Summary

The number of people entering the United States increased sharply starting in 2021 and peaked in 2023 before slowing in 2024. That surge in immigration imposed a fiscal burden on state and local governments.

In this report, the Congressional Budget Office examines the fiscal impact of that surge in 2023. State and local tax revenues grew, especially sales tax revenues, but the costs of providing services grew more. The largest increases in costs were for primary and secondary education, shelter and related services, and border security. Costs for incarceration and income security programs also increased.

The Surge in Immigration

Most immigrants in the surge were people whom CBO categorizes as other foreign nationals—immigrants who were not lawful permanent residents, were not eligible to apply for lawful permanent residency on the basis of their current status, and were not admitted on a temporary basis under the Immigration and Nationality Act (INA). Some of those immigrants received permission to enter or remain in the country (for example, through the use of parole authority), and some did not.1 Some had newly arrived, and some had been residing in the United States under a temporary status that expired.

All told, the surge in immigration that started in 2021 added an estimated 4.4 million people to the U.S. resident population in 2023. That increase represents a 1.3 percent gain over what the population would have been if the net immigration of other foreign nationals had remained at historical levels. (The net immigration of other foreign nationals is the number of new people in that category in a given year minus those who emigrated or obtained a different status. It does not reflect deaths of people in the category.)

The surge continued beyond 2023, but it slowed starting in June 2024, when an executive order suspended the entry of most noncitizens at the southern border, and has slowed further in 2025. CBO does not produce 10-year baseline projections for state and local budgets. Therefore, this analysis examines the effects of the surge on state and local budgets in 2023, the most recent year for which data from the national income and product accounts, which report government expenditures and receipts at the state and local level, are available.

CBO’s Analysis

This analysis focuses on the incremental impact of that immigration surge, not the broader effects of all immigrants who arrived between 2021 and 2023 or were already living in the United States. CBO used a similar approach to estimate the effects of the surge on the federal budget and the economy in a previous report.2 In that report, CBO projected that the increase in immigration would lead to larger increases in federal revenues than federal costs over the 2024–2034 period.

To isolate the effects of the surge in immigration, CBO compared estimates of state and local governments’ actual tax revenues and spending with estimates of the revenues and spending that would have occurred in the absence of the surge—that is, if the net immigration of other foreign nationals had totaled about 200,000 people per year from 2021 to 2023, in line with pre-2020 trends. In 2023, state and local governments’ tax revenues totaled $2.5 trillion, and their spending, net of federal grants-in-aid, totaled $2.8 trillion. For a detailed explanation of CBO’s estimation methods, see Appendix A.

Effects of the Surge in Immigration on State and Local Budgets in 2023

State and local governments saw both their tax revenues and their spending increase in 2023 as a result of the surge in immigration. In CBO’s estimation, the increase in spending was greater than the increase in taxes:

• Immigrants who were part of the surge that started in 2021 paid $10.1 billion in state and local taxes in 2023—primarily sales taxes, followed by income and property taxes.
• State and local governments spent $19.3 billion on goods and services for those immigrants. More than half of that spending was for elementary and secondary education, shelter and related services, and border security. Most of the remaining spending was for general services whose use could not be attributed to individuals and whose costs CBO expected to rise immediately in proportion to population growth.3
• As a result, the surge led to a direct net cost of $9.2 billion, amounting to 0.3 percent of state and local spending, net of federal grants-in-aid.

Changes in actual taxes and spending in 2023 differ from potential broader or longer-term effects and from costs that are borne without adding to spending. Therefore, CBO also calculated an alternative measure to estimate the potential fiscal pressures from the surge population.

In addition to the direct budgetary effects, that measure of potential effects reflects expected adjustments to state and local budgets that would take time to implement, such as property tax collections from new housing construction, and indirect increases in tax revenues from growth in the overall economy. It also captures non­budgetary costs—for example, more crowded classrooms in public schools and more crowded public transportation systems in places where spending did not keep pace with increases in the population. CBO’s measure of potential effects is largely consistent with measures used by other studies on the fiscal impact of immigration.4

The net cost of those potential effects is similar to the net cost of the direct effects of the surge in immigration. In CBO’s assessment, the potential effects of the immigration surge on state and local budgets in 2023 were as follows:

• The potential increase in state and local tax revenues due to the surge in immigration totaled $18.8 billion. In addition to the $10.1 billion in taxes paid directly by immigrants, that total includes $0.8 billion in revenues from potentially higher property taxes and $8.0 billion in tax revenues from greater economic activity due to the surge.
• The potential increase in state and local costs for goods and services totaled $28.6 billion. The difference between that amount and the direct effect is driven by the potentially higher costs of increased enrollment in public elementary and secondary schools and by increases in spending for additional categories of general services.5
• As a result, the potential net cost of the surge population to state and local governments was $9.8 billion in 2023.

The sizes of the direct and potential effects probably varied substantially among states, depending on the size of the surge population in the state and on state and local policies. This report focuses on the aggregate effects for the whole country.

This report’s estimates of the direct and potential effects of the surge population on state and local budgets are consistent with findings from other studies that used methods similar to CBO’s. Like this report, those studies found that immigration led to greater increases in costs than revenues for state and local governments. For more details about how the present report compares with those studies, see Appendix B.

The Surge Population in 2023

CBO estimates that by the end of 2023, the surge in immigration that started in 2021 had increased the net immigration of people in the other-foreign-national category by 4.3 million.6 Immigrants in that category are not lawful permanent residents, are not eligible to apply for that status, and have not been admitted as students, temporary workers, or other nonimmigrants under the terms of the INA.7 (For information about categories of immigrants in CBO’s analysis, see Box 1.) When births and deaths among people who were part of the immigration surge are accounted for, the net increase in the population amounted to 4.4 million.

Of the 4.3 million immigrants in the surge who were still present in 2023, about half arrived that year, about a third arrived in 2022, and the rest arrived in 2021. Most members of the surge population who arrived in 2023 were in the United States for only part of the year—six months, on average. Estimates of spending in this report are adjusted to account for those immigrants’ part-year residence in 2023. (Estimates of tax revenues reflect only the amounts immigrants paid while they resided in the country, so they are also lower for those who were part-year residents in 2023.)

CBO used estimates of the surge population’s demographic and economic characteristics to estimate certain categories of revenues and spending. Similar estimates of the population’s characteristics are discussed in detail in CBO’s earlier report on the federal budgetary effects of the immigration surge.8 For this report, CBO updated those estimates using data from the 2023 American Community Survey (ACS) and adjusted the sample of foreign-born noncitizens who had arrived since 2021 to reflect the surge population’s countries of origin.

In CBO’s assessment, members of the surge population were younger, on average, than members of the U.S. population as a whole and were more likely to be of working age. About a quarter were under age 18. Of those who were 18 or older, 56 percent had a high school diploma or less. Over half of the surge population resided in six states: Florida, Texas, California, New York, New Jersey, and Illinois. (By comparison, 40 percent of U.S. residents lived in those states in 2023.) The share of 18- to 54-year-olds in the surge population who were married was similar to the share in the general population.

How Immigration Affects State and Local Budgets

When immigration increases, state and local governments typically collect more in taxes but also face greater demand for services, which drives up spending. CBO estimates that in 2023, the increase in state and local governments’ spending due to the immigration surge was greater than the increase in tax revenues, resulting in a net cost.

The size of that net cost depends on the method used to estimate taxes and spending. Since state and local governments do not directly attribute revenues and spending amounts to individuals, the fiscal impact of new immigrants on budgets cannot be directly observed. Instead, analysts must estimate that impact. Considerable uncertainty surrounds the choice of estimation methods.

CBO used two measures to estimate net costs: direct effects, which reflect changes in tax revenues and spending that were directly attributable to members of the surge, and potential effects, which reflect broader economic impacts and potential fiscal pressures on state and local governments in addition to those direct effects. By either measure, the surge in immigration resulted in increases in both tax revenues and spending for state and local governments (see Figure 1).

CBO estimates that the increase in spending due to the surge was greater than the increase in taxes in 2023, resulting in a direct net cost of $9.2 billion. Accounting for broader effects of the immigration surge—the nonbudgetary costs for the rest of the population and revenues from increases in overall economic activity—CBO estimates that the potential net cost was $9.8 billion.

The direct net cost captures the direct effects of new immigrants on state and local government budgets. That measure reflects the fact that increases in the population may not immediately translate into proportional changes in tax revenues and spending: Tax policies and spending programs may take time to change, and policymakers may respond to the increased demand for goods and services by shifting resources from other areas or providing fewer goods and services to existing residents—allowing public school classrooms or public transportation systems to become more crowded, for instance.9 Therefore, the cost of providing goods and services to an additional person may be below the average cost for the population.

The potential net cost reflects fiscal pressures created by immigrants, whether or not those pressures resulted in immediate budget changes. That measure accounts for the fact that increases in the demand for services due to the presence of immigrants impose a cost on governments, even if total spending does not increase in proportion to the demand. For example, a government may respond to increased demand by reducing the services provided to existing residents and holding spending constant. CBO estimated the potential net cost of the immigration surge by assigning the average cost of services for the total population to immigrants who were part of the surge and allocating taxes to them in proportion to their share of the tax base.10

The measure of direct effects differs from the measure of potential effects in several ways:

• The direct effect on spending for public elementary and secondary education is based on the marginal cost of increased enrollment—that is, the observed change in average spending per student when enrollment changes. By contrast, the potential effect is estimated by applying average per-student spending for the general population to children in the surge population.
• Spending for some categories of general services is included in the potential effect but excluded from the direct effect because CBO expected that such spending would not respond to changes in the population immediately. (General services are government services, such as public transportation, for which the estimated costs are allocated equally to all members of the population.)
• CBO attributed a smaller amount of property taxes to the surge population in its measure of the direct effect than it did in its measure of the potential effect, on the view that increased demand for housing did not immediately translate into higher assessed property values and, therefore, did not immediately increase tax revenues.
• The potential effect includes increases in tax revenues caused by changes in overall economic activity due to the surge.

CBO’s analysis focuses only on 2023. The net cost of the immigration surge in later years might be different, for several reasons. For example, the net budgetary effects would depend on whether members of the surge and their children remained in the country or left, whether changes in the fiscal positions of state and local governments prompted changes to their tax policies and spending programs, and whether governments’ decisions to provide services for the surge population affected those immigrants’ future earnings or their use of government services over the longer term. The state and local taxes paid by members of the surge population who remained in the country would probably increase over time, but the costs of services for those immigrants could either rise or fall.

Effects of the Surge in Immigration on State and Local Tax Revenues

The amount of state and local tax revenues attributable to the immigration surge depends on several factors, including state and local tax laws, the income and spending behavior of members of the surge population, their demographic characteristics (such as their age, family structure, and state of residence), and their compliance with the tax system. The surge in immigration also affected revenues indirectly by influencing economic activity among the nonsurge population.

CBO estimates that in 2023, the direct increase in state and local revenues from the immigration surge was $10.1 billion (see Table 1). Of that amount, $7.1 billion came from general and selective sales taxes. (Selective sales taxes, also called excise taxes, are taxes on purchases of specific goods or services, such as motor fuels or tobacco products.) The direct effect includes only the property taxes attributable to rent payments made by members of the surge population; it does not reflect the full potential increase in taxes from such payments, which CBO expected to lag behind the increase in housing demand.

By comparison, CBO estimates that the immigration surge had the potential to increase state and local revenues in 2023 by $18.8 billion. That potential increase includes $10.9 billion in tax revenues attributable to the surge population—the $10.1 billion directly paid by those immigrants plus additional property taxes they would have paid if property tax assessments had fully adjusted. The potential effect also includes $8.0 billion in tax revenues that resulted from growth in economic activity. Those revenues came from several sources: corporate taxes on additional firm profits, individual income taxes on additional wage and business income, sales taxes on additional spending, and additional property taxes, including the portion of property taxes borne by owners of rental properties.

Taxation and Immigration

State and local governments receive most of their revenues from sales taxes, property taxes, and individual income taxes.11 In 2023, 34 percent of state and local tax revenues across the entire country came from general and selective sales taxes, 29 percent came from property taxes, and 22 percent came from individual income taxes.12

Because immigrants are generally subject to the same tax rules as citizens, most of the taxes paid by immigrants to state and local governments are also sales, property, and individual income taxes. Sales and income tax revenues responded directly to the surge population’s spending and earnings. Property tax revenues also responded directly to the increase in housing demand from the immigration surge. In CBO’s assessment, however, because it takes time for increases in property values to be reflected in property tax assessments, the increased demand did not fully translate into an increase in revenues from property taxes.

Immigrants’ Characteristics and Taxes. CBO’s estimates of immigrants’ state and local tax liabilities are based on several expected characteristics of members of the surge population, including their income, age, sex, marital status, and state of residence. According to CBO’s estimates, those immigrants earned less income, on average, than members of the general population. The surge population was also younger, and a larger proportion was male. Similar shares of adults were married in the surge population and the general population.

Because members of the surge population had relatively low earnings and had spent only a short time in the United States, CBO simplified its estimates by calculating tax liabilities as if all individuals and families in that population were renters rather than homeowners.

Immigrants’ Compliance With the Tax System. The amount of state and local individual income and sales taxes paid by immigrants depends on how many of those immigrants receive authorization to work, how those who are not authorized to work participate in the labor market, how many authorized and unauthorized workers file tax returns, and how they spend their earnings.

CBO expects that immigrants who are authorized to work are as likely to pay taxes on their income as other members of the population. Immigrants who work without authorization are less compliant with the income tax system, but they still pay some state and local taxes on their earnings. For example, some of those immigrants will choose to file taxes if their income has been reported to the tax authorities or if they expect that doing so will increase their chances of a favorable outcome at a future immigration hearing.

In CBO’s estimation, rates of compliance for income taxes were 15 percent lower among the surge population than among the total population because those immigrants were more likely to work without authorization. (A recent data-sharing agreement between tax and immigration authorities may make immigrants less likely to file individual income taxes, but that agreement was not in place in 2023.)13 CBO expects that compliance rates for sales taxes were the same for the surge population as for the total population.

CBO’s Estimation Method. To estimate the taxes paid by immigrants who were part of the surge, CBO drew from the method it used to estimate federal taxes paid by those immigrants in its previous report.14

As in that earlier analysis, CBO projected the income and taxes of the surge population by applying their expected demographic characteristics to a representative sample of taxpayers in the agency’s microsimulation tax model.15 CBO estimated each taxpayer’s spending on goods subject to sales taxes and on rental expenses that affect property taxes, then calculated how much of those taxes immigrants owed. Each taxpayer in the model was assigned a probability of being part of the immigration surge. Those probabilities were based on the agency’s demographic projections with and without the surge and reflected the likely demographic characteristics of members of the surge population. CBO then reweighted the probabilities so that the income distribution of the surge population matched the expected distribution of income among members of that population and across states.

Sales Taxes

CBO estimates that members of the surge population paid $7.1 billion in state and local sales taxes (general plus selective sales taxes) in 2023, or 7.3 percent of their adjusted gross income (AGI).16 Sales taxes made up the largest share of taxes paid by those immigrants. In CBO’s assessment, immigrants in the surge sent 15 percent of their earnings to family members in their home countries and therefore spent less of their income and paid less in sales taxes than they otherwise would have.

Across all 50 states and the District of Columbia, U.S. residents’ sales taxes averaged 5.3 percent of their AGI in 2023. The higher rate paid by immigrants who were part of the surge reflects their lower average income. People with less income tend to spend a larger share of it, and they also tend to spend a larger share of it on goods that are subject to general and selective sales taxes. As a result, they pay proportionally more in sales taxes.

Property Taxes

The surge in immigration boosted the demand for housing, which increased rental costs, property values, and assessed values (that is, estimates of property values that state and local governments use to levy property taxes). In CBO’s estimation, the revenues from taxes on rental properties that were directly attributable to the surge population totaled $0.7 billion, or 0.8 percent of those immigrants’ AGI. The potential increase in revenues amounted to $1.5 billion, or 1.6 percent of their AGI. The direct effect is smaller than the potential effect because, in CBO’s assessment, although immigrants’ demand for housing increased property values, assessed values were slower to increase.

The effect of the surge in immigration on property tax revenues is difficult to estimate for several reasons. First, the timing of increases in assessed values and property taxes is uncertain. Many new immigrants initially stay in transitional housing, such as shelters or the homes of relatives, which delays their effect on the housing market. As they move into rental housing, prices tend to rise, which increases assessed property values, but often with a lag. New housing construction eventually increases to meet the demand, further expanding the property tax base, but those effects may also take time to materialize.

Second, CBO estimates that virtually all immigrants in the surge were renters in 2023—but the proportion of property taxes paid by renters, as opposed to landlords, is difficult to determine, and experts disagree about what that proportion is.17 In this analysis, CBO assigned half of the burden of property taxes on rental properties to renters and the other half to property owners. The portion of property taxes borne by rental property owners is included in the potential effect on revenues from the increase in overall economic activity.

Finally, growth in economic activity due to the immigration surge also increased the property taxes paid by businesses. Those additional taxes were passed on to the broader population through higher prices or reduced profits. Therefore, like the share of rental property taxes borne by property owners, those taxes are included in the potential effect on revenues from the increase in overall economic activity.

Income Taxes

CBO estimates that in 2023, immigrants who were part of the surge population paid a total of $1.6 billion in state and local income taxes, or 1.6 percent of their AGI, on average. By comparison, state and local individual income taxes equaled 3.4 percent of the AGI of the general population in 2023.

The lower income tax rate paid by those immigrants reflects several factors, in CBO’s assessment. First, immigrants in the surge earned less income than other U.S. residents, especially during their first few years in the country. Second, on average, immigrants in the surge were less compliant with the U.S. tax system than other residents. And third, immigrants in the surge were more likely to reside in states without an income tax, such as Florida and Texas.

Other Taxes

The surge population is estimated to have paid $0.7 billion in other state and local taxes in 2023. Those taxes included workers’ compensation taxes, insurance trust taxes (mostly in California), and motor vehicle license taxes.

Increases in Tax Revenues From Overall Changes in the Economy

The surge in immigration also affected tax revenues through its effects on the economy, such as increases in productivity (and thus income) and interest rates. Those indirect effects raised revenues from corporate taxes on firms’ profits, individual income taxes on wage and business income among the nonsurge population, sales taxes on spending by the nonsurge population, and property taxes, including those borne by owners of rental properties and those paid by businesses but passed on to others through increases in prices or reductions in profits.

CBO includes those additional taxes in its estimate of potential effects because they were not directly paid by immigrants in the surge. In CBO’s projections, growth in the economy stemming from the surge increased state and local governments’ tax revenues by $8.0 billion.

Effects of the Surge in Immigration on State and Local Spending

State and local governments’ costs to provide goods and services to the surge population in 2023 depended on how many members of that population and their children received those goods and services and on the costs per person. CBO used survey data, administrative data, and state or local budget data to estimate the number of people who received each type of benefit and the total amount of benefits. State and local governments’ costs also included spending for border security in response to the immigration surge.

To measure the direct effect of the immigration surge on state and local spending, CBO estimated the marginal cost per person; to measure the potential effect, CBO used the average cost per person. For most categories of spending, the estimates of marginal and average costs were the same.

When possible, CBO used national data on noncitizens or recently arrived immigrants to allocate state and local governments’ spending to members of the surge population. For example, the Bureau of Justice Statistics (BJS) collects data on the citizenship status of inmates in state prisons. CBO used that information to allocate state and local governments’ spending for incarceration to members of the surge population in proportion to their share of the inmate population. (CBO estimates that in 2023, noncitizens made up 4 percent of inmates in state prisons and 7 percent of the U.S. population.) BJS does not have data on the citizenship status of people arrested by state police or prosecuted in state courts, so CBO allocated spending for police protection and courts equally to all residents in each state.

In CBO’s estimation, the surge in immigration directly increased state and local spending by $19.3 billion in 2023; the potential increase was much larger, at $28.6 billion (see Table 2). Two categories of spending account for most of that difference: spending for public primary and secondary education, and spending for general services whose use cannot be attributed to specific individuals. Those two categories also account for most of the direct and potential costs.

For each of those categories, CBO expected the marginal costs per person to differ from the average costs per person. CBO estimates that changes in state and local spending for public primary and secondary education were not proportional to state-level changes in enrollment. Therefore, the direct increase in spending was smaller than the potential cost projected using the average spending per student for the total population. The direct effect on spending for general services was also smaller than the potential cost because it excluded categories of services for which spending was unlikely to respond to changes in the population in the short run.

The estimated effects are nationwide estimates, but they reflect differences in costs per person and in policies across states. To capture that variation, CBO researched state and local polices that determined whether members of the surge population were eligible for financial support for postsecondary education, shelter services, and health insurance and income security programs as well as how the costs of border security and general services varied across states.

CBO’s estimates also reflect differences between the surge population and the general population in their use of goods and services provided by state and local governments. When possible, CBO relied on its prior estimates of the use of public benefits among the surge population. When such estimates were unavailable, CBO used public data, such as survey data, to estimate the use of goods and services by the surge population.

Education

All states have laws making elementary and secondary education compulsory, and children are entitled to public education regardless of their immigration status. Upon arrival, immigrant children may enroll in schools, and schools must assign those children to classrooms and provide equal access to educational services, including transportation. Those services are funded jointly by state and local governments. CBO estimates that in 2023, 550,000 children in public schools, or 1.1 percent of public school students, were part of the surge population.

By contrast, CBO estimates that few immigrants in the surge population received financial support for postsecondary education from state and local governments in 2023.18 Those governments provide direct financial support to students for postsecondary education through two channels—reduced tuition for public institutions for in-state residents and financial aid—but few members of the surge population would have been eligible for either form of support.

Public Primary and Secondary Education. In CBO’s estimation, the surge in immigration directly increased spending for public primary and secondary education by $5.7 billion, or 0.7 percent, in 2023. The potential effect on costs was much larger, amounting to an increase of $9.4 billion, or 1.2 percent. Both of those estimates include higher costs due to lower English proficiency among the surge population: Because recent immigrants are often English-language learners, they tend to need additional instructional and support services. CBO estimates that those services cost state and local governments $1.2 billion in 2023.19

State and local governments adjust the amounts of funding they provide to schools when enrollment increases, but the direct effect is smaller than the potential effect because in the short run, that adjustment is not proportional to the increase. (That difference can arise for many reasons, including state and local budget processes.) As a result, the marginal cost of an additional student is less than the average cost for all students. To calculate both the direct effect and the potential effect, CBO multiplied a measure of the change in spending per student by the change in public school enrollment due to the surge.

The direct effect reflects the statistical relationship between state-level changes in fall enrollment and state-level changes in state and local governments’ spending per student, net of federal grants-in-aid. CBO found that spending per student declined when enrollment increased and increased when enrollment declined. Some school districts, for example, did not hire new teachers in proportion to enrollment growth, so class sizes grew and teachers’ workloads increased. The immigration surge thus might have affected public education in ways that were not reflected in state and local budgets.

Growth in enrollment was measured against CBO’s estimate of what enrollment would have been without the surge in immigration. Given research indicating that one student transferred from public to private school for every five to seven immigrants who entered public school, CBO estimated that the net increase in public school enrollment was five-sixths of the total number of immigrant children who enrolled.20

The potential effect is intended to reflect nonbudgetary costs, such as more crowded classrooms and increased workloads for teachers and other school staff, that lead some parents to move their children to private schools. For that reason, CBO’s estimate of the potential cost is not adjusted to account for any change in spending per student in response to changes in enrollment or for any shifting between public and private schools.

Postsecondary Education. CBO estimates that state and local governments spent a negligible amount on public postsecondary education for the surge population in 2023, because almost no members of that population were eligible for in-state tuition benefits or financial aid from state governments. To simplify its analysis, CBO allocated no state or local spending for postsecondary education to the surge population.

Many states provide a pathway for undocumented immigrants to qualify for in-state tuition if they attend high school in the state for two or three years and receive a high school diploma or equivalent. Few members of the immigration surge had been in the country long enough to use that pathway in 2023. Some states provide a pathway that requires establishing one year of residency and meeting specific eligibility criteria that are consistent with the requirements for federal financial aid. CBO estimates that few members of the immigration surge could have met those requirements.

Shelter and Related Services

Using publicly available information from state and local governments, CBO estimated that state and local governments in four states—New York, Massachusetts, Illinois, and Colorado—spent a total of $3.3 billion to provide shelter and related services, including food and legal support, to the surge population in 2023. Although those services were not restricted to members of the surge population, CBO attributed their costs to the surge population because it is likely that many people who received such services were part of the surge.

CBO could not find public data on such costs in other states—but in CBO’s assessment, spending in those four states accounted for most of the state and local spending on shelter and related services for the surge population. Estimates from a recent study suggest that New York City, Massachusetts (primarily suburban Boston), Chicago, and metropolitan Denver together accounted for about 90 percent of the national increase in asylum seekers in shelters from 2022 to 2024.21

New York City and Massachusetts have right-to-shelter laws that require the states to provide shelter or housing assistance to certain individuals or families.22 CBO estimates that in 2023, New York City spent $2.6 billion to provide shelter and related services to asylum seekers.23 Massachusetts spent $0.5 billion on emergency shelter services in 2023, up from $0.2 billion in fiscal year 2019; CBO attributed that $0.3 billion increase in spending to the surge population.24

Illinois and Colorado do not have right-to-shelter laws, but they also used state and local funding to provide shelter and related services to members of the surge population. State and local governments in Illinois spent $0.4 billion in 2023 on an emergency program that provided support to asylum seekers; $0.3 billion of that amount was spent by the City of Chicago.25 Spending for shelter and related services in Colorado was less than $50 million.

Shelter and related services that were not explicitly designated for the surge population are included in spending for general services and are allocated to those immigrants in proportion to their share of the resident population.

Border Security

CBO estimates that in 2023, four states—Texas, Arizona, Florida, and California—spent a total of $2.7 billion on border security in response to the immigration surge. That amount includes spending for law enforcement efforts to prevent illegal immigration, drug smuggling, and human trafficking; the installation of physical barriers; and the interstate transportation of immigrants.

CBO attributes $2.5 billion of Texas’s border security spending in 2023 to the surge in immigration. That estimate reflects the difference between the state’s average annual appropriations for border security before the surge ($0.4 billion) and its total spending for border security in 2023 ($2.9 billion).26

Arizona spent $174 million on border security in 2023, CBO estimates. In 2021, Arizona established a border security fund, which was intended to bolster border security efforts within the state. Arizona spent about $300 million from that fund in fiscal year 2023 and less than $50 million in fiscal year 2024.27

CBO estimates that Florida and California spent very small amounts for border security in 2023 in response to the surge. Florida spent $15 million to support the state National Guard’s border security efforts during the first half of the year.28 In addition, CBO estimates that Florida spent $6 million in 2023 to transport immigrants to other states.29 California spent $15 million that year to expand the state military department’s drug interdiction efforts.30

Health Insurance and Income Security Programs

Some federal health insurance and income security programs provide benefits to immigrants that are partially funded by states, and some states provide similar benefits to immigrants through their own programs, which are sometimes called replacement programs. CBO identified five federal programs that states used to partially fund benefits to the surge population or that they supplemented with similar benefits: Medicaid, the Children’s Health Insurance Program (CHIP), Temporary Assistance for Needy Families (TANF), the Supplemental Nutrition Assistance Program (SNAP), and Supplemental Security Income (SSI). In 2023, state and local spending on those programs totaled about $300 billion. In addition, CBO estimated workers’ compensation benefits received by the surge population.

To simplify its analysis, CBO did not consider state and local funding for benefits through other federal health insurance or income security programs.31 In addition, because the Current Population Survey (CPS) does not identify recipients of certain benefits, and because synthesizing eligibility rules and benefit amounts across states is complicated, this analysis does not separately identify all forms of direct assistance from state and local governments.32 Spending on direct assistance that was not attributable to individuals is included in spending for general services and distributed equally across the population.

CBO estimates that in 2023, states spent a total of about $1 billion on Medicaid, CHIP, and TANF benefits and similar state-funded benefits for members of the surge population. About half of that spending, or $0.5 billion, was for Medicaid, CHIP, or state-funded programs offering Medicaid-like benefits; $0.4 billion was for TANF. States spent a very small amount on state-funded replacement programs that offered benefits similar to SNAP and SSI benefits and an even smaller amount on workers’ compensation for the surge population.

Medicaid and CHIP. Certain groups in the surge population were eligible to receive Medicaid and CHIP coverage. That coverage is typically limited to qualified aliens after a five-year waiting period, but states can choose to provide Medicaid to—and receive the federal share of Medicaid funding for—children and pregnant individuals who would qualify if they were not still within that waiting period. (Medicaid and CHIP are considered together in this analysis because the two programs offer similar benefits.) Limited Medicaid coverage for emergency medical services is also available for people who would qualify for full Medicaid benefits if not for their immigration status.33 CBO estimates that states’ share of spending for those federal benefits for the surge population totaled $0.3 billion in 2023.

Additionally, some low-income immigrants who do not qualify for Medicaid or CHIP because of their immigration status may receive similar benefits through state-funded programs.34 States often restrict eligibility for those programs to certain age groups, and in rare cases, they limit the coverage provided to certain demographic groups.35 To simplify its analysis, CBO treated all state-funded programs as though they provided the same type of benefits as the federal Medicaid program.36 CBO estimates that about one-tenth of the surge population lived in states with such programs and met the age-based eligibility requirements. The cost of those state-funded programs for the surge population totaled $0.2 billion in 2023, CBO estimates.

Other Programs. Like Medicaid, TANF provides benefits that are jointly funded by federal and state governments. To avoid penalties, such as reduced federal grants, states must use some of their own funds to fulfill that program’s maintenance-of-effort (MOE) requirements. MOE dollars generally must achieve one of TANF’s objectives: assisting needy families so that children can be cared for in their own homes or the homes of relatives; encouraging job preparation, marriage, and work; preventing out-of-wedlock pregnancy; or promoting the formation of two-parent families.37 CBO attributed $0.4 billion of states’ MOE spending in 2023 to the surge population.

CBO also estimated how much it cost state and local governments to provide benefits similar to SNAP and SSI benefits, as well as state supplements for immigrants who were eligible for federal SSI benefits, to the surge population.38 Only a few state-funded programs provide SNAP- or SSI-like benefits to certain groups of immigrants who are ineligible for federal benefits because of their immigration status.39 In 2023, few members of the surge population were eligible for federal SSI benefits and, therefore, state supplements. As a result, CBO estimates that state and local governments spent less than $50 million on SNAP- and SSI-like benefits for the surge population.

State and local governments spent a very small amount on workers’ compensation for the surge population. Data from the national income and product accounts (NIPAs) indicate that state and local spending on workers’ compensation totaled $8.5 billion in 2023. CBO used self-reported benefits from a sample of CPS respondents to estimate the share of benefits that went to members of the surge population. In CBO’s estimation, immigrants in the surge population received less than one-half of one percent of the total state and local spending on workers’ compensation.

Incarceration

In CBO’s estimation, state and local governments’ spending for the incarceration of members of the surge population in 2023 totaled $0.77 billion. That value is based on the estimated number of people in the surge population in state prisons and local jails for violations of state or local laws and the estimated costs per inmate in 2022 and 2023; it excludes those in federal detention or federal prisons for violations of immigration laws or for other reasons. (Local jails typically hold individuals who are awaiting trial or have short sentences—usually less than a year; they sometimes temporarily hold inmates from state prisons where space is limited. State prisons, by contrast, house individuals with longer sentences, often for felonies.)

CBO estimates that about 13,500 members of the surge population were in a state prison or local jail at the end of 2023 for violations of state and local laws.40 That estimate is uncertain, however. Moreover, the share of the surge population that is incarcerated could change as members of the group spend additional years in the United States, because their individual circumstances may change and because the time required to prosecute criminal cases may delay changes in incarceration rates.

To arrive at that estimate, CBO first estimated the total number of surge immigrants who were incarcerated in 2023. Using data from the 2023 ACS and data from the Bureau of Justice Statistics, CBO estimated that about 18,000 surge immigrants were incarcerated at the end of that year.41 Other BJS data indicated that about two-thirds of that group, or roughly 11,500 immigrants, were in prison rather than jail, and about two-thirds of those in prison, or approximately 7,500 immigrants, were in a state prison.42 Because comparable data about inmates in jail were not available, and because there are very few federal jails, CBO estimated that all of the roughly 6,000 remaining surge immigrants in jail were in a local, rather than federal, jail.

CBO next estimated the cost of those immigrants’ incarceration. In 2023, according to NIPA data, state and local governments spent a total of $107 billion on incarceration (net of federal grants provided for the incarceration of prisoners regardless of their immigration status), or about $60,000 per inmate. To estimate that average, CBO divided the total amount by the number of inmates in state prisons and local jails at the end of 2023. Multiplying that average by the number of incarcerated immigrants who were part of the surge yields an estimated total cost to state and local governments of $0.83 billion.

CBO subtracted about $60 million from that total to account for federal grants provided by the Department of Justice’s State Criminal Alien Assistance Program. That program provides federal aid to states for the incarceration of undocumented immigrants who have been convicted of misdemeanors or felonies. The program awarded about $200 million to states in 2023.43 CBO estimated that about one-third of the undocumented immigrants in the country in 2023 were part of the surge and therefore attributed that share of the awarded amount to incarcerated members of the surge population.

Research has consistently found that immigrants are less likely than native-born citizens to be arrested, convicted, or incarcerated.44 Studies using indirect methods to associate crime rates over time in a city or state with the presence of undocumented immigrants have also concluded that the presence of those immigrants was unconnected to crime rates.45

General Services Allocated Equally to All

Many general services provided by state and local governments (for example, police protection) cannot be attributed to individual users. Other services provide direct benefits to people. The costs for some of those services—education and worker’s compensation—are described above. For others, the available data do not allow benefits to be linked to individual users. CBO allocated the costs for those general services in each state equally across the population of that state. In other words, for each state, the share of costs allocated to the surge population was consistent with that population’s share of the total population. In CBO’s estimation, the direct effect of the surge in immigration on state and local spending for those services was $6.1 billion, and the potential effect was $11.6 billion.

CBO expected that spending for some categories of general services would respond proportionally to changes in the population. To measure those direct effects, CBO allocated the average cost of the following services for the total population to members of the surge population:

• Public safety services, including police and fire protection and courts;
• Housing and community services, including sanitation;
• Health care; and
• Income security.46

Spending for other categories of general services was not expected to respond to population changes in the short run, although increases in the population would put pressure on those services—for example, in the form of more crowded public transportation. To measure those potential effects, CBO allocated the costs of the following services for the total population to members of the surge population in proportion to their share of the total population:

• General government services, including executive and legislative functions, tax collection, and financial management;
• Economic affairs, including transportation, agriculture, energy, and natural resources; and
• Recreation, culture, and libraries.

Spending was allocated equally to all residents in a state. To account for differences among states in spending on general services in 2023, CBO used data from the Census of Governments and the NIPAs to adjust its allocations. States’ total per capita spending on general services ranged from 60 percent to 500 percent of the U.S. average. Among California, Florida, New York, and Texas, it ranged from 70 percent to 150 percent.

Sources of Uncertainty in CBO’s Estimates

CBO’s estimates of the effects of the immigration surge on state and local budgets are highly uncertain. That uncertainty stems from difficulties in determining the number of immigrants in the surge and their immigration status, the methods that were used to estimate those immigrants’ characteristics and geographic locations, the methods used to estimate effects on state and local governments, and expectations about the behaviors of those immigrants and the responses of state and local governments.

Number of Other Foreign Nationals in the Immigration Surge

Because no single dataset identifies immigrants by when they arrived and provides details about their immigration status, CBO relied on various administrative data sources to estimate the net immigration of other foreign nationals—including the number of people who became other foreign nationals and the number who left that group because their immigration status changed or because they emigrated or died. To estimate the total number of immigrants residing in the country in a given year, CBO extrapolated mortality and emigration rates from historical data. As a result, CBO’s estimate of the number of other foreign nationals in the immigration surge is uncertain. The status of those immigrants is also uncertain because it depends on administrative policies, practices, and resources, as well as on decisions made by the immigrants themselves.

The number of people in the immigration surge directly affects estimates of state and local tax revenues and spending. The statuses of those immigrants affect estimates of the taxes they paid, because immigration status is correlated with income. Immigrants’ statuses also affect estimates of state and local spending because residents’ immigration status determines their eligibility for government assistance.

Characteristics of the Surge Population

The datasets that CBO used to analyze the effects of the immigration surge did not identify individuals in the surge population. Such individuals had to be identified using statistical methods, which are subject to error. Individuals in the data were classified as members of the surge population on the basis of demographic and economic characteristics that CBO expected that group to have. Those characteristics included year of arrival, age, education, family composition, and earnings. Any errors in the characteristics of the surge population in the tax data would lead to mismeasurement of the effects of the surge on tax revenues. Similar errors in the survey data would lead to mismeasurement of the effects on spending for services ranging from education to incarceration.

Geographic Distribution of the Surge Population

CBO’s estimates are sensitive to the distribution of the surge population across states because there is significant variation in state and local tax and spending policies. However, the agency’s estimates of the geographic distribution of those immigrants are highly uncertain. In addition, states’ budget processes are complex and varied, and information about how states administer different benefit programs for immigrants is limited.

Revenue sources and spending could also vary between rural and urban areas or across substate administrative areas (such as school districts). Because substate geographic information is typically suppressed in tax and survey data to protect individuals’ privacy, CBO cannot estimate how many members of the surge population resided in those areas.

In CBO’s estimates, the distribution of the surge population across metropolitan and rural areas within states matches that of the broader population. A different distribution could lead to significant variation in the costs of the surge population across localities within states. For example, the direct costs of public primary and secondary education would be lower in districts with declining enrollment, because immigrants could fill empty spots without the districts’ having to immediately increase resources. The opposite is true for districts with rising enrollment. For example, in New York, enrollment was declining before the immigration surge, but in some parts of New York City where immigrants who were part of the surge may have later located, enrollment was rising. If the surge population was concentrated in such areas, the direct costs would be higher than CBO’s estimates based on state-level data.

Estimation Methods

Because state and local governments do not directly attribute revenues and spending amounts to individuals, the fiscal impact of new immigrants cannot be directly observed. Instead, analysts must estimate that impact by allocating state and local tax revenues and spending across the population. Considerable uncertainty surrounds the choice of allocation methods, and different methods produce different results, which may affect conclusions about the fiscal impact of new immigrants.

For example, when analyzing the direct effect of the immigration surge in 2023, CBO expected that increases in the population that began in 2021 would affect some categories of general services (such as police and fire protection and courts) but not others (such as transportation and energy). Incorrectly including a category in the first group could lead to an overestimate of the direct change in spending; incorrectly including one in the second group could lead to an underestimate.

Similarly, in its revenue estimates, CBO allocated half of property taxes to renters and half to property owners. That approach affected the shares of property tax revenues that CBO attributed to the immigration surge and to broad economic growth. The property tax burdens on owner-occupied and rental properties are similar, but for rental properties, the burden is shared between renters and owners. CBO included the share paid by owners of rental properties in the potential effect on revenues from the increase in overall economic activity.

Immigrants’ Behaviors and Governments’ Responses

Estimates of the effects of the immigration surge on state and local budgets further depend on the expected behaviors of the surge population. In CBO’s revenue estimates, those expected behaviors include the extent to which immigrants without work authorization complied with the tax system, the amount of money they sent back to family members in their home country, and how their consumption varied across the income distribution. In the agency’s spending estimates, expected behaviors include decisions to enroll children in school and use government services.

There is also uncertainty about how quickly increases in a locality’s population lead to increases in state and local governments’ spending and property tax revenues. CBO concluded that by 2023, spending on primary and secondary education would not have adjusted fully to the additional immigration observed since 2021. When estimating revenues from property taxes, CBO projected that members of the surge population would start paying the average amounts paid by the general population one year after their arrival.

---

1. 1. Under section 212(d)(5) of the Immigration and Nationality Act, 8 U.S.C. § 1182(d)(5), the Secretary of Homeland Security may parole people into the United States, which allows them to enter temporarily without formal admission to the country only on a case-by-case basis for urgent humanitarian reasons or significant public benefit and under conditions the Secretary prescribes.

2. 2. Congressional Budget Office, Effects of the Immigration Surge on the Federal Budget and the Economy (July 2024), www.cbo.gov/publication/60165.

3. 3. Those services are police and fire protection, courts, health care, housing and community services, and income security programs not included in other categories.

4. 4. See National Academies of Sciences, Engineering, and Medicine, The Economic and Fiscal Consequences of Immigration (National Academies Press, 2017), Chapter 9, pp. 495–566, http://tinyurl.com/3k3ycafj.

5. 5. The categories included in the potential effect, but not the direct effect, are the executive and legislative functions of government, tax collection and financial management, transportation, other economic affairs, recreation and culture, and libraries.

6. 6. CBO’s estimates of the size of the immigration surge are derived from its projections of the Social Security area population. For more details, see Congressional Budget Office, The Demographic Outlook: 2024 to 2054 (January 2024), pp. 5–6, www.cbo.gov/publication/59697.

7. 7. Over the same period, net immigration totaled 2.1 million for lawful permanent residents and immigrants eligible for that status and 0.2 million for residents classified as nonimmigrants under the INA.

8. 8. Congressional Budget Office, Effects of the Immigration Surge on the Federal Budget and the Economy (July 2024), www.cbo.gov/publication/60165.

9. 9. Every state in the country except Vermont has some stipulations to balance its operating budget, but the rules and levels of enforcement vary. See Tax Policy Center, “What Are State Balanced Budget Requirements and How Do They Work?” (January 2024), https://tinyurl.com/yf3j7rxp. In the case of spending for public education, instead of accounting for those rules explicitly, CBO used historical data to capture how long it typically takes for state and local governments’ spending to change in response to changes in school enrollment.

10. 10. That approach is commonly used in research on the fiscal effects of immigration. See, for example, National Academies of Sciences, Engineering, and Medicine, The Economic and Fiscal Consequences of Immigration (2017), https://doi.org/10.17226/23550.

11. 11. Five states have no general sales tax, and nine have no broad-based income tax. See Janelle Fritts, Facts and Figures 2022: How Does Your State Compare? (Tax Foundation, March 2022), https://tinyurl.com/yc4fc5jy.

12. 12. Bureau of Economic Analysis, “Table 3.3. State and Local Government Current Receipts and Expenditures” (accessed April 10, 2025), https://tinyurl.com/22hucdtm.

13. 13. See Internal Revenue Service, Memorandum of Understanding Between the U.S. Department of the Treasury, Internal Revenue Service, and the U.S. Department of Homeland Security, U.S. Immigration and Customs and Enforcement, for the Exchange of Nontax Criminal Enforcement (April 7, 2025), https://tinyurl.com/mw9c5c8x.

14. 14. Congressional Budget Office, Effects of the Immigration Surge on the Federal Budget and the Economy (July 2024), www.cbo.gov/publication/60165. For the analysis in the present report, CBO used publicly available tax data rather than tax return data because estimating state and local (as opposed to federal) revenues is outside of CBO’s core responsibilities, and the agency is not authorized to use tax return data for that purpose.

15. 15. For information about that model, see Congressional Budget Office, “An Overview of CBO’s Microsimulation Tax Model” (June 2018), www.cbo.gov/publication/54096.

16. 16. AGI consists of income from all sources not specifically excluded by the tax code, minus certain deductions (for example, of a portion of the self-employment tax and certain contributions to retirement accounts).

17. 17. See, for example, Sarah Baker, “Property Tax Pass-Through to Renters: A Quasi-Experimental Approach” (draft, University of California, Berkeley, October 2024), http://tinyurl.com/mr4ca8ax; Carl Davis, Marco Guzman, and Emma Sifre, Tax Payments by Undocumented Immigrants (Institute on Taxation and Economic Policy, July 2024), https://tinyurl.com/2zfja4zb; and David J. Schwegman and John Yinger, The Shifting of the Property Tax on Urban Renters: Evidence From New York State’s Homestead Tax Option (Census Bureau, Center for Economic Studies, December, 2020), https://tinyurl.com/372sjydr.

18. 18. In this analysis, postsecondary institutions are universities, colleges, and trade and vocational schools that people attend after graduating high school or passing the General Educational Development (GED) test.

19. 19. CBO estimated those costs using formulas compiled by the Education Commission of the States. See Education Commission of the States, “K-12 Funding 2024, English Language Learning Funding” (March 2024), http://tinyurl.com/46urxpcf.

20. 20. See Thomas J. Murray, “Public or Private? The Influence of Immigration on Native Schooling Choices in the United States,” Economics of Education Review, vol. 53 (August 2016), pp. 268–283, https://doi.org/10.1016/j.econedurev.2016.04.003.

21. 21. Bruce D. Meyer, Angela Wyse, and Douglas Williams, Asylum Seekers and the Rise in Homelessness, Working Paper 33655 (National Bureau of Economic Research, April 2025), www.nber.org/papers/w33655.

22. 22. New York City is required by a court-ordered agreement to provide shelter and meals to homeless individuals and families; see Callahan v. Hugh L. Carey, 909 N.E.2d 1229 (N.Y. 2009). In Massachusetts, a state law calls for emergency housing aid for low-income pregnant women and families if funding is approved in the state budget; see Mass. Gen. Laws. ch. 23B, § 30 (2023).

23. 23. The estimate of New York City’s shelter costs in 2023 is the average for fiscal year 2023 (which ended in June 2023) and fiscal year 2024 (which started in July 2023). See Office of the New York City Comptroller, “Accounting for Asylum Seeker Services: Fiscal Impacts” (accessed February 18, 2025), http://tinyurl.com/4hnsnat3.

24. 24. Public data indicate that Massachusetts spent $1 billion on shelter costs for its emergency shelter program by the end of 2024; CBO attributed half of that amount to spending for the surge population in 2023. See Shia Kapos and Kelly Garrity, “Democratic-Led States Still Grapple With Housing Migrants,” Politico (January 2, 2025), https://tinyurl.com/mry5fjbj. For details about the state’s spending on emergency shelter services in fiscal year 2019, see Commonwealth of Massachusetts, “Emergency Assistance Family Shelters and Services” (accessed May 13, 2025), https://tinyurl.com/2pt4dd4c. For more information about the state’s spending on shelter for asylum seekers, see Special Commission on Emergency Housing Assistance Programs, Report of the Special Commission on Emergency Housing Assistance Programs (November 19, 2024), https://tinyurl.com/yu3zsh4x.

25. 25. Felix Winnekens, Jane H. Ridley, and Alyssa B. Farrell, “Migrants and Asylum Seekers Pose Budgetary Challenges in New York City, Chicago, and Denver,” S&P Global (February 13, 2024), http://tinyurl.com/2p96kmaw. For details about the funding for Illinois’s program that provides shelter and related services to asylum seekers, see Illinois State Comptroller, “Welcoming With Dignity FY23 and FY24 Allocated Funding Breakdown” (accessed May 12, 2025), https://tinyurl.com/mr3sss24.

26. 26. Texas started Operation Lone Star in 2021 in response to the surge in immigration. To estimate the state’s border security spending in 2023, CBO used documents provided by the Texas Legislative Budget Board with department-level border security expenditures from the second quarter of fiscal year 2023 (December 2022 to February 2023) through the second quarter of fiscal year 2024 (December 2023 to February 2024) and assigned the following amounts to calendar year 2023: two-thirds of the amount expended in the second quarter of fiscal year 2023, the total amount spent in the third and fourth quarters of fiscal year 2023 (March 2023 to August 2023), and two-thirds of the amount spent in the first and second quarters of fiscal year 2024 (September 2023 to February 2024). In the two-year budget cycle that ended in August 2019, Texas appropriated $0.8 billion—or about $0.4 billion per year—for border security. See Texas Legislative Budget Board, Fiscal Size-Up: 2018–19 Biennium (September 2018), https://tinyurl.com/3t45kkv8.

27. 27. Arizona’s fiscal year 2023 started in July 2022 and ended in June 2023. For details about spending from Arizona’s border security fund in fiscal year 2023, see State of Arizona, Annual Financial Report for the Year Ended June 30, 2023 (December 1, 2023), p. 162, https://tinyurl.com/4kyphum5. For details about spending in fiscal year 2024, see State of Arizona, Annual Financial Report for the Year Ended June 30, 2024 (November 22, 2024), p. 169, https://tinyurl.com/hyhw3he7.

28. 28. Ana Ceballos, “DeSantis Spent Over $15 Million in Six Months on ‘Illegal Migration’ in Texas, Keys,” Miami Herald (July 14, 2023), https://tinyurl.com/4rc3jr5a.

29. 29. Florida appropriated $12 million to its Unauthorized Alien Transport Program for fiscal year 2023, which ended on June 30, 2023. CBO assigned half of that amount to calendar year 2023. See S.B. 1718, 2023 Leg., Reg. Sess. (Fla. 2023).

30. 30. The state appropriated a total of $30 million for two years. See State of California, California State Budget: 2022–23 (June 27, 2022), https://tinyurl.com/mpufn4uu. California does not provide funding for immigration enforcement. S.B. 54, 2017 Leg., Reg. Sess. (Cal. 2017).

31. 31. For example, Colorado and Washington offer state-funded subsidies through the health insurance marketplaces established by the Affordable Care Act to some immigrants who are not eligible for the federal benefits. See Claire Heyison and Shelby Gonzales, States Are Providing Affordable Health Coverage to People Barred From Certain Health Programs Due to Immigration Status (Center on Budget and Policy Priorities, February 1, 2024), https://tinyurl.com/yxbjxk45.

32. 32. For example, some states offer support to immigrants who are ineligible for federal benefits through general assistance or similar programs. See National Immigration Law Center, “Table 9: State-Funded SSI Replacement Programs” (September 2002), http://tinyurl.com/5msxkd8u.

33. 33. For more details about Medicaid coverage for emergency medical services, see Congressional Budget Office, letter to the Honorable Jodey Arrington regarding emergency Medicaid services for certain non-U.S. nationals (October 2, 2024), www.cbo.gov/publication/60805. States can choose to exempt children, pregnant individuals, and postpartum individuals from the five-year waiting period for Medicaid and CHIP.

34. 34. As of November 2023, 11 states and the District of Columbia provide Medicaid-like benefits to certain immigrants who are ineligible for federal Medicaid benefits. For information about those programs, see Akash Pillai, Drishti Pillai, and Samantha Artiga, State Health Coverage for Immigrants and Implications for Health Coverage and Care (KFF, May 1, 2024, updated April 10, 2025), http://tinyurl.com/yc454abk; and Claire Heyison and Shelby Gonzales, States Are Providing Affordable Health Coverage to People Barred From Certain Health Programs Due to Immigration Status (Center on Budget and Policy Priorities, February 1, 2024), https://tinyurl.com/yxbjxk45

35. 35. For example, Illinois provides comprehensive Medicaid-like benefits to residents under age 19 but does not cover long-term care for adults age 42 or older.

36. 36. This analysis excludes state-funded benefits to more limited populations. For example, Hawaii provides Medicaid-like benefits to qualified immigrants who are age 65 or older, blind, or disabled and in the five-year waiting period for Medicaid. See Claire Heyison and Shelby Gonzales, States Are Providing Affordable Health Coverage to People Barred From Certain Health Programs Due to Immigration Status (Center on Budget and Policy Priorities, February 1, 2024), p. 13, footnote 45, https://tinyurl.com/yxbjxk45.

37. 37. Gene Falk, The Temporary Assistance for Needy Families (TANF) Block Grant: A Primer on TANF Financing and Federal Requirements, Report RL32748, version 45 (Congressional Research Service, August 14, 2023), www.congress.gov/crs-product/RL32748.

38. 38. All but six states offer supplemental payments through either federal- or state-administered programs. See Social Security Administration, “Supplemental Security Income (SSI),” in Understanding Supplemental Security Income SSI Benefits—2024 Edition (accessed March 11, 2025), www.ssa.gov/ssi/text-benefits-ussi.htm.

39. 39. California, Illinois, Maine, Minnesota, and Washington have state-funded food assistance programs for certain groups of immigrants who are ineligible for federal SNAP benefits. Those programs provide benefits at the SNAP benefit level. See National Immigration Law Center, “Table 12: State-Funded Food Assistance Programs” (updated January 2024), https://tinyurl.com/25u47yzc. Connecticut has a similar program but does not provide benefits at the full SNAP benefit level, so it is excluded from this analysis. As of September 2022, California, Hawaii, Illinois, Maine, New Hampshire, and Washington have state-funded SSI programs for certain immigrants who are ineligible for federal SSI benefits. Information about Hawaii’s and Washington’s programs suggests that state benefit levels are lower than federal benefit levels. See National Immigration Law Center, “Table 9: State-Funded SSI Replacement Programs” (September 2022), http://tinyurl.com/5msxkd8u.

40. 40. Arrested people stay in a jail for about one month, on average; see Zhen Zeng, Jail Inmates in 2022—Statistical Tables (Bureau of Justice Statistics, December 2023), https://tinyurl.com/366rdc89. Thus, the number of members of the surge population who were in jail at the end of 2023 is smaller than the number who were in jail at any point during that year. The same applies to inmates in prison, though inmates generally remain in prison for more than a year; see Danielle Kaeble, Time Served in State Prison, 2018 (Bureau of Justice Statistics, March 2021), https://tinyurl.com/pt9pddtc.

41. 41. CBO used the ACS data to determine the number of people under age 65 living in institutional group quarters and the BJS data to estimate the share of such people who were incarcerated. For the BJS data, see Bureau of Justice Statistics, Corrections Statistical Analysis Tool (CSAT) Prisoners, “Advanced Query” (accessed February 17, 2025), http://tinyurl.com/3x692vuw.

42. 42. Four states (Alaska, California, Oregon, and Rhode Island) did not report any data on citizenship status. For those states, CBO imputed the number of noncitizens in state prisons. The imputed numbers were calculated using data from other states and were based on the statistical relationship between the number of noncitizens multiplied by the overall incarceration rate and the number of incarcerated noncitizens. See E. Ann Carson and Rich Kluckow, Prisoners in 2022—Statistical Tables (Bureau of Justice Statistics, November 2023), Table 22, https://tinyurl.com/mu7y49nt.

43. 43. Bureau of Justice Assistance, “BJA FY 2023 State Criminal Alien Assistance Program (SCAAP) Award Details,” https://tinyurl.com/2tsj5b2j, and “BJA FY 2024 State Criminal Alien Assistance Program (SCAAP) Award Details,” https://tinyurl.com/2zss7dy6. The estimate for calendar year 2023 is three-quarters of the award for federal fiscal year 2023 (which ended in September 2023) and one-quarter of the award for federal fiscal year 2024 (which began in October 2023).

44. 44. Michelangelo Landgrave and Alex Nowrasteh, Illegal Immigrant Incarceration Rates, 2010–2023 (Cato Institute, April 2025), https://tinyurl.com/564dtpcj; Alex Nowrasteh, Illegal Immigrant Murderers in Texas, 2013–2022 (Cato Institute, June 2024), https://tinyurl.com/bdhs73m3; and Michael T. Light, Jingying He, and Jason P. Robey, “Comparing Crime Rates Between Undocumented Immigrants, Legal Immigrants, and Native-Born U.S. Citizens in Texas,” Proceedings of the National Academy of Sciences, vol. 117, no. 51 (December 2020), pp. 32340–32347, https://tinyurl.com/3ucekbph.

45. 45. Christian Gunadi, “On the Association Between Undocumented Immigration and Crime in the United States,” Oxford Economic Papers, vol. 73, no. 1 (January 2021), pp. 200–224, https://tinyurl.com/dyzmm65w; Michael T. Light and Ty Miller, “Does Undocumented Immigration Increase Violent Crime?” Criminology, vol. 56, no. 2 (March 2018), pp. 370–401, https://tinyurl.com/4ze3kyfy; and Kristin F. Butcher and Anne Morrison Piehl, “Cross-City Evidence on the Relationship Between Immigration and Crime,” Journal of Policy Analysis and Management, vol. 17, no. 3 (Summer 1998), pp. 457–493, https://tinyurl.com/y5k3n5kb.

46. 46. The last two categories exclude the health insurance and income security programs that are described separately above.

Appendix A: Methods Used to Estimate the Effects of the Immigration Surge

This appendix provides additional details about how the Congressional Budget Office estimated the effects of the surge in immigration that began in 2021 on state and local budgets in 2023. This analysis focuses on the tax liabilities of and spending in that year for the people who arrived as part of the surge and their children.

CBO presents two measures of the effects of the immigration surge on state and local budgets: direct and potential effects. Direct effects reflect changes in state and local governments’ revenues and spending that were directly attributable to people in the surge population. Potential effects reflect those direct effects plus increases in property tax revenues, additional tax revenues from greater economic activity, and nonbudgetary costs associated with greater demand for government services.

The first section of this appendix describes CBO’s estimates of characteristics of the surge population—namely, the age, educational attainment, state of residence, sex, marital status, and family income of immigrants in that population. Those characteristics were used to identify the people who were most likely part of the surge population in data from the American Community Survey (ACS) and Customs and Border Protection (CBP) and to calculate taxes paid by them and spending for them. The next two sections provide more detailed discussion of how CBO allocated total state and local tax receipts and expenditures in 2023 to those people.

Characteristics of the Surge Population in 2023

CBO has documented an increase starting in 2021 in the flow of people immigrating to the United States whom the agency refers to as other foreign nationals. Some of them received permission to enter or remain in the country, and some did not. (For a discussion of the categories that CBO used for immigrants in this analysis, see Box 1.) The net immigration of other foreign nationals from 2021 to 2023 significantly exceeded the rates observed in previous years and ultimately added an estimated 4.3 million more immigrants to the U.S. population by the end of that period than there would have been if net immigration had remained at historical levels. When births and deaths among that group of immigrants are accounted for, the net increase in the population totals 4.4 million people.

In this analysis, the agency refers to those 4.4 million people as the surge population. Although the immigration surge continued into 2024, this analysis focuses on spending and revenues by state and local governments in 2023 because, at the time of publication, that is the most recent year for which the necessary data were available.

Using the 2023 ACS and information from CBP about immigrants’ encounters with CBP officials, CBO compared the characteristics of the other foreign nationals who arrived from 2021 to 2023 (and their native-­born children) with those of the rest of the population (see Table A-1).1 For example, the encounters data from CBP showed that many of the immigrants in the surge population were from Central or South America. The agency found that the surge population in 2023 differed from the general population in the following ways:

• People in the surge population were younger than the general population, on average.
• People in the surge population were less educated, on average, than those in the general population.
• The surge population was concentrated in several states. Immigrants residing in four states accounted for half of the surge population in 2023.
• A larger portion of the surge population than of the general population was male.
• The percentage of people ages 18 to 54 who were married was similar for the two populations.
• The average income of people in the surge population was less than that of the rest of the U.S. population.

  Table A-1.

  Characteristics of the Surge Population and the U.S. Population in 2023

  Percent

  

  Notes

  Data source: Congressional Budget Office, using data from the American Community Survey (ACS) and Customs and Border Protection (CBP). See www.cbo.gov/publication/61256#data.

  The characteristics of the surge population were estimated using the sample of noncitizen immigrants who arrived between 2021 and 2023 (and their native-born children) in the 2023 ACS weighted by the country of origin derived from data about encounters with CBP officers. The characteristics of the U.S. population were estimated using the 2023 ACS.

  a. All children under age 15 are included as having less than a high school diploma. People who were enrolled in school in 2023 are included at the highest level of attainment they had achieved, even if they expected to attain additional education.

  b. For consistency with previous work, the income distribution for the surge population in 2023 was projected using the 2021 and 2022 ACS along with the 2019 Current Population Survey’s Annual Social and Economic Supplement, as described in Congressional Budget Office, Effects of the Immigration Surge on the Federal Budget and the Economy (July 2024), www.cbo.gov/publication/60165. The federal poverty guidelines are developed by the Department of Health and Human Services and are used to determine financial eligibility for certain programs. See Department of Health and Human Services, “2023 HHS Poverty Guidelines” (accessed March 17, 2025), https://tinyurl.com/yp99npzw.

Revenues

To estimate the effects that immigrants who were part of the surge had on state and local revenues, CBO used methods similar to those used in its analysis of their effects on the federal budget.2 CBO estimated the largest revenue sources using its tax simulation model. For state and local income taxes, CBO applied a simplified state income tax calculator to the expected income of immigrants in the surge. For sales and property taxes, CBO first allocated the total revenues from those taxes to a nationally representative sample of taxpayers on the basis of estimated spending on various goods and services and then estimated the amount of those revenues that could be attributed to the immigration surge. Revenues from other types of state and local taxes attributable to the surge population were estimated using various simplified methods, such as per capita allocation. CBO estimated the indirect revenue effects arising from the additional economic activity brought on by the surge by applying average state and local revenues as a percentage of gross domestic product (GDP).

Imputing Immigrant Status and Income

CBO identified immigrants who were part of the surge using the same approach it used to produce its estimates of the effects of the surge on the federal budget. For that analysis, CBO estimated which taxpayers among the representative sample that the agency uses as an input to its individual tax microsimulation model were part of the immigration surge and used that model to project their income and tax payments into the future.3 CBO is permitted to use that tax data only for its baseline projections of federal receipts, so for this analysis, the agency took the same approach but used public-use tax data that are statistically altered to protect the confidentiality of individual taxpayers. CBO assigned a probability of being part of the immigration surge to each taxpayer in the model on the basis of differences between CBO’s baseline population projections (which include the immigration surge) and projections under a counterfactual scenario in which the surge did not occur. The probabilities reflect the expected age, sex, marital status, and income distribution of the surge population.

Sources of State and Local Revenues

To simulate state and local revenues, CBO first decided which revenue sources to include in its analysis. In general, the analysis included revenues from state and local taxes, which change with additional economic activity, but excluded most nontax revenue sources, either because the surge in immigration was unlikely to affect them (such as grants from the federal government) or because changes in revenues from a source were matched by changes in spending from the source that were also omitted from the analysis (such as revenues and spending from nongovernmental enterprises).

CBO included the following revenue sources in its analysis: individual income taxes, general sales taxes, selective sales taxes (also known as excise taxes), property taxes, motor vehicle license taxes, corporate income taxes, other taxes, workers’ compensation revenues, and other insurance trust revenues (see Table A-2).4 To derive the appropriate revenue totals, CBO used data from the Census Bureau’s Annual Survey of State and Local Government Finances and from the Bureau of Economic Analysis’s (BEA’s) national income and product accounts (NIPAs)—specifically, NIPA Tables 3.3 and 3.19, which provide information about state and local governments’ current receipts and expenditures.5 The Census Bureau provides detailed, state-by-state, general revenues for state and local governments, but BEA excludes some nontax revenue sources (such as charges and user fees).

CBO used the detailed information in the Census Bureau’s tables only for the categories that are included in the NIPAs as current tax receipts and as contributions for government social insurance. Thus, CBO omitted from its analysis revenues from several of the sources in the Census Bureau’s dataset, including all current charges, miscellaneous general revenues (other than special assessments), utility revenues, and revenues of state and local employee pension funds. CBO also excluded from its analysis transfers from the federal government and unemployment compensation revenues (which are included in the federal budget). CBO included net liquor store revenues (that is, revenues minus the stores’ expenditures) in alcohol sales tax receipts, and it combined special assessments with property tax receipts.

CBO’s method of allocating revenues relied on estimates of state and local revenues by state. The Census Bureau’s data about state and local revenues are published with a significant lag, and CBO made several adjustments to those data to project calendar year 2023 tax liabilities. The most recent data that include details about state and local governments’ receipts are for fiscal year 2022.

The Census Bureau publishes more timely data about state governments’ tax receipts.6 CBO used those data for fiscal years 2022 and 2023 to project local revenues for fiscal year 2023 on the basis of the assumption that growth in revenues collected by local governments in a state equaled growth in revenues collected by the state government. CBO then created estimates for each state of tax liabilities for calendar year 2022 for each of the revenue sources by combining the fiscal year 2022 and 2023 tax receipts, taking into account the dates of each state’s fiscal year. The agency then scaled those targets to match the estimates for calendar year 2023 that BEA published in NIPA Table 3.3 (see Table A-3).

Estimates of Revenues From Specific Sources

The rest of this section provides additional information about the estimates of revenues from five different sources: individual income taxes, general sales taxes, selective sales taxes, property taxes, and other taxes. The indirect effects on state and local governments’ tax revenues from growth in the overall economy attributable to the surge are also discussed.

State and Local Individual Income Taxes. CBO estimated state and local individual income taxes for all taxpayers in its tax microsimulation model by using a simplified, table-based state and local tax calculator. The agency created that simplified calculator by using a more detailed state and local tax calculator to simulate tax liabilities in each state for tax units with varying numbers of taxpayers and dependents and with incomes ranging from zero to $500,000 in $1,000 increments.7 CBO then applied the average state and local income tax rate from that simplified calculator to the income of each taxpayer in its sample. The total amounts of simulated taxes align well with the aggregate amounts in the Census Bureau’s data and the NIPAs. CBO then used the probabilities of taxpayers’ being part of the surge to determine the state and local income taxes that those immigrants paid.

To account for the geographic distribution of immigrants, CBO calculated state and local income tax revenues in each state for each record with a nonzero probability of representing an immigrant in the surge population and weighted each state revenue amount by CBO’s expected distribution of such immigrants among states, thus creating a 50-state weighted value. That reweighting lowered estimated state and local income tax receipts because a disproportionate share of immigrants were in Florida and Texas, which do not have state income taxes. The agency’s total estimate summed those state-weighted amounts for all taxpayers weighted by the estimated probability of their being part of the surge.

In CBO’s estimation, rates of compliance for income taxes were 15 percent lower among the surge population than among the total population because a larger share of the surge population worked without authorization. The same factor was applied in CBO’s analysis of the immigration surge’s effects on the federal budget. CBO expects that only about half of the immigrants in the surge population received authorization to work, and in the agency’s estimation, immigrants without work authorization are less likely to comply with the tax system than people who are authorized to work.

State and Local General Sales Taxes. CBO used the same probabilities of taxpayers’ being part of the surge population, income distribution, and distribution of that population across states to produce the estimates of revenues from state and local general sales taxes that it used to estimate revenues from state and local income taxes. The agency began with its estimates of total revenues from general sales taxes from each state and then, using estimates by others of state and local business tax receipts, split those amounts into the share paid by individuals and the share paid by businesses on intermediate goods and services.8 Next, CBO used different approaches to allocate those two components of general sales tax receipts to all taxpayers in its tax microsimulation model.

Allocating the general sales taxes paid by individuals involved several steps. First, CBO used the “2023 Optional State Sales Tax Tables” that the Internal Revenue Service (IRS) provided in the instructions for Schedule A (itemized deductions) to estimate a sales tax amount for each record in its tax microsimulation model.9 Those tables are based on the Consumer Expenditure Survey (CEX), and the sales tax amounts vary with income, family size, and the sales tax rate and base in each state. CBO next increased those amounts to account for revenues from local general sales taxes and vehicle sales taxes.10 To account for revenues from vehicle sales taxes, CBO used national data from the 2023 CEX to estimate the share of total consumption spent on vehicle purchases by income and increased the sales tax for each record accordingly.11 To account for revenues from local sales taxes, CBO added estimates of revenues from those taxes to its state sales tax amounts by using ratios of the sum of state and average local general sales tax rates to state general sales tax rates that were calculated using data from the Tax Foundation.12 Finally, CBO adjusted those amounts by state so that the state totals matched CBO’s targets.

Businesses pay state and local general sales taxes on intermediate goods and services that they purchase from other businesses. CBO allocated the cost of those business sales taxes to consumers. Because businesses sell products both in state and across state lines, CBO allocated two-thirds of each state’s business sales tax revenues to in-state residents and one-third to out-of-state residents.13 Both the in-state and out-of-state allocations were based on total estimated consumption. Specifically, CBO used data from the 2023 CEX to estimate consumption-to-income ratios for the different income brackets. Those ratios were used to estimate consumption for each record in the tax microsimulation model. The estimates do not include any lost sales tax revenues stemming from certain cities’ housing immigrants who were part of the surge population in hotels. In CBO’s estimation, the negative effects on tourism and business travel to those cities were small given that hotel occupancy rates remained below pre­pandemic levels.14

CBO then adjusted for the expected geographic distribution of immigrants in the surge by estimating general sales taxes paid by individuals and businesses for each record in each state on the basis of consumption (as described above) and weighting by the distribution of the surge population across states. CBO reduced its general sales tax estimates for the surge population by 15 percent to reflect a reduction in consumption because, in the agency’s estimation, those immigrants are likely to have sent a portion of their income to family members who reside in other countries.

State and Local Selective Sales Taxes. CBO estimated revenues from state and local selective sales taxes by using the same immigrant probabilities, income distribution, and distribution of the surge population across states that it used to estimate revenues from state and local income taxes. The agency began with its estimate of total revenues in each state from sales taxes on motor fuels, utilities, tobacco, alcohol, and other goods subject to selective sales taxes.

Selective sales taxes were allocated on the basis of the consumption of goods subject to selective sales taxes. CBO used CEX data for 2023 to estimate ratios of consumption of several categories of goods—motor fuels, utilities, tobacco, alcohol, and all goods (used for all other selective sales taxes)—to income for different income groups. The agency used those ratios to estimate each type of consumption for each record in the tax microsimulation model. CBO then allocated selective sales taxes to each record on the basis of state and type of consumption to match the target amount for each state. Next, the agency estimated the amount of selective sales taxes paid by immigrants by estimating selective sales taxes for each record in each state and weighting those amounts by the distribution of the surge population across states. Finally, CBO reduced its estimates of revenues from state and local selective sales tax attributable to the surge population by 15 percent to reflect the expectation that those immigrants would have sent a portion of their income to family members who resided in other countries, reducing their consumption.

State and Local Property Taxes. CBO used a multistep process to estimate property taxes. First, the agency divided property taxes into those levied on owner-­occupied housing, rental housing, and commercial property, and then it allocated each of those components separately:

• The estimated property tax payments for owner-occupied housing were based on the itemized deductions for those taxes reported on federal tax returns; for filers in CBO’s model who did not itemize, amounts were estimated on the basis of those paid by similar taxpayers.
• Property taxes on rentals were estimated using personal consumption expenditures for owner-occupied and rental housing in NIPA Table 2.4.5. To allocate the rental property component, CBO estimated rental consumption of taxpayers who did not own a home using 2023 CEX data and a procedure similar to the one it used to estimate revenues from selective sales taxes.
• The remaining property taxes were estimated to have been levied on properties owned by businesses. CBO allocated those property taxes paid by business to individual taxpayers in proportion to their estimated total consumption.

Next, CBO estimated the revenues from rental and business property taxes that were attributable to people in the immigration surge by estimating those taxes for each record in each state and then weighting those amounts by the distribution of the surge population across states. Because the immigration surge started in 2021 and immigrants in the surge population had been in the country for three years at most in 2023, CBO simplified its estimates by calculating them as if everyone in the surge population was a renter in 2023.

CBO created two measures of revenues from property taxes—one for direct effects and the other for potential effects. The property tax effect attributable to immigrants in the surge reflects only the property tax revenues stemming from rent payments made by those individuals. CBO’s estimate of the immigration surge’s direct effect on property taxes is smaller than its estimate of the potential effect because, in the agency’s estimation, the increase in population resulting from the surge did not immediately lead to a proportionate increase in property values and tax assessments—that is, the agency expected the increase in revenues from property taxes to lag behind the increase in demand for housing and the subsequent increase in property values brought on by the immigration surge.

The property taxes paid by owners of rental properties (estimated to be half of the collections of property taxes on rental properties) are included in the effects attributable to the increase in overall economic activity stemming from the surge, and thus part of the potential effects of the surge.15 The additional property taxes paid by businesses were not, in CBO’s assessment, directly paid by the surge population but instead were passed on to the broader population through higher prices or to owners of capital through lower returns. Like the revenues from property taxes borne by owners of rental properties, collections of property taxes paid by businesses are included in the estimate of the effects of the increase in overall economic activity.

Other State and Local Taxes. CBO used various reduced-form methods to estimate revenues from other types of state and local taxes. For example, the agency estimated the average motor vehicle license taxes paid by people over the age of 17 and then applied that average amount to the agency’s estimate of the number of immigrants in the surge who were over the age of 17 in 2023. Similarly, CBO estimated the average amount of workers’ compensation taxes paid by employed people and then multiplied that average amount by its estimate of the number of immigrants in the surge who worked. CBO expects that collections of many other types of taxes—including license taxes, death and gift taxes, and oil and mineral extraction taxes (known as severance taxes)—did not change in response to the increase in immigration. By contrast, corporate income taxes are estimated to have increased because of the increased economic activity attributable to the immigration surge, and those additional taxes were paid by taxpayers who were not part of the surge; thus, they are included in CBO’s estimate of the potential effects of the immigration surge on state and local revenues.

Indirect Effects on State and Local Tax Revenues. In addition to the direct effects that resulted from the increase in the number of people paying taxes, CBO estimates that the changes in the overall economy stemming from the immigration surge also affected state and local budgets. For example, businesses that saw increased demand for their products because of the immigration surge are likely to have hired more workers and invested in additional machinery, causing their profits to rise. Furthermore, innovation-related productivity is estimated to be higher because of the surge, leading to higher income from both labor and capital, though those effects may take time to materialize. Those indirect effects increased state and local tax receipts from the nonsurge population. CBO used a multistep process to estimate those effects.

• The agency first estimated the effect of the immigration surge on GDP by comparing its baseline economic forecast (which incorporates the effects of the surge) with its forecast of outcomes under a counterfactual economic scenario in which the surge did not occur.
• Next, the agency subtracted from that estimate of the immigration surge’s overall effect on GDP its estimate of the surge population’s earnings to estimate the growth in GDP accruing to the nonsurge population.
• Finally, CBO applied an overall state and local tax rate (equal to the ratio of total state and local revenues to GDP) to estimate the additional revenues state and local governments received as a result of the additional GDP.

Spending

To estimate the amounts that state and local governments spent on the surge population, CBO relied on its own estimates of the surge population, survey data, and other public data. Although the goal was to estimate the share of the total costs of goods and services provided by state and local governments that was attributable to the surge in immigration—similar to the approach the agency used for revenues—CBO first needed to develop an allocation method. That method involved distributing the total costs of goods and services across the entire population by identifying all users and estimating the average cost per user. When applicable, CBO adjusted the estimates for the surge population that differed from those for the rest of the population to account for direct effects of the surge population on state and local budgets (as the agency did in its analysis of revenues from property taxes).

The method involved three major steps. First, CBO estimated the total cost of various goods and services provided by the state and local governments by using data from the NIPAs. Second, CBO used data from the Current Population Survey’s Annual Social and Economic Supplement (CPS ASEC) to construct a sample of the U.S. population in 2023 that included a subsample of people who were likely to be in the surge population. Third, using the previous two components, CBO developed a method to allocate the total spending amounts in the NIPAs to people in the sample. The sum of the amounts allocated to people in the sample who were estimated to have been part of the surge population is the effect of the surge in immigration on state and local spending in 2023. For some categories of spending, the allocation method distinguishes between the direct and potential effects of the immigration surge on state and local budgets.

Revised Estimates of Total Spending by Category

The total state and local spending included in this analysis is consistent with the estimate of current expenditures minus spending on interest payments on debt and spending on subsidies in the NIPAs. To estimate the total spending by state and local governments for each category, CBO started with the total current expenditures by state and local governments in 2023 in the NIPAs—$3,762 billion (Table 3.16, line 80). That total includes federal grants-in-aid of $951 billion (Table 3.17) and spending on interest payments and subsidies of $277 billion (Table 3.16) that is unrelated to the surge population. Both of those amounts were subtracted from the total current expenditures, resulting in total state and local spending of $2,535 billion in 2023 in the categories included in CBO’s analysis (see Table A-4).

Because federal grants supporting state and local programs are outside the scope of this analysis, the total for each spending category should not include such grants. The estimates for spending categories in NIPA Table 3.16 include any grants-in-aid that the state or local government received. To exclude such grants from each spending category, CBO either used data from a source that did not include that aid (Table 3.12, for example) or subtracted the amount of grants-in-aid given in Table 3.17 from the corresponding category in Table 3.16.

In some cases, NIPA Table 3.17 provides a total that must be allocated to subtotals provided in Table 3.16. CBO used other sources of data to allocate those amounts to different categories. For example, Table 3.16 provides a total for state and local governments’ spending for education and subtotals for elementary and secondary education, higher education, and “libraries and other.” CBO used data from the National Center for Education Statistics and the Institute of Museum and Library Services to allocate federal grants-in-aid for education to those components of spending for education.16

CBO expects that some categories of state and local spending for general services will take time to respond to the immigration surge, so two sets of totals were created—one for the direct effect of the surge population on such spending and one for the potential effect. In CBO’s assessment, for at least the first three years of the immigration surge, the increase in population had no effect on spending for the following categories included in NIPA Table 3.16: the executive and legislative functions of government, tax collection and financial management, transportation (air, water, highway, transit, and railroads), other economic affairs (general, labor, agriculture, energy, and natural resources), recreation and culture, and libraries. The spending on general services allocated to calculate the direct effect of the immigration surge was $603 billion, and the amount allocated to calculate the potential effect was $1,200 billion. Thus, the spending for general services attributable to the immigration surge that, in CBO’s assessment, did not immediately respond to the increase in population that resulted from the surge amounted to $597 billion.

In all, after federal grants-in-aid and spending on interest payments and subsidies were removed, $2,535 billion was allocated to estimate the potential effect of the surge population. After the portion of spending for general services that, in CBO’s assessment, did not immediately respond to the surge population ($597 billion) was subtracted, $1,938 billion remained to be allocated to the U.S. population to estimate the direct effect of the surge population.

A Sample of the Surge Population

To create a representative sample of the surge population in 2023, CBO used the Current Population Survey’s Annual Social and Economic Supplements. In addition to the data for 2023, CBO used data from the CPS ASECs for 2021 and 2022 to increase its sample size and improve the accuracy of its estimates. The combined CPS ASEC sample provides a representative sample of the U.S. civilian noninstitutionalized population over those three years. To create a sample of the surge population, CBO used a sample of noncitizens who had arrived within three years of the survey and who were identified by the agency as other foreign nationals on the basis of several variables in the CPS ASEC data.17

CBO estimates that immigrants in the surge population differ significantly from immigrants who arrived earlier in terms of their age, educational attainment, and state of residence. To match the characteristics of the surge population, the sampling weights for the sample of other foreign nationals were adjusted to match the distribution of the surge population by age, educational attainment, state of residence, and year of immigration. Corresponding changes were made to the weights for records representing people not in the surge sample.

As a result of those adjustments, the sum of the weights for the records in the surge sample is equal to CBO’s estimate of the increase in population in 2023 attributable to the immigration surge (4.4 million people), and the sum of all sample weights equals CBO’s estimate of the resident population in 2023. For the sample of immigrants in the surge, the distributions by age, educational attainment, state of residence, and year of immigration match those discussed above, and the distributions for the whole sample match those in the original CPS ASEC sample.

To estimate other characteristics of the surge population (such as the percentage residing in group quarters), CBO used a sample of recent immigrants in the ACS reweighted on the basis of CBP data. The ACS has a larger sample size and broader set of populations than the CPS ASEC, which results in more precise estimates.

Allocation of Spending to the Surge Population

In the last step, CBO allocated the total spending by state and local governments to individuals in its sample (including immigrants in the surge population). The agency used data from multiple sources to inform that allocation:

• Survey data were used to identify people in the surge population who incurred costs for education, state prisons and local jails, and workers’ compensation. CBO multiplied the number of users of goods and services by the average spending per user to calculate the total increase in the costs to state and local governments resulting from the surge in immigration. The agency expects that the average cost of those services per user (excluding the costs of public primary and secondary education) was the same for the surge population as it was for the general population.
• Additional information was used to estimate the number of people in the surge population who incurred costs in other spending categories for which separate totals were either provided in the NIPAs or estimated on the basis of other administrative data. Those categories consist of states’ shares of spending for certain federal health insurance and income security programs, state-funded health insurance and income security programs for immigrants, shelter and related services, and border security measures.
• The rest of the spending by state and local governments cannot be attributed to individuals because neither survey nor administrative data provide enough information to differentiate the use of state and local government services by people in the surge population from the use of those services by people in the general population. CBO distributed such spending equally across the general population.

Because, in CBO’s estimation, state and local governments’ spending lags behind the increase in population brought on by the immigration surge, the agency adjusted the costs per user for certain goods and services to distinguish between the direct and potential effects. CBO used lower costs per user for the direct effect than for the potential effect of the surge on public primary and secondary education to reflect the historical relationship between changes in enrollment and changes in spending. The costs per user were lower for the direct effect on spending for general services than they were for the potential effect of such spending because CBO removed some categories of spending from its calculations of the direct effect.

Estimates of Spending for Specific Categories

The rest of this section provides additional information about the estimates of four categories of spending: public primary and secondary education, postsecondary education, Medicaid and the Children’s Health Insurance Program (CHIP), and Temporary Assistance for Needy Families (TANF).

Public Primary and Secondary Education. In 2023, CBO estimates, state and local governments spent $5.7 billion more on public primary and secondary education than they would have if the immigration surge had not occurred. To estimate state and local spending for primary and secondary education attributable to the surge, CBO calculated the difference between its estimate of such spending under the baseline scenario and its estimate of such spending under the counterfactual scenario.

For each scenario, CBO estimated the cost as the product of spending in 2022, the growth in enrollment from the 2022–2023 school year to the 2023–2024 school year, and the growth in spending per student over that period. The estimated direct effect and potential effect differed by CBO’s estimate of the growth in enrollment and in spending per student from 2022 to 2023.

The estimated number of students enrolled in primary and secondary schools is the total number of children ages 5 to 15 and children older than 15 in the CPS ASEC sample who reported enrollment in high school.18 That enrollment number was further adjusted to account for the fact that the 2023 arrival cohort was enrolled for half a year, on average. In the CPS ASEC sample, 0.60 million full-year equivalent enrollees in primary and secondary schools in 2023 were members of the surge population. Using the ACS, CBO estimated that about 90 percent of the immigrants in the surge population who were enrolled in primary and secondary schools that year were enrolled in public schools. Thus, CBO estimated that in 2023, the full-year equivalent enrollment in public primary and secondary schools of people in the surge population was 0.55 million.

For the estimate of the direct effect, estimates of spending per student were based on the statistical relationship between changes in enrollment and changes in spending per student; for the estimate of the potential effect, the costs per student were the total spending on public primary and secondary education divided by the number of children enrolled in those schools in 2023.

Like property taxes, state and local governments’ spending on primary and secondary education did not immediately rise in proportion to the increase in the number of students arriving as part of the surge; rather, the increase in spending lagged behind the change in population. To quantify the possible response by state and local governments to the surge in immigration, CBO estimated the relationship between state-level enrollment changes and state-level changes in spending per student by using information from the Common Core of Data (CCD) for the 1999–2000 school year to the 2019–2020 school year.19 If state and local governments adjusted the spending in proportion to changes in enrollment, the coefficient of enrollment growth would be expected to be zero. But CBO found that an increase of one percentage point in enrollment growth was associated with a 0.37 percentage-point decrease in growth of spending per student; conversely, a one percentage-point decrease in the enrollment growth was associated with a 0.34 percentage-point increase in growth of spending per student. Those relationships suggest that the spending on public primary and secondary education did not immediately adjust in proportion to the change in enrollment. Thus, an increase in enrollment leads to a decline in average spending, and a decrease in enrollment, to an increase in average spending.

In addition, evidence suggests that as the number of immigrant children in public schools increases, some students transfer from public to private schools.20 Thus, in CBO’s assessment, some students who were not part of the surge population would have moved to private schools in response to the surge. As a result, the net increase in enrollment in public schools was smaller than the number of children in the surge population who enrolled in public schools. On the basis of limited evidence, CBO estimated that for every six students in the surge population who attended public schools, one student from the nonsurge population moved from a public to a private school.

• For the baseline scenario, which includes the surge population, the enrollment change for each state in 2023 was calculated using data from the CCD on enrollment for the 2022–2023 and 2023–2024 school years. The growth in spending per student was estimated on the basis of that change in enrollment and the statistical relationship between enrollment and spending per student described above.
• For the counterfactual scenario in which the immigration surge did not occur, the enrollment change in 2023 was calculated by subtracting five-sixths of the estimated enrollment of students in the surge population from enrollment in the baseline scenario (to account for some of the nonsurge students’ shift to private schools). The growth in spending per student was estimated in the same way as above except for that it was estimated on the basis of that modified measure of enrollment growth in 2023.

The potential costs to state and local governments are greater than the direct costs—$9.4 billion compared with $5.7 billion in 2023. The approach that CBO used to estimate the potential effects of the surge population on state and local governments’ spending on public primary and secondary education was similar to the one it used to estimate the direct effects, with two differences: No adjustment was made to account for the possibility that some nonsurge students who were enrolled in public schools would enroll in private schools, and spending per student was not adjusted to account for the change in enrollment.

The estimates of both the direct and potential effects account for costs incurred by schools to provide additional services for English-language learners. Because most immigrants in the surge population come from non-English-speaking countries, state and local governments incurred an additional $1.2 billion to provide those services to the surge population, CBO estimates. That amount was included in the agency’s estimates of both the direct and potential effects of the surge population on state and local spending.

That estimate of the cost of additional services was based on the share of the surge population with limited English proficiency and the additional funding provided to schools for English learners. Using the 2023 ACS, CBO estimated that in that year, about half of the surge population did not speak English or did not speak it well. That share was larger for more recent arrivals—about 60 percent for people who arrived in 2023 and about 40 percent for those who arrived in 2021. By contrast, only 1 percent of the people in the nonsurge sample did not speak English or did not speak it well in 2023.

Many states use a formula to allocate funding for English-language learners to schools and school districts in proportion to their fall enrollment. The formulas give additional weight to students who require additional services, such as those with disabilities and those who have limited English proficiency. CBO calculated a weighted average of the additional funding that states give English-language learners in their formulas.21 CBO estimated that the average additional weight for English-language learners, based on the state formulas for which such a calculation was feasible, was just over 25 percent.22

Postsecondary Education. Immigrants in the surge population could be eligible for in-state tuition and state or local financial aid through two pathways. First, at least half of the surge population lived in states in which other foreign nationals could qualify for in-state tuition at public institutions if they attended high school in the state for two or three years and graduated.23 Second, in certain states, members of the surge population with certain federal immigration statuses would have been eligible for in-state tuition through the same channel as U.S. citizens and permanent residents: by establishing legal domicile (or residence), which typically requires evidence that a person has lived in the state for at least one year and, depending on the state, has some type of lawful status or presence. To simplify the analysis, CBO applied the immigration and residency requirements for in-state tuition to estimate financial aid provided by state and local governments.24

CBO estimates that only a small percentage of the surge population was eligible for financial support for postsecondary education from state and local governments in 2023. The first pathway was unavailable to the vast majority of the surge population in that year. On the basis of its analysis of the six states with the largest shares of the surge population, CBO estimates that only a very small number of those immigrants who arrived in 2021 lived in a state with a two-year high school attendance and graduation requirement and would have been able to meet the requirement and enroll in an in-state public postsecondary institution by the end of 2023. Members of the surge population living in states with a three-year high school attendance and graduation requirement would not have been eligible for state and local support for postsecondary education. (Of the six states analyzed, only New York had a two-year requirement; the other states had a three-year requirement.) Consequently, to simplify the analysis, CBO did not estimate that members of the surge qualified for the first pathway.

The second pathway was also unavailable to nearly all of the surge population in 2023. CBO estimates that very few members of that population could have met the Higher Education Act’s requirement that a person must “demonstrate the intention of becoming a citizen or permanent resident” to be eligible for federal financial aid.25 Some states that provide the second pathway rely on federal regulations to determine eligibility for in-state tuition.26 Although some other states that provide the second pathway do not rely on those regulations, CBO estimates that the number of immigrants in the surge population who qualified for the second pathway in those states was negligible.27

Medicaid and State-Funded Programs That Offer Medicaid-Like Benefits. The first component of the Medicaid estimate is the total amounts paid for Medicaid or the Children’s Health Insurance Program by states for people in the surge population who were eligible for those programs—$0.3 billion.28 That estimate is based on CBO’s estimate of federal Medicaid and CHIP spending for the surge population.29 To estimate the states’ share of spending for Medicaid and CHIP for the surge population, CBO used its estimate of the federal share of such spending and its estimate of the federal medical assistance percentage (that is, the percentage of costs paid by the federal government, which varies by state) for that population, which averaged about 70 percent across states.

On the basis of the CPS ASEC sample of the surge population described above, CBO estimated that about one-third of children and less than 10 percent of adults in the surge population resided in states that funded programs that provided Medicaid-like benefits in 2023. As of November 2023, 11 states and the District of Columbia had created state-funded Medicaid-like programs that provide health care coverage to children regardless of their immigration status.30 Four states and the District of Columbia offered state-funded Medicaid-like benefits to adults.31 To simplify its analysis, CBO did not estimate the costs of the state-funded coverage for limited populations in Connecticut, Hawaii, Maine, Massachusetts, New York, Pennsylvania, Vermont, and Washington.32

By applying the shares discussed above to the estimated 3.5 million people in the surge population who were ineligible for federal Medicaid benefits in 2023, CBO estimated that less than half a million people in the surge population were eligible for state-funded Medicaid-like benefits that year. In CBO’s assessment, people in the surge population who were eligible for state-funded programs had the same likelihood of enrolling in those programs as people in the surge population who were eligible for federal benefits had of enrolling in the federal programs, and the costs per person for people in the surge population enrolled in the state-funded programs were the same as the costs per person for federal benefits for that population. On the basis of those assessments and after accounting for part-year residency among the 2023 arrival cohort, CBO estimated that the costs of state-funded Medicaid-like programs attributable to the surge population were $0.2 billion in 2023.

Temporary Assistance for Needy Families. TANF is a joint federal-state program that provides cash assistance, work support, and other services to low-income families with children. The federal government provides grants to states, and states contribute their own dollars to fulfill the maintenance-of-effort (MOE) requirement. The discussion here focuses on the MOE spending that could be attributed to the surge population, which CBO estimated by comparing states’ MOE spending from fiscal year 2019 to fiscal year 2023.

CBO’s estimate of states’ MOE spending on cash assistance is based on a tabulation of the TANF national data files for cash assistance by the source of funding (federal versus state) and the citizenship of all family members.33 The agency estimated that between fiscal years 2019 and 2023, states’ MOE spending on TANF cash assistance increased by about a third for families with citizens only and more than doubled for families with at least one member who was not a citizen or whose status was unknown. CBO attributed $0.1 billion of states’ MOE spending on TANF cash assistance to the surge population in fiscal year 2023. That estimate reflects the rate of growth from fiscal year 2019 to fiscal year 2023 in states’ MOE spending on TANF cash assistance for families with at least one member who was not a citizen or whose status was unknown, minus the growth rate over those years for families that consisted only of citizens.

Between fiscal years 2019 and 2023, the increase in states’ MOE spending on cash assistance accounted for about one-third of the increase in total states’ MOE spending on TANF, so CBO attributed $0.4 billion of states’ MOE spending to the surge population in fiscal year 2023.34 The caseload was stable between fiscal year 2023 and the first quarter of fiscal year 2024.35 CBO therefore applied the estimate for fiscal year 2023 to calendar year 2023.

---

1. 1. The CBP data were obtained from two sources: the Transactional Records Access Clearinghouse, which uses the Freedom of Information Act to obtain records from CBP about immigrants’ encounters with CBP officials, and the Department of Homeland Security, which provides information about CBP’s apprehensions of people by their country of citizenship. See Transactional Records Access Clearinghouse, “Stopping ‘Inadmissibles’ at U.S. Ports of Entry” (accessed March 17, 2025), https://tinyurl.com/mr46y5n2; and Customs and Border Protection, “Nationwide Encounters” (accessed March 17, 2025), https://tinyurl.com/49y9keeh.

2. 2. Congressional Budget Office, Effects of the Immigration Surge on the Federal Budget and the Economy (July 2024), www.cbo.gov/publication/60165.

3. 3. For more details about the model, see Congressional Budget Office, An Overview of CBO’s Microsimulation Tax Model (June 2018), www.cbo.gov/publication/54096.

4. 4. A health insurance fund in California accounted for nearly all the other insurance trust revenues.

5. 5. Census Bureau, “Annual Survey of State and Local Government Finances (ALFIN)” (accessed March 17, 2025), https://tinyurl.com/4jap75x8; Bureau of Economic Analysis, “National Data: National Income and Product Accounts” (accessed March 17, 2025), https://tinyurl.com/5h8n87u6.

6. 6. Census Bureau, “Annual Survey of State Government Tax Collections (STC)” (accessed March 17, 2025), https://tinyurl.com/ykhztv4k.

7. 7. For information about the detailed calculator that informed CBO’s estimates, see Jon Bakija, Documentation for a Comprehensive Historical U.S. Federal and State Income Tax Calculator Program, working paper (Williams College Department of Economics, March 2019), https://tinyurl.com/jj5kfj2u.

8. 8. EY, Total State and Local Business Taxes: State-by-State Estimates for FY22 (December 2023), https://tinyurl.com/4srjas6e.

9. 9. Internal Revenue Service, 2023 Instructions for Schedule A: Itemized Deductions (December 2023), www.irs.gov/pub/irs-prior/i1040sca--2023.pdf.

10. 10. The IRS produces a table for local general sales taxes, but CBO does not have enough geographic detail in its model to use that table effectively. Vehicle sales taxes are excluded from the IRS tables.

11. 11. Bureau of Labor Statistics, Consumer Expenditure Survey, 2023, Table 1203, “Income Before Taxes” (September 2024), https://tinyurl.com/3k2bpuyy.

12. 12. Janelle Fritts, Facts and Figures 2022: How Does Your State Compare? (Tax Foundation, March 2022), Table 18, https://tinyurl.com/yc4fc5jy.

13. 13. The Minnesota Department of Revenue’s Tax Research Division reported that, in 2021, two-thirds of general sales taxes paid by businesses were borne by in-state residents. The other one-third were borne by out-of-state residents. See Minnesota Department of Revenue, Tax Research Division, 2024 Minnesota Tax Incidence Study: An Analysis of Minnesota’s Household and Business Taxes (March 2024), Table B-2, https://tinyurl.com/3wta2raz.

14. 14. See, for example, Office of the New York State Comptroller, Tracking the Return: The Tourism Industry in New York City, Report 04-2025 (May 2024), https://tinyurl.com/2kw77vjd.

15. 15. Researchers do not agree on who bears the burden of property taxes on rental properties. See, for example, David J. Schwegman and John Yinger, The Shifting of the Property Tax on Urban Renters: Evidence From New York State’s Homestead Tax Option (Census Bureau, Center for Economic Studies, December 2020), https://tinyurl.com/372sjydr; and Sarah Baker, “Property Tax Pass-Through to Renters: A Quasi-Experimental Approach” (draft, University of California, Berkeley, October 2024), http://tinyurl.com/mr4ca8ax. Other researchers have, like CBO, estimated the burden to be shared equally by renters and owners of rental properties. See Carl Davis, Marco Guzman, and Emma Sifre, Tax Payments by Undocumented Immigrants (Institute on Taxation and Economic Policy, July 2024), https://tinyurl.com/2zfja4zb.

16. 16. National Center for Education Statistics, “ElSi (Elementary/Secondary Information System) Table Generator” (accessed April 21, 2025), https://nces.ed.gov/ccd/elsi/tableGenerator.aspx, and Integrated Postsecondary Education Data System (accessed March 17, 2025), https://tinyurl.com/57adz5nk; and Institute of Museum and Library Services, “IMLS Interactive Map” (accessed April 23, 2025), https://tinyurl.com/2s4uw3r8.

17. 17. For details about the method used to identify other foreign nationals in the CPS ASEC, see Julia Heinzel, Rebecca Heller, and Natalie Tawil, Estimating the Legal Status of Foreign-Born People, Working Paper 2021-02 (Congressional Budget Office, March 2021), www.cbo.gov/publication/57022. In that paper, other foreign nationals are referred to as people without legal status.

18. 18. Students reporting half-time enrollment in high school were discounted by one-half.

19. 19. Changes in both enrollment and spending per student for each state were normalized by dividing the changes at the national level. Enrollment is based on fall enrollment data, and spending is based on current total spending for the school year. The regression also included a set of state fixed effects. Data on enrollment and spending were drawn from National Center for Education Statistics, “ElSi (Elementary/Secondary Information System) Table Generator” (accessed April 21, 2025), https://nces.ed.gov/ccd/elsi/tableGenerator.aspx.

20. 20. Evidence suggests that one student transferred from public to private school for every five to seven immigrants who entered public school. Thomas J. Murray, “Public or Private? The Influence of Immigration on Native Schooling Choices in the United States,” Economics of Education Review, vol. 53 (August 2016), pp. 268–283, https://doi.org/10.1016/j.econedurev.2016.04.003.

21. 21. The weights were based on the share of the surge population residing in each state as estimated using the ACS. Education Commission of the States, “K–12 Funding 2024: English Learner Funding” (March 2024), http://tinyurl.com/46urxpcf.

22. 22. Although states set the allocation formula, most states require local governments to contribute to the instruction costs as calculated by the funding formula. CBO’s estimate does not include estimates of additional local funding for English-­language learners.

23. 23. That estimate is based on an analysis of the six states with the largest shares of the surge population—Florida, Texas, California, New York, New Jersey, and Illinois. New York has a two-year residency requirement, and the other five states have a three-year residency requirement.

24. 24. Despite that simplification, in some states, the immigration and residency requirements for in-state tuition do not align with requirements for state and local financial aid. For example, since 2023, Florida has provided a channel for eligibility for in-state tuition to undocumented students. See H.B. 851 § 5, 2014 Leg., Reg. Sess. (Fla. 2014). However, undocumented immigrants do not qualify for state aid in Florida. See University of Florida, Student Financial Aid and Scholarships, “Undocumented Students” (accessed February 28, 2025), http://tinyurl.com/3bppjbpe.

25. 25. That is one requirement among many. For the additional requirements, see Department of Education, Federal Student Aid, 2024–2025 Federal Student Aid Handbook, Chapter 2, “U.S. Citizenship and Eligible Noncitizens” (January 2024), http://tinyurl.com/4xa7ktdf.

26. 26. For example, to determine eligibility for in-state tuition, the University of Alaska Board of Regents uses the same definition of eligible noncitizen that is used to determine eligibility for federal financial aid. See University of Alaska Board of Regents, Regents’ Policy and University Regulation (accessed March 5, 2025), Chapter 5.10, www.alaska.edu/bor/policy-regulations.

27. 27. For example, without explicitly relying on federal regulations, Oregon grants eligibility for in-state tuition to humanitarian parolees. See Or. Rev. Stat. § 352.287(9)(c) (2023).

28. 28. Qualified aliens can generally participate in Medicaid after a five-year waiting period. That waiting period is waived for some groups, such as asylees, people from Cuba or Haiti, and certain Afghan and Ukrainian parolees. In addition, states can choose to provide full benefits under Medicaid or CHIP to some people. Medicaid covers emergency medical treatment for people who would qualify for full benefits if not for their immigration status or the five-year waiting period.

29. 29. Congressional Budget Office, Effects of the Immigration Surge on the Federal Budget and the Economy (July 2024), www.cbo.gov/publication/60165.

30. 30. Those 11 states are California, Connecticut, Illinois, Maine, Massachusetts, New Jersey, New York, Oregon, Rhode Island, Vermont, and Washington. The age ranges for childhood eligibility vary among states. For example, Maine covers children under 21, whereas New Jersey covers children under 19. CBO accounted for the variation in age eligibility among states for both children and adults. Massachusetts’s and Vermont’s coverage for children is more limited than that provided by other states. (Vermont’s coverage, for example, excludes long-term care.) Nevertheless, children in those states were considered to be age-eligible for state-funded Medicaid-like benefits in CBO’s estimates. See Claire Heyison and Shelby Gonzales, States Are Providing Affordable Health Coverage to People Barred From Certain Health Programs Due to Immigration Status (Center on Budget and Policy Priorities, February 1, 2024), https://tinyurl.com/yxbjxk45.

31. 31. Those four states are California, Illinois, New York, and Oregon. California offered benefits for many adults, but those ages 26 to 49 were ineligible; Illinois offered benefits for adults age 42 or older; New York covered adults age 65 or older; and Oregon and the District of Columbia offered benefits to adults regardless of their age or pregnancy status. See Akash Pillai, Drishti Pillai, and Samantha Artiga, State Health Coverage for Immigrants and Implications for Health Coverage and Care (KFF, May 1, 2024, updated April 10, 2025), http://tinyurl.com/yc454abk. Some of those states’ coverage was extended after 2023. For example, in January 2024, California expanded eligibility to include income-eligible adults ages 26 to 49. In addition, Oregon did not offer comprehensive coverage to adults ages 26 to 54 until July 1, 2023; to account for that midyear expansion, half of the immigrants in the surge population ages 26 to 54 in Oregon were considered to be eligible in 2023. See Oregon Health Authority, “OHP Benefits Available to People of All Ages and Immigration Status,” Oregon Health News Blog (June 28, 2023), https://tinyurl.com/c44ck4b9. Illinois paused enrollment for adults ages 42 to 64 on July 1, 2023. To account for that pause, half of the immigrants in the surge population ages 42 to 64 who arrived in 2023 and resided in Illinois were considered to be eligible that year. Illinois Department of Human Services, “Health Benefit Coverage for Immigrant Adults: Enrollment Pause” (June 28, 2023), https://tinyurl.com/mszy2w2b.

32. 32. Connecticut, Maine, Massachusetts, New York, and Washington provide Medicaid-like benefits to postpartum people; Vermont provides coverage to pregnant people; and Hawaii, New York, and Pennsylvania provide coverage to other limited populations. Hawaii, for example, provides Medicaid-like benefits to people who are age 65 or older, blind, or disabled. For information about the state-funded programs that provide Medicaid-like benefits, see Claire Heyison and Shelby Gonzales, States Are Providing Affordable Health Coverage to People Barred From Certain Health Programs Due to Immigration Status (Center on Budget and Policy Priorities, February 1, 2024), p. 13, https://tinyurl.com/yxbjxk45.

33. 33. The Congressional Research Service provided the tabulation.

34. 34. See Administration for Children and Families, Office of Family Assistance, “TANF Financial Data - FY 2023” (November 7, 2024), https://tinyurl.com/bdh5ycb8, and “TANF Financial Data - FY 2019” (February 22, 2021), https://tinyurl.com/y5mw5hvr.

35. 35. Administration for Children and Families, Office of Family Assistance, “State TANF Data and Reports” (March 20, 2025), https://tinyurl.com/y4wsvfst.

Appendix B: Comparison With Estimates From Other Studies

Most other studies of the effects of immigration on state and local revenues and spending have examined the net fiscal costs of all immigrants, not just newly arrived other foreign nationals.1 Although estimates of the state and local fiscal impact of immigrants vary in the literature, studies using methods similar to those that the Congressional Budget Office used have generally found greater increases in spending than in revenues, a pattern consistent with this report’s estimates of two different measures of the immigration surge’s effects on state and local budgets:

• Direct effects—changes in state and local governments’ revenues and spending that were directly attributable to people in the surge population, and
• Potential effects—the direct effects plus increases in property tax revenues, additional tax revenues from greater economic activity, and nonbudgetary costs associated with greater demand for government services.

Those other studies have typically assigned average costs for the general population to the immigrant population—a measure closer to the potential effects than the direct effects—so CBO compared their estimates with its estimates of the potential effects of the immigration surge.

CBO’s 2007 Review of Relevant Studies

In a report published in 2007, CBO concluded that unauthorized immigrants paid less in state and local taxes than state and local governments spent to provide services to that population.2 CBO also concluded that unauthorized immigrants accounted for only a small share of spending on those services. The focus on unauthorized immigrants in that report differs from the focus on other foreign nationals in this one in that some other foreign nationals are lawfully present. (For a discussion about the categories that CBO used for immigrants in this report, see Box 1.) That earlier report also considered the total population of unauthorized immigrants, whereas this report focuses only on immigrants who arrived from 2021 to 2023.

In the 2007 report, CBO reviewed a number of studies that estimated taxes paid by immigrants, costs incurred by state and local governments to provide various services to immigrants, and certain types of federal assistance for states that provided such services. Given some limitations of those studies—including a lack of information about unauthorized immigrants’ tax compliance and use of benefits—the report cautioned against generalizing the estimates from those studies to the national level.

More Recent Studies

More recent studies on the fiscal impact of immigrants have generally used one of two methods:

• The accounting method, which involves summing taxes and spending for all immigrants in the data, or
• The econometric method, which uses models to estimate the impact of immigration on governments’ budgets.

Because CBO’s analysis of the potential effects follows the accounting method, this discussion is limited to studies that used the first approach.3

Studies relying on the accounting method have yielded mixed results on the impact of first-generation immigrants on state and local budgets. The different results are due in part to different analytical choices. For example, a study by the Cato Institute, which found a positive fiscal impact of first-generation immigrants, allocated spending for all native-born dependents of those immigrants to the second generation.4 That study also included indirect effects on revenues, such as increases in taxable income from additional business profits. By contrast, a study by the National Academies of Sciences, which found a negative fiscal impact of first-generation immigrants, assigned all native- and foreign-born dependents of first-generation parents to the first generation and did not account for indirect effects on revenues.5

CBO’s analysis focuses on the surge in immigration that began in 2021 and its effects on revenues and spending in 2023. Only a small share of the immigrants in the surge population had native-born children by 2023, and none of those children were old enough to incur costs for public education. As a result, excluding immigrants’ native-born children would have had little effect on CBO’s estimates—less of an effect than doing so would have on studies that covered the entire immigrant population—and CBO’s estimates are more comparable to those from studies that excluded such children.6

CBO’s estimates of potential effects include revenues from increases in economic activity due to the surge in immigration. Population growth leads to higher gross domestic product and, therefore, higher income and spending. Those indirect effects on overall economic activity increase tax revenues beyond the amounts directly attributable to immigration. In CBO’s analysis, those indirect effects account for almost half of the potential effect of the surge in immigration on state and local tax revenues.7

Additionally, CBO’s analysis differs from others in that it incorporates spending that was a unique consequence of the surge. For example, CBO’s estimates of state and local spending include spending on shelter and related services in response to the influx of asylum seekers to urban centers; the estimates also include spending on state-funded health insurance and income security programs that provide services to people who are ineligible for federal benefits because of their immigration status.

Although CBO’s methods for estimating the effects of immigrants on tax revenues are broadly similar to those used in other studies, CBO’s estimates of state and local revenues tend to be lower. That difference reflects CBO’s assessment that the surge population had lower income, on average, than the overall immigrant population. CBO also adjusted its estimates to reflect expected differences in the geographic distribution of the surge population and the general U.S. population. The heavy concentration of immigrants in the surge population in Florida and Texas tends to reduce revenues attributable to immigrants because residents of those states have a lower-than-average tax burden.

---

1. 1. One exception is a study that analyzed the fiscal impact of refugees, asylees, and their immediate families. See Robin Ghertner, Suzanne Macartney, and Meredith Dost, The Fiscal Impact of Refugees and Asylees at the Federal, State, and Local Levels From 2005 to 2019 (Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, February 2024), http://tinyurl.com/c5sm692k.

2. 2. Congressional Budget Office, The Impact of Unauthorized Immigrants on the Budgets of State and Local Governments (December 2007), www.cbo.gov/publication/41645.

3. 3. CBO’s estimates of direct effects are more conceptually similar than its estimates of potential effects to estimates produced using the econometric method, which aims to measure actual changes in taxes and spending stemming from immigration with statistical models, such as regressions. For example, CBO’s estimate of the direct cost of primary and secondary education was produced by measuring a statistical relationship between changes in enrollment and changes in spending per student. Because CBO’s estimation methods for direct and potential effects differ for only a few categories of taxes and spending, this report limits the discussion to studies that used the accounting method.

4. 4. Alex Nowrasteh, Sarah Eckhardt, and Michael Howard, The Fiscal Impact of Immigration in the United States (Cato Institute, March 2023), https://tinyurl.com/fd5aa9em.

5. 5. National Academies of Sciences, Engineering, and Medicine, The Economic and Fiscal Consequences of Immigration (National Academies Press, 2017), Chapter 9, pp. 495–566, http://tinyurl.com/3k3ycafj.

6. 6. For example, the Cato Institute study excluded the fiscal impact of native-born children from its estimate of first-generation effects, attributing that impact to the second generation. See Alex Nowrasteh, Sarah Eckhardt, and Michael Howard, The Fiscal Impact of Immigration in the United States (Cato Institute, March 2023), https://tinyurl.com/fd5aa9em.

7. 7. The National Academies of Sciences study did not include indirect effects, but the Cato Institute study did. For an example of another study that included indirect effects on revenues, see Michael A. Clemens, The Fiscal Effect of Immigration: Reducing Bias in Influential Estimates, Working Paper 9464 (CESifo, December 2021), http://tinyurl.com/5yxrrx3k.

About This Document

This report was prepared at the request of the Chairman of the House Budget Committee. In keeping with the Congressional Budget Office’s mandate to provide objective, impartial analysis, the report makes no recommendations.

Nabeel Alsalam (a consultant to CBO) and James Pearce prepared the report with contributions from YeJin Ahn, Rebecca Heller, and Ian Shayne and with guidance from Edward Harris, John McClelland, Xiaotong Niu, and Julie Topoleski. Daniel Crown, Gene Falk (of the Congressional Research Service), Justin Falk, Jennifer Gray, Emma Kugelmass, Justin Latus, Susanne Mehlman, Kim Rueben (a consultant to CBO), Emily Stern, and Carolyn Ugolino contributed to the analysis. Rachel Austin, Jeremy Crimm, Elizabeth Cove Delisle, Sean Dunbar, Kathleen FitzGerald, Kevin Laden, Sam Papenfuss, and Robert Sunshine (a consultant to CBO) offered comments. Daniel Page fact-checked the report.

Analysts from the Arizona Joint Legislative Budget Committee, the California Legislative Analyst’s Office (Ann Hollingshead, Seth Kerstein, and Brian Uhler), the Cato Institute (David J. Bier, Michael Howard, and Alex Nowrasteh), and the Manhattan Institute (Daniel Di Martino and Reihan Salam) commented on CBO’s analytical approach, as did Steven Camarota (of the Center for Immigration Studies), Michael Clemens (of George Mason University), Carl Davis (of the Institute on Taxation and Economic Policy), Joyce Manchester (of Tree House Economics), and Robert Rector (of the Heritage Foundation). The assistance of external reviewers implies no responsibility for the final product; that responsibility rests solely with CBO.

Jeffrey Kling reviewed the report, Christine Browne and Bo Peery edited the report, and Casey Labrack created the graphics and prepared the text for publication. The report is available at www.cbo.gov/publication/61256.

CBO seeks feedback to make its work as useful as possible. Please send comments to communications@cbo.gov.

Phillip L. Swagel

Director

June 2025