of the chambers. The crucial point is the President's position when
a bill is in conference. At that stage, the administration hangs
tough on some items and acquiesces on others. As the report later
states, "the SAP-based estimates might have overstated the poten-
tial savings from a presidential line item veto. For example, a
President might have chosen not to exercise the veto on all items
to which objections were raised in the SAPs" (p. 8).
2. Theoretical vs. Realistic Savings. The report estimates savings
that "might have occurred" or spending that "could have been re-
duced" (p. 1). This choice of "might" and "could" tilts the analysis
toward the maximum highest number. Available data clearly indi-
cates that a $70 billion saving over a six-year period is unrealistic.
The report acknowledges that other administration documents re-
veal that an analysis based on SAPs "may overstate the savings
that would have occurred" (p. 2). There is a substantial difference
in moving from might/could (theoretically possible) to would (likely
to occur).
The report notes that an 0MB report in 1988 "indicated that the
President would have vetoed much smaller amounts than those the
SAPs identified as objectionable for that year" (p. 2). The 0MB re-
port is a valuable guide to what Presidents are likely to do with
item-veto authority. The SAP-based estimate of line item veto sav-
ings for 1988 is $12.6 billion in budget authority. The 0MB report
identified only $540 million in potential savings from item vetoes
(p. 9). The GAO study admits that the SAP-based estimates "may
overstate" the potential savings from a line item veto (p. 9).
308
To be precise, SAP-based estimates overstate savings by a factor
of 23 for 1988. If that ratio is applied to the six-year period, likely
savings drop from $70.6 billion to $3.03 billion.
Curiously, the report "judged that the SAPs are a reasonable in-
dicator of the maximum savings that might have been achieved if
a President had used line item veto authority in the period we
studied" (p. 9). From its own analysis, SAPs appear to be an unrea-
sonable indicator, unless they are used solely for the purpose of es-
timating "maximum" savings rather than likely savings. Also on
page 9, the report states that "it is impossible to determine conclu-
sively whether or not the SAP-based estimates developed for this
report accurately reflect the way a President who had actually had
line item veto authority in the period 1984 through 1989 would
have used that authority." If the analysis is that difficult to prove
conclusively, why release a report that gives readers the impression
that $70 billion could have been saved over a six-year period?
3. Double-counting (rescissions). Even a figure of $3 billion over
the six-year period probably overstates what might have been
saved through an item veto. The report does not deduct from its
$70 billion estimate the savings that result from the President's
current authority to rescind appropriations. For the years in ques-
tion. President Reagan asked Congress to rescind $18.6 billion from
fiscal years 1984 through 1989. Congress rescinded $0.4 billion.
However, over that same period of time, Congress initiated and en-
acted 144 rescission actions totaling $24 billion. It can be assumed
that some of the items rescinded appeared earlier in SAPs. The re-
port therefore credits the item veto for some savings that were
achieved by existing rescission procedures.
The potential of rescission authority for deleting appropriations
items is borne out by the first three years of the Reagan adminis-
tration. From fiscal 1981 through fiscal 1983, President Reagan
proposed $24.8 billion in rescissions and Congress approved $16.1
billion. In addition to rescissions proposed by the President, Con-
gress has initiated and enacted a total of $36.2 billion in rescis-
sions since the Budget Act of 1974.
4. Double-counting (Program Terminations). The report estimates
that 71 federal programs would have been terminated with an item
veto, including the Economic Development Administration, Legal
Services Corporation, and Amtrak. Those programs were "repeat-
edly proposed" for termination in SAPs during that period (page 8).
To the extent that programs were recommended for termination in
more than one of the six years of SAPs, did the report rely on dou-
ble-counting?
If the President had item-vetoed Amtrak in fiscal 1984 and Con-
gress failed to override, it might be proper to credit the President
with $716.4 million in savings for that year. But is it proper to
credit the President with savings for the next five years (fiscal
1985 through fiscal 1989)? Suppose the President recommended no
funds for Amtrak in his fiscal 1985 budget, Congress inserted the
money against his wishes, the President item vetoed that amount
and Congress failed to override. Again the President is credited
with savings for fiscal 1985. Will that scenario be repeated for the
next four years? It is reasonable to assume that Congress will al-
ways reintroduce funds for programs that had been terminated?
309
That assumption seems unreasonable. Operating under that as-
sumption, a President receives credit for a savings in one year, no
matter how long ago, and receives perpetual credit thereafter. Ac-
cording to that scenario, a President could terminate a program in
1812 and receive credit every year after that.
It is not even clear that the President should be credited with
$716.4 million in savings for the first year. In terminating an agen-
cy like Amtrak, are there no termination costs for outstanding con-
tracts and severance pay for agency personnel? Can those costs be
absorbed by the previous appropriation or will supplemental appro-
priations be needed for the phase-out? In case of an agency like the
Economic Development Administration, if it is legally impossible to
fire all of the employees, will other agencies be required to absorb
these people? Because of these considerations, net savings will be
less than the report indicates.
5. Assuming that "Savings" are Permanent. The report assumes
that each presidential saving, obtained through the item veto, is
permanent and will remain untouched by other governmental pres-
sures. That assumption is contradicted by the experience of the
budget process. Under Section 302(b) of the Budget Act of 1974,
Congress allocates ceilings to the appropriations subcommittees. It
is well-known that if the subcommittees report a bill substantially
under the allocation, it invites amendments on the floor that bring
the aggregate back toward the ceiling. Thus, a "savings" by the
subcommittee is quite temporary and is unlikely to last.
Why assume that "savings" from a presidential item veto will be
any more permanent? It is more likely that a successful item veto
(say of Amtrak in the above example) will unleash spending pro-
posals by the executive and legislative branches. The savings might
be transitory, quickly neutralized by a spending initiative in a
forthcoming supplemental appropriations bill.
6. Presidential Spending Initiatives. The figure of $3 billion also
overstates savings because the study assumes that Presidents are
interested only in reduced federal expenditures. Yet Presidents
have their own programs and activities that they advocate, and the
availability of an item veto could be an important weapon in coerc-
ing legislators to support White House spending priorities. Armed
with an item veto, a President could tell legislators that a project
or program in their district or state will be item-vetoed unless they
support the President's spending goals. If the legislators and the
President reach an amicable agreement, legislative add-ons would
be preserved along with presidential add-ons. Since these
interbranch conversations would likely remain confidential, the
public would never know that the item veto can increase federal
spending. A balanced assessment of the item veto must take into
account this dynamic in executive-legislative relations.
7. Studies at the State Level. Appendix III of the report summa-
rizes the studies at the state level that estimate spending reduc-
tions from an item veto. The report states that this literature "ex-
hibits no apparent consensus" on the budgetary impact of an item
veto, and yet the consensus in Appendix III seems clearly that the
item veto yields no fiscal restraint. Of the eight studies summa-
rized, seven conclude that the item veto is not a tool for fiscal re-
straint. Instead, it is used primarily to advance partisan interests
310
and executive spending programs. The only study that is optimistic
about potential savings from an item veto was coauthored by
James C. Miller III, who served as 0MB Director in the Reagan
administration. These studies should have cautioned against an-
nouncing a $70 billion federal saving over a six-year period.
U.S. Senate,
Committee on Appropriations,
Washington, DC, March 17, 1992.
Mr. Joseph Ross,
Director, Congressional Research Service, the Library of Congress,
Washington, DC.
Dear Mr. Ross: This is to request that the Congressional Re-
search Service provide an evaluation of a recent General Account-
ing Office report entitled "Line Item Veto — Estimating Potential
Savings". I have enclosed a copy of this report and a subsequent
letter that I sent to the Greneral Accounting Office expressing my
concerns about the report, to which I have not yet received a reply.
If you have any questions regarding this request, please do not
hesitate to contact me or Jim English, Staff Director of the Appro-
priations Committee, at 224-7200.
With kind regards, I am
Sincerely,
Robert C. Byrd,
Chairman.
STATEMENT ON CBO'S LETTER RESPONDING TO SENATOR
BYRD'S CONCERNS ABOUT THE CBO STUDY ON REDUCED
DEFENSE SPENDING
Mr. BYRD. Mr. President, last February, the Congressional
Budget Office released a study, entitled "The Economic Effects of
Reduced Defense Spending," which omitted several important
points. I raised these points with the CBO Director, Dr. Robert D.
Reischauer, in a letter on March 9. On March 17, Dr. Reischauer
responded to my concerns promptly and forthrightly, for which I
commend him.
The study estimated the economic impact of two hypothetical de-
fense spending reductions. It concluded that real GNP would rise
permanently by the end of the next decade by about $50 billion a
year, in 1992 dollars, if defense spending were cut 20 percent by
fiscal 1997. However, in the short run, it estimated the loss of
600,000 defense related jobs and described worst case scenarios for
three selected communities heavily dependent upon defense indus-
try.
My letter of March 9 listed several concerns. First, the study ig-
nored the expressed intent of the Budget Enforcement Act of 1990
by assuming future defense spending reductions will be used for
deficit reduction. The act allows defense spending reductions in fis-
cal year 1994 and fiscal year 1995 to be used for domestic discre-
tionary spending, as long as the overall spending caps are met.
Second, the study lumps together defense spending reductions
enacted in fiscal years 1991 and 1992 with the reductions under
consideration now for fiscal years 1993 through 1997. This gives
311
the appearance of larger economic impact than would result from
the reductions in fiscal years 1993 through 1997 alone.
Third, the study ignores already enacted programs which will
ease the economic impact of defense spending reductions. As noted
in a February 6, 1992, Congressional Research Service Issue Brief,
"Defense Budget Cuts and the Economy," economic adjustment as-
sistance programs already in existence under present law include:
over half a billion dollars each year set aside specifically to help
military and defense workers through the Economic Dislocation
and Worker Adjustment Assistance [EDWAA] Program; job train-
ing under title III of the Job Training Partnership Act; unemploy-
ment insurance; and support for impacted communities under title
DC of the Public Works and Economic Development Act of 1965, in-
cluding $50 million appropriated under the Defense Authorization
and Appropriations Acts of 1991.
Fourth, the study could better explain that most defense workers
threatened with job loss will switch to civilian production, retrain,
or retire without entering the ranks of the long-term unemployed.
Fifth, and finally, the study takes a worst case look at defense
spending reductions without considering a best case.
In his response to my concerns. Dr. Reischauer agreed that, even
though the CBO study assumed defense reductions would be used
for deficit reduction, defense spending reductions may be used for
domestic discretionary spending in fiscal year 1994 and fiscal year
1995. In fact, he observed that the defense savings contemplated in
the CBO study "would be required simply to avoid real reductions
in nondefense discretionary spending." He added, "In the long run,
increased spending on carefully chosen public investment projects
would work to increase the potential growth of the economy in just
the same way as a reduction in the Federal deficit."" * * * on av-
erage, public investments in the past do seem to have been as
worthwhile as private investments. * * * "
Dr. Reischauer also said that CBO "should have acknowledged
existing Federal programs aimed at mitigating the effects of de-
fense cutbacks and provided more discussion of other actions that
could be taken to mitigate the effects of defense spending cut-
backs." He reiterated the study's finding that "growth in
nondefense jobs would eventually offset the adverse effects of de-
fense cutbacks." Finally, Dr. Reischauer noted that "the study
clearly acknowledged that the calculations reflected a worst-case
assessment. * * *"
I thank Dr. Reischauer for his timely response. His letter casts
the CBO study in a more balanced light, and I commend it to my
colleagues for their consideration.
I ask unanimous consent that this correspondence be entered
into the Record, so that my colleagues and other interested read-
ers might be better informed about this study.
There being no objection, the letters were ordered to be printed
in the Record, as follows:
312
Congressional Budget Office,
Washington, DC, March 17, 1992.
Hon. Robert C. Byrd,
Chairman, Committee on Appropriations,
U.S. Senate, Washington, DC.
Dear Mr. Chairman: Your recent letter noted that several topics
of interest and concern to you were omitted from our February
1992 study entitled, "The Economic Effects of Reduced Defense
Spending." Overall, I believe the study represented a balanced re-
sponse to the Minority Leader's request. But, as you suggest, sev-
eral aspects of the analysis could have been explained more fully.
The study focused on the economic effects associated with cutting
defense spending and using the savings to reduce the federal defi-
cit. The peace dividend could, of course, be put to other uses. As
you note, under the provisions of the current Budget Enforcement
Act [BE A], defense cuts in 1994 and 1995 can be devoted to aug-
menting nondefense discretionary spending, including spending on
public investment, so long as overall limits on discretionary spend-
ing are met. Our study discussed the effects of devoting the peace
dividend to public investment in general terms, but did not analyze
those effects in detail.
We chose this focus because the size of the defense options ana-
lyzed in our study seemed consistent with the overall spending lim-
its in the BEA. The BEA requires rather substantial reductions in
total federal discretionary spending, particularly in 1994 and 1995.
Compared with 1992 levels, the real cuts in defense spending dis-
cussed in our study are no larger in 1994 and 1995 than the overall
cuts in discretionary spending mandated in the BEA. Thus, the de-
fense savings analyzed in our study would be required simply to
avoid real reductions in nondefense discretionary spending. This
reasoning was not adequately explained in the study, however, and
therefore your criticism is well taken.
Leaving aside issues of compliance with the BEA limits, how
would devoting the peace dividend to public investments affect the
economy? In the long run, increased spending on carefully-chosen
public investment projects would work to increase the potential
growth of the economy in just the same way as a reduction in the
federal deficit, as we stated in our report (see page 6). In the short
run, devoting defense spending cuts to public investment might
avoid the temporary GNP loss that is likely to occur if the deficit
is cut. Whether this favorable short-run outcome could be achieved
depends on how quickly governments could arrange to spend addi-
tional funds on investment projects, as those funds are withdrawn
from the defense sector.
The favorable long-run effects of investment spending also de-
pend on how carefully projects are chosen. Additions to the already
extensive infrastructure of roads, rivers, and airports, for example,
are not likely to have such a favorable payoff as those undertaken
in the past, and some may not easily pass a careful cost-benefit
analysis. And some investments, such as additional federal spend-
ing on education, may prove worthwhile in the long run but take
a long time to yield benefits. But on average, public investments
in the past do seem to have been as worthwhile as private invest-
313
ments, and with sufficient care, could continue to contribute to pro-
ductivity growth.
You also expressed concern that the estimates in our study in-
cluded job losses associated with cuts enacted in 1990 and 1991,
rather than focusing on the losses associated with the cut that may
be enacted for fiscal 1993. At the time the detailed computer sim-
ulations used in the study were completed, 1991 was the latest
year for which enacted appropriations were available. Thus, we
used that year as a base. If you wish, we would be glad to update
our macroeconomic analyses for you.
Finally, you note several changes that could have been made in
our study that might have resulted in a less gloomy short-run pic-
ture. These changes include more mention of federal programs to
alleviate the impact of defense cutbacks on local economies, better
explanation of the ability of defense workers to switch to civilian
jobs, and less emphasis on worst-case analyses of local area im-
pacts.
The best solution for a displaced defense worker is a new job,
and our study emphasized that growth in nondefense jobs would
eventually offset the adverse effects of defense cutbacks. Indeed, we
argued that defense spending cuts could eventually benefit the
economy. Thus, I think we did emphasize that displaced defense
workers could be absorbed into the civilian sector. As you note, our
analyses of local-area effects began with a worst-case assessment.
Such an assessment is analytically feasible and suggests the mag-
nitude of the short-term problems facing local communities after a
major base closes or defense companies scale back production. But
the study clearly acknowledged that the calculations reflected a
worst-case assessment and noted factors that might ameliorate
short-term problems (see pages 33 and 41). In addition, our study
was generally positive about the long-term prospects for recovery
in communities affected by defense cuts.
These points notwithstanding, I understand the concern in the
Congress about the job losses associated with defense spending cut-
backs, particularly in a period of recession. I accept your point that
we should have acknowledged existing federal programs aimed at
mitigating the effects of defense cutbacks and provided more dis-
cussion of other actions that could be taken to mitigate the effects
of defense spending cutbacks.
I appreciate constructive criticism of the sort that you offered. It
helps to improve the quality and clarity of our analysis. I hope my
response is an adequate explanation of our reasoning and provides
some additional information. If I can be of further assistance,
please let me know.
Sincerely,
Robert D. Reischauer, Director.
314
U.S. Senate,
Committee on Appropriations,
Washington, DC, March 9, 1992.
Dr. Robert D. Reischauer,
Director, Congressional^Budget Office, Washington, DC.
Dear Dr. Reischauer: Recently, the Congressional Budget Of-
fice published a study, "The Economic Effects of Reduced Defense
Spending." Some key areas in which I have interest and concern
were omitted from your analysis.
First, the study assumes that all future defense spending reduc-
tions will be devoted to deficit reduction. Rather, for fiscal 1994
and 1995, Congress will determine the allocation of defense and
other discretionary funds under one spending cap. Beyond fiscal
1995, there is no cap at all. Therefore, your assumption regarding
the use of defense reductions is just that — an assumption. That fact
makes it impossible for you to predict with any certainty the eco-
nomic effects. This assumption puts other uses of the defense re-
duction, like public investment, at a disadvantage in future debate.
Second, the study lumps together defense reductions enacted in
1990 and 1991 with those which may be enacted this year. No
analysis is presented of the potential job loss attributable to just
the defense reduction which may be enacted for fiscal 1993.
Third, the study makes no mention of the previously enacted fed-
eral programs to alleviate the impact of defense reductions upon
local economies. Aside from unemployment benefits, dislocated de-
fense workers qualify for job training under Title III of the Job
Training Partnership Act (JTPA), as amended by the Omnibus
Trade and Competitiveness Act of 1988. The fiscal 1991 Defense
Authorization and Appropriations Acts (P.L. 101-510 and P.L. 101-
511) provided $150 million of adjustment assistance under JTPA
for the Department of Defense. These Acts also provided $50 mil-
lion specifically for funding Title K assistance to communities im-
pacted by defense cuts under the Public Works and Economic De-
velopment Act of 1965. The Office of Economic Adjustment within
the Defense Department and the President's Economic Adjustment
Committee will both help minimize economic dislocation from de-
fense reductions.
Fourth, the study could better explain that most threatened de-
fense workers will switch to civilian production, retrain, or retire
without entering the ranks of the long-term unemployed. This
country experienced far larger defense cutbacks following World
War II, Korea, and Vietnam. Much could be learned from the suc-
cess we had in transforming our economy following those conflicts,
but the report makes no mention of this.
Fifth and finally, certain parts of the study "represent a worst
case." When analyzing uncertain future economic events in re-
sponse to defense reductions, the results would be more fairly pre-
sented if they were accompanied by a sensitivity analysis which
also assumes a "best case." By focusing on three local economies,
the study gives the impression of devastating impact despite state-
ments to the effect that the nationwide effect is small.
315
I would like to have your views on these points as soon as pos-
sible.
Sincerely,
Robert C. Byrd, Chairman.
[From the Congress Daily page 3 September 18, 1992]
Budget — Bowsher Apologizes To Byrd For GAO Report
Comptroller General Charles Bowsher has apologized to Senate
Appropriations Chairman Byrd for a line-item veto report issued
earlier this year that Byrd characterized Thursday as a "piece of
trash." Byrd said Bowsher had written him a letter "taking a dif-
ferent position" on the line-item veto than the GAO had taken ear-
lier this year in an unsolicited report, "Line Item Veto: Estimating
Potential Savings." Byrd's office Thursday released the exchange of
letters between Byrd and Bowsher earlier this year regarding the
report.
The Byrd letter, dated March 2, criticized GAO for doing an un-
solicited report, failing to confer with Byrd before issuing it and
reaching conclusions Byrd said were unfounded. The report con-
tended that, had the line-item veto been used between 1984-89 and
had the president vetoed all items the OMB's "Statement of Admin-
istration Policy" found objectionable, six-year savings could have
reached $70 billion. But the GAO report acknowledged that the
president might not have vetoed all objectionable items and that
Congress might have overridden some vetoes. "Armed with this