Cyclical Fluctuations in the Share of Corporate Profits in National Income

April 1977

B\OY

CYCLICAL FLUCTUATIONS IN THE SHARE OF CORPORATE PROFITS IN NATIONAL INCOME

by —

Scott B. Brown

NOTE: International Finance Discussion Papers are preliminary materials circulated to stimulate discussion and critical conment. References in publications to International Finance Discussion Papers (other than an acknowledgment by a writer that he has had access to unpublished material) should be cleared with the author or authors.

Cyclical Fluctuations in the Share of Corporate Profits in National Income

by

Scott B. Brown”

During the period since 1960, the share of total national income attributable to corporate profits has fallen in most Western industrial countries. This fall has been especially pronounced during the 1970's, and has stimulated considerable public discussion of trends in business profitability and their implications for the future course of capital formation. In this country the discussion has centered on the likelihood of a "capital shortage,"

a scenario in which low rates of return to capital lead to continued slack in business investment, and hence to low rates of growth of real output.

This paper focuses on explanations for the observed decline in the profits share which do not require the assumption of a secular decline in rates of return to capital. Using data for six industrial countries, the first part of this paper presents evidence that much of the observed decline in the profits share has resulted from changes in the distribution of gross corporate profits among dividends, interest, and retained earnings, rather than a decline in profitability. The second part attempts to identify the extent to which recent variations in the gross profits share have resulted from predictable "normal" responses to the cyclical pattern of economic activity, by comparing profits shares predicted from simple reduced form equations to the actual profits shares, In the last part, it appears that many countries have not experienced profits

shares recently which are consistent with "full-employment" growth paths.

*Division of International Finance, Board of Governors of the Federal Reserve System. The author is grateful to Robert M. Dunn, Robert S. Goldfarb, Raymond Lubitz, and Larry J. Promisel for valuable comments on this draft. This paper is based on an M.A. thesis at the George Washington University.

1. Observed Trends in the Distribution of National Income. in Six Industrial Countries, and Computation of the Gross Profits Share.

National income is generally divided into at least four functional categories in national accounts statistics: compensation of employees, entrepreneurial income, personal property income (including net interest income), and corporate profits. Compensation of employees includes wages and salaries, payments and supplements in cash or in kind, and social security contributions by employers. This aggregate corresponds roughly to "labor income" as the term is generally used in economic theory; however, in practice many large salaries are derived from a managerial relationship to the firm much more akin to ownership than to employment, and contain elements of both rental and investment income. Entrepreneurial income is

a very mixed bag, as much of it accrues to individuals as implicit labor income,

Profits and personal property income may be categorized reasonably as income from the ownership of capital, Where possible, corporate profits are defined in this study as profits of private corporations before deduction of taxes or dividends, Adjustments are made to the data where necessary in order to deduct from factor incomes the portion of income which is attributable to revaluation of inventories through price inflation’ -

>

thus, corporate profits are net of inventory valuation adjustment (IVA).

1/ Ina recent set of articles, Shoven and Bulow argue that there can be equally significant adjustments of the opposite sign due to the net gain to corporate debtors from depreciation of financial assets through price inflation. See Shoven and Bulow (1975) and Shoven and Bulow (1976).

-3-

Due to institutional differences among the six countries included in this study, the United States, Canada, Germany, Italy, Japan, and the United Kingdom, factor income categories are not strictly comparable. Some differences are noted in the following brief discussion of trends in the distribution of national income. Where possible, adjustments are made in the process of computing the gross profits share in order to render the data more comparable,

The observed share of corporate profits in U.S. national income has declined since the years immediately following World War II. This ratio fell from its 1946-1955 average of 12.4 per cent to about 7.6 per cent in 1975, with most of the decline occurring during the 1970's. The share of rental income and entrepreneurial income also declined over this period, while the shares of compensation of employees and interest income rose substantially. Much of these trends can be explained by changes in the organization of American business: small farmers and small businessmen have found it advantageous to enter the ranks of the employed, and many individual professional workers have joined service-oriented corporations and government agencies. This does not serve to explain the falling share of profits, however; it merely indicates a switch between the income of employees and the incone of individual entrepreneurs.

In a study of the falling share of profits focusing on the rate of return to capital in the United States, William Nordhaus points out that during periods of inflation the tax system enhances the advantages of debt financing of corporate investment over the use of retained earnings, Assuming that average corporate tax rates have risen since World War II, it

has become advantageous for corporations to finance investment through the

-4-

issue of debt ttam increasing extent (although this advantage depends on the twin assumptions that average tax rates rise in spite of changes in the taxation system, and that lenders do not fully anticipate inflation in setting their interest rates for corporate borrowers) .2/ Such a result is in keeping with the findings of Modigliani and Miller, that under an unintegrated system of income taxation a firm's debt/equity ratio will be higher than that which would prevail in a system without "double taxation" of dividends, In an earlier study of the profits share, Okun and Perry observe that a shift by corporations toward a greater degree of debt financing would be reflected in a lower ratio of profits to total income, but not a lower ratio of profits to total equity (the rate of return) “1!

For this reason, gross profits data are constructed for each country in this study by including net private interest income (for widest possible comparability, data correspond with the U.S. accounts definition, net of payments on government debt and consumer credit). Gross profits shares and observed corporate profits shares are summarized in Table 1 at the end of this chapter, and presented in more detail in appendix tables 3 and 4,

There has been no consistent decline in the gross U.S. profits share during the 1960's and 1970's: in fact, gross profits since 1960 have averaged about 14,8 per cent of national income, while they averaged 13.5 per cent from 1946 to 1959. The gross profits share was depressed in individual years, such as 1970, 1974 and 1975; however, it recovered to its previous level in each instance (data for the first three quarters of 1976

indicate that the profits share rose in that year to about 14.9 per cent).

2/ Nordhaus (1974), pp 169-173. 3/ Modigliani and Miller (1958).

4/ Okun and Perry (1970), p. 469.

~-5-

Thus, the observed decline in the share of corporate profits in national income during the 1960's and 1970's seems to reflect only a rise in interest payments attributable to a rising debt/equity ratio, rather than a secularly declining rate of return to capital.

As in the United States, the share of compensation of employees in Canadian national income has risen substantially since the first decade after World War II, growing from an average of 64 per cent to 73 per cent in 1975. The share of farm operators and nonfarm unincorporated business has fallen, and the share of interest payments has risen. However, the share of corporate profits has fallen only slightly, from 16.2 per cent to 14.7 per cent in 1975 (and in fact was quite high by historical standards, at 17 per cent, in 1974).

Canada's system of income taxation is partially integrated: individual corporate shareholders receive a 33.3 per cent dividend tax credit, while taxes on "investment income" (including-interest payments, rent, and royalties) are refunded to a corporation on-the basis of one dollar per three dollars of dividend payments, 2! Integration of the tax system serves to reduce considerably the advantages of debt finance over retained earnings for a corporation, and results in a lower debt/equity ratio than that found under an unintegrated tax system, ceteris paribus, Because the high degree of integration of the tax system destroys most of the incentive for a high corporate debt/equity ratio in times of high imflation rates, it was not necessary to adjust the Canadian data to include net interest payments.

5/ In fact, the dividend tax credit is due to be raised to 50 per cent on

January 1, 1978, and other incentives for;equity investment were announced in the current Canadian budget.

-6-

German national accounts do not give a detailed breakdown of national income by factor shares. They indicate compensation of employees defined in a way comparable with other countries' employment income categories, but aggregate all other factor incomes into "property and business income." The share of compensation of employees has risen steadily since the 1950's, with a corresponding decline in the share of property and business incomes from its 1950-1955 average of 41.3 per cent to 28.6 per cent in 1975, Because of the extremely broad level of aggregation of property incomes, I attempted no adjustments to the data. I use property and business incomes as the best available proxy for corporate profits. (The ‘analogous U.S. factor share, defined to include all non-labor incomes, declined from a 1946-1955 average of 32.8 per cent to 23.1 per cent in 1975.)

As is the case in Germany, Italian national accounts do not now permit detailed analysis of factor incomes. The only consistent time series data available in recent years (a better breakdown is available until the 1970's) aggregate all types of non-labor income. The share of income other than compensation of employees in Italian national imcome has declined steadily since the early 1950's, from its 1951-1955 average of 51.4 per cent to 29.8 per cent in 1975,

The 1970's have seen a drastic decline in the share of unadjusted profits (defined roughly as in the United States) in Japanese national income, from 15.5 per cent to 8.6 per cent in 1975. At the same time, the share of compensation of employees rose from 54.5 per cent to 63.4 per cent. The share of personal property income (rent, dividends, and interest) has risen since 1952, and the share of income from unincorporated business has fallen considerably. When net interest payments are added to corporate pro-

fits, the gross profit hare shows no decline in the 1960's or 1970's.

-7-

However, the gross profits share was depressed in 1974.

British national accounts show a decline in the share of corporate profits in national income since the first decade after World War II: this share fell from its 1946-1955 average of 17.2 per cent to 10.4 per cent in 1975. As in other countries, the share of income from employment has risen; however, there has been an atypical rise in the share of rent and selfemployment income in national income as well. On the basis of the rise in the share of rent and self-employment income in national income and labor force data which show a disproportionate increase in the self-employed labor force, it would appear that corporate employment has become less attractive relative to self-employment during the 1960's and 1970's.2/

“Gross trading profits of private corporations" are defined in the British national accounts to exclude corporate property income and rent, do not include IVA, and are gross of dividend payments, tax, and interest. Prior to 1973, the British tax system was unintegrated. Thus, in order to render U.K. gross profits data comparable with U.S. gross profits, corporate rental and property income were added and the corporate IVA subtracted from the British data. Even these adjusted data show a steady decline in the profits share, from its 1946-1955 average of 18.7 per cent to 16.4 per cent in 1973 (and, in spite of the institution of a dividend tax credit for private holders of corporate shares in 1973, this share declined further to

12.3 per cent in 1975).

6/ While the dependent labor force of the United Kingdom grew 4.5 per cent between 1960 and 1974, the entrepreneurial labor force grew 14.3 per cent.

-8-

For four countries for which gross profits data were constructed, the gross profits share has declined much less than the unadjusted corporate profits share, as is shown in Table 1 below. Most of the evidence of a declining share during the 1960's and 1970's was eliminated for the United States, Canada, and Japan, Evidence of a considerable decline persisted for the United Kingdom; a considerable decline also occurred in the ratio of total non-labor income to national income in Germany and Italy.

eee TABLE TI

A. Share of Unadjusted Corporate Profits in National Income (per cent):

United United States Canada Japan Kingdom Average to 1955 12.4 16.2 9.0 17,2 1960 - 1964 11,9 14,1 13.3 15.3 1965 - 1969 12.3 14.4 12.7 13.9 1970 - 1975 8.6 14.2 11.3 12.6 B, Share of Gross Corporate Profits * in National Income (per cent); United United States Canada Germany italy Japan Kingdom Average to 1955 13.5 16.2 41,3 51,4 10.4 18.7 1960 - 1964 14.7 14,1 36.9 45.6 17.2 18.4 1965 - 1969 16.0 14.4 34.9 43.4 17.2 17,2 1970 - 1975 13.9 14.2 30.6 35.4 16.7 15,0

eee

* Non-labor income for Germany and Italy, (See also appendix tables 3 and 4,)

-~9-

2. Cyclical Variation in the Gross Profits Share.

Profits differ rather fundamentally from labor income in their cyclical volatility. Labor income is constrained both in cyclical upturns and downturns to stay near its trend value by adjustment lags. — For instance, collective bargaining fixes the wage scale in unionized sectors for long periods of time, so that there is little room for wage adjustment during the business cycle. In the past, firms have also been reluctant to adjust their labor force fully downward during recessions, due to training and other fixed costs of hiring new workers once the cyclical upturn begins.

However, profits are a residual, determined by the interaction of costs and revenues. Labor costs adjust incompletely to depressed demand in a recession; similarly, as fixed capital cannot be varied without a considerable lag, the rate of firms' capacity utilization drops ina recession. Each of these influences ensures that costs will not decline in proportion with a drop in revenues: average fixed costs rise, the wage bill does not adjust fully downward, and unit labor costs rise further due to reduced productivity at low levels of output. By the same token, the existence of underemployed labor and excess capacity in the cyclical upturn ensures that profits will then show a disproportionate increase,

This behavior of profits during the business cycle corresponds with a reduced form equation used by Arthur Okun and George Perry to relate the share of corporate profits to economic activity. The equation was estimated with annual data for 1954-1966, with these results:

(1) $= 0.1052 - 0.287 £+ 0.076 (_,

(15.11)* (4.60)

R = 0,956

- 10 -

where $ is the share of gross corporate profits in GNP, G is the gap between potential and actual GNP as estimated by "Okun's Law," and g is the ratio of the GNP gap to actual GNP. On the basis of the profits shares predicted from this equation for 1967-1970, Okun and Perry described a "profits squeeze" in 1969 and 1970, a situation in which the actual share of gross profits in GNP was consistently about one per cent lower than the predicted share_/, The signs of the coefficients of g and G4 are those which are predicted by our discussion of the cyclical volatility of profits. € rises in cyclical downturns as actual output falls further short of potential, and falls during a recovery as actual output becomes more nearly equal to potential. Thus, the immediate effect of changes in corporate revenues during the business cycle should cause the first coefficient to be negative. The positive sign of the second coefficient reflects the lagging adjustments in capacity and labor force which firms make in order to improve profitability. The opposite signs also reflect the fact that business cycles in the United States typically have

G

been about two years long, so that Yy is large when ©). is small (and vice

1

versa).

The Okun-Perry equation did not perform nearly as well when I extended its sample period to 1975, using the Cochrane-Orcutt method to correct for first-order serial correlation of residuals. Reestimated for 1954-1975, the equation had a smaller coefficient of determination, and, more importantly,

had coefficient signs the opposite of those estimated for 1954-1966:

(2) é = 0.1426 + 0.001 < - 0.002 @) ; (2.69) (4.86) 7 R= 9.7612; DW = 1.3102; 3 = 0.7669

> i/ Okun and Perry (1970), pp. 466-469.

il -

It is not surprising that the explanatory power of Okun's GNP-gap term has declined since the 1960's. The variable was constructed according to "Okun's

‘ which assumes that potential GNP equals actual GNP at a four per cent

Law,' unemployment rate, Numerous studies of labor market developments during the past several years suggest that this is no longer the appropriate structural unemployment rate: labor force participation rates among groups with higher unemployment rates (such as women and the young) have risen during the 1970's, and the generally high unemployment rates of the period have reflected this. In order to "correct" the Okun- Perry equation for a higher structural unemployment rate, I substituted a time series of unemployment rates normalized for a roughly constant age-sex composition which was computed by Michael Wachter for the constant four per cent rate assumed in "Okun's Law." Wachter's normalized unemployment rate was about four per cent from 1953 to 1958, and rose to over 5.5 per cent in 1974 before subsiding to its current 8) estan

estimated level of 5.3 per cent= This series was used to reestimate

equation (1) for 1954-1975:

(3) G = 0.1423 - 0,0007$ + 0.0021 &)_, (2.35) (6.71) R° = 0.7861; pw = 1.2082; 6 = 0.8723

Here the estimated coefficients have the predicted signs.

Unfortunately, the estimation-of Wachter's normalized unemployment rate for the other five countries requires more detailed data on labor markets than is readily available. The discussion of eyelical volatility of profits suggested an alternative specification which involves data available for all six countries:

R

(4) Y > ag + a, CAPU +

8/ Wachter (1976), pp. 126-133. _

-~ 12 -

where CAPU is the Wharton Economic Research Unit index of capacity utilization and DU is the change in the unemployment rate (used here as a proxy for firms’ adjustment of labor force). Because profits were assumed to vary directly with the rate of capacity utilization, ay should be positive. The adjustments to labor force for which DU is a proxy are the lagged responses which firms make in order to improve profitability, so a should also be positive. Estimated through 1975, equation (4) gave highly unsatisfactory results for U.S. data, with an incorrect coefficient sign for DU and a CAPU coefficient which was not significantly different from zero. Results for the other five countries were generally even worse than for U.S. data.

I therefore chose to estimate a reduced-form equation substituting an indirect indicator of the rate of capacity utilization for CAPU in equation (4). Because adjustments to capacity lag the business cycle, high real GNP growth rates should be associated with high rates of capacity utilization, and low real GNP growth with low capacity utilization, Equation (4) was reestimated substituting the real GNP growth rate (y¥/y) for CAPU:

65) 2 = a 4 ay (y/y) + a, DU

Y oO

The unemployment term DU had an estimated coefficient not significantly different from zero, but (y/y) appeared to have.considerable explanatory power. Again reestimating with only the real GNP growth rate as an explanatory

variable:

i b=)

(6) = 0.1399 + 0.0025 (y/y) - (7.69)

—2 “i R = 0.8414; DW = 1.0948; p = 0.8502

Equation (6) seemed to fit the data quite well, although the Durbin-Watson

statistic indicated that there was still autocorrelation present at a 5 per

-13-

cent significance level. More important, however, is the fact that it generated very similar fitted values of the profits share to those obtained from equation (3), which used Wachter's normalized unemployment rate to estimate the GNP gap. When (3) and (6) were reestimated for 1951-1969 and the estimated coefficients were used to Predict profits shares for 1970-1975, the two sets of predictions were also very similar. Not only were the point estimates nearly the same, but when a 95 per cent confidence interval was estimated for each predicted value, the predicted values were found to be not significantly different from the actual profits share for any year.

I therefore chose to use equation (6) as a simple predictor of "eyclically normal" profits shares for the six countries included in this study, as (6) is easily estimated for all of the countries. The estimated parameters for the sample period to 1969 were used to predict "normal" profits shares for 1970-1975. The results are summarized in appendix table 5.

For five of the six countires, equation (6) fit data for the sample period to 1969 quite well; for the sixth, the United Kingdom, the economic relationships were rendered unstable by wide fluctuations in government policy stance during the sample period. For the United States, Germany, and Japan it appeared that actual profits shares have not been significantly different from the predicted "normal" profits shares during the 1970's. (Ina recent article, Feldstein also concluded that the recent decline in the U.S. profits share was a purely cyclical phenomenon2’ , ) The Canadian profits share oscillated about the "normal" profits share, most recently tending to be significantly greater than the "normal" share. The Italian share was greater than "normal" in 1970, and has ghown only the decline consistent with

economic activity in Italy since then. The British profits share has tended

in recent years to be consistently smaller than its "cyclically normal" level. 9/ Feldstein (1977), pp.116-117.

- 14 -

3. "Full-Employment'" Profits Shares.

Another potential application of this exercise is the use of the estimated historical relationships between the profits share and economic activity to estimate "full-employment" (gross) profits shares. Once again, the use of equation (6) in this context requires some justification. Equation (3) provides an easy measure of the full-employment share: since potential GNP equals actual GNP at full employment, g equals zero, and the full-employment profits share is simply the intercept term for equation (3) (for the United States, this implies a full-employment profits share of 14.65 per cent). Equation (6) could be interpreted as providing the fullemployment profits share when its fitted coefficients are applied to the trend rate of growth of potential output. Using the average annual growth rate of real GNP from 1960 to 1972 as a proxy for the trend rate of growth of potential output, the U.S. full-employment share implied by (6) is 14.8 per cent, nearly equal to that implied by (3). With this further assurance that the results of (3) and (6) were not inconsistent, I used (6) to estimate full-employment (gross) profits shares for the other five countries, again using the average annual real GNP growth rate for 1960-1972 as a proxy for the trend rate of growth of potential output. The estimated full-employment shares were not particularly sensitive to the exact growth rate chosen for potential output: the full-employment shares implied by rates one per cent lower than the average annual real GNP growth rates for 1960-1972 are nearly the same. Both sets of estimated full-employment profits shares are shown

in table 2 below.

-15eee

TABLE 2

Full-Employment and Actual Profits Shares (Per cent)

United United States Canada Germany Italy Japan Kingdom

"Full-Employment" Share 14.8 14,3 31.8 41.9 17.0 18.0

"Full-Employment" Share for 1% Lower Growth Rate 14.5 14.1 31.5 41.5 16.7 17.9

1970-1975 Actual Average Share 13.9 14.2 30.6 35.4 16.7 15.0

According to either set of estimates, the observed share of (gross) profits in national income during the 1970's has been about equal to the "full-employment" share, on average, in Canada and Japan. The average observed share was slightly lower than the "full-employment" share in the United States and Germany, and considerably lower in Italy and the United

Kingdom.

CONCLUSION

The first part of this paper presented evidence that much of the decline in the share of corporate profits in national income observed in Western industrial countries during the 1970's has resulted from changes in the distribution of gross corporate profits among dividends, interest, and retained earnings, rather than from a secular decline in corporate profitability. The next part indicated that much of the variation in the gross profits share during the 1970's has resulted from the predictable responses

of the profits share to the cyclical pattern of economic activity. It also

- 16-

appeared that many countries have not experienced profits shares consistent with "full-employment" growth paths.

The fact that low profits shares recently experienced in some industrial countries are consistent with normal responses to business cycles in those countries may be interpreted as implying that low profits shares are a transitory cyclical phenomenon. However, this need be the case only if the relatively sluggish pace of economic activity which caused the low profits shares is itself a transitory phenomenon, The high explanatory power of the real GNP growth rate in accounting for changes in the profits share points this out clearly. If growth rates below historical trend rates were to persist in response to structural changes in industrial economies, low profits shares would persist as well unless the institutional arrangements under which corpo-

rations operate were to be altered,

APPENDIX

Table 3, The Share of Corporate Profits in National Income: Unadjusted Data

United United States Canada Germany Italy Japan Kingdom Average for decade

to 1955 12,18/ 16,2b/ 41.3o/ 51,L4/ 9, 08/ . 17,28/ 1960 11.3 13.4 39,4 48 ,2 14.4 16.5 1961 11.1 13.7 37.8 48,1 14,4. 15,0 1962 12,0 13.7 36,0 46 4 13.3 14,2 1963 12.3 14,2 35.6 43,2 12,0 15.3 1964, 12,9 15.5 35,7 42,0 12.3 15.5 1965 13.6 15.3 3543 42,9 10.6 15.2 1966 13.3 14.5 34.3 43,7 11,2 13.9 1967 12,1 13.7 34,1 4345 12.8 13.3 1968 12,0 14.7 36,1 43,4 14.3 14,1 1969 10.6 13.7 34.8 43.5 14.7 13.1 1970 8.5 12,0 33.3 40,7 15.5 12.6 1971 9,0 12.3 31.7 37.5 13.9 12.5 1972 9.7 13.5 31.3 36.9 12.5 12.7 1973 9.3 15.6 30,2 3563 10,2 13.8 1974. 7.5 17,0 28.7 32,4 7,0 13.5 1975 7.6 14.7 28.6 29.8 8.6 10,4

a/ 1946-1955 average, b/ 1947-1955 average, c/ 1950-1955 average for all non-labor incomes, d/ 1951-1955 average for all non-labor incomes, e/ 1952-1955 average,

Table 4, The Share of . Gross Corporate Profits in National Income

United United States Canada Germany italy Japan Kingdom

Average for decade

to 1955 13,58/ 16,2b/ 41,39/ 51,43/ 10,42/ 18, 78/ 1960 13.7 13.4 394. 48,2 17.9 ~ 19,9 1961 13.7 13.7 37.8 48,1 18,2 18.3 1962 14.8 13.7 36,0 46.4 17.2 17.4 1963 15.3 14,2 35.6 43,2 16,1 18,1 1964 16,0 15.5 35.7 42,0 16.6 18,3 1965 16.9 15.3 35.3 429 15.2 18.2 1966 16.8 14.5 34.3 43.7 15.8 17,0 1967 15.8 13.7 34.1 43.5 17.3 17.1 1968 15,8 14.7 36.1 - 43,4 18.7 17.4 1969, 14.6 13,7 34.68 43.5 19,2 16.4 1970 13.2 12,0 33.3 40,7 20,2 15.4 1971 14.0 12,3 31.7 37.5 18.8 15.7 1972 14.6 13.5 31.3 36.9 17.5 16,1 1973 14,2 15.6 30,2 35.3 15.3 16.4 1974. 13,4 17,0 28,7 32,4 13.1 14,1£/ 1975 13.8 14.7 28.6 29,8 15.2 12,3£/

1946-1955 average,

1947-1955 average,

1950-1955 average for all non-labor incomes,

1951-1955 average for all non-labor incomes.

1952-1955 average,

£/ As a result of partial integration of the British tax system in 1973, adjusted profits data for 1974 and 1975 are not strictly comparable with the earlier data,

tele ele ie

Table 5

A. SUMMARY OF REGRESSION RESULTS FOR EQUATION (6)

Coefficient -2 Number of Country Constant of (y/y) R DW 6 Observations

United States - 1387 20023 28292 1.3913 6856 21 (17.64) (7.15)

Canada » 1304 0024. 5791 1.7717 462 14. (29,79) (3.20)

Germany* 23099 20025 .9102 1.3853 -928 13 (11,89) (4,10)

Italy 44006 . 0037 .8905 1.8537 2857 13 (18,28) (2,02)

Japan 1451 . 0024, 25835 1.6715 ATA 13 (11.58) (2.32)

United Kingdom .1769 0013 4640 2.3294 2753 18 (28,30) (1.75)

B. COMPARISON OF PREDICTED AND ACTUAL PROFITS SHARES (Indicates whether the actual profits share is less than (-) or greater than (+) the 95 per cent confidence interval about the predicted share.)

United United States Canada Germany Italy Japan Kingdom 1970 (-) (+) 1971 (-) 1972 1973 (+) 1974 (+) (=) 1975 t=)

* The German equation included a vector with value zero until 1960, and value

one from 1960 to 1975, This is intended to correct for the fact that German national accounts prior to 1960 exclude data for Berlin and the Saar, and include them thereafter,

BIBLIOGRAPHY

Feldstein, Martin, 'Does the United States Save Too Little?" American Economic Review 67 (February 1977): 116-121,

Modigiiani, Franco, and Miller, Merton H,, "The Cost of Capital, Corporation

Finance, and the Theory of Investment,''! American Economic Review 48 (June 1958): 261-297.

"Corporate Income Taxes and the Cost of Capital: A Correction," American Economic Review 53 (June 1963): 433-443.

Nordhaus, William D., "The Falling Share of Profits," Brookings Papers on Economic Activity (1: 1974): 169-208. ;

Okun, Arthur M., "Potential GNP: Its Measurement and Significance," American

Statistical Association, Proceedings of the Business and Economic Statistics Section (1962): 98-104,

Okun, Arthur M., and Perry, George L., "Notes and Numbers on the Profits Squeeze," Brookings Papers on Economic Activity (3: 1970): 465-472.

Shoven, John B., and Bulow, Jeremy I., "Inflation Accounting and Nonfinancial

Corporate Profits: Physical Assets," Brookings Papers on Economic Activity (3: 1975): 557-611.

» "Inflation Accounting and Nonfinancial Corporate Profits: Financial

Assets and Liabilities," Brookings Papers on Economic Activity (1: 1976): 15-57.

Wachter, Michael L., "The Changing Cyclical Responsiveness of Wage Inflation," Brookings Papers on Economic Activity (1: 1976): 115-159.