Immigrant public finance transfers: a comparative analysis by city.
Subject: Transfer payments (Comparative analysis)
Immigrants (Economic aspects)
Immigrants (Comparative analysis)
Public finance
Authors: DeVoretz, Don
Pivnenko, Sergiy
Pub Date: 06/22/2004
Publication: Name: Canadian Journal of Urban Research Publisher: Institute of Urban Studies Audience: Academic Format: Magazine/Journal Subject: Social sciences Copyright: COPYRIGHT 2004 Institute of Urban Studies ISSN: 1188-3774
Issue: Date: Summer, 2004 Source Volume: 13 Source Issue: 1
Product: Product Code: 9000140 Public Finance-Total Govt; 9100001 Public Administration & Finance; 9000100 Public Admin & Finance-Total Govt; E198450 Immigrants NAICS Code: 92113 Public Finance Activities
Geographic: Geographic Scope: British Columbia; Ontario; Quebec Geographic Name: Montreal, Quebec; Toronto, Ontario; Vancouver, British Columbia Geographic Code: 1CQUE Quebec; 1CONT Ontario; 1CBRI British Columbia
Accession Number: 115498150
Full Text: Abstract

A comparison of immigrant tax payments and government transfers in Vancouver, other Canadian cities, and the nation as a whole is the focus of this study. A life-cycle model is employed to predict the lifetime patterns of selected public transfers and tax payments by Canadians and immigrant households, with a particular focus on Vancouver. Considering only the selected financial measures, the results indicate that immigrants in Vancouver, Toronto and Montreal make a positive public finance contribution, although the foreign born contribute less than the Canadian born and there are substantial differences in this gap across cities. Moreover, the gap is larger in each of these cities than is the national average.

Keywords: Public Finance, Immigrants

Resume

Darts cette etude, les auteurs comparent les versements en impots des immigrants et les paiements de transfert gouvernementaux pour Vancouver, d'autres villes canadiennes et tout le Canada. Un modele de cycle de vie permet de predire les profils de certains paiements de transfert publics et des versements en impets des Canadiens et des menages d'immigrants, en particulier ceux des residents de Vancouver. Compte tenu de mesures fiscales precises, les resultats montrent que les immigrants des regions de Vancouver, Toronto et Montreal, contribuent de facon positive aux finances publiques bien que leur contribution soit inferieure a celle des Canadiens de souche et que des differences importantes se manifestent d'une ville a l'autre. Ce fosse entre Canadiens de souche et immigrants est d'ailleurs plus marque pour chacune des villes que pour les moyennes nationales.

Mots cles : Finance publique, immigrants

Introduction

Immigrants affect the public finances of Canada as soon as they arrive in the country, as they may bring tax subsidies in the form of embodied formal education from their origin country (Coulson and DeVoretz 1993). They may continue to subsidize the Canadian taxpayer if their tax contributions exceed their use of public services (Akbari 1995). Of course, the opposite may hold. Immigrants may consume more services than they pay in taxes, especially soon after their arrival (Baker and Benjamin 1995). Current Canadian literature addresses the public transfer question exclusively at the aggregate level (Akbari 1989, 1995). It argues that the representative immigrant, regardless of location in Canada, makes a modest net contribution over a lifetime to the treasury. This contribution may vary across Canada for a variety of reasons.

Since the major metropolitan areas of Canada receive differential proportions of immigrants across entry classes, the effects on individual city economies differ by immigrant employment outcomes. Figure 1 shows that, over the recent past, Montreal, Toronto and Vancouver have respectively received about 60, 50 and 40 percent or more of their immigrants in the non-economically assessed refugee and family classes. (1) In the same period Vancouver received over 60 percent of its immigrants in the economically assessed groups and only 4 percent of new arrivals in the refugee class. Given such differential entry and economic conditions in Canada's three main immigrant-receiving cities, it is the central task of this paper to measure the possible variations in public transfers across cities.

Theory

A crucial and reoccurring question in the minds of immigration critics is: "Do the foreign born make a net positive contribution to the treasury?" A seminal article by Simon (1984), in which he argues that it is necessary to consider the life-cycle consumption of public services and tax payments of the foreign born, sets a theoretical framework to answer this question. Simon suggested that income and, therefore, tax payments are concave in age, while public transfers consumption is convex in age. Hence, the theoretical construct for a life-cycle model for tax payments and goods consumption is assured if these conditions hold. (2) This is illustrated in Figure 2.

[FIGURE 2 OMITTED]

In addition, since many Canadian immigrants are doubly selected, first by themselves and secondly by Canada's point system, the hypothesized age transfer patterns may vary by birth status. The effect of this double selection is observed in many dimensions. First, Canadian born residents start consuming public finance transfers (denoted Cn-Cn in Figure 2) for health services at birth and then, at the age of six, receive educational subsidies. Between age 19 and 22, the Canadian born population simultaneously exits the education sector and enters the labour force. This reduces their consumption of subsidized education and associated government transfers. At this point in the life-cycle, the Canadian born begin to pay taxes (Tn-Tn). In contrast to this life-cycle pattern, the foreign born public good consumption curve (Ci-Ci) begins later, after entry to Canada (i.e. age 26) and may lie everywhere below (or above) the public good consumption curve for its Canadian born cohort, depending upon their economic and demographic circumstances. This ambiguity in the level of consumption of public transfers by the foreign born at every age results from two countervailing forces. On the one hand, the immigrant's initial lack of eligibility to consume some subsidized services, such as health, are tied to residency requirements. Other programs, such as (Un)Employment Insurance, are income contingent and, thus, may reduce immigrant access to public transfers (Benjamin and Baker 1995; for a detailed discussion of immigrants' use of Employment Insurance see Sweetman, 2001). On the other hand, the greater initial risk faced by the immigrants in the labour market could lead to their greater use of entitlement programs, such as social assistance, upon arrival. It is hypothesized that foreign born tax payments (Ti-Ti) which are initially low accelerate to a "'crossover" point (at X), where tax payments by the foreign born exceed those of the Canadian born as a result of higher foreign born earnings and the progressive nature of Canada's tax system.

Figure 2 best represents the optimistic case. A pessimistic case would have the foreign born earn less and never "catch-up" in earnings to the Canadian born, which would result in lower tax payments. This would delay or completely forestall their tax "crossover." Under this scenario, given a low household income, the foreign born consumption of government transfers would lie above that predicted in Figure 2. This is shown in Figure 3, where immigrant tax payments never exceed their consumption of public services. The life-cycle model represents a flexible framework to analyze differential tax and benefit transfers by foreign birth status.

[FIGURE 3 OMITTED]

Data

The Survey of Consumer Finances (SCF) for Census Families, conducted annually by Statistics Canada until 1997, is a standard cross-sectional data set used to analyze Canadian government transfer consumption. This data set contains many of the relevant variables outlined in the life-cycle model. Direct observations on both tax payments and consumption of particular government transfers are provided in detail at the census metropolitan area (CMA) level for both the Canadian-and foreign born populations. In order to obtain reliable sample statistics and to minimize business cycle effects on consumer finances, we use nine SCFs from 1989 through 1997. Because of the survey design and response rates, each microdata record was weighted using the weights provided by Statistics Canada.

A study such as this is limited by the data available. While the SCF is arguably the best data for this purpose, it only contains a limited number of measures from the full set of both government transfers and taxes paid. With respect to the consumption of government transfers, it contains only measures of the transfers listed in Table 1. It does not include measures of other types of transfers, such as subsidized housing. Further, it does not include the value of government services consumed by individuals, such as healthcare, education and the like. Finally, it does not contain measures of the value of the consumption of public goods such as parks, roads and policing.

Looking at the taxes paid side of the ledger, the "Total Income Tax" variable in SCF microdata file represents the total federal and provincial income tax payable by all individuals in the family for the reference taxation year. It excludes any CPP/QPP contributions on self-employment earnings and other earnings, social benefits repayment and supplementary CPP/QPP contributions, and UI/El premiums. In our analysis, tax payments are the total amounts stated on each individual income tax return (i.e., the income tax actually paid less the refund due or the tax paid plus the portion still owing). It is important to note that our tax variable does not include the GST, provincial sales taxes, property taxes and the like, and therefore understates total taxes paid.

The value of the study, therefore, is not in estimating the absolute value, that is the levels, of tax contributions and receipts from government, since there are very important omissions on each side. Rather, it is in comparing the relative levels of what income taxes and government transfers paid we can observe across cities by immigrants and the Canadian born. If other taxes and transfers are highly correlated with these, then it can give us some sense of the bigger picture.

The Vancouver foreign born and Canadian populations differ with respect to key demographic, economic and government transfers consumption variables. The Vancouver immigrant population is estimated to be more than four years older and earns almost $3,000 less than the Canadian comparison group. In addition, the immigrant population consumes slightly more in government transfers ($4,594) than the Canadian born population ($4,304), while paying less in taxes ($7,168) than Vancouver's Canadian born residents ($8,625). We investigate these patterns in detail below.

Figure 4 depicts the distribution of consumption by foreign birth status for those with positive public transfer receipt (i.e., those with zero government transfer receipt are excluded from this graph). First, it should be noted that the average transfers to the Canadian born population ($5,683) only slightly exceed those to the foreign born ($5,588) in Vancouver's CMA. Broadly similar patterns of receipt exist between the two groups, with the exception that the immigrant population receives almost twice as much in child tax benefits as the Canadian born. Of course, this graph does not take into account differences in age and other characteristics between the immigrant and Canadian born populations.

It is possible to array the total reported government transfers by age and foreign birth status to create a life-cycle diagram, although we only start the profile at age 20. This is presented in Figure 5. Total government transfers by age follow the predicted pattern of the life-cycle hypothesis. Benefit receipts increase dramatically for both the foreign born and Canadian born Vancouver populations around retirement and the Canadian born receive somewhat greater transfers beyond about age 70.

[FIGURE 5 OMITTED]

It is possible to disaggregate the consumption of public finance transfers into the two most contentious public consumption items, (Un)Employment Insurance and social assistance. Figures 6 and 7 respectively present the lifecycle consumption patterns for these two government transfers for Vancouver foreign born and Canadian born households circa 1989-1997.

[FIGURES 6&7 OMITTED]

Several points are clear. First, unemployment insurance consumption reaches its maximum ($1,100) at age 30 and then slowly declines to age 60, followed by a dramatic drop for the Canadian born household in Vancouver. Thus, the Canadian born assimilate into employment insurance consumption between the ages 20 to 30. Their consumption then almost flattens until their late 50s, when their labour force attachment becomes more tenuous. The foreign born household in Vancouver follows a different pattern. Lifetime employment insurance consumption peaks twice--at age 35 ($1.200) and at age 60 ($1,300), with a decline to $500 between ages 40 and 45.

Figure 7 portrays the use of social assistance by Vancouver households over the life cycle. Between the ages 20 to 30, Vancouver households receive an approximately average $600 in social assistance transfers, regardless of birth status. Then, for both Canadian--and foreign born households between the ages 30 to 60, social assistance payments display a convex pattern with occasional fluctuations. The rise in payments after age 55 for both groups conforms to that predicted by the life-cycle model, given their weaker commitment to the labour market.

Net Transfers: Vancouver

Figures 8 and 9 summarize the results from combining the tax and consumption amounts to yield net government financial transfers by birth status in Vancouver for the limited set of taxes and transfers considered.

[FIGURES 8&9 OMITTED]

For Canadian born households in Vancouver, the life-cycle model is strongly supported by the evidence depicted in Figure 8. The total net fiscal transfer over the simulated lifetime of the representative Canadian born in Vancouver is $129,791 (1992 dollars). The computation of the total net transfer uses a 3 percent discount rate. The present value calculation is based each age cohort experiences the same pattern as that of the other age cohorts in the sample. (3) The combination of the convex age consumption profile and the concave tax payment profile supports the prediction, illustrated in Figure 2, that a positive public finance transfer will occur over the working life (age 25 to 65) of the representative head of household.

The foreign born Vancouver household follows a similar life-cycle pattern, contributing a positive, albeit smaller, lifetime transfer of $73,549. (4) This is approximately 57 percent of the amount transferred by Vancouver's representative Canadian born household.

Figures 10 and 11 place the Vancouver net financial transfer payments in comparative perspective and indicates the sensitivity of the results. For Canada as a whole, the average Canadian born household contributes a modest but positive lifetime financial transfer of $87,300. Lifetime financial transfers for the average foreign born household in Canada were lower at $73,438. In fact, the mean foreign born household for all of Canada transfers about 84 percent of the amount transferred by the mean Canadian born household. This is in sharp contrast to the size of the public finance transfers by the Canadian born in Vancouver, which greatly exceeds the national average for all Canadians.

[FIGURES 10&11 OMITTED]

Conclusion

Table 3 presents aggregate statistics supporting our thesis on the geographical dispersion of public finance transfers. Row (1) reports the Canada-wide results. It reflects the typical pattern reported in past Canada-wide research (Akbari 1995). The results indicate that, for the limited set of taxes and government transfers considered, the contribution to the treasury by Canadian born households across Canada exceed the contribution of foreign born households. The foreign born transfer 84 percent of the Canadian born payments (col. 4).

The Vancouver CMSA results, shown in Row (2), clearly reflect several city-specific effects. First, in Vancouver, the net financial transfer from Canadian born households is $129,791, while the net transfer from foreign born households is only $73,549. Thus, the ratio of payments by foreign birth status (col. 4) collapses in Vancouver, with the foreign born paying only 57 percent of the total paid by the Canadian born.

These trends are even more pronounced in Toronto, where net transfers are disproportionately higher for the Canadian born ($185,510) than for the foreign born ($84,194). The ratio of net transfers (col.3) becomes even smaller between the Toronto foreign born and Canadian born households (45 percent). Thus, on average, the foreign born household's transfer for the factors considered, while strongly positive, is $101,316 less than the lifetime net contribution of a representative Canadian born household in Toronto circa 1989-1997 (col. 5).

The public finance transfers in Montreal are the smallest observed. The net lifetime public finance transfer for a foreign born household in Montreal is only $15,940, while the Canadian born in the same city contribute, on average, $98,450.

In sum, it must first be noted that the foreign born in Toronto and Vancouver transfer at least as much as the national average, whereas immigrants in Montreal contribute only about one-fifth of the national average. Secondly, the Canadian born contribute more than the foreign born. However, the educational costs of neither group are included here and these are much higher for the Canadian born.

Comparative studies are available on the net fiscal impact of immigrants both in Europe and the United States to help place these results in perspective. The most comprehensive study to date, undertaken by Smith and Edmondson (1997), includes local, state and federal transfers for immigrants and the American born. The set of goods and taxes in their study is intended to be exhaustive. Table 4 reports the total immigrant net annual fiscal transfer (at the local, state and federal levels) for the United States and reveals that in 1996 immigrants were on average a drain on the U.S. treasury. As in this study, the immigrant burden varied by study area with immigrants resident in New Jersey imposing a negative fiscal impact of $166 per native household while in California the negative impact was $226 per native household. (5)

It is not possible to "import" these U.S. results to Canada, since Canada's taxes (and government service costs) are greater, and Canada's government transfer programs are much more generous. Moreover, this study is not nearly as comprehensive as the U.S. one. However, there is evidence of dramatic differences across Canadian cities in the impact of immigrants on the treasury and there is clearly much more work to be done on this topic.

Notes

(1) Montreal actually retained few of its business class immigrants who later moved to Toronto and especially Vancouver, according to IMDB records.

(2) There are however, two necessary adaptations to this theory. First, Canada has a progressive income tax system. This fact should theoretically yield a greater divergence between tax payments and (public) transfer consumption than would typically appear in the (private) life-cycle wealth accumulation literature (Shamsuddin and DeVoretz 1999). However, note that many of the non-measured taxes, such as property taxes and sales taxes, are not progressive. In addition, unlike most private consumption, the consumption of important publicly provided services is age specific. For example, education and health are used most at the earliest and latest part of the life cycle respectively.

(3) For example, the 20 30 year old cohort in 1995 must replicate the earnings and consumption pattern of the 3040 year old cohort contained in the 1995 sample in 2005. Since productivity increases over time, which raises incomes and tax transfers, these present value calculations are downward biased. This is certainly incorrect for both groups, as illustrated by Warman and Worswick in this issue. The declining earnings of more recent cohorts implies that these estimates are probably optimistic, with respect to immigrant transfers to the government treasury.

(4) This predicted pattern is a consequence of the convex shape for the consumption of government transfers and the concave shape for the tax payments, as represented in Figure 2.

References

Akbari, A. H. 1989. The benefits of immigrants to Canada: Evidence on tax and public services. Canadian Public Policy 15(4): 424-435.

Akbari, A. H. 1995. The impact of Immigrants on Canada's Treasury, circa 1990. In Diminishing Returns." The Economics of Canada 'S Recent Immigration Policy, ed. D. J. DeVoretz, 111-127. Toronto: C.D. Howe Institute; and Vancouver: The Laurier Institution.

Baker, M. and D. Benjamin. 1995. Labour market outcomes and the participation of immigrant women in Canadian transfer programs. In Diminishing Returns." The Economics of Canada's Recent Immigration Policy, ed. D. J. DeVoretz, 209-242. Toronto: C.D. Howe Institute; and Vancouver: The Laurier Institution.

Clune, M. 1997. The fiscal impacts of immigrants: A California case study. Paper prepared for panel. Berkeley: Department of Demography, University of California.

Coulson, R. G. and D. J. DeVoretz. 1993. Human capital content of Canadian immigration: 1966-1987. Canadian Public Policy 19(4): 357-366.

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. Princeton, N.J.: Office of Population Research, Princeton University.

Shamsuddin, A. F. M. and D. J. DeVoretz. 1999. Wealth accumulation of Canadian and foreign-born households in Canada. Review of Income and Wealth (December): 515-553.

Simon, J. 1984. Immigrants, taxes, and welfare in the United States. Population and Development Review 10 (1): 55-69.

Smith, J. P. and B. Edmondson. 1997. Do immigrants impose a net fiscal burden? Annual estimates. The new Americans: Economic, demographic and fiscal effects of immigration, eds J. P. Smith and B. Edmondson, 254-296. Washington, D.C.: National Academy Press.

Statistics Canada. 1996. 1995 Survey of Consumer Finances. Microdata File: Census Families. Ottawa: Government of Canada.

Sweetman, A. 2001. Immigrants and Employment Insurance. In Essays on the Repeat Use of Unemployment Insurance, eds. Saul Schwartz and Abdurranhman Aydemir. Social Research and Demonstration Corporation: Ottawa, 123-154.
Table 1: Descriptive Statistics: Vancouver Foreign-born Population
in 1989-1997, 1992 dollars

               Age       Income      Total      Child      Old Age
                                    Gov. Tr    Tax Ben

Man          48,66      $40,077     $4,594     $370        $1,318
Min          17         -$17,265    $0         $0          $0
Max          80         $298,044    $50,668    $10,008     $16,723
Number *     2465713    2465713     2465713    2465713     2465713

             Canada       E.I.      Social       Tax
             Pension                Assist

Man          $1,014     $751        $523       $7,168
Min          $0         $0          $0         $0
Max          $20,547    $22,473     $28,694    $112,799
Number *     2465713    2465713     2465713    2465713

Source: Surveys of Consumer Finances for Census Families 1989-1997.

Sample size: 3333 observations

* SCF revised weights used

Table 2: Descriptive Statistics: Vancouver Canadian-born
Population in 1989-1997, 1992 dollars

              Age       Income      Total       Child     Old Age
                                   Gov. Tr.    Tax Ben

Mean        44.24      $43,035     $4,304      $205       $1,247
Min         17         -$85,343    $0          $0         $0
Max         80         $289,035    $42,640     $11,088    $16,164
Nmnber *    4861720    4861720     4861720     4861720    4861720

            Canada       E.I.       Social       Tax
            Pension                Assist.

Mean        $1,073     $741        $547        $8,624
Min         $0         $0          $0          $0
Max         $18,685    $22,331     $22,729     $143,953
Nmnber *    4861720    4861720     4861720     4861720

Source: Surveys of Consumer Finances for Census Families

Sample size: 6373 observations

* SCF revised weights used

Table 3: Net Present Value of Public Finance Transfers by
CMSAs: 1989-1997, 1992 dollars

Location          Canadian-born        Foregn-born    Col (3)/Col. (2)
(1)               (2)                  (3)            (4)

Canada            $87,300              $73,438        0.84
Vancouver         $129,791             $73,549        0.57
Toronto           $185,510             $84,194        0.45
Montreal          $98,450              $15,940        0.16

Location          Col. (3) - Col (2)
(1)               (5)

Canada            -$13,862
Vancouver         -$56,242
Toronto           -$101,316
Montreal          -$82,510

Source: Surveys of Consumer Finances for Census families: Head of
Household.

Figure 1: Distribution of Immigrants by Entry Class Across Canadian
Cities: 1990-1995 arrivals

                   Toronto     Montreal    Vancouver

Skilled Workers     19.0%       29.0%       17.0%

Other               17.0%        6.0%       20.0%

Business             5.0%       15.0%       24.0%

Refugee             16.0%       20.0%        6.0%

Family Class        43.0%       30.0%       34.0%

Source: Landed Immigrant Data System (LIDS), CIC Canada

Note: Table made from bar graph.

Figure 4: Federal Government Transfers in Vancouver by Type and Birth
Status in 1989-1997

                          Foreign-born              Canadian-born
                     Total: $11,336,100,886     Total: $20,927, 952,947
                      Per Recipient: $5,588     Per Recipient: $5,683

Other                         11.7%                      9.9%

Social Assistance             11.6%                     12.9%

Employment Insurance          16.7%                     17.5%

Pension Plan                  22.5%                     25.4%

Old Age Benefits              29.2%                     29.5%

Child Tax Benefits             8.3%                      4.8%

Source:

Surveys of Consumer Finances (SCF) for Census Families 1989-1997.
All cases were weighted according to SCF Revised Weights

Note: Table made from bar graph.
Gale Copyright: Copyright 2004 Gale, Cengage Learning. All rights reserved.