the impact of the chst on interprovincial redistribution...

The Impact of the CHST on Interprovincial Redistribution in Canada49

CANADIAN PUBLIC POLICY – ANALYSE DE POLITIQUES, VOL. XXIV , NO. 1 1998

The Impact of the CHST onInterprovincial Redistributionin CanadaTRACY R. SNODDON

Department of EconomicsWilfrid Laurier UniversityWaterloo, Ontario

En 1996, les transferts effectués sous le Financement des programmes établis (FPE) et le Régime d’assistancepublique du Canada (RAPC) ont été remplacés par le Transfert canadien en matière de santé et de programmessociaux (TCSPS). Cet article analyse l’impact du TCSPS sur la redistribution nette des revenus entre lesprovinces lors des premières années de sa mise en application. Les résultats démontrent que le TCSPSmodifie les tendances récentes dans la redistribution nette mais que les effets sont modestes lorsqu’on lescompare à ceux des mesures antérieures comme le gèle des droits per capita du FPE et le plafond du RAPC,imposés en 1990. La décision récente du gouvernement fédéral d’augmenter le plancher du TCSPS réduitd’avantage l’effet du TCSPS sur la redistribution par rapport au statu quo tout comme le fait la décisiond’ajuster l’allocation du TCSPS graduellement (plutôt qu’immédiatement) en vue d’obtenir des parts percapita égales. Finalement, la distinction entre les transferts monétaires et les droits est importante lorsquel’on mesure la redistribution nette des transferts fédéraux vers les provinces.

In 1996, transfers made under Established Programs Financing and the Canada Assistance Plan were replacedwith the Canada Health and Social Transfer (CHST). This paper examines the impact of the CHST on thenet redistribution of revenues across provinces in its first few years of operation. The results show that,while the CHST alters recent trends in net redistribution, the effects are modest when compared to theimpacts of earlier measures such as the freeze in per capita EPF entitlements and the “cap on CAP,” im-posed in 1990. The federal government’s recent decision to raise the CHST cash floor further reduces theCHST’s effect on redistribution relative to the status quo as does the decision to adjust the allocation of theCHST slowly (rather than immediately) in the direction of equal per capita shares. Finally, the distinctionbetween cash transfers and entitlements is important when measuring net redistribution from federal trans-fers to provinces.

INTRODUCTION

Prior to 1996, the federal government transferredrevenues to provinces under three major pro-

grams: equalization, Established Programs Financ-ing (EPF), and the Canada Assistance Plan (CAP).

These fiscal arrangements involved not only a sub-stantial amount of revenue sharing from the federalgovernment to the provinces but a significant amountof redistribution across provinces. In 1995, follow-ing 15 controversial years of federal governmentcuts, freezes, and caps in transfers to provinces,

50 Tracy R. Snoddon


Ottawa replaced EPF and CAP with the CanadaHealth and Social Transfer (CHST) effective 1996-97. The CHST affects the allocation of revenueacross provinces and the overall scale of transfers.To what extent is the redistribution of revenuesacross provinces affected by the switch to this newtransfer?

To address this question, this paper documentsrecent trends in interprovincial sharing from fed-eral transfers to provinces and estimates the impactof the CHST on interprovincial sharing in its firstfew years of operation. Interprovincial sharing ismeasured using net redistribution, a particularlyuseful measure since it includes not only transfersreceived by provinces but also takes into account aprovince’s indirect contribution to financing trans-fers through federal government revenues collectedin that province.1 If a province receives a share oftotal transfers equal to the proportion of federalgovernment tax revenue raised in that province, thentransfers involve revenue sharing and do not redis-tribute revenues across provinces. When transfershares do not correspond to the proportion of fed-eral revenue raised in a province, net redistributionis positive or negative and transfers permit somerevenue sharing but are also redistributive.

The paper begins with a brief description of EPF,CAP, equalization, and the CHST. While a numberof excellent descriptions of Canada’s fiscal arrange-ments are available elsewhere, a basic understand-ing of these transfers and recent changes to them isnecessary to understand recent trends in net redis-tribution and the implications of the new CHST forinterprovincial redistribution in Canada.2 Followinga summary of the data and methodology, net redis-tribution from federal transfers to provinces is cal-culated for the period 1987-88 to 1998-99. For com-parison, net redistribution under alternative sce-narios such as a continuation of the old system oftransfers or an equal per capita allocation rule forthe CHST are also considered. Given the importanceattached to cash transfers rather than transfer enti-tlements, net redistribution is calculated on both an

entitlements and a cash basis to determine whetherthe distinction is important in the context of inter-provincial sharing.

The results demonstrate that while the replace-ment of EPF and CAP with the CHST affects theextent of net redistribution from cash transfers, thiseffect is modest when compared to the impact ofother changes to transfers. In particular, the effectson net redistribution of the freeze in per capita EPFentitlements and the selective cap on CAP transfersin 1990 are much bigger than the effects of theswitch to the CHST. Relative to a continuation ofthe old system of EPF and CAP, the level of inter-provincial sharing is lower under the CHST. Therecently announced increase in the CHST cash floorfrom $11 billion to $12.5 billion to take effect in1997-98, however, reduces the CHST’s negativeimpact on net redistribution. Sensitivity analysisshows that a CHST with an equal per capita alloca-tion formula (in contrast to the gradual adjustmentof CHST shares in this direction) has a predictablylarger effect on net redistribution. Finally, net re-distribution is sensitive to whether transfers aremeasured on a cash or entitlements basis. In percapita terms, the difference is small but increasesover time. Thus, the distinction between entitlementsand cash is important for discussions of interpro-vincial sharing, especially when forecasting net re-distribution in the future.

EPF, CAP AND EQUALIZATION

Established Programs FinancingPrior to 1977, the federal government transferred asignificant amount of revenue in the form of condi-tional, cost-sharing grants to the provinces to helpfund postsecondary education and health care. Theopen-ended nature of these transfers as well as thedeterministic link between provincial spending de-cisions and federal expenditures under this programprompted Ottawa to shift to a block grant system in1977. The new EPF grant consisted of two independ-ent components: a cash payment and a tax transfer.3



The per capita cash transfer for 1977-78 was basedon half the national average per capita transfersmade for postsecondary education and health in1975-76. Each province received an equal per capitacash transfer under EPF which, from this point for-ward, escalated based on the rate of growth in GNPper capita. The second component involved a trans-fer of 13.5 personal income tax (PIT) points and onecorporate income tax point from the federal to pro-vincial governments. Since the revenue-generatingpotential associated with the transferred tax pointsdiffered across provinces, the tax transfer was equal-ized under the equalization program. All provinceswith below average revenue-generating potentialreceived the same equalized tax point transfer whilethe value of the tax transfer for provinces with aboveaverage revenue-generating potential was higher andvaried across these provinces.

Quebec received some of its EPF cash as an ad-ditional unequalized tax abatement of 8.5 PIT points.In 1995-96, for example, this abatement equalled$1.2 billion. While Quebec received the same percapita cash transfer as all other provinces on paper,its actual cash payment was lower by the amount ofthe province’s special tax abatement. Historically,this abatement has been treated as cash rather thana tax transfer by both Quebec and Ottawa.

In 1982, several changes were made to EPF, themain effect of which was to link the tax and cashcomponents. Under the new scheme, a province’sEPF cash transfer was calculated as the differencebetween its entitlement and the value of the taxtransfer including associated equalization. Whilegenerated by the tax transfer, associated equaliza-tion is paid out as part of the province’s equaliza-tion transfers and, to avoid double counting, mustbe deducted from the EPF entitlement when deter-mining the cash transfer. Per capita EPF entitlements(as opposed to cash) were to grow based on the rateof growth in GDP. With these changes, the federalgovernment treated the tax point transfer as if it was,in effect, transferred anew each year. Much of theprovincial discontent over fiscal arrangements has

been rooted in these changes. For example, the fed-eral government’s decision to freeze the per capitaEPF entitlement from 1991 to 1996 implied an even-tual decline in cash transfers. This outcome sharplycontrasted with the original 1977 arrangement ofannual growth in EPF cash. For Quebec, the issueof declining cash is more pressing because of itsspecial tax abatement in lieu of cash.

In 1995-96, EPF entitlements equalled approxi-mately $21.8 billion of which $805 million took theform of associated equalization.4 EPF cash trans-fers equalled $10.6 billion in 1995-96. EPF entitle-ments and associated equalization are provided forselected years in Tables A.1 and A.2 of the Appen-dix.5 A chronological summary of the federal meas-ures pertaining to EPF since 1982-83 is also givenin the Appendix in Table A.3. In the 1995 federalbudget, EPF was discontinued and replaced with theCHST.

Canada Assistance Plan (CAP)Under CAP, the federal government transferred rev-enues on a cost-sharing basis to provide financialsupport to provinces and territories for social as-sistance and welfare. Until recently, the federal gov-ernment was responsible for 50 percent of the costsof eligible provincial expenditures and CAP trans-fers took the form of open-ended, matching grantsequally available to all provinces. While adminis-tration of CAP-eligible programs was the exclusivedomain of the provinces, the objectives of CAP wereto provide assistance to needy individuals and toaddress poverty. In 1990 the federal governmentimposed a limit of 5 percent growth in CAP pay-ments to Alberta, Ontario, and British Columbia, of-ten referred to as the “have” provinces. The cap wasextended until 1994-95 and, in 1995-96, CAP trans-fers to all provinces were frozen at 1994-95 levels.Both Ontario and British Columbia hit the limit inall years the cap was in place while, in Alberta, thecap was binding only in the early years owing tothe reductions in welfare spending undertaken bythe Klein government during the early 1990s.

52 Tracy R. Snoddon


Total CAP entitlements for all provinces equalled$7.8 billion in 1995-96.6 Owing to the cost-sharingnature of this program, per capita CAP transfersvaried across provinces. With the exception of Que-bec, each province received the full value of theirentitlement as a cash transfer. Quebec received anadditional tax abatement of 5 PIT points in lieu ofcash which, in 1995-96, reduced its $2.7 billion CAPentitlement by $700 million. As of 1996-97, CAPwas discontinued and replaced with the CHST.

EqualizationCanada’s equalization program, introduced in 1957,is designed to unconditionally transfer revenues toprovinces with low revenue-generating capacity andis the only transfer where interprovincial redistri-bution is an explicit goal. Section 36 (2) of the Con-stitution Act 1982 commits the federal governmentto the principle of equalization and to ensuring thatprovincial governments can provide reasonablycomparable levels of public services at reasonablycomparable levels of taxation. Entitlements are de-termined by an explicitly legislated formula, usu-ally in place for five-year intervals, based on theaverage revenue-generating potential in Ontario,Quebec, Saskatchewan, Manitoba, and British Co-lumbia. The only provinces that do not receiveequalization are Alberta, British Columbia, and On-tario. A ceiling on the rate of growth in total equali-zation protects the federal government against dra-matic increases in equalization and floor provisionsprotect provinces to varying degrees for year overyear reductions in entitlements.7 Current funding ar-rangements for equalization are in effect from 1994-95 to 1998-99. Total equalization amounted to $8.6billion in 1995-96.8

THE CHST

In 1996-97, the federal government eliminated EPFand CAP transfers and introduced the CHST, argu-ing that block funding offers provinces more flex-ibility in the provision of social assistance and, at

the same time, provides greater predictability andfinancial sustainability of transfers. Like EPF, theCHST consists of both a cash and a tax transfer.Thus, the estimated value of the tax transfer madein 1977 is included as part of the CHST entitlementand cash is again determined residually. Associatedequalization resulting from the tax point transfercontinues to be paid as part of a province’s equali-zation entitlement.

With the CHST, both the scale and the allocationof transfers across provinces are different as com-pared to the case for EPF and CAP combined. TotalCHST entitlements are set at $26.9 billion and $25.1billion for 1996-97 and 1997-98 respectively. Thelegislated, five-year CHST funding arrangement forthe period 1998-99 to 2002-03 freezes total entitle-ments at $25.1 billion until the year 2000 after whichthe aggregate entitlement increases by a GDP-basedescalator. Two additional features of the CHST areimportant to note. Unlike its predecessor, the CHSTincludes a cash floor of $12.5 billion below whichtotal cash transfers to provinces cannot fall.9 For1996-97, each province receives a CHST share equalto its combined share of EPF and CAP in 1995-96.From 1997-98 and onwards, however, the allocationformula adjusts annually in the direction of equalper capita entitlements to reflect population growthand migration among provinces with the goal of re-ducing existing funding disparities in per capitaentitlements by half by the fiscal year 2002-03.10

METHODOLOGY AND DATA

The major transfer programs are most often dis-cussed on an individual basis and in absolute or percapita dollar terms or as a percentage of the nationalaverage. To understand interprovincial redistributionfrom transfers, however, an aggregate measure isappropriate. This paper uses the concept of net re-distribution developed in Reid and Snoddon (1992)to measure interprovincial sharing where net redis-tribution for a province is defined as total transfers



received net of the province’s contribution to thefinancing of transfer costs, summed over all majortransfers.11 Transfer costs are assumed to equal thetransfers made in total to all provinces, and so is-sues of deficits, administration costs, and the wel-fare costs of taxation are ignored. Provinces are as-sumed to share in the costs of transfers in propor-tion to the share of federal government revenue gen-erated in that province. For example, Ontario’s con-tribution to the financing of transfers or cost sharein year t is calculated as the federal government’srevenue from Ontario in that year divided by thefederal government’s total revenue for year t.

To illustrate, consider Ontario’s net redistributivegain for 1995-96. Ontario receives $10.7 billionfrom EPF and CAP transfers (Ontario does not re-ceive equalization). Ontario’s cost share in 1995-96 is 0.42283 indicating that roughly 42 percent ofthe $37.3 billion in transfers are financed by fed-eral government revenue collected in Ontario. There-fore, net redistribution for Ontario is:

Net gain = $10.7 billion – ($37.3 billion x 0.42283)= – $5.092 billion

This represents a net per capita “contribution” toother provinces of $459. Note that net redistribu-tion for a given province can be positive or negative.

In this example, net redistribution is calculatedusing entitlements rather than cash payments. “En-titlement” refers to the total value of the transfer ascalculated by the Department of Finance. For equali-zation and CAP transfers there is no distinction be-tween entitlements and cash transfers.12 For EPF andthe CHST, entitlements are distinct from cash trans-fers since a portion of these entitlements are “paid”out as a tax point transfer.

Net redistribution can also be calculated usingcash transfers. For province i, the difference in netredistribution from entitlements, NR i , and fromcash, NR i , is:

where TAXi and TAXT represent province i’s taxpoint transfer and the aggregate tax transfer respec-tively, NR i is province i’s net gain from the taxtransfer, and TAX sharei and cost sharei are prov-ince i’s share of the tax transfer and of transfer costsrespectively. NR i – NR i measures a province’s netgain from the tax transfer and is positive if a prov-ince’s share of the aggregate tax transfer exceedsits cost share. In Ontario, for example, the net percapita contribution is $459 when measured on anentitlements basis as compared to $464 when meas-ured on a cash basis, indicating that Ontario is a netgainer from the tax transfer.

There is a good deal of debate on the issue ofdeclining cash transfers and whether transfers toprovinces should be measured using entitlements orcash. The main argument in favour of using entitle-ments and including the tax transfer is that entitle-ments represent forgone tax revenues (actual andpotential) by the federal government. Those whofocus on cash transfers argue that the inclusion oftax transfers is confusing and, since no cash is ex-changed, an entitlements measure overstates the fed-eral government’s contributions to provincial cof-fers. If a province’s cost share is close to its tax trans-fer share, then the distinction between cash and en-titlements is of little importance, at least with re-spect to the concept of net redistribution. The re-sults presented below, however, indicate that this isprobably not the case.

Data on cash transfers, entitlements, and popu-lation, as well as projections for equalization, asso-ciated equalization, CHST transfers and populationwere supplied by the Department of Finance. Priorto 1995-96, provincial cost shares can be calculatedusing data on federal government revenue by

E

C

X

E C

= ä ã å å å

TAXi

TAXT TAXT ë

í ì ì ì − cost sharei H TAXT

= ( TAX sharei − cost sharei ) H TAXT

NR i

E − NR i

C = NR i

X

54 Tracy R. Snoddon


province from the Provincial Economic Accountsand from CANSIM. Beyond 1994-95, however, thenecessary data to calculate these shares are unavail-able. Provincial cost shares for 1995-96 to 1998-99are approximated using existing trends in federalgovernment revenue by province, projections forgrowth in provincial tax transfers, and the histori-cal relationship between a province’s tax and costshares. Appendix Table A.7 shows provincial costshares for selected years. Two caveats are noted. Asboth Quebec and the federal government treat Que-bec’s special abatements under EPF and CAP, andunder the CHST, as cash, this is the approachadopted in this paper. Finally, since the Yukon andthe Northwest Territories are excluded from theanalysis and aggregate CHST entitlements includetransfers to these territories, entitlements for theseterritories are netted out of the net redistributioncalculations.

NET REDISTRIBUTION FROM FEDERAL

TRANSFERS TO PROVINCES, 1987-1988 TO

1998-1999

Each province’s net redistributive gain, calculatedon both an entitlements and cash basis, is summa-rized in Table 1 for selected years over the period1987-88 to 1998-99. On an entitlements basis, Que-bec receives the largest net gain and the largest con-tribution comes from Ontario in every year consid-ered. Table 1 also shows that federal-provincialtransfers generate significant positive net gains forall provinces except Ontario, Alberta, and BritishColumbia, or the “have” provinces. A similar pat-tern of interprovincial sharing emerges when netredistribution is calculated using cash transfers. InTable 2, net redistribution from entitlements andcash transfers is shown in per capita terms. New-foundland receives the largest per person gain while

TABLE 1Net Redistribution from Total Entitlements (Cash)

Nfld. PEI NS NB Que. Ont. Man. Sask. Alta. BC

1987-88 886 168 586 717 2618 -4918 665 292 -693 -321(883) (169) (608) (732) (2639) (-4910) (664) (288) (-678) (-395)

1990-91 993 196 784 829 2953 -5546 850 481 -805 -734(990) (198) (817) (851) (2990) (-5500) (849) (482) (-826) (-851)

1995-96 1013 169 857 743 3681 -5092 926 275 -1308 -1264(1020) (175) (922) (791) (3646) (-5143) (932) (276) (-1310) (-1311)

1996-97 1006 166 846 752 3448 -4823 919 267 -1322 -1258(1014) (174) (922) (805) (3403) (-4903) (928) (269) (-1310) (-1301)

1998-99 998 172 885 776 3355 -4785 964 225 -1297 -1293(1008) (181) (968) (833) (3311) (-4886) (968) (231) (-1293) (-1321)

Note: All numbers are in millions of dollars.Source: Author’s calculations based on data provided by the Department of Finance.



Saskatchewan traditionally receives the smallest netper capita gain from both entitlements and cashtransfers. The per capita contributions in the “have”provinces range between $350 to $477 for 1995-96.

From Tables 1 and 2, it is clear that interprovin-cial sharing as measured by net redistribution dif-fers depending on whether cash transfers or entitle-ments are used. For example, the difference in On-tario’s net contribution from entitlements and fromcash in 1993-94 is $63 million or $6 per person.While the difference for any one province is gener-ally not greater than ± $90 per capita for the entireperiod, the discrepancy is growing over time formost provinces. If not for this evidence, the distinc-tion between entitlements and cash would be unim-portant at least in a discussion of interprovincialsharing.

The argument in favour of including the tax pointtransfer to reflect cash foregone by the federal gov-ernment is not very compelling. This argumenthinges on the belief that EPF entitlements are paidout by the federal government entirely as cash inthe absence of a tax transfer. Recent experience sug-gest this is unlikely. Even with tax transfers, the fed-eral government has, by reducing the EPF escalatorand eventually freezing per capita EPF entitlements,reduced its EPF cash obligations. The pressures toreduce cash obligations are likely to be even greaterif EPF entitlements are paid out entirely as cash.One additional argument can be made in favour ofcash transfers. Forecasts of net redistribution in thefuture are based on projections for populationgrowth and the growth in the value of the tax pointtransfer in each province. For a given value of theaggregate tax transfer under the CHST, predictions

TABLE 2Net Per Capita Redistribution from Total Entitlements (Cash)


1987-88 1560 1320 667 1007 397 -531 616 287 -291 -110(1555) (1330) (692) (1028) (400) (-530) (616) (283) (-285) (-135)

1990-91 1754 1515 878 1151 436 -566 784 485 -324 -232(1748) (1534) (915) (1182) (441) (-561) (783) (468) (-332) (-269)

1995-96 1758 1241 914 979 502 -459 814 271 -476 -336(1771) (1289) (984) (1042) (497) (-464) (820) (272) (-477) (-349)

1996-97 1761 1215 899 986 467 -429 804 261 -474 -327(1777) (1271) (979) (1055) (461) (-436) (812) (263) (-470) (-338)

1998-99 1748 1238 931 1012 448 -413 842 214 -446 -326(1766) (1301) (1018) (1087) (442) (-422) (845) (220) (-444) (-333)

Note: All numbers are in dollars.Source: Author’s calculations based on data provided by the Department of Finance.

56 Tracy R. Snoddon


for provinces’ net redistributive gains from entitle-ments are not sensitive to projections for how thistax transfer is shared across provinces. Estimatesof net redistribution from cash are, however, sensi-tive to differential growth rates across provinces thataffect their share of a given aggregate tax transfer.For these reasons, and given the evidence on thegrowing disparity between net redistribution fromentitlements and from cash, this paper focuses onnet redistribution from cash transfers.

Figure 1 shows the trends in net per capita redis-tribution for the Atlantic provinces, Quebec, Mani-toba, and Saskatchewan, the “have-not” provinces.The net per capita gain in Quebec and Saskatchewan,declining since the mid 1990s, continues to do so at

least in the first few years the CHST is in place. ForManitoba, Nova Scotia, and New Brunswick, themodest growth in the net per capita gain in recentyears continues under the CHST. There has been lit-tle change in the net per gain from cash transfersfor Prince Edward Island and Newfoundland since1993-94 even with the introduction of the CHST.

Figure 2 shows the trends in net per capita redis-tribution for the have provinces. For Alberta andBritish Columbia, the switch to the CHST clearlyreduces the net per capita contribution reversing thehistorical trend of increasing, and then relativelyconstant, net per capita contributions. The declinein Ontario’s net per capita contribution in recentyears also continues under the new CHST.

FIGURE 1Net Per Capita Redistribution from Cash Transfers, Have-Not Provinces



These figures show that the most dramaticchanges in net per capita redistribution over the pe-riod do not coincide with the introduction of theCHST but occur in 1990-91. The freeze in per capitaEPF entitlements in 1990 combined with the growthin the value of tax transfers cause EPF cash and thenet per capita gain to fall in the following one totwo years for most “have-not” provinces. The freezeand the cap on growth in CAP transfers in 1990 bothcontribute to an increase in the net per capita loss inAlberta and British Columbia. During the same pe-riod, Ontario’s net per capita contribution declineslargely as a result of a reduction in its cost shareduring the 1990-91 recession. Alberta’s net percapita contribution increases from 1992-93 to 1994-95 more quickly than British Columbia’s and at the

same time that Ontario’s net per capita contributionis falling. This increase in net per capita contribu-tion corresponds to Alberta’s significant cuts towelfare spending and the subsequent reduction inthe province’s CAP transfer.

A number of factors influence the level and pat-tern of net redistribution. The cap on CAP, the freezein EPF, the 1990-91 recession, the introduction ofthe CHST, and Alberta’s welfare reform clearly af-fect net per capita redistribution across provinces.The single most important determinant of net redis-tribution at any time is, however, equalization. Ta-ble 3 shows the net per capita redistribution by trans-fer for 1987-88, 1995-96, and 1998-99. Equaliza-tion accounts for over 75 percent of the net per capita

FIGURE 2Net Per Capita Redistribution from Cash Transfers, Have Provinces

58 Tracy R. Snoddon


gain in the Atlantic provinces and Manitoba and asomewhat smaller percentage in Quebec and Sas-katchewan. Equalization also accounts for the larg-est percentage contribution from the three “have”provinces. By 1998-99, the relative importance ofequalization to a province’s net per capita gain orloss increases for all provinces except Alberta andBritish Columbia. That equalization dominates thelevel of interprovincial sharing in Canada is not sur-prising since, unlike other federal transfers to prov-inces, the explicit intention of this program is re-distribution. With the reduction in the scale of trans-fers under the CHST relative to EPF and CAP, it isalso not surprising that equalization accounts for aneven larger share of net redistribution in the future.

NET REDISTRIBUTION: THE CHST VERSUS

THE STATUS QUO

The replacement of EPF and CAP transfers with theCHST has two distinct effects: a reduction in thescale of transfers (a scale effect) and an adjustmentin the share of transfers across provinces (a shareeffect). The scale effect reduces the extent of netredistribution across provinces if aggregate trans-fers are lower under the CHST. The share effect alsoreduces the extent of redistribution since the adjust-ments in provincial shares of transfers favour themore populous and wealthier provinces, Ontario,British Columbia, and Alberta. To fully assess theimpact of the CHST on redistribution, net redistri-

TABLE 3Net Per Capita Redistribution by Transfer


1987-88EPF Cash 205 159 82 122 69 -127 113 140 -2 60CAP 61 51 -10 78 78 -82 23 25 -9 42Equalization 1289 1120 620 828 253 -321 480 118 -274 -237

Total 1555 1330 692 1028 400 -530 616 283 -285 -135

1995-96EPF Cash 147 94 38 58 70 -66 87 116 -51 -10CAP 156 45 32 23 139 -73 42 37 -105 -40Equalization 1468 1150 914 961 288 -325 691 119 -321 -299

Total 1771 1289 984 1042 497 -464 820 272 -477 -349

1998-99CHST Cash 211 102 49 61 155 -99 106 105 -122 -31Equalization 1555 1199 969 1026 287 -323 739 115 -322 -302

Total 1766 1301 1018 1087 442 -422 845 220 -444 -333

All numbers are in dollars.Source: Author’s calculations based on data provided by the Department of Finance.



bution under a continuation of the old system oftransfers, or the “status quo,” is calculated and com-pared to net redistribution given the CHST.

To estimate net redistribution in the status quo,the level of and an allocation for EPF/CAP trans-fers must be determined for 1996-97, 1997-98, and1998-99. The 1995 Budget Plan estimates of theaggregate level of EPF/CAP entitlements under acontinuation of the status quo are summarized inTable 4. Since in 1995-96 both aggregate EPF/CAPentitlements and provincial entitlement shares werefrozen at 1994-95 levels, it is not unreasonable toassume shares remain relatively constant over thenext few years. Provinces are, therefore, assumedto maintain their 1995-96 shares of EPF/CAP enti-tlements for the next three years. Province i’s EPF/CAP entitlement in the status quo, EPF/CAP i , iscalculated as the province’s share of EPF and CAPcombined in 1995-96 multiplied by the aggregateentitlement for EPF and CAP, EPF/CAPT , or,

The cash transfer is determined by subtracting theestimated value of the province’s tax transfer fromEPF/CAP i . Equalization transfers are, of course,independent of the EPF/CAP or CHST variant in place.

A comparison of net per capita redistributiongiven the CHST versus the status quo captures boththe CHST’s scale and share effects. To separate theseeffects, net redistribution under a “revenue-neutralstatus quo” is estimated. This scenario measures theeffects on net redistribution if EPF and CAP arecontinued but aggregate entitlements are scaled backto the level of entitlements projected for the CHSTand provincial shares are assumed to equal the com-bined shares of EPF and CAP in 1995-96.13 As sug-gested in Boessenkool (1996), the announced totalsfor the CHST may provide an indication of theamount the federal government is willing to trans-fer under any EPF/CAP variant. The revenue-neutralcase thus captures the impact of adjusting provin-cial CHST shares in the direction of equal per capitatransfers. Province i’s EPF/CAP entitlement in therevenue-neutral status quo is:

where CHSTT represents the aggregate CHST enti-tlement. Province i’s cash transfer is then found bysubtracting the value of the CHST tax transfer esti-mated for that province from its EPF/CAP entitle-ment in the revenue-neutral case. The net redistri-bution calculations for cash transfers given the sta-tus quo, the revenue-neutral status quo, and the

TABLE 4Total Entitlements, Tax Point Transfers and Cash Under the CHST and the Status Quo

Entitlements Tax Point Transfer Cash

CHST and EPF and CAP in Revenue-Neutral Status Quo 1996-97 26.8 11.7 15.11997-98 25.0 12.4 12.61998-99 25.6 13.1 12.5

Status Quo EPF and CAP 1996-97 29.4 11.7 17.71997-98 29.6 12.4 17.21998-99 29.9a 13.1 16.8

aNo estimate is provided in the 1995 Budget. EPF + CAP are estimated to be 1 percent higher than 1997-98 level.Note: All numbers are in billions of dollars.

E

E

E

EEPF/ CAP i

E = ( EPF/ CAP share i95/ 96) H EPF/ CAPT

E

( EPF/ CAP i

E ) RNSQ = ( EPF/ CAP share i95/ 96) H CHSTT

E

60 Tracy R. Snoddon


CHST are summarized in Table 5. Since equaliza-tion is the same in all three cases, differences in netper capita redistribution reflect differences in non-equalization transfers. A comparison of net redis-tribution under the CHST with the revenue-neutralcase measures the share effect of the CHST. Thescale effect from the CHST is seen by comparingnet redistribution under the revenue-neutral statusquo and the status quo.

Net per capita redistribution under these alterna-tives for all provinces are illustrated in Figures 3 to8. These figures show that, while the extent of re-distribution is lower under the revenue-neutral caseas compared to the status quo, the revenue-neutralcase involves a greater degree of sharing than pre-

dicted for the CHST. Notice that net per capita re-distribution in 1996-97 under the CHST and therevenue-neutral case are identical. This is to be ex-pected since aggregate transfers and provincialshares are the same in both cases. From this pointforward, provincial CHST shares are adjusted toreflect population growth and migration and so dif-fer from the shares in the revenue-neutral status quo.Both the reduction in the scale of transfers and thechange in the allocation of transfers across provinceswork to reduce net per capita gains (and losses) fromcash transfers although, for Saskatchewan and PrinceEdward Island, the share effect is quite small.14

A comparison of net per capita redistributionbetween the CHST and the status quo indicates that

TABLE 5Net Per Capita Redistribution: Alternative ScenariosCash Transfers

Scenario Nfld. PEI NS NB Que. Ont. Man. Sask. Alta. BC

With CHST1996-97 1777 1271 979 1055 461 -436 812 263 -470 -3381997-98 1741 1255 989 1059 441 -422 818 237 -452 -3261998-99 1766 1301 1018 1087 442 -422 845 220 -444 -333

Status Quo1996-97 1815 1287 983 1064 482 -450 827 284 -488 -3441997-98 1840 1279 1010 1088 488 -450 861 276 -493 -3511998-99 1880 1331 1047 1124 499 -454 897 253 -499 -364

Revenue-Neutral Status Quo1996-97 1777 1271 979 1055 461 -436 812 263 -470 -3381997-98 1772 1254 1001 1072 450 -426 831 241 -461 -3391998-99 1816 1305 1037 1108 464 -432 867 221 -468 -352

Equal Per Capita CHST1996-97 1695 1269 954 1048 362 -399 806 283 -387 -3001997-98 1686 1254 972 1058 355 -391 818 260 -377 -3011998-99 1788 1297 1002 1086 363 -393 844 241 -377 -311

Note: All numbers are in dollars.Source: Author’s calculations based on data provided by the Department of Finance.



FIGURE 4Net Per Capita Redistribution, New Brunswick and Nova Scotia

FIGURE 3Net Per Capita Redistribution, Newfoundland and Prince Edward Island

62 Tracy R. Snoddon


FIGURE 6Net Per Capita Redistribution, Alberta and British Columbia

FIGURE 5Net Per Capita Redistribution, Manitoba and Saskatchewan



FIGURE 8Net Per Capita Redistribution, Quebec

FIGURE 7Net Per Capita Redistribution, Ontario

64 Tracy R. Snoddon


not all provinces are affected equally. For example,in 1998-99, the net per capita gain for Saskatchewanand Quebec is estimated to be 17.1 percent and 13.9percent higher respectively in the status quo rela-tive to the CHST. Ontario’s net per capita contribu-tion is 8.4 percent higher under the status quo and,in Alberta and British Columbia, the net per capitacontribution in the status quo is estimated to be 13.2percent and 10 percent higher as compared to netper capita contributions when the CHST is intro-duced. In comparison, the impact on net per capitaredistribution for Prince Edward Island, NovaScotia, and New Brunswick are much smaller. Forthese provinces, net per capita redistribution wouldhave been 2 percent to 3.6 percent higher under thestatus quo.

NET REDISTRIBUTION: CHANGING THE

CHST CASH FLOOR

Only one year after the CHST was introduced, thefederal government increased the cash floor from$11 billion to $12.5 billion, effective for the 1997-98 transfers. Based on Department of Finance pro-jections for the value of the tax point transfers,CHST cash is estimated to fall below $12.5 billionin 1998-99.15 With an $11 billion cash floor, the netper capita loss from transfers in Ontario, Alberta,and British Columbia and the net per capita gain inall other provinces are smaller relative to net redis-tribution in 1998-99 with the $12.5 billion floor.Thus, the move by the federal government to raisethe cash floor further mitigates against the adverseeffects on net redistribution brought about by theswitch to the CHST.

SENSITIVITY ANALYSIS

The results are sensitive to the provincial cost shareestimates for 1995-96 to 1998-99. A reasonable al-ternative is to assume cost shares return to sharesbased on some historical average or to 1990-91shares. Under both alternatives, net per capita re-

distribution is more favourable for most provincesexcept Ontario and Quebec. Boessenkool (1996)explores the impact on alternative growth scenarioson CHST cash transfers to provinces. Since differ-ential rates of growth in tax bases and tax transfersacross provinces can affect net redistribution, netredistribution is recalculated assuming the rate ofgrowth in the value of the tax transfer for each ofthe “have” provinces is one percentage point higherthan projected by the Department of Finance. In thiscase, all “have-not” provinces receive higher net percapita gains and each “have” province makes highernet per capita contributions. It is interesting to notethat if net redistribution is calculated using entitle-ments, a different allocation of the same aggregatevalue of the tax transfer across provinces would notalter the predictions of the net redistributive gain orloss for any province.

Before the CHST was revealed, much discussioncentred around provincial shares of any transfermeant to replace EPF and CAP. Under the CHST,provincial shares gradually adjust over time in thedirection of equal per capita CHST entitlements toreduce funding disparities. What is the impact onnet per capita redistribution if, for example, theCHST is instead introduced with an equal per capitaallocation rule in 1996-97? The net per capita re-distribution given an equal per capita allocation rulefor the CHST is shown in Table 5. In this scenario,interprovincial sharing is largely the result of equali-zation and the financing of equalization from fed-eral government revenue. To the extent that popula-tion shares differ somewhat from provincial costshares, some additional sharing comes about as aresult of financing CHST transfers from federalgovernment general revenue. Not surprisingly, thethree “have” provinces do better with equal percapita CHST entitlements as does Saskatchewan.The net per capita gain in all other provinces is loweras compared to the case where CHST shares gradu-ally adjust. In 1998-99, for example, Quebec suf-fers the biggest loss from the switch to equal percapita CHST entitlements with its net per capitaredistributive gain falling by over 17.5 percent. If



Quebec’s special tax abatements are taken into ac-count, the loss is probably greater.

Like EPF, the CHST includes both a tax transferwith the cash payment determined residually as thedifference between the CHST entitlement and thetax point transfer. In the absence of a cash floor,growth in the value of tax point transfer over timeimplies an eventual decline in cash transfers underthe CHST. A cash floor prevents this cash erosionsince, once the $12.5 billion floor becomes binding(as early as 1997-98), future growth in the value ofthe tax transfer leaves cash unaffected. Suppose thefederal government considers a separation of the taxpoint transfers from CHST cash transfers and rede-fines a $12.5 billion cash entitlement shared on anequal per capita basis as suggested in Courchene(1995). Under this alternative, net redistribution isreduced and Ontario, Alberta, and British Colum-bia are the clear winners. Another alternative is toconvert the equivalent of $12.5 billion in cash intoa PIT tax point transfer with provincial shares equalto the proportion of federal government PIT tax col-lected in each province. In this case, the additionaltax transfer is equalized up to the five provincestandard and the value of the transfer to each prov-ince fluctuates based on the growth of provinces’PIT tax bases. This option reduces the overall ex-tent of net redistribution in Canada as compared withnet redistribution given the existing CHST or giventhe equal per capita cash option discussed abovewith provinces like Ontario and British Columbiathe clear beneficiaries.

CONCLUSIONS

This paper examines recent trends in interprovin-cial sharing from federal transfers to provinces since1987-88 and estimates the impact on redistributionacross provinces of replacing EPF and CAP trans-fers with the CHST in the next few years. The re-sults show that the CHST’s effects on net redistri-bution are modest and that the federal government’srecent move to increase the CHST cash floor fur-

ther reduces the CHST’s impact. The switch to theCHST reduces the amount of net redistribution awayfrom Alberta and British Columbia, reversing thehistorical trend of increasing, and then relativelyconstant, net per capita redistribution. Netredistributive gains for Newfoundland and PrinceEdward Island fall relative to recent levels with theintroduction of the CHST. For the other provinces,however, existing trends are unchanged with the in-troduction of the CHST. The paper shows that theCHST’s impact on net redistribution is greater ifintroduced with equal per capita entitlements ratherthan the current allocation formula that adjust sharesonly very slowly over time in this direction. In re-cent years, the most significant changes in trends ininterprovincial sharing do not coincide with the re-placement of EPF and CAP with the CHST. Thefreeze in per capita EPF entitlements and the capon CAP transfers for Alberta, Ontario, and BritishColumbia in 1990 have a much bigger impact onnet redistribution across provinces than does theCHST in its first three years of operation.

The level of interprovincial sharing is higher un-der a continuation of the old system of EPF and CAPas compared with net redistribution given the CHST.The results also show that the net redistributionmeasure of interprovincial sharing is sensitive towhether transfers are measured on a cash or entitle-ments basis. In per capita terms, the difference issmall but has been increasing over time. Thus, thedistinction between entitlements and cash is impor-tant for discussions of interprovincial sharing, es-pecially when forecasting net redistribution in thefuture.

NOTES

The author would like to thank participants at the Cana-dian Economic Association meetings in June 1996 forhelpful comments and Rob Sager for research assistance.Support provided by an internal research grant from theOffice of Research, Wilfrid Laurier University is grate-fully acknowledged.

66 Tracy R. Snoddon


1This paper deals with intergovernmental transfers.Interpersonal sharing and federal transfers to persons arenot addressed.

2See Courchene (1984, 1994) and Boadway andHobson (1993).

3See Courchene (1995) for a detailed description ofcash versus tax point transfers.

4Unless otherwise noted, associated equalization isincluded under equalization and deducted from total EPFand CHST entitlements to avoid double counting.

5Full results for all years are available upon request.

6Provincial CAP entitlements for selected years areshown in Table A.4 in the Appendix.

7See Courchene (1994) for an easy-to-understandguide to equalization.

8Appendix Table A.5 shows provincial equalization en-titlements for selected years.

9The CHST was originally introduced with a cash floorof $11 billion. In December 1997, the federal governmenttabled legislation to raise the cash floor to $12.5 billioneffective 1997-98.

10Canada. Department of Finance (1996), p. 59.

11This methodology has been used elsewhere. For ex-ample, see Boadway and Hobson (1993).

12In practice there is a distinction under CAP for Que-bec. However, since Quebec’s special abatements underCAP, EPF, and the CHST are treated as cash, this distinc-tion is ignored.

13Per capita EPF entitlements have been frozen since1990 and the federal government froze CAP plus EPFtransfers in 1995-96 at 1994-95 levels. The continuationof EPF and CAP assumes a continuation of the status quosince 1994-95; it does not capture net redistribution had EPFand CAP been unrestrained by the federal government.

14Similar results are found when net redistribution ismeasured using entitlements.

15Even more recent estimates indicate cash transfersfor 1997-98 of less that $12.5 billion which explains whythe federal government made the new cash floor effec-tive for 1997-98 rather than 1998-99 as originally in-tended. This paper works with the original estimates forCHST entitlements as given in Tables 4 and A.6.

REFERENCES

Boadway, R. and P. Hobson (1993), IntergovernmentalFiscal Relations in Canada (Toronto: Canadian TaxFoundation).

Boessenkool, K. (1996), The Illusion of Equality: Pro-vincial Distribution of the Canada Health and SocialTransfer. C.D. Howe Institute Commentary 80 (To-ronto: C.D. Howe Institute).

Canada. Department of Finance (1995), Budget Plan.Ottawa. February 27.

____ (1996), Budget Plan. Ottawa. March 6.Courchene, T.J. (1984), Equalization Payments: Past,

Present and Future (Toronto: Ontario Economic Coun-cil).

____ (1994), Social Canada in the Millennium: ReformImperatives and Restructuring Principles, The SocialPolicy Challenge 4 (Toronto: C.D. Howe Institute).

____ (1995), Redistributing Money and Power: A Guideto the Canada Health and Social Transfer (Toronto:C.D. Howe Institute).

Hobson, P. and F. St. Hilaire (1994), Reforming Federal-Provincial Fiscal Arrangements: Towards SustainableFederalism (Montreal: Institute for Research on Pub-lic Policy).

Reid, B. and T. Snoddon (1992), “Redistribution UnderAlternative Constitutional Arrangements for Canada,”in Alberta and the Economics of ConstitutionalChange, ed. P. Boothe (Edmonton: Western Centre forEconomic Research).



APPENDIX

TABLE A.1Total EPF Entitlements

Nfld. PEI NS NB Que.a Ont. Man. Sask. Alta. BC Canada

1987-88 392 88 606 491 4548 6391 744 701 1640 2018 17619

1990-91 429 98 677 546 5136 7423 821 751 1883 2400 20165

1995-96 426 101 693 561 5418 8193 840 750 2028 2778 21788

aIncludes Quebec’s abatement of 8.5 PIT points.Notes:1. All numbers are in millions of dollars.2. Includes associated equalization.Source: Department of Finance.

TABLE A.2Associated Equalization

Nfld. PEI NS NB Que. Man. Sask. Canada

1987-88 82.8 16.0 72.2 78.2 331.7 86.4 98.5 765.8

1990-91 88.1 18.0 83.1 85.4 370.6 97.5 113.7 856.4

1995-96 81.9 16.9 85.9 79.4 354.9 84.9 100.8 804.7

1998-99 91.0 19.0 98.0 90.0 399.0 88.0 117.0 902.0

Note: All numbers are in millions of dollars.Source: Department of Finance.

68 Tracy R. Snoddon


TABLE A.3Federal Measures Pertaining to EPF

Year Federal Measures

1982-83 technical changes made to EPF

1983-84 federal government places a cap on growth of the postsecondary component of EPF equal to6 percent over 1982-83 level

1984-85 federal government places a cap on growth of the postsecondary component of EPF equal to5 percent over 1983-84 level

1986-87 in 1985-86 federal government announces EPF escalator will be limited to the rate of growthin GDP - 2 percent (GDP - 2 percent) for subsequent years

1987-88 to 1989-90 escalator limited to GDP - 2 percent

1990-91 in 1989 federal government announces that EPF escalator will be further reduced to GDP - 3percent

1991-92 in 1990 federal government announces per capita EPF entitlement will be frozen at 1989-1990 levels for two years

1992-93 in 1991 federal government announces that the freeze on EPF will continue until 1994-95

1993-94 to 1994-95 freeze continues

1995-96 EPF escalator reverts to GDP - 3 percent

TABLE A.4Transfers to Provinces Under the Canada Assistance Plan

Nfld. PEI NS NB Que.a Ont. Man. Sask. Alta. BC Canada

1987-88 89 22 130 153 1602 1422 179 157 457 632 4842

1990-91 120 29 191 185 1951 2062 230 159 542 726 6195

1995-96 190 35 267 199 2704 2508 296 236 520 882 7837

aIncludes Quebec’s abatement of 5 PIT points.Note: All numbers are in millions of dollars.Source: Department of Finance.



TABLE A.5Equalization Entitlements

Nfld. PEI NS NB Que. Man. Sask. Canada

1987-88 807 163 734 724 3151 727 299 6605

1990-91 919 194 949 868 3627 915 531 8002

1995-96 954 188 1115 927 3952 1056 337 8529

1998-99 997 200 1191 989 4037 1118 349 8881

Note: All numbers are in millions of dollars.Source: Department of Finance.

TABLE A.6CHST Entitlements and CHST Cash

Nfld. PEI NS NB Que. Ont. Man. Sask. Alta. BC Canada

1996-97Entitlements 559 123 869 690 7359 9683 1032 898 2273 3311 26797Cash 347 72 519 407 4612 5012 606 518 1147 1822 15062

1997-98Entitlements 504 115 800 634 6799 9081 948 834 2147 3142 25004Cash 282 61 431 336 3902 4121 506 431 941 1558 12569

1998-99Entitlementsa 506 117 812 643 6866 9362 961 857 2242 3241 25607Cashb 275 61 426 332 3825 4120 504 432 960 1565 12500



aEntitlements given binding $12.5 billion cash floor.bAdditional cash is added to ensure that the difference between aggregate CHST entitlements and the total tax transferequals $12.5 billion and is distributed based on province’s share of total CHST entitlements in that year.Note: Entitlements include associated equalization, are expressed in millions, and exclude the Yukon and NorthwestTerritories.Source: Department of Finance.

70 Tracy R. Snoddon


TABLE A.7Provincial Shares of Federal Government Revenue


1987-88 1.13% 0.31% 2.87% 2.03% 22.4% 45.0% 3.18% 2.71% 9.86% 10.50%

1990-91 1.15% 0.32% 2.83% 2.04% 22.1% 44.9% 3.04% 2.53% 9.64% 11.52%

1995-96a 1.27% 0.37% 3.03% 2.31% 21.5% 42.3% 3.16% 2.54% 10.32% 13.19%

1998-99a 1.23% 0.38% 3.04% 2.28% 21.3% 42.1% 3.06% 2.57% 10.54% 13.50%

aEstimates based on Department of Finance projections for CHST tax transfers and historical trends in the value ofprovincial tax transfers and federal revenue shares. Rounded to two decimal places for presentation only.Note: Shares may not add to 100 due to rounding.Source: Provincial Economic Accounts and CANSIM matrices 6757 to 6766.

the impact of the chst on interprovincial redistribution...

Documents