Revised September 13, 1996


by Richard Kogan

On August 5, 1996, Mr. Dole released his tax plan in the form of a fact sheet and a major address. A second fact sheet released on August 27 provides some supplemental information. Because Mr. Dole's address focused primarily on questions of tax rates and economic growth, the accompanying proposals to reduce federal programs enough to offset the revenue losses from the tax cuts and balance the budget have received less attention. This analysis examines those budget-cutting proposals.


There are four significant points to be made about Mr. Dole's six-year budget reduction plan. Each of these points is discussed in this analysis.

Dole Plan is Built on the Congressional Budget Plan

Mr. Dole's budget plan starts with the budget plan Congress agreed to this June. His August 27 fact sheet states, "As Senate Majority Leader, Bob Dole voted for the plan and supports its provisions." Thus, the Dole plan includes six-year reductions (relative to current law) of $168 billion in Medicare, $72 billion in Medicaid, $19 billion in the Earned Income Tax Credit, $8 billion in federal employee retirement, and $5 billion in veterans benefits, as well as other smaller cuts and user fee increases. It should be noted that a number of these entitlement cuts are controversial; their enactment should not be considered a foregone conclusion under any administration.

The congressional budget plan also includes still deeper, but generally unspecified, reductions in non-defense discretionary (i.e., non-entitlement) programs. In addition, it includes the continued but slow "build-down" of the defense establishment that has been occurring since the end of the Cold War. This latter aspect of the congressional plan is inconsistent with recent congressional funding decisions that seem to commit the Defense Department to acquiring large and expensive advanced weapons systems that would cost more, rather than less, to operate in the future.

Dole Plan Calls for Additional Spending Cuts

In addition to the spending cuts called for in the Congressional budget plan, the Dole plan calls for further spending cuts totaling $208 billion over six years, plus another $9 billion in debt service savings. Based primarily on the first fact sheet, those cuts are as follows:

Table 1: Additional Budget Cuts in Dole Plan
1997-2002 total, in billions
Non-entitlement programs: (excluding defense)
Energy Dept.$32
Commerce Dept.15
"10% cut in administrative costs"90
1% across-the-board cut23
Entitlement programs: (excluding Social Security and Medicare)
1% across-the-board cut23
Asset Sales: auctioning the electromagnetic spectrum34
Debt service: lower interest payments on the debt9

Non-entitlement programs spread across many areas of the government and include: highway construction and repair; national parks and forests; environmental cleanup; education and job training; housing; the Census Bureau; the FBI, DEA, and INS; federal courts; veterans hospitals and doctors; medical, agricultural, high-tech, and basic scientific research; air traffic controllers; NASA; and the employees of the Social Security Administration and the Unemployment Compensation offices. Under the Dole plan, these programs, excluding defense, would be cut more deeply than under the congressional budget plan in four ways:



Table 2: Non-defense Discretionary Spending
(dollars in billions)
Based on spending levels for:
this Congresslast Congress
Base level (CBO August 1996 est.)270278
Base level in 2002, adjusting for inflation322340
Congressional budget plan, 2002245245
Dole plan, 2002203203
Dollar cut from base level adjusted for inflation-119-137
Percent cut from base level adjusted for inflation-37%-40%


Neither the Congressional plan nor the Dole plan provides much information on where these cuts would occur. (The President's budget, which could require non-defense discretionary cuts in 2002 almost as deep as those in the congressional plan — but less deep than those in the Dole plan — also contains ambiguities about where such cuts would occur.) Because of their magnitude, it is not plausible that the Dole spending cuts can be achieved simply by squeezing fat.

Entitlement programs would be cut an additional $23 billion over six years beyond the cuts called for in the congressional budget plan. The first Dole fact sheet is silent on where this $23 billion in cuts would occur, other than to say that Social Security and Medicare are off the table.5 If the additional, unspecified cuts are spread equally across the board, however, the effects would be as follows:

It may be controversial to deepen the cuts on programs assisting low-income people and broaden the reach of entitlement cuts to previously untouched programs such as unemployment compensation.

Asset sales, such as the proposal to auction more of the electromagnetic spectrum, have two drawbacks. The first is that the deficit reduction they accomplish is only temporary. It is risky to finance permanent tax cuts with temporary asset sales; that is a recipe for later, growing deficits. The second drawback is that, unlike other deficit reduction, asset sales do not increase net national saving and so do not have the beneficial effects on economic growth of real deficit reduction. Asset sales do not increase net national saving because the purchase of federal assets by the private sector is financed entirely out of funds already saved — the retained earnings of businesses or the amounts that businesses borrow from the credit markets. For these reasons, asset sales do not count as deficit reduction for purposes of the Budget Enforcement Act of 1990.

The Dole plan calls for $55 billion in asset sales over six years. Of that amount, $53 billion is from auctioning rights to the electromagnetic spectrum (i.e., broadcast and other radio band frequencies to be used by high-tech electronic information devices such as cellular phones and interactive television). The congressional budget plan assumed $19 billion in proceeds from such auctions; Mr. Dole assumes an additional $34 billion. Auctioning is certainly a reasonable way to allocate this limited resource and is likely to be superior to the pre-1993 approach of giving away broadcast rights to selected users. But it is much less certain that the private markets will choose to bid $53 billion for even the most expansive distribution of the electromagnetic spectrum.

The Plan in Total

Combining the existing congressional budget plan and the further budget reductions proposed by Mr. Dole yields the reductions in federal spending shown in Table 3 (see below). These reductions total $858 billion over six years and $268 billion in 2002, measured relative to CBO's "capped baseline," which is a projection of spending under existing law.6

In assessing the depth of the reductions, it may be best to think in percentage terms. Some have implied that Mr. Dole's proposed budget cuts are mild because they would reduce federal programs by only six percent. Such a figure creates mis-impressions.

When this analysis is done in a more meaningful fashion, examining the percentage reduction called for in 2002 by the Dole plan as a whole (including the reductions in the congressional budget plan), the result is the non-defense figure of 37 percent after adjusting for inflation, as noted above. The average reduction for all programs other than Social Security, Medicare, and defense would be more than 20 percent. (Less than one tenth of these proposed reductions have been achieved, largely through enactment of the welfare bill.)

The Dole Plan for Budget Reductions
Combines the Congressional budget plan and additional Dole reductions
(in billions of dollars)
1997199819992000200120026-yr Total
A.Non-entitlement programs (discretionary appropriations):
B.Entitlement Programs:
Social Security*0000000
Low-income (including Medicaid and EITC)
Enacted in "welfare" bill38910111355
Remaining cuts3613192737104
Other entitlements52445827
C.Asset sales:46810121555

* Same amount as in Congressional budget plan; all other reductions are deeper.
NOTES: 1) Figures may not add due to rounding. All cuts are measured from CBO's baseline projection of spending under current law, with minor baseline adjustments consistent with the congressional budget plan.
2) CBO's baseline projection for non-entitlement programs assumes that 1996 appropriations will be adjusted only for inflation except when the statutory "cap" on such programs is even more restrictive. This analysis assumes that, for non-entitlement programs, the defense and non-defense baseline projections each grow at the same rate.
3) Sen. Dole's cut of one percent in "other spending programs" is assumed to apply across the board except to Social Security, Medicare, and defense, as specified in his plan. His one percent cut therefore falls partly on entitlement programs, partly on non-entitlement programs.
4) The first Dole fact sheet specifies the year-by-year path of the $217 billion in additional spending cuts he calls for but does not specify the year-by-year path of individual components (such as the spectrum auction or the cuts in the Department of Energy). This analysis assumes that each component of the additional cuts follows the same year-by-year path as the total $217 billion reduction once the $9 billion in extra debt service savings are subtracted from the $217 billion path.
5) In addition to the program cuts shown here, any plan to reduce deficits below CBO's projected levels will also reduce projected interest payments.

Dole Plan Depends on Optimistic Revenue Estimates

The revenue estimates in Mr. Dole's budget plan are optimistic in two ways. First, the starting point — the estimate of revenues under current law — is $80 billion higher than the estimates CBO uses. The Dole plan seems to have treated an upward blip in 1996 revenues as though it were a permanent increase in the revenue base.7 Such an adjustment is inconsistent with the congressional leadership's contention of last winter that balanced budget plans are valid only if based entirely on CBO estimates.

Second, the Dole budget assumes that lower marginal income tax rates — both on earned income and investment income — will induce people: a) to work more hours; and b) to save more of their income. Higher personal saving means that more capital will be available for investment.

In effect, the tax cuts are intended to increase the supply of both labor and capital and thereby to increase economic growth. (This is why a cut in marginal tax rates is called a "supply side" tax cut.) The figures in Mr. Dole's plan show that he expects to recoup 33 percent of the revenues otherwise lost from supply-side tax cuts through increased economic growth; that is, he expects economic "feedback" of 33 percent.8

Most experts think this is a serious overestimate. For example, Charles Schultze of the Brookings Institution testified September 5 that a study he had undertaken of across-the-board cuts in personal and corporate income tax rates suggested a "relatively generous estimate" of the supply-side revenue feedback would be about half the 33 percent that the fact sheets assume.9

In summary, if the $80 billion upward adjustment to CBO's revenue baseline is not justified and if the economic feedback is half as much as anticipated (which is still a generous estimate of feedback), then deficits over the six-year period would be about $150 billion higher than Mr. Dole has forecast. This possibility could necessitate another $150 billion in budget cuts on top of those called for in the Dole plan.

Furthermore, to balance the budget using CBO estimates would mean that neither the $80 billion adjustment to CBO's revenue baseline nor the $147 billion in revenue feedback would be counted. As a result, $227 billion in additional spending cuts could be needed.

Table 4: Planned and Possible Budget Cuts
relative to CBO capped baseline; excluding debt service
1997-2002 Total
Called for in the Congressional Plan$650
Additional cuts called for by Sen. Dole+$208
Dole plan$858
Possible further cuts if revenue optimism is unwarranted+$227
Possible total$1,085

In short, the tax cuts proposed by Mr. Dole require increasing the six-year budget cuts in the congressional budget plan by at least $208 billion and by as much as $435 billion if the budget is to be balanced by 2002 using CBO estimates. Under those circumstances, the total budget cuts would reach $1.1 trillion and would be two-thirds larger than those in the congressional budget plan. (See Table 4.)

Effect of Deficits

Mr. Dole's plan rests on several optimistic economic assumptions: 1) that the "economic bonus" of $254 billion from lower interest rates and higher growth that CBO expects if a balanced budget plan is agreed to will fully materialize; 2) that the 1996 upward blip in revenues represents a permanent trend; and 3) that economic feedback will be noticeably larger than most economists expect.

Table 5: Economic Optimism in Dole Plan

Cause of economic improvementSource of estimate6-year amount
From balancing the budgetCBO$254 billion
From higher 1996 revenues1st fact sheet$ 80
From economic feedback of tax cuts1st fact sheet$147



In addition, and perhaps most significantly, the plan requires that all $858 billion in budget cuts be accomplished.

If some or all of these four assumptions do not fully materialize, there will be deficits rather than a balanced budget for two reasons. First, revenues will be lower and/or spending higher than planned. Second, the higher deficits will subtract from national saving. The result of lower national saving is a reduced supply of capital for investment; therefore, economic growth would likely be lower than projected.

This possibility raises an important point about growth strategies. Lower marginal tax rates might or might not induce individuals to save more (if lower income tax rates do have such an effect, it appears to be small). But if these tax cuts result in higher deficits, those deficits would certainly decrease national saving, thereby slowing growth. For this reason, cutting income tax rates is generally an ill-advised way to try to increase the rate of economic growth.

This point was made forcefully in an analysis issued in 1991 by the Republican staff of the House Budget Committee under the direction of the Committee's ranking Republican member at that time, Rep. Willis Gradison. "Because of the effect of government borrowing on investment," the analysis stated, "there is no tax incentive that promotes growth as efficiently as deficit reduction"10 (see box below; emphasis in original). Charles Schultze of the Brookings Institution recently calculated that "deficit reduction is some two-and-one-half times as effective as a tax cut in raising national output and income." 11 Similarly, CBO has written that "reducing the deficit is the single most reliable way to improve national saving. Over the long run, a permanently higher rate of saving would...raise the nation's standard of living." 12

For this reason, proposed tax cuts — including the President's proposals, as well as the larger cuts in the congressional budget plan and the still larger ones in Mr. Dole's plan — are generally unwise. They should be deferred until the long-run budget picture appears considerably brighter than CBO now projects.

Tax Incentives versus Deficit Reduction:

In 1991, the Republican staff of the House Budget Committee issued an analysis on tax incentives and deficit reduction.13 The following is an excerpt from the analysis' conclusion.

"Whether aimed at increasing efficiency or growth, many "growth enhancements" backfire. This is due to two factors. First, few incentives are very powerful. They simply do not result in huge increases in output. Second, they typically lose revenues, increasing government borrowing as a consequence, and reducing the accumulation of private capital because of the government's increased claim on savings.

"The generally weak power of incentives to influence growth is easily illustrated. Over the past decade, long-run output grew roughly 2.5 percent a year. About one percentage point of this was due to productivity growth (with the rest due to growth in the labor force). Most scholarly research indicates only a quarter to a half of productivity growth is due to capital accumulation (.25 to .5 percentage points per year). This means that capital formation would have to at least double and perhaps quadruple just to add .5 percentage points to the overall growth rate. The additional capital would have to come from saving. Total saving would not have to increase as much as capital formation (since three-quarters of saving is used to replace worn-out capital). But still, just to boost growth from 2.5 percent to 3 percent a year, total saving would have to increase 25 to 50 percent.

"Sadly, most evidence suggests that saving is unresponsive to any tax incentives designed to increase it. And capital gains tax cuts and IRAs only affect a small part of saving. Even the most optimistic estimates of the responsiveness of saving to taxes are too low to support the argument that such incentives significantly boost saving and growth.

"Worse, by losing revenue, many tax incentives would slow growth. The government's additional borrowing demands would use up saving that would be made available for private investment. Thus, many "growth enhancements" would actually do just the opposite of what they are intended to do: they would reduce growth through their effects on the deficit.

"The incentives could be implemented in combination with cuts in outlays or additional taxes elsewhere in the budget to make up the lost revenue. But, of course, the net result of the tax incentive and offsetting revenue increase on national saving would not be as great as the effect of the revenue increase or spending cut alone.

"In short, there is no tax-based growth enhancement that will increase growth as much as deficit reduction. If growth is to be increased, revenue increases and spending cuts will serve best if used to reduce the deficit instead of to pay for tax incentives."

13 "Tax Incentives, Growth, and the Deficit Reduction," op. cit.


1. Judy Shelton, writing in an op-ed in the Wall Street Journal of August 27, 1996. According to the WSJ, "Ms. Shelton is an economic policy advisor to the Dole-Kemp campaign."

2. When the deficit is reduced below projected levels, the debt will also be lower than projected and federal payments for interest on the debt will be correspondingly lower. The vast bulk of such debt service savings are already incorporated into the congressional budget plan, which (if implemented) would result in a balanced budget in 2002 if CBO's current estimating assumptions are correct. In general, the additional spending cuts proposed by Mr. Dole only offset his additional tax cuts, and so do not further reduce the deficit. Because the Dole tax cuts occur more gradually than the tax cuts in the congressional budget plan, however, the Dole budget plan aims for lower deficits than the congressional plan in the first two years and slightly higher ones thereafter. Based on CBO methodology, the Dole deficit path would reduce federal interest savings an additional $9 billion over six years, relative to the congressional path. The existence of extra debt service savings is mentioned in the second Dole fact sheet.

3. The total level of outlays over 1997-2002 in the Dole plan for non-defense programs before this $37 billion cut (including non-entitlement programs and entitlements other than Social Security and Medicare) is $3.8 trillion. Therefore, a $37 billion cut is almost exactly one percent, just as the first fact sheet says. If the $37 billion were meant to apply only to non-defense discretionary programs and a one percent reduction in entitlement programs were not involved, the reduction in non-defense discretionary programs would be almost three percent. Since this cut is described as "1%," the most reasonable conclusion is that it applies both to discretionary and entitlement programs.

4. This analysis is consistent with the August 5 fact sheet, except for the attribution of $9 billion in savings to reduced debt service costs. That fact sheet specifies each of the four elements listed: $90 billion in non-defense, non-entitlement "administrative" reductions, $32 billion from the Energy Department, $15 billion from the Commerce Department, and $46 billion (of which $9 billion is debt service) from "other" programs except defense, Social Security, and Medicare. The second fact sheet does not fully match the first. Like the first, it states that the Dole plan includes $217 billion more in six-year savings than the congressional budget plan. But unlike the first fact sheet, the August 27 fact sheet lists only three items and they total only $199 billion. Moreover, the second fact sheet does not include a dollar estimate for "administrative" savings or cuts in the Energy and Commerce Departments, does not mention the "one-percent reduction" in other programs, and implies that some unspecified portion of the amounts listed as spending reductions in the first fact sheet may in fact be revenue increases (e.g. through closing "corporate loopholes" and through "tax amnesty"). Because the first fact sheet is more specific and provides six-year estimates for each element, while the numbers in the second fact sheet do not add up, this analysis of necessity relies on the first (i.e., the August 5) fact sheet.

5. The second fact sheet suggests two possible areas for such reform. But one of those proposals — an extension of expiring patent and trademark fees — is included in the congressional budget plan; the fact sheet appears to double-count those savings, since it lists them as additions to the congressional plan.

6. CBO's "capped baseline" projects the cost of entitlement programs based on existing law and CBO's economic, demographic, and related assumptions. It projects non-entitlement programs by assuming that the statutory cap on such spending will constrain those programs to grow by less than the full amount needed to adjust current funding levels for inflation. As a result, CBO's "capped baseline" assumes discretionary spending levels in 2002 that are $13 billion lower than current levels adjusted for inflation. In this analysis, therefore, the Dole plan for non-entitlement, non-defense spending in 2002 is $119 billion (or 37 percent) lower than 1996 levels adjusted for inflation, but only $106 billion (or 34 percent) below CBO's "capped" baseline.

7. The second Dole fact sheet defends an upward adjustment to CBO's revenue baseline on the grounds that tax receipts this past April were higher than either CBO or the Office of Management and Budget expected. The Wall Street Journal has suggested a reason for the extra spurt of revenues: "A surge in revenues from 1995 tax returns may also reflect a curious one-time windfall. Experts speculate that some investors who had been holding off taking stock-market gains, thinking the new GOP Congress might cut the capital-gains tax rate, gave up on this late last year and went ahead and took their gains — producing a tax bill that had to be paid in April." (Jackie Calmes, "Scary Deficit Forecasts for Clinton Years Fade as Tax Revenue Grows," Wall Street Journal, August 1, 1996.) If this explanation is correct, the spurt in revenues should not be extrapolated into the future.

Suppose, though, that Mr. Dole is correct in increasing CBO's revenue estimates by $80 billion. If this is because of a permanently higher level of capital gains realizations, the cut in capital gains tax rates proposed by Mr. Dole will lose more revenue than he assumes. In short, the Dole assumptions of revenue collections under current law and the cost of a capital gains cut are inconsistent with each other. They cannot both be true.

8. The figure of 33 percent for economic feedback is derived as follows. The plan calls for supply-side tax cuts of $446 billion over six years and assumes $147 billion in extra revenues from economic feedback: 147/446 = 33 percent. (The plan calls for total tax cuts of $548 billion over six years. However, only $446 billion can reasonably be classified as supply side. The plan calls for $75 billion in revenue losses from the child tax credit; because that credit does not reduce marginal tax rates, it does nothing to change incentives to work or save. The plan also calls for $27 billion from reducing income tax rates on Social Security benefits for the wealthiest 13 percent of retirees. Since these people are already retired, this tax cut cannot be expected to have any significant effect on the hours they work. This $27 billion cut is not a supply side tax cut.)

9. Charles L. Schultze, The Effect of a 15 Percent Tax Cut on the American Economy, September 5, 1996.

10. "Tax Incentives, Growth, and Deficit Reduction," House Committee on the Budget: Republican Staff Report, November 6, 1991.

11. Charles L. Schultze, The Effect of a 15 Percent Tax Cut on the American Economy, September 5, 1996.

12. Congressional Budget Office, Reducing the Deficit: Spending and Revenue Options, February 1995.

Analysis of Dole's child tax credit proposal.