Factor Prices in Industrial Countries

International and intertemporal comparisons of prices of factors of production are warranted for several reasons. Under certain conditions, factor price comparisons may help to ascertain the revealed comparative advantage of a given country vis-à-vis other countries, as well as the changes in advantage over time. Also, such comparisons may shed light on the degree of worldwide or regional economic integration brought about by deliberate efforts aimed at trade and payments liberalization, harmonization of tax and regulatory measures, and overall coordination of monetary, fiscal, and exchange rate policies and by the expanded role of multinational corporations in promoting commodity and factor movements across country boundaries.

Abstract

International and intertemporal comparisons of prices of factors of production are warranted for several reasons. Under certain conditions, factor price comparisons may help to ascertain the revealed comparative advantage of a given country vis-à-vis other countries, as well as the changes in advantage over time. Also, such comparisons may shed light on the degree of worldwide or regional economic integration brought about by deliberate efforts aimed at trade and payments liberalization, harmonization of tax and regulatory measures, and overall coordination of monetary, fiscal, and exchange rate policies and by the expanded role of multinational corporations in promoting commodity and factor movements across country boundaries.

International and intertemporal comparisons of prices of factors of production are warranted for several reasons. Under certain conditions, factor price comparisons may help to ascertain the revealed comparative advantage of a given country vis-à-vis other countries, as well as the changes in advantage over time. Also, such comparisons may shed light on the degree of worldwide or regional economic integration brought about by deliberate efforts aimed at trade and payments liberalization, harmonization of tax and regulatory measures, and overall coordination of monetary, fiscal, and exchange rate policies and by the expanded role of multinational corporations in promoting commodity and factor movements across country boundaries.

From a somewhat different perspective, international comparisons of factor prices and their components may be used to explain differences in competitiveness among countries in world markets and, it is hoped, to identify the contribution of specific policy instruments to those differences. Indeed, the ordinarily adopted indicators of competitiveness—changes in unit labor costs, average labor productivity, wholesale prices, or export prices—are of limited usefulness for policymaking purposes.

In view of the wider variation in factor prices and their components among countries at a given time than within a country over time, international factor price comparisons may provide indirect evidence of the responsiveness of the demand for factor inputs to fluctuations in factor costs. Such evidence, of course, is of interest to policymakers in a domestic context insofar as they may consider adopting selective fiscal, financial, or regulatory measures (some of them popularized as “supply-side” policies) to cut factor costs and thus induce capital accumulation and employment.

Notwithstanding the widespread interest in these issues, until recently the paucity of appropriate data has hindered intercountry factor price comparisons. However, the present availability of relevant data on key variables makes possible this attempt to measure the prices of capital and labor inputs in eight major industrial countries on a consistent basis for the years 1973 and 1978. The first section provides the theoretical basis for the comparison; the second discusses measurement methods and data sources; the third analyzes the basic results; the fourth examines the relationship between factor prices and factor proportions and presents calculations of unit factor costs; the fifth discusses the role of taxes on corporate income and payroll; and the final section contains the summary and conclusions.

I. Theoretical Background

In broad terms, factor demand of business enterprises is determined by output demand and by the real prices of capital and labor inputs. In the corporate sector, the rental price of fixed capital input, pk, consists of the product of the price of fixed assets and the required rate of return on invested funds:1

pk=qρ=q(rΔq*/q+δ)(1kuz)/(1u)(1)

where

q = price of fixed assets

ρ = required rate of return (gross of depreciation)

r = nominal cost of funds

δ = rate of economic depreciation

Δq*/q = expected rate of change in the price of fixed assets

k = rate of investment of tax credit or cash grant

u = corporation income tax rate

z = present value of tax depreciation deductions

Further, the nominal cost of funds for the corporate investor is given by2

r=f(1u)rd+(1f)re(2)

where f and (1 − f) denote, respectively, the proportion of debt and equity financing of corporate assets, rd is the interest rate on debt (fully deductible from the tax base), and re is the rate of return on equity. The latter is defined as

re=d+g*(3)

where d is the dividend (before personal income tax and dividend credit) and g* expected capital gains, both expressed as a proportion of shareholders’ equity. The present value of depreciation for income tax purposes is

z=Σt=0xt(1+r)t(4)

where xt, is the depreciation deduction, as a proportion of the original or historical price of the asset, allowed in time t on an asset purchased in time 0. If, instead, the depreciation deduction were adjusted each period for price changes since the date of acquisition of the asset, the discount rate in equation (4) would have to be reduced by the expected rate of change in asset price.

The price of labor input, pL, is equivalent to total compensation of labor:

pL=(v+ymw)(1+yns)+h(5)

where v represents the money wage and salary earnings of employees per time worked (after the employee’s share of payroll taxes but before personal income tax), so that

v=w(1ym)(6)

and

w = wages and salaries per time worked

y = payroll tax rate (contributions to social security and other government funds) divided between ym, the employee’s share, and yn, the employer’s share, as a proportion of gross wage

s = wage subsidy rate paid to the employer, as a proportion of gross wage

h = nonwage labor costs (benefits in kind, bonuses, etc.) per time worked

Assuming profit maximizing behavior and perfectly competitive markets, equilibrium obtains between the marginal product and real price of each factor input:3

Q/K=pk/p(7a)
Q/L=pL/p(7b)

where Q is the quantity of output, K and L denote the quantity of capital and labor inputs, respectively, and p is the price of output. However, in the presence of market imperfections and various regulatory constraints, the factor’s marginal product is likely to exceed its real price.4 By the same token, the desired quantity of each factor input tends to fall short of the equilibrium quantity that would obtain in a frictionless, competitive world, namely,

K=K(Q,pk/p)(8a)
L=L(Q,pL/p)(8b)

the precise nature of these relationships being contingent on the characteristics of the production function. Alternatively, if firms maximize both profits and sales, the marginal product of each factor may be depressed below its respective price, so that factor demand lies above the amount determined through profit maximization alone.5

In contrast to this partial equilibrium approach, in a general equilibrium framework factor supplies determine factor prices. According to the traditional trade model, commodity flows tend to equalize factor prices among countries even if factors of production are immobile, assuming no specialization in production and uniform technology. However, relaxation of these assumptions (except that no country can influence world commodity prices) results in intercountry variation of factor prices according to relative factor endowment:6

pK/pL=Φ(L/K)(9)

Obviously, the degree of variation tends to narrow with increased international mobility of capital or labor.

Therefore, observed factor prices may involve a two-way causality. Relative factor endowment tends to determine relative factor prices; conversely, factor price changes brought about by policy measures or certain other exogenous events (e.g., technological innovation) may alter the stocks of factor inputs in a particular country or region, a process that is accelerated by international factor movements. It follows, then, that a close crosscountry correlation between factor price ratios and factor proportions may be interpreted as evidence of near-equilibrium market conditions within each country, a high degree of responsiveness of factor inputs to price changes and to underlying policy changes, and/or incomplete international factor mobility.

II. Measurement

Comparisons involving eight major industrial countries (Belgium, France, the Federal Republic of Germany, Italy, Japan, the Netherlands, the United Kingdom, and the United States) were undertaken for the years 1973 and 1978, when these countries were operating at close to full capacity and prices were relatively stable. The measurement encompasses the prices of capital and labor inputs faced by nonfinancial corporations at two levels of aggregation: all industries and manufacturing.

The price level of output (p) was measured in terms of the purchasing power parity for gross domestic product (GDP) in each country’s national currency (i.e., number of currency units that have the same purchasing power as one 1973 U.S. dollar), based on 1973 benchmark estimates from the United Nations International Comparison Project.7 The 1973 price data were updated to 1978 on the basis of the implicit price deflator for each country’s GDP given in the national accounts statistics of the Organization for Economic Cooperation and Development—OECD—(1981). No separate calculation of the price of manufacturing output was possible, since there is no breakdown of purchasing power parity for the value added of each industry.

The price of fixed assets (q) is the purchasing power parity for nonresidential fixed capital formation covering 29 categories of productive assets,8 from the United Nations International Comparison Project. The 1973 data were updated to 1978 by applying the OECD price deflators for investment in four broad groups of assets (nonresidential buildings, other construction except land improvement, transport equipment, and machinery and other equipment) to the corresponding asset category. A separate price variable was constructed for assets used predominantly in manufacturing (comprised of industrial buildings, metalworking machinery, other special industry machinery, and general industry machinery), which was extrapolated to 1978 with the price deflator for manufacturing investment.9

The expected rate of change in the price of fixed assets (Δq*/q) was measured by the average annual rate of change in the price deflators for investment in broad groups of assets and for manufacturing investment over the three years ended in 1973 and 1978. The data source is the OECD (1981) national accounts statistics.

The rate of economic depreciation (δ) reflects a geometrically declining pattern,10 and is based on the service lives used by the U.S. Department of Commerce in the Bureau of Economic Analysis Capital Stock Study.11 The rate for each asset category was weighted by the 1973 share of that category in nonresidential fixed investment to obtain the average depreciation rate for all industries and separately for the manufacturing sector.

The rates of investment tax credit or cash grant (k) and of annual depreciation deductions allowed for tax purposes over the life of the asset (x) were weighted by the 1973 share of each asset category in nonresidential fixed investment in all industries or manufacturing. Only the capital cost recovery allowances available to the average investor were taken into account, excluding those extended under special eligibility criteria (by region, economic activity, or time period) within each country.12 The corporation income tax rate (u) is the average of rates on retained and distributed earnings, assuming a before-tax dividend payout ratio of one half. The main source for these tax variables is the 1972 OECD Tax Depreciation Survey,13 extended to 1973 and 1978 on the basis of various national sources.

The nominal cost of funds (r) is a weighted average of the market rate of interest on long-term nonfinancial corporate bonds (rd)14 and the rate of return on corporate equity (re), consisting of the yield per share of equity (d) plus the expected rate of capital gains in the value of equity (g*) measured by the three-year average annual rate of change in industrial share prices ending in 1973 and 1978. The respective weights are the ratio of liabilities (f) and shareholders’ equity (1 − f) to total assets of nonfinancial corporations. Data on financial variables were obtained from the OECD Financial Statistics, except for the index of industrial share prices published in International Monetary Fund (1981).

The price of labor input (pL) was measured by compensation of employees (v) per hour worked, including wages and salaries (w), employee and employer payroll tax payments for social security and other government funds (at rates ym and yn) less subsidies (s), and other cash and noncash benefits (h), but excluding certain expenditures of a social nature such as food and medical services for employees and expenditures on recruiting and training. For greater international comparability, data from labor cost surveys rather than from national income accounts have been used. Thus, it was necessary to limit the sectoral coverage to compensation of all employees to major production industries (mining, manufacturing, and construction), reported periodically by the European Communities for member countries, and in national sources for Japan and the United States.15 Estimates of hourly compensation of production workers, prepared by the U.S. Department of Labor (1981), were used as the price of labor input in manufacturing. Unlike the other main components of compensation that are broken down in labor cost surveys, data on the distribution of the payroll tax liability between employers and employees were checked against published statutory information from the U.S. Department of Health, Education and Welfare (1973 and 1977). Although in a number of ways the hourly compensation data are comparable across countries (as to the nature and size of establishments, etc.) they do not take into account differences in the quality of labor input. In an attempt to correct for such differences, the price of labor was adjusted by a cross-country index of educational attainment.16

III. Basic Results

Table 1 presents the prices of capital and labor inputs in all industries, divided by the overall price level, in 1973 and 1978, relative to 1973 U.S. prices. In 1973, the real price of capital (shown in column 3) ranged from 37 per cent of the U.S. price in Italy to more than 108 per cent in Belgium and Japan; by 1978, the price of capital fell in all countries except the Federal Republic of Germany and the United Kingdom. International variations in the price of capital seem to have been determined primarily by differences in the required rate of return on investment (column 1). In 1973, observed rates of return varied from 8 per cent of the value of investment in Italy to rates in excess of 20 per cent in Belgium, Japan, the Netherlands, and the United States. In 1978, both Italy and the Netherlands displayed rates of return below 10 per cent, whereas Japan’s 19 per cent rate was the highest. In contrast, the contribution of the price of fixed assets (column 2) seems to be substantially weaker and its variance across countries is much smaller. Asset prices differ by less than 20 per cent above and below the U.S. price in both years (an exception being the unusually high 1978 price in the United Kingdom). A plausible explanation for this similarity of prices is the relatively active international trade in many producer goods among industrial countries.

Table 1.

Prices of Factor Inputs, 1973 and 1978

article image

The dispersion of the required rate of return among countries and between the two years can be best explained by focusing on the outlying observations. Specifically, Italy’s required rate of return and thus its price of capital input are consistently the lowest among industrial countries. The main reason behind this result appears to be the relatively high expected rate of inflation (10 per cent and 18 per cent in 1973 and 1978, respectively), which was more than twice as high as the nominal opportunity cost of funds, owing largely to depressed nominal rates of interest and anticipated losses in equity values. Analogously, in the Netherlands, in 1978 the nominal cost of funds was less than one half of the expected rate of inflation, owing especially to the high rate of anticipated capital losses which was magnified by the relatively large proportion of equity in the financial structure of corporations in that country. At the other end of the spectrum, the high required rates of return in Japan in both years and in the United States in 1973 can be ascribed to the combination of low inflationary expectations and the high nominal cost of finance.

The overall decline in the required rate of return, as well as the price of capital input, from 1973 to 1978 is attributable partly to the increasing differential between the expected rate of inflation and the nominal cost of funds (pulled down by slipping equity values) in all countries except the Federal Republic of Germany. In the United Kingdom, the fall in the rate of return was more than offset by the actual rise in the price of producer goods, resulting in an increased price of capital services.

All things considered, the expected rate of inflation is a major determinant of the required rate of return; therefore, obvious difficulties in measuring inflationary expectations cannot be ignored in evaluating the required rate-of-return calculations. Nevertheless, the method used here for measuring the expected rate of change in capital goods prices has not been proved inferior to alternative methods that would be applicable uniformly to industrial countries, except for a possible overestimate of the expected rate of inflation for Italy and the United Kingdom in 1978 (both experiencing a relatively high inflation rate at that time) and thus an underestimate of their required rate of return and real price of capital input.17

The real price of labor unadjusted for international differences in quality (column 4) was higher for the United States than for the other countries in both years. Adjustment for quality increases the variance of the price of labor (column 5). In 1973, it ranged from 47 per cent of the U.S. price in Japan to 111 per cent in the Federal Republic of Germany and Italy; five years later, these extreme values increased to 59 per cent in Japan and to nearly 139 per cent in Germany and also in Belgium. Overall, the country ranking is affected considerably by the quality adjustment index, which may be distorted by the exclusion of experience and other noneducational factors.18

The opposite trends in factor prices between 1973 and 1978 are highlighted in Table 2. In practically all countries, capital-labor price ratios dropped dramatically over the period. Notably, in Belgium and the Netherlands, and in Italy’s manufacturing sector, the ratios fell by at least one half. The only exception is the United Kingdom, where the price ratio stayed virtually unchanged thanks to the increase in producer goods prices.

Table 2.

Factor Price Ratios (Pk/Pl), 1973 and 1978

(Index, U.S. 1973 = 100)

article image

The price of capital relative to that of labor was consistently lower in the manufacturing sector (columns 3 and 4) than in the broader industrial aggregate (columns 1 and 2)—except in Japan—the country ranking being roughly the same under either coverage. Japan lies at the top, with a factor price ratio more than twice as high as the U.S. ratio, followed by the United Kingdom, and then by the United States and most EEC member countries; Italy had the lowest ratio, equivalent to not more than one third of the adjusted U.S. ratio.

IV. Factor Prices and Factor Proportions

Factor price comparisons become all the more meaningful when considered together with estimates of factor proportions, as suggested in the theoretical section. Drawing on independent calculations of intercountry quantity indexes of output, and capital and labor inputs, for seven countries included in the sample (i.e., excluding Belgium)19 and on the foregoing measures of factor prices, it was possible to plot the scatter of observations for 1973 and 1978 shown in Chart 1 adjusted for labor quality differences.

Chart 1.
Chart 1.

Factor Price Ratio and Quantity Ratio, 1973 and 19781

Citation: IMF Staff Papers 1982, 003; 10.5089/9781451972702.024.A004

1 France (FR), the Federal Republic of Germany (GR), Italy (IT), Japan (JP), the Netherlands (NT), the United Kingdom (UK), and the United States (US).

Despite some measurement errors (arising partly from discrepancies in sectoral coverage), the capital-labor price and labor- capital quantity ratios exhibit a remarkable positive relationship—with the exception of Italy, for reasons discussed later in this section. Accordingly, inverse relationships (consistent with demand functions (8a) and (8b)) can be detected between price and quantity ratios of each factor input to output, with a more robust nexus for labor than for capital (Charts 2 and 3). These findings seem to have at least three implications.

Chart 2.
Chart 2.

Real Price of Capital and Capital-Output Ratio, 1973 and 19781

Citation: IMF Staff Papers 1982, 003; 10.5089/9781451972702.024.A004

1 France (FR), the Federal Republic of Germany (GR), Italy (IT), Japan (JP), the Netherlands (NT), the United Kingdom (UK), and the United States (US).
Chart 3.
Chart 3.

Real Price of Labor and Labor-Output Ratio, 1973 and 19781

Citation: IMF Staff Papers 1982, 003; 10.5089/9781451972702.024.A004

1 France (FR), the Federal Republic of Germany (GR), Italy (IT), Japan (JP), the Netherlands (NT), the United Kingdom (UK), and the United States (US).

First, the evidence is broadly consistent with the view that factor price ratios reflect relative factor endowment, suggesting that each country (or group of countries) possesses a comparative advantage over at least several others in the production of commodities that require the more intensive use of a given factor input. Specifically, in both years Japan enjoyed an advantage relative to other industrial countries in the production of labor- intensive goods, as opposed to the Federal Republic of Germany’s advantage in producing capital-intensive products. France and the United States, and particularly the Netherlands in 1978, were closer to Germany, while the United Kingdom leaned toward Japan.

Second, the relatively large and systematic variation in factor prices reflects insufficient mobility of factors of production across country boundaries. Interestingly, the absence of fully integrated factor markets seems to be prevalent even among EEC member countries. This inference must be qualified insofar as certain factors are less mobile than others. The stronger price-quantity relationship for labor than for capital indicates that probably there are greater impediments (including intercountry tax or subsidy rate differentials) to labor than to capital movements.20

Third, on the basis of the above evidence, it is not possible to reject the hypothesis that, beyond the short run, factor inputs respond to changes in factor prices.21 This implies that policy measures that alter factor prices tend to be effective in influencing capital formation and employment at the margin. In this respect, the weaker price-quantity correlation for capital can be attributed to a large extent to the slower adjustment to the new equilibrium stock in response to a change in the price of capital.

None of these implications seems to be equally valid in the case of Italy, which may not belong to the same universe as the other countries in the sample. According to the pattern of the other country observations, Italy’s low real price of capital (the lowest in the group) should have been associated with the highest capital- output ratio or the lowest labor-capital ratio; yet Italy ranked closer to the middle of the range of quantity ratios. Compared with other countries, the marginal product of capital seems to have been substantially higher than its real price. The disequilibrium was probably due to a relatively high risk premium, symptomatic partly of the variance of inflationary expectations, not reflected adequately in the observed cost of funds. As mentioned earlier, however, Italy’s real price of capital may be understated for 1978 because of possible error in the measurement of inflationary expectations.

Comparative estimates of factor proportions are also useful to calculate factor costs per unit of output. In essence, unit factor costs consist of the summation of the product of real factor prices and the respective factor-output ratios. They can be expressed and interpreted in two alternative ways. If factor prices are deflated by the price level of output, unit factor costs show the real resource cost of production in each country, relative to the base country and period. If, on the other hand, factor prices are converted at the market exchange rate into a single national currency, unit factor costs can serve as an indicator of the relative competitiveness of each country’s products in world markets. As a corollary, any difference between relative dollar factor costs and relative real factor costs can be interpreted as the magnitude of the adjustment that would be required in the exchange rate so as to reflect the prevailing intercountry cost structure in a purchasing-power-parity sense.

Both sets of unit factor costs in all industries are presented in Table 3 for seven countries (excluding Belgium).22 In 1973, unit factor costs in real terms (column 1) were about 30 per cent lower in France, Japan, and the United States than they were in the Federal Republic of Germany, Italy, and the Netherlands, while the United Kingdom was between the two groups. Whereas costs in France, the Netherlands, and the United States declined between 1973 and 1978, the remaining countries experienced a moderate increase. Relative to the United States, dollar unit factor costs23 (column 2) in 1973 were more than 40 per cent higher in Germany and the Netherlands, while they were 10 per cent lower in Japan and the United Kingdom. But the spread in dollar factor costs was much larger in 1978. Compared with the United States, which had the lowest costs, costs were nearly twice as high in Germany, two thirds higher in the Netherlands, and roughly 20 to 40 per cent higher in France, Italy, and Japan.

Table 3.

Unit Factor Costs and Unit Labor Costs, 1973 and 19781

article image

Column (1): (pK/p)K/Q + (pL/p)L/Q

Column (2): (pK/e)K/Q + (pL/e)L/Q

Column (4): (pL/p)L/Q

Column (4): (pL/e)L/Q

Comparisons of unit labor costs (columns 3 and 4) reveal a country ranking broadly similar to that of unit factor costs, which is not surprising given the relatively large share of the wage bill in global costs of production. However, in Italy, unit labor costs were higher (highest in real terms in 1978) relative to other countries than were unit factor costs, partly because of the low estimates of the price of capital. Further, by 1978, the United States had become the lowest-cost producer among industrial countries in terms of all factor costs and second lowest in terms of labor costs.

V. Role of Taxation

The corporation income tax acts as a wedge between the actual price of capital input and the price that would obtain in its absence. Broadly speaking, the extent and nature of the wedge is contingent on the sectoral coverage of the tax, the tax treatment of capital cost recovery, interest expenses, capital gains and losses, and price changes. In general, depreciation of fixed assets is deducted from the tax base according to prescribed rates, applied to the historical cost of the asset. In a number of countries, depreciation rates have become quite liberal compared with what is believed to be the rate of economic depreciation, and are complemented with investment credits or grants. Nominal interest expenses are deductible, while capital gains and losses are included in the tax base only upon realization and are often taxed at preferential rates. None of these countries provides an automatic adjustment of the tax base for inflation.

Since in industrial countries the corporation income tax can be regarded almost as an economy-wide tax on income from capital, it is plausible to assume for analytical purposes that the tax is borne by the owners of capital. This working assumption sets the conditions for tax neutrality, under which the price of capital remains invariant with respect to changes in the tax rate, as if the tax rate was zero. Full or zero-rate neutrality requires deductibility of economic depreciation of fixed assets, deductibility of interest costs, and taxation of accrued net capital gains as ordinary income, all adjusted for the rate of inflation.24 Following this approach, the required rate of return on investment in equation (1) is restated as

ρ=r/(1u)Δq*/q+δ(10)

Alternatively, in view of practical difficulties in taxing capital gains on an accrual basis, neutrality may be stated in a partial form, namely, by allowing only the deductibility of inflation- adjusted economic depreciation and of real interest costs.25 Under these conditions, the required rate of return becomes

ρ=[r(1fu)Δq*/q]/(1u)+δ(11)

Either one of these neutrality standards can be used to ascertain the tax wedge or contribution of the tax system to the price of capital input, assuming full absorption of the tax by the investor.

In addition, it can be of interest to focus separately on the role of the existing method of capital cost recovery in determining the price of capital in each country. There are two specific features that are worth examining: tax depreciation rules and historical cost accounting. These features become particularly relevant in an inflationary situation when countries ordinarily attempt to offset the penalty imposed by inflation by providing tax subsidies (in the form of accelerated depreciation, tax credits, etc.) The more efficient approach would, instead, consist of allowing economic depreciation adjusted automatically by the rate of inflation. The required rate of return under such an approach would be

ρ=(rΔq*/q)/(1u)+δ(12)

If, however, economic depreciation is applied to the historical cost of the asset, then

ρ=(rΔq*/q+δ)[1uδ/(r+δ)]/(1u)(13)

Table 4 shows the required rate of return on invested funds calculated under the present tax system (column 1), full neutrality (column 2), partial neutrality (column 3), inflation-adjusted or indexed economic depreciation (column 4), and historical-cost economic depreciation (column 5) applying the definitions given in equations (1), (10), (11), (12), and (13), respectively. Differences between the rate of return under the existing system and the rate that would prevail under the hypothetical system in a given column illustrate the magnitude of the immediate impact—without allowing for secondary repercussions—of the existing system, or a particular characteristic thereof, on the rate of return in each country.26

Table 4.

Required Rate of Return on Investment (ρ) Under Alternative Corporation Income Tax Systems, 1973 and 1978

(In per cent of asset value)

article image

An interesting outcome of this experiment is that, in both 1973 and 1978, for most countries actual rates of return were very close to the rates calculated under the partial standard of neutrality. The notable exceptions were Italy and the United Kingdom, both experiencing relatively high expected inflation rates. Furthermore, the lower rate of return under the present system than under either neutrality criterion suggests that the rewards conferred (through tax deferral on capital gains, deductibility of nominal interest payments, and liberalized capital cost recovery) outweigh the burden imposed (through historical-cost depreciation) by the existing system under inflation.27 The much higher rate of return obtained under the full neutrality criterion than under the partial one in all countries reflects the effect of taxing accrued gains (or deducting accrued losses) under full neutrality.

The calculations assuming inflation-adjusted and historical-cost economic depreciation for tax purposes resulted as expected in the lowest (in 1978 a negative 1 per cent for Italy!) and somewhat higher than actual rates of return, respectively, for each country. It is no wonder that spokesmen for business interests often advocate indexation of the depreciation base along with retention of accelerated writeoffs and other investment incentives. Also, the greatest benefit from inflation-adjusted depreciation would arise in countries experiencing high rates of inflation.

More generally, positive differences between rates of return assuming historical-cost economic depreciation and those calculated under the present system reflect largely the magnitude of tax subsidies extended in the form of generous capital cost recovery.28 The United Kingdom displays the largest deviation, owing to instantaneous depreciation (expensing) of most equipment and machinery purchases.29

Unlike the taxation of capital, it is far easier to gauge the magnitude of the effective tax rate (to be distinguished from y, the statutory rate) on labor. In fact, with the available data, it is relatively simple to calculate the tax liability as a proportion of the price of labor input. From equation (5), it follows that the effective payroll tax (less subsidy) rate is

a=(ym+yns)w/pL(14)

where the net tax consists of compulsory social security contributions of employers and employees and other taxes levied, less subsidies received, on labor compensation.30

Table 5 shows the effective payroll tax rate, broken down by major type of tax, for all industries in 1973 and 1978. Obviously, variations in the payroll tax rate across countries and over time reflect not only differences in legal rates but also variations in the ceiling on taxable wages, in other components of labor compensation, and in the composition of the taxpaying population. In global terms, in 1973, payroll taxes represented between one fourth and one third of the price of labor input in industrial countries in continental Europe and not more than one tenth of it in the other countries. Subsequently, effective payroll taxation increased in all countries, mainly as a result of discretionary increases in statutory rates of social security contributions and ceilings on taxable wages—made necessary by escalating social welfare costs brought about by the high rates of inflation and the rising proportion of beneficiaries to taxpayers in most of these countries.

Table 5.

Effective Payroll Tax Rate (a), 1973 and 1978

(In per cent of labor compensation)

article image

Further perusal of the results reveals a strong correlation between the effective payroll tax rate and the price of labor input in both years, as shown in Chart 4. This outcome runs counter to the view that over the long run payroll taxes are absorbed by labor.31 Although no insight into the shifting mechanism itself can be obtained from the evidence, it seems clear that payroll taxation can exercise a powerful influence on the price of labor, and thus on unit factor costs.32

Chart 4.
Chart 4.

Real Price of Labor and Effective Payroll Tax Rate, 1973 and 19781

Citation: IMF Staff Papers 1982, 003; 10.5089/9781451972702.024.A004

1 Belgium (BL), France (FR), the Federal Republic of Germany (GR), Italy (IT), the Netherlands (NT), the United Kingdom (UK), and the United States (US).

Apparently, to counteract the ensuing adverse effect of rising payroll taxes on employment, on the one hand, and international competitiveness,33 on the other, in the 1970s several industrial countries introduced wage subsidies often targeted to hard-to-employ groups.34 Along these lines, for example, Italy adopted the fiscalizzazione approach of offsetting a proportion of the pay-roll tax for social security with an increase in indirect tax revenue. As shown in Table 5, this offset amounted to 5 per cent of labor compensation in 1978.

VI. Summary and Conclusion

This paper represents an attempt to measure factor prices that are both comprehensive and internationally comparable. Capital and labor prices in 1973 and 1978 are estimated for eight major industrial countries: Belgium, France, the Federal Republic of Germany, Italy, Japan, the Netherlands, the United Kingdom, and the United States. The estimates explicitly incorporate key components of factor prices: the price of capital goods, the rate of depreciation, the cost of equity and debt finance, the expected rate of inflation, and the company income tax system are taken into account in the calculation of the price of capital input, and the wage rate and payroll taxes and subsidies are elements of the price of labor input. The conceptual underpinning of the estimates is drawn partly from the neoclassical theory of the firm and partly from the neoclassical theory of exchange.

Although the evidence that emerges from this international comparison of factor prices—involving a limited number of countries—cannot be interpreted as conclusive corroboration of any particular set of. hypotheses, it can serve as the basis for selecting those that may be useful for policy formulation and be subject to further testing. The empirical results, the hypotheses to which they lend support, and the ensuing policy implications are summarized below.

1. In 1973, the real price of capital in Japan, the Netherlands, and the United States was almost one and one half times as large as in France and the Federal Republic of Germany and about twice as large as in Italy. Although by 1978 these differences seem to have narrowed somewhat, they still remained substantial. In both years, the real price of labor (adjusted for quality differentials) in Belgium, Germany, Italy, and the Netherlands was roughly twice as large as in Japan, while the prices in the other countries were between these two extremes. The wide and systematic dispersion of factor prices suggests that capital and labor markets are far from being fully integrated in the industrialized world, and even within the EEC, despite concerted efforts to reduce barriers to trade and factor movements and to coordinate macroeconomic policies. Stated differently, prevailing factor price differentials point to significant welfare gains to be derived from further relaxation of existing impediments to trade and factor flows.

2. In 1973, Japan had the highest capital-labor price ratio and the lowest capital-labor quantity ratio, followed by the United Kingdom with a lower price ratio and higher quantity ratio; at the other end, the Federal Republic of Germany displayed a combination of the lowest factor price ratio and the highest factor quantity ratio; and France, the Netherlands, and the United States were scattered between the two extremes. The correlation between factor price and quantity ratios and country ranking remained broadly unchanged in 1978, except for the Netherlands, which, together with Germany, became the most capital-intensive producer. (Italy was the only country that deviated from this general pattern; its relatively low capital-labor price ratio was not matched by capital abundance.) For both years, the evidence shows that factor prices are inversely related to factor endowment across countries that have attained a given stage of industrialization and that possess broadly similiar risk characteristics. According to this finding, factor prices can be viewed as a reliable indicator of comparative advantage. From another perspective, the evidence can lead to the interpretation that policy instruments (regulations, taxes, subsidies) that alter factor prices are likely to be effective in influencing capital formation and employment.

3. In 1973, unit factor costs (expressed in U.S. dollars) were more than 40 per cent higher in the Federal Republic of Germany and the Netherlands, and about 10 per cent lower in Japan and the United Kingdom, than in the United States. However, by 1978, the United States had become the most competitive producer; unit factor costs were almost twice as high in Germany, two thirds higher in the Netherlands, and about 20 to 40 per cent higher in France, Italy, and Japan. Given the relatively large share of the wage bill in factor costs, wage-related policies (such as incomes policy, minimum wage regulation, wage indexation, and payroll taxes and subsidies), along with the exchange rate, are probably the most important determinants of a country’s international competitiveness in the near term. Policies affecting the price of capital (fiscal incentives, interest rate regulation, credit rationing) are likely to have a significant impact on competitiveness only over the longer term—through their effect on capital formation, and thus on factor productivity.

4. During the 1970s, the price of capital input would have been higher under a completely neutral company tax structure (i.e., allowing economic depreciation for tax purposes, taxation of capital gains as ordinary income, and full adjustment for inflation). The advantage conferred by the tax system was particularly pronounced in countries that offered liberal investment incentives or where inflationary expectations were relatively high, such as Italy and the United Kingdom, as against the Federal Republic of Germany and Japan, which have had relatively stable prices and which have provided the least favorable investment subsidies. The price of labor has been strongly correlated with the effective rate of payroll taxation (consisting almost entirely of social security contributions), which in 1978 amounted to more than 30 per cent of labor costs in Germany, Italy, and the Netherlands. In sum, the combination of increased payroll tax rates and liberalized investment incentives seems to have contributed significantly in many countries to the decline in the price of capital relative to the price of labor. Thus, aside from the obvious effect on factor choice—biased against labor—the increased payroll tax burden appears to pose a severe drag on these countries’ international competitiveness.

REFERERNCES

  • Brittain, John A., The Payroll Tax for Social Security, The Brookings Institution (Washington, 1972).

  • Brown, M., and N. Revankar, A Generalized Theory of the Firm: An Integration of the Sales and Profit Maximization Hypotheses,” Kyklos, Vol. 24 (No. 3, 1971), pp. 42743.

    • Search Google Scholar
    • Export Citation
  • Christensen, L. R., D. Cummings, and D.W. Jorgenson (1980), “Economic Growth, 1947–73: An International Comparison,” in New Developments in Productivity Measurement and Analysis, ed. by John W. Kendrick and Beatrice N. Vaccara (University of Chicago Press, 1980), pp. 595698.

    • Search Google Scholar
    • Export Citation
  • Christensen, L. R., D. Cummings, and D.W. Jorgenson (1981), “Relative Productivity Levels, 1947–1973: An International Comparison,” European Economic Review, International Seminar in Macroeconomics, Vol. 16 (May 1981), pp. 6194.

    • Search Google Scholar
    • Export Citation
  • Christensen, Laurits R., and Dale W. Jorgenson, The Measurement of U.S. Real Capital Input, 1929–1967,” Review of Income and Wealth, Series 15 (December 1969), pp. 293320.

    • Search Google Scholar
    • Export Citation
  • Denison, Edward F., Why Growth Rates Differ: Postwar Experience in Nine Western Countries, The Brookings Institution (Washington, 1967).

    • Search Google Scholar
    • Export Citation
  • European Communities, Statistical Office (1976), Social Statistics: Labor Costs in Industry, 1972–75, Vol. 6/1975 (Luxembourg, 1976).

  • European Communities, Statistical Office (1981), Hourly Earnings/Hours of Work, IV 1980, Vol. 1/1981 (Luxembourg, 1981).

  • Feldstein, Martin S., Tax Incidence in a Growing Economy with Variable Factor Supply,” Quarterly Journal of Economics, Vol. 88 (November 1974), pp. 55173.

    • Search Google Scholar
    • Export Citation
  • Feldstein, Martin S., and C. Horioka, Domestic Saving and International Capital Flows,” Economic Journal, Vol. 90 (June 1980), pp. 31429.

    • Search Google Scholar
    • Export Citation
  • Hall, R. E., and D. W. Jorgenson, Application of the Theory of Optimum Capital Accumulation,” in Tax Incentives and Capital Spending, ed. by Gary Fromm, The Brookings Institution (Washington, 1971), pp. 960.

    • Search Google Scholar
    • Export Citation
  • Harberger, Arnold C., Vignettes on the World Capital Market,” American Economic Review, Papers and Proceedings, Vol. 70 (May 1980), pp. 33137.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, International Financial Statistics Yearbook, 1981, Vol. 34 (Washington, 1981).

  • Japan, Ministry of Labour, Statistics and Information Department, Year Book of Labour Statistics, 1973 and 1978 (Tokyo, 1974 and 1979).

  • Jones, Ronald W., International Trade: Essays in Theory (Amsterdam and New York, 1979).

  • King, Mervyn A., Public Policy and the Corporation, Cambridge Studies in Applied Econometrics, No. 3 (London and New York, 1977).

  • Kopits, George (1975), International Comparison of Tax Depreciation Practices, Organization for Economic Cooperation and Development (Paris, 1975).

    • Search Google Scholar
    • Export Citation
  • Kopits, George (1978), “Wage Subsidies and Employment: An Analysis of the French Experience,” Staff Papers, Vol. 25 (September 1978), pp. 494527.

    • Search Google Scholar
    • Export Citation
  • Kopits, George (1980), “Effects of Tax Changes on Direct Investment Abroad,” in United States Taxation and Developing Countries, ed. by Robert Hellawell (Columbia University Press, 1980), pp. 22362.

    • Search Google Scholar
    • Export Citation
  • Kopits, George (1981), “Fiscal Incentives for Investment in Industrial Countries,” Bulletin for International Fiscal Documentation, Vol. 35 (July 1981), pp. 29194.

    • Search Google Scholar
    • Export Citation
  • Kopits, George (forthcoming), “Inflation, Income Taxation, and Economic Behavior,” in Comparative Tax Studies: Essays in Honor of Richard Goode, ed. by S. Cnossen. It is scheduled to be published by North Holland.

    • Search Google Scholar
    • Export Citation
  • Kravis, Irvin B., Alan Heston, and Robert Summers, in collaboration with others, International Comparisons of Real Product and Purchasing Power, United Nations International Comparison Project: Phase II (published for the World Bank by Johns Hopkins University Press, 1978).

    • Search Google Scholar
    • Export Citation
  • Organization for Economic Cooperation and Development (1973 and 1978), Economic Outlook, Vols. 13 and 23 (Paris, July 1973 and July 1978).

    • Search Google Scholar
    • Export Citation
  • Organization for Economic Cooperation and Development (1973 and 1978) (1980), Labour Force Statistics, 1967–1978 (Paris, 1980).

  • Organization for Economic Cooperation and Development (1973 and 1978) (1981), National Accounts of OECD Countries, 1962–1979, Vol. 2 (Paris, 1981).

    • Search Google Scholar
    • Export Citation
  • Organization for Economic Cooperation and Development (1973 and 1978) , OECD Financial Statistics (Paris, various issues).

  • Samuelson, Paul A., Tax Deductibility of Economic Depreciation to Insure Invariant Valuations,” Journal of Political Economy, Vol. 72 (December 1964), pp. 604606.

    • Search Google Scholar
    • Export Citation
  • Tideman, Nicolaus, and Donald P. Tucker, The Tax Treatment of Business Profits Under Inflationary Conditions,” in Inflation and the Income Tax, ed. by Harry J. Aaron, The Brookings Institution (Washington, 1976), pp. 3374.

    • Search Google Scholar
    • Export Citation
  • U.S. Department of Commerce, Bureau of Economic Analysis, Survey of Current Business, Vols. 57 and 59 (Washington, July 1977 and July 1979).

    • Search Google Scholar
    • Export Citation
  • U.S. Department of Health, Education and Welfare, Social Security Administration,Social Security Programs Throughout the World, 1973 and 1977 (Washington, 1973 and 1977).

    • Search Google Scholar
    • Export Citation
  • U.S. Department of Labor, Bureau of Labor Statistics (1980), Handbook of Labor Statistics (Washington, December 1980).

  • U.S. Department of Health, Education and Welfare, Social Security Administration (1981), “Estimated Hourly Compensation of Production Workers in Manufacturing, Ten Countries, 1970–80” (unpublished data, Washington, March 1981).

    • Search Google Scholar
    • Export Citation
  • Young, A. H., and J. C. Musgrave, Estimation of Capital Stock in the United States,” in The Measurement of Capital, ed. by Dan Usher (University of Chicago Press, 1980), pp. 2381.

    • Search Google Scholar
    • Export Citation
*

Mr. Kopits, Senior Economist in the European Department, holds degrees from Georgetown University. Formerly he was associated with The Brookings Institution and the U.S. Treasury Department, and taught at the Johns Hopkins School of Advanced International Studies. He is indebted to Harry Grubert, Thomas Horst, Gary Hufbauer, and Duncan MacRae for helpful comments.

1

The basic formulation of the rental price of domestic capital input in the absence of inflation is given in Hall and Jorgenson (1971). For extensions to an inflationary situation, see Tideman and Tucker (1976), and to capital located abroad, Kopits (1980).

2

According to an alternative view that assumes equilibrium in the financial (debt and equity) market, the cost of funds is determined by the rate of interest modified explicitly by the income tax treatment of individual (or institutional) shareholders, as well as by the corporate income tax. See King (1977, Ch. 8).

3

In countries that impose a general sales tax, or a value-added tax, at a rate i, the market price of output p should be replaced by p(1−i). Also, insofar as the value-added tax levied on capital goods at rate j is credited on a current basis against the tax paid at the next stage of processing or distribution, the market price of asset q in pK should be adjusted to q(1 −j). This, in fact, is the general practice followed by member countries of the European Economic Community.

4
A more general statement of equation (7a) for capital is
Q/K=pk(1+1/є)/p(11/η)
where є is the price elasticity of supply of capital and η is the price elasticity of demand for output. Clearly, as either elasticity value declines, there is an increasing gap between the marginal product and price of capital. An analogous statement of equation (7b) is applicable to labor input.
5

For example, for capital, equation (7a) would be replaced by

Q/K=(pk/p)/λ
where λ ≥ 0 to the extent that the marginal utility of revenue exceeds zero. For a derivation, see Brown and Revankar (1971).
7

See Kravis, Heston, and Summers (1978). The author is grateful to Alan Heston for making available detailed printouts of the multilateral price and quantity comparisons.

8

The following assets are specifically excluded: educational and hospital buildings; buildings for cultural, religious, sports, and social purposes; roads, streets, and highways; and land improvement and plantation and orchard developments.

9

The price deflator for France includes mining and utilities; for Italy and the Netherlands it also includes construction; and for Japan the deflator for aggregate investment was used because of the lack of any sectoral breakdown of investment.

10

The approach used by Christensen and Jorgenson (1969), among others, has been adopted here.

12

It is assumed that the enterprises’ tax liability is sufficiently large to make full use of available depreciation allowances and tax credits.

14

Except for Italy and Japan where, because of the relatively large proportion of short-term bank financing of corporate debt, the prime rate charged on bank loans was used.

15

European Communities (1976) and (1981) and Japan (1974) and (1979). The U.S. hourly compensation rates were obtained from national income accounts reported by the U.S. Department of Commerce (1977; 1979) supplemented with information from the U.S. Department of Labor (1980). The author has benefited from discussions with Arthur Neef of the U.S. Bureau of Labor Statistics on labor cost data comparability.

16

The 1970 index was extrapolated to 1973 and 1978 on the basis of each country’s annual growth rate in educational attainment; see Christensen, Cummings, and Jorgenson (1980) and (1981). Lacking an index value for Belgium, it is assumed that educational attainment in that country was the same as in the Netherlands in both years.

17

Compared with forecasts of the rate of change in the gross national product deflator by the OECD (1973 and 1978) and the Fund in its staff surveys of the world economic outlook, the three-year average rate of change in the deflator was within ½ of 1 percentage point of those alternative forecasts for five out of eight countries in both 1973 and 1978. Only for Italy and the United Kingdom in 1978 was the three-year average outside that range—above the OECD and Fund forecasts. (Unfortunately, no such forecasts are available for the rate of change in the fixed investment deflator.)

18

In an earlier comparative study of the United States and European countries, differences in educational level were found to be smaller. However, educational differences were considerably more significant than differences in experience or age-sex composition of the labor force. (See Denison (1967), Chs. 7, 8, and 9.)

19
The quantity indices for 1973 were computed by Christensen, Cummings, and Jorgenson (1981) in the context of a transcendental logarithmic production function. The capital quantity index has been adjusted to include only nonresidential structures and producer durables. The labor-output ratio was updated to 1978 on the basis of changes in manhours worked as a proportion of real GDP. The 1978 capital-output ratio was generated by grafting onto the 1973 benchmark the following relationship (as a proportion of real GDP) for each country:
Kt=It+(1δ)Kt1
where this year’s stock of fixed assets is defined as the summation of gross fixed investment (I) and last year’s capital stock after depreciation (all in real terms) over the years t = 1973, … 1978. The data sources for manhours are the same as for hourly compensation in all production industries and OECD (1980); for GDP and investment the source is OECD (1981).
20

As regards fixed assets, buildings and other structures are virtually immobile, while machinery and equipment are traded actively. Short-term portfolio investment is highly mobile in contrast to direct investment, although many multinational enterprises have ready access to the international bond market and to internal funds pooled among a large number of affiliates located in different countries. For further arguments and evidence on the degree of worldwide integration of capital markets, see the debate between Harberger (1980) and Feldstein and Horioka (1980).

21
The extent of this response is reflected by the least-squares estimate of the relationship between the factor quantity ratio and the factor price ratio (with quality-adjusted labor input) for the pooled country/year observations (excluding Belgium and Italy) shown in Chart 1:
ln(L/K)=4.502(0.661)+0.944(0.146)ln(pK/pL)R¯2=0.788n=12
where the second regression coefficient can be interpreted as an estimate of the constant elasticity of factor substitution, R2 is the coefficient of determination adjusted for loss in degrees of freedom, n is the number of observations, and the standard error of each regression coefficient is shown in parentheses. The estimated equation can be derived from the division of equation (8b) by equation (8a).
22

None of the measures requires adjustment for labor quality differentials, since any adjustment would have an offsetting impact through the price and the quantity of labor.

23

The average annual exchange rate of each country’s currency per U.S. dollar (e) was obtained from International Monetary Fund (1981).

24

This approach follows from the neutrality definition given by Samuelson (1964).

26

In the absence of inflation and capital gains on equity, the rate of return under any one of the tax principles applied in columns (2) through (5) would be the same. If, however, the expected rate of change in the price of capital goods was equivalent to the anticipated rate of change in the value of shareholders’ equity, then the rate of return calculated under either definition of neutrality in columns (2) and (3) would be identical.

27

For further discussion, see Kopits (forthcoming).

28

A more direct estimate of the tax subsidy rate on fixed assets is obtained by multiplying the income tax rate by the difference between the present value of depreciation deductions allowed for tax purposes (plus credits and grants) and the present value of depreciation under a neutral system; see Kopits (1981).

29

Instantaneous depreciation leads to full tax neutrality if the after-tax required rate of return stays unchanged regardless of the tax rate. This presupposes that the corporation income tax is fully shifted and interest payments are not deductible.

30

It may be argued that measurement of the tax rate on labor services should include the personal income tax if it has the same incidence as payroll taxes. This argument seems to be valid, for example, in the Netherlands, where wage negotiations are normally conducted in reference to the workers’ real disposable income (after payroll and income taxes).

32
The shifting of payroll taxes away from labor is confirmed by the application of a more formal test used by Brittain (1972, Ch. 3). Specifically, the following least-squares estimate of a wage equation consistent with a CES production function was obtained for the pooled sample of observations (excluding Belgium):
ln[pL(100a)/p]=8.148(2.287)+0.817(0.148) ln(Q/L)+0.185(0.474) ln(100+a)R¯2=0.762n=14
where the after-tax price of labor is hypothesized as being determined by the average productivity of labor and by the effective payroll tax rate, R2 is the coefficient of determination adjusted for loss in degrees of freedom, n is the number of observations, and the standard error of each regression coefficient is shown in parentheses. The insignificance of the coefficient of the tax variable leads one to reject the hypothesis that changes in the payroll tax rate have a long-run impact on the after-tax real wage rate (which would have required a significant coefficient approaching −1). The coefficient was equally insignificant when the employee’s share of social security contributions was excluded from the tax rate.
33

The adverse impact on competitiveness stems from the absence of border tax adjustment (on exports and imports) based on the destination principle as regards payroll tax payments.

IMF Staff papers: Volume 29 No. 3
Author: International Monetary Fund. Research Dept.