Ten Principles of the Economics of Capital

Capital theory is one of the most difficult topics in all of economics. This may explain why political leaders who want to “get the country moving again” so often adopt policies that lead to economic stagnation. In what follows, we briefly describe the principles of the economics of capital and their implications for public policy.

Principle 1: Capital is the single most important determinant of real wages. In a very real sense, the amount of capital in our economy determines how much wage income we earn, even if we do not personally own any capital. Workers’ wages and the capital stock are inextricably linked. The only way that the real wages, and thus the well-being, of workers can rise is if there is more capital per worker. In general:

• About 98 percent of the variation in real wages over the past 37 years can be explained by the capital-to-labor ratio alone, without reference to any other economic factor.

• For every 10 percent increase in the average amount of capital per worker, the real wage rate increases by 11.9 percent. [See Figure II.]

• In recent years, there has been considerable discussion about which policies will raise the average income of the average worker. The answer is those policies that encourage more capital formation.

• For every additional dollar of income produced by a larger capital stock, two-thirds goes to labor and only one-third to capital.

• After taxes and depreciation, the discrepancy is even greater; labor receives 43.7 cents of each additional dollar of sales, while owners of capital receive only 3.7 cents.

• In other words, workers get to keep $12 in aftertax wages for every $1 of additional aftertax income to investors.

These facts have dramatic public policy implications. In general, public policies that promote capital accumulation primarily benefit wage earners, while policies that discourage capital accumulation primarily penalize wage earners.

In today’s political debate, it is common for politicians to assert or imply that taxes on income from capital only affect the well-being of the rich. For example, those who argue for a higher tax rate on capital gains income frequently imply that average-income families will be better off, since the rich will bear a larger share of the burden of government. They conveniently ignore the fact that less capital means lower wages for everyone, including those who own no capital.

Principle 3: The amount of capital is determined by investment. The nation’s capital stock is the sum total of all of its capital goods. Because these goods lose value over time, some level of investment is necessary to maintain the capital stock at its current size. Beyond that level, additional investment will cause the capital stock to grow, whereas less investment will cause it to shrink. Where do new investment funds come from? One source is increased savings by U.S. citizens. A second is investment by foreigners. A third is the repatriation of funds invested in other countries by U.S. citizens. Just as foreign investors have part of their portfolios in U.S. assets, so many U.S. citizens have investments overseas. When investment opportunities in the United States become more attractive relative to options in the rest of the world, U.S. investors will allocate more of their funds to the United States and less to other countries.

Principle 4: The amount of investment is determined by the real aftertax rate of return on capital. The amount of physical capital available in our economy depends on the willingness of people to invest in productive capital goods. In making these decisions, investors are guided by the return they will receive. The income to the investor must be adjusted for inflation, depreciation, taxes and the riskiness of the investment. After these adjustments are made, the investor can assess the aftertax real rate of return on the investment. For as long as economists have worried about macroeconomic fluctuations, they have attempted to explain the behavior of investment spending. For John M. Keynes, investment decisions were largely irrational acts. For economists in the Keynesian tradition, investment decisions are primarily determined by market rates of interest. Both these views are inconsistent with the facts. The evidence shows that investment decisions are based on the rate of return investors expect to earn.

• As Figure IV shows, changes in investment spending are very closely related to the rate of return on capital.

• In fact, a simple correlation can explain about 75 percent of the changes in investment spending in terms of changes in the return on capital alone.

Principle 5: Because of changes in investment spending, the aftertax rate of return on capital tends to be constant. Suppose something happens to cause the rate of return on capital to rise above its historic average. There will be an increase in investment, adding to the current stock of capital. As the capital stock expands, the rate of return on capital will fall. Conversely, when the rate of return on capital is below its historical level, there will be a decrease in investment. As the capital stock shrinks, the rate of return on capital will rise. In this way, changes in investment spending tend to keep the rate of return on capital constant over time. As Figure V shows:

• Over the past 37 years, the rate of return on capital in the U.S. economy has tended to be remarkably stable — averaging about 3.3 percent per year.

• This stability has persisted despite radical changes in the structure of the economy, significant changes in technology and substantial changes in the taxation of income from capital.

• Events that change the rate of return on capital, such as changes in the tax law, rarely cause variations of more than 1 percentage point above or below the long-term average.

• A return to the rate of 3.3 percent usually occurs within five years following a significant deflection, and 60 percent of the adjustment occurs within two years.

Measuring the Rate of Return on Capital. In our complex economy, measuring the economy-wide rate of return on capital is difficult. The U.S. Department of Commerce has carefully recorded 37 different types of physical capital in the United States since 1929. We have used this information to construct estimates of investment in each of 73 industries. Because a particular type of capital may have a different productivity and a different useful life depending on the industry in which it is used, there are in principle 2,701 discrete types of capital on which data are maintained. Each type is also affected by significant differences in the tax treatment of capital income in corporations and unincorporated businesses.

Based on the Department of Commerce data set and the U.S. tax law, we have calculated an economy-wide aftertax rate of return on capital for each of the last 37 years. The method used to make these calculations adjusts the gross return to capital for depreciation, inflation, taxes and risk. The results of the calculations are shown in Table I.

Other Measures of the Return on Capital. Table II shows the gross and aftertax returns to capital in manufacturing, calculated by the Office of Business Analysis, U.S. Department of Commerce. These data, constructed from survey information collected by the Census Bureau, are consistent with the economy-wide estimates for the real aftertax rate of return to capital presented in Table I. The average aftertax rate of return for manufacturing is somewhat higher than for the economy as a whole, reflecting the somewhat higher risk.

The aftertax rate of return in manufacturing has been remarkably stable over the past 30 years, although less stable than the rate of return for the whole economy. As one sector of the economy adjusts to changing market conditions, the economy-wide average reflects a much smaller variation.

Why the Rate of Return on Capital Tends To Be Constant. The empirical evidence summarized above suggests that the aftertax real rate of return on capital is one of the most constant relationships found in all of economics. But why is that so? Traditional neoclassical economics teaches that the rate of return on capital must reflect people’s preference for future rather than current consumption. In other words, to be induced to save (forgo current consumption) and invest (with the expectation of greater, future consumption), people must receive a minimum rate of return on their investment. Because the time preferences of people are unlikely to change very much over time, the rate of return on capital will remain roughly constant. Furthermore, the sheer size of the capital stock means that an increase in the rate of return for one asset will have little effect on the economy-wide rate. The average return on capital, therefore, moves very slowly.

In a modern, open economy another consideration comes into play — the international flow of capital among countries. U.S. firms compete for capital in an international marketplace. Thus the aftertax rate of return on capital in the United States is determined by the time preferences of people all over the world. Because the United States is a safe haven for capital, the rate of return here will be much lower than it is where investments are riskier, such as most Latin American countries. A 3.3 percent rate is the return necessary to induce international investors to invest in the United States rather than in other countries around the world.

Principle 6: Taxes on capital do not affect the aftertax rate of return on capital but instead affect the amount of capital available.

Although an increase in a tax on capital causes a one-time reduction in wealth for owners of capital, it does not permanently affect the future aftertax rate of return on capital. After such an increase, the aftertax rate of return on capital will be below its historical average. Investors will respond by lowering their rate of investment. The capital stock will shrink (relative to what it would have been) until the rate of return reaches 3.3 percent. After the adjustment has taken place, the owners of capital will receive the same aftertax rate of return they received before the tax increase. This does not mean that owners of capital will be indifferent to taxes on capital. These taxes lower the aftertax future income stream on existing capital assets. Thus a tax on capital lowers the value of capital assets and makes current owners of capital less wealthy. For any new purchase of an asset, however, capitalists will and can expect the normal rate of return of 3.3 percent.

Principle 7: Taxes on capital raise the cost of capital to business and make U.S. industry less competitive in international markets.

In order to supply capital, investors must receive a minimum aftertax real rate of return. In the long run, as we have seen, this rate is 3.3 percent for investments in the United States. The users of capital must pay a much higher rate, however. In addition to paying the suppliers a normal rate of return, the users must pay the cost of economic depreciation (the loss in value of physical assets) and the cost of taxes on capital. The total of these costs is called the cost of capital.

The Cost of Capital in the United States. Table III shows the components of the cost of capital for all real assets in the U.S. economy over the period 1955 to 1990. As the table shows:

• The economy-wide average cost of capital today is 9.7 percent, an 8 percent increase from what it was in 1955.

• Each of the components of the cost of capital — depreciation, taxes and the rate of return to the investors — contributes about one-third of the total cost.

• Corporate income, personal income and property taxes each account for between 10 and 11 percent of capital costs.Fig. IV

The Tax Wedge. Taxes on capital create a wedge between the return to the suppliers of capital and the rate paid by the users. Because in the long run the suppliers of capital adjust their supply to cover all taxes, the full cost of this wedge must be paid by those who use capital in the process of production. Taxes on capital, therefore, raise the cost of capital to business enterprises. In so doing, they also raise production costs and make domestic producers less competitive in international markets.

Taxes, Inflation and Risk. In addition to the direct effect of taxes on the cost of capital, there are indirect effects. For example, in the absence of taxes, a 2 percent rate of inflation would increase the cost of capital (in nominal terms) by 2 percentage points. When taxes are present, the nominal cost of capital will increase by more than 2 percentage points, however. The reason is that the presence of taxes magnifies the effects of inflation. Similarly, the presence of taxes magnifies the risk premium required to induce investors to make riskier investments. Because of taxes, therefore, inflation will choke off additional investment opportunities and investors will avoid riskier investments, including investments in research and development.

Principle 8: The structure of U.S. capital taxes encourages investment in short-lived assets and discourages investment in long-lived assets.

Taxes on capital do not merely discourage investment. They also alter the types of investments that are attractive. Even if the tax rates themselves do not directly discriminate against any particular type of investment, the interaction of taxes with risk and with inflation distorts investment decisions.

We have already seen how the presence of taxes causes investors to avoid riskier (but valuable and important) investments. Inflation creates even more perverse results. In general, income from capital is not indexed in the U.S. tax code. As a result, people with capital gains income pay taxes on inflationary gains, and investors in long-term projects are allowed to deduct only historical (depreciated) costs against inflationary revenues. To avoid the damaging effects of taxes and inflation, investors in the United States are encouraged to choose short-term over long-term assets and assets that can be depreciated quickly over those with a longer life. Evidence of this perverse effect can be seen in Table III. As the table shows, depreciation as a percent of the total cost of capital has been rising over the past 40 years:

• In the 1950s, depreciation constituted between one-fifth and one-fourth of the total cost of capital.

• Today, however, depreciation is almost one-third of the cost of capital.

The structure of U.S. taxes combined with inflationary monetary policies have discriminated against the manufacturing industries (where assets depreciate more slowly) and in favor of services and retail trade (where assets depreciate more quickly). This may be one of the most important reasons for the decline of the rust belt industries in the United States and the inability of many manufacturing concerns to compete successfully in international markets.

Principle 9: Because capital taxes lower the nation’s output, an increase in capital taxes almost always results in less revenue for government.

Because of their effects on investment and on the size of the capital stock, taxes on capital have severe effects on the economy as a whole. This is why governments that impose new taxes on capital almost always collect much less total revenue than they anticipated. Not only does the tax base (the capital stock) shrink, but so does aggregate output and national income. This means that government tax collections will be smaller for almost every other type of tax, including labor taxes. In general, almost any (targeted) reduction in taxes on capital will result in a net gain in revenue for government. For example:

• A 10 percent cut in the tax rate on all capital income will cost government about $10 billion a year in lost revenue.

• Because of the tax cut, however, the cost of capital will fall by about 5 percent and, in response, new investment will increase the capital stock by 5 percent.

• This means that the economy as a whole will receive an infusion of about $750 billion in new capital, leading to an increase in the nation’s output of goods and services of about $250 billion (net of depreciation).

• Of the $250 billion in increased national income, about $120 billion will go to government in taxes and about $120 billion will consist of an increase in aftertax wages.

Thus in return for giving up $10 billion in annual taxes, government revenues will rise $120 billion a year — a $110 billion profit — and wage earners will also realize a $120 billion increase in aftertax income.

As an example of this relationship, consider the capital gains tax. Historically, there has been a negative relationship between capital gains tax rates and capital gains revenue collected by the federal government. Whenever tax rates have been increased, tax revenues have dropped, and vice versa.

Perhaps because capital gains taxes have become a political issue, the major forecasting agencies of Congress have (1) denied the existence of this historical relationship, (2) gone to great lengths to explain it away or (3) maintained that the past relationship would not hold in the future. For example, both the CBO and the JCT resolutely maintained that the 40 percent increase in the capital gains tax rate in the 1986 Tax Reform Act would increase rather than reduce government revenue. This prediction has proved to be a considerable embarrassment to both agencies.

Before the tax hike took effect, there was a huge jump in capital gains income in 1986. For the three years following the tax increase, however, capital gains income was lower than it was in 1985. [See Figure VI.] The CBO prediction, by contrast, was way off:

• The CBO forecast that capital gains income would be $225 billion in 1989 — 50 percent higher than it actually was.

• Since the $75 billion error is repeated in each year of a five-year CBO forecast, the total mistake amounted to about $375 billion.

• As result, annual tax revenue from capital gains was about $20 billion less than the CBO projected.

Similarly, as late as 1990 the JCT staff was still predicting that capital gains realizations would continue to increase annually during the 1990s. Their forecast for 1990 of $237 billion turned out to be $87 billion too high.

Principle 10: By moving to a more efficient tax system, the United States could collect the same amount of government revenue with much less harm to the private sector. Every tax has a distorting effect on the private sector. For example, a tax on labor discourages people from working and, with less labor, capital is less productive. As we have seen, a tax on capital leads to less capital and, with less capital, labor is less productive. How we collect these taxes, however, makes a big difference. In principle, two additional distortions characterize the U.S. tax system.

Distortion Due to Marginal Tax Rates. One type of distortion received a great deal of attention in the 1980s: the difference between average and marginal tax rates. When all income is not treated equally in the tax code, marginal tax rates have to be higher, relative to the average tax burden. Yet marginal tax rates are the ones that determine people’s decisions about whether to supply capital and labor. When all capital (including residential houses) and all taxes (including sales and property taxes) are considered:

• In 1992 the average tax rate on labor income was 36.1 percent (including all taxes), and the marginal tax rate was 44.2 percent.

• The average tax rate on capital income was 44.6 percent, and the marginal tax rate was 55.7 percent.

• Because of these large discrepancies between average and marginal tax rates, the output of goods and services for the year was $411.1 billion lower. [See Table VI.]

Distortions Due to the Mix of Capital and Labor Taxes. During the 1986 tax reform debate, considerable attention was given to the desirability of removing distortions in the tax system by taxing all capital assets in the same way. Unfortunately, the Reagan administration and the Congress ignored an even more important distortion: the unequal treatment of capital and labor.

Business enterprises combine capital and labor in production based on their aftertax prices. In an ideal tax system, the ratio of the aftertax prices of capital and labor would equal their before-tax ratio. To the degree that these ratios are not the same, firms are encouraged to combine capital and labor in inefficient ways — leading to less total output. Unfortunately, while eliminating one type of distortion, tax reform led to a worse one.

• Whereas in 1986 the average tax on capital income was 20.2 percent greater than the average tax on labor income, by 1992 the distortion rose to 25.8 percent.

• The resulting efficiency loss cost the U.S. economy about $213.2 billion in reduced production. [See Table VI.]

Total Efficiency Loss. As Table V shows, the first round of Reagan tax cuts moved us toward a more efficient tax system. The efficiency loss from the way we impose taxes fell from 16.9 percent of output in the last year of the Carter administration to 12.9 percent in 1983 and reached 11.9 percent in 1987. Since then, a series of tax changes has reversed the direction:

• In 1990, the total efficiency loss caused by the way we impose taxes was 13.5 percent of output.

• In 1992, the total distortion cost Americans $642.7 billion — about $2,650 for every man, woman and child in the country.