Payroll Taxation in Canada an Overview

7/31/2019 Payroll Taxation in Canada an Overview

http://slidepdf.com/reader/full/payroll-taxation-in-canada-an-overview 1/18

Payroll Taxation in Canada:

An Overview

Livio Di Matteo

Michael Shannon*

Department of Economics

Lakehead University

Despite constituting the third most im-portant source of revenue for Canadian

governments, payroll taxes receive relativelylittle attention. This is particularly surprisingon two counts: first, there are important con-

troversies regarding the employment effectsof these taxes and, second, payroll taxeshave grown substantially over the last 30years and are likely to continue to increase inthe future.

Payroll taxes rose from 2.9 per cent of totalgovernment revenue in 1965 to 12.9 per centin 1993. As a share of total labour income,they increased from 1.9 per cent in 1961 to10.3 per cent in 1992. Payments for the threemajor programs funded through payroll taxes— Unemployment Insurance, the Canada and

Quebec Pension Plans, and provincial work-ers’ compensation — rose from 3.6 per cent of GDP in 1985 to 5.1 per cent in 1992.

The goals of this article are to provide a com-prehensive overview of payroll taxation, toconsider its future direction, and to discuss itspotential impacts. We begin by outlining thecurrent state of payroll taxation in Canada,chronicling its recent growth as a revenuesource. We also consider the range of mar-ginal payroll tax rates across provinces anduse earnings distribution data to determine

the groups facing the highest payroll tax rates.We then discuss anticipated future trends.Following this, the impact of payroll taxationon wages and employment is considered. Weconclude that these taxes do have a negativeimpact on employment; however, there are awide range of estimates as to the size of thiseffect.

Payroll Taxes inCanadaThe federal and provincial governments im-

pose payroll taxes with levies on both employ-ers and employees. The federal governmenttaxes employers and employees to finance un-employment insurance and pensions. Prov-inces tax employers to fund workers’ com-pensation and, more recently, health care andeducation; Quebec levies both employers andemployees for its own pension plan.

Over the last 30 years, the proportion of totalgovernment revenue accounted for by payrolltaxes has quadrupled (Table 1). The averagepayroll tax rate (as a percentage of total labour

income) rate jumped dramatically in 1966with the onset of the Canada and Quebec Pen-sion Plans (C/QPP) and has risen steadilyever since (Figure 1).

Employer contribution rates for C/QPP and

Unemployment Insurance (UI) have risensubstantially in recent years (Table 2). Therate applied to contributable earnings forC/QPP was 1.8 per cent for both employeesand employers from 1966 to 1986. Since then,they have risen 0.1 percentage points per year,reaching 2.6 per cent in 1994. These rates are

levied on wage and salary earnings betweenmaximum pensionable earnings ($34,400 in1994) and the basic exemption ($3,400 in1994).

UI premiums are charged on earnings belowa weekly maximum ($815 in 1994) which islinked to changes in the average industrialwage over an eight-year period. The employerrate is 1.4 times that for employees.1 Undercurrent arrangements, those working less

Summer 1995 Canadian Business Economics 5



than 15 hours per week who earn below 20per cent of the weekly maximum are exempt.

As Table 2 indicates, UI premiums have fol-lowed an upward trend since 1966. This re-flects rising unemployment, increases in self-financing requirements due to the generosityof benefits and, since 1990, the end of contri-

butions to the UI fund from general reve-nues.2 Current premium levels are more thanthree times higher than they were in 1966.There have also been significant cyclicalmovements in UI premium rates around theupward trend. Premiums peaked followingthe recession of the early 1980s, declining

once the debt accrued during the recessionhad been paid off. Rates rose once again withthe recession of the early 1990s in order to meetthe growing cumulative deficit in the fund.

Workers’ compensation is financed by pro-vincial levies on employer payrolls below aspecified maximum earnings per employee.The tax or assessment rate applied to this basevaries across industry rate groups.3 Theserates are generally set at levels sufficient tocover the projected costs of accidents incurredby the rate group in each year. Additional sur-charges may be added to cover costs of past

unfunded accidents. Average rates have risensubstantially in several provinces (Table 2).No province experienced a fall in its averageassessment rate between 1966 and 1994.

Quebec, Manitoba, Newfoundland, and On-tario also levy additional payroll taxes to helpfinance health care and/or postsecondary

Figure 1 Payroll Taxes as a Percentage of

Labour Income

Source: Canadian Economic Observer, Kesselman (1994).

Notes: Payroll Taxes are defined as total contributions to Workers’Compensation, CPP/QPP and UI and Provincial Payroll Taxes forHealth & Education. Labour income is wages, salaries andsupplementary labour income.

Table 1 Trends in Payroll Taxes as a Percentage of Total Government Revenue1

YearUnemployment

InsuranceWorkers’

Compensation

Canada/

QuebecPension

ProvincialPayroll Taxes2 All

1950 2.3 1.2 - - 3.5

1955 2.2 1.0 - - 3.2

1960 2.5 1.1 - - 3.6

1965 1.9 1.0 - - 2.9

1970 1.5 0.8 3.3 0.1 5.7

1975 3.0 1.2 3.0 0.3 7.5

1980 2.7 1.3 3.0 0.5 7.51985 4.6 1.4 3.0 0.9 9.9

1990 4.6 1.7 3.5 1.9 11.7

1992 5.9 1.6 3.8 1.8 13.1

1993 5.8 1.3 3.9 1.8 12.9

Sources: Canadian Tax Foundation (1993); Kesselman (1994); data from Statistics Canada.

1 Includes all three levels of government.

2 Refers to provincial payroll taxes brought in by Quebec, Manitoba, Ontario, and Newfoundland to finance health andeducation.

Per cent

0

2

4

6

8

10

12

26 2 9 3 2 35 38 41 44 47 50 53 56 59 62 65 68 7 1 74 77 80 83 86 89 92

6 Canadian Business Economics Summer 1995



education4. In Ontario, the Employer HealthLevy was introduced in 1990. It is applied at agraduated rate to total annual gross wages,salaries, and other remuneration paid by theemployer. In 1994, the rate was 0.98 per centup to a $200,000 annual payroll, and rose to amaximum of 1.95 per cent for payrolls over

$400,000.

Manitoba introduced a payroll tax in 1982.The initial tax rate was 0.75 per cent, whichrose to 2 per cent in 1987; however, the payrollexemption was also increased from $50,000 to$100,000. A payroll tax credit introduced in1991 can reduce the rate by up to 0.3 per centfor firms with qualifying training costs.

Quebec’s payroll tax was introduced with itsMedicare plan in 1970 at an original rate of 0.8

per cent. Now Quebec also has a business sur-tax on the payroll tax which stands at 15 percent of the payroll tax. In 1994, the effectivepayroll tax rate was 3.75 per cent.

Newfoundland brought in its payroll tax in1990 to fund health care and postsecondaryeducation. It was imposed at a rate of 1.5 percent on payrolls in excess of $300,000. In1992, the commercial property tax for schoolswas replaced by broadening the base for thepayroll tax. The exemption was lowered from$300,000 to $100,000 of the annual payrolland the rate was increased from 1.5 to 2 percent. The fishery, agricultural, and forestry in-dustries which were originally exempted,now are taxed at 1 per cent.

Table 2 Payroll Tax Rates, Workers’ Compensation Average Assessment Rates, EmployerUI and C/QPP Premiums, Provincial Payroll Taxes, 1966-941

Workers’ Compensation Average Assessment Rates Other Payroll Tax Rates

Nfld. P.E.I. N.S. N.B. Que. Ont. Man. Sask. Alta. B.C. UI2

CPP/

QPP3

Provincial Taxes

Que. Man. Ont.4 Nfld.

1966 1.62 n.a. 1.16 1.97 1.39 1.23 0.67 1.53 1.44 1.30 1.33 1.80 0 0 0 0

1971 1.62 n.a. 1.15 1.54 1.09 1.19 1.06 0.93 1.32 1.38 1.5 1.80 0.8 0 0 0

1976 1.42 n.a. 1.18 1.48 1.86 1.75 1.10 2.21 1.47 1.83 2.31 1.80 1.15 0 0 0

1981 1.49 2.44 1.26 1.38 2.10 1.66 0.89 1.31 1.60 2.52 2.52 1.80 2.63 0 0 0

1986 1.79 1.32 1.19 1.77 2.04 2.59 1.67 1.37 1.59 2.19 3.29 1.80 3.11 1.50 0 0

1991 2.92 1.95 1.66 2.04 2.32 3.20 2.25 1.63 1.85 1.82 3.54 2.30 3.55 2.25 .98/1.95

1.50

1992 3.00 2.00 1.98 2.13 2.50 3.16 2.16 1.65 1.98 2.02 4.20 2.40 3.75 2.25 .98/1.95

1.75

1993 3.23 2.22 2.28 2.19 2.75 2.95 2.16 1.60 2.04 2.16 4.20 2.50 3.75 2.25 .98/1.95 2.00

1994 3.18 2.07 2.54 2.15 2.75 3.01 2.12 1.67 2.13 2.42 4.30 2.60 3.75 2.25 .98/1.95

2.00

Sources: Association of Workers Compensation Boards of Canada; Kesselman (1994); Statistics Canada, Unemployment Insurance Statistics ,Health and Welfare Canada (1986).

1. In years where rates changed an average is reported.

2. Since 1971 this is 1.4 times the rate paid by employees. Pre-1971, employers and employees paid matching contributions which variedwith earnings.

3. Employees pay a matching rate. The 1994 Worker’s Compensation Assesment Rates are preliminary.

4. Rate varies by payroll size.




Table 3 Payroll Tax Rates by Province: Employer Contributions Only

Nfld. P.E.I. N. S. N.B. Que. Ont.1 Man. Sask. Alta. B.C.

1966 4.75 n.a 4.29 5.10 4.52 4.36 3.80 4.66 4.57 4.43

1971 4.92 n.a. 4.45 4.84 5.19 4.49 4.36 4.23 4.62 4.68

1976 5.53 n.a 5.29 5.59 7.12 5.86 5.21 6.32 5.58 5.94

1981 5.81 6.76 5.58 5.70 9.05 5.98 5.21 5.63 5.92 6.84

1986 6.88 6.41 6.28 6.86 10.24 7.68 8.26 6.46 6.68 7.28

1991 8.76 7.79 7.50 7.88 11.70 10.99 10.34 7.47 7.69 7.66

1992 9.60 8.60 8.58 8.73 12.85 11.71 11.01 8.25 8.58 8.62

1993 9.93 8.92 8.98 8.89 13.20 11.60 11.11 8.30 8.74 8.86

1994 10.08 8.97 9.44 9.05 13.40 11.86 11.27 8.57 9.03 9.32

1. Ontario uses 1.95 as the rate of the Provincial payroll tax for 1987-1993.

Table 4 Marginal Payroll Tax Rates, 1991

Nfld. P.E.I. N.S. N.B. Que. Ont. Man. Sask. Alta. B.C.

Employer Rates Only

Under CPP

maximum*

10.26 7.79 7.50 7.88 11.70 10.99 10.33 7.47 7.69 7.66

Under UI

maximum

(over CPP)

8.16 5.69 5.40 5.78 9.60 8.89 8.24 5.37 5.59 5.56

Under WCmaximum

(over UI, CPP)

9.45 1.95 1.66 2.04 5.87 5.15 4.5 1.63 1.85 1.82

Above all

ceilings

1.5 0 0 0 3.55 1.95 2.25 0 0 0

Combined Employer and Employee Rates

Under CPP

maximum1

15.08 12.61 12.32 12.7 16.53 15.81 15.16 12.29 12.51 12.48

Under UI

maximum

(over CPP)

12.98 10.51 10.22 10.6 14.43 13.71 13.06 10.19 10.41 10.38

Under WC

maximum

(over UI, CPP)

4.42 1.95 1.66 2.04 5.87 5.15 4.5 1.63 1.85 1.82

Above all

ceilings

1.5 0 0 0 3.55 1.95 2.25 0 0 0

1. Above CPP minimum insurable earnings and with sufficient hours and earnings to be covered by UI.




The combined employer payroll tax ratesfrom 1966 to 1994 are reported for each prov-ince in Table 3. In most provinces, these ratesdoubled or even tripled during the period. In1994, Quebec had the highest combined em-ployer rate, at 13.4 per cent (compared to 3.13per cent in 1966). The lowest rate was in Sas-

katchewan (8.57 per cent in 1994 compared to4.66 per cent in 1966).5

Impacts of PayrollTaxes on DifferentEmployee GroupsSince economic models typically focus on

decisions at the margin, marginal payroll taxrates are most relevant in evaluating the im-pacts of these taxes. Marginal payroll tax rates

differ from the tax rates reported in Tables 3(plus employee contributions) because UI,C/QPP, and workers’ compensation applyonly to earnings below a specified maximumand because of our use of an average assess-ment rate for workers’ compensation.6

The ceilings for UI, C/QPP, and workers’compensation, above which no tax ischarged, mean that the marginal (and aver-age) tax rates will eventually fall as a worker’searnings rise. In 1991, for example, the maxi-mum marginal combined employer tax ratefor an Ontario worker was 10.99 per cent forthose paid less than $30,500 who were UI in-surable. This fell to 8.89 per cent for thoseearning between $30,500 and $35,360 (i.e.,those whose additional wage income was tax-able under all payroll taxes except C/QPP),then to 5.15 per cent for those earning be-tween $35,360 and $42,000 (above the UImaximum insurable earnings), and finally to1.95 per cent for those earning more than$42,000. Marginal rates by earnings level foreach province in 1991 are reported in Table 4.

Table 5 draws upon 1991 data on CanadaPension Plan contributors to show the distri-

bution of the workforce (excluding Quebec)by marginal rate level. The annual payroll taxceiling for C/QPP was $30,500; in 1991, 64 percent of workers had earnings below this leveland consequently faced the maximum possi-ble marginal payroll tax rate (assuming theywere all UI insurable). Another 7.9 per centhad earnings between the C/QPP ceiling and

the (annual) UI ceiling of $35,360 and conse-quently had marginal rates equal to the sumof all tax rates excluding C/QPP. Workers’compensation ceilings vary by province. In1991, about 20 per cent of workers in mostprovinces were paid more than the workers’compensation ceiling and consequently had

the lowest possible marginal payroll tax rate.This minimum rate equaled the provincialpayroll tax rate (or 0 per cent in provinceswith no provincial payroll tax). In summary,about 20 per cent of workers faced the mini-mum marginal rate and 64 per cent faced themaximum.

Data from the 1989 Labour Market ActivitySurvey (LMAS) provide a similar picture. Be-tween 63 per cent and 66 per cent of workershad earnings below the 1989 C/QPP ceiling of $27,700.7 Estimates from the LMAS indicatethat 55-58 per cent faced the maximum mar-

ginal rate. Another 4-6 per cent were betweenUI and C/QPP ceilings and 82-84 per cent werebelow the workers’ compensation cutoff.

Table 5 Percentage of

Employees Above

Earnings Ceilings, 1991

Earnings

Ceiling

Per Cent

Above

Maximum

CPP $30,500 35.9

UI1 $35,360 28.0

Workers’ Compensation

Nfld. $45,500 9.0

PEI. $25,000 29.1

N.S. $37,300 19.9

N.B. $36,000 19.3

Ont. $42,000 21.1

Man. $38,000 17.5

Sask. $39,000 17.5

Alta. $40,000 21.2

BC. $45,800 16.5

Source: Based on data from Health and WelfareCanada (1991).

1. Full-year equivalent




The LMAS dataset can be used to determinethe composition of the groups facing differentmarginal tax levels and, consequently, it indi-cates where the impacts of the taxes are likelygreatest. As Table 6 indicates, the proportionof those facing the top marginal rate is rela-tively high for women, young workers, the

less educated, and those in the Atlantic Prov-inces. Those employed in non-union jobs,with small employers, in the retail trade andservice industries (especially accommodationand food), and in clerical, sales, or service oc-cupations are all substantially more likely tobe at the top rate than the average worker. Inshort, the low-wage sectors face the highestmarginal tax rates because of tax ceilings. It isthe workers in these sectors who bear thebrunt of these taxes8

Future Payroll TaxTrendsIn this section, we consider whether the up-

ward trend in payroll tax rates is likely to con-tinue. Each tax is examined in turn.

Canada Pension PlanThe increases in the C/QPP rates which

have occurred since 1986 are part of aplanned long-term rate increase required to

maintain the viability of the plan. The rateschedule projected sees combined employer-employee rates rising gradually from 5.4 percent in 1995 to 12.80 per cent in 2021 (Cana-dian Institute of Actuaries, 1993).

While this projection suggests substantial fu-ture rate increases, it is not certain that thisrate schedule will actually be realized. At onelevel, the projection is based upon assump-tions regarding future growth in wages andsalaries, life expectancy, and other variables,which are themselves uncertain. At a more

fundamental level, the projection assumesthat program benefit levels and eligibility willremain the same. Less drastic rate increaseswould be required if the program were to bejudged too costly, with the result that substan-tial reductions in benefits and eligibility oc-curred.

Workers’ CompensationThe likely course of future workers’ compen-

sation assessments varies substantially from

province to province, largely because of dif-ferences in financial circumstances. Data col-lected by the Financial Executives Institute of Canada (1995) for 1993 show that Saskatche-wan ran a surplus while its pension liabilitieswere actually overfunded (Table 7). Threeother provinces had surpluses but had plans

with unfunded liabilities. Six provinces haddeficits in 1993; of these, Ontario, Quebec,and Nova Scotia also had substantial un-funded liabilities.

These financial positions suggest that as-sessment rates in Ontario, Quebec, and NovaScotia may have to rise substantially in the fu-ture. In fact, since 1984, Ontario’s assessmentrates have included a surcharge intended toamortize the unfunded liability by 2014. TheWorkers’ Compensation Board of Ontario(1992) suggested that its financial positionwould require the average assessment rate to

rise to 3.98 per cent by 1996 and remain atthat level for the rest of the amortization pe-riod. However, because of the slow economicrecovery in Ontario since the 1990-91 reces-sion, the assessable payroll has not grown asfast as projected and the post-1995 rate is nowlikely to be higher than the 3.98 per cent tar-get.9. Assuming a sustained one per cent risein average assessment rates seems reasonablefor Ontario. The problem is less severe in theother two provinces with large unfunded li-abilities and their required rate increases arelikely to be smaller than those in Ontario.10

The other two provinces with deficits, BritishColumbia and Prince Edward Island, mayalso require rate increases but, given theirrelatively small unfunded liabilities, the re-quired increases could be quite small or sim-ply temporary. Some rate reductions mayeventually be possible in those provinces run-

ning su rpluses; however, Alberta, NewBrunswick, and Newfoundland do have sub-

stantial unfunded liabilities and rates may notcome down for some time. Only Saskatche-wan appears to be in a position to immedi-ately reduce rates.11

Unemployment InsuranceUnemployment insurance premiums are

currently at a cyclical high and can be ex-pected to fall as the debt accrued during thelast recession is paid off and a cumulative sur-plus accumulates. The combined employer-




Table 6 Percentage of Employees at the Maximum Payroll Tax Rate

by Selected Characteristic, 1989

All Employees 58.7

Men 47.4

Women 71.9

Age

16-24 72.2

25-34 61.3

35-44 48.7

45-54 52.6

55-64 54.2

Province

Newfoundland 67.2

Prince Edward Island 76.3

Nova Scotia 65.3

New Brunswick 67.2

Quebec 62.9

Ontario 55.6

Manitoba 63.9

Saskatchewan 62.6

Alberta 57.4

British Columbia 53.0

Education

Less than 8 years 69.0

Some secondary 63.5

High School 66.3

Some Post-secondary 63.9

Post-secondary Certificate 55.8

University Degree 38.0

Trades certificate 56.4

Employment Status

Full-time 57.9

Part-time 62.4




Union Status

Union 50.2

Non-union 63.3

Industry

Primary 52.3

Manufacturing 53.8

Construction 57.3

Transportation, Communications andother Utilities

43.1

Trade 68.2

Finance 64.0

Services 63.3

Public administration 47.4

Occupation

Managerial/ Professional 45.3

Clerical 75.9

Sales 64.3

Service 70.5

Primary 61.7

Processing 56.3

Metals, Machining 49.4

Product Fabrication 66.1

Mechanics/Repair 41.7

Construction 52.3

Transport 52.7

Material Handling/ Craft 57.9

Firm Size (Number of employees)

Less than 20 70.2

20-99 64.6

100-499 58.5

500 or more 48.1

Source: Based on Labour Market Activity Survey data.




employee rate is currently projected to fallfrom 7.2 per cent in 1995 to 6.72 per cent in1996 and 5.28 per cent in 1997 (Freeman,1995). This projected rate decline is slowerthan that which would have occurred if the1995 federal budget had not required the fundto build a cumulative surplus of $5 billion

prior to the next downturn rather than merelyeliminate the cumulative debt.12 Rates arealso unlikely to return to pre-recession lowsdue to changes in financing made in 1990which eliminated general revenue contribu-tions to the UI fund. Through the 1980s, theshare of UI benefits funded out of generalrevenues was roughly 25 per cent (Employ-ment and Immigration Canada, variousyears).

Since all costs must now be covered by pre-miums, this change in financing will lead to asubstantial, permanent upward shift in pre-

miums. However, this upward pressure willbe partly offset by cuts to benefits made in1993 and 1994. The present combined rate of 7.37 per cent stands out markedly comparedto the pre-recession low of 4.35 per cent. Add-ing 25 per cent to the latter figure (i.e., the for-mer general revenue contribution) suggeststhat rates could fall as low as 5.44 per cent,which is near the currently projected 1997rate. This may prove to be the new rate floor.

Naturally, if more cuts are made to UI bene-fits, and these savings are used to reduce pre-miums, then payroll taxes will fall further. Hu-man Resources Development Canada (1994)has provided estimates of savings under a va-riety of reforms to UI. The elimination of Re-gional Extended Benefits would provide the

single largest saving of the options costed($4.5 billion). This would reduce combinedemployer-employee premiums by 1.63 per-centage points. This is likely an upper boundas it assumes the largest possible cuts and thatthe savings are realized to their full extent aspremium reductions. It should be noted thatthere could be other competing uses for thesesavings, including new labour adjustment orincome support programs.

Provincial Payroll Taxes

Substantial provincial debt (estimated at$155 billion in 1994) will likely create pres-sure for new payroll taxes or for further risesin existing ones. Adding to these pressures arethe cuts to federal transfers brought about bythe replacement of the Established ProgramFinancing and the Canada Assistance Planwith the Canada Social and Health Transfer,as announced in the 1995 federal budget.

Payroll taxes may be an attractive source of additional revenue from a political perspec-

Table 7 Financial Positions of the Workers’ Compensation Boards,1993(millions of dollars)

Deficit/Surplus Unfunded Liability Assessable Payroll

Newfoundland 16 -119 2,921

Prince Edward

Island

-1 -10 547

Nova Scotia -12 -461 4,640

New Brunswick 17 -66 N.A.

Quebec -65 -3,516 61,900

Ontario -504 -11,532 83,423

Manitoba 19 -72 5,856

Saskatchewan 4 62 6,860

Alberta 299 -277 23,010

British Columbia -94 -191 37,523

Source: Financial Executives of Canada Institute (1995)




tive. Witness, for example, the relativelymuted response to the introduction of the Em-ployer Health Tax in Ontario in 1990, com-pared to the introduction of the Goods andServices Tax a year later. The existing provin-cial taxes are nominally earmarked for par-ticular expenditures (health in Ontario and

Quebec, and health and education in Mani-toba and Newfoundland). This makes ittempting to argue that pressure for rises inthese taxes may be created by increases inhealth spending, as the population continuesto age. However, it is probably more accurateto view these taxes as just another source of general revenue and, therefore, to think of fu-ture trends as being tied to the provinces’overall fiscal situation.

Quebec will have a new one per cent payrolltax in 1996 to raise funds for training. Levy-grant schemes to fund training, which typi-

cally use a payroll tax as the levy, have alsobeen discussed as a possibility for Ontario inthe recent past although this may end with therecent change in government.

Moreover, relative to most types of taxation,a major attraction of payroll taxes is their rela-tively low cost of administration and compli-ance. Once the tax is in place and the collec-tion mechanisms established, raising furtherrevenues from rate increases is a relatively

simple matter. Ontario, for example, reportedthat for the fiscal year 1992-93, the adminis-trative costs of its payroll taxes amounted to0.38 percent of the associated tax revenueswhich are substantially lower than for mostother taxes (Kesselman, 1994: 182).13

Another attraction of payroll taxation for pro-vincial governments is the ability to shiftsome of the tax burden onto the federal gov-ernment. This effect operates through twochannels. First, federal employees and agen-cies pay provincial payroll taxes in an effort to

be consistent with other employers eventhough, technically speaking, the Crown isimmune from taxation. Second, for federal in-come tax purposes, incorporated and unin-corporated businesses are able to deduct pay-roll taxes paid when calculating their incomefor the purposes of remitting federal incometax.14

SummaryIn conclusion, it seems likely that the long-

term upward trend in payroll tax rates willcontinue. This will be driven by rising C/QPPcontribution rates and reinforced by increasesin other payroll taxes, particularly increases in

workers’ compensation assessments in someprovinces. A possible source of relief is unem-ployment insurance, with premiums declin-ing as the expansion continues, as the debt in-curred during the last recession is paid off,and as UI benefits are cut and eligibility re-quirements tightened. The combined effectsof falling unemployment and UI reform couldmore than offset rises in other payroll taxes inthe near term. However, as the economymoves through the business cycle, UI premi-ums will stabilize and then rise, while C/QPP

premiums will continue to climb, with the re-sult that the overall payroll tax rate will prob-ably increase.

The Impact of PayrollTaxation

Economist and BusinessViews

In simple economic models, a payroll taxlevied on employers lowers labour demand,

inducing a decline in wages which will par-tially or, if labour supply is perfectly inelastic,wholly offset the effects of the tax on employ-ment.15 Alternatively, a payroll tax levied onemployees will decrease labour supply, lead-ing to an increase in wages which will againcounter the initial effects of the tax on em-ployment. In either case, the employer’s wagecosts (tax inclusive) will rise while the net-of-tax wage received by workers will fall. Em-ployment will decline in either case (if neithercurve is perfectly inelastic) but no unemploy-

ment will be generated by the tax since work-ers will remain on their labour supply curveand choose to work less at the lower net-of-tax wage. Thus, we have the result, familiar toeconomists, that the effect of the tax is thesame regardless of whether it is levied on theemployer or the employee.

Economists have placed the greatest empha-sis on whether business or labour bears theburden of the tax. The answer depends upon




how responsive the market wage rate is to theintroduction of the tax. If an employer payrolltax eventually leads to a large offsetting fall inwages, then workers will bear much of theburden of the tax. The same result is obtainedif an employee payroll tax induces only asmall increase in the wage. In either case, the

labour supply and labour demand elasticitiesare the critical parameters in determining theextent to which wages change and the degreeto which employment will change in responseto the wage change.

In Canada, especially for the approximately35 per cent of the workforce covered by col-lective agreements, the simple model of sup-ply and demand may not accurately reflecthow wages are determined. Non-cooperativebargaining theory suggests that the effects of payroll taxes in a collective bargaining contextwill be in the same direction as in the simple

model (Wilton and Prescott, 1993). Employ-ers will attempt to share the cost of an em-ployer payroll tax with their unionized work-ers by trying to negotiate smaller wageincreases. Unions will try to shift the em-ployee portion of a payroll tax to employers bynegotiating higher wage increases.

There is a contrast between what might becalled the “business view” or “public view” of the effects of payroll taxes and the econo-

mist’s view. Business groups view payrolltaxes as a “tax on jobs” and emphasize theiremployment effects. Quoting the CanadianChamber of Commerce: “Rising payroll taxesleave companies with little choice other thanto cut jobs or hire fewer people.” (Globe and

Mail, February 10, 1994). This view also has afollowing in government circles. Paul Martinhas referred to payroll taxes as “a cancer onjobs” (Freeman, 1995). Statements concern-

ing the 1994 roll-back of UI premiums (Fi-nance Canada, 1994) and arguments made

for UI reform by Human Resources Develop-ment Canada (1994: 42-43) also emphasizeemployment effects.

Can these two views be reconciled? Stressingemployment impacts makes most sense toeconomists if labour demand and supplycurves are elastic or if wage rates are rigid andthe tax is levied on employers. In the formercase, any wage change induced by the tax willresult in relatively large changes in employ-ment. In the latter case, labour demand falls

with an increase in the tax but, as there is nooffsetting decline in wages, employment fallsin the rigid-wage sector.

Furthermore, the fall in employment in therigid-wage case has the potential to increaseunemployment since those losing their jobswill still want to work at the constant wage.The “potential” proviso is necessary since thedisemployment effect may be specific to therigid-wage sector. The existence of sectorswith flexible wages in which those losing jobsin rigid-wage sectors can work will mean thatsome of these unemployed workers will enterthe flexible-wage sector, driving its wageslower than the payroll tax would alone, untilthe unemployed are absorbed. Unemploy-ment may still occur, but it will be of the“wait” variety, as workers queue for jobs inthe rigid-wage sector.16

If there is no other flexible-wage market towhich workers losing their jobs in the rigid-wage market can move, as is the case in mini-mum-wage labour markets, the effects of pay-roll taxes will be solely on employment andwill definitely create unemployment.17 Therigid-wage case has received little attentionfrom economists when examining payrolltaxes despite the common use of this assump-tion in the macroeconomics literature.18

It could also be argued that the business

view reflects a focus on the short-run effectsof payroll taxes. This can be a legitimate viewif the wage adjustments stressed by econo-mists are slow, or if those employers that nowbear the costs of the tax may not be around tobenefit from any longer-term wage declinewhich the tax may bring. Dahlby (1993) notesthat there is support in the empirical literaturefor slow dynamic adjustment; some evidenceon this point is presented in the next section.It is also possible that business is not con-cerned with employment effects per se, but

that emphasizing them is the best way to gar-ner support against a policy which creates awelfare loss for employers via reduced pro-ducer’s surplus and added costs of adminis-tering the tax.

Another surprising result from the business

point of view is the economist’s predictionthat it is irrelevant to the tax’s eventual impactwhether it is levied on employers or employ-ees. It is hard to imagine, for example, that




employer groups would be indifferent to apayroll tax levied on their wage costs or to asurtax on wage income levied through the in-come tax system. Yet the simple model sug-gests that the effects would be much thesame. Again, wage inflexibility may help toreconcile the two positions. In the presence of

rigid wages above market-clearing levels, atax on labour income will have no effect onemployer wage costs or on employment; i.e.there will be no costs to employers.

Offsetting BenefitsSome payroll taxes provide a benefit to the

employer which may, to some degree, offsetthe cost of the tax itself. Workers’ compensa-tion premiums are paid by employers to fi-nance payments to injured workers. This sys-tem provides two possibl e benefits to

employers: first, workers give up the right tosue their employers in the event of a work-place injury and, second, since workers knowthey will be compensated in the event of aninjury, this reduces any compensating wagedifferential needed in more dangerous jobs.19

It is clearly inappropriate, then, to treat the en-tire workers’ compensation premium as a costto employers. Indeed, Vaillancourt and Mar-ceau (1990) find that the effects of workers’

compensation assessments are in the oppo-site direction to the effects of other payroll

taxes. They suggest that offsetting benefitsmay help explain this.

The offsetting benefit argument can also beapplied to unemployment insurance andC/QPP premiums. The greater security pro-vided by UI and the post-retirement pensionbenefits through C/QPP may make workerswilling to accept lower wages, thus offsettingthe employer portion of the tax. Only the pro-vincial payroll taxes have no apparent offset-ting benefit.

The lack of UI experience rating and the im-perfect experience rating systems in workers’compensation programs weaken the benefit-tax link for individual employers. The pay-as-you-go nature of C/QPP has the same effect.Unemployment insurance may allow employ-ers with unstable labour demand patterns topay a lower wage than would be necessarywithout UI. Consequently, the net effect of UIon employment could actually be positiveamong such employers. Of course, firms with

stable work patterns receive no such offset-ting benefit and, given the lack of experiencerating, end up subsidizing other employers.

These arguments suggest that payroll taxeswith offsetting benefits should be analyseddifferently from other payroll taxes. However,the differences between the two types of taxesmay not be as stark as it first seems. In prac-tice, the effect of a change in tax rates may besimilar between the two types of taxes eventhough the consequences of introducing thetwo different types of tax from scratch may bevery different. This is because the link be-tween a change in the tax rate and changes inoffsetting benefits of the program funded maybe weak. For example, current increases in theworkers’ compensation assessments to payfor unfunded benefits awarded in the pastprovide no offsetting benefit to current em-ployees. Similarly, increases in C/QPP premi-

ums for today’s young workers to pay pensionliabilities of tomorrow’s retiring baby boom-ers provides little offsetting benefit for the for-mer. Reforms which strengthen links betweentax rate changes and benefit changes will re-duce the distortionary effects of these taxes.

The introduction of a payroll tax may pro-vide another type of offsetting benefit to em-ployers if it, or the program financed by thetax, replaces an existing employer tax or dis-

places a fringe benefit commonly provided byemployers. The Ontario Employer Health Tax,for example, replaced Ontario Health Insur-ance Plan premiums which were commonlypaid by employers (as a fringe benefit).20

Ceilings and FloorsAn important feature of UI, C/QPP, and

workers’ compensation taxes is the existenceof a ceiling above which earnings are nottaxed. Such ceilings lead to the differences inmarginal tax rates recorded in Table 4. At the

extreme, the marginal employer tax rate on anextra hour worked by a worker with earningsabove the ceilings of all payroll taxes is zerocompared to at least 8.57 per cent and as highas 13.4 per cent for a low-wage worker in1994. These ceilings will make higher-wagelabour relatively cheaper at the margin, withthe consequence that wage and employmenteffects will be more severe for lower-wageworkers. This can create a bias in favour of hiring highly-paid workers. It also provides an




incentive to have current workers work morehours rather than hire additional workers andso favors full-time over part-time employ-ment. The effects of ceilings also make thesetaxes more regressive.

These impacts will be less important for a taxwith offsetting benefits. With UI, for example,earnings above the insurable ceiling are nottaxed but, also, are uninsured. Consequently,wages for higher-paid workers may be greaterthan if insurance applied to all earnings.

The existence of minimum pensionable andinsurable earnings under C/QPP and UI, andthe 15 hour per week requirement for UI, cre-ate biases in the opposite direction to ceilings.However, figures reported in an earlier sectionsuggest that ceilings are empirically more im-portant than floors.

Exempt or Low-Tax GroupsIf certain groups are exempt from the payroll

tax — for example, the self-employed or somesmall businesses under provincial payrolltaxes — then such taxes will tend to depresswages (incomes) in the exempt group as wellas the covered groups. In effect, the fall in la-bour demand in the covered group representsa fall in labour demand for all workers of thatparticular skill level regardless of group andwill therefore lower wages across groups.

Empirical Work onthe Effects of PayrollTaxationDahlby (1993: 133), in reviewing the litera-

ture on the incidence of payroll taxes, notesthat “empirical studies....suggest that labourbears over 80 per cent of the employer payrolltax burden in the long run”.21 The usual find-ing is that, consistent with the simple model,

wage rates adjust to offset payroll taxes result-ing in a significant shifting of the tax to la-bour.22 This result suggests that it is incorrectto ignore the wage effects of payroll taxes, asthe business view tends to do.

A number of Canadian studies have exam-ined the “shifting” issue. Using collectiveagreements data in Quebec, Vaillancourt andMarceau (1990) found the standard result thatincreases in general payroll taxes reducedwage growth with some of the tax, shifted to

workers. Increases in workers’ compensationassessments, however, actually appeared toincrease wage growth. The authors use an off-setting-benefit argument to explain this latterresult, noting the close tie between tax ratesand injury costs for workers’ compensation.Beach, Lin, and Picot (1995) estimate a labour

demand equation derived from a CES produc-tion function on pooled provincial employ-ment and wage data for 1966-1993. Their re-sults are consistent with full shifting of theemployer payroll tax to workers through wagereductions. Wilton and Prescott (1993) exam-ine the impact of income, sales, and payrolltaxes on wage costs using data on private-sec-tor collective agreements from 1979-1992. Us-ing several alternative specifications, theyfind that higher employer payroll taxes actu-ally increase real wages. This is contrary to

the standard result from the simple supply-demand model or the bargaining framework,which lies behind Wilton and Prescott’sspecification.

The macroeconomic-labour literature,which aims to explain unemployment pat-terns by estimating aggregate labour marketmodels, can also be used to assess the effectsof payroll taxes.23 In these models, employ-ment is negatively related to real wage costsand real wages depend negatively on the un-employment rate and positively on the“wedge”, which is the difference between thewage cost to the employer (including tax) andthe net wage of the worker. The payroll tax en-ters the model as part of the wedge. As in thesimple model, a rise in payroll taxes increasesthe wedge, and therefore tax-inclusive wagecosts, and this lowers employment. However,unlike the simple model, there is no assump-tion that employment is determined along thelabour supply curve. Consequently, risingpayroll taxes raise wage costs and lower em-ployment, but can also lead to unemploy-ment. These unemployment effects of payrolltaxes create downward pressure on wagerates which partially offsets the original effectof payroll taxes on wage costs.

Our calculations based on Bean, Layard, andNickell (1986) indicate that a one per cent risein payroll tax rates results in long-run in-creases of 0.5 per cent in real wage costs anda -0.2 per cent decrease in employment. Bean,Layard and Nickell’s own simulations suggest




that increases in the wedge raised the unem-ployment rate by 1.34 points between theearly-1960s and the early 1980s — a period inwhich the unemployment rate rose by about4.5 percentage points. Of the 1.34 point riseattributed to the wedge, our calculations sug-gest that about 25 per cent can be attributed to

the rise in payroll taxes alone. Simulations byKeil and Symons (1990) indicate that 0.5points of the 3 percentage point rise in unem-ployment between 1981 and 1985 was due toincreases in the wedge. Again, based on ourcalculations, about 22 per cent of this 0.5point increase can be attributed to the rise inpayroll taxes.

Both Bean, Layard, and Nickell and Keil andSymons adopt dynamic specifications. Re-sults from the former show a mean lag for em-ployment adjustment of 1.5 years. In the Keiland Symons model where dynamics are

slower, while a change in the payroll tax doesnot affect long-run employment, it can havelingering negative effects as the economymoves toward equilibrium. This pace of la-bour market adjustment lends some credenceto the business view of the payroll-tax effect.Both models, though, are consistent with theeconomist’s story in the long run.

In general, then, studies support the stand-ard model suggesting that the effects of pay-

roll taxes on employer wage costs are moder-ated by changes in wage rates. Only the studyby Wilton and Prescott is at odds with thisconclusion. This suggests that an importanteffect of the rise in payroll taxes, then, haslikely been slower wage growth. In addition,the macroeconomic-labour studies suggestthat payroll taxes will reduce employmentand raise unemployment though these effectsmay be transitional. Another interesting point

arising from this literature and the study byWilton and Prescott is that the impact of a

payroll tax on labour markets can be expectedto be much the same as other taxes (i.e. salesand income taxes) which drive a wedge be-tween the supply and demand prices for labour.

Finally, the literature has also considered thedistributional impact of Canadian payrolltaxes. Recent work by Gillespie, Vermaeten,and Vermaeten (1995) has found that payrolltaxes were generally regressive with respectto family income from 1951-1969 but by 1988,they had acquired an inverted U-shape. Pay-

roll taxes were progressive until about the 5thdecile, remained proportional for the nexttwo, and then began to decline. Most likely,the increased progressivity of the taxes at thelow to middle income ranges does not reflectchanges in tax design but rather the growth inthe importance of transfers as an income

source in the lowest deciles.

New Estimates of theImpact of PayrollTaxationMeasuring the impact of payroll taxes on em-

ployment and wages requires estimates of la-bour demand and supply elasticities. How-ever, there are no generally acc eptedestimates of these elasticities for the Canadian

economy. Past studies report large differ-ences.24

Table 8 presents estimates of the wage andemployment effects of a one per cent rise inmarginal payroll tax rates based upon variouscombinations of labour supply and labour de-

mand elasticities. The elasticities used arereasonable by standards of the literature, withthe exception of the -2.5 labour demand elas-ticity (cases IV and VII) used by Keil and Sy-mons (1990), which is provided for contrast.As the labour demand and supply elasticitiesincrease, the effects of payroll taxes on em-ployment increase; however, the results showthat for reasonable labour demand elastici-ties, these effects are not substantial.

The table also illustrates the sensitivity of thewage-shifting results to even the small rangeof elasticities considered. A labour demandelasticity of -0.3 and labour supply elasticityof 0.15 implies that 33 per cent of the tax riseis realized as a rise in labour costs and the restas a fall in employee pay. Elasticities of -0.15

and 0.5 more than reverse this result with theemployer cost share at 77 per cent and the em-ployee burden at 23 per cent. Therefore, the

wage effects of the taxes may be particularlyhard to predict.25

The top panel of Table 9 presents the long-run effects of a one per cent rise in average(rather then marginal) payroll taxes, as de-rived from the Bean, Layard, and Nickellmodel. These are closest to those obtained for




case VI in the previous table (for a labour de-mand elasticity of -0.3 and a labour supplyelasticity of 0.5). The actual long-run labourdemand elasticity in Bean, Layard, andNickell is -0.42.

The bottom panel of Table 9 reports long-runimpacts based upon our own estimates usingannual data for the period 1958-1992. Webase our model on Bean, Layard, andNickell’s aggregate labour market model so asto facilitate comparison as well as make use of a standard and accepted modelling approach.This involves specification of a real wageequation and an employment equation (bothin log-linear form). In the model, the realwage is explained by the wedge (includingpayroll taxes), an indicator of labour markettightness (the employment rate), the capital-labour ratio, and the unemployment insur-ance coverage rate. Employment depends

upon the real wage, GDP, and the capital-la-bour ratio. Both equations include quadratictrend terms and allow for dynamic adjust-ment by including a one-period lag of the de-pendent variable as a regressor. Modificationswere made to reflect the current article’s inter-est in payroll taxes.26 Both level and first-dif-ference versions of the model were estimatedusing two-stage least squares.27

The results broadly parallel those of Bean,

Layard, and Nickell. The coefficient on thewedge was positive and reveals that there is a

positive relationship between the averagepayroll tax rate and real wages; however, thiscoefficient is only marginally significant in thelevels version and insignificant in the first dif-ferences version. Moreover, there is a negativeand significant relationship between the realwage variable and changes in employment

suggesting that increases in the average pay-roll tax rate can lead to a reduction in employ-ment via changes in the real wage. The long-run labour demand elasticity was -0.56.28

These results suggest that the employmenteffects of payroll taxes are non-trivial. Solvingthe estimated equations for the long-run ef-fect of a one per cent rise in the average pay-roll tax rate on employment indicates a rise inreal wage costs of 0.56 per cent and a decreasein employment of 0.32 per cent in the long run(Table 9). The wage received by workerswould fall by 0.44 per cent. Using 1994 as the

base year, in the long run about 40,600 jobswould have been lost by such a policy action.

We emphasize that there are a wide range of estimates of the actual magnitude of the joblosses from payroll taxation because, ulti-mately, the size of the losses depends on esti-mates of labour demand and supply. At thesame time, the overall evidence is consistentwith the view that payroll taxes reduce em-ployment. The fact that since the 1960s the

persistent upward trend in unemploymenthas been accompanied by rising payroll tax

Table 8 Estimated Effects of a 1 Per Cent Rise in Marginal Payroll

Tax RateVarious Labour Supply and Demand Scenarios

I II III IV V VI VII

Elasticity Assumptions

Labour Supply 0 0.15 0.15 0.15 0.50 0.50 0.50

Labour Demand all -0.15 -0.3 -2.5 -0.15 -0.3 -2.5

Effects of a 1 Per Cent Rise in Marginal Payroll Tax Rate

Employer Wage

Costs/share

0 0.50 0.33 0.06 0.77 0.63 0.17

Worker Wage/share -1 -0.50 -0.67 -0.94 -0.23 -0.37 -0.83

Employment 0 -0.08 -0.10 -0.14 -0.12 -0.19 -0.42

Jobs1 (000s) 0 9.5 12.7 18.2 14.6 23.9 53.0

1. Assumes employment before tax increase of 12,700,000.




rates is probably not an entirely coincidentalrelationship.

ConclusionPayroll taxes have grown to become the third

most important source of government reve-

nue after personal income taxes and con-sumption taxes. Given the unfunded liabilityproblems in C/QPP and the workers’ com-pensation schemes in some provinces as wellas fiscal problems at both the provincial andfederal level, governments may attempt tofurther exploit this tax base. In addition, therelatively hidden nature of employer payrolltaxes provides a further incentive for their usegiven the current climate of resistance to vis-ible tax increases.

Payroll tax rates vary across provinces andindustry groups because of the number of dif-ferent taxes as well as differences in rates. Wefound that when the composition of thegroups facing tax levels is examined, low-in-come groups faced higher marginal rates.Women, young workers, those with less edu-cation, employees in retail, clerical, and serv-ices, and in small business generally, arelikely to be at high marginal rates and conse-quently bear many of the employment andwage costs these taxes may produce.

The empirical literature suggests that payrolltaxes raise employer wage costs and have im-

portant negative consequences for take-homewages. There is also some evidence of nega-

tive employment effects. In estimating the ef-fects of payroll taxes, we emphasize that thesize of these impacts depends on the labourdemand and supply elasticities utilized. How-ever, there is no consensus estimate of theseelasticities. Our own estimates imply that aone per cent rise in the average tax rate will

lead to a 0.56 per cent rise in wage costs and a-0.32 per cent fall in employment.

Notes

* The authors wish to acknowledge the helpful com-

ments of Jon Kesselman and two anonymous referees.

1. Prior to 1971, premiums differed by earnings class.

The 1966 figure reported in Table 2 is an average.

2. The 1994 system is more generous than the 1966

system in terms of level and duration of benefits.

Generosity, however, peaked in the 1970s.

3. In some provinces, partial experience rating allows

rates to vary among firms within an industry rate

group as well.

4. These taxes are examined in detail in Kesselman

(1994).

5. The summed rates in Table 3 apply to wage and

salary earnings below the maximums specified for

workers’ compensation, C/QPP, and UI. Rates ap-

plicable to other types of labour income, such as

fringe benefits, differ as UI and C/QPP exclude these

while the provincial taxes do not. In each province,

employee contributions represented 5.67 per cent of

total labour income in 1994.

6. The use of average workers’ compensation assess-

ments is unavoidable if figures are to be reported insummary form. Quebec, for example, had 341 indus-

try rate groups in 1993. Note, however, that the use

of an average rate masks substantial variation in

Table 9 Estimated Effects of a One Per Cent Increase in Average

Payroll Tax Rate

Bean, Layard, and Nickell1

Wage Costs 0.5

Employment -0.2

Job Loss2

(000s) 25.4Di Matteo and Shannon3

Wage Costs 0.56

Employment -0.32

Job Loss2 (000s) 40.6

1. Based upon estimates reported in Bean, Layard and Nickell (1986)

2. Assumes employment before tax increase of 12,700,000.

3. Details of estimates available from the authors.




rates between industry groups. For example, work-

ers’ compensation assessment rates ranged from

0.15 per cent to 23.6 per cent across Ontario rate

groups in 1992.

7. The range arises from the use of alternative meas-

ures of earnings.

8. Again, the picture, especially by industry, is impre-

cise because of the use of average assessment rates

for workers’ compensation.

9. The possibility of extending the amortization period

to 2024 or abandoning full funding for 70 per cent

funding were also examined. These options would

still require rate increases, but on a smaller scale

than those necessary under the current plan.

10. The unfunded liability was 9.8 per cent of assessable

payrolls in Nova Scotia, 5.7 per cent in Quebec, and

13.3 per cent in Ontario in 1993.

11. This discussion refers to rate adjustments required

to clear past liabilities. Considerable uncertainty

with respect to these future rate projections is intro-

duced by changes in future accident rates and bene-

fits. The recent introduction of a wage-loss systemin Ontario, for example, makes future benefits for

permanent disability pensions less predictable.

12. The changes in the budget will tend to dampen cy-

clical rate swings requiring rates to be higher during

upswings and lower during recessions than was the

case in the past.

13. On the other hand, Vaillancourt (1986: 83) argues

that the total administrative and compliance costs

of the personal income tax and payroll tax system

in Canada account for about 6.9 per cent of tax reve-

nues collected.

14. The federal government has moved to curtail the

deductibility of these provincial payroll taxes.

15. Dahlby (1992) provides a thorough review of the

theoretical literature on payroll tax impacts.

16. See Mincer (1976) for a model with this type of

unemployment.

17. This argument, combined with the observation that

marginal payroll tax rates are highest for low-wage

workers, suggests that employment impacts may be

especially large for these workers.

18. Mitchell (1993) provides a recent survey of the wage

rigidity literature.

19. See Meng (1989) for measures of compensating dif-

ferentials for dangerous jobs in Canada.

20. Dahlby (1993: 81) notes that 65 per cent of premi-

ums were paid by employers.

21. Hamermesh (1993: 172), however, is skeptical of

these “not very satisfactory studies”.

22. The supply-demand model suggests that an em-

ployer payroll tax should decrease net-of-tax wages

(or slow wage growth) while a tax on employees

should increase wage growth.

23. Layard, Nickell, and Jackman (1991) and Lindbeck

(1993) provide recent examples.

24. For example, Woodland (1975) reports labour de-

mand elasticities by industry that range from -0.351

(finance) to -0.009 (trade) while Merrilees (1982)

reports small labour demand elasticities that are ac-

tually positive for certain age groups. The long-run

labour demand elasticity for Canada in the Keil and

Symons (1990) model is -2.5 compared to -0.42 in

Bean, Layard, and Nickell (1986). Beach, Lin, and

Picot (1995) provide an estimate of roughly -.3. Hum

and Simpson (1991) summarize the Canadian la-

bour supply literature reporting that uncompen-

sated labour supply elasticities are almost always

below 0.5 with a concentration of studies around

0.4-0.5 and another group concentrated in the 0.1-

0.2 range. Some studies have even reported negative

labour supply elasticities.

25. Hamermesh (1993) concludes that near full-shifting

to labour is a reasonable assumption on the basis of

low labour-supply elasticity estimates (mainly for

the United States).

26. Specifica lly, two changes were made: (1) the

“wedge” was separated into payroll tax and non-

payroll tax components and (2) rather than model

the change in the real employer wage as a function

of the level of the wedge, we considered it more

natural to express the level of the real wage as func-

tion of the level of the wedge. Additional detail re-

garding the specification, data sources, and

estimation are available from the authors.

27. This is in recognition of the simultaneity of the two

equations. Employment depends upon the real wage

and the real wage depends upon employment

through the employment rate.

28. These results from Table 9 are from the levels version

of the model because it is not possible to simulate

long-run effects from the first- differences version of

the model.

References

Beach, Charles, Zhengi Lin, and Garnett Picot(1995) “The Employer Payroll Tax in Can-ada and its Effects on the Demand for La-bour,” paper presented at the Conferenceon Transition and Structural Change inthe North American Labour Market, JohnDeutsch Institute and the Industrial Rela-tions Centre, Queen’s University, King-ston.

Bean, C., R. Layard, and S. Nickell (1986)“The Rise in Unemployment: a Multi-country Study,” Economica, Vol. 53, pp.S1-S22.

Canadian Institute of Actuaries (1993) Cana-dian Retirement Income Social Security

Programs, Report of the Task Force on So-cial Security Financing (Toronto).

Canadian Tax Foundation (1993) The Na-

tional Finances (Toronto).




Dahlby, Bev (1992) “Taxation and Social In-surance,” in R.M. Bird and J. Mintz (eds.)Taxation to 2000 and Beyond, CanadianTax Paper No. 93, Canadian Tax Founda-tion, pp. 110-165.

Dahlby, Bev (1993) “Payroll Taxes,” in A.Maslove (ed.) Business Taxation in On-

tario (Toronto: University of TorontoPress), Chapter 3.

Employment and Immigration Canada (vari-ous years) Annual Report (Ottawa: Sup-ply and Services Canada).

Finance Canada (1994) Proposed Changes to

the Unemployment Insurance Program,

Ottawa.

Financial Executives Institute of Canada(1995) Workers’ Compensation in Can-

ada 1993 (Toronto).

Freeman, Alan (1995) “UI Premium Cuts on

the Horizon,” Globe and Mail, March 7.Gillespie, W. Irwin, Frank Vermaeten, and

Arndt Vermaeten (1995) “Who Paid theTaxes in Canada, 1951-1988,” CarletonEconomics Paper, No. 95-02, Ottawa.

Hamermesh, D. (1993) Labor Demand

(Princeton: Princeton University Press).

Health and Welfare Canada (1986) Income Se-curity Programs (Ottawa).

Health and Welfare Canada (1991) Canada Pension Plan Contributors (Ottawa).

Hum, D. and Simpson, W. (1991) Income

Maintenance, Work Effort, and the Cana-dian Mincome Experiment , study pre-pared for the Economic Council of Canada (Ottawa: Supply and ServicesCanada).

Human Resources Development Canada(1994) From Unemployment Insurance to Employment Insurance (Ottawa: Supplyand Services Canada).

Keil, M.W. and J.S.V. Symons (1990) “AnAnalysis of Canadian Unemployment,”Canadian Public Policy, Vol. XVI, No. 1,

pp. 1-16.

Kesselman, Jonathan R. (1994) “CanadianProvincial Payroll Taxation: A Structuraland Policy Analysis,” Canadian Tax Jour-nal, Vol. 42, No. 1, pp. 150-200.

Layard, R., S. Nickell, and R. Jackman (1991)Unemployment - Macroeconomic Perform-

ance and the Labor Market (Oxford: Ox-

ford University Press).Lindbeck, A. (1993) Unemployment and Mac-

roeconomics (Cambridge: MIT Press).

Meng, R. (1989) “Compensating Differentialsin the Canadian Labour Market,” Cana-

dian Journal of Economics, Vol. XXII, No.2, pp. 413-424.

Merrilees, W. (1982) “Labour Market Seg-mentation in Canada: an EconometricApproach,” Canadian Journal of Eco-

nomics, Vol. XV, pp. 458-473.

Mincer, J. (1976) “Unemployment Effects of

Minimum Wages,” Journal of Political Economy, Vol. 84, pp. S87-S104.

Mitchell, D.J.B. (1993) “Keynesian, OldKeynesian, and New Keynesian WageNominalism,” Industrial Relations, Vol.32, No. 1, pp. 1-29.

Vaillancourt, F. (1986) The Administrativeand Compliance Costs of the Personal In-come Tax and Payroll Tax System in Can-

ada, 1986, Canadian Tax Foundation,Canadian Tax Paper no. 86.

Vaillancourt, F. and N. Marceau (1990) “On

General Firm-specific Payroll Taxes,” Eco-nomics Letters, Vol. 34, pp. 175-181.

Wilton, D. and D. Prescott (1993) “The Ef-fects of Tax Increases on Labour Costs,”Government and Competitiveness Pro-ject, Discussion Paper 93-29, School of Policy Studies, Queen’s University, King-ston.

Woodland, A. (1975) “Substitution of Struc-tures, Equipment and Labor in CanadianProduction,” International Economic Re-view, Vol. 16, pp. 171-187.

Workers’ Compensation Board of Ontario(1992) WC Funding Strategy (Toronto).

Payroll Taxation in Canada an Overview

Documents