immigrants by the end of 2034. Enhanced spending on border security prevents unauthorized immigration during the first 10 years and beyond.
In both scenarios, immigration laws concerning authorized workers remain unchanged.
Without the mass deportation policies, including those policies in the 2025 OBBBA, Table 1 shows that the population is projected to rise from 337.2 million at the end of 2025 to 379 million by 2054, with the number of unauthorized immigrants projected to increase from 11.5 million in 2025 to 12.3 million in 2054.1
| 2025 | 2026 | 2027 | 2028 | 2029 | 2030 | 2031 | 2032 | 2033 | 2034 | 2044 | 2054 | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Total population | 337.2 | 339.3 | 341.2 | 343.2 | 345.1 | 347.0 | 349.0 | 350.7 | 352.5 | 354.1 | 368.2 | 379.0 | |
| Unauthorized population | 11.5 | 11.5 | 11.5 | 11.5 | 11.6 | 11.6 | 11.7 | 11.7 | 11.8 | 11.8 | 12.2 | 12.3 | |
| 4-Year Deportation Policy | |||||||||||||
| Total population | 336.0 | 336.8 | 337.4 | 338.2 | 340.2 | 342.1 | 344.1 | 346.0 | 347.8 | 349.6 | 364.1 | 375.3 | |
| Unauthorized population | 10.3 | 9.1 | 8.0 | 6.8 | 7.1 | 7.3 | 7.5 | 7.7 | 7.9 | 8.1 | 9.5 | 10.4 | |
| Difference in the total population relative to current-law baseline | -1.2 | -2.5 | -3.8 | -5.0 | -5.0 | -5.0 | -4.9 | -4.7 | -4.7 | -4.5 | -4.1 | -3.7 | |
| Difference in the unauthorized-immigrant population relative to current-law baseline | -1.2 | -2.4 | -3.5 | -4.7 | -4.5 | -4.3 | -4.2 | -4.0 | -3.9 | -3.7 | -2.7 | -1.9 | |
| 10-Year Deportation Policy | |||||||||||||
| Total population | 336.0 | 336.8 | 337.4 | 338.2 | 338.7 | 339.2 | 339.9 | 340.4 | 340.7 | 341.3 | 353.1 | 360.2 | |
| Unauthorized population | 10.3 | 9.1 | 8.0 | 6.8 | 5.7 | 4.5 | 3.3 | 2.2 | 1.0 | 0.0 | 0.0 | 0.0 | |
| Difference in the total population relative to current-law baseline | -1.2 | -2.5 | -3.8 | -5.0 | -6.4 | -7.9 | -9.1 | -10.4 | -11.9 | -12.8 | -15.1 | -18.8 | |
| Difference in the unauthorized-immigrant population relative to current-law baseline | -1.2 | -2.4 | -3.5 | -4.7 | -5.9 | -7.2 | -8.4 | -9.5 | -10.8 | -11.8 | -12.2 | -12.3 |
Note: Numbers reflect end-of-year values. The decrease in the total population is smaller than the decrease in the unauthorized population because, once unauthorized immigrants leave the United States, they no longer have U.S.-born children who would have added to the overall count.
Source: Penn Wharton Budget Model
Under the 4-year deportation scenario, the total U.S. population decreases by 1.2 million in the first year and by 5 million by the end of 2028. As the annual flow of unauthorized immigration returns to current-law levels in 2029, the gap gradually narrows; by 2054, the population is only 3.7 million smaller than the baseline. Much of this difference reflects timing rather than permanence, as some unauthorized immigrants deported under the policy would have emigrated voluntarily by 2054 even under current law. Additionally, the gap between the unauthorized immigrant population and the total population at the end of 2054 also accounts for the number of "missing" U.S.-born children—births that do not occur because their immigrant parents were removed and are no longer present in the U.S.
The 10-year deportation scenario results in larger and enduring demographic changes. The population falls by 5.0 million by the end of 2028, falls by 12.8 million by 2034, and falls by 18.8 million by 2054. Almost 12.3 million of the 18.8 million reduction is due to the reduction in unauthorized immigrants.
To analyze the policy’s impact on different populations, we categorize households into three distinct groups based on their education levels and legal status.
High-skilled households are native-born citizens, permanent residents, and visa-holding immigrants who have at least some college education. They pay taxes on all their income and are eligible for Social Security, Medicare, and Medicaid.2 These workers occupy positions with high marginal productivity and high wages.
Authorized low-skilled households include native-born citizens, permanent residents, and visa-holding immigrants with a high school diploma or less. They also pay taxes on all their income and are eligible for the same programs as high-skilled households. Their tasks require less formal schooling.
Unauthorized low-skilled households consist of unauthorized immigrants or individuals with only temporary permissions like DACA, Temporary Protected Status, pending asylum claims, or humanitarian parole. These workers rely on more informal employment relationships, reflecting the narrower set of occupations open to them. About 44 percent of the unauthorized low-skilled group pays income and payroll taxes,3 but they are not eligible for Social Security, Medicare, or Medicaid benefits.4 However, unauthorized workers often draw upon emergency Medicaid, public health programs, school lunch programs, WIC (Women, Infants, and Children) and local community health centers that receive federal grants.
By separating workers according to education, occupation, and legal status, the framework recognizes that labor contributes heterogeneously to output. Interactions among these groups amplify or dampen various economic effects.
A deportation program requires enforcement expenditures at the border and inside the border that increase primary deficits in addition to the lost tax revenue that would have otherwise been paid by unauthorized households subject to deportation.
In Table 2, we combine estimates from two studies: one from the American Immigration Council (AIC)5 and another from the American Action Forum (AAF).6 Both studies break the deportation process into the four stages: arresting, detaining, processing, and removing large numbers of undocumented immigrants.7 Averaging the two estimates yields a total cost of $70,236 per deportee.8 We calculate total deportation cost per year by multiplying the number of deportees in that year times the average cost per deportee.
| Stage of Enforcement | Per-Deportee Cost | ||
|---|---|---|---|
| AIC | AAF | Average | |
| Arrests | 8,750 | 21,397 | 15,074 |
| Detention / Monitoring | 82,750 | 5,324 | 44,037 |
| Legal Processing | 15,750 | 1,999 | 8,874 |
| Removal / Transport | 2,630 | 1,872 | 2,251 |
| Subtotal (deportation) | 109,880 | 30,591 | 70,236 |
Notes: All numbers are expressed in 2024 dollars, deflated with the GDP price index; unless otherwise specified, amounts are stated in U.S. dollars.
Sources: Penn Wharton Budget Model
AIC refers to American Immigration Council, "Mass Deportation: Devastating Costs to America, Its Budget and Economy," October 2924, available at https://www.americanimmigrationcouncil.org/sites/default/files/research/mass_deportation_report_2024.pdf
AAF refers to American Action Forum, "The Budgetary and Economic Costs of Addressing Unauthorized Immigration: Alternative Strategies," March 2015, available at https://www.americanactionforum.org/research/the-budgetary-and-economic-costs-of-addressing-unauthorized-immigration-alt/
The AAF study includes the cost of maintaining zero unauthorized immigration, estimated at $21.1 billion annually. Each year, we adjust the enforcement cost by the GDP deflator and the projected number of deportees.
| Provision | 2025 | 2026 | 2027 | 2028 | 2029 | 2030 | 2031 | 2032 | 2033 | 2034 | 2025 - 2034 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Outlays | 77.1 | 65.4 | 50.6 | 46.5 | -43.0 | -38.1 | -28.9 | -21.5 | -14.0 | -11.5 | 82.7 |
| Revenues | 1.4 | -4.2 | -7.4 | -17.0 | -23.6 | -24.7 | -24.5 | -28.0 | -29.2 | -30.1 | -187.4 |
| Change in Primary Deficit | 75.7 | 69.6 | 58.1 | 63.5 | -19.4 | -13.4 | -4.4 | 6.5 | 15.2 | 18.6 | 270.1 |
| Memorandum: | |||||||||||
| Increase (+) or decrease (-) in primary deficit with dynamic effects | 69.0 | 66.6 | 57.9 | 64.9 | -8.5 | -2.8 | 9.1 | 21.1 | 33.6 | 38.7 | 349.6 |
Source: Penn Wharton Budget Model
| Provision | 2025 | 2026 | 2027 | 2028 | 2029 | 2030 | 2031 | 2032 | 2033 | 2034 | 2025 - 2034 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Outlays | 77.1 | 65.4 | 50.6 | 46.5 | 42.7 | 54.2 | 50.5 | 63.7 | 69.2 | 41.4 | 561.4 |
| Revenues | 1.4 | -4.2 | -7.4 | -17.0 | -20.0 | -29.5 | -39.1 | -49.4 | -60.6 | -74.5 | -300.4 |
| Change in Primary Deficit | 75.7 | 69.6 | 58.1 | 63.5 | 62.6 | 83.6 | 89.6 | 113.2 | 129.8 | 116.0 | 861.8 |
| Memorandum: | |||||||||||
| Increase or decrease (-) in primary deficit with dynamic effects | 70.2 | 67.8 | 59.5 | 64.9 | 67.5 | 92.0 | 106.0 | 133.9 | 161.0 | 164.0 | 986.8 |
Source: Penn Wharton Budget Model
Table 3 outlines the fiscal impacts of the two policies.
The 4-year policy (Table 3A) leads to increased expenditures due to the costs associated with deporting unauthorized immigrants and enhanced border security measures. Despite reductions on Medicaid emergencies and other spending outlined above, total outlays would need to rise by $82.7 billion over the 10-year budget window. These new required outlays are in addition to the $170 billion already allocated in the 2025 OBBBA for enhanced border security, interior enforcement and deportation. Notice that outlays turn negative in the year 2030, helping to reduce some of these 10-year outlays. These values reflect reductions in federal spending, including on Medicaid emergencies and other spending outlined above, due to a lower level of authorized households (about $9 billion per year) after four years. But most of these cost savings come from recapture of outlays that are already financed under 2025 OBBBA that are no longer used for enhanced border security, interior enforcement and deportation associated with mass deportation after 4 years. Without this recapture, the 10-year outlay costs would increase to $184 billion.
As 44 percent of unauthorized immigrants contribute to income and payroll taxes, their removal results in a decrease in tax revenue, equal to $187.4 billion from 2025 to 2034. Consequently, primary deficits would increase by $270.1 billion over ten years. Factoring in dynamic effects, primary deficits increase by $349.6 billion, mainly due to lower economy-wide average wages that reduce tax revenue. As discussed below, wages fall for high-skilled workers who face the highest marginal tax rates.
The 10-year policy (Table 3B) extends deportation through 2034, increasing outlays by $561.4 billion over ten years. These new required outlays are in addition to the $170 billion already allocated in the 2025 OBBBA for enhanced border security, interior enforcement, and deportation. With a greater reduction in population compared to the 4-year policy, the loss in revenue grows proportionately to the number of unauthorized immigrants no longer paying taxes. Revenues would decline by $300.4 billion from 2025 to 2034. Primary deficits would increase by more $861.8 billion over the 10-year budget window. With dynamic effects, primary deficits increase to $986.8 billion.
Table 4 presents the economic impacts of both deportation policies.
| 2029 | 2034 | 2039 | 2044 | 2049 | 2054 | |
|---|---|---|---|---|---|---|
| Gross domestic product | -1.0 | -1.0 | -0.9 | -0.9 | -0.9 | -1.0 |
| Capital stock | -1.0 | -1.3 | -1.3 | -1.4 | -1.5 | -1.6 |
| Hours worked | -1.2 | -1.0 | -0.8 | -0.5 | -0.5 | -0.5 |
| Average wage | 0.0 | -0.2 | -0.3 | -0.4 | -0.5 | -0.5 |
| Consumption | -0.7 | -0.7 | -0.6 | -0.6 | -0.6 | -0.6 |
| Debt held by the public | 1.0 | 1.2 | 1.4 | 1.5 | 1.6 | 1.7 |
| GDP per capita | 0.2 | 0.0 | -0.1 | -0.3 | -0.3 | -0.5 |
Source: Penn Wharton Budget Model
| 2029 | 2034 | 2039 | 2044 | 2049 | 2054 | |
|---|---|---|---|---|---|---|
| Gross domestic product | -1.1 | -3.3 | -3.8 | -4.0 | -4.4 | -4.9 |
| Capital stock | -1.1 | -3.7 | -4.7 | -5.3 | -6.0 | -6.9 |
| Hours worked | -1.5 | -3.4 | -3.6 | -3.5 | -3.7 | -4.0 |
| Average wage | 0.0 | -0.4 | -0.8 | -1.1 | -1.3 | -1.7 |
| Consumption | -0.8 | -2.2 | -2.5 | -2.6 | -2.8 | -3.0 |
| Debt held by the public | 1.2 | 2.5 | 3.5 | 4.3 | 4.9 | 5.4 |
| GDP per capita | 0.2 | 0.1 | -0.2 | -0.4 | -0.6 | -1.1 |
Source: Penn Wharton Budget Model
As shown in Table 4A, under the 4-year policy, most aggregate economics variables (GDP, capital stock, hours worked and consumption) fall with the general resizing of the population base. However, the more interesting results are per capita. The general wage rate, averaged across all workers, fall slowly over time relative to current policy before deporting, eventually falling by 0.5 percent. So does GDP per capita. Intuitively, while unauthorized low-skilled workers are mostly substitutes for authorized low-skilled workers, unauthorized low-skilled workers are complements to high-skilled workers. Higher-skill workers, despite being fewer in number, still have a larger impact on the wage base and GDP. They are generally harmed by deportation more than authorized lower-skill workers are helped. A higher-skill worker also contributes more to taxes than lower-skill workers, especially in the presence of a progressive income tax system.
Under the 10-year policy, the total population decreases by five times more relative to the 4-year policy over the 2025-2054 period. Likewise, changes in aggregate variables are roughly five times larger by 2054, as shown in Table 4B. Per-capita variables also fall. The general average wage declines by over three times as much, eventually falling by 1.7 percent. Output per capita declines by twice as much, eventually falling by 1.1 percent.
| Deportation Policy | |||||
|---|---|---|---|---|---|
| 2034 | 2039 | 2044 | 2049 | 2054 | |
| 4-Year Policy | |||||
| High-Skilled | -0.4 | -0.5 | -0.5 | -0.5 | -0.5 |
| Authorized, low-skilled | 1.1 | 0.6 | -0.1 | -0.3 | -0.6 |
| Unauthorized, low-skilled | 2.5 | 1.6 | 0.6 | 0.2 | -0.4 |
| 10-Year Policy | |||||
| High-Skilled | -1.5 | -1.9 | -2.1 | -2.5 | -2.8 |
| Authorized, low-skilled | 5.0 | 5.0 | 4.5 | 4.7 | 4.7 |
| Unauthorized, low-skilled | - | - | - | - | - |
Note: Starting in 2034, there are no more unauthorized immigrants in the U.S. under the 10-year policy, so we cannot calculate a change in wages for those years.
Source: Penn Wharton Budget Model
Table 5 presents projected wage changes for workers categorized by skill level and authorization status under the two policy scenarios. The Appendix tests the sensitivity of the results to key parameter combinations.
Under the 4-year policy scenario, high-skilled wages decrease by 0.5 over time due to a reduction in the number of low-skilled workers that are available to complement the productivity of high-skilled workers as well as a reduction in the size of the capital stock. Wages for authorized low-skilled workers, however, increase by 1.1 percent by the year 2034, but those wages subsequently decline by 0.6 percent by the year 2054 relative to no mass deportations. Intuitively, authorized low-skilled workers eventually face more competition from the slow return of unauthorized low-skilled workers. But the size of the capital stock is smaller in 2054 (Table 4).
The 10-year removal policy results in a more substantial cumulative adjustment. High-skilled wages fall by 1.5 percent by 2034 and fall by 2.8 percent by 2054, reflecting a greater reduction in both low-skilled labor and capital. However, the elimination of unauthorized workers leads to a 4.7 percent increase in wages for authorized low-skilled workers. Despite the fall in aggregate capital, authorized low-skilled workers do not face the competition from unauthorized low-skilled workers.
Dynamic distributional analysis considers how a policy affects households across the income and age distribution, including the unborn (represented by a negative age index at the time of the reform). It asks how much, on average, households in each (income, age) bucket value the proposed policy change over their entire lifetime, represented as a one-time transfer at the time of the policy change. Dynamic distributional analysis is the standard in academic research, where conventional analysis is rarely used due to several key limitations that dynamic analysis addresses. It is also the natural welfare metric of the OLG lifecycle model with uninsurable risk, which is the paradigm model in economics for modeling the bidirectional equilibrium relationship between microeconomic and macroeconomic responses to a policy change.
Note: "Gross Income" refers to each household's income in the year of the policy change. We categorize households not yet in the labor force (ages 20 and younger) by their gross income in the year they enter the labor force. We calculate the gross income distribution across all groups of workers.
Source: Penn Wharton Budget Model
Table 6 presents policy "equivalent variations" for the cases and versions listed in Table 4. A positive equivalent variation indicates that a person would be better off under the policy reform, while a negative equivalent variation indicates that a person would be worse off. For instance, as shown in Table 6A, a high-skilled household aged 30 in the 20th to 40th percentile of income distribution loses about $4,900 of value from this policy, indicated by the negative value. In other words, this household is indifferent between adopting this policy bundle and losing a one-time payment of $4,900 without this policy change. Conversely, a household aged 30 in the top 20th percentile of income loses an equivalent of $10,400, meaning they would be indifferent between adopting this policy bundle and paying $10,400 to avoid it.
Table 6A provides the distributional analysis for the 4-year policy. Two major factors are the decrease in capital and the loss of unauthorized low-skilled labor. Panel A shows the impact on high-skilled workers. High-income retirees benefit from this policy due to increased returns to capital over the next decades from the decreased capital stock. The working-age and future generations face adverse effects due to increased debt and related capital crowding out, along with the loss of low-skilled workers who complement their productivity. Losses among the working-age households average about $2,200, while future generations average almost $5,000 in losses. Retirees gain, on average, $6,600.
Panel B of Table 6A reports results for authorized low-skilled workers. As the policy reduces the number of workers competing directly with the authorized low-skilled population, their wages increase over almost the first 20 years of the policy. Consequently, the working-age population generally benefits from this policy, with an average gain of $6,600. However, the unborn population is significantly worse off compared to the unborn high-skilled group, with an average loss of $11,200. This occurs because wages for the authorized low-skilled population decrease over time more than those of the high-skilled due to increased debt, lowering the capital stock. Retirees experience similar gains to their highly skilled counterparts, averaging about $6,200.
Table 6B provides the distributional analysis for the 10-year policy. Panel A shows the effects on high-skilled workers. With fewer unauthorized low-skilled workers and a lower capital stock, almost all working-age households are worse off, experiencing an average loss of $9,000—almost five times higher than under the 4-year policy. Increased debt crowds out more capital, making the unborn significantly worse off, with average losses of about $31,000, nearly six times more than under the 4-year policy. Retirees fare better under the 10-year policy, by gaining an average of $19,000.
Panel B of Table 6B reports results for authorized low-skilled workers. With significant wage gains following the removal of the entire unauthorized population from the country, almost all working-age and future generations benefit. Working-age individuals gain $31,600 on average, while future generations gain $58,000 on average. Retirees gain nearly $20,000 on average.
These projections use a Cobb-Douglas production function in which aggregate output is a function of the capital stock 
, total factor productivity (TFP) 
and a labor aggregate 
:
where labor combines high-skilled (subscript H) and low-skilled (subscript L) composites using a using a constant elasticity of substitution (CES) aggregator:
The elasticity 
between high-skilled and low-skilled labor is set equal to 2.0 following Ottaviano and Peri (2012). The parameter 
is calibrated numerically to match the relative wages between high-skilled labor and low-skilled labor. Low-skilled labor is further decomposed into authorized and unauthorized low-skilled labor using a CES aggregator:
where 
is the elasticity of substitution between low-skilled authorized (LA) and unauthorized (LU) workers. The weighting parameter 
is calibrated to match the relative wages between authorized and unauthorized low-skilled labor.
Our previous work modeled the relationship between TFP and the number of high-skilled labor STEM workers. For the current estimates, however, the TFP channel is minimal since the number of high-skilled labor STEM workers remains largely unchanged. Hence, TFP grows with exogenous technological change. The capital stock dynamics are also standard, as it evolves due to investment net of depreciation. But the labor aggregate represents a combination of the three types of labor we consider in the exercise: authorized high-skilled, authorized low-skilled and unauthorized low-skilled.
The parameter is key to our results:
Higher substitutability (larger
) means authorized and unauthorized low-skilled workers are highly interchangeable. Removing undocumented labor raises wages for the remaining low-skilled workers but reduces output more modestly because authorized workers can readily fill the gap.Lower substitutability (smaller
) implies stronger complementarity. Deportations then generate smaller wage gains or even wage losses. They impose a larger output loss, because unauthorized low-skilled labor cannot be easily replaced by authorized workers.
| 10-Year Deportation Policy | ||||||
|---|---|---|---|---|---|---|
| High-Skilled Wages (change from baseline) | ||||||
| 2034 | 2039 | 2044 | 2049 | 2054 | ||
| Elasticity of Substitution | ||||||
| Cobb-Douglas | 1.00 | -5.6 | -8.6 | -11.3 | -14.2 | -17.8 |
| alternative value for sensitivity reporting | 2.00 | -3.3 | -4.6 | -5.6 | -6.8 | -8.2 |
| alternative value for sensitivity reporting | 3.00 | -2.5 | -3.3 | -4.0 | -4.7 | -5.6 |
| Caiumi and Perri (2024) (some high school) | 6.76 | -1.8 | -2.3 | -2.7 | -3.1 | -3.5 |
| PWBM's estimate (some high school) | 12.35 | -1.6 | -2.0 | -2.3 | -2.6 | -3.0 |
| Main specification | 20.00 | -1.5 | -1.9 | -2.1 | -2.5 | -2.8 |
| PWBM's estimate (high school degree) | 83.30 | -1.4 | -1.7 | -1.9 | -2.2 | -2.5 |
| Perfect substitutes | 1,000,000 | -1.1 | -1.4 | -1.5 | -1.7 | -2.1 |
| Authorized Low-Skilled Wages (change from baseline) | ||||||
| 2034 | 2039 | 2044 | 2049 | 2054 | ||
| Elasticity of Substitution | ||||||
| Cobb-Douglas | 1.00 | -40.2 | -56.3 | -57.9 | -61.3 | -59.7 |
| alternative value for sensitivity reporting | 2.00 | -3.5 | -5.0 | -6.1 | -6.8 | -7.7 |
| alternative value for sensitivity reporting | 3.00 | 2.3 | 2.3 | 1.5 | 1.3 | 0.9 |
| Caiumi and Perri (2024) (some high school) | 6.76 | 4.7 | 4.8 | 4.2 | 4.3 | 4.3 |
| PWBM's estimate (some high school) | 12.35 | 5.0 | 5.0 | 4.5 | 4.6 | 4.7 |
| Main specification | 20.00 | 5.0 | 5.0 | 4.5 | 4.7 | 4.7 |
| PWBM's estimate (high school degree) | 83.30 | 5.0 | 5.0 | 4.5 | 4.7 | 4.9 |
| Perfect substitutes | 1,000,000 | 5.3 | 5.3 | 4.9 | 5.2 | 5.2 |
Source: Penn Wharton Budget Model
Table 7 presents the projected changes in wages for authorized high- and low-skilled workers under the 10-year deportation policy relative to the baseline, across a range of values for the elasticity of substitution between authorized low-skilled and unauthorized low-skilled labor. In addition to our own estimates—based on different education levels among low-skilled workers—we include the lowest estimate available in the literature, reported by Caiumi and Perri (2024), as well as a set of reference elasticity values ranging from the Cobb-Douglas case (elasticity near 1) to the case of perfect substitutes (elasticity approaching infinity). At elasticity values below 1, the model fails to converge under the policy, indicating that strong complementarity leads to labor market dislocations that the model cannot accommodate.
These comparisons highlight how both economy-wide losses, and the distributional consequences of deportation depend on how easily employers can substitute between authorized and unauthorized low-skilled workers.
Starting from an elasticity of substitution close to 1, we see that the high complementarity between authorized and unauthorized low-skilled implies large losses for all authorized workers across all skill levels. As the elasticity of substitution increases, authorized low-skilled workers can substitute more easily for unauthorized low-skilled and thus the impact on the wages of the complementary high-skilled workers tends to be more muted. In our main specification, wages of high-skilled workers go down by 2.8 percent in 2054 as compared to the baseline scenario. As a basis for comparison, Chassamboulli and Peri (2015), using a search model where authorized and unauthorized low-skilled workers are perfect substitutes, but differ across other dimensions, estimate that deporting 50 percent of Mexican undocumented workers reduces high-skilled wages by 0.35 percent in the long run. Scaling their estimate to a policy that removes 100 percent of undocumented workers regardless of origin implies a long-run effect on wages of –1.4 percent, which is about half the decline projected in our model.
The Cobb-Douglas case implies even larger effects on the wages of authorized low-skilled workers, the effect is about –60 percent by 2054. As the elasticity increases the effect becomes less negative and even turns positive as the elasticity passes a threshold somewhere between elasticity values 2 and 3. For the full range of empirically plausible elasticities—from Caiumi and Perri’s lowest estimate of 6.76 to our highest estimate of 83.3 based on low-skilled workers with a high school degree—the projected wage gains for authorized low-skilled workers fall between 4 and 5 percent. Under our main specification, the increase reaches 4.7 percent by 2054, which is more than twice the scaled-up long-run gain of 2.2 percent reported by Chassamboulli and Peri (2015).
Felipe Ruiz Mazin and Felix Reichling produced this analysis under the direction of Kent Smetters. Mariko Paulson prepared the brief for the website.
Media (only): For the fastest response, email us at inquiries-pwbm@wharton.upenn.edu.
All other responses: Please use our Contact Us.
Sign up for PWBM Breaking News, Alerts and Newsletter. Unsubscribe anytime.
-
The Census Bureau collects data on foreign-born residents but not their legal status. Estimates of the number of unauthorized immigrants range from 10.5 million to 13 million by other entities, but it could be as high as 15 million. To the extent the actual value exceeds 11.5 million, the budgetary costs of mass deportation will be larger than what we report. However, President Trump has set a goal to deport 1 million per year. As such, our analysis already deports at a slightly larger rate than his stated goal. ↩
-
For Social Security, everyone who works in covered employment—citizen, green-card holder, or visa holder—earns credits. Cash benefits later depend on having 40 credits and being lawfully present (or living in a country with an SSA agreement) when you claim. For Medicare, only green-card holders can qualify, and only after five continuous years of legal permanent residency status (plus work history or premium requirement). Temporary visa holders generally cannot. For Medicaid, green-card holders face a federal 5-year wait (with some exemptions); visa holders are limited to emergency coverage unless their state offers a special program. ↩
-
See Stephen Goss, Alice Wade, J. Patrick Skirvin, Michael Morris, K. Mark Bye, and Danielle Huston, 2013, “Effects of Unauthorized Immigration on the Actuarial Status of the Social Security Trust Funds”, Actuarial Note 151, available at https://www.ssa.gov/oact/NOTES/pdf_notes/note151.pdf. ↩
-
Under the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996, unauthorized immigrants are generally ineligible for most federal public benefits, including Medicaid (except for emergency services), Supplemental Nutrition Assistance Program (SNAP), Temporary Assistance for Needy Families (TANF), Supplemental Security Income (SSI), and Social Security benefits. ↩
-
American Immigration Council, "Mass Deportation: Devastating Costs to America, Its Budget and Economy," October 2924, available at https://www.americanimmigrationcouncil.org/sites/default/files/research/mass_deportation_report_2024.pdf. ↩
-
American Action Forum, "The Budgetary and Economic Costs of Addressing Unauthorized Immigration: Alternative Strategies," March 2015, available at https://www.americanactionforum.org/research/the-budgetary-and-economic-costs-of-addressing-unauthorized-immigration-alt/ ↩
-
We depart from the AIC's assumptions by not allowing for any self-deportation (they assume 20 percent of unauthorized immigrants would self-deport) and by considering the deportation of 10 percent of all undocumented immigrants per year. We also adjust the AAF figures, originally reported in 2013 dollars, to current-year values using the GDP deflator. ↩
-
Since January 2025, actual cost per deportee has been substantially exceeded this value. However, some of those costs are fixed expenses that can be used across a larger base of deportees. ↩
Age 0 to 20 20 to 40 40 to 60 60 to 80 80 to 100 -20 -1800 -4100 -6800 -7200 -6300 -15 -1800 -4200 -6700 -7200 -6400 -10 -1700 -3900 -6200 -6500 -6000 -5 -1600 -3700 -6000 -6400 -5600 0 -1500 -3500 -5900 -6200 -5300 5 -1400 -3600 -5700 -6000 -4100 10 -1400 -3500 -5600 -5900 -3700 15 -1300 -3500 -5400 -4900 -3400 20 -1100 -2900 -4500 -3700 -2900 25 -1500 -3400 -4900 -8900 -12600 30 -2100 -4900 -7800 -8600 -10400 35 -3500 -5500 -6900 -7900 -8300 40 -3000 -4800 -6100 -6500 -5700 45 -3200 -3800 -4900 -5200 -3000 50 -2200 -2900 -4400 -3600 0 55 -1600 -1900 -2200 -1600 3000 60 -1500 -900 -600 1600 8100 65 0 1600 3500 6700 46300 70 -1200 1500 3000 11000 32800 75 -1100 1000 1800 8800 27700 80 -600 900 1200 6200 20800 85 -300 500 600 4100 14500
Age 0 to 20 20 to 40 40 to 60 60 to 80 80 to 100 -20 -6800 -14000 -15300 -15800 -16500 -15 -6800 -13400 -14800 -15100 -16200 -10 -6200 -11800 -12900 -13100 -13400 -5 -5700 -10100 -10700 -10800 -10700 0 -4900 -8700 -8600 -8600 -8100 5 -4300 -6500 -6100 -5900 -4800 10 -3200 -3500 -2400 -2200 -100 15 -2600 -500 1700 2100 6400 20 -600 1900 4800 8400 9400 25 0 3200 3600 7000 12400 30 -200 2300 5500 9800 17100 35 2400 4000 7400 9100 14700 40 2600 5200 7600 9600 16500 45 1700 4600 6500 10100 17900 50 1700 4700 6700 10800 19600 55 2400 4500 6600 10500 18600 60 1000 2200 4000 6000 11300 65 -500 1200 2500 3700 7800 70 -1400 1500 2600 10300 30700 75 -1000 1100 1400 7500 25400 80 -400 600 1200 5600 19500 85 -200 400 900 3900 14200
Age 0 to 20 20 to 40 40 to 60 60 to 80 80 to 100 -20 -11200 -26700 -43600 -46500 -41900 -15 -10900 -26900 -42500 -46400 -41700 -10 -10600 -24400 -39600 -42400 -38000 -5 -9400 -22300 -36400 -39200 -34500 0 -8400 -19900 -33600 -35300 -30200 5 -7200 -19300 -30800 -32100 -21500 10 -6800 -17600 -27700 -29000 -17800 15 -5700 -14800 -23500 -21300 -14200 20 -4200 -11300 -17800 -14100 -10400 25 -5800 -12800 -17900 -23100 -40200 30 -8000 -17400 -24800 -28200 -33500 35 -10900 -18000 -22700 -25600 -24400 40 -9300 -15500 -19400 -20400 -15100 45 -10300 -13800 -19000 -20600 -9900 50 -7000 -8600 -10600 -9100 4500 55 -4900 -5700 -6500 -4300 12500 60 -3700 -3300 -3200 1700 23200 65 2100 6500 12600 23300 59300 70 -100 5600 11000 35600 104900 75 -400 3900 7100 26800 81500 80 -200 3100 3500 7100 36900 85 -100 1700 2500 11800 39800
Age 0 to 20 20 to 40 40 to 60 60 to 80 80 to 100 -20 18900 47300 66800 69500 97500 -15 18200 45600 64600 67300 94400 -10 18600 46400 65000 67600 93900 -5 18800 46200 64400 66900 92300 0 19100 46200 63600 66000 91000 5 19300 46400 63800 66400 91100 10 19400 46800 66100 68000 94800 15 17800 44800 62800 64200 87300 20 14000 36200 51600 65600 68800 25 18100 40100 41200 51800 72200 30 20700 33300 42700 55700 83800 35 30500 32700 41300 46200 66300 40 26800 30300 35700 41600 65100 45 16500 22800 25900 34700 58000 50 10200 15800 20600 30500 52500 55 5000 9500 14500 22500 40700 60 -700 1200 5000 8800 20200 65 -1100 5400 9600 13300 25700 70 -400 5600 10200 33400 96900 75 -400 4200 6100 24400 76200 80 -100 2600 4900 17400 56500 85 0 1500 3100 11200 38800
