Printer Friendly

Investment, valuation, and dynamics: a case study in the law of supply and demand.

The author seeks to explain the significant increase in dividends paid by nonfinancial corporations between 1988 and 1990 while the overall business situation was deteriorating. He demonstrates that managements were in effect writing up the values of their corporations due to a swing to a higher rate of investment, leading to increased employment and also increased profitability.

THE BEHAVIOR OF corporate managements sometimes seems to defy explanation. Between 1988 and 1990, dividends paid by nonfinancial corporations rose from $82 billion to $112 billion, or by 35 percent in just two years. How can we make sense out of such generosity to stockholders while the overall business situation was deteriorating, while the outstanding debts of these corporations were swelling by 24 percent, and while their interest payments were climbing from 47 percent to 64 percent of pretax profits? This leap in dividend payments was all the more remarkable compared to the much slower 8 percent annual growth rate of dividends over the preceding ten years.

Yet this extraordinary behavior was no flash-in-the pan. We can make sense out of it. I shall demonstrate that the dividend surge signifies a profound change in the corporate environment, with important and positive implications for the outlook for business investment and labor productivity in the United States. Furthermore, I shall base my case on nothing fancier than the Law of Supply and Demand, which, slowly but surely, worked its magic on all the developments that led to this striking shift in dividend payouts.

Decisions to raise dividends are never taken lightly. The penalties from having to reverse those decisions take a heavy toll on a company's stock price and the credibility of its management and board of directors. The seriousness surrounding the dividend decision means that dividend payments are the most reliable reflection of corporate management's view of long-run earning power. This is the sense in which we must examine the unusual developments at the end of the 1980s.


That big increase in dividend payments tells us that managements were in effect writing up the value of their companies. It would be interesting to see how corporate values as seen by managements compare with the judgments of the owners of the companies -- the stockholders.

The most meaningful method for calibrating management valuations is to relate their dividend decisions, or their view of long-run earning power, to the assets that generate that earning power. One of the great lessons of the 1980s was that assets matter and that earnings over the long pull are not merely a function of today's profits or the growth rate of the past x-number of years. Therefore, I have calculated the ratio of nonfinancial corporate dividend payments to the current, or replacement, cost of their stock of productive capital, basing my calculations on the Bureau of Economic Analysis constant cost and current cost estimates of the gross stocks of business capital owned by nonfinancial corporations(1)

Figure 1 shows the thirty-eight-year history of this dividend/GSBC ratio and compares it to the price that investors in the stock market have been willing to pay for those same dividends.(2) The two ratios trace roughly similar paths, but the divergences are important. The stock market was optimistic about the economy of the 1950s earlier than corporate managements, who waited until Camelot to show their enthusiasm. But business executives smelled trouble well ahead of stock market investors in the late 1960s, while both groups shared the gloom of the 1970s. Managements' subsequent reluctance to join the market in celebrating the ebullience of the 1980s makes their change of heart after 1988 all the more remarkable.


The sharp rise in the ratio of dividends to capital stocks during 1989 and 1990 must imply something more than just an improved outlook for profits. Even pressure to shed cash as a result of the takeover craze is an insufficient explanation. The outpouring of dividends suggests that managements were beginning to take a more hopeful view of the productivity of the capital stock itself.

We can discover the sources of this optimism by examining variations in the capital stock in relation to the number of employees hired to work on it. I begin that examination with the history of the capital/labor ratio that appears in Figure 2, which shows the changing relative inputs of the two primary factors of production since 1947. Labor input equals total hours worked in the nonfarm corporate sector; capital input equals the inflation-adjusted, constant cost value of the gross stocks of business capital of nonfinancial corporations.

Note that capital intensity rises over time, but in an irregular fashion. Steep upward moves in the capital/labor ratio occur during recessions, as employees are laid off while fixed assets remain in place. The ratio then stabilizes in the early stages of expansions, as laid off workers are put back on the payroll, after which the ratio drifts upward as prosperity encourages an accelerated rate of business investment.

As in so many other areas, the experience of the 1980s does not fit neatly into this mold. There is nothing in earlier expansions that compares with the extended sideways movement of the capital/labor ratio from 1982 to 1990. From 1947 to 1982, the capital stock grew nearly three times as fast as labor hours worked. Since 1982, the capital stock has grown only 40 percent faster than labor input, which increased at a trend rate of 2.4 percent a year, nearly double the 1.3 percent annual rate recorded from 1947 to 1982.

This curious behavior of the production function has a simple explanation, which will be the first time I shall invoke the Law of Supply and Demand to explain what was happening. That step requires us to transform the physical measurement of labor and capital inputs as shown in Figure 2, into monetary measurements of the input of labor and capital as shown in Figure 3. This chart traces the ratio of employee compensation in nonfinancial corporations to the current cost, replacement value of their capital stocks.

The object of the exercise at this point is to determine changes in the relative costs of the two factors. Although employee compensation is a straightforward measurement of the cost of labor inputs, the current cost value of the capital stock is an unconventional measure of the cost of capital. Business school graduates today use elaborate models to calculate the cost of capital. But in a cruder, if no less relevant, sense, managers deciding on how to structure the productive process between labor and capital goods may simply consider such matters as the size of the current payroll, the number of people around the place compared with the cost and number of machines or square feet of floor space, and how to price a product to meet with customer acceptance and still earn a profit. The ratio that I have used here is a quick-and-dirty measure, but it does combine the difference in inflation rates between the two factors with the total physical volume of capital equipment and the total number of bodies on the payroll.

The most noticeable feature of Figure 3 is the way that labor costs raced ahead of capital costs during the 1960s. The bulge in labor costs was the consequence of an economy that was expanding at a much faster pace than the labor force, which still reflected the low birthrates of the years from 1929 to 1945 -- another example of the Law of Supply and Demand at work. Unemployment was 6.5 percent of the labor force in 1961, but it had shrunk to only 3.5 percent by 1969.

The situation started to shift after 1969, although not because labor costs were slowing down in an absolute sense -- they were not. Change came about as inflation began to have an increasing impact on the current cost value of the capital stock. In addition, as the war babies matured, the supply of labor, instead of growing more slowly than total population, started to balloon; in 1969, 40 percent of the population was in the labor force, but that number had risen to 47 percent by 1980.

By the end of the 1970s, the more rapid growth in the supply of labor had pushed employee compensation to a point that was once again low relative to my estimate of the cost of capital (which, to satisfy business school techniques of measurement, was also elevated at that point by historically high interest rates on debt plus single-digit price/earnings ratios on equities). The Law of Supply and Demand had decreed that labor was to become the cheaper factor of production.

But we can see that the relative positions began to change in the opposite direction during the business expansion that got under way in 1983. Labor has gradually become more costly than capital over the past seven or eight years, in part because labor force growth has eased off relative to the growth of the total population, although the recent development has been at a far more leisurely pace than during the 1960s.

The financial developments reflected in Figure 3 explain the variations in the physical measurements of the capital/labor ratio displayed in Figure 2. It was the inflation in labor costs in the 1960s that led to the steepening climb in the capital/labor ratio in the 1970s. The whole objective was to economize on labor, and the unemployment rate followed suit by rising from an average of 4.5 percent during 1963-72 to 6.9 percent from 1973 to 1982. The objective was achieved, and the relative cost of labor declined. By the time the 1980s rolled around, all the incentives pointed toward hiring more labor and economizing on capital goods.

The result was that unusually flat trend of the capital/labor ratio in the 1980s. This pattern reflects the acceleration in labor input, but it also reflects the wave of "restructurings" that amputated the obsolete and nonproductive pieces from a bloated capital stock, slowing the overall growth of the capital stock during that decade.

But as the cycle repeats itself, and as labor becomes the more expensive factor of production, more recent headlines are bellowing once again about slashing reductions in employment. Plants are still closing down, but now in relatively few industries. Unemployment, on the other hand, is widely dispersed throughout the economy and hitting in some areas that had never been troubled with layoffs in the past. Much of the unemployment shows signs of being structural -- the job losses are permanent -- rather than cyclical. If labor is the problem, will capital be the solution?


The answer to this question will also come from the Law of Supply and Demand. As one factor of production becomes less costly than the other, we should expect to see its use increase. As the cost relationships shift, so should the inputs.

To be realistic, we should not expect the response to be instantaneous: some delay is inevitable. When capital is the expensive factor, business firms will not immediately discard hunks of the capital stock just because they are costly. The productive process is too integrated for such hasty decisions, as each factory, computer, refinery, commercial building, or communications system becomes a nucleus for all kinds of organizational structure deployed around it. When the relative positions go into reverse and labor is the more expensive factor, managements will not hurry to add fixed capital at the first sign that they should do so. Each addition to fixed assets is sunk money, not readily reversed, which lends high risk to hasty action. The more prudent decision is to wait until the indications have lasted long enough to show signs of durability, after which the whole extended process of planning, budgeting, building, and organizing has to take place.(3)

If the response of the production function to changing cost relationships is slow, it is also sure. I experimented with a series of regressions of the four-year growth rates in the capital/labor ratio on varying lags of the ratio of employment compensation to the current cost capital stock that we saw in Figure 3. The resulting model, displayed in Figure 4 and based on a lag of four years, estimates that an increase of 1 percent in the dollar-based employee compensation ratio leads four years later to an increase of about 0.288 percent in the physical capital/labor ratio. The estimates are robust: the coefficient in the regression of 1951 to 1979 varied by less than .01 percent from the coefficient for the period from 1979 to 1990, with [R .sup. 2]s of 0.44 to 0.57. "t" statistics exceed 5.0.

Some of the model's estimation errors are the result of cyclical swings in employment, which, as I mentioned above, are more volatile than changes in the capital stock. When I introduced the unemployment rate as a second independent variable in the model, the [R.sub.2] jumped to over 0.60 and the errors in the 1980s become more muted. The model in Figure 4 omits the unemployment rate as a variable, in order to focus the discussion on the impressive power of the compensation/capital stock ratio by itself in determining the relative inputs of labor and capital.

The model has overestimated the growth in the capital/labor ratio since 1986. The same tendency toward overestimation is apparent in the 1960s. In the present instance, the peeling off of redundant employees has been a delayed reaction, but it is now rolling like a snowball. The impact in the service industries is most striking, especially as this is the sector in which productivity improvement has been such a serious problem. A most dynamic process is now at work in that part of our economy.


A reduction in the number of employees can rise the capital/labor ratio simply by reducing the size of the denominator. That is what made the ratio jump so noticeably in 1991. There is nothing inevitable that requires the numerator -- the capital stock -- to accelerate at the same time. Yet I stated at the outset of this article that the impressive dividend increases of 1989-90 have "important and positive implications for the outlook for business investment."

I justify this assertion on the basis of Figure 5, which compares four-year growth rates in the real capital stock with the dividend/current cost capital stock ratio from Figure 1. The chart looks like mush at first glance, but it supports the central message of the earlier charts that the Law of Supply and Demand is alive and well. Quantities and resource allocations respond to changes in relative prices.

Free enterprise economies have natural tendencies to overshoot, and that is the essence of the story told by Figure 5. The great investment boom of the euphoric 1960s and early 1970s was fed in part by management strategies that put more emphasis on market share than profitability. In an increasingly competitive global economy shocked by the surge in oil prices, the result was a glut of capital goods that turned out to be uneconomical and prematurely obsolete. As the supply of capital overwhelmed the demand for it, profitability was impaired and managements began to lower their estimates of the long-term earning power of their assets. A less optimistic view of earning power naturally led to a more subdued rate of growth in the capital stock.

The slowdown in the growth of capital was gradual at first, as the inflation of the 1970s served to validate temporarily much of the unproductive capital. Major surgery had to wait until the disinflation of the 1980s. The primary story of the years since 1981 has been the elimination of the ill-conceived investments of the preceding decade as "restructure" became the key business buzzword.

The surgery has been so protracted and the cuts so deep and so widespread that the redundancy of fixed assets may at long last be nearing an end. The overhang of commercial construction is still with us, but in manufacturing, transportation, and even in retailing, huge amounts of the capital stock have simply disappeared from the scene.

There is some interesting statistical confirmation of what is so vividly apparent on an anecdotal basis. From 1962 to 1981, an average of thirty-eight cents out of every dollar of real nonresidential investment was associated with an increase of real capital stock of nonfinancial corporations, with a standard deviation of four cents; the balance was offset by withdrawals and discards from the stock. There was an abrupt and striking change in this ratio after 1981. From 1982 to 1991, an average of only twenty-six cents per dollar of new investment was associated with an increase in the real capital stock, with a standard deviation of two cents.

These calculations suggest that new investment was replacing a significantly larger share of the old capital stock during the 1980s than it replaced during the two preceding decades. At first glance, one might deduce from the data that the average life of the capital stock has simply grown shorter over time, which has come about as more and more nonresidential investment has taken the form of equipment rather than structures. But that process has little to do with the figures I cite here: the share of equipment in total fixed nonresidential investment started to shrink way back in the early 1960s and, abstracting from the commercial building boom of the first half of the 1980s, has not accelerated in recent years. Furthermore, as the standard deviations reveal, the downshift in the ratio of capital stock growth to new investment was abrupt after 1981, not a gradual process spread out over time. There are two entirely different environments for the growth in the capital stock relative to investment, one before 1981 and one after.


All of this analysis supports the conclusion that the process of adjusting the capital stock to the investment errors and overenthusiasm of the past is well advanced. The inevitable oscillations of our capitalistic, free enterprise system give every sign of swinging back toward higher rates of new investment and a more rapid growth rate in the stock of capital goods. The dividend increases of 1989 and 1990, in a less than exuberant business environment, are a loud and clear message of what lies ahead.

The good news in this forecast extends beyond the outlook for investment. The capital/labor ratio may be heading for an increase, but that process is likely to lead to more jobs at the same time. The rise in the capital/labor ratio during the 1960s was persistent and strong, but it was also accompanied by 2 percent a year growth in civilian employment.

The prospects for an acceleration in business investment described here receive further support from the growing importance of corporate cash flows as a stable and reliable source of funds, as capital consumption allowances have grown in importance relative to retained profits. The coefficient of variation -- the ratio of standard deviation to mean -- of corporate cash flows averaged 0.31 during the 1960s and 1970s but then fell to only 0.20 during the 1980s. Capital consumption allowances as a percentage of total cash flows jumped from an average of 60 percent during the 1960s and 1970s to 82 percent in the 1980s.

The precise timing of the forecast given here is difficult because the recession has interrupted the ordained sequence of events, and upturns in business investment always lag cycle troughs. Nevertheless, the dividend increases of 1989 and 1990 predict that the revival in investment as recovery develops is likely to exceed most current expectations. That prediction has solid underpinnings in the Law of Supply and Demand.


(1)As the BEA has not yet published estimates for 1991, I have assumed that the capital stock continued to grow at ten basis points below the rate during the preceding four years. If this assumption is wrong, I have probably overstated the capital stock for that year.

(2)That sentence is less than completely accurate. The price/dividend ratio is for the Standard & Poor's Composite Index of 500 stocks, but this index serves as a satisfactory proxy for the present purpose.

(3)For an illuminating and systematic analysis of this view of the investment process, see Robert Pindyck, "Irreversibility, Uncertainty, and Investment," Journal of Economic Literature, September 1991.
COPYRIGHT 1992 The National Association for Business Economists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Bernstein, Peter L.
Publication:Business Economics
Date:Jul 1, 1992
Previous Article:The emerging bank structure of the 1990s.
Next Article:Replacing GNP: the updated system of national economic accounts.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters