6
Do Immigrants Impose a Net Fiscal Burden? Annual Estimates

Introduction

Are immigrants an asset or a liability in the provision and financing of public services in the United States? Judging by the 1996 welfare reform legislation restricting the access of legal and illegal immigrants to a variety of federally funded transfer programs, citizen approval in 1994 of Proposition 187 in California denying funding for public services to illegal immigrants, and recent suits by Arizona, California, Florida, New York, New Jersey, and Texas to recover additional funding from the federal government for immigrant services, many people believe the effect of immigrants is both negative and large. This chapter outlines how the fiscal impacts of immigrants on U.S. citizens in a single fiscal year should be measured and provides estimates of that annual impact for residents of California, New Jersey, and the nation as a whole. Chapter 7 extends this analysis to provide estimates of the long-run or dynamic fiscal effects of new immigrant families. 1

Estimates of the annual and dynamic fiscal consequences of immigration are important for three reasons. First, from the perspective of those wishing to redesign immigration policy, estimates of the annual and lifetime fiscal impacts of new immigrants can be combined with estimates of their effects on the current

1  

Comprehensive surveys of the many studies estimating the fiscal impact of new immigrants on government spending and revenues are Rothman and Espenshade (1992), Vernez and McCarthy (1995, 1996), and MaCurdy et al. (1996).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 254
--> 6 Do Immigrants Impose a Net Fiscal Burden? Annual Estimates Introduction Are immigrants an asset or a liability in the provision and financing of public services in the United States? Judging by the 1996 welfare reform legislation restricting the access of legal and illegal immigrants to a variety of federally funded transfer programs, citizen approval in 1994 of Proposition 187 in California denying funding for public services to illegal immigrants, and recent suits by Arizona, California, Florida, New York, New Jersey, and Texas to recover additional funding from the federal government for immigrant services, many people believe the effect of immigrants is both negative and large. This chapter outlines how the fiscal impacts of immigrants on U.S. citizens in a single fiscal year should be measured and provides estimates of that annual impact for residents of California, New Jersey, and the nation as a whole. Chapter 7 extends this analysis to provide estimates of the long-run or dynamic fiscal effects of new immigrant families. 1 Estimates of the annual and dynamic fiscal consequences of immigration are important for three reasons. First, from the perspective of those wishing to redesign immigration policy, estimates of the annual and lifetime fiscal impacts of new immigrants can be combined with estimates of their effects on the current 1   Comprehensive surveys of the many studies estimating the fiscal impact of new immigrants on government spending and revenues are Rothman and Espenshade (1992), Vernez and McCarthy (1995, 1996), and MaCurdy et al. (1996).

OCR for page 254
--> and future earnings of domestic workers (discussed in Chapter 5) to determine whether they will be net economic contributors to society. Second, because our system of public finance is a federal system and new immigrants tend to concentrate in particular regions of the country, they may create taxpayer inequities. Regions that receive a high concentration of new immigrants may bear relatively higher fiscal burdens in the short run if new immigrants contribute less in revenues—inclusive of federal and, for local governments, state aid—than they receive in additional public services. If so, residents in these regions shoulder a disproportionate share of the nation's annual fiscal burden of immigration. Apart from these tax inequities for native residents, a disproportionate allocation of immigration's fiscal burdens may also induce an inefficient allocation of labor. Facing relatively high tax burdens because of concentrated immigration, workers may leave productive employment in the fiscally disadvantaged region for less productive jobs elsewhere. Third, although estimates of the annual fiscal impacts are important for understanding the economic consequences of immigration for the current year, estimates of how an immigrant family consumes public services and pays taxes over time are also important in order to know the full consequences of admitting additional immigrants into the United States. Almost no family stays just one-year. On one hand, new immigrants, even those receiving a net fiscal transfer from residents in the annual accounting (those with children in school, for example), may ultimately be net contributors to the public-sector over their lifetimes, as they pass into years of productive labor force participation. In the dynamic fiscal accounting, native residents will then be net fiscal beneficiaries. On the other hand, new immigrants who help solve our "annual" funding problem for Social Security and Medicare by increasing the population of payroll taxpayers (young adults, for example), are likely to become recipients of those programs later in life. In that case, an annual fiscal gain for native residents may eventually become a long-run fiscal burden. Furthermore, the long-run burdens or contributions of new immigrants can be reallocated among native residents through the choice of tax and debt policies. Today's burdens can be shifted onto future native residents (and immigrants) through increased government borrowing. Only a dynamic fiscal accounting can reveal these redistributions of fiscal burdens or benefits across generations and the true long-run consequences on native residents of new immigration. Finally, only dynamic fiscal accounting allows us to calculate the effects of new immigrants on the long-run economic sustainability of current fiscal policies. Both annual and dynamic estimates of the fiscal effects of immigration will require decisions along five dimensions. The first four decisions must be made for both annual and dynamic accounting, the fifth for dynamic accounting only. First, what is the appropriate demographic unit of analysis? The choice will depend on whether the study seeks to provide static or dynamic estimates of fiscal burdens. Since the household is the primary unit through which public services

OCR for page 254
--> are consumed and taxes paid, it is the most appropriate unit as a general rule and is recommended for static analysis. Dynamic fiscal accounting using households becomes exceedingly difficult, however, as (often arbitrary) forecasts of family dissolution and formation become necessary. Dynamic studies are best completed by estimating fiscal impacts for individuals. If household estimates are required, hypothetical households can then be constructed directly from the individual fiscal profiles. Second, what is the legal status of new immigrants? Legal status determines what services immigrants can receive and what taxes they will be required to pay. Third, given immigrants' household composition and legal access to public services, what will be the effects of additional immigrants on the costs of providing those services? Care must be taken to specify the appropriate technology of public services. Some services are "public" goods—national defense is the easiest example—and do not require additional service spending to accommodate new residents; other services are more like private goods—for example, income transfers—and new residents must receive the same spending as current residents if service levels for all are to be held constant. In addition, new immigrants may utilize public services at different rates than current residents.2 Fourth, what will be the contribution of new immigrants to public revenues? For certain taxes—income and payroll—immigrant contributions seem clear. Less clear are the property tax contributions of immigrants who own or rent their residences, the corporate tax contributions of immigrants who buy corporate shares, and sales and excise taxes paid by immigrants who remit an unknown portion of their income to their country of origin. Fifth—and unique to dynamic fiscal accounting—what are the future paths of income and population growth for new immigrants and for current residents, and what are the future paths of government spending and tax rates? In addition, dynamic fiscal accounting requires specification of a social rate of discount, so that future tax revenues and spending needs can be compared in terms of current dollars. The fact that different scholars give different answers to these five questions goes a long way to explaining the wide range of estimates now available for the fiscal impacts of new immigration (see MaCurdy et al., 1996). The next section outlines the appropriate methodology for measuring the annual fiscal impact of immigrants on native households.3 Afterward comes a section that applies this methodology to estimate the annual fiscal impact of immigrants in two states with a heavy concentration of immigrants: New Jersey 2   Immigrants may impose a variety of external costs and benefits on society, some of which are not described here. Immigration may be associated with more diverse restaurants in an area, a benefit to some residents that is not directly measured in the fiscal analysis. Some of these social consequences of immigration are discussed in Chapter 8. 3   Attention is limited in this chapter to the fiscal effects of immigration. The effects of immigration on labor market aspects of economic welfare of native residents are addressed in Chapters 4 and 5.

OCR for page 254
--> and California. In addition, estimates of the annual fiscal impact of immigrants on the federal budget are provided. The chapter's final section summarizes the results. We emphasize here, and again in the conclusion, that any estimates of the annual fiscal impact of new immigrants, our own included, must be used with care. Annual estimates provide only a one-year snapshot of the how immigrant-headed households contribute revenues and withdraw resources from the public treasury. Such numbers do not tell us what the final fiscal impacts of a change in national immigration policy will be. To estimate the full fiscal impact of new immigrants, we need to know how those immigrants and their descendants will use services and pay taxes over their lifetimes while in the United States. Those estimates can come only from a dynamic fiscal analysis, the task of Chapter 7. The static analysis presented here provides the building blocks for the dynamic analysis, however, much as snapshots provide the images necessary for a moving picture. Estimating the Annual Fiscal Impact of New Immigrants The annual fiscal impact of immigrants on the economic well-being of native residents can be estimated in two steps.4 Step one approximates the effect of the added tax burden imposed on native residents to fund the current level of services received by natives and now extended to immigrants. The reader should note that this analysis is carried out under the assumption that there are no behavioral changes in response to changes in the number of immigrants—it is what is known as a "partial-equilibrium" exercise. Step two shows how this added tax burden can be estimated from published data on government spending, taxes, family incomes, and family program participation. Fiscal Accounting: Step One Before the admission of immigrants, native residents enjoy consumption of government-provided public services (denoted by gN) and after-tax income (denoted as yN) that together determine their economic well-being. Higher levels of public services and more after-tax income improve the well-being of current residents, and reductions in services and after-tax income reduce it. Immigrants may affect the level of services provided to native residents, the level of their taxes and thus their after-tax incomes, or both. The final fiscal effects of immigrants on a typical native resident's economic welfare can be approximated by 4   The methodology outlined here is a general approach—known as economic-incidence analysis—and can be applied to estimate the fiscal effects of any group on another.

OCR for page 254
--> the following monetary measure of the net annual fiscal impact of immigration on native residents (denoted as NAFIN): (1) where ΔgN is the change in public services received by the native household because new immigrants now share these services, µN is the economic value (measured in dollars) the native resident places on these public services, and ΔTN is the change in taxes paid by the native resident because of new immigration. If NAFIN is positive, then the fiscal effects of immigration make the typical native resident better off economically. If NAFIN is negative, then the fiscal effects of immigration make the native resident worse off economically.5 5   NAFIN provides a first-order approximation of the change in native residents' economic welfare when new immigrants affect the public services that natives receive and the taxes that natives pay. This specification of the fiscal effects of immigration on native residents' economic well-being is part of a more general analysis, one that combines the labor market effects of immigration with the fiscal effects. If UN = U(yN, gN) represents the economic welfare of a native resident from public services (gN) and after-tax income (yN) then a small change in the number of immigrants (denoted as ΔM) affects native resident welfare through changes in yN and gN as: (i) The monetary equivalent of ∂UN is obtained by dividing through the change in utility by the marginal utility of an extra dollar of income (∂UN/∂yN) to a native resident. That is: (ii) where µN is now defined to equal (∂UN/∂gN)/(∂UN/∂yN) and represents the native resident's willingness to pay (= ΔyN/ΔgN) for public services. Changes in a native resident's after-tax income come from two effects: the effects of new immigrants on resident incomes and the effects of new immigrants on resident taxes. Defining after-tax income (yN ) as pretax income (IN) minus taxes paid (TN) implies: (iii) Changes in a native resident's after-tax income (ΔyN) equal the change due to new immigrants' effects on the native resident's pre-tax earnings (ΔIN) minus the change due to new immigrants' effects on a native resident's tax payments (ΔTN). Substituting (iii) into (ii) and defining ΔgN = (∂gN/∂M)ΔM now specifies the first-order approximation to the final effect of new immigrants on a native resident's economic welfare as: (iv)

OCR for page 254
--> Most studies of the fiscal impact of new immigrants make one important assumption, at least implicitly, before using NAFIN as a measure of the change in current residents' well-being. The studies typically assume that all governments respond to the flow of new immigrants by holding fixed the level of government services provided to current residents—in other words, they assume government policies are adjusted so that ΔgN ≡ 0. If immigrants share government facilities and resources with native residents, then new facilities may be needed to provide services to the immigrants and to protect the level of services now provided to native residents. If so, government spending must increase and government taxes must rise. If native residents help pay for this increase in spending, then their taxes must rise. Using this approach implicitly values all changes in government services at their input costs. All fiscal impacts are now felt through changes in native residents' tax payments. With this additional assumption that all changes are made through taxes, the monetary measure for the overall fiscal effects of new immigration on current residents reduces to: (2) where it must be understood that government spending and residents' taxes are adjusted to protect the level of services now provided to current residents (i.e., ΔgN = 0).6 If ΔTN > 0 and native taxes rise, then NAFIN < 0 and native residents are worse off for fiscal reasons following immigration. Conversely, if ΔTN < 0 and native residents receive a net fiscal transfer from the new immigrants, then their taxes fall, NAFIN > 0, and native residents are better off for fiscal reasons following immigration.     which can be rewritten more simply as: where ΔV is the monetary value to a native resident of the change in utility caused by new immigration, ΔIN is the change in pretax incomes, and NAFIN is the net annual fiscal impact of new immigrants defined as the effect of new immigrants on public services received by native residents (µNΔgN) minus the effect of new immigrants on native resident taxes (ΔTN). Chapters 4 and 5 have provided estimates of ΔIN. This chapter provides estimates of NAFIN. 6   The analysis assumes that changing immigration has no effect on the fiscal policies of state and local governments. On the expenditure side of the equation, the analysis assumes that public goods and services are perfectly elastically supplied (with constant marginal and average costs) by state and local governments. This implies that an additional immigrant (or native) household residing in a state or locality has no effect other than as a scale multiplier on the level of public service provision. The analysis also implicitly assumes that the components of state and local revenue do not change as a result of immigration (or in-migration of native-headed families) to various localities. The assumption of exogenous fiscal policies provides useful short-run estimates for state and local government effects. Future work in this area could examine how much immigration affects fiscal policies at various levels of government and incorporate such endogenous effects into the modeling exercise.

OCR for page 254
--> Fiscal Accounting: Step Two The annual fiscal impact of new immigrants on the tax payments of current native residents—ΔTN—can be estimated from the government budget constraint that equates government spending to government revenues. Government revenues come from taxes on current native residents measured as the average tax per native resident times the number of current native residents (= TN ∙ N); from taxes on current immigrant residents measured as the average tax per immigrant times the number of current immigrants (= TM ∙ M); and from nonresident revenues such as government borrowing, intergovernmental transfers, proceeds from previously accumulated government wealth, and business taxes not already allocated to resident households equal to AN per native resident times the number of residents (= AN ∙ N). Similarly, government spending is allocated to provide services to native residents at a cost of EN per resident times the number of residents (= EN ∙ N); to previous immigrants at a cost of EM per immigrant times the number of immigrants (= EM ∙ M); and to businesses and nonresidents, including holders of current government debt raised to finance past deficits at a cost of XN per native resident times the number of native residents (= XN N). The government's cash-flow balance requires revenues in any year to equal spending in that year: (3) or alternatively, (4) Any changes in the budget must also show a cash-flow balance. If we assume that the number of native residents does not change following immigration and that government spending allocated to native residents remains fixed at current levels, so that service flows to those residents remain constant too, then changes in the public budget following an inflow of ΔM new immigrants becomes:7 7   The full effect of new immigration on the cash-flow balance is represented by: and this becomes the expression in the text when ΔN ≡ 0 and ΔEN ≡ 0. Note that this decomposition ignores second-order terms (such as ΔM ∙ [ΔTM - ΔEM]) because they will be small relative to the first-order terms.

OCR for page 254
--> (5) The effect on native residents' taxes of all budget changes that follow from immigration can therefore be specified as:8 (6) This change in native taxes measures the stock of dollars needed to protect the current flow of public services provided to native households following immigration minus any new revenues generated by the new immigrants themselves. This difference, ΔTN, is what the average native household must pay to protect government services after immigration and - ΔTN is the net annual fiscal impact (NAFIN ) of this tax change. If ΔTN > 0 and NAFIN < 0, then new immigrants impose a net fiscal burden on native residents. If ΔTN < 0 and NAFI N > 0, then new immigrants offer a net fiscal benefit to native residents. Finally, if ΔTN = 0, then the new immigrants are ''fiscally neutral" for native residents.9 This budget accounting clarifies the three possible avenues through which new immigrants can change a native resident's tax burden. First, if new immigrants receive the same level of services and pay the same level of taxes as previous immigrants, and spending exceeds (or is less than) taxes, then there is new tax burden (or relief) equal to the net fiscal balance of each immigrant (= TM 8   This impact need not be borne only by current native residents. If native residents share the burden of new immigrants with previous immigrants (M), then the measure of the fiscal effects of new immigration on native residents will be smaller and measured by: The discussion that follows is valid for this adjusted measure of - ΔTN as well. 9   The measure of - ΔTN specified here is for a single government. The analysis can be extended to a federalist public system by measuring - ΔTN for each level of government, denoted as - ΔTNL for the local fiscal burden to be paid by local native residents in the jurisdiction receiving the new immigrants, by - ΔTNS for the state fiscal burden to be paid by native state residents in the state receiving the new immigrants, and by - ΔTNF for the federal fiscal burden to be paid by all native residents throughout the country. The total fiscal impact of new immigrants will be: where: for i = L, S, and F.

OCR for page 254
--> - EM) times the number of new immigrants (ΔM). This aggregate fiscal balance is then shared among all native residents (N). Second, if the new immigrants are poorer or less well educated than previous immigrants, then their utilization of public services—particularly, redistributive services and perhaps public education—may well be greater than utilization by previous immigrants. Furthermore, their tax contributions are likely to be lower. If so, then ΔEM (the expenditures on immigrants) > 0 and ΔTM (the taxes immigrants pay) < 0, and there will be a need for an upward fiscal adjustment to the initial fiscal balance estimate of (TM - EM). Of course, if new immigrants are richer and better educated than previous immigrants, then average spending on immigrants may fall (ΔEM < 0) and average immigrant taxes may rise (ΔTM > 0), reducing the average tax burden or increasing the average level of tax relief for native residents. Current spending per immigrant may also decline with new immigration, even if the same level of immigrant services is provided, if there are significant economies of scale in the provision of public services to immigrants. For example, if a public service is a pure public good, such that the same level of services can continue to be supplied without an increase in spending even in face of new immigration, then average spending per immigrant following new immigration should be reduced (ΔEM < 0) to reflect this savings. This savings is assigned to the current native resident as a reduction in the tax burden imposed on them by new immigration. These adjustments to the average fiscal impact of immigration caused by admitting new immigrants (ΔTM - ΔE M) are then multiplied by all previous immigrants (M), and this total adjustment is then shared by native residents (N). Finally, new immigrants may mean changes in nonresident revenues (ΔAN), or nonresident spending obligations per native resident (ΔX N), or both. The most likely direct source of new nonresident monies (ΔAN > 0) for state and local governments is additional federal aid to states and additional federal and state aid to localities when new immigrants qualify for federally supported public programs. Welfare programs at the state level and education at the local level are the most important examples. New immigrants may also have a positive effect on business tax revenues (ΔAN > 0) at all levels of government if workers' wages are lowered and returns to capital or to land increase and those higher business profits are then taxed. Offsetting these potential gains is the fact that new immigrants will share the proceeds from the stock of existing national public wealth.10 Finally, nonresident spending at the state or local level may change (ΔXN) if new immigrants affect the ability of government to provide services to the business sector, requiring compensating increases in government spending so as to hold 10   For estimates of the value of federal government holdings of land and mineral wealth, see Boskin et al. (1985). See note 19 below.

OCR for page 254
--> business services fixed (ΔXN > 0). To the extent that the increase in nonresident revenues is larger than any increase in nonresident spending, the additional tax burden of new immigration on the average native resident will decline. Fiscal Accounting: Implementation Steps one and two detail how to measure, in principle, the annual fiscal impact of new immigrants on current residents. Turning theory into practice requires data that first measure the number of current native residents (N), previous immigrants (M), and new immigrants (ΔM); average spending for (EM) and revenues from (TM) previous immigrants; the adjustments to average immigrant expenditures (ΔEM) and revenues (ΔTM) needed to reflect attributes of the new immigrant population; and changes in nonresident expenditures (ΔXN) and revenues (ΔAN) because of new immigration. There are guiding principles for estimating each of these needed quantities. Measuring Population and Population Changes Since most government programs will be decided for households, it is appropriate to measure N, M, and ΔM as the number of households in a given governmental jurisdiction.11 The decennial census defines a household as family members living in the same home.12 Furthermore, households should be distinguished by their legal status to receive government services and to pay taxes.13 11   The calculation of spending and revenue includes all persons residing in the state on the date of the study. By definition, the static analysis is based on period estimates. All persons residing in the state, including native-born and foreign-born persons who moved into the state from another state, are included in the analysis. The static fiscal analysis takes into account interstate migrants, and the estimates are influenced by the number and characteristics of these migrants as well as of residents who were born in the state. 12   The fiscal implications of this census definition should be appreciated when estimating the fiscal impact of previous immigrant households on native residents. For example, a previous immigrant couple with four U.S.-born children may, in the year in which fiscal effects are measured, have only two of those children living at home. The other two older children may now be in the workforce. The two older children will be counted in the census definition as native (not foreign-born) residents. In estimating the fiscal costs of previous immigration policies, it would be an error to count taxes and spending for the two immigrant parents and the two children at home and not count the taxes and spending of the two older children now in the workforce. It is most likely that these older children are contributing more in taxes than they are withdrawing in spending (particularly if their own children are not yet born). These positive contributions should be balanced against the net fiscal impacts of the parents and the two children still at home. Only a dynamic fiscal analysis can correctly account for these life-cycle changes; see Chapter 7. 13   Available census and survey data do not distinguish the legal status of foreign-born respondents. The participation of illegal immigrants in the census and surveys, however, affects the reported public welfare use because they are ineligible for such programs.

OCR for page 254
--> Measuring Spending and Spending Changes Spending per immigrant on government services (EM) equals the sum of all spending needed to provide the current level of public services or government assistance to previous immigrant households now provided divided by the previous immigrant population. Spending must include the wages and benefits of public employees, the costs of materials and supplies, the payment of land rents, and the depreciation of public capital.14 If the provided service is a transfer payment, then the costs of that service should equal the transfer paid plus administrative costs required to verify and service the recipient household. The average level of spending per immigrant can change with new immigrants for one or more of four reasons: (1) New immigrant households are provided with lower (ΔEM < 0) or higher (ΔEM > 0) levels of services than previous immigrant households;15 (2) new immigrant households permit the use of a more (ΔEM < 0) or less (ΔEM > 0) productive technology for producing services;16 (3) new immigrant households use the public facilities less (ΔEM < 0) or more (ΔEM > 0) intensely than previous immigrant households;17 or (4) new immigrant households may permit economies of scale through the sharing of public facilities with previous immigrants allowing a reduction (ΔEM < 0) in overall spending per immigrant. Most studies of the fiscal costs of immigration assume that new immigrants are provided with services similar to those for previous immigrants, benefit from an equally productive service technology, and consume those services at the same rate as previous immigrants. These assumptions are reasonable provided care is taken to estimate the average spending on new immigrants from spending on previous immigrants with comparable incomes and family demographics. Even with matching demographic cohorts, however, there may remain economies of scale in public service provisions. If so, then ΔEM must be estimated, allowing for potential savings because services can be shared. Average spending declines by: 14   All of these costs are included in most government balance sheets, except perhaps land rents and the depreciation costs of capital used to provide services. One possible measure of land rents and depreciation costs is current-period principal and interest repayments for infrastructure debt. Most important, principal and interest payments to cover prior fiscal deficits should not be included as part of the estimate of the expenditures necessary to provide public services today to new immigrants and residents. These past deficit borrowings did not support capital and land acquisition and therefore have no relationship to the costs of providing current-period public services. New immigrant taxes, however, do contribute to these deficit repayments. 15   For example, the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 restricts immigrant access to federal redistributive programs. 16   For evidence of the costs of using different production technologies when providing immigrant services in education and welfare, see Clark (1994). 17   For evidence that new immigrants may use welfare services more intensely than current immigrants, see Borjas and Trejo (1991) and Borjas and Hilton (1996).

OCR for page 254
--> Immigrant households in New Jersey impose a total fiscal burden on native residents of -$232 per household through the state and local sectors, about four-tenths of 1 percent of a typical native family's income (.0037 = $232/$61,966; see Table 6.1). From Table 6.3, an average immigrant household in California is shown to have a negative average fiscal balance (TM - EM) of -$831 per household at the local level and -$2,632 per household at the state level. Again, treating all current immigrant-headed households as new immigrants (ΔM = M) requires multiplying these estimates of (TM - EM) by the ratio of current immigrants to native households (=M/N = 2.851m/8.385m = .34) to estimate the NAFIN of California immigrants on native residents. For California's local sector, NAFIN equals -$283 per native household, and for California's state sector NAFI N equals -$895 per native household) (see Table 6.5, panel A). The total fiscal burden imposed on California native residents through the state and local public-sectors because of immigrant-headed households is -$1,178 per household, or about 2.3 percent of a typical native household's annual income (.023 = $1,178/$50,518; see Table 6.1). Immigrant households do make a positive contribution to the fiscal position of native households through the federal budget, however. Table 6.4 allows us to estimate the NAFIN from the federal budget for residents of New Jersey and California because of the presence of current immigrants in their states. First, the average federal fiscal balance of (TM - EM) -$2,682 per immigrant household reported in Table 6.4 is adjusted for the fact that national defense is a pure public good; defense spending of $2,809 per immigrant household should not be included in new expenditures and is therefore added to the negative fiscal balance ($127 = -$2,682 + $2,809). Second, the resulting $127 per immigrant household is multiplied by the number of immigrant households in California (= M = 2.851m) and then divided by the national native household population (= N = 89.018m), since all native households in the country share the federal contributions and costs of California immigrants. The resulting estimate of NAFI N from the federal budget for all California immigrant households on California native households is $4 per native household (= $127 x (2.851m/89.018m); see Table 6.5, panel A). The estimates in Table 6.4 for how California immigrants contribute to the national treasury can be used to approximate the contributions of New Jersey immigrants, assuming that they contribute in the same way as their country-of-origin counterparts in California. If so, then the net fiscal contribution of a New Jersey immigrant family, adjusted for defense spending, equals $520 per immigrant household. 41 Again, this contribution should be aggregated over all New 41   This estimate follows from applying the New Jersey distribution of immigrant households by country of origin (Table 6.1) to the estimated average fiscal balance adjusted for defense spending for each immigrant cohort in Table 6.4: $520 per immigrant household = (.063/.135) x (-$2,486 + $2,809) per European-Canadian household) + (.027/.135) x ($863 + $2,809) per Asian household) +

OCR for page 254
--> Jersey immigrant families and then shared by all U.S. native households. In the end, a New Jersey native household benefits from a positive NAFIN of about $3 (see Table 6.5, panel A).42 Combining the local, state, and federal estimates of the net fiscal burdens imposed by immigrant households on native residents in New Jersey and California shows that the average native household bears an overall fiscal burden of $229 in New Jersey and $1,174 in California. The fiscal burden borne by Californians is larger primarily because of the larger burdens imposed at the local and particularly the state levels. Causing these differences are the larger ratio of immigrant to native households in California,43 California's large share of immigrants in the economically poorer Latin American cohort,44 and the relatively more generous welfare programs available to California residents.45 The analysis in Tables 6.2, 6.3, and 6.4 also allows estimates of the aggregate net fiscal impact on all U.S. native households of the current national immigrant population (Table 6.5, panel B). Two estimates are presented here, using the local and state budgets of New Jersey and California to estimate the net annual fiscal impact of immigrants through the state and local sectors. Because California has relatively more generous welfare programs, the national estimate using     (.038/.135) x (-$5,038 + $2,809) per Latin American household) + (.007/.135) x ($2,245 + $2,809) per "other" immigrant household). 42   This share of the net fiscal contribution of New Jersey immigrant households is estimated as (net fiscal contribution/New Jersey immigrant household) ¥ (number of immigrant households in New Jersey) ÷ (number of U.S. native households) = ($520) ¥ (.439 m) ÷ (89.018 m) = $2.56/U.S. native household, including California native families. To be consistent with the California estimates, the 1994/95 New Jersey immigrant household population of 438,738 was used. 43   From Table 6.1, the ratio of immigrant to native households in California is .34 (= 2.851m/ 8.385m). In New Jersey the ratio of immigrant to native households is .16 (= .392m/2.503m). 44   In both California and New Jersey, immigrants from Latin America impose the largest net fiscal burdens on native households. Among the immigrants in California, more than half (= .56 = 1.592m/ 2.851m; Table 6.1) come from Latin America; in New Jersey the share of all immigrants from Latin America is about one-quarter (= .28 = . 111m/.392m; Table 6.1). 45   Comparing "transfers to households" in Tables 6.2 and 6.3 reveals that California immigrant households in each country-of-origin cohort receive from two to four times the transfers per household in New Jersey. More generally, the aggregate sizes of the per-household local-plus-state-sector transfers in California and New Jersey are $8,657 per household (= $5,523 per household + $3,134 per household; Table 6.3) and $5,856 per household (= $3,141 per household + $2,715 per household; Table 6.2), respectively. The state and local public sector is 48 percent larger in California than in New Jersey ($8,657/$5,856 = 1.48). Comparing the net fiscal contribution/transfer to immigrants by country of origin for California and New Jersey shows the transfer to Latin Americans in California is a 47 percent larger net fiscal benefit than that to their counterparts in New Jersey (-$4,977/ -$3,396 = 1.47) and the transfer to Asians in California is 57 percent larger than to Asians in New Jersey (-$2,591/-$1,650 = 1.57). "Other" immigrant households in the two states receive about equal net fiscal transfers, but they are only a small percentage of all immigrant households. The California fiscal disadvantage from the deficit immigrant groups is offset somewhat by the larger net contribution from the Europe/Canada immigrants, presumably because of more progressive state taxation in California.

OCR for page 254
--> the California budget can be viewed as an upper-bound estimate of the national net annual fiscal impact. Table 6.5, panel B, reports the total (local plus state plus federal) net annual fiscal impact (NAFIN) imposed by each immigrant cohort for both the New Jersey and California specifications of the state and local budgets.46 In fact, except for immigrant households from Latin America, today's current immigrants are net fiscal contributors to the overall fiscal position of native U.S. households, primarily because of their large positive net contributions to the federal treasury to help pay for defense spending. The Latin American immigrant cohort is large, however, and their negative net annual fiscal impact more than offsets the aggregate fiscal contributions paid to natives by Canadians, Europeans, Asians, and other immigrants. Weighting each cohort's contribution by its share in the national immigrant population provides an estimate of the net annual fiscal impact imposed by a national average immigrant receiving either the New Jersey or the California state and local budgets. The national average immigrant imposes a net annual fiscal impact of -$1,613 per immigrant household when receiving the New Jersey budget and a net fiscal burden of -$2,206 per immigrant household when receiving the California budget. The aggregate NAFIN on native residents for all U.S. immigrants is estimated by multiplying these per-immigrant burdens by the number of immigrant households in the nation as whole. In 1994-95, there were 9,156,000 immigrant-headed households in the United States.47 The aggregate net annual fiscal impact imposed on native households by all immigrant-headed households in the United States is therefore estimated to range from -$14.77 billion (New Jersey budgets) to perhaps as high as -$20.16 billion (California budgets) (see Table 6.5, panel B). Sharing this aggregate burden over all 89,019,000 native households in the United States in 1994-95 would imply a net annual fiscal impact per native household ranging from -$166 (New Jersey budgets) to perhaps as high as -$226 (California budgets). This is an annual fiscal burden imposed on a typical native U.S. household by the current stock of immigrant-headed households now in the United States. The burden ranges from about four-tenths of 1 percent to half of 1 percent of the average household income of $45,000 in 1996.48 46   For example, the total average fiscal balance for immigrants from Europe/Canada living in New Jersey is the sum of Table 6.2's estimates of that cohort's local and state net fiscal transfer plus the Table 6.4 estimate of that cohort's federal average fiscal balance adjusted for removing defense spending: -$124 + $250 + [-$2,486 + $2,809] = $449. The total average fiscal balance for immigrants from Europe/Canada living in California is the sum of Table 6.3's estimates of that cohort's local and state net fiscal transfer plus the Table 6.4 estimate of that cohort's federal net fiscal transfers: $548 + $760 + [-$2,486 + $2,809] = $1,631. All other cells in Table 6.5, panel B, are calculated in a similar way. 47   Estimates calculated from the 1994-95 Current Population Survey. 48   Average household income is estimated as the average for the most recent year available (= $41,428 in 1993) adjusted for inflation.

OCR for page 254
--> It is important to stress what the NAFIN estimates here represent. They are estimates of the annual fiscal burdens imposed on native households by current immigrant-headed households in the early 1990s. They are not estimates of the annual costs we could expect in all future years from admitting new immigrant families, and they are not estimates of the annual fiscal costs today of past immigration policies. To estimate the future fiscal costs, or benefits, of immigration, one must allow for today's fiscally costly young immigrants to leave school, take jobs, and contribute taxes; this requires a dynamic fiscal accounting. To estimate the annual cost today of past immigration policies also requires a dynamic analysis, one that looks back in time. Children born to immigrants in the United States who are now living on their own and earning incomes must be included in this historical evaluation. These contributing, second-generation children are here because of past immigration policies; they are not, however, included in the average fiscal balance for immigrant-headed households calculated here, because these children no longer live at home. Again, only a dynamic analysis can accurately account for the contributions of these individuals. Chapter 7 outlines one approach to dynamic fiscal accounting for immigration. The Net Annual Fiscal Impact of New Immigrants Although it is perhaps interesting to know the annual fiscal burden of current immigrant-headed households on U.S. native households, more relevant for contemporary policy debates over immigration policy are estimates of the net fiscal burdens of new immigrants on native U.S. households. Tables 6.6 and 6.7 offer estimates of the net annual fiscal impact of new immigrants on native families living in New Jersey (Table 6.6 ) and California (Table 6.7), under the assumption that new immigrant families are just like the current stock of immigrants now residing in the state. The estimates reported in Tables 6.6 and 6.7 answer this question: What would be the fiscal burden on native residents if the number of immigrant-headed households now in the state were to exactly double? The net fiscal impact imposed by a doubling of the number of immigrant-headed households is simply equal to the net annual fiscal impact from current immigrant-headed households, now shared by all households in the state or nation (see note 6). Thus, the net fiscal impact of doubling "all immigrants" will be slightly smaller than the burden of previous immigrants because the tax base over which to share the costs of the new immigrants will be larger. Whereas, as we've said, the current stock of immigrant households imposes a NAFIN on native households of about -$229 per household in New Jersey and -$1,174 per household in California, doubling "all immigrants" will impose a net annual fiscal impact on native New Jersey households of $199 per household (Table 6.6) or

OCR for page 254
--> TABLE 6.6 Net Annual Fiscal Impact (NAFI) Imposed by Additional Immigrant-Headed Households on Native Residents: New Jersey (1996 dollars per current New Jersey household) Immigrant Group Local NAFI State NAFI Federal NAFI Total NAFI All immigrants -$125 -$76 $2 -$199 Contribution by age of household head < 65 -$150 -$74 $12 -$212 65+ $25 -$2 -$10 $13 Contribution by region of origin Europe/Canada -$9 $18 $1 $10 Asia -$42 -$6 $3 -$45 Latin America -$65 -$81 -$3 -$149 Other -$9 -$7 $1 -$15 Note: Estimates of the NAFI imposed by New Jersey immigrants through the federal budget are based on Clune's estimates of the average federal fiscal balance by a typical immigrant household living in California in each age group and from each region of origin; see Table 6.4. The per immigrant household estimates were then weighted by the New Jersey immigrant population shares to obtain the federal NAFI estimates reported above. Other = Africa and Oceania. Source: Calculations based on Tables 6.2 and 6.4, adjusted for the share of native households in the New Jersey population (= .865) for local and state NAFIs and in the national population (= .0045) for the federal NAFI. about three-tenths of 1 percent of the state's average household income (.03 = $199/$66,371) and will impose a NAFIN on native California households of -$876 per household (Table 6.7), or about 1.8 percent of the state's average household income (.018 = $876/$48,347). Tables 6.6 and 6.7 also decompose the net annual fiscal impacts from doubling "all immigrants" into the contributions made by doubling each of various subgroups of immigrant-headed households. For example, doubling all immigrant households doubles the number whose heads are under age 65 and the number whose heads are over age 65. What contributions do these two age groups make to the "all immigrants" totals? In both New Jersey and California, older immigrant households make a positive fiscal contribution at the local government level, paying more in taxes than they use in services (Tables 6.6 and 6.7). These positive contributions partially offset the negative net annual fiscal impact imposed by immigrant-headed households whose heads are younger than 65. At the state level, both age groups contribute to the negative NAFIN imposed on

OCR for page 254
--> TABLE 6.7 Net Annual Fiscal Impact (NAFI) Imposed by Additional Immigrant-Headed Households on Native Residents: California (1996 dollars per current California household) Immigrant Group Local NAFI State NAFI Federal NAFI Total NAFI All immigrants -$211 -$669 $4 -$876 Contribution by age of household head 8 65 -$221 -$635 $77 -$779 65+ $10 -$34 -$73 -$97 Contribution by region of origin Europe/Canada $17 $24 $2 $43 Asia -$30 -$141 $26 -$145 Latin America -$189 -$528 -$36 -$753 Other -$9 -$24 $12 -$21 Note: Other = Africa and Oceania. Source: Calculations based on Tables 6.3 and 6.4, adjusted for the share of native households in the California population (= .746) for local and state NAFIs and in the national population (= .029) for the federal NAFI. natives by immigrant households. Younger households, however, are responsible for 98 percent (.98 = -$74/-$76; Table 6.6) of the "all immigrant" total in New Jersey (.98 = -$74/-$76; Table 6.6), and for 95 percent of the total in California (.95 = -$635/-$669; Table 6.7). The reason, of course, is the larger number of younger immigrants. The relative contributions of the two age groups to the federal net annual fiscal impact shows that the younger immigrant groups make a net contribution in both New Jersey and California; older immigrants impose a fiscal burden on the federal budget. Tables 6.6 and 6.7 also report the results of a decomposition by country of origin of the net annual fiscal impact from doubling immigrant-headed households. New immigrants matching the economic and demographic attributes of current immigrants from Europe/Canada will be overall net fiscal contributors (NAFIN > 0). New immigrants matching the attributes of current immigrants from Asia, Latin America, and "other" regions will impose a net fiscal burden on natives (NAFIN < 0). The single biggest group contribution to new net fiscal burdens—75 percent in New Jersey (.75 = -$149/-$199; Table 6.6) and 86 percent in California (.86 = -$753/-$876; Table 6.7)—will be by new immigrants matching the economic and demographic attributes and numbers of current Latin American immigrants.

OCR for page 254
--> Finally, the net annual fiscal impact imposed on an average native household nationally from a doubling of the current immigrant population can be estimated. If all states use the New Jersey state and local budget, then the aggregate NAFIN from a doubling of today's immigrant-headed households will be -$14.77 billion (see Table 6.5, panel B). If all states used the California state and local budget, then the aggregate NAFIN from doubling today's immigrant household population will be -$20.16 billion. These new fiscal burdens will be shared by all 98.175 million current native and immigrant households, implying an additional net fiscal cost per household ranging from $150 per household (New Jersey budgets) to $205 per household (California budgets), or a burden of between three-tenths and four-tenths of 1 percent of current national household income. These estimates of the net annual fiscal impact of doubling the current immigrant-headed household population can be adjusted to reflect the fiscal burdens on native residents of more modest population changes. For example, what will be the net annual fiscal impact on natives of a 5 percent increase in the current immigrant household population—an increase of about 460,000 new households? Such an increase represents an approximate continuation of current immigration policies. Such an increase will raise next year's net fiscal burden of immigrant-headed households by about $10 per household in New Jersey (-$9.95 = .05∙ -$199 per household; see Table 6.6) and by about $45 per household in California (-$43.80 = .05 ∙ -$876 per household; see Table 6.7). Nationally, a 5 percent increase in today's immigrant population will lead to an increase in next year's net fiscal burden for all U.S. native households of from $7.50 per household (-$7.50 = .05 ∙ -$150 per household for New Jersey budgets) to $10 per household (-$10.25 = .05 ∙ -$205 per household for California budgets). Conclusions On the basis of this analysis of the annual net fiscal burden of immigrant households on native families, the panel reaches six conclusions: The state and local net annual fiscal impact of current immigrant-headed households on native residents measured as the difference between the costs of state and local services received and state and local taxes paid for New Jersey residents (for fiscal year 1989-90 adjusted to 1996 dollars) is estimated at $232 per native household. Similarly constructed estimates for the net annual fiscal impact of current immigrant-headed households on native residents for California residents (for fiscal year 1994-95 adjusted to 1996 dollars) is estimated at $1,178 per native household. There are three central causes for the negative fiscal impact of immigrants on native residents at the state and local levels: (1) immigrant-headed households have more children than native households on average and therefore consume

OCR for page 254
--> more educational services, (2) immigrant-headed households are poorer than native households on average and therefore receive more state and locally funded income transfers, and (3) immigrant-headed households have lower incomes than native households on average and thus pay lower state and local taxes. There are, however, important variations within the immigrant population in the size of the net annual fiscal impact imposed on native residents. Current immigrants over the age of 65 are net fiscal contributors to native residents in New Jersey, but they are a small net fiscal burden on native residents in California. However, almost all of the fiscal burdens imposed on native households by immigrants come from immigrant households whose head is younger than 65. Among the younger immigrant households, the net burden imposed on natives is by far the greatest for those immigrants from Latin American countries. In fact, immigrant households of European or Canadian origin make a net fiscal contribution to the native households of New Jersey and California through the state and local budgets in those states. On average, immigrant-headed households in California made a net fiscal contribution to the federal budget in fiscal year 1994-95, receiving less in services and transfers than they paid in taxes. The positive fiscal impact of immigrant households at the federal level arises because they are assumed to impose no additional burden on the federal budget for national defense, specified here as a ''pure" public good. The one exception to this pattern is the immigrant-headed households from Latin America; those households were a fiscal burden even at the federal level. New Jersey and California are both states with high immigration, and as a consequence, the net annual fiscal impact of immigrant-headed households on native residents in those states—particularly from the services provided through state and local governments—are high as well. If these net fiscal burdens from immigrant households were shared not just within the states but nationally, then the burden per native resident would fall significantly. Estimates of the net fiscal burden imposed on all 89,019,000 U.S. native households by all 9,156,000 U.S. immigrant-headed households through all levels of government range from $166 per native household to $226 per native household. The lower estimate gives all U.S. immigrant households the New Jersey state and local budget; the higher estimate gives all U.S. immigrant households the California state and local budget. A decision to admit 460,000 new immigrant-headed households—assuming those households match the economic and demographic attributes of immigrant-headed households now in the United States—would add about $10 per household to next year's net fiscal burdens for New Jersey residents and about $45 per household to next year's net fiscal burden for California residents. Nationally, admitting an additional 460,000 immigrant-headed households would lead to an increase in next year's net fiscal burden on all U.S. native households

OCR for page 254
--> of about $7.50 per household (New Jersey budgets) to $10 per household (California budgets). The estimates of the net annual fiscal impact provided in Tables 6.2 to 6.7 provide a useful snapshot of the current fiscal consequences of today's immigrant-headed households on native residents in the United States. But also like a snapshot, these annual estimates cannot be used to criticize the past nor to predict the future. A simplistic use of the net annual fiscal impacts estimated here will be misleading for at least two reasons. First, both the native and current immigrant populations and the populations of newly admitted immigrants grow over time. The annual fiscal impact estimates provided here must be adjusted for these changing demographics if we are to accurately judge the fiscal burdens or benefits today of prior policies, or to predict the future burdens or benefits of today's choices. Second, annual estimates take people as they are today, but in the future native residents, current immigrants, and newly admitted immigrants will be people who differ both demographically and economically; their fiscal contributions or fiscal burdens will be different too. Children who consume services and pay no taxes today become contributing taxpayers tomorrow. Today's contributing adults will retire in the future and become net beneficiaries of government programs. Estimates of the current net annual fiscal impact of today's immigrant-headed households are not likely to give us very accurate information about the fiscal impact of today's immigrants 20 or more years from now. The analysis here of the annual fiscal impact of today's immigrant households provides a starting point for understanding the future fiscal consequences of immigration. Predictions as to the long-term fiscal consequences of current or new immigration policies, however, must be based on a truly dynamic analysis of the fiscal incidence of immigration. Such a study must project the demographic and economic futures of current residents and new immigrants and the future paths of government spending, taxes, and debt policies. Only then will we have an honest representation of the long-run consequences of national immigration policies. Such estimates are provided in Chapter 7. References Aaron, H. 1975 Who Pays the Property Tax: A New View. Washington, D.C.: The Brookings Institution. Angrist, J., and V. Lavy 1997 Using Maimonides' Rule to Estimate the Effect of Class Size on Scholastic Achievement. National Bureau of Economic Research Working Paper No. 5888. Borjas, G., and S. Trejo 1991 Immigrant participation in the welfare system. Industrial and Labor Relations Review 44(Jan.): 195-211. Borjas, G., and L. Hilton 1996 Immigration and the welfare state: Immigrant participation in means-tested entitlement programs. The Quarterly Journal of Economics 111 (May):575-604.

OCR for page 254
--> Boskin, M.J., M.S. Robinson, T. O'Reilly, and P. Kumar 1985 New estimates of the value of federal mineral rights and land. American Economic Review 75(5):923-936. Brueckner, J. 1981 Congested public goods: The case of fire protection. Journal of Public Economics February:45-58. Clark, R.L. 1994 The Costs of Providing Public Assistance and Education to Immigrants. Washington, D.C.: The Urban Institute. Clune, M. 1997 The Fiscal Impacts of Immigrants: A California Case Study. Paper prepared for panel. Department of Demography, University of California, Berkeley, January. Courant, P.N. 1977 A general equilibrium model of heterogeneous local property taxes. Journal of Public Economics 8(3):313-327. Craig, S. 1987 The impact of congestion on local public good production. Journal of Public Economics 32(Apr.):331-353. Duncombe, W., and J. Yinger 1993 An analysis of returns to scale in public production, with an application to fire protection. Journal of Public Economics 52(Aug.):49-72. Garvey, D., and T. Espenshade 1996 Fiscal Impacts of New Jersey's Immigrant and Native Households on State and Local Governments: A New Approach and New Estimates. Office of Population Research, Princeton University, September. Grieson, R. 1974 The economics of property taxes and land values: The elasticity of supply of structures. Journal of Urban Economics (Oct.):367-381. Hanushek, E. 1986 The economics of schooling: Production and efficiency in public schools. Journal of Economic Literature 24(September), 1986:1141-1177. Inman, R. 1978 A generalized congestion function for highway travel. Journal of Urban Economics 5(Jan):21-34. MaCurdy, T. 1992 Work disincentive effects of taxes: A reexamination of some evidence. American Economic Review 82(May):243-249. MaCurdy, T., T. Nechyba, and J. Bhattacharya 1996 An economic framework for assessing the fiscal impacts of immigration. Department of Economics, Stanford University, September. McMillan, M. 1989 On measuring congestion of local public goods. Journal of Urban Economics 26(Jan): 131-137. Mieszkowski, P. 1972 The property tax: An excise or profits tax? Journal of Public Economics 1(April):73-97. Pechman, J. 1985 Who Paid the Taxes, 1966-85. Washington, D.C.: The Brookings Institution. Rothman, E., and T. Espenshade 1992 Fiscal impacts of immigration to the United States. Population Index 58(Fall):381-415.

OCR for page 254
--> Shoven, J., and J. Whalley 1972 A general equilibrium calculation of the effects of differential taxation of income from capital in the U.S. Journal of Public Economics 1(Apr.):281-321. Vernez, G., and K. McCarthy 1995 The Fiscal Costs of Immigration: Analytical and Policy Issues. DRU-958-1-IF. Center for Research on Immigration Policy, RAND. 1996 The Costs of Immigration to Taxpayers: Analytic and Policy Issues. Center for Research on Immigration Policy, RAND.