[An updated version of this article can be found at Investment in the 2nd edition.]
In the United States, investment accounts for about one-sixth of gross national product. It was 16.6 percent of GNP in 1990. Yet investment has occupied a much more important role in policy discussions than this share of production might suggest. The two main reasons for this are that investment is volatile and, therefore, a cause of business fluctuations and that investment contributes to economic growth.
Concern with these issues of business cycles and growth has led to very active tax policy toward investment during the postwar years, as a succession of governments has tried to influence the level, pattern, and timing of investment spending. The evidence suggests that such policies have been effective, but that many other, uncontrollable factors continue to influence investment. In the short term, investment decisions still appear to be strongly driven by the "animal spirits" to which Keynes attributed investment fluctuations. Over the longer term, the nature of investment has been affected by changes in the demographic makeup of society and in the composition of industrial production.
What Is Investment?
Although in general parlance investment may connote many types of economic activity, economists normally use the term to describe the purchase of durable goods by households, businesses, and governments. Private (nongovernmental) investment is commonly divided into three broad categories: residential investment, which accounts for about a quarter of all private investment (25.7 percent in 1990); nonresidential, or business, fixed investment, which accounts for most of the remainder; and inventory investment, which is small but volatile. Indeed, inventory investment is often negative (it was in 1990, and in three years during the eighties). Business fixed investment, in turn, is composed of equipment and nonresidential structures. Equipment now makes up over three-quarters of business investment.
Because of the decline in manufacturing and agriculture and the rise in services in the United States, the composition of private investment has changed considerably during the postwar period. The biggest single change has been the increased investment in computers and information-processing equipment. In 1953, spending on computers and related equipment was only 1 percent of nonresidential investment spending. By 1989 this figure had grown to 25 percent.
Governments invest, too—in schools, roads, and other components of economic infrastructure. But one important development in recent years has been the decline of government investment. Between 1982 and 1989, the net (of depreciation) private nonresidential capital stock rose by 20 percent, adjusted for inflation. Residential capital rose by 21 percent over the same period. In contrast, the nonmilitary government capital stock fell by 3 percent. Because, as past evidence has shown, many government capital assets (such as roads and waterways) facilitate private business activities, this decline suggests that private productivity will suffer as a result.
Investment and Business Fluctuations
One reason for so much interest in investment behavior is its apparent role in causing or exacerbating business cycles. Investment is a volatile component of GNP, falling sharply during recessions and rising just as sharply during booms. As the economy went into a deep recession in the early eighties, for example, real GNP fell 3 percent between 1981 and 1982, but investment fell in real terms by 18 percent. In the following year, as the expansion began, GNP rose 4 percent while investment rose 13 percent.
Why is investment so volatile? The key lies in the nature of the investment process. Investment decisions often require long lead times, and their consequences are as durable as the investment goods themselves. Consider, for example, the case of commercial construction, which declined in the late eighties. Office buildings planned during a period of strong demand for space may be completed during a recession, when demand even for existing space is weak. Such a shift in fortunes causes a decline in investment for two reasons. First, the need for office space has declined. Second, the amount of office space has risen, so that subsequent investment must fall not only to keep pace with slower demand, but also to eliminate the "overhang" of empty space.
Economists call this magnification of the impact of declines in product demand the "accelerator" model of investment. As industrial production shifts away from such strongly cyclical industries as manufacturing, the strength of cycles in business fixed investment may weaken. This moderating process is also likely to be helped by the shift toward less durable investments requiring shorter planning and construction periods, such as computers.
Investment and the Cost of Capital
While fluctuations in output exert a strong influence on investment behavior, the costs of investing matter, too. These costs include the prices of capital goods themselves, as well as interest rates, required returns to equity owners, and the taxes that firms must pay on the profits that the investments generate.
A convenient summary of the effects of these different cost components is the "user cost of capital," a term introduced in the sixties by Harvard economist Dale Jorgenson. The user cost of capital shows how each of these factors influences investment, and has proved particularly useful in evaluating a variety of tax changes that have been introduced during the past few decades.
Activist tax policy toward investment began in 1954 with the introduction of accelerated depreciation. By permitting investors earlier deductions for depreciation, the 1954 changes increased the present value of these tax deductions and lowered the user cost of capital. An even more powerful incentive, the investment tax credit (ITC), was introduced in 1962. Although its provisions were frequently changed, the credit was in force for most of the period between 1962 and 1986, when it was repealed in the Tax Reform Act of 1986.
By permitting a 10 percent credit for qualifying investments (primarily in machinery and equipment), the ITC lowered the effective cost of investing, the user cost, by roughly the same percentage. The effects of this cost reduction are evident in the increasing share of equipment investment after the credit's introduction. In 1961, the year before the ITC was introduced, expenditures on equipment accounted for 51 percent of business fixed investment. This share rose to 52 percent in 1962, 54 percent in 1963, 55 percent in 1964 and 1965, and 57 percent by 1966. Although the repeal of the credit in 1986 did not cripple investment in machinery and equipment, some studies have found that the investment would have been even higher—by perhaps as much as 1 percent of GNP—had the credit not been repealed.
Other Determinants of Investment
Investment is influenced by demand conditions, the effects of which (including profitability) can be represented by the accelerator effect, and cost conditions, as summarized by the user cost of capital. Researchers have found that another independent determinant of investment behavior is liquidity—the liquid assets a company has on hand plus the cash flow it is currently generating. While the user cost of capital varies with the cost of funds in credit markets or to firms issuing equity, many firms are limited in their access to these markets. Because they must rely primarily on internal funds to finance investment, liquidity matters. The liquidity constraint seems particularly to affect smaller corporations and (along with accelerator effects) helps explain investment volatility, because cash flow itself (after-tax profits plus book depreciation) is very cyclical.
Liquidity also plays an important role in residential investment. About two-thirds of all residential investment takes the form of owner-occupied housing. The reliance of home buyers on the mortgage market has wrought havoc on residential construction during periods of tight credit that typically have accompanied recessions. As a result residential investment has often been even more cyclical than other forms of fixed investment. However, the reliance on borrowed money has, in certain periods, actually encouraged investment in owner-occupied housing, through the interaction of inflation and the tax system.
Inflation encourages housing investment in two ways. First, mortgage interest payments are tax deductible. As interest rates rise with inflation, so do tax deductions, even if the real interest rate, defined as the interest rate less the inflation rate, does not. For example, if the interest rate is 8 percent and the inflation rate 4 percent, the real interest rate is 4 percent, and an investor in the 28 percent tax bracket gains a tax reduction of 2.24 cents (28 percent of 8 percent) per dollar of debt. If the real interest rate remains at 4 percent, but the inflation rate rises to 8 percent, the nominal interest rate rises to 12 percent and the value of tax deductions rises to 3.36 cents (28 percent of 12 percent) per dollar of debt.
Second, the increased value of a home that accompanies inflation is essentially untaxed, because of provisions that allow a tax-free rollover if another house is purchased and a one-time exclusion of gain (currently $125,000) after age 55. [Editor's note: this provision of the tax code changed dramatically in 1997.] This favors investment in owner-occupied housing over other types of investment with taxable gains. Another disadvantage of other types of fixed investment is that the depreciation allowances that investors receive are based on original asset cost, which may fall well short of true replacement cost in the presence of inflation. (This is not a problem faced by owner-occupiers, who do not receive depreciation allowances.)
The evidence in favor of the hypothesis of inflation-induced investment in owner-occupied housing comes primarily from the late seventies, when the hypothesis arose. During the economic expansion from 1976 to 1979, when inflation averaged 7.3 percent (measured using the GNP deflator, based on the prices of all goods and services included in GNP), investment in owner-occupied housing accounted for 33.5 percent of fixed investment. During the comparable expansion period of 1983 to 1986 with inflation averaging just 3.3 percent, residential investment's share fell to 29.1 percent.
An alternative explanation, for which there is evidence, is that the increase in housing investment in the late seventies was caused by the increase in family formation during the period—the coming of age of the baby boomers. If this explanation is correct, then housing prices and demand are likely to decline into the next century, well past the housing slump of the 1990-91 recession.
Why Is U.S. Investment So Low?
Investment helps increase productivity by raising the level of capital per worker and, perhaps, hastening the adoption of new technologies. As a share of GNP, gross private investment has been relatively stable over the past few decades. This stability, however, masks a disturbing trend. Because additions to the productive capital stock equal gross investment less depreciation, the shift toward equipment and, particularly, short-lived and rapidly depreciating equipment, means that the ratio of net investment to GNP has fallen over time. While gross investment's share of GNP went from 15.9 percent to 15.3 percent between 1969 and 1988 (both relatively strong investment years) net investment's share fell from 7.4 percent to 4.8 percent.
By the early nineties the dollar value of investment in the United States was less than in Japan, a country with roughly half the population. The main explanation for this difference is that U.S. investors faced a higher cost of capital than their Japanese counterparts. Comparative studies of the United States and Japan suggest that a lower Japanese cost of funds (as opposed to differences in the tax treatment of corporations, for example) is the major source of a cost-of-capital gap that appears to exist, or at least to have existed in recent decades, between the two countries. The cost of funds is higher in the United States because the low saving rate in the United States makes the supply of funds low and drives up interest rates.
The liberalization of international capital markets can offset the effects of low saving rates in particular countries. Investors living in countries with high saving rates can invest in countries, like the United States, that have low saving rates. This has happened and is the main reason for the large inflows of capital to the United States in recent years. Yet there remains a correlation across countries between levels of national saving and investment.
Alan J. Auerbach is the Robert D. Burch Professor of Economics and Law at the University of California, Berkeley and a research associate at the National Bureau of Economic Research in Cambridge, Massachusetts.
Ando, Albert, and Alan J. Auerbach. "The Cost of Capital in the United States and Japan: A Comparison." Journal of Japanese and International Economies 1 (June 1988): 134-58.
Auerbach, Alan J. "The Tax Reform Act of 1986 and the Cost of Capital." Journal of Economic Perspectives 1, no. 1 (Summer 1987): 73-86.
Feldstein, Martin, ed. The Effects of Taxation on Capital Accumulation. 1987.
Mankiw, N. Gregory, and David N. Weil. "The Baby Boom, the Baby Bust and the Housing Market." Regional Science and Urban Economics 19 (May 1989): 235-58.
Munnell, Alicia H., ed. Is There a Shortfall in Public Capital Investment? 1990.