Historical Data. Table 1 shows gross investment and depreciation from 1975 to 1997 for two definitions of capital:

• Direct physical capital includes investment in physical assets that are owned by the Federal Government. This includes spending for construction and rehabilitation (e.g., office buildings, dams, veterans hospitals, and military bases) and purchase of major equipment (e.g., defense weapons systems and air traffic control).

• Nondefense direct physical capital excludes the defense items from the first definition.

The magnitude of net investment in direct physical capital has been much smaller. Including defense, net investment was less than 2% of outlays until the 1980s defense buildup, rising to almost 4% during the buildup and more recently becoming negative as the depreciation on the buildup has exceeded new defense investment. Excluding defense, net investment in direct physical capital has fluctuated in a more narrow range, between 0.4% and 0.9% of outlays.

Budget Impact. If shifting to a capital budget were limited to redefining the calculation of the deficit, then the budget impact of the shift appears directly in Table 1.  Outlays and the deficit in each year would change by the amounts shown as net investment in the table. As noted in the preceding paragraph, the effect of this redefinition on outlays would be relatively small.

However, shifting to a capital budget could involve more than simply redefining the deficit. Many advocates of a capital budget framework argue that a capital budget would allow taxes to be set so that investment would be financed over time by those who enjoy the benefits of the investment, rather than by those who paid taxes in the year the investment was made. This would be accomplished by setting taxes to cover operating budget outlays, defined to exclude spending on new investment and to include depreciation on prior investment.

Compared to financing investments with current taxes, setting taxes to cover depreciation would allow a tax cut if net investment were positive, on the rationale that the addition to the capital stock should be financed by future taxpayers, not current ones. On the other hand, this basis of setting taxes would require a tax increase if net investment were negative, in order to charge current taxpayers for drawing down the capital stock.

Table 2 shows the hypothetical impact if the budget had shifted to setting taxes high enough to cover depreciation rather than investment outlays, starting in 1975. The analysis assumes that the unified budget financed all investment with taxes over this period (i.e. that all borrowing was to support current spending, rather than investment).⁽¹⁾ Thus, the impact of the shift would have been to change receipts in each year by the amount of net investment removed from the budget, plus the debt service from prior changes in receipts. The table shows that if capital is defined as total direct physical investment, receipts would have had to rise in 1975 and 1976 to pay for the negative net investment in those two years. Later, as the defense buildup rose, tax cuts would have been allowed. Eventually, the interest on those tax cuts would have grown about as large as net investment, however. By 1997, the required $33.5 billion increase in receipts would have been $17.7 billion to pay for that year's negative net investment and $15.8 billion to pay for the higher interest from previous tax cuts.

If the shift in policy was for capital defined as nondefense direct physical investment, the change in receipts would have been far smaller. Net investment would have been positive in all years, which would allow small tax cuts. However, the interest on these tax cuts would have grown steadily over time. In 1997, it would have exceeded the amount of net investment, requiring a small tax increase.

The results in Table 2 arbitrarily assume that a capital budget was implemented in 1975. The choice of starting point would affect interest, since interest depends on the cumulative changes in receipts since the start of the capital budget. For example, if the capital budget for nondefense direct physical investment had started in 1980, interest would not have exceeded net investment in 1997, allowing a continued tax cut for that year.

The accompanying chart shows the changes in receipts for this policy under the two definitions of capital, as a percent of actual unified budget receipts. During the peak of the defense buildup, setting taxes to cover the depreciation on direct physical capital would have allowed a tax cut peaking at a little more than 3% of receipts. In recent years, the capital budget would have required a tax increase of around 2%, since defense depreciation now significantly exceeds investment. If the definition of capital had excluded defense, it would have allowed a tax cut of a little more than 1% of receipts in the late seventies, and would have had a negligible impact since then.

Footnote:
1.  The extent to which historical borrowing financed investment versus non-investment spending is impossible to determine. An argument might be made that the large deficits in the 1980s were allowed in part because of the defense buildup. To the extent that one assumed that part of historical investment was financed by borrowing, the changes in receipts shown in Table 2 are overstated.

Attachments: