Mandatory Spending: Using Budget Triggers to Constrain Growth
(31-JAN-06, GAO-06-276).
Prepared as part of GAO's basic statutory responsibility for
monitoring the condition of the nation's finances, the objectives
of this report were to (1) determine the feasibility of designing
and using trigger mechanisms to constrain growth in mandatory
spending programs and (2) provide an analysis of the factors that
led to differences between estimated and actual outlays in seven
mandatory budget accounts during fiscal years 2000 through 2004.
-------------------------Indexing Terms-------------------------
REPORTNUM: GAO-06-276
ACCNO: A46083
TITLE: Mandatory Spending: Using Budget Triggers to Constrain
Growth
DATE: 01/31/2006
SUBJECT: Budget activities
Budget administration
Budget controllability
Budget deficit
Budget outlays
Budget scorekeeping
Economic analysis
Economic policies
Financial management systems
Fiscal policies
Policy evaluation
Spending legislation
Strategic planning
******************************************************************
** This file contains an ASCII representation of the text of a **
** GAO Product. **
** **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced. Tables are included, but **
** may not resemble those in the printed version. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
******************************************************************
GAO-06-276
* Report to Congress
* January 2006
* MANDATORY SPENDING
* Using Budget Triggers to Constrain Growth
* Contents
* Results in Brief
* Background
* Objectives, Scope, and Methodology
* Trigger Mechanisms Could Help Constrain Mandatory Spending but
Must Be Carefully Designed
* Issues to Consider in Constructing a Trigger
* Issues to Consider in Designing the Triggered Response
* Expert Views on Trigger Mechanisms Are Mixed
* Reasons for Differences between Estimated and Actual Outlays in
Selected Accounts Varied
* Legislation Enacted After Original Estimates Explained Many
Differences between Estimated and Actual Outlays
* Economic Factors Were Especially Important in Some Programs'
Differences
* Technical Factors Explained a Broad Spectrum of Differences
* Conclusions
* Matter for Congressional Consideration
* Agency Comments
* Illustrative Examples of Triggers and Responses for Case Study
Accounts
* Account Name
* Administering Organization
* Program Description
* Funding Source
* Differences between Estimated and Actual Outlays
* Explanation of Key Differences
* Illustrative Triggers and Response
* Account Name
* Administering Organization
* Program Description
* Funding Source
* Differences between Estimated and Actual Outlays
* Explanation of Key Differences
* Ideas for Improving the Accuracy of Estimates
* Account Name
* Administering Organization
* Program Description
* Funding Source
* Differences between Estimated and Actual Outlays
* Explanation of Key Differences
* Illustrative Trigger and Response
* Account Name
* Administering Organization
* Program Description
* Funding Source
* Differences between Estimated and Actual Outlays
* Explanation of Key Differences
* Account Name
* Administering Organization
* Program Description
* Funding Source
* Differences between Estimated and Actual Outlays
* Explanation of Key Differences
* Currently Existing Program Trigger and Response
* Illustrative Trigger and Response
* Account Name
* Administering Organization
* Program Description
* Funding Source
* Differences between Estimated and Actual Outlays
* Explanation of Key Differences
* Currently Existing Program Triggers and Responses
* Illustrative Trigger and Response
* Account Name
* Administering Organization
* Program Description
* Funding Source
* Differences between Estimated and Actual Outlays
* Explanation of Key Differences
* Illustrative Trigger and Response
* Account Name
* Administering Organization
* Program Description
* Funding Source
* Differences between Estimated and Actual Outlays
* Explanation of Key Differences
* Illustrative Triggers and Responses
* Analysis of Total Outlays, Receipts, and Fiscal Position
* Aggregate Mandatory Spending Estimates Were Close to Actual
Outlays but Large Differences Appear at the Account Level
* Differences between Estimated and Actual Mandatory Outlays Had
Limited Effect on the Unified Deficit/Surplus
* Mandatory Budget Accounts
* GAO Contact and Acknowledgments
Report to Congress
January 2006
MANDATORY SPENDING
Using Budget Triggers to Constrain Growth
Contents
Tables
Figures
Abbreviations
January 31, 2006Letter
The President of the Senate The Speaker of the House of Representatives
Over the next few decades as the baby boom generation retires and health
care costs continue to rise, federal spending on retirement and health
programs-Social Security, Medicare, Medicaid, and other federal pension,
health, and disability programs-will grow dramatically. Absent policy
changes on the spending and/or revenue sides of the budget, a growing
imbalance between expected federal spending and tax revenues will mean
escalating and ultimately unsustainable federal deficits and debt that
threaten our future economy and national security as well as the standard
of living for the American people.1
Given rising deficits, the expiration of the Budget Enforcement Act (BEA)
of 1990,2 and the long-term fiscal outlook, new budget control mechanisms
are needed. Accordingly, there have been calls for the reintroduction of
discretionary spending caps and PAYGO rules. Although PAYGO was effective
in preventing legislative actions that increased the deficit, it did not
address increases that occurred absent legislative action. Constraining
the growth of existing mandatory spending programs requires additional
action.
In our 1994 report on capping mandatory spending,3 we noted that an
alternative method to prompt congressional review of mandatory spending
trends would be to require Congress to vote periodically on whether or not
to make program changes when mandatory spending exceeds certain targets.
One way to do this and potentially achieve greater fiscal responsibility
would be to create triggers for individual mandatory
programs-predetermined spending or revenue thresholds-that signal the need
for some type of action to be taken on the program. Once tripped, the
trigger could drive either a review or an automatic action. It could, for
example, trigger a requirement for Congress to either review or reaffirm
acceptance of the unexpected increase in actual program spending or
projections. Alternatively, it could trigger previously specified changes
to the program that automatically take effect to reduce spending or
increase program revenue.
Insufficient transparency regarding both the expected and actual cost path
for spending and revenue decisions hampers the ability of decision makers
to make informed choices. In previous work, we have called for increased
disclosure and recognition of long-term costs of proposed policies and
programs.4 The ability to monitor actual spending paths can also play an
important role in decisions about both the overall fiscal position and the
allocation of scarce resources. Moreover, as we reported in 1994, a cap on
mandatory spending would have little if any effect on the longer-term
growth trends until and unless issues of underlying program eligibility
and benefits are addressed.5 Thus, efforts to constrain growth in
mandatory programs need to be focused on and tailored to individual
programs. One way to assess mandatory spending is to analyze growth by
examining the estimated and actual outlays for each program. Because
budget estimates can be linked to achieving fiscal responsibility in the
government, identifying and understanding recurring patterns between
mandatory account budget estimates and actual results can facilitate
future budget decisions.
This report, prepared as part of our basic statutory responsibility for
monitoring the condition of the nation's finances, examines issues related
to using such triggers on mandatory programs. The objectives were to (1)
determine the feasibility of designing and using trigger mechanisms to
constrain growth in mandatory spending and (2) provide an analysis of the
factors that led to differences between estimated and actual outlays in
seven mandatory budget accounts during fiscal years 2000 through 2004.
This report does not deal with the question of projected costs at the time
decisions are made but instead with the need for responses when there is
significant growth. Analogous analyses could be applied to the revenue
side of the budget (e.g., tax expenditures).
Results in Brief
One idea to constrain growth in mandatory programs is to develop triggers
that, when tripped, prompt a response. A trigger could result in a "hard"
or automatic response, unless Congress and the President acted to override
or alter it. Alternatively, reaching a trigger could require a "soft"
response, such as a report on the causes of the overage, development of a
plan to address it, or an explicit and formal decision to accept or reject
a proposed action or increase. By identifying significant increases in the
spending path of a mandatory program relatively early and acting to
constrain it, Congress may avert larger financial challenges in the
future. However, both in establishing triggers and in designing the
subsequent actions to be triggered, the integrity of program goals needs
to be preserved.
The budget experts we consulted had mixed views of triggers. Some
expressed strong support for budget triggers. These proponents of triggers
noted that mandatory spending is currently unconstrained and a mechanism
that causes decision makers to at least periodically reevaluate spending
is better than allowing spending to rise unchecked. Others, however,
expressed considerable skepticism about the effectiveness of triggers;
many felt they would be circumvented or ignored. For example, one expert
pointed to "accounting tricks" that have resulted from triggers with hard
responses, such as when Congress mandated certain costs not be counted
against spending limits so as to avoid across-the-board cuts. Others
worried that applying budget triggers to various mandatory programs
diverts attention from what they see as the real source of the nation's
fiscal woes-health care spending. Further, they felt that establishing
triggers on such programs could mislead the public into thinking that the
long-term fiscal problem had been addressed, thus delaying efforts to
appropriately address it.
Any discussion to create triggered responses and their design must
recognize that unlike controls on discretionary spending, there is some
tension between the idea of triggers and the nature of entitlement and
other mandatory spending programs. These programs-as with tax provisions
such as tax expenditures-were designed to provide benefits based on
eligibility formulas or actions as opposed to an annual decision regarding
spending. This tension makes it more challenging to constrain costs and to
design both triggers and triggered responses. At the same time, with only
about one-third of the budget under the control of the annual
appropriations process, considering ways to increase transparency,
oversight, and control of mandatory programs must be part of addressing
the nation's long-term fiscal challenges.
Ignoring significant growth in mandatory accounts is inconsistent with
evaluation of programs and their costs. While we appreciate the concerns
raised by budget experts, we believe that, if carefully designed, budget
constraint mechanisms such as triggers should be considered as existing
programs are reexamined or reauthorized and when new programs are created.
Each program would need to be considered individually to ensure that any
actions that are triggered preserve program goals. The seven mandatory
accounts we examined helped inform our thinking about budget constraint
mechanisms, and we present illustrative examples of how growth could be
constrained in many of the accounts discussed in appendix I.
For seven case study accounts,6 we categorized the reasons provided by
agencies for differences between estimated and actual outlays during
fiscal years 2000 through 2004 as the result of (1) legislative changes
enacted after original estimates were submitted, (2) economic changes such
as interest and unemployment rates, or (3) technical changes, which is a
residual category that represents revisions to budget estimates that
cannot be attributed to legislative or economic factors. Our analysis of
the reasons for differences between estimated and actual outlays showed
that out of 40 differences, legislative changes7 were the primary reason
for 19, economic changes for 7, and technical changes for 13. In one case,
it was unclear which factors most significantly caused the difference
between estimated and actual outlays. In many cases, a combination of
factors caused the differences.
OMB and agencies responsible for the seven case study accounts either did
not have comments or provided comments that were clarifying and/or
technical in nature. These comments were incorporated as appropriate.
Background
BEA8 divided federal spending into two broad categories: discretionary and
mandatory. Discretionary spending refers to outlays from budget authority
that is provided in and controlled by appropriation acts; it can and has
been controlled through annual, adjustable dollar limits (spending caps)
that permanently lower the base for future appropriations. Mandatory
spending9 refers to outlays resulting from budget authority that is
provided in laws other than appropriation acts, for example, entitlement
programs such as Medicare, Food Stamps, and veterans' pensions. Mandatory
spending-like tax expenditures-is governed by eligibility rules and
benefit formulas, which means that funds are spent as required to provide
benefits to those who are eligible and wish to participate. Therefore,
unforeseen events such as changes in the economy or additional demands for
services can translate into unanticipated additional program outlays.
Congress controls spending for these programs indirectly by defining
eligibility and setting the benefit or payment rules rather than directly
through appropriation acts. On an annual basis, however, mandatory
spending is relatively uncontrollable since Congress and the President
must change substantive law in order to further increase or decrease
outlays. This makes it more challenging to constrain costs and to design
both triggers and triggered responses.
Over the past 4 decades, we have seen mandatory spending grow as a share
of the total federal budget. For example, figure 1 shows that spending on
mandatory programs rose from approximately 42 percent of total federal
spending in 1984 to about 49 percent in 1994, and to 54 percent in 2004.
This growth is projected to continue with mandatory programs claiming
about 58 percent of total federal spending in 2010.
Figure 1: Federal Spending for Mandatory and Discretionary Programs
The nation's long-term fiscal outlook is daunting under many different
policy scenarios and assumptions. For instance, under a fiscally
restrained scenario, if discretionary spending grew only with inflation
over the next 10 years and all existing tax cuts expire when scheduled
under current law, spending for Social Security and health care programs
would grow to consume over three-quarters of federal revenues by 2040 (see
fig. 2). On the other hand, if discretionary spending grew at the same
rate as the economy-measured by Gross Domestic Product (GDP)-in the near
term and if all tax cuts were extended, federal revenues may just be
adequate to pay interest on the growing federal debt by 2040 (see fig. 3).
Numerous alternative scenarios can be developed incorporating different
combinations of possible policy choices and economic assumptions, but
these two scenarios can be viewed as "bookends" showing a range of
possible outcomes.
Figure 2: Composition of Spending as a Share of GDP under Baseline
Extended
Note: In addition to the expiration of tax cuts, revenue as a share of GDP
increases through 2015 due to (1) real bracket creep, (2) more taxpayers
becoming subject to the alternative minimum tax (AMT), and (3) increased
revenue from tax-deferred retirement accounts. After 2015, revenue as a
share of GDP is held constant.
Figure 3: Composition of Spending as a Share of GDP Assuming Discretionary
Spending Grows with GDP After 2005 and All Expiring Tax Provisions Are
Extended
Note: Although expiring tax provisions are extended, revenue as a share of
GDP increases through 2015 due to (1) real bracket creep, (2) more
taxpayers becoming subject to the AMT, and (3) increased revenue from
tax-deferred retirement accounts. After 2015, revenue as a share of GDP is
held constant.
As both these simulations illustrate, absent policy changes on the
spending and/or revenue side of the budget, the growth in spending on
federal retirement and health entitlements will encumber an escalating
share of the government's resources. Neither slowing the growth in
discretionary spending nor allowing the tax provisions to expire-nor both
together-would eliminate the imbalance. Although revenues will be part of
the debate about our fiscal future, making no changes to Social Security,
Medicare, Medicaid, and other drivers of the long-term fiscal gap would
require at least a doubling of taxes-and that seems implausible.
Accordingly, substantive reform of Social Security and our major health
programs remains critical to recapturing our future fiscal flexibility.
These long-term spending projections can largely be attributed to the
aging population and increased health care costs. This does not, however,
mean that the rest of the budget should be exempt from review. It is
important to periodically look at mandatory accounts in order to determine
possible ways to constrain spending and ensure a more accurate and
responsible federal budget process.
Congressional interest in fiscal discipline and the adoption of budget
tools to control mandatory spending are not new. The Balanced Budget and
Emergency Deficit Control Act of 1985, commonly referred to as
Gramm-Rudman-Hollings (GRH), established declining deficit targets and a
sequestration procedure to reduce spending if those targets were exceeded.
GRH was amended several times, most significantly by BEA in 1990. One
important reason for BEA's success in reducing the deficit during the
1990s was that the process enforced a previously reached agreement to
reduce the deficit. However, recurring surpluses at the end of the decade
caused a new debate to emerge and undermined the acceptance of BEA's
spending caps and PAYGO enforcement. BEA rules were not extended beyond
their scheduled expiration date at the end of fiscal year 2002.
In the past, mandatory spending caps were proposed as a way to control the
growth of mandatory programs. This idea was discussed in a report we
issued in 1994.10 Mandatory caps fail to address underlying eligibility
and benefits formulas-which drive spending. In addition, if caps were
imposed in the context of a control requiring across-the-board spending
cuts, they would present agencies with difficulties in successfully
reducing their program spending to stay within limits, and perhaps lead to
a cycle of continual sequestrations. This difficulty is because in such a
regime, any shortfalls in savings or growth in spending that occurred
despite agency efforts would be added to the amount of cuts required in
the next year. Moreover, the mandatory programs that would be most
affected by a cap-because of their high and/or volatile growth rates-are
also the programs for which a cap would be hardest to implement.
In the mid-1990s, there was a period when the idea of constraining
greater- than-expected growth through the use of triggers surfaced.
However, it coincided with a period when actual growth generally was less
than expected. Recently, with the reappearance of large deficits, there
has been a resurgence of interest in restoring budget controls and
containing the growth in both discretionary and mandatory spending. For
example, in
2005, numerous bills to reinstate fiscal discipline were proposed.11
Moreover, in May 2005, OMB issued a memo to agencies that required them to
propose offsets to any administrative action that would increase mandatory
spending.
Budget estimates and actual outlays are determined over a period that
spans nearly 2 years: from the time the President's budget is formulated,
about a year before the start of the fiscal year in question, to the
completion of that fiscal year. Within this 2-year lag period between
original estimates and actual outlays, legislative, economic, and
technical factors can affect program outlays. Budget estimates are revised
part way through the fiscal year and included in the budget request for
the following fiscal year. These revisions reflect updated technical and
economic assumptions as well as any legislative changes. Also, midsession
reviews conducted during the summer, usually in July, update budget
estimates prior to the completion of the fiscal year. In addition, both
CBO and OMB estimate the cost of bills that affect mandatory spending.
Objectives, Scope, and Methodology
The objectives of this study were to (1) determine the feasibility of
designing and using trigger mechanisms to constrain growth in mandatory
spending and (2) provide an analysis of the factors (legislative,
economic, and technical) that led to differences between estimated and
actual outlays in seven mandatory budget accounts during fiscal years 2000
through 2004. This second objective contributed to our understanding of
programs, helped us better appreciate the reasons behind growth in
mandatory accounts that experienced relatively large dollar changes, and
more fully informed our thinking about triggers.
To accomplish our first objective, we performed a literature search on
mechanisms to constrain mandatory spending and had discussions with
numerous budget experts from OMB, CBO, the Senate Budget Committee staff,
and various policy research organizations. Based on our research,
interviews at agencies, and discussions with experts, we then considered
possible approaches for budgetary constraint within each account.
To accomplish our second objective we extracted from OMB's budget database
mandatory outlays of accounts where 50 percent or more of the outlays were
mandatory. We analyzed these data for fiscal years 2000 through 2004. To
determine the estimated and actual outlays for each year, we used the
original budget estimate and the actual outlays reported 2 years later,
after the end of the fiscal year. For example, when determining the
difference between estimated and actual outlays for fiscal year 2000, we
compared the fiscal year 2000 budget estimates published in February 1999
to the actual outlays published in February 2001.
From the 534 accounts with outlays at least half mandatory, we selected
the top 10 accounts that experienced the greatest average dollar change
between original estimate and actual outlays in absolute value terms for 5
fiscal years (2000-2004). The complete list of these accounts is included
as appendix III. These 10 accounts, which represent approximately 50
percent of total average mandatory outlays, include (1) Interest on
Treasury Debt Securities, (2) Unemployment Trust Fund, (3) Commodity
Credit Corporation Fund,12 (4) Federal Supplementary Medical Insurance
Trust Fund (Medicare Part B), (5) Federal Hospital Insurance Trust Fund
(Medicare Part A), (6) Grants to States for Medicaid,13 (7) Rail Industry
Pension Fund, (8) Federal Direct Student Loan Program (FDLP) Account, (9)
Payments to Health Care Trust Funds, and (10) Mutual Mortgage Insurance
Program Account (MMI). Because many of the programs we selected are
relatively big, large dollar increases may represent small percentage
increases relative to program size.
After initial analysis, we excluded three of these accounts from further
analysis: Interest on Treasury Debt Securities, MMI,14 and Payments to
Health Care Trust Funds. We eliminated the U.S. Treasury account because
interest payments are a function of all other funding decisions and thus
provide little insight into trigger design.15 We excluded the MMI account
because the program itself is discretionary-only the large mandatory
reestimates of its credit subsidy required by the Federal Credit Reform
Act of 1990 caused it to fall into our original sample. Because decisions
about the size of this program are annually made in the appropriations
process and can be informed by the reestimates of previous years' loans,
there is no need for separate triggers. Finally, we excluded the Payments
to Health Care Trust Funds account because the payments are classified as
intragovernmental transfers and therefore do not affect overall budget
outlay data. Moreover, these transfers are captured within other accounts
in our sample.
Figure 4 below shows the 5-year average difference between estimated and
actual mandatory outlays in absolute value terms for the seven accounts we
reviewed. These differences ranged from $9.4 billion in the Unemployment
Trust Fund to $2.6 billion in FDLP.
Figure 4: Five-Year Average Differences between Estimated and Actual
Mandatory Outlays, Fiscal Years 2000 through 2005
To gain more perspective on what factors contributed to the differences
between estimated and actual outlays in the remaining seven accounts, we
met with officials from the cognizant agencies to determine if the reasons
behind the differences were (1) legislative, (2) economic, (3) technical,
or a combination of the three. We did not independently verify the
explanations agencies provided for differences.
Our work was done between May 2005 and January 2006 in Washington, D.C.,
in accordance with generally accepted government auditing standards.
Trigger Mechanisms Could Help Constrain Mandatory Spending but Must Be
Carefully Designed
The purpose of a budget trigger is to either automatically cause some
action to occur or to prompt decision makers to evaluate and consider
responding to rising costs. For example, where differences between
expected and actual growth in a program exceed a specified amount,
Congress could decide explicitly-by voting-whether to accept the slippage
or could take action to bring the spending path closer to the original
goal by recouping some or all of the slippage through changes in the
program. Our background research, work in case study agencies, and
discussions with budget experts highlighted several issues to consider
when designing triggers and their resulting actions, such as the extent of
agreement among decision makers about underlying fiscal goals, measures
selected to trip the trigger, and the triggered response.
While a budget process can surface important issues, it is not a
substitute for substantive debate-no process can force agreement where one
does not exist. Accordingly, the success of any effort to constrain growth
depends on whether there is widespread agreement on the underlying goals;
absent such agreement, any trigger would likely be circumvented. For
example, underlying the successful budget enforcement mechanisms embodied
in BEA was the broadly accepted goal of deficit reduction and an agreement
on a specific set of legislative changes to reach that goal. Its triggers
were centered around measures that Congress could control-discretionary
spending caps and changes to entitlement and tax laws. However, once the
budget moved into surplus in the late 1990s and there was no longer
agreement on fiscal goals, actions were taken to bypass BEA controls. For
example, the consolidated appropriations acts for both fiscal years 2000
and 2001 mandated that OMB change the PAYGO scorecard balance to zero.
Both OMB and CBO estimated that without instructions to change the
scorecard, sequestrations would have been required in 2001.
Other countries we have studied have sought to address national priorities
by developing explicit goals to guide fiscal policy and justifying their
goals with compelling rationales that often pointed out the potential
fiscal and economic benefits of budgetary discipline. In a 2000 report,16
we noted that having fiscal goals anchored by a rationale that is
compelling enough to make continued restraint acceptable is critical to
sustain support for budgetary discipline.
Issues to Consider in Constructing a Trigger
One of the reasons for the success of BEA was its link to congressional
action.17 Discretionary spending caps and PAYGO constrained congressional
action-BEA held Congress accountable only for things it could control and
not for the effect of economic or technical factors on spending or
revenues. This was both the strength and the limitation of PAYGO. Triggers
seek to go beyond the PAYGO regime by subjecting program growth to
scrutiny even where that growth is the result of economic, population, or
other factors outside congressional control. Triggers recognize that even
the best estimates can turn out to be wrong and that decision makers who
expected one path might wish to consider changes in a program where the
path is significantly different from what was anticipated.
In general, there are two types of responses to budget triggers-soft and
hard-depending on what type of action results when the trigger is tripped.
A "soft" response prompts special consideration of a program or a proposal
for action when a certain threshold or target is breached. Examples of
soft responses that could be triggered include requiring the administering
agency to prepare a special report explaining why the trigger's threshold
was breached, or requiring the President to submit a proposal for reform.
An example of a soft response already exists in the Medicare program,
which requires the President to submit a proposal to Congress for action
if the Medicare Trustees determine in 2 consecutive years that the general
revenue share of Medicare spending is projected to exceed 45 percent
during a 7-year period.18 In addition, a few Social Security reform
proposals have included language requiring presidential and congressional
action if the Social Security Board of Trustees determines that the
balance ratio of either of the Social Security trust funds will be zero
for any calendar year during the succeeding 75 years.19
Soft responses can help in alerting decision makers of potential problems
but they do not ensure that action to decrease spending or increase
revenue is taken. With soft responses, the fiscal path continues unless
Congress and the President take action. In contrast, a trigger could lead
to "hard" responses requiring a predetermined, program-specific action to
take place, such as changes in eligibility criteria and benefit formulas,
automatic revenue increases, or automatic spending cuts. With hard
responses, spending is automatically constrained, revenue is automatically
increased, or both, unless Congress takes action to override. Figure 5
below illustrates the conceptual differences between hard and soft
responses of a budget trigger.
Figure 5: Conceptual Differences between Hard and Soft Responses
In establishing triggers, both near- and long-term perspectives need be
considered. For some programs it might be appropriate to tie triggers to
historical data. For example, unexpected spending growth in student loans
might be measured against past historical spending data. However, for
other programs that expose the government to long-term commitments-such as
Medicare or Social Security-it might be more appropriate to tie the
trigger to projections of future spending. Social Security, however,
represents a large long-term commitment of future resources. Thus, growth
for this program might be measured against changes in actuarial
projections of Social Security's 75-year outlook. Such an approach could
be used for other programs with long-term commitments, such as pension
insurance, if good long-term projections become available.
Since all estimates are subject to some uncertainty, the triggering
mechanism should not be so tight that it is overly sensitive to normal
variation in budget estimation. One way to address this concern is to
establish a normal or expected range of budget uncertainty and set a
trigger level that falls outside this range. For example, if a program's
actual outlays historically fall within plus or minus 5 percent of
estimated outlays, a trigger set at a level greater than 5 percent would
best signal unexpected growth. This approach resembles one CBO uses for
certain programs to analyze the budgetary effects of legislative
proposals.20 Using a probabilistic model, CBO estimates the weighted
average of the effects associated with all possible sets of circumstances,
taking into account their respective probabilities. Such an approach could
be adapted to establish a range of uncertainty around a budget estimate.
Triggers also could be used to ensure that policy changes actually achieve
intended reductions in spending growth. Such triggers could address
concerns that some budget constraint mechanisms create the false
impression that long-term problems have been addressed.
Although any hard response can be overridden by congressional action, it
could be important to incorporate a more automatic escape clause into
budget enforcement mechanisms such as triggers. Effective budget
enforcement mechanisms need to be able to accommodate changing budget
policy and political environments in which future outcomes are difficult
to predict. For example, periods of economic growth may be brief or
sustained, but inevitably are followed by periods of economic downturn
that may be shallow or deep. Escape mechanisms, such as expiration dates,
allow budget policies and procedures to be renegotiated later. In addition
to expiration dates, House or Senate rules can provide flexibility. For
example, any Senator may raise a point of order against legislation
violating PAYGO rules prohibiting consideration of revenue or direct
spending legislation that is not deficit-neutral. However, the point of
order may be waived if there is broad consensus on the need to do so-that
is, if there is an affirmative vote of three-fifths of the membership.
Although they provide important flexibility, escape clauses can be
overused. For example, in fiscal year 2002, the Department of Defense and
Emergency Supplemental Appropriations Act21 instructed that $130.3 billion
in costs be eliminated from the PAYGO scorecard. Both OMB and CBO
estimated that without instructions to change the scorecard, a
sequester-across-the-board spending cuts-would have been required in 2002.
In addition, many programs were exempt from PAYGO's sequestration
requirement. These exemptions meant that the full brunt of any sequester
was concentrated in the remaining programs, resulting in cuts so draconian
that Congress and the President changed the targets rather than impose the
required cuts.
Issues to Consider in Designing the Triggered Response
Whether a triggered response is soft, hard, or a combination of the two,
efforts to constrain growth in mandatory programs need to be focused at
the program level. The experience with GRH highlights the importance of
individually designed triggers and responses. The deficit-neutrality
targets under GRH triggered a hard response-across-the-board spending
cuts-if they were not met. The deficit targets under GRH were not achieved
due to the inability of Congress and the President to control all of the
factors-mainly economic factors-that affected whether the trigger would be
breached and their unwillingness to accept the across-the-board cuts that
would have been necessary to meet the deficit targets.
In developing program-specific triggers and responses, proposed changes in
underlying benefits structure and design of mandatory programs can be
considered in the context of the factors that drove the growth and the
goals and objectives of specific programs. For example, certain programs
such as unemployment insurance and crop assistance are designed and
intended to have a countercyclical effect on the economy. That is, they
are aimed at reducing the size and duration of swings in economic activity
in order to keep economic growth closer to a pace consistent with low
inflation and high employment. Thus, a triggered response in these
programs needs to be sensitive to whether growth is being driven by
automatic budget stabilizers. For example, a rise in the unemployment rate
would by design increase outlays in federal unemployment insurance not
only to provide assistance to the unemployed but also to stabilize the
economy. If a trigger were established that resulted in a contractionary
response, it could undermine these important goals and exacerbate the
effects of unemployment on the economy. In a January 2002 report,22 the
Congressional Research Service (CRS) suggested one option to avoid
procyclical triggers would be to delegate to some entity-for instance
Congress or an executive department-the responsibility for evaluating each
year whether deteriorating economic conditions would make a trigger
detrimental. If conditions were found to be deteriorating, decisions would
need to be made about whether and how to implement any reduction. CRS
acknowledged, however, that this type of proposal could be criticized on
the grounds that it is based on a subjective decision and thus could be
prey to the sort of political pressures that critics fear would undermine
a trigger. Indeed, one budget expert we met with expressed concern that in
devising a budget trigger, it would be helpful to acknowledge political
pressures by considering who would judge progress against the trigger and
the neutrality of the judging entity.
The programs and agencies we reviewed have objectives and missions that
contribute to the achievement of public policy goals such as income
security, feeding the nation, fostering higher education, and providing
health care. To these ends, these programs are designed to provide
entitlements-benefits and assistance-to eligible recipients. While
striving to meet these commitments, our nation is faced with a daunting
long-term fiscal outlook based on the challenges of an aging population,
unsustainable deficits, and mounting debt while also ensuring truth and
transparency. Figure 6 depicts the inherent tension in balancing public
policy goals and long-term fiscal challenges.
Figure 6: Balancing Public Policy Goals and Long-term Fiscal Challenges
Addressing this tension invariably entails difficult political choices
among competing programs that promise benefits to many Americans but are
collectively unaffordable and unsustainable at current revenue levels. In
February 2005 we highlighted the size of fiscal imbalances looming in the
future and the challenge of our policy process to act with more foresight
to take early action on problems that may not constitute an urgent crisis
but pose important longer-term threats to the nation's fiscal, economic,
security, and societal future.23 Budget triggers are mechanisms that can
encourage and facilitate such action.
To help us better consider the implications of establishing triggers, we
looked at seven mandatory accounts with relatively large differences
between estimated and actual outlays: Commodity Credit Corporation (CCC),
Federal Direct Student Loan Program Account, Grants to States for
Medicaid, Federal Hospital Insurance (HI) Trust Fund (Medicare Part A),
Federal Supplementary Medical Insurance (SMI) Trust Fund (Medicare Part
B), Rail Industry Pension Fund, and the Unemployment Trust Fund. We
explored ways in which existing triggers and their corresponding actions
could be revised, as well as an array of new trigger mechanisms that take
into consideration the issues just discussed and could be adopted to
promote better budgeting in light of the nation's long-term fiscal
outlook.
It is important to consider the data upon which the trigger will
hinge-future projections based on historical data, growth as a percent of
GDP, total growth, or another measure altogether. For example, Congress
has established a trigger to constrain growth in Medicare spending for
physicians' services. The sustainable growth rate (SGR) is a statutorily
set formula that estimates the allowed rate of increase in spending for
physicians' services; that rate is used to construct the spending target
for the following calendar year. If actual spending exceeds the cumulative
SGR targets, fee updates in future years must be lowered sufficiently both
to offset the accumulated excess spending and to slow expected spending
for the coming year.24 The Medicare Prescription Drug, Improvement, and
Modernization Act of 2003 (MMA) established another trigger-the general
revenue share of Medicare spending. If the Medicare Trustees determine in
2 consecutive years that the general revenue share is projected to exceed
45 percent during a 7-year projection period, the President must submit a
proposal to Congress for action. To date, this threshold has not been
breached and thus no response has been triggered. However, Medicare
Trustees are expected to determine the first breach in their upcoming 2006
report as the trigger is projected to be tripped in 2012, which falls
within the 7-year projection period captured in that report. For
unemployment insurance, a trigger was established around balances in the
Unemployment Trust Fund. When funds accumulating in federal unemployment
accounts reach statutorily set limits, a distribution of the "excess"
funds from the
trust fund to individual states' accounts-called "Reed
Distributions"25-are automatically triggered based on each state's share
of covered wages. One way to constrain federal spending would be to
increase the statutory cap on federal unemployment accounts, thus making
it more difficult to trigger Reed Distributions to states. By making it
more difficult to trip the trigger, funds could continue to build during
economic prosperity and be available to states when truly needed to
counter rising unemployment.
Our analysis allowed us to develop a list of illustrative examples, which
analyze the related trade-offs involved in balancing restraint with
optimization of program goals. These are shown in appendix I, along with a
brief description of the program and account. Finally, where appropriate
we present illustrative examples of hard responses that could be
established to constrain spending. We do not specifically advocate any of
these approaches-they are presented for illustrative purposes only to
provide a sense of the types of trigger and resulting actions that could
be established. Although the illustrative examples we developed apply
specifically to the seven case study accounts that we reviewed, we believe
the information can further the larger policy conversation about how to
increase oversight of the path of mandatory spending and advance and
encourage budgetary discipline.
Expert Views on Trigger Mechanisms Are Mixed
We interviewed budget experts from OMB, CBO, the Senate Budget Committee
staff, and various policy research organizations to discuss views on using
triggers to constrain mandatory spending. Overall, views were mixed. While
some were more in favor of triggers than others, many expressed concern
that they would be circumvented or ignored, thereby questioning their
effectiveness. In addition, many were concerned that triggers could
jeopardize the underlying intent of mandatory programs. Several experts
also pointed to the need to ensure that any triggers developed be
carefully designed to avoid procyclical effects.
Some of the experts expressed strong support for budget triggers. These
individuals believed that triggers with hard responses had the potential
to constrain mandatory spending and that the accountability added by
triggers would be preferable to the current unconstrained environment. For
example, one expressed concern about the debt burden being permitted to
mount for future generations in order to avoid the reduction in benefits
or increase in taxes needed to finance current benefits. Linking revenues
and spending with GDP, she argued, would help avoid such generational
inequities. Another added that under current policy, spending grows
automatically, by default, faster than tax revenues as the population ages
and health costs soar. He argued that only by changing the budget's
autopilot programming can we gain the flexibility needed to continually
improve government policies and services.
Others, however, said that triggers reduced accountability because they
enable decision makers to publicly extol budget constraint but quietly
continue to increase spending. One pointed to "accounting tricks" that
have resulted from triggers with hard responses, such as when Congress
mandated certain costs not be counted against spending limits so as to
avoid across-the-board cuts. However, as discussed previously, triggers
also could be used to ensure that policy changes actually achieve intended
reductions in spending growth. Such triggers could address concerns that
some budget constraint mechanisms create the false impression that
long-term problems have been addressed.
Many expressed skepticism that budget constraint mechanisms such as
triggers would be adhered to; one cited Medicare's SGR as an example. The
SGR system is designed to apply financial brakes whenever actual spending
for physicians' services exceeds predefined spending targets. It does this
by reducing physician fees or limiting their annual increase. Because the
actual versus target spending comparison is cumulative, future fee updates
are reduced to lower future actual spending below future target spending
until total cumulative actual spending is the same as total cumulative
target spending. However, fee declines were averted for 2003, 2004, and
2005 by administrative and legislative actions that modified or overrode
the SGR system.
Some experts worried that applying budget triggers to various mandatory
spending programs would divert attention from the real source of the
nation's fiscal woes-health care-whose costs continue to rise faster than
GDP. They pointed to CBO data as evidence that, outside of health care and
to a lesser extent Social Security, virtually all other mandatory programs
are decreasing or holding steady as a percent of GDP. Accordingly, they
expressed concern that establishing triggers on such programs could
mislead the public into thinking that the long-term fiscal problem had
been addressed, thus delaying efforts to appropriately address it.
Many of the budget experts raised concerns about triggers jeopardizing the
important underlying missions and program goals financed by mandatory
accounts. In particular, concerns were raised about undermining
countercyclical effects of programs such as unemployment insurance, Food
Stamps, and the Earned Income Tax Credit. Some noted that the desire to
preserve program goals is the reason why triggers with hard responses have
not worked in the past. With respect to the SGR, for example, one expert
explained that the reason Congress overrides the trigger is to ensure
doctors do not stop accepting Medicare patients.
Finally, a couple of experts pointed out that triggers need not only apply
to spending; the revenue side of the budget should also be addressed. One
noted, for example, that an increase in taxes to cover spending growth
would increase visibility to the public and thus permit the American
people to be more aware of how much they are paying for services. Applying
triggers to tax cuts was an issue considered in 2001 when the budget was
in surplus and tax cuts were proposed. For example, Federal Reserve
Chairman Greenspan at that time expressed his preference for a trigger
that would make tax cuts contingent on the realized net debt level.
Comptroller General Walker also raised the possibility of using a trigger
to return a "surplus dividend" if actual surpluses occurred in excess of
specific levels. Ultimately, however, triggers were not adopted. Instead,
tax cuts were enacted through 2010 even though substantial deficits have
reappeared. In addition, as we reported in a February 2005 testimony,26
there has been an extensive use of tax incentives, rather than direct
spending authority, to fund social objectives. As we reported in September
2005,27 the sum of revenue loss estimates associated with tax
expenditures-such as tax exclusions, credits, and deductions-was nearly
$730 billion in 2004.28 Many tax expenditures operate like mandatory
spending programs and generally are not subject to reauthorization. Such
tax expenditures are embedded in the tax system and are off the radar
screen for the most part. This is a concern from a budgetary standpoint
because federal dollars committed to fund these expenditures do not
compete in the annual appropriations process and are effectively "fully
funded" before any discretionary spending is considered. The analysis we
applied to spending in this report would also be useful in examining tax
expenditures. However, challenges in defining and measuring tax
expenditures, to some extent, would affect any effort to curtail revenues
foregone through tax expenditures. For example, after taxpayers have taken
advantage of tax expenditures, the federal government still may not know,
with much certainty, how much tax revenue was foregone, who benefited, and
what results were achieved.29
Reasons for Differences between Estimated and Actual Outlays in Selected
Accounts Varied
To better appreciate the reasons behind growth in mandatory accounts and
thus inform our thinking on triggers, we examined the reasons for
differences between originally estimated and actual outlays for seven
mandatory accounts that experienced relatively large dollar changes.30
Based on agencies' explanations of differences between estimated and
actual outlays of the case study accounts we examined, we found that
legislation enacted after original estimates were submitted was the
primary driver in 19 out of 40 differences during fiscal years 2000
through 2004. Economic factors, such as changes in interest and
unemployment rates, were primarily responsible for 7 differences. Finally,
technical factors, which cover a broad spectrum, most significantly drove
13 out of 40 differences. In one case, it was unclear which factors most
significantly caused the difference between estimated and actual outlays.
In many cases, a combination of factors resulted in differences.
In categorizing agencies' explanations for differences between estimated
and actual outlays, we applied criteria similar to those that CBO uses in
its annual budget and economic outlook reports to categorize changes as
legislative, technical, and economic. However, in our report, legislative
action was classified in a somewhat different manner from the method that
CBO applies. Whereas we examined the actual budgetary effect that resulted
from the legislation, CBO projects the anticipated future
budgetary effect of legislation.31 Figure 7 describes the criteria that we
applied to categorize agencies' explanations into three factors. While
this framework is helpful in evaluating changes in the federal budget, it
is not precise and should be viewed as indicative as opposed to
determinative.
Figure 7: Factors Affecting Budget Estimates
Table 1 summarizes the factors-legislative, economic, and technical-that
most significantly resulted in differences between estimated and actual
outlays by fiscal year and account. The factors that were major drivers
for differences between estimated and actual outlays are denoted with "."
Other factors that affected the difference are denoted with "x." In one
case, it was unclear which factors most significantly caused the
difference between estimated and actual outlays. In that case, both
relevant factors are marked with an "x." Detailed explanations supporting
this summary are presented in appendix I.
Table 1: Reasons for Differences between Estimated and Actual Outlays
Account Fiscal Reason for
year differences
Legislative Economic Technical
Commodity Credit Corporation-Corn 2000 x
2001 x x
2002
2003 x
2004 x
Commodity Credit Corporation-Crop 2000
Disaster Assistance 2001
2002
2003
2004
Federal Direct Student Loan Program 2000
Account 2001 x
2002 x
2003 x
2004
Grants to States for Medicaid 2000
2001
2002
2003 x
2004 x
Medicare Part A-Federal Hospital 2000
Insurance Trust Fund 2001
2002
2003 x x
2004 x
Medicare Part B-Federal 2000
Supplementary Medical Insurance 2001
Trust Fund 2002
2003 x x
2004 x
Rail Industry Pension Fund 2000
2001
2002 x
2003 x
2004 x
Unemployment Trust Fund 2000
2001
2002 x
2003 x
2004 x
Total major drivers of differences 19 7 13
Source: GAO analysis of agencies' explanations of differences between
estimated and actual outlays.
Note: Factors that were major drivers for the difference between estimated
and actual outlays are denoted with "." Other factors that affected the
difference are denoted with "x." In one case, it was unclear which factors
most significantly affected the difference. In that case, both relevant
factors are marked with an "x."
Legislation Enacted After Original Estimates Explained Many Differences
between Estimated and Actual Outlays
As seen above in table 1, most of the accounts we reviewed were directly
affected by legislation that was enacted after original estimates were
developed and significantly contributed to differences between expected
and actual outlays in 19 out of 40 instances.32 For example, the Temporary
Extended Unemployment Compensation Act (TEUC) of 2002 led to the
disbursement of greater-than-expected unemployment benefits. Supplemental
appropriations for crop disaster assistance and Agricultural Market
Transition Act payments largely contributed to additional outlays that
were not assumed in original CCC budget projections. Similarly, the MMA
and the Railroad Retirement and Survivors' Improvement Act of 2001,
respectively, increased Medicare outlays and Rail Industry Pension
outlays.
TEUC was enacted to provide up to 13 weeks of federally funded
unemployment insurance benefits to workers in all states who had exhausted
their entitlement to regular state unemployment benefits. Furthermore, the
Act provided up to 13 additional weeks of federally funded benefits to
workers in states with especially high unemployment rates. Congress
renewed this extension in April 2003, which allowed qualified individuals
to file for federal extensions through December 2003 and collect on those
extensions through December 2004. As a result, program outlays exceeded
estimates by $7.9 billion in 2002, $11 billion in 2003, and $4.3 billion
in 2004.
Outlays in both CCC programs that we reviewed also were directly affected
by subsequent legislative action that occurred after original budget
estimates were formulated. For example, Crop Disaster Assistance programs
are funded through supplemental appropriations every year throughout the
5-year period that we reviewed, which led to an additional total of $6
billion in program outlays. According to OMB officials, the Administration
prefers not to include estimates in the budget for relatively
unpredictable disaster-related programs such as crop disaster assistance.
Instead, such funding is typically initiated by Congress through
supplemental appropriations.33 Accordingly, for all 5 years we examined,
no estimates were provided and all of the outlays were as a result of
supplemental appropriations.
Legislative action that increased market loss assistance payments to corn
producers largely contributed to the greatest underestimates of outlays
for that particular commodity-nearly $9 billion in 2000 and 2001 together.
These payments were authorized on an ad hoc basis and, in fiscal year
2000, were paid out for both 1999 and 2000.
The Consolidated Appropriations Resolution of 2003 provided substantially
higher Medicare34 payments to physicians than estimated in original budget
projections and contributed to the largest discrepancy-over $13
billion-between estimated and actual SMI outlays throughout the 5 years
that we reviewed. Furthermore, both the HI and SMI trust funds incurred
unanticipated additional outlays as a result of MMA. Several of the
provisions under MMA were implemented in 2004 and directly affected that
year's outlays; however, officials from the Centers for Medicare &
Medicaid Services (CMS) said that the largest factors that led to
additional HI outlays of approximately $4.4 billion and additional SMI
outlays of nearly $12.3 billion were the substantially increased payments
to private health plans and rural health providers, as well as the
increased physician payment update-all of which were provided for under
MMA.
Finally, the Railroad Retirement and Survivors' Improvement Act of 2001
changed a number of benefit and eligibility criteria, which led to a sharp
rise in retirements. For example, the enactment of this law (1) eliminated
benefit reductions to early retirees, (2) eliminated the maximum threshold
on the amount of combined monthly employee and spouse benefit payments,
(3) lowered the minimum eligibility requirement for railroad retirement
annuities, and (4) increased benefit payments for widow(er)s. Under this
legislation, funds in excess of those needed for current benefit payments
and administrative expenses were transferred to the National Railroad
Retirement Investment Trust. As a result, rail industry retirements
increased, and pension fund outlays increased by almost $20 billion in
2002 and 2003 collectively.
Economic Factors Were Especially Important in Some Programs' Differences
Case study agencies cited economic factors as primary reasons for
differences between estimated and actual outlays for 7 out of 40
differences. This was especially true for agricultural commodities,
student loans, and unemployment insurance. For example, market prices for
commodities affected federal subsidy payments to farmers, changes in
interest rates affected revenues received from student loan borrowers, and
unemployment affected outlays of federal unemployment insurance. Economic
factors also affected the hospital market basket, which contributed to
greater-than-expected Medicare outlays.
For CCC's corn program,35 market prices were both underestimated and
overestimated over the 5-year period. According to a Department of
Agriculture (USDA) official, corn prices are extremely volatile and highly
dependent on weather conditions and global food production. In addition,
the countercyclical design of federal commodity subsidies results in
outlays that are highly sensitive to changes in price. This official
explained that a 1 cent change in estimated corn prices results in about
an $85 million change in federal outlays.
The historically low interest rates that prevailed in recent years were
below levels previously forecasted, which affected estimated student loan
subsidy costs. Subsidy cost estimates for FDLP are highly sensitive to
changes between projected and actual interest rates because borrower
interest rates are variable. The decline in interest rates resulted in
lower-than-expected interest payments to the government from FDLP
borrowers, thus increasing reestimated subsidy costs for these loans.36
Concurrently, the volume of student consolidation loans, which allow
borrowers to lock in fixed interest rates, increased as interest rates
declined. In consolidating their loans, borrowers effectively paid off
their underlying loans, thereby lowering anticipated interest payments to
the government on the loans and, in turn, increasing the estimated subsidy
costs of the underlying loans.
Discrepancies between estimated and actual unemployment insurance outlays
are partially attributed to economic factors such as unanticipated changes
in both the unemployment rate and benefit recipiency rates. For example,
Department of Labor officials said that most of the outlay overestimate in
2000 resulted from a lower-than-expected unemployment rate-the ratio of
the total number of unemployed individuals to the total workforce-which
translated into lower-than-expected outlays. In subsequent years, the
unemployment rate was underestimated and thus contributed to
greater-than-expected outlays. Inaccurate assumptions about the benefit
recipiency rates, that is, the ratio of the total number of unemployed
individuals filing for or receiving benefits to the total number of
unemployed, further contributed to the agency's errors in accurately
estimating program outlays. According to agency officials, these economic
factors tend to be key drivers affecting budget estimates, albeit to a
somewhat lesser degree during the timeframe we reviewed given the
significance of the temporary extended unemployment compensation
legislation that substantially increased outlays in 2002 through 2004.
To a lesser extent, economic factors affected Medicare outlays. In 2003, a
higher-than-expected market basket,37 which is basically a price index
representing the cost of providing health care services to patients, was
part of the explanation behind higher-than-originally-estimated Medicare
outlays, according to CMS officials. This increase in the market basket
led to greater-than-expected inpatient and outpatient hospital
expenditures in the HI and SMI funds respectively.
Technical Factors Explained a Broad Spectrum of Differences
Technical factors, which encompass a somewhat wide-ranging residual
category, significantly explained outlay differences in 13 out of 40
instances. Generally, technical factors account for differences between
budget estimates and actual outlays that cannot be attributed to
legislative or economic factors. For example, delayed implementation and
difficulty in predicting the behavior of providers under new payment
systems, an increased case mix, and the deferral of adjusting payments for
skilled nursing facilities utilization led to differences between
estimated and actual Medicare outlays. Increases in administrative costs
and revised assumptions of the amount of loan defaults and collections
caused some of the direct student loan outlays to differ from original
estimates. Similar to the diversity of the programs we reviewed, there was
great variability among the technical factors that affected account
outlays.
Actual outlays for both Medicare Parts A and B differed from estimates
primarily due to a number of technical factors, which accounted for both
some of the largest and some of the smallest discrepancies. For example,
the largest discrepancy in the HI fund (Medicare Part A) occurred in
fiscal year 2000 for which outlays were lower-than-originally estimated by
nearly $16 billion. According to CMS, the majority of this inaccuracy is
attributed to lower-than-expected benefit payments as a result of the
agency's difficulty in predicting the behavior of providers under newly
implemented payment systems for skilled nursing facility (SNF) services
and home health services. CMS officials said that these payment systems
were very new at the time fiscal year 2000 budget estimates were done and
the effect of these new systems was unknown. Similarly, SMI outlays were
$4.6 billion less-than-originally estimated due largely to the delayed
implementation of and unfamiliarity with a new outpatient hospital
prospective payment system. Other technical factors CMS cited to explain
the differences between estimated and actual Medicare outlays included
case mixes that were more complex than expected and deferred payment
refinements for SNF utilization. Case mix refers to the average complexity
of inpatient admissions for Medicare beneficiaries. A change in the mix of
cases causes the amount of benefit payments to change. The deferral of
payment adjustments for SNF utilization contributed to
greater-than-expected outlays in both fiscal years 2003 and 2004. These
adjustments would have reduced payment rates that had previously been
increased on a temporary basis under the Medicare, Medicaid, and SCHIP
Balanced Budget Refinement Act of 1999 (BBRA).38 CMS included the
budgetary effects of these adjustments in their HI estimates for 2003 and
2004, but later decided not to implement them citing the need for
additional time to review and analyze the implications of implementing
hospital case mix refinements.
Differences between estimated and actual outlays for federal direct
student loans were most frequently explained by technical factors,
including revised assumptions in the Department of Education's loan
subsidy model, increased administrative costs, and Congress's decision not
to adopt a budget proposal to shift administrative expenses to a
discretionary account. Moreover, because of the way federal credit
programs are budgeted, original estimates include a loan subsidy amount
for one fiscal year but actual outlays include a loan subsidy reestimate
for all prior fiscal years-in the case of FDLP, up to 8 years for fiscal
year 2004.
Conclusions
Given that unsustainable federal deficits and debt threaten our future
economy and national security as well as the standard of living for the
American people, renewed emphasis on increasing fiscal discipline is
crucial. Mandatory spending represents an increasing percentage of the
federal budget (e.g., about 54 percent in 2004, up from about 42 percent
in 1984). Unexpected growth in individual programs-especially certain very
large programs-can significantly change the nation's fiscal position. By
identifying significant increases in mandatory spending relatively early
and acting to constrain it, Congress may avert even larger fiscal
challenges in the future.
The notion of establishing budget triggers to constrain growth is not new
and has been used in the past with varying degrees of success. Given that
spending for mandatory programs is driven by underlying benefit and
eligibility formulas, serious efforts to constrain spending would require
substantive changes to current law. Such changes should consider program
goals and objectives and be enacted as programs are created, reexamined,
or reauthorized. While budget triggers certainly are neither a panacea nor
a substitute for deliberate consideration by stakeholders and decision
makers, they can help to prompt action and enhance fiscal responsibility.
Ignoring significant growth in mandatory accounts is inconsistent with
evaluation of programs and their costs. While we appreciate the concerns
raised by budget experts, in our opinion, establishing budget triggers
warrants serious consideration in order to constrain growth in mandatory
spending programs. However, it is clear that how the triggers are designed
must be carefully considered. For example, once widespread agreement on
underlying public policy goals has been achieved, it needs to be decided
whether a soft or hard response to a trigger-or a combination
thereof-would be most appropriate. Also, it is important to consider the
data upon which the trigger will hinge-future projections of historical
data, growth as a percent of GDP, total growth, or another measure
altogether. Moreover, this trigger concept might also be useful in
examining tax expenditure growth. Calculating a normal range of
uncertainty for a program could help avoid triggering an action
prematurely or unnecessarily. In addition, it is important to strike an
appropriate balance between responses that constrain spending or increase
revenues. We recognize that automatic responses pose much more difficult
trade-offs. Ensuring countercyclical effects are not undermined is of
particular importance. In any case, recognizing the natural tension in
balancing both long-term fiscal challenges and other public policy goals,
each program needs to be considered individually to ensure that any
responses triggered strike the appropriate balance between the long-term
fiscal challenge and the program goals. Considering ways to increase
transparency, oversight, and control of mandatory spending programs must
be part of addressing the nation's long-term fiscal challenges.
Matter for Congressional Consideration
To promote explicit scrutiny of significant growth in mandatory accounts,
as mandatory spending programs are created, reexamined, or reauthorized,
Congress should consider incorporating budget triggers that would signal
the need for action. Further, it should determine whether in some cases it
might be appropriate to consider automatically causing some action to be
taken when the trigger is exceeded. Once a trigger is tripped, Congress
could either accept or reject all or a portion of the response to the
spending growth.
Agency Comments
We requested comments on a draft of this report from OMB; the Departments
of Agriculture, Education, Health and Human Services, Labor; and the
Railroad Retirement Board. OMB and the Departments of Education and Labor
had no comments. The Departments of Agriculture, Health and Human
Services, and the Railroad Retirement Board provided clarifying and/or
technical comments, which we incorporated as appropriate.
This report was prepared under the direction of Susan J. Irving, Director,
Federal Budget Analysis, Strategic Issues, who can be reached at (202)
512-9142 or [email protected] . Other key contributors are listed in
appendix IV.
David M. Walker Comptroller General of the United States
Illustrative Examples of Triggers and Responses for Case Study
Accounts Appendix I
Addressing growth in mandatory spending is an important but complicated
matter that requires looking below the aggregate and into specific
programs. Mandatory spending is governed by eligibility rules and benefit
formulas, which means that funds are spent as required to meet the needs
of all those who are eligible and wish to participate. Accordingly,
spending in mandatory programs cannot be constrained through the
application of simple caps/limits. Rather, it requires changes in the
underlying benefit structure and design of programs. As a result,
constraints of individual programs that look at the specific economic and
other factors that drive spending are likely to be most effective.
One idea to constrain growth in mandatory programs is to develop triggers
that, when tripped, would cause some automatic cost-cutting or revenue-
increasing response-such as changes in eligibility criteria, benefit
formulas, or fees-automatically to go into effect unless Congress and the
President act to make other changes. An alternative approach would replace
such a "hard" response with a "soft" one such as requiring special
consideration of a program or a proposal for action when the trigger's
threshhold is breached. Examples of soft responses include raising a point
of order, requiring the administering agency to prepare a special report
explaining why the trigger was breached, or submitting a proposal for
reform. Soft responses may be helpful in alerting decision makers of
potential problems but do not ensure that such action is taken.
Especially in designing hard responses, careful consideration must be
given to avoid counteracting the program's goals and objectives. For
example, a rise in the unemployment rate would by design increase outlays
in federal unemployment insurance not only to provide assistance to the
unemployed but also to stabilize the economy. If a trigger were
established that resulted in a contractionary response, it could undermine
these important goals and exacerbate the effects of unemployment on the
economy.
We selected seven mandatory budget accounts to examine in order to inform
our thinking about budget trigger responses and the design issues that
need to be considered. These seven accounts were selected because of their
relatively large 5-year average differences between estimated and actual
outlays. These accounts are the
(1) Commodity Credit Corporation1
o Corn
o Crop Disaster Assistance,
(2) Federal Direct Student Loan Program Account,
(3) Grants to States for Medicaid,
(4) Medicare Part A: Federal Hospital Insurance Trust Fund,
(5) Medicare Part B: Federal Supplementary Medical Insurance Trust Fund,
(6) Rail Industry Pension Fund, and
(7) Unemployment Trust Fund.
In this appendix, for each case study account we present contextual
information such as the administering agency, program description, and
source of funding. Also we provide the agency's explanation of key
differences between estimated and actual outlays and, as appropriate,
other relevant information. Finally, where appropriate we present
illustrative examples of hard responses that could be established to
constrain spending.2 In some cases these illustrative examples involve
revising currently existing triggers and their corresponding actions. In
other cases new triggers and responses are presented. We do not
specifically advocate any of these approaches as Congress would need to
balance the program and national objectives sought with the long-term
fiscal challenges facing our nation. The approaches we present are for
illustrative purposes only to provide a sense of the types of trigger and
resulting actions that could be established.
Account Name
Commodity Credit Corporation Fund-Corn
Administering Organization
Primarily the Farm Service Agency (FSA), U.S. Department of Agriculture
Program Description
The Commodity Credit Corporation (CCC) is a government-owned and
government-operated entity that was created in 1933 to stabilize, support,
and protect farm income and prices. CCC also helps maintain balanced and
adequate supplies of agricultural commodities and aids in their orderly
distribution.
For fiscal years 2000-2002 (under 1996 Farm Bill provisions), CCC provided
corn-related subsidies primarily through two types of payments available
to supplement farmers' incomes: (1) production flexibility payments to
historical producers of corn and (2) nonrecourse loans, which allow
farmers to store production and use loan proceeds to meet cash flow needs
without selling the crop. Ad hoc legislation provided additional payments
in the form of market loss assistance payments to compensate producers for
low prices.
For fiscal years 2003-2004 (under 2002 Farm Bill provisions), CCC provided
corn-related subsidies through three types of payments available to
supplement farmers' incomes: (1) direct payments to historical producers
of corn; (2) countercyclical payments, which provide a safety net in the
event of low crop prices; and (3) nonrecourse loans.
Funding Source
CCC has an authorized capital stock of $100 million held by the United
States and the authority to have outstanding borrowings of up to $30
billion at any one time. Funds are borrowed from the U.S. Treasury.
Differences between Estimated and Actual Outlays
Based on a 5-year average, estimated outlays for corn differed from actual
outlays by about $1.9 billion per year, or 63.4 percent, in absolute value
terms. However, the actual annual differences varied between an
overestimate of $388 million and an underestimate of $7 billion. Table 2
presents the estimated and actual outlays associated with CCC's corn
program, by fiscal year.
Table 2: Estimated and Actual Corn Outlays, by Fiscal Year
Nominal dollars in millions
Fiscal year Original outlay Actual outlays Difference
estimate
2000 $3,087 $10,136 $-7,049
2001 4,444 6,297 -1,853
2002 3,013 2,959 54
2003 1,803 1,415 388
2004 2,695 2,504 191
5-year average dollar difference $-1,654
5-year average difference as a -55.5%
percent of average estimated
outlays
5-year average dollar difference $1,907 63.4%
(absolute value) 5-year
percentage difference (absolute
value)
Source: GAO analysis of FSA budget data.
Explanation of Key Differences
According to the Farm Service Agency, legislative action and economic
changes were the primary reasons behind differences between estimated and
actual outlays for CCC's corn program during fiscal years 2000 through
2004. In general, weather and natural disasters are the key drivers of
differences between estimated and actual outlays, which are highly
sensitive to changes in the price of corn. Outlays increase when the corn
price decreases. A 1 cent drop in the price of a bushel of corn can lead
to about $85 million increase in countercyclical payments. Participation
also affects costs. Farm program costs depend on market prices and farm
production, which in turn are influenced by world weather, the condition
of the general economy, the foreign and trade policies of the United
States and other food-exporting nations, the rate of inflation, and the
value of the dollar, among other variables. Detailed explanations are
shown in table 3.
Table 3: Explanation of Differences between Estimated and Actual Corn
Outlays
Nominal
dollars in
billions
Fiscal year FSA's explanation
and dollar of differences
differencea Legislative Economic Technical
2000 Additional $5.1 Loan deficiency payments
billion fixed were underestimated by
$-7.0 payments for $1.6 billion due to a
producers of sharp drop in prices
grains and cotton ($0.20 and $0.25 per
were authorized in bushel for 1999 and 2000
Oct. 1999 (Pub. L. projections). Remaining
106-78 S: 802). difference due to lower
loan repayments since
more loans were repaid at
lower rates due to weak
market conditions,
representing marketing
loan gains for producers.
2001 $2.1 billion Underpayment was A small drop in
Market Loss moderated by a small net loan
$-1.9 Assistance reduction in loan expenditures, as
payments were deficiency payments less corn was
authorized in Aug. resulting from a slight placed under loan
2001 (Pub. L. rise in the average than projected.
107-25 S: 1). market price and a change
in the seasonal pattern
from projections.
2002 Three cent per bushel
drop in the 2001 price of
$0.05 corn, which triggered
higher loan deficiency
payments (LDP). As more
producers opted for LDPs,
fewer placed corn under
loan, thus reducing net
outlays.
2003 Mandated policy Increase in net loan
change saved $1.9 outlays, reflecting a
$0.39 billion by change in the loan rates.
eliminating The rate for 2002 and
production 2003 corn was assumed at
flexibility $1.67 per bushel under
contract payments previous legislation but
unless requested increased to $1.98 under
by producers who 2002 legislation. Thus,
are parties to the the face value of loans
contract (Pub. L. made went up.
107-171 S: 1107).
This was partially
offset by $1.4
billion for the
new direct payment
program.
2004 An increase in prices The CCP decline
pushed corn above the outweighed
$0.19 countercyclical payment increases of $130
(CCP) trigger level, thus million for
reducing CCPs by $397 direct payments,
million. The CCP decline which were due to
outweighed increases of higher base acres
$100 million for LDPs, than were assumed
which reflect a decline before the actual
in the 2004 crop price, sign-up.
raising the LDP rate and
quantity.
Source: Farm Service Agency.
Notes: Primary drivers of outlay differences are marked in bold.
aA negative difference means that actual outlays were higher than
originally estimated. A positive difference means that actual outlays were
less than originally estimated.
Illustrative Triggers and Response
The 2002 Farm Bill3 guaranteed historical producers of corn and other
commodities a minimum price per bushel, known as a target price, which
they can expect to earn. To constrain spending, one possible trigger could
be when the target price exceeds the market price by some historically
average percentage, the legislated target price could be reduced. However,
to avoid price shocks to the industry and possible procyclical effects,
the price reduction could be deferred to the following year.
The Farm Bill also established a formula for fixed, direct payments to
historical producers of corn and other commodities. To limit spending on
this income-support program, one idea for a trigger could be to link
direct payments to farm sector production prices. For example, if
production prices drop by more than 3 percent,4 Congress could redefine
the formula to be less generous.
Alternatively, Congress could limit the guarantee of direct payments to
current producers of corn rather than historical producers.
Account Name
Commodity Credit Corporation Fund-Crop Disaster Assistance
Administering Organization
Primarily the Farm Service Agency (FSA), U.S. Department of Agriculture
Program Description
CCC is a government-owned and government-operated entity that was created
in 1933 to stabilize, support, and protect farm income and prices. CCC
also helps maintain balanced and adequate supplies of agricultural
commodities and aids in their orderly distribution.
Crop Disaster Assistance programs reimburse producers for qualifying
losses to agricultural commodities (other than sugar cane or cotton seed)
due to damaging weather or related conditions. The damages must be in
excess of 35 percent of the established price of crops for lost production
or 20 percent for lost quality. Crop disaster programs cover insured,
uninsured, and noninsurable crops. The program has no set funding
limitation, however, payments are limited to $80,000 per person, and
producers with incomes greater than $2.5 million are ineligible. This crop
disaster assistance program is not permanently authorized.
Funding Source
CCC has an authorized capital stock of $100 million held by the United
States and the authority to have outstanding borrowings of up to $30
billion at any one time. Funds are borrowed from the U.S. Treasury.
Although Crop Disaster Assistance programs are provided through
appropriations acts,5 the Department of Agriculture considers and applies
funding for the programs in a manner similar to mandatory programs.
According to an FSA official, funding is provided to all eligible
applications for assistance by prorating available funding if necessary.
The Office of Management and Budget (OMB) also considers crop disaster
assistance programs to be mandatory in that all eligible applicants may
receive benefits.
Differences between Estimated and Actual Outlays
Based on a 5-year average, estimated outlays for crop disaster assistance
differed from actual outlays by about $1.2 billion per year in absolute
value terms. However, the actual annual differences varied between $230
million and $1.9 billion. Table 4 presents the estimated and actual
outlays associated with CCC's crop disaster assistance programs, by fiscal
year.
Table 4: Estimated and Actual Crop Disaster Assistance Outlays, by Fiscal
Year
Nominal dollars in millions
Fiscal year Original outlay Actual outlays Difference
estimate
2000 $0 $1,251 $-1,251
2001 0 1,848 -1,848
2002 0 230 -230
2003 0 1,867 -1,867
2004 0 804 -804
5-year average dollar difference $-1,200
5-year average difference as a N/A
percent of average estimated
outlays
5-year average dollar difference $1,200 N/A
(absolute value) 5-year
percentage difference (absolute
value)
Source: GAO analysis of FSA budget data.
Explanation of Key Differences
According to OMB staff, it is not OMB's policy to include an estimate for
disaster assistance in the President's budget. Instead, these programs are
typically funded through subsequent legislation.
Ideas for Improving the Accuracy of Estimates
Although OMB typically does not include an estimate for crop disaster
assistance in the President's budget, we have reported in the past that
shifting the budget timing to an up-front recognition of emergency costs
through reserves may promote a more comprehensive and transparent debate
over federal budgetary priorities during the regular budget process.6 For
example, we suggested that federal governmentwide emergency reserves could
set aside budget authority in advance for expected yet unpredictable
events as part of the annual resource-allocation process. Another approach
would be to establish agency-specific reserve funds for those agencies
that regularly respond to federal emergencies. Funds would be appropriated
to these agencies on a contingent basis, meaning that certain
agency-specific criteria would have to be met before the funds could be
used. While these approaches are not of the trigger/response variety that
is the subject of this report, they would help accomplish a goal of
constraining spending if the emergency budget authority provided in
advance is assumed to be within a constrained total budget authority.
Account Name
Federal Direct Student Loan Program Account
Administering Organization
Office of Federal Student Aid, U.S. Department of Education
Program Description
The Department of Education (Education) provides financial aid in part to
increase access to college. Education's first direct loans were made in
the fourth quarter of fiscal year 1994.7 Through its William D. Ford
Federal Direct Loan Program (FDLP), students and/or their parents borrow
money directly from the federal government through the vocational,
undergraduate, or graduate schools the students attend. As is the case
under the Federal Family Education Loan Program (FFELP), or "guaranteed"
student loan program, there are four types of direct loans.8
Stafford Loans-variable rate loans available to students. The federal
government pays the interest on behalf of borrowers while the student is
in school, during a 6-month grace period when the student first leaves
school, and during statutory deferment periods related to borrower
unemployment and economic hardship.
Unsubsidized Stafford Loans-variable rate loans to students with the same
terms as Stafford Loans except that the government does not pay interest
costs during in-school, grace, and deferment periods.
PLUS Loans-variable rate loans made to parents. The borrower pays all
interest costs.
Consolidation Loans-borrowers may combine multiple federal student loans
into a single, fixed rate loan. The interest rate is based on the weighted
average of the interest rates in effect on the loans being consolidated or
a fixed percentage.
Funding Source
Education finances FDLP through a combination of appropriations and
borrowing from Treasury. Education receives permanent, indefinite budget
authority for estimated subsidy costs-the amount expected not to be repaid
by borrowers-of its loans. These costs are generally updated, or
reestimated, annually. The portion of direct loans that Education predicts
will ultimately be repaid by borrowers is financed by borrowing from
Treasury and is not considered a cost to the government because it is
expected to be returned to the government in future years.
Differences between Estimated and Actual Outlays
Based on a 5-year average, estimated outlays for direct student loans
differed from actual outlays by about $2.6 billion per year, or 702
percent, in absolute value terms. However, the actual annual differences
varied between an overestimate of $2.8 billion and an underestimate of
$5.3 billion. A large component of these differences reflects the fact
that initial estimates do not include reestimates of prior year costs,
which are reflected in actual outlays. In addition, initial estimates
reflect proposed policies, many of which were not enacted and so were not
reflected in subsequent actual outlays. Table 5 presents the estimated and
actual outlays associated with the federal direct student loan program, by
fiscal year.
Table 5: Estimated and Actual Direct Student Loan Outlays, by Fiscal Year
Nominal dollars in millions
Fiscal year Original outlay Actual outlays Difference
estimate
2000 $-42 $-2,862 $2,820
2001 115 257 -142
2002 -635 97 -732
2003 -283 5,055 -5,338
2004 -786 3,246 -4,032
5-year average dollar difference $-1,485
5-year average difference as a 455.2%
percent of average estimated
outlays
5-year average dollar difference $2,613 702%
(absolute value) 5-year
percentage difference (absolute
value)
Source: GAO analysis of President's budget data.
Note: A negative outlay amount indicates a positive collection of revenue.
Also, for credit programs, the term actual is misleading because
reestimates will continue until all the loans in that cohort have been
repaid.
Explanation of Key Differences
Because FDLP is a relatively new program, it has a short history of
repayment activity and little historical data are available. Accordingly,
Education initially relied heavily on data from the guaranteed student
loan program to develop estimates for most key cash flow assumptions in
its FDLP cash flow model, which is used to estimate the subsidy cost of
the program. Over the past few years, Education has incorporated FDLP data
into many cash flow assumptions; as more data become available, Education
plans to completely phase out the use of guaranteed loan data for FDLP
assumptions.9
Drops in interest rates have been a key driver behind differences in
estimated versus actual outlays. Not only are loans being paid off at
lower rates than anticipated but the drop in rates has also led to a
dramatic increase in consolidations (which are prepayments). Detailed
explanations are shown in table 6.
Table 6: Explanation of Differences between Estimated and Actual Direct
Student Loan Outlays
Nominal dollars
in billions
Education's
explanation
of
differences
Fiscal year and Legislative Economic Technical
dollar
differencesa
2000 Actual includes about $2.4
billion in prior year
$2.8 reestimates of loans made
in fiscal years (FY) 1994
through 1999. Net downward
reestimate of prior
cohorts primarily due to
revised assumptions about
a drop in defaults and an
increase in collections.
FY 2000 cohort subsidy
decreased $442 million.
Administrative costs
decreased $30 million.
2001 Changes in interest Actual includes about $481
rates resulted in million in prior year
$-0.14 $481 million upward reestimates of loans made
reestimate of prior in FYs 1994 through 2000.
year cohorts. FY 2001 cohort subsidy
decreased $432 million.
Administrative costs
increased $94 million
2002 Drop in interest Administrative costs
rates caused FY increased $42 million.
$-0.73 2002 cohort subsidy
to increase $694
million due to
lower projected
borrower
repayments.
2003 Revised assumptions Actual includes about $4.6
on interest rates, billion for 2-years worth
$-5.3 prepayments through of prior-year reestimates
consolidations, and of loans made in FYs 1994
defaults resulted through 2002. (No
in upward reestimate was executed in
reestimate of prior FY 2002.) $15 million
year cohorts of policy proposal to shift
$4.6 billion. FY administrative expenses to
2003 cohort subsidy a discretionary account
increased $250 not enacted by Congress.
million.
2004 $710 million policy
proposal to shift
$-4.0 administrative expenses to
a discretionary account
not enacted by Congress.
About $2.6 billion in
upward reestimates of
prior year cohorts (loans
made during FYs 1994
through 2003) reflects
technical changes to model
assumptions, including
higher level of
prepayments, which lower
future interest income,
and higher defaults for
borrowers choosing
income-contingent loan
repayment.
Source: Department of Education.
Notes: Primary drivers of outlay differences are marked in bold.
aA negative difference means that actual outlays were higher than
originally estimated. A positive difference means that actual outlays were
less than originally estimated.
Illustrative Trigger and Response
Congress could decrease the subsidy cost to the government by, among other
things, increasing the amount of fees borrowers must pay to obtain a loan
or increasing borrowers' interest rate. For example, continued differences
between estimated and actual outlays could be used as a trigger, resulting
in higher origination fees or interest rates for new FDLP loans. In
implementing such a trigger and response, Congress would need to consider
whether FFELP borrowers should similarly be affected. Under current law,
loans made to borrowers, unless otherwise specified, are to have the same
terms, conditions, and benefits and be made available in the same amounts
under both FDLP and FFELP.10
Account Name
Grants to States for Medicaid
Administering Organization
Centers for Medicare & Medicaid Services (CMS), U.S. Department of Health
and Human Services
Program Description
Medicaid is a health-financing program for eligible low-income individuals
and families. Federal statute defines over 50 population groups that are
potentially eligible for states' programs. In general, eligibility is
limited to low-income children, pregnant women, parents of dependent
children, people with disabilities, and the elderly. Although Medicaid is
one federal program, it consists of 56 distinct state-level programs-one
for each state, territory, Puerto Rico, and the District of Columbia.11
Each of the states has a designated Medicaid agency that administers the
program. In accordance with the Medicaid statute and within broad federal
guidelines, each state establishes its own eligibility standards;
determines the type, amount, duration, and scope of covered services; sets
payment rates; and develops its administrative structure.
Funding Source
The federal government matches state Medicaid spending for medical
assistance according to a formula that compares each state's average per
capita income-a proxy reflecting the health of the state's economy and its
response to economic changes-to the national per capita income. Therefore,
states with a high per capita income receive less federal funds than
states with a low per capita income. As economic conditions improve or
decline in a particular state, so does the amount of federal matching
funds granted to that state. The federal share, known as the Federal
Medical Assistance Percentage (FMAP), can range from 50 to 83 percent.
States are required to describe the nature and scope of their programs in
comprehensive written plans submitted to CMS-with federal funding for
state Medicaid services contingent upon CMS approval of the plans. This
approval hinges on whether CMS determines that state plans meet all
applicable federal laws and regulations.
Although the source of Medicaid funding is through an annual appropriation
act, Medicaid is not considered a discretionary spending program. Because
Medicaid is an entitlement created by the operation of law, if Congress
fails to appropriate money necessary to fund payments and benefits,
eligible recipients may seek legal recourse. In such case, necessary
payments may be made through the indefinite judgment fund pursuant to 31
U.S.C. S: 1304.
Differences between Estimated and Actual Outlays
Based on a 5-year average, estimated Medicaid outlays differed from actual
outlays by about $4.2 billion per year, or 2.9 percent, in absolute value
terms. Actual annual differences ranged from an underestimate of $5.1
billion to an overestimate of $6.3 billion. Table 7 presents the estimated
and actual Medicaid outlays for fiscal years 2000 through 2004.
Table 7: Estimated and Actual Medicaid Outlays, by Fiscal Year
Nominal dollars in millions
Fiscal year Original outlay Actual outlays Difference
estimate
2000 $114,660 $117,921 $-3,261
2001 124,838 129,374 -4,536
2002 142,423 147,512 -5,089
2003 158,790 160,693 -1,903
2004 182,543 176,231 6,312
5-year average dollar difference $-1,695
5-year average difference as a -1.2%
percent of average estimated
outlays
5-year average dollar difference $4,220 2.9%
(absolute value) 5-year
percentage difference (absolute
value)
Source: GAO analysis of President's budget data.
Explanation of Key Differences
Both legislative and technical factors led to differences in estimated and
actual Medicaid outlays. For example, the Jobs and Growth Tax Relief
Reconciliation Act of 2003,12 which temporarily changed federal matching
rates for benefits and provided fiscal relief to states, affected
estimated Medicaid outlays in both fiscal years 2003 and 2004. Technical
factors included misestimates of medical assistance payments,
administrative costs, vaccines for children, and collections. Also, there
were a number of legislative proposals that were not adopted. It is not
clear if economic factors also contributed to the differences, although it
is likely so given the economic downturn that occurred during this time
period. Changing economic conditions could have led to differences in the
number of individuals eligible for and receiving benefits, and therefore
total program outlays.
According to CMS officials, Medicaid estimates are based primarily on
state estimates and may be adjusted by CMS' Office of the Actuary to
reflect recent trends in how state estimates have changed over time or how
they have compared with actual expenditures in recent years. Agency
officials were unable to accurately identify and quantify the effects of
any of these factors and explained that the difficulty lies with the
variability of program structure across states. Each state is allowed the
discretion to structure and modify its program, including the
establishment of eligibility criteria and payment rates. Similarly, state
legislative actions and economic conditions vary across the country and
could have varying effects on program outlays. Consequently, aggregating
state data into a single Medicaid figure would mask estimating
inaccuracies and challenges since the negative effect in one might be
offset by positive effect in another. In the event that the difference
between estimated and actual spending is very large, CMS said it would
then investigate and seek explanations from the states. Although CMS did
not consider the differences evident throughout the 5-year period we
reviewed to be large enough to prompt such an evaluation, they did provide
some explanation behind misestimates as shown in table 8 below.
Table 8: Explanation of Differences between Estimated and Actual Medicaid
Outlays
Nominal dollars
in billions
CMS' explanation of
differences
Fiscal year and Legislative Economic Technical
dollar
differencea
2000 Underestimated medical
assistance payments by
-$3.3 over $3 billion,
administrative costs by
$71 million, and vaccines
for children by $2
million. The resulting
outlay underestimate was
further increased by a
legislative proposal
expected to save $161
million that had been
included in the original
estimate but was not
accepted.
2001 Underestimated medical
assistance payments by
-$4.6 $7.1 billion and vaccines
for children by $357
million. These
underestimates were
partially offset by a $977
million overestimate of
administrative costs, an
underestimate of almost
$1.3 billion in
collections, and a
legislative proposal
expected to cost $663
million that had been
included in the original
estimate but was not
accepted.
2002 Underestimated medical
assistance payments by
-$5.1 $5.3 billion. This
underestimate was
partially offset by a $138
million underestimate of
collections and
overestimates of
administrative costs and
vaccines for children by
$722 million and $4
million respectively. The
resulting outlay
underestimate was further
increased by a legislative
proposal expected to save
$606 million that had been
included in the original
estimate but was not
accepted.
2003 The Jobs and Growth Underestimated medical
Tax Relief assistance payments by
-$1.9 Reconciliation Act of nearly $3 billion and
2003 was enacted in vaccines for children by
late May 2003, after $241 million. These
original FY 2003 underestimates were
estimates were made. partially offset by a $1.1
CMS said this billion overestimate of
legislation accounted administrative costs, a
for approximately $4 $112 million underestimate
billion in of collections, and a
unanticipated outlays. legislative proposal
expected to cost $98
million that had been
included in the original
estimate but was not
accepted.
2004 The Jobs and Growth Outlays were overestimated
Tax Relief as a result of a
$6.3 Reconciliation Act of legislative proposal
2003 was enacted in expected to cost $5.8
late May 2003, after billion, which had been
original FY 2004 included in the original
estimates were made. estimate but was not
CMS said this adopted, a $1 billion
legislation accounted overestimate of
for approximately $6 administrative costs, and
billion in a $168 million
unanticipated outlays. underestimate of
collections. The resulting
outlay overestimate was
partially offset by
underestimates of medical
assistance payments and
vaccines for children by
$520 million and $133
million respectively.
Source: Centers for Medicare & Medicaid Services.
Notes: Primary drivers of outlay differences are marked in bold.
aA negative difference means that actual outlays were higher than
originally estimated. A positive difference means that actual outlays were
less than originally estimated.
Account Name
Federal Hospital Insurance (HI) Trust Fund (Medicare Part A)
Administering Organization
Centers for Medicare & Medicaid Services, U.S. Department of Health and
Human Services
Program Description
The account funds the Medicare Part A program which partially covers the
costs of, among other things, home health care, inpatient care in
hospitals and skilled nursing facilities, and hospice care. Based on their
work history, most U.S. citizens and permanent residents and their spouses
are eligible for Medicare Part A if they are 65 years of age or older.
Also, certain persons under 65 years old who are disabled or have
end-stage renal disease are eligible for coverage. Enrollees or their
spouses who have contributed to Medicare through payroll taxes for at
least 10 years of employment are automatically enrolled at age 65 and need
not pay premiums to receive coverage. Individuals who have not met this
eligibility requirement may pay a monthly premium to purchase Part A
coverage.13
Funding Source
The primary funding source for Medicare Part A comes from payroll taxes.
Other relevant revenue sources include interest on investments in
government securities held by the fund,14 income from taxation of Old Age,
Survivors, and Disability Insurance (Social Security) benefits, and
premiums collected from voluntary participants.
Differences between Estimated and Actual Outlays
Based on a 5-year average, estimated Medicare Part A outlays differed from
actual outlays by about $5.6 billion per year, or 3.8 percent, in absolute
value terms. Actual annual differences ranged from an underestimate of
$4.3 billion to an overestimate of $15.9 billion. Table 9 presents the
estimated and actual HI outlays for fiscal years 2000 through 2004.
Table 9: Estimated and Actual HI Outlays, by Fiscal Year
Nominal dollars in millions
Fiscal year Original outlay Actual outlays Difference
estimate
2000 $143,898 $127,973 $15,925
2001 143,427 140,573 2,854
2002 144,674 145,606 -932
2003 147,295 151,308 -4,013
2004 159,750 164,136 -4,386
5-year average dollar difference $1,890
5-year average difference as a 1.3%
percent of average estimated
outlays
5-year average dollar difference $5,622 3.8%
(absolute value) 5-year
percentage difference (absolute
value)
Source: GAO analysis of President's budget data.
Explanation of Key Differences
Both legislative and technical factors led to differences between
estimated and actual HI outlays. For example, the Medicare Prescription
Drug, Improvement, and Modernization Act of 2003 (MMA)15 led to
greater-than-expected outlays in fiscal year 2004. Technical factors
included difficulty in predicting the behavior of providers under new
payment systems, misestimates of home health transfers to and from the
Supplementary Medical Insurance (SMI) Trust Fund, and misestimates of
service usage. Economic factors, specifically the hospital market basket,
also contributed to differences. The hospital market basket is an input
price index that represents the cost of the mix of goods and services that
comprise routine, ancillary, and special-care unit inpatient hospital
services. Detailed explanations of the differences are shown in table 10.
Table 10: Explanation of Differences between Estimated and Actual HI
Outlays
Nominal
dollars in
billions
CMS'
explanation of
differences
Fiscal year Legislative Economic Technical
and dollar
differencea
2000 Overestimated benefit
payments by $10.1
$15.9 billion, which CMS
attributed to its
difficulty in predicting
the behavior of providers
under new payment systems
for skilled nursing
facility (SNF) services
and home health services.
Also, a $6.6 billion
overestimate of home
health transfers to the
SMI fundb further widened
the gap between estimated
and actual outlays. CMS
attributed this to
discrepancies in
implementing the payment
system, particularly by
home health agencies.
Specifically, the cap on
average per-beneficiary
home health expenditures
was treated as an
absolute cap, thereby
cutting services to
patients requiring
numerous visits per
episode of care. This
resulted in unexpected
additional savings. The
resulting outlay
overestimate was
partially offset by a
legislative proposal
expected to save $808
million that had been
included in the original
estimate but was not
accepted.
2001 Overestimated benefit
payments by almost $3.9
$2.9 billion, which CMS
attributed to the
discrepancy between what
they assumed service
usage to be and actual
usage. Overestimates of
$242 million in home
health transfers to the
SMI fund and $84 million
in quality improvement
organizations (QIO)c
further added to the
difference. The resulting
outlay overestimate was
partially offset by a
$1.2 billion quinquennial
adjustment as required by
lawd and a legislative
proposal expected to save
$185 million that had
been included in the
original estimate but was
not adopted.
2002 Underestimated benefit
payments by $2.3 billion,
-$0.9 which CMS attributed to
the discrepancy between
what they assumed service
usage to be and actual
usage. The resulting
outlay underestimate was
partially offset by
overestimates of $1.3
billion and $77 million
in home health transfers
to the SMI fund and QIOs
respectively.
2003 Inpatient hospital Inpatient hospital
expenditures were expenditures were higher
-$4.0 higher than expected than expected also
due to a because of a hospital
higher-than-expected case mix increase. Case
"market basket" mix refers to the average
payment update. complexity of inpatient
Market basket refers admissions for Medicare
to the input price beneficiaries. Payments
index based on the are based on the type of
cost of a particular case, so if the mix of
type of health cases changes, payments
provider (e.g., also change. Expenditures
hospital, skilled were also higher because
nursing facility, expected SNF resource
home health agency) utilization group (RUG)
to provide services refinements were not
to patients. By law made, which would have
these indexes are reduced payments. SNFs
used to update received higher payments
Medicare payments. due to the introduction
of a new administrative
policy to adjust payment
updates for past
differences between
actual and estimated
market basket increases.
The $7.3 billion
underestimate in benefit
payments was further
increased by an $18
million underestimate of
administrative costs, but
was partially offset by
overestimates of $2.8
billion in home health
transfers to the SMI
fund, $43 million in
QIOs, and a legislative
proposal expected to cost
$410 million that was
included in the original
estimate but was not
adopted.
2004 Underestimated Once again, expected SNF
benefit RUG refinements were not
-$4.4 payments by made, resulting in
$4.3 billion. higher-than-estimated
CMS attributed expenditures. Hospice
this to several expenditures were also
MMA provisions higher than estimated.
that were The $4.3 billion
enacted and underestimate in benefit
implemented payments was further
after original increased by
estimates were underestimates of $23
made, which led million in QIOs and $17
to higher million in administrative
actual costs.
expenditures.
In particular,
payments to
private health
plans
contracting
with Medicare
were increased
substantially
as were
payments to
rural health
providers.
Source: Centers for Medicare & Medicaid Services.
Notes: Primary drivers of outlay differences are marked in bold. In 2003
it was unclear which factor most significantly affected the difference. In
that case, there is no primary driver marked in bold.
aA negative difference means that actual outlays were higher than
originally estimated. A positive difference means that actual outlays were
less than originally estimated.
bHome health agency transfers occur between the HI and SMI trust funds and
total billions of dollars throughout the 5-year period. However, the
positive variance in one fund is equally offset by the negative variance
in the other. As a result, when the Medicare trust funds are taken
together, this intertrust fund activity has no cumulative impact on the
federal surplus/deficit.
cQuality improvement organizations are groups of practicing doctors and
other health care experts paid by the federal government to check and
improve the care given to Medicare patients.
dSection 217(g) of the Social Security Act provides for periodic transfers
between the general fund of the Treasury and the HI trust fund, if needed
to adjust prior payments for the costs arising from wage credits granted
for military service before 1957.
Currently Existing Program Trigger and Response
MMA established a trigger with a soft response to constrain growth in
Medicare; it requires the President to submit a proposal to Congress for
action if the Medicare Trustees determine in 2 consecutive years that the
general revenue share16 of Medicare is projected to exceed 45 percent
during a 7-year projection period. To date, this threshold has not been
breached and thus no response has been triggered. According to the 2005
Medicare Trustees' report, the trigger is expected to be breached in 2012,
which falls within the 7-year projection period that will be covered in
the 2006 Medicare Trustees' report. If the 45 percent threshold is
projected to be breached again in the next consecutive 7-year projection
period, the President will be required to propose legislation, within 15
days of submitting the fiscal year 2009 budget, to respond to the funding
warning.
Illustrative Trigger and Response
Using the trigger of general revenue exceeding 45 percent in 2 consecutive
years during a 7-year period, hard responses could also be developed.
Possible responses are to adjust taxes, benefit formulas, or eligibility
criteria. For example, Medicare payroll taxes could automatically be
increased unless Congress took action to prevent the increase.
Alternatively, reaching the trigger could cause automatic changes to
benefit formulas or eligibility criteria, or a combination of benefit
changes and tax increases. Of course congressional action could change the
automatic response if it was deemed inappropriate at that time.
Account Name
Federal Supplementary Medical Insurance (SMI) TrustFund (Medicare Part B)
Administering Organization
Centers for Medicare & Medicaid Services, U.S. Department of Health and
Human Services
Program Description
This account, also known as Medicare Part B, partially covers the cost of
doctors' services, clinical laboratory services, outpatient hospital
services, some physical and occupational therapy services, and some home
health care. Eligibility requirements for Medicare Part B are similar to
those for Part A. However, unlike for Medicare Part A, enrollment is
voluntary. Enrollees must pay a monthly premium to receive Part B
coverage. In 2005, premiums were $78.20 per month and the deductible was
$110. Premium and deductible rates may change every year.
Most Part B services are paid based on a fee schedule. Physicians, the
largest Part B service type, are paid under the sustainable growth rate
(SGR) system,17 which determines the increase in payments per service for
the physician fee schedule for each year based on a statutory formula.
Under the SGR system, actual physician-related spending is compared with
target physician-related spending levels. If actual spending exceeds
target spending, then future physician fee schedule updates are reduced.
Funding Source
SMI is financed from general revenues (approximately 75 percent) and
beneficiary premiums (approximately 25 percent).
Differences between Estimated and Actual Outlays
Based on a 5-year average, estimated Medicare Part B outlays differed from
actual outlays by about $6.4 billion per year, or 6.1 percent, in absolute
value terms. Actual annual differences ranged from an underestimate of
$13.4 billion to an overestimate of $4.6 billion. Table 11 presents the
estimated and actual SMI outlays by fiscal year.
Table 11: Estimated and Actual SMI Outlays, by Fiscal Year
Nominal dollars in billions
Fiscal year Original outlay Actual outlays Difference
estimate
2000 $91,795 $87,216 $4,579
2001 96,372 97,531 -1,159
2002 107,830 107,113 717
2003 108,416 121,816 -13,400
2004 119,353 131,632 -12,279
5-year average dollar difference $-4,308
5-year average difference as a -4.1%
percent of average estimated
outlays
5-year average dollar difference $6,427 6.1%
(absolute value) 5-year
percentage difference (absolute
value)
Source: GAO analysis of President's budget data.
Congress has overridden the statutory updates for the 2003, 2004, and 2005
physician fee schedules. Although the SGR system called for negative
updates in these years, Congress instead granted increases in physician
payments per service. For several years the law was changed to specify
higher spending for physicians after the budget estimates were already
done. Consequently, this contributed to actual outlays that were higher
than estimated.
Explanation of Key Differences
Both legislative and technical factors led to differences between
estimated and actual SMI outlays. For example, the Consolidated
Appropriations Resolution of 2003 and MMA led to greater-than-expected
outlays for spending for physicians' services. Technical factors included
delayed implementation and difficulty in predicting the behavior of
providers under a new outpatient hospital prospective payment system,
misestimates of home health transfers to and from the HI fund, and
misestimates of service usage. Similar to the HI fund, changes in the
hospital market basket also contributed to differences. Detailed
explanations of the differences are shown in table 12.
Table 12: Explanation of Differences between Estimated and Actual SMI
Outlays
Nominal
dollars in
billions
CMS' explanation
of differences
Fiscal year Legislative Economic Technical
and dollar
differencea
2000 Benefit payments were
$11.6 billion lower
$4.6 than expected, which
CMS attributed to the
delayed implementation
and difficulty in
predicting the behavior
of providers under a
new outpatient hospital
prospective payment
system. The system was
being created from
scratch and little
research had been done
on what type of system
would work best.
Original projections
had an earlier start
date, which caused
higher expenditures to
be estimated. This
overestimate was
further increased by
overestimates of $90
million and $17 million
in transfers to
Medicaid and QIOs
respectively. The
resulting overestimate
was partially offset by
a $6.6 billion
overestimate of home
health transfers
received from the HI
fund and a legislative
proposal expected to
save $570 million that
had been included in
the original estimate
but was not adopted.
2001 Most of the difference
is attributed to a
-$1.2 legislative proposal
expected to save $685
million that was not
adopted. Benefit
payments were $241
million greater than
expected, which CMS
attributed to the
discrepancy between
what they assumed
service usage to be and
actual usage. In
addition, home health
transfers received from
the HI fund were $242
million less than
expected.
2002 Most of the difference
resulted from a $2.1
$0.7 billion overestimate of
benefit payments, which
CMS attributed to the
discrepancy between
what they assumed
service usage to be and
actual usage. This
overestimate was
partially offset by a
$1.3 billion
overestimate of home
health transfers
received from the HI
fund and a $42 million
underestimate of
transfers to Medicaid.
2003 Consolidated A higher hospital The underestimate of
Appropriations market basket (as benefit payments was
-$13.4 Resolution 2003 mentioned in the further increased by a
was passed after Medicare Part A $112 million
the original section) also caused underestimate of
budget estimates, higher-than-expected transfers to Medicaid,
which caused a outpatient hospital but was partially
substantially expenditures. offset by a legislative
higher physician proposal expected to
fee update than cost $70 million that
estimated in the was not adopted and a
projections. $16 million
overestimate in QIOs.
2004 As noted in the The underestimate of
Medicare Part A benefit payments was
-$12.3 section, MMA was further increased by a
passed in 2003 $168 million
and some of its underestimate of
provisions were transfers to Medicaid,
implemented in but was partially
2004, causing offset by a legislative
higher private proposal expected to
plan and rural cost $55 million that
provider was included in the
expenditures. original estimate but
Also, the not adopted.
physician fee
update was much
higher than had
been originally
estimated due to
the MMA
legislation that
was enacted after
the 2004 budget
estimates were
made.
Consequently,
benefit payments
were $12.2
billion greater
than expected.
Source: Centers for Medicare & Medicaid Services.
Notes: Primary drivers of outlay differences are marked in bold.
aA negative difference means that actual outlays were higher than
originally estimated. A positive difference means that actual outlays were
less than originally estimated.
Currently Existing Program Triggers and Responses
Congress has established two triggers with soft and hard responses to
constrain growth in SMI. First, under the SGR system, if actual
physician-related spending exceeds target physician-related spending then
future physician fee schedule updates are reduced. Because the actual
versus target spending comparison is cumulative, future fee updates are
reduced to lower future actual spending below future target spending until
total cumulative actual spending is the same as total cumulative target
spending. Although the SGR system was designed to encourage fiscal
discipline, Congress has chosen to modify or override this constraint a
number of times. We have previously reported on concerns about the SGR
system and
considerations for reform.18 Second, MMA established a trigger with a soft
response; it requires the President to submit a proposal to Congress for
action if the Medicare Trustees determine in 2 consecutive years that the
general revenue share of Medicare is projected to exceed 45 percent during
a 7-year projection period. To date, this threshold has not been breached
and thus no response has been triggered. As mentioned in the Medicare Part
A section of this appendix, the trigger is expected to be breached in
2012, which falls within the specified 7-year projection period that will
be covered in the 2006 Medicare Trustees' report. If the 45 percent
threshold is projected to be breached again in the next consecutive 7-year
projection period, the President will be required to propose legislation,
within 15 days of submitting the fiscal year 2009 budget, to respond to
the funding warning.
Illustrative Trigger and Response
Using the trigger of general revenue exceeding 45 percent in 2 consecutive
years during a 7-year period, hard responses could also be developed.
Possible responses are to adjust premiums,19 benefit formulas, or
eligibility criteria. For example, Part B premiums could automatically be
increased unless Congress took action to prevent the increase.
Alternatively, reaching the trigger could cause automatic changes to
benefit formulas or eligibility criteria, or a combination of benefit
changes and premium increases. Of course congressional action could change
the automatic response if it was deemed inappropriate at that time.
Account Name
Rail Industry Pension Fund
Administering Organization
Railroad Retirement Board (RRB)
Program Description
The RRB administers a Federal retirement-survivor benefit program for the
nation's railroad workers and their families, under the Railroad
Retirement Act. In connection with this retirement program, the RRB has
administrative responsibilities under the Social Security Act for certain
benefit payments and railroad workers' Medicare coverage.
Under the Railroad Retirement Act, retirement and disability annuities are
paid to railroad workers with at least 10 years of service, or 5 years if
performed after 1995. Annuities are also payable to spouses and divorced
spouses of retired workers and to widow(er)s, surviving divorced spouses,
remarried widow(er)s, children, and parents of deceased railroad workers.
Qualified railroad retirement beneficiaries are covered by Medicare in the
same way as Social Security beneficiaries.
Railroad retirement benefits are calculated under a two-tier formula. Tier
I is based on combined railroad retirement and Social Security credits,
using Social Security benefit formulas. Tier II is based on railroad
service only and is similar to the defined benefit pensions paid
over-and-above Social Security benefits in other industries. In addition,
some annuitants may also be qualified for supplemental benefits and vested
dual benefits. Cost-of-living adjustments on the Tier I portion of
annuities are paid similarly to those for Social Security. However, the
adjustment for the Tier II portion is limited to 32.5 percent of the
previous year's increase in the Consumer Price Index. Supplemental
annuities and vested dual benefits are not subject to cost-of-living
adjustments.
Funding Source
Payroll taxes paid by railroad employers and their employees are the
primary source of funding for the railroad retirement benefit program.
Corresponding to the two-tier benefit structure, railroad retirement taxes
are levied on a two-tier basis. Railroad retirement Tier I payroll taxes
are coordinated with Social Security taxes so that employees and employers
pay Tier I taxes at the same rate as Social Security taxes. In addition,
both employees and employers pay Tier II taxes, which are used to finance
railroad retirement benefit payments over-and-above Social Security
equivalent levels. These Tier II taxes are based on the ratio of certain
asset balances to the sum of benefit payments and administrative expenses.
While the railroad retirement system has remained separate from the Social
Security system, the two systems are closely coordinated with regard to
earnings credits, benefit payments, and taxes. The financing of the two
systems is linked through a financial interchange under which, in effect,
the portion of railroad retirement annuities that is equivalent to Social
Security benefits is coordinated with the Social Security system. The
purpose of this financial coordination is to place the Social Security
trust funds in the same position they would be in if railroad service were
covered by the Social Security program instead of the railroad retirement
program.
Starting in fiscal year 2002, revenues in excess of benefit payments are
invested to provide additional trust fund income. The National Railroad
Retirement Investment Trust (NRRIT), established by the Railroad
Retirement and Survivors' Improvement Act of 2001, manages and invests
railroad retirement assets. The trust is a tax-exempt entity independent
from the federal government. Railroad retirement funds are invested in
nongovernmental assets, as well as in governmental securities. Prior to
the Act, investment of Railroad Retirement Account assets was limited to
U.S. government securities.
Additional trust fund income is derived from revenues from federal income
taxes on railroad retirement benefits, and appropriations from general
Treasury revenues provided after 1974 as part of a phase-out of certain
vested dual benefits.
Differences between Estimated and Actual Outlays
Based on a 5-year average, estimated outlays from the Rail Industry
Pension Fund differed from actual outlays by about $4.1 billion per year,
or 125.7 percent, in absolute value terms. The actual annual differences
between estimated and actual outlays varied between an overestimate of $77
million and an underestimate of about $17.9 billion. The majority of the
underestimate was a result of legislation that resulted in funds being
transferred out of the account and into a nongovernmental investment trust
fund. Table 13 presents the estimated and actual outlays associated with
the Rail Industry Pension Fund, by fiscal year.
Table 13: Estimated and Actual Rail Industry Pension Fund Outlays, by
Fiscal Year
Nominal dollars in millions
Fiscal year Original outlay Actual outlays Difference
estimate
2000 $3,038 $2,961 $77
2001 3,044 2,967 77
2002 3,078 4,814 -1,736
2003 3,416 21,326 -17,910
2004 3,639 4,225 -586
5-year average dollar difference $-4,016
5-year average difference as a -123.8%
percent of average estimated
outlays
5-year average dollar difference $4,077 125.7%
(absolute value) 5-year
percentage difference (absolute
value)
Source: GAO analysis of President's budget data.
Explanation of Key Differences
The discrepancies between estimated and actual outlays in fiscal years
2002 through 2004 can be attributed to the enactment of the Railroad
Retirement and Survivors' Improvement Act of 2001, which was signed into
law on December 21, 2001. This legislation lowered eligibility
requirements for annuitants and eliminated reductions that previously
applied to annuities of 30-year employees retiring between ages 60 and 62.
The law also lowered the minimum eligibility requirement to receive
regular annuities from 10 to 5 years of service after 1995 and increased
the Tier II amount paid to a widow(er) from 50 percent to 100 percent.
Additionally, the maximum limit on monthly railroad retirement benefits
was eliminated. The law reduced the Tier II tax rate on rail employers in
2002 and 2003, and in 2004 provided automatic Tier II tax rate adjustments
for both employers and employees. Lastly, funds in excess of those needed
for current payment of benefits and administrative expenses were
transferred to the National Railroad Retirement Investment Trust.
Agency officials indicated that the level of employment in the rail
industry is the most difficult factor to predict when estimating revenue
because it directly affects payroll tax income. Employment only affects
estimates in the long term, not short term. When reporting budget
estimates to OMB, the agency uses middle-range estimates that assume
employment will decrease gradually over time. Additionally, financial
interchanges of the estimated allocation of benefits between the Railroad
Retirement Account and Social Security Equivalent Benefit Account make it
difficult to estimate exact outlays as they are continually changing. A
detailed explanation of the differences is shown in table 14.
Table 14: Explanation of Differences between Estimated and Actual Rail
Industry Pension Fund Outlays
Nominal dollars
in billions
RRB's explanation of
differences
Fiscal year and Legislative Economic Technical
dollar
differencea
2000 Difference results
from changes in the
$0.08 estimated allocation
of benefits between
the Railroad
Retirement Account
and Social Security
Equivalent Benefit
Account. The actual
allocation of
benefits between
these accounts for a
given calendar year
is not known until
the financial
interchange
determination is
completed some 16
months after the end
of a calendar year
(which implies about
19 months after the
end of the fiscal
year ending in the
given calendar year).
2001 Difference results
from changes in the
$0.08 estimated allocation
of benefits between
the Railroad
Retirement Account
and Social Security
Equivalent Benefit
Account.
2002 The number of retirements The original outlay
increased due to the estimates are for
$-1.7 enactment of the Railroad benefit payments
Retirement and Survivors' only.
Improvement Act of 2001,
which lowered eligibility
requirements for annuitants
and eliminated reductions
that previously applied to
annuities of 30-year
employees retiring at age
60. The Act also lowered
the minimum eligibility
requirement to receive
regular annuities from 10
to 5 years of service after
1995 and increased the Tier
II amount paid to a
widow(er) from 50 percent
to 100 percent.
Additionally, the maximum
limit on monthly railroad
retirement benefits was
eliminated. The Act reduced
the Tier II tax rate on
rail employers in 2002 and
2003, and in 2004 provided
automatic Tier II tax rate
adjustments for both
employers and employees.
Funds in excess of those
required for current
payment of benefits and
administrative expenses,
$1.432 billion, were
transferred to the NRRIT.
2003 The number of retirements The original outlay
increased due to the estimates are for
$-17.9 enactment of the Railroad benefit payments
Retirement and Survivors' only.
Improvement Act of 2001.
Funds in excess of those
required for current
payment of benefits and
administrative expenses,
$17.75 billion, were
transferred to the NRRIT.
2004 Funds in excess of those The original outlay
required for current estimates are for
$-0.59 payment of benefits and benefit payments
administrative expenses, only.
$586 million, were
transferred to the NRRIT.
Source: Railroad Retirement Board.
Notes: Primary drivers of outlay differences are marked in bold.
aA negative difference means that actual outlays were higher than
originally estimated. A positive difference means that actual outlays were
less than originally estimated.
Illustrative Trigger and Response
If actual outlays exceeded estimates by more than the historical average,
Congress could reduce retirement benefits across the board. For example,
if estimated outlays historically differed from actual outlays by a
specified percent, increases in outlays above that specified percent could
automatically result in an across-the-board increase in retirement
contributions or a cut in retirement benefits. To determine an appropriate
threshold, rail officials would need to look at long-term historical
differences to minimize the effects of events such as the legislative
change in fiscal years 2002 and 2003.
Account Name
Unemployment Trust Fund
Administering Organization
Employment and Training Administration, U.S. Department of Labor
50 states, District of Columbia, Puerto Rico, and the Virgin Islands
Program Description
Unemployment insurance is designed to serve as a "counter-cyclical" remedy
to the effects of recessions by putting more dollars in the pockets of the
labor force, thereby increasing the demand for goods and services and
stabilizing the U.S. economy.
The Unemployment Trust Fund (UTF) finances unemployment insurance-a joint
federal-state program that provides temporary cash benefits to eligible
workers who become unemployed through no fault of their own and helps to
stabilize the economy in times of economic recession. Guided by federal
law, unemployed workers must meet certain criteria set by their state in
order to receive these benefits. Unemployment insurance is administered by
state employees under state law.
Extended benefits are paid during periods of high state unemployment.
Extended benefits are financed one-half by state payroll taxes and
one-half by the federal unemployment payroll tax. The federal tax also
pays for the cost of federal and state administration of unemployment
insurance, labor-market information programs, veterans' employment
services, and 97 percent of the costs of the employment service. States
may receive repayable advances from the UTF when their balances in the
fund are insufficient to pay benefits.
Federal unemployment payroll taxes accumulate in three accounts: (1) the
Employment Security Administration Account (ESAA), which covers both
federal and state administrative costs; (2) the Extended Unemployment
Compensation Account (EUCA), which covers the federal share of extended
unemployment benefits and has been used to fund temporary extended
unemployment compensation benefits; and (3) the Federal Unemployment
Account (FUA), which funds loans to insolvent state accounts. There is a
statutory ceiling on the size of each of these accounts, the amounts of
which are calculated each September. The ceiling for the ESAA account is
40 percent of the appropriated amounts during the fiscal year for which
the ceiling is being calculated. For the EUCA and FUA accounts, this
ceiling is 0.5 percent of the total covered wages in the prior calendar
year.
Funding Source
The UTF is funded by employer contributions (payroll taxes) and benefit
reimbursements from nonprofit entities and governmental units that are
paid in lieu of payroll taxes. The UTF may receive repayable advances from
the general fund of the Treasury when it has insufficient balances to make
advances to states or to pay the federal share of extended benefits.
The UTF invests its receipts in U.S. government securities and then draws
on them when the government needs to pay unemployment benefits and/or
cover administrative costs. In addition, the Treasury maintains a trust
fund account for each state that it can use to build up reserves in times
of economic stability. Forty-nine states have triggers that automatically
raise state employer taxes when UTF balances fall below a specific level.
States finance the costs of regular unemployment insurance benefits and
their half of the permanent Extended Benefits Program with employer
payroll taxes imposed on at least the first $7,000 paid annually to each
employee.
Differences between Estimated and Actual Outlays
Based on a 5-year average, estimated outlays from the UTF differed from
actual outlays by about $9.4 billion per year, or 29.6 percent, in
absolute value terms. However, the actual annual differences varied
between an overestimate of about $5 billion and an underestimate of about
$22 billion. Table 15 presents the estimated and actual outlays associated
with the unemployment program by fiscal year.
Table 15: Estimated and Actual Unemployment Trust Fund Outlays, by Fiscal
Year
Nominal dollars in millions
Fiscal year Original outlay Actual outlays Difference
estimate
2000 $25,773 $20,790 $4,983
2001 24,708 27,989 -3,281
2002 28,443 50,841 -22,398
2003 40,795 54,617 -13,822
2004 39,830 42,525 -2,695
5-year average dollar difference $-7,443
5-year average difference as a -23.3%
percent of average estimated
outlays
5-year average dollar difference $9,436 29.6%
(absolute value) 5-year
percentage difference (absolute
value)
Source: GAO analysis of President's budget data.
Explanation of Key Differences
Because the overall unemployment rate increased over the 5 fiscal years,
actual UTF outlays also increased as would be expected. UTF outlays are
highly sensitive to changes in the unemployment rate. For example, between
2000 and 2001 the 17.5 percent increase in the unemployment rate was
associated with a 34.6 percent increase in actual UTF outlays. This
relationship is best illustrated by referring to figure 8.
Figure 8: Percent Change in Unemployment Rate versus Percent Change in
Actual UTF Outlays
Between 2001 and 2002, UTF outlays increased 81.6 percent in response to a
23.4 percent increase in the unemployment rate. In 2002, part of the
outlay increase was due to legislation extending federally-funded
unemployment insurance benefits through the Temporary Employment
Compensation Act of 2002 (TEUC) which resulted in unanticipated UTF
outlays. The unemployment rate continued to rise during this time as
130,000 workers were displaced after the events on September 11, 2001, and
the economic recession persisted. Between 2002 and 2003, TEUC benefits
were extended and the unemployment rate continued to increase but did so
at a decreasing rate. The 3.4 percent increase in the unemployment rate
and the subsequent extension of TEUC led to the 7.4 percent increase in
UTF outlays. Between 2003 and 2004, the unemployment rate decreased by 8.3
percent and outlays decreased by about 22 percent. Table 16 presents the
Department of Labor's (Labor) explanation for differences between
estimated and actual outlays.
Table 16: Explanation of Differences between Estimated and Actual
Unemployment Trust Fund Outlays
Nominal dollars
in billions
Labor's explanation
of differences
Fiscal year and Legislative Economic Technical
dollar
differencesa
2000 An overestimate of the
unemployment rate (5 percent
$5.0 estimated versus 4 percent
actual) explained $4.7
billion of the difference.
2001 An underestimate of the
recipiency rateb (38 percent
$-3.3 estimated versus 42 percent
actual) explained $2.9
billion of the difference.
The recipiency rate
increased from 37 percent
the prior year, which is
typical for a recession, but
was unanticipated.c
2002 TEUCd enactment A 1.1 percent underestimate
resulted in $7.9 of the unemployment rate due
$-22.4 billion of to recession resulted in a
unanticipated $6.6 billion difference.
outlays. Underestimates of the
recipiency rate and the
average weekly benefit
accounted for about $4.8
billion and $2.4 billion in
outlays, respectively.e
2003 TEUC extension An underestimate of the
resulted in $11 unemployment rate accounted
$-13.9 billion of for an additional $2.9
unanticipated billion.
outlays.
2004 TEUC extension Overestimates of the
resulted in $4.3 recipiency rate and average
$-2.7 billion of weekly benefit partly offset
unanticipated the TEUC extension.
outlays.
Source: Department of Labor.
Notes: Primary drivers of outlay differences are marked in bold.
aA negative difference means that actual outlays were higher than
originally estimated. A positive difference means that actual outlays were
less than originally estimated.
bThe recipiency rate refers to the number of benefit claims and, more
specifically, is the ratio of the insured unemployed (claimants) to the
total number of unemployed. The rate tends to vary between 35 and 45
percent.
cAccording to the CRS, the terrorist attacks of September 11, 2001 are
directly attributed to displacing 130,000 employees.
dThe Temporary Extended Unemployment Compensation Act (TEUC), as amended,
temporarily extended unemployment benefits from March 2002 through
December 2004.
eDepartment of Labor officials did not include the $8 billion Reed
Distribution in 2002 as part of the explanation of the difference between
estimated and actual outlays in fiscal year 2002 because it was considered
an intragovernmental transfer and was not recorded until the states used
the money held in the U.S. Treasury. Reed Distributions to states'
accounts occur when funds accumulating in federal unemployment accounts
reach statutorily set limits.
Illustrative Triggers and Responses
Currently, when funds accumulating in federal unemployment accounts reach
statutorily set limits, a distribution of the "excess" funds from the UTF
to individual states' accounts in the U.S. Treasury is automatically
triggered based on each state's share of covered wages. These
distributions are known as "Reed Distributions."20 Congress can also
legislatively trigger a special distribution21 as it did in March 2002,
which provided $8 billion in distributions to all 50 states, the District
of Columbia, Puerto Rico, and the Virgin Islands and extended UTF benefits
up to an additional 13 weeks longer than the maximum 26 weeks previously
allowed by most states.
One potential option to constrain federal spending would be to increase
the statutory cap on federal unemployment accounts, thus making it more
difficult to trigger Reed Distributions to states. By making it more
difficult to trip the trigger, funds could continue to build during
economic prosperity and be available to states when truly needed to
counter rising unemployment.
A different alternative for constraining growth would be to establish a
trigger using a measure of economic prosperity-such as GDP growth in a
specified number of consecutive quarters. If this trigger was reached,
federal unemployment taxes would automatically increase, allowing trust
fund balances to rise. To avoid procyclical effects, these taxes could be
automatically reduced again using periods of rising unemployment or
recession as the trigger for that action.
Analysis of Total Outlays, Receipts, and Fiscal Position Appendix II
While the focus of this report is on budget triggers as they relate to
selected case study accounts, we have included our analysis of aggregate
receipts, outlays, and surplus/deficit measures to provide broader
context. Findings related to our seven case study accounts and the reasons
for differences between estimated and actual outlays are discussed in the
body of this report. More detailed summaries of each account are included
in appendix I.
Aggregate Mandatory Spending Estimates Were Close to Actual Outlays but
Large Differences Appear at the Account Level
In the aggregate, original estimates of total mandatory spending were
fairly close to actual results, however large discrepancies were evident
at the account level. During fiscal years 2000 through 2004, estimated
total mandatory outlays differed from actuals by no more than about 2
percent, or $24 billion. However at the account level, average estimated
and actual outlays varied greatly. While the largest difference was in the
Interest on the Public Debt account-a result of other changes-other
accounts also showed significant changes between estimated and actual
outlays. Alternatively, there are many mandatory accounts with virtually
no differences between estimated and actual outlays. The variation among
individual accounts was not apparent at the aggregate level because the
combination of positive and negative differences offset each other.
Figure 9 shows that total spending on mandatory programs was expected to
rise throughout the 5-year period and that resulting outlays were just
slightly higher than expected.
Figure 9: Estimated and Actual Total Mandatory Outlays for FYs 2000-2004,
constant 2004 dollars
Although aggregate estimates were close to actual estimates, the continued
actual and forecasted growth in mandatory programs has raised concerns
about the government's long-term fiscal outlook. Addressing growth in
mandatory spending is an important but complicated matter that requires
looking below the aggregate and into specific programs.
Differences between Estimated and Actual Mandatory Outlays Had Limited
Effect on the Unified Deficit/Surplus
The unified budget deficit/surplus measures federal fiscal position, that
is, the difference between total annual receipts and outlays. Not
surprisingly, the relatively small differences between total estimated and
actual mandatory outlays had a limited effect on the unified budget
surplus/deficit. In most cases throughout fiscal years 2000 through 2004,
the difference between estimated and actual mandatory outlays accounted
for approximately 7 percent or less of the difference between the
estimated and actual fiscal position. Despite the fact that mandatory
outlays were close to expectations, surplus/deficit measures proved
difficult to estimate throughout the 5-year period, primarily because of
misestimates of federal receipts.1
During fiscal years 2000 through 2004, deficit/surplus projections were
generally more optimistic than reality. Figure 10 illustrates the
estimated and actual fiscal position (surplus/deficit) throughout the
5-year period. Although increasing surpluses were projected for the first
three years followed by growing deficits, actual results show that the
nation's fiscal position in fact declined throughout the 5-year timeframe.
In addition, projections for fiscal years 2003 and 2004 show that the
deficit was expected to grow but not to the magnitude that ultimately
resulted.
Figure 10: Estimated and Actual Surplus/Deficit, Fiscal Years 2000-2004,
constant 2004 dollars
The fiscal position represents the difference between total federal
revenues and outlays in a given year. Although mandatory spending
constitutes more than half of total federal spending, misestimates of the
amount of mandatory spending did not contribute significantly to the
differences between the predicted and actual fiscal position. According to
the detailed receipt and outlay data shown in table 17, the mandatory
outlay difference in most cases accounted for less than 7 percent of the
difference between the estimated and actual fiscal position with one
exception. In fiscal year 2001, the mandatory outlay estimating error had
a larger than usual effect-approximately 29 percent-on the fiscal position
estimating error. While this particular year stands out in the analysis,
it is a reasonable result given that the total amount of error in
surplus/deficit projections was much smaller-approximately 30 percent or
$60 billion-compared with any other year during the 5-year period. For
example, a $242 billion surplus was projected for 2002 when in fact the
nation's fiscal position changed from surplus to deficit, resulting in a
$165 billion deficit for that year.2 This discrepancy represented a
misestimate of approximately 168 percent. In both fiscal years 2001 and
2002, mandatory outlay estimates differed from actual outlays by
approximately 2 percent. This relatively small difference accounted for
over one quarter of the resulting error in the surplus projection for 2001
because the difference between estimated and actual receipts also was
relatively small. It accounted for less than one-tenth of the total error
in the fiscal position projection for 2002 because the difference between
estimated and actual receipts was much larger. Effects similar to the
latter occurred more frequently throughout the 5-year period, indicating
that estimation errors in mandatory outlays had a limited effect on fiscal
position.
Table 17: Aggregate Estimated and Actual Outlays and Receipts for Fiscal
Years 2000-2004
Constant 2004 dollars
in billions
Original Actual Difference Percent of Percent of
estimate original difference
estimate
Fiscal Year 2000
Receipts $2,055.7 $2,210.9 -$155.3 -7.6% 119.7%
Outlays 1,927.6 1,953.1 -25.5 -1.3 19.7
Discretionary spending 645.7 671.2 -25.4 -3.9 19.6
Mandatory spending 1,093.7 1,085.0 8.6 0.8 -6.6
Offsetting receipts -46.8 -46.5 -0.3 0.7 0.3
Net interest 234.9 243.3 -8.4 -3.6 6.5
Surplus/Deficit $128.1 $257.8 -$129.7 -101.3% 100.0%
Fiscal Year 2001
Receipts $2,154.1 $2,124.4 $29.7 1.4% 49.9%
Outlays 1,957.8 1,987.7 -29.9 -1.5 -50.1
Discretionary spending 676.3 692.7 -16.4 -2.4 -27.6
Mandatory spending 1,107.6 1,125.1 -17.5 -1.6 -29.4
Offsetting receipts -48.4 -50.1 1.7 -3.5 2.9
Net interest 222.2 220.0 2.2 1.0 3.8
Surplus/Deficit $196.3 $136.7 $59.6 30.3% 100.0%
Fiscal Year 2002
Receipts $2,296.5 $1,941.7 $354.7 15.4% 87.0%
Outlays 2,054.2 2,107.1 -52.8 -2.6 -13.0
Discretionary spending 724.7 769.4 -44.6 -6.2 -11.0
Mandatory spending 1,184.1 1,208.2 -24.1 -2.0 -5.9
Offsetting receipts -51.8 -49.7 -2.1 4.0 -0.5
Net interest 197.1 179.1 18.0 9.1 4.4
Surplus/Deficit $242.2 -$165.3 $407.6 168.3% 100.0%
Fiscal Year 2003
Receipts $2,093.5 $1,821.9 $271.6 13.0% 89.3%
Outlays 2,175.4 2,207.8 -32.4 -1.5 -10.7
Discretionary spending 806.5 843.7 -37.2 -4.6 -12.2
Mandatory spending 1,260.0 1,263.2 -3.2 -0.3 -1.0
Offsetting receipts -75.7 -55.6 -20.1 26.6 -6.6
Net interest 184.7 156.5 28.2 15.3 9.3
Surplus/Deficit -$81.9 -$386.0 $304.0 -371.0% 100.0%
Fiscal Year 2004
Receipts $1,922.0 $1,880.1 $42.0 2.2% 40.1%
Outlays 2,229.4 2,292.2 -62.8 -2.8 -59.9
Discretionary spending 818.8 895.4 -76.6 -9.4 -73.1
Mandatory spending 1,287.9 1,295.1 -7.2 -0.6 -6.9
Offsetting receipts -53.7 -58.5 4.8 -8.9 4.6
Net interest 176.4 160.2 16.2 9.2 15.5
Surplus/Deficit -$307.4 -$412.1 $104.7 -34.1% 100.0%
Source: GAO analysis of President's budget data.
In contrast, revenue estimate inaccuracies proved to have a greater effect
on projections of the nation's fiscal position. Throughout the 5-year
period, total estimated outlays differed from actual outlays by no more
than 3 percent while total estimated receipts differed from actual
receipts by up to 15 percent in absolute value terms. This suggests that
revenue, rather than outlay estimates, led most significantly to the
discrepancies in surplus/deficit projections. Figure 11 shows the total
estimated and actual federal receipts in dollar terms for each year we
reviewed.3
Figure 11: Total Estimated and Actual Receipts, Fiscal Years 2000-2004,
constant 2004 dollars
As mentioned earlier in this report, the greatest revenue estimating
errors occurred in 2000, 2002, and 2003, which correlate with the years in
which the fiscal position projections were the most inaccurate. For
example, in fiscal year 2002, an approximate 2.6 percent underestimate in
total outlays coupled with an approximate 15.4 percent overestimate of
receipts translated into a large shift in fiscal position from surplus to
deficit. Similar effects occurred in 2000 and 2003. As shown in table 18,
the driving source of revenue misestimates in any given year varied, but
individual and corporate income taxes often proved difficult to estimate.
Table 18: Revenue Estimates and Actual Results by Source and Fiscal Year
Constant 2004 dollars in
millions
Component of revenue Original Actual Actual minus Percent of
estimate original original
Fiscal Year 2000
Individual income taxes $982,250 $1,096,574 -$114,324 -11.6%
Corporate income taxes 206,721 226,298 -19,578 -9.5
Social insurance taxes and 694,901 712,721 -17,820 -2.6
contributions
Excise taxes 76,312 75,180 1,132 1.5
Estate and gift taxes 29,445 31,670 -2,225 -7.6
Customs duties 20,048 21,740 -1,692 -8.4
Miscellaneous receipts 45,991 46,753 -762 -1.7
Total $2,055,668 $2,210,937 -$155,269 -7.6%
Fiscal Year 2001
Individual income taxes $1,037,459 $1,060,855 -$23,396 -2.3%
Corporate income taxes 207,799 161,181 46,618 22.4
Social insurance taxes and 727,707 740,389 -12,682 -1.7
contributions
Excise taxes 81,805 70,663 11,143 13.6
Estate and gift taxes 34,465 30,300 4,165 12.1
Customs duties 22,267 20,665 1,602 7.2
Miscellaneous receipts 42,590 40,341 2,249 5.3
Total $2,154,093 $2,124,393 $29,699 1.4%
Fiscal Year 2002
Individual income taxes $1,130,332 $899,356 $230,977 20.4%
Corporate income taxes 229,239 155,117 74,122 32.3
Social insurance taxes and 760,476 734,241 26,234 3.4
contributions
Excise taxes 77,557 70,190 7,367 9.5
Estate and gift taxes 30,070 27,773 2,297 7.6
Customs duties 23,614 19,491 4,123 17.5
Miscellaneous receipts 45,165 35,547 9,618 21.3
Total $2,296,452 $1,941,715 $354,737 15.4%
Fiscal Year 2003
Individual income taxes $1,028,676 $811,304 $217,372 21.1%
Corporate income taxes 210,047 134,701 75,346 35.9
Social insurance taxes and 765,831 728,793 37,038 4.8
contributions
Excise taxes 70,552 69,022 1,530 2.2
Estate and gift taxes 23,509 22,446 1,063 4.5
Customs duties 20,244 20,303 -58 -0.3
Miscellaneous receipts -25,371 35,308 -60,679 239.2
Total $2,093,489 $1,821,877 $271,612 13.0%
Fiscal Year 2004
Individual income taxes $849,880 $808,959 $40,921 4.8%
Corporate income taxes 169,060 189,371 -20,311 -12.0
Social insurance taxes and 764,548 733,407 31,141 4.1
contributions
Excise taxes 70,905 69,855 1,050 1.5
Estate and gift taxes 23,379 24,831 -1,452 -6.2
Customs duties 20,713 21,083 -370 -1.8
Miscellaneous receipts 38,540 32,565 5,975 15.5
Total $1,937,025 $1,880,071 $56,954 2.9%
Source: GAO analysis of President's budget data.
Mandatory Budget Accounts Appendix III
Table 19: Budget Accounts with Greater than 50 percent Mandatory Outlays
Dollars in
millions
Obs. # Agency Account 5-year avg.
dollar
change
(absolute
value)
1 Treasury Interest on Treasury debt $20,016
securities (gross)
2 Veterans Affairs Disability compensation 9,977
benefits*
3 Labor Unemployment trust fund 9,436
4 Veterans Affairs Compensation* 7,423
5 Agriculture Commodity Credit 6,944
Corporation fund
6 Health and Human Federal supplementary 6,427
Services medical insurance trust
fund
7 Health and Human Federal hospital insurance 5,622
Services trust fund
8 Office of Personnel Employees health benefits 5,065
Management fund*
9 Health and Human Grants to States for 4,220
Services Medicaid
10 Railroad Retirement Rail industry pension fund 4,077
Board
11 Health and Human Payments to health care 3,800
Services trust funds
12 Housing and Urban FHA-mutual mortgage 2,892
Development insurance program account*
13 Veterans Affairs Pensions benefits* 2,528
14 Education Federal direct student 2,370
loan program account
15 Health and Human Immediate helping hand 2,240
Services prescription drug plan*
16 Postal Service Postal Service fund 2,129
17 Housing and Urban FHA-mutual mortgage and 2,067
Development cooperative housing
insurance funds
liquidating account
18 Agriculture Food stamp program 2,058
19 Treasury Temporary State fiscal 2,000
assistance fund*
20 Social Security Federal old-age and 1,979
Administration survivors insurance trust
fund
21 Office of Personnel Payment to civil service 1,862
Management retirement and disability
fund
22 Allowances Bipartisan economic 1,600
security plan*
23 Treasury Payment where child credit 1,582
exceeds liability for tax
24 Education Federal family education 1,573
loan program account
25 Federal Communications Universal service fund 1,539
Commission
26 Health and Human Temporary assistance for 1,322
Services needy families
27 Health and Human Allowance for Medicare 1,200
Services modernization*
28 Justice September 11th victim 1,192
compensation (general
fund)*
29 Social Security Payments to social 1,168
Administration security trust funds
30 Labor Advances to the 1,154
Unemployment trust fund
and other* funds
31 Treasury Refunding internal revenue 1,153
collections, interest
32 Office of Personnel Government payment for 1,081
Management annuitants, employees
health benefits
33 Treasury Payment where earned 1,004
income credit exceeds
liability for tax
34 Office of Personnel Civil service retirement 988
Management and disability fund
35 Labor Black lung disability 987
trust fund
36 Housing and Urban FHA-general and special 964
Development risk insurance funds
liquidating account
37 Health and Human State children's health 922
Services insurance fund
38 Transportation Compensation for air 910
carriers*
39 Social Security Federal disability 865
Administration insurance trust fund
40 International Assistance Foreign military sales 850
Programs trust fund
41 Federal Deposit Bank insurance fund* 844
Insurance Corporation
42 Treasury Payment to the Resolution 799
Funding Corporation
43 International Assistance United States quota, 793
Programs International Monetary
Fund*
44 Federal Communications Spectrum auction program 785
Commission account*
45 Education Federal family education 778
loan liquidating account
46 Office of Personnel Employees and retired 775
Management employees health benefits
funds*
47 Social Security Supplemental security 759
Administration income program
48 Agriculture Rural electrification and 681
telecommunications
liquidating account
49 Justice Immigration support* 652
50 Small Business Business loan program 625
Administration account*
51 Treasury Interest paid to credit 625
financing accounts
52 Justice Crime victims fund 609
53 Veterans Affairs Housing program account 594
54 Labor Pension benefit guaranty 581
corporation fund
55 Treasury Claims, judgments, and 569
relief acts
56 Agriculture Commodity Credit 553
Corporation export loans
program account
57 Housing and Urban FHA-mutual mortgage 543
Development insurance capital reserve
account*
58 Housing and Urban FHA-general and special 536
Development risk program account*
59 Export-Import Bank of Export-Import Bank loans 516
the United States program account*
60 Health and Human Child care entitlement to 499
Services States
61 Treasury Federal Financing Bank 488
62 Homeland Security Citizenship and 471
Immigration Services*
63 Agriculture Agricultural credit 467
insurance fund program
account*
64 Treasury Exchange stabilization 440
fund
65 Justice Immigration services* 418
66 Labor Reemployment accounts* 400
67 Health and Human Payments to States for 392
Services foster care and adoption
assistance
68 Other Defense Civil Payment to Department of 390
Programs Defense Medicare-eligible
retiree health care fund*
69 Other Defense Civil Payment to military 376
Programs retirement fund
70 Railroad Retirement National railroad 367
Board retirement investment
trust*
71 Railroad Retirement Railroad social security 363
Board equivalent benefit account
72 Other Defense Civil Department of Defense 361
Programs Medicare-Eligible retiree
health care fund*
73 Labor Welfare to work jobs 353
74 Homeland Security Retired Pay* 352
75 Department of Allied contributions and 346
Defense-Military cooperation account
76 Veterans Affairs Education benefits 324
77 Tennessee Valley Tennessee Valley Authority 323
Authority fund
78 Small Business Disaster loans program 319
Administration account*
79 Energy Bonneville Power 317
Administration fund
80 Department of Pentagon reservation 317
Defense-Military maintenance revolving
fund*
81 Other Defense Civil Military retirement fund 312
Programs
82 Department of Iraq relief and 310
Defense-Military reconstruction fund, Army*
83 Agriculture Federal crop insurance 310
corporation fund
84 Labor Payments to the 305
Unemployment trust fund*
85 Federal Deposit Savings association 295
Insurance Corporation insurance fund*
86 Health and Human Payments to States for 284
Services child support enforcement
and family support
programs
87 Interior Mineral leasing and 279
associated payments
88 Treasury Air transportation 273
stabilization program
account*
89 Homeland Security National Flood Insurance 269
Fund*
90 Interior Interior Franchise Fund* 260
91 Agriculture Farm security and rural 254
investment programs*
92 Agriculture Funds for strengthening 241
markets, income, and
supply (section 32)
93 Health and Human Allowance for transitional 240
Services Medicare low-income drug
assistance*
94 Federal Emergency National flood insurance 227
Management Agency fund*
95 Housing and Urban Guarantees of 223
Development mortgage-backed securities
liquidating account*
96 Veterans Affairs Vocational rehabilitation 202
and employment benefits*
97 Department of National defense stockpile 197
Defense-Military transaction fund*
98 National Credit Union Credit union share 195
Administration insurance fund
99 Federal Deposit FSLIC resolution fund 189
Insurance Corporation
100 Export-Import Bank of Export-Import Bank of the 188
the United States United States liquidating
account
101 Office of Personnel Employees life insurance 187
Management fund
102 Education Federal student loan 182
reserve fund*
103 Agriculture Rural electrification and 180
telecommunications loans
program account*
104 Transportation Coast Guard military 178
retirement fund*
105 Agriculture Child nutrition programs 177
106 Health and Human Social services block 177
Services grant
107 Treasury Payment where health care 177
credit exceeds liability
for tax*
108 Transportation Retired pay* 174
109 Agriculture Rural development 174
insurance fund liquidating
account
110 Labor Energy employees 160
occupational illness
compensation fund*
111 Labor Federal unemployment 159
benefits and allowances
112 Veterans Affairs Supply fund* 159
113 Health and Human Ricky Ray hemophilia 150
Services relief fund*
114 Transportation Payment to Coast Guard 147
military retirement fund*
115 Social Security Payment to social security 146
Administration trust funds post-1956
military service wage
credits*
116 Agriculture Forest Service trust funds 139
117 Agriculture Rural housing insurance 131
fund liquidating account
118 Treasury Restitution of forgone 129
interest*
119 International Assistance Economic assistance loans 127
Programs liquidating account
120 Agriculture Agricultural credit 125
insurance fund liquidating
account
121 Interior Tribal special fund 123
122 Interior Working capital fund* 118
123 Housing and Urban Housing for the elderly or 118
Development handicapped fund
liquidating account
124 District of Columbia Federal payment to the 117
District of Columbia
pension fund
125 Agriculture Forest Service permanent 116
appropriations
126 Small Business Disaster loan fund 116
Administration liquidating account
127 Health and Human Payment to the Ricky Ray 116
Services hemophilia relief fund*
128 Farm Credit System Financial Assistance 116
Financial Assistance Corporation assistance
Corporation fund liquidating account
129 Transportation Ocean freight differential 115
130 Legislative Branch Payments to copyright 113
owners
131 Veterans Affairs Housing liquidating 104
account
132 Treasury Contribution for annuity 103
benefits*
133 Social Security Special benefits for 95
Administration disabled coal miners*
134 Labor Special benefits for 94
disabled coal miners*
135 International Assistance Foreign military loan 93
Programs liquidating account
136 Treasury Treasury forfeiture fund* 88
137 Small Business Business loan fund 87
Administration liquidating account
138 Labor Special benefits 86
139 International Assistance Overseas Private 86
Programs Investment Corporation
program account*
140 Interior Miscellaneous permanent 82
payment accounts
141 Agriculture Commodity Credit 81
Corporation guaranteed
loans liquidating account
142 Treasury Payment of anti-terrorism 80
judgments*
143 Judicial Branch Judiciary filing fees 80
144 Education Rehabilitation services 79
and disability research
145 Treasury Payment where alternative 76
to failing school credit
exceeds liability for tax*
146 International Assistance Foreign military financing 70
Programs loan program account*
147 Veterans Affairs National service life 70
insurance fund
148 Treasury Internal revenue 68
collections for Puerto
Rico
149 Corps of Engineers-Civil Rivers and harbors 67
Works contributed funds
150 Health and Human Retirement pay and medical 67
Services benefits for commissioned
officers
151 Transportation Miscellaneous trust funds 65
152 United Mine Workers of United Mine Workers of 63
America Benefit Funds America combined benefit
fund
153 Agriculture Healthy investments in 63
rural environments*
154 Railroad Retirement Federal payments to the 61
Board railroad retirement
accounts
155 Transportation Maritime guaranteed loan 60
(Title XI) program
account*
156 Agriculture Rural telephone bank 60
liquidating account
157 General Services General supply fund* 60
Administration
158 Veterans Affairs Burial benefits* 59
159 Treasury Continued dumping and 56
subsidy offset*
160 Transportation Aviation insurance 56
revolving fund*
161 Agriculture Fund for rural America 55
162 Agriculture Payments to states 54
stabilization*
163 Health and Human Program management* 52
Services
164 Veterans Affairs Burial benefits and 51
miscellaneous assistance*
165 Treasury Refunds, transfers, and 51
expenses of operation,
Puerto Rico*
166 Federal Communications Pioneer's preference 50
Commission settlement*
167 Treasury Restoration of lost 49
interest, Medicare trust
funds*
168 Corps of Engineers-Civil Revolving fund* 48
Works
169 Agriculture Expenses, Public Law 480, 48
foreign assistance
programs, Agriculture
liquidating account
170 Agriculture Expenses and refunds, 48
inspection and grading of
farm products*
171 Justice Working capital fund* 48
172 Department of Army conventional 48
Defense-Military ammunition working capital
fund*
173 United States Enrichment United States Enrichment 47
Corporation Fund Corporation Fund*
174 Health and Human Public Health Service 47
Services Commissioned Corps
retirement fund*
175 Department of Surcharge collections, 46
Defense-Military sales of commissary
stores, Defense*
176 Interior Recreation fee permanent 45
appropriations
177 Housing and Urban Working capital fund* 44
Development
178 Health and Human Transitional drug 43
Services assistance, Federal
supplementary medical
insurance trust fund*
179 Treasury Confiscated and vested 42
Iraqi property and assets*
180 Labor Administrative expenses, 40
Energy employees
occupational illness
compensation fund*
181 Interior Lower Colorado River Basin 40
development fund
182 Treasury Federal Reserve Bank 38
reimbursement fund
183 Legislative Branch Government Printing Office 38
revolving fund
184 Justice Public safety officers' 37
benefits*
185 Interior Permanent operating funds 37
186 Department of State Working capital fund* 36
187 Treasury Financial agent services* 36
188 Commerce Census working capital 36
fund*
189 Justice Federal Prison Industries, 36
Incorporated*
190 Federal Deposit Federal deposit insurance 36
Insurance Corporation fund*
191 Interior Upper Colorado River Basin 35
fund*
192 Farm Credit System Farm credit system 35
Insurance Corporation insurance fund
193 Health and Human Payment to health care 35
Services trust funds for post-1956
military service wage
credits*
194 Justice Assets forfeiture fund 35
195 Health and Human Health care fraud and 34
Services abuse control account
196 Health and Human Health education 33
Services assistance loans program
account*
197 International Assistance Housing and other credit 33
Programs guaranty programs
liquidating account
198 Agriculture Initiative for future 33
agriculture and food
systems*
199 Justice Radiation exposure 32
compensation trust fund*
200 Homeland Security Boat Safety* 30
201 Agriculture Payments to States, 29
northern spotted owl
guarantee, Forest Service*
202 Treasury Working capital fund* 29
203 Corps of Engineers-Civil Coastal wetlands 29
Works restoration trust fund
204 Transportation Boat safety* 28
205 Other Defense Civil Contributions* 28
Programs
206 Homeland Security Oil Spill Recovery* 28
207 Transportation Oil spill recovery* 28
208 Interior Tribal trust fund 27
209 Interior Compact of free 27
association
210 Commerce Promote and develop 26
fishery products and
research pertaining to
American fisheries
211 Health and Human Payment to Public Health 26
Services Service Commissioned Corps
retirement system*
212 Central Intelligence Payment to Central 26
Agency Intelligence Agency
retirement and disability
system fund
213 Housing and Urban Low-rent public 25
Development housing-loans and other
expenses
214 Environmental Protection Re-registration and 24
Agency expedited processing
revolving fund*
215 Health and Human HHS service and supply 23
Services fund*
216 Interior Abandoned mine reclamation 23
fund*
217 Federal Emergency Disaster assistance direct 23
Management Agency loan program account*
218 Justice Fees and expenses of 23
witnesses
219 Judicial Branch Judiciary information 22
technology fund
220 Health and Human Vaccine injury 22
Services compensation program trust
fund
221 Interior Natural resource damage 22
assessment fund
222 Agriculture Miscellaneous trust funds* 21
223 Health and Human Promoting safe and stable 21
Services families
224 Interior Payments to the United 21
States territories, fiscal
assistance
225 Agriculture McGovern-Dole 20
international food for
education and child
nutrition program*
226 Treasury Assessment funds 20
227 Transportation Essential air service and 20
rural airport improvement
fund*
228 Housing and Urban Revolving fund 18
Development (liquidating programs)*
229 Agriculture Perishable Agricultural 18
Commodities Act fund
230 International Assistance Urban and environmental 18
Programs credit program account*
231 Justice Commissary funds, Federal 18
prisons (trust revolving
fund)*
232 Veterans Affairs Veterans special life 18
insurance fund
233 Agriculture Working capital fund* 17
234 Federal Emergency Flood map modernization 17
Management Agency fund*
235 Agriculture Trade adjustment 16
assistance for farmers*
236 Interior Federal aid in wildlife 16
restoration
237 Homeland Security Refunds, transfers, and 16
expenses of operation,
Puerto Rico*
238 Agriculture Milk market orders 16
assessment fund*
239 Health and Human Health education 15
Services assistance loans
liquidating account
240 Transportation Right-of-way revolving 15
fund liquidating account
241 Justice Payment to radiation 15
exposure compensation
trust fund*
242 Interior Working capital fund 15
243 Railroad Retirement Railroad unemployment 15
Board insurance trust fund
244 Department of Other DOD trust funds 14
Defense-Military
245 Veterans Affairs Franchise fund* 14
246 Treasury Office of Thrift 14
Supervision*
247 Justice Diversion control fee 14
account
248 General Services Panama Canal revolving 14
Administration fund*
249 National Science Donations 13
Foundation
250 Farm Credit System Financial assistance 13
Financial Assistance corporation trust fund*
Corporation
251 District of Columbia District of Columbia 12
Federal pension liability
trust fund
252 Interior Colorado River dam fund, 12
Boulder Canyon project
253 Interior Sport fish restoration 12
254 Treasury Presidential election 12
campaign fund*
255 United Mine Workers of United Mine Workers of 12
America Benefit Funds America 1992 benefit plan
256 Department of State Payment to Foreign Service 12
retirement and disability
fund
257 Health and Human Miscellaneous trust funds 11
Services
258 Commerce Emergency steel guaranteed 11
loan program account*
259 Health and Human State grants and 11
Services demonstrations*
260 Health and Human Children's research and 10
Services technical assistance
261 Agriculture Rural strategic investment 10
program grants*
262 Health and Human Contingency fund* 10
Services
263 Federal Emergency National flood mitigation 10
Management Agency fund*
264 Agriculture Rural economic development 10
grants
265 Interior Everglades watershed 10
protection*
266 Interior Other permanent 10
appropriations
267 Health and Human Vaccine injury 9
Services compensation*
268 Federal Emergency Disaster assistance direct 9
Management Agency loan liquidating account*
269 Interior Interior Franchise Fund* 9
270 Federal Deposit Office of Inspector 9
Insurance Corporation General
271 Veterans Affairs Post-Vietnam era veterans 9
education account
272 Federal Retirement Program expenses 9
Thrift Investment Board
273 Interior Assistance to territories* 9
274 Commerce Working capital fund* 8
275 Department of Buildings maintenance 8
Defense-Military fund*
276 Department of State Miscellaneous trust funds 8
277 Interior Helium fund 8
278 Other Defense Civil Education benefits fund 8
Programs
279 Justice Independent counsel 8
280 Corps of Engineers-Civil Washington aqueduct* 8
Works
281 Interior Reclamation trust funds* 8
282 Agriculture Rural cooperative 8
development grants*
283 Energy Emergency fund, Western 8
Area Power Administration*
284 United Mine Workers of Federal payment to United 8
America Benefit Funds Mine Workers of America
combined benefit fund*
285 Agriculture Rural business investment 7
program account*
286 Education College housing and 7
academic facilities loans
liquidating account
287 Interior Miscellaneous trust funds* 7
288 Veterans Affairs Service-disabled veterans 7
insurance fund
289 Legislative Branch Gift and trust fund 7
accounts
290 Agriculture Local television loan 7
guarantee program account*
291 Housing and Urban Community development loan 7
Development guarantees liquidating
account*
292 Transportation Working Capital Fund* 7
293 Department of State Foreign Service retirement 7
and disability fund
294 Public Company Public Company Accounting 7
Accounting Oversight Oversight Board*
Board
295 Agriculture Miscellaneous trust funds 7
296 Department of Foreign national employees 7
Defense-Military separation pay
297 Transportation Federal ship financing 7
fund liquidating account
298 Interior Cooperative fund (Papago)* 6
299 Commerce Coastal zone management 6
fund
300 Energy Continuing fund, 6
Southeastern Power
Administration*
301 International Assistance Overseas Private 6
Programs Investment Corporation
liquidating account*
302 Housing and Urban Rental housing assistance 6
Development fund*
303 Interior Miscellaneous trust funds 6
304 Treasury Federal interest 6
liabilities to States
305 Department of Voluntary separation 5
Defense-Military incentive fund
306 Labor Special workers' 5
compensation expenses
307 Health and Human State grants and 5
Services demonstrations*
308 Housing and Urban Community development loan 5
Development guarantees program
account*
309 Interior Miscellaneous permanent 5
appropriations
310 Interior Contribution for annuity 5
benefits*
311 Health and Human Job opportunities and 5
Services basic skills training
program*
312 Veterans Affairs Miscellaneous veterans 5
housing loans program
account*
313 Agriculture Miscellaneous contributed 5
funds*
314 Commerce Economic development 5
revolving fund liquidating
account*
315 Commerce Environmental improvement 5
and restoration fund*
316 Veterans Affairs Veterans reopened 5
insurance fund
317 Agriculture Conservation reserve 4
program*
318 Veterans Affairs Canteen service revolving 4
fund*
319 Legislative Branch U.S. Capitol Preservation 4
Commission*
320 Commerce Federal ship financing 4
fund fishing vessels
liquidating account*
321 Interior Payments for trust 4
accounting deficiencies*
322 National Credit Union Operating fund* 4
Administration
323 Railroad Retirement Supplemental annuity 4
Board pension fund*
324 Agriculture National Sheep Industry 4
Improvement Center*
325 Agriculture Road and trail fund* 4
326 Commerce Payments to NOAA 4
commissioned officer corps
retirement fund*
327 Treasury Administering the public 4
debt*
328 International Assistance Loan guarantees to Israel 4
Programs program account*
329 Agriculture Renewable energy program 4
account*
330 District of Columbia Federal payment for water 4
and sewer services*
331 Department of Host nation support fund 4
Defense-Military for relocation
332 Standard Setting Body Payment to standard 4
setting body*
333 Agriculture Miscellaneous contributed 3
funds
334 Interior Bureau of Reclamation loan 3
liquidating account
335 General Services Disposal of surplus real 3
Administration and related personal
property
336 Legislative Branch Gifts and donations 3
337 Commerce National Oceanic and 3
Atmospheric Administration
Commissioned Officer Corps
retirement*
338 Treasury Informant payments* 3
339 Legislative Branch Judiciary office building 3
development and operations
fund*
340 Department of State Foreign Service national 3
separation liability trust
fund
341 Treasury Refunds, transfers and 3
expenses, Unclaimed and
abandoned goods*
342 Agriculture Wetlands reserve program 3
343 Agriculture Expenses and refunds, 3
inspection and grading of
farm products
344 Transportation Operations and maintenance 3
(Harbor services fee
collections)*
345 Veterans Affairs General post fund, 3
national homes
346 Judicial Branch Judicial officers' 3
retirement fund
347 National Archives and National archives gift 3
Records Administration fund*
348 Office of Personnel Government payment for 3
Management annuitants, employee life
insurance
349 Vietnam Education Vietnam debt repayment 3
Foundation fund*
350 Agriculture Miscellaneous contributed 2
funds
351 Commerce Damage assessment and 2
restoration revolving fund
352 Housing and Urban Manufactured home 2
Development inspection and monitoring*
353 Interior Migratory bird 2
conservation account
354 Interior Indian direct loan program 2
account*
355 Transportation Operations and 2
maintenance*
356 General Services Expenses of transportation 2
Administration audit contracts and
contract administration
357 International Assistance Miscellaneous trust funds, 2
Programs AID*
358 Commerce Fisheries finance program 2
account*
359 Department of Other DOD trust revolving 2
Defense-Military funds*
360 Health and Human Medical facilities 2
Services guarantee and loan fund
361 Veterans Affairs Service members' group 2
life insurance fund*
362 Farm Credit Revolving fund for 2
Administration administrative expenses*
363 Federal Housing Finance Federal housing finance 2
Board board*
364 Judicial Branch Registry Administration 2
365 Legislative Branch Compensation of members 2
and related administrative
expenses
366 Small Business Pollution control 2
Administration equipment fund liquidating
account*
367 Corps of Engineers-Civil Permanent appropriations 2
Works
368 Corps of Engineers-Civil Payment to South Dakota 2
Works terrestrial wildlife
habitat restoration trust
fund*
369 Commerce Limited access system 2
administration fund*
370 Homeland Security US Customs Refunds, 2
Transfers and Expenses,
Unclaimed and Abandoned
Goods*
371 Agriculture Wildlife habitat 2
incentives program
372 Veterans Affairs Insurance benefits 2
373 District of Columbia Federal payment to the 2
District of Columbia
judicial retirement and
survivors annuity fund*
374 Morris K. Udall Environmental dispute 2
Scholarship and resolution fund*
Excellence in National
Environmental Policy
Foundation
375 Agriculture Farm storage facility 2
loans program account*
376 Department of Miscellaneous special 2
Defense-Military funds*
377 Housing and Urban Elderly vouchers* 2
Development
378 Department of State International litigation 2
fund
379 Interior National Indian Gaming 2
Commission, Gaming
activity fees
380 Treasury Interest on uninvested 2
funds
381 Transportation Railroad rehabilitation 2
and improvement
liquidating account
382 Environmental Protection Abatement, control, and 2
Agency compliance direct loan
liquidating account*
383 Panama Canal Commission Panama Canal Commission 2
dissolution fund*
384 Interior White Earth settlement 1
fund
385 Interior Indian loan guaranty and 1
insurance fund liquidating
account*
386 Transportation Emergency preparedness 1
grants
387 Legislative Branch Congressional use of 1
foreign currency, House of
Representatives
388 Agriculture Rural communication 1
development fund
liquidating account*
389 Commerce Franchise fund* 1
390 Justice United States trustee 1
system fund*
391 Interior Revolving fund for loans 1
liquidating account
392 Interior National forests fund, 1
Payment to States
393 Interior Recreational fee program 1
394 Interior Miscellaneous trust funds 1
395 Treasury Federal tax lien revolving 1
fund*
396 International Assistance Microenterprise and small 1
Programs enterprise development
program account*
397 Interior Contributed funds 1
398 Interior Miscellaneous permanent 1
appropriations
399 Treasury Check forgery insurance 1
fund
400 Treasury Payment to terrestrial 1
wildlife habitat
restoration trust fund*
401 District of Columbia District of Columbia 1
judicial retirement and
survivors annuity fund
402 Environmental Protection Abatement, control, and 1
Agency compliance loan program
account*
403 Harry S. Truman Harry S. Truman memorial 1
Scholarship Foundation scholarship trust fund
404 James Madison Memorial James Madison Memorial 1
Fellowship Foundation Fellowship trust fund
405 Legislative Branch John C. Stennis Center for 1
Public Service Training
and Development trust fund
406 National Archives and National archives trust 1
Records Administration fund*
407 Agriculture Agricultural resource 1
conservation demonstration
program account*
408 Agriculture Rural economic development 1
loans program account*
409 Housing and Urban Manufactured housing fees 1
Development trust fund*
410 Labor Panama Canal Commission 1
compensation fund
411 Department of State Miscellaneous trust funds, 1
information and exchange
programs*
412 Interior Operation and maintenance 1
of quarters
413 Veterans Affairs Special therapeutic and 1
rehabilitation activities
fund*
414 Equal Employment EEOC education, technical 1
Opportunity Commission assistance, and training
revolving fund*
415 International Assistance Peace Corps miscellaneous 1
Programs trust fund
416 Judicial Branch Judicial survivors' 1
annuities fund
417 Legislative Branch Congressional use of 1
foreign currency, Senate
418 Social Security Special benefits for 1
Administration certain World War II
veterans*
419 Appalachian Regional Miscellaneous trust funds 1
Commission
420 Broadcasting Board of Foreign Service national 1
Governors separation liability trust
fund*
421 Christopher Columbus Christopher Columbus 1
Fellowship Foundation Fellowship Foundation*
422 Agriculture Limitation on inspection 1
and weighing services
expenses*
423 Agriculture Rural development loan 1
fund liquidating account
424 Agriculture Biomass research and 1
development*
425 Energy Payments to States under 1
Federal Power Act
426 Health and Human Revolving fund for 1
Services certification and other
services*
427 Labor Working capital fund* 1
428 Interior Range improvements 1
429 Interior Cooperative endangered 1
species conservation fund*
430 Interior Everglades restoration 1
account
431 Interior Leases of lands acquired 1
for flood control,
navigation, and allied
purposes
432 Interior Contributed funds* 1
433 Treasury Payment of Government 1
losses in shipment*
434 Treasury Terrorism insurance 1
program*
435 Transportation Amtrak corridor 1
improvement loans
liquidating account*
436 Federal Emergency National insurance 1
Management Agency development fund*
437 International Assistance Property management fund* 1
Programs
438 International Assistance Foreign Service national 1
Programs separation liability trust
fund
439 Judicial Branch Gifts and donations, 1
Federal Judicial Center
Foundation
440 Legislative Branch Senate revolving funds* 1
441 Morris K. Udall Morris K. Udall 1
Scholarship and Scholarship and Excellence
Excellence in National in National Environmental
Environmental Policy Policy Foundation*
Foundation
442 Other Defense Civil White House commission on 1
Programs the national moment of
remembrance*
443 Telecommunications Telecommunications 1
Development Fund development fund*
444 Agriculture Distance learning, 0
telemedicine, and
broadband program*
445 Agriculture Rural economic development 0
loans liquidating account*
446 Agriculture Facilities acquisition and 0
enhancement fund*
447 Housing and Urban Homeownership assistance 0
Development fund*
448 Housing and Urban Consolidated fee fund* 0
Development
449 Department of State International Center, 0
Washington, D.C.*
450 Interior Donations and contributed 0
funds*
451 Treasury Collection Contractor 0
Support*
452 Veterans Affairs United States Government 0
life insurance fund
453 Veterans Affairs Medical facilities 0
revolving fund*
454 Veterans Affairs Veterans extended care 0
revolving fund*
455 International Assistance Kuwait civil 0
Programs reconstruction trust fund*
456 Judicial Branch United States Court of 0
Federal Claims Judges'
retirement fund
457 Legislative Branch Compensation of members, 0
Senate
458 Legislative Branch Tax Court judges survivors 0
annuity fund*
459 Agriculture Emergency boll weevil loan 0
program account*
460 Agriculture Gifts and bequests 0
461 Agriculture Apple loans program 0
account*
462 Agriculture Rural community fire 0
protection grants*
463 Agriculture National sheep industry 0
improvement center
revolving fund*
464 Agriculture Land acquisition 0
reinvestment fund*
465 Energy Continuing fund, 0
Southwestern Power
Administration*
466 Energy Advances for cooperative 0
work*
467 Housing and Urban Interstate land sales* 0
Development
468 Department of State Foreign service national 0
defined contributions
retirement fund*
469 Department of State USIA Foreign Service 0
national separation
liability trust fund*
470 Interior Dutch John community 0
assistance*
471 Transportation Saint Lawrence Seaway 0
Development Corporation*
472 Transportation Minority business resource 0
center program*
473 Federal Financial Registry fees 0
Institutions Examination
Council Appraisal
Subcommittee
474 International Assistance Private sector revolving 0
Programs fund liquidating account*
475 Japan-United States Japan-United States 0
Friendship Commission Friendship trust fund
476 National Credit Union Community development 0
Administration credit union revolving
loan fund*
477 Other Defense Civil Wildlife conservation 0
Programs
478 Other Defense Civil Soldiers' and airmen's 0
Programs home revolving fund*
479 Other Defense Civil White House commission on 0
Programs the national moment of
remembrance*
480 Allowances Contingent offset for the 0
refundable portion of the
health care tax credit*
481 Barry Goldwater Barry Goldwater 0
Scholarship and Scholarship and Excellence
Excellence in Education in Education Foundation
Foundation
482 Agriculture P.L. 480 title I food for 0
progress credits, program
account*
483 Commerce Gifts and bequests 0
484 Department of Concurrent receipt accrual 0
Defense-Military payments to the Military
Retirement Fund*
485 Department of Restoration of the Rocky 0
Defense-Military Mountain Arsenal*
486 Education Reading excellence* 0
487 Education School construction* 0
488 Education Class size reduction and 0
teacher financing*
489 Education Perkins loan revolving 0
fund*
490 Education Federal family education 0
loan insurance fund*
491 Health and Human State legalization impact 0
Services assistance grants*
492 Health and Human Health maintenance 0
Services organization loan and loan
guarantee fund*
493 Homeland Security Disaster assistance direct 0
loan program account*
494 Housing and Urban Empowerment 0
Development zones/enterprise
communities*
495 Justice Civil liberties public 0
education fund*
496 Labor Foreign labor 0
certification processing*
497 Interior Bureau of Reclamation loan 0
program account*
498 Interior Miscellaneous permanent 0
appropriations*
499 Interior Payment to tribe, Lower 0
Brule Sioux Trust Fund*
500 Interior Miscellaneous Indian trust 0
payments*
501 Interior Operation and maintenance 0
of quarters*
502 Interior Operation and maintenance 0
of quarters*
503 Interior Fee collection support, 0
national park system*
504 Interior National park renewal 0
fund*
505 Interior Concessions improvement 0
accounts*
506 Interior Park concessions franchise 0
fees*
507 Interior African elephant 0
conservation fund*
508 Interior Miscellaneous permanent 0
appropriations*
509 Treasury Payment to Justice, FIRREA 0
related claims*
510 Treasury Payments to the farm 0
credit system financial
assistance corporation
liquidating account*
511 Treasury Miscellaneous activities 0
to be authorized in
tobacco legislation*
512 Treasury Miscellaneous permanent 0
appropriations*
513 Transportation Railroad rehabilitation 0
and improvement program*
514 Transportation Aviation user fees 0
515 Veterans Affairs Veterans housing benefit 0
program fund*
516 Veterans Affairs Reinstated entitlement 0
program for survivors
under P.L. 97-377*
517 Veterans Affairs Miscellaneous veterans 0
housing loans guaranteed
loan financing account*
518 Veterans Affairs Medical care cost recovery 0
fund*
519 District of Columbia District of Columbia 0
Federal pension fund*
520 District of Columbia Federal payment for water 0
and sewer services*
521 Environmental Protection Revolving fund for 0
Agency certification and other
services*
522 General Services Working capital fund* 0
Administration
523 General Services Acquisition workforce 0
Administration training fund*
524 General Services Pennsylvania Avenue 0
Administration activities*
525 General Services Land acquisition and 0
Administration development fund*
526 International Assistance Payment to the Foreign 0
Programs Service retirement and
disability fund
527 Judicial Branch Payment to judiciary trust 0
funds
528 Legislative Branch Gifts and donations* 0
529 Legislative Branch United States Capitol 0
Police memorial fund*
530 National Aeronautics and National Space Grant 0
Space Administration Program*
531 National Aeronautics and Science, space, and 0
Space Administration technology education trust
fund
532 Other Independent Foreign service national 0
Agencies separation liability trust
fund*
533 Other Independent Miscellaneous trust funds* 0
Agencies
534 Tennessee Valley Tennessee Valley Authority 0
Authority Office of the Inspector
General*
Total $162,987
Source: GAO analysis of President's budget data.
Note: The shaded rows indicate the 7 case study accounts discussed in
appendix I. Accounts with fewer than 5 years of data are denoted with an
"*."
GAO Contact and Acknowledgments Appendix IV
Susan J. Irving, (202) 512-9142, [email protected]
In addition to the individual named above, Christine Bonham, Assistant
Director, as well as Carol Henn, Richard Krashevski, Leah Nash, Sheila
Rajabiun, Paul Posner, and Stephanie Wade made key contributions to this
report.
(450410)
www.gao.gov/cgi-bin/getrpt? GAO-06-276 .
To view the full product, including the scope
and methodology, click on the link above.
For more information, contact Susan J. Irving, 202-512-9142,
[email protected].
Highlights of GAO-06-276 , a report to Congress
January 2006
MANDATORY SPENDING
Using Budget Triggers to Constrain Growth
Prepared as part of GAO's basic statutory responsibility for monitoring
the condition of the nation's finances, the objectives of this report were
to (1) determine the feasibility of designing and using trigger mechanisms
to constrain growth in mandatory spending programs and (2) provide an
analysis of the factors that led to differences between estimated and
actual outlays in seven mandatory budget accounts during fiscal years 2000
through 2004.
What GAO Recommends
To promote explicit scrutiny of significant growth in mandatory accounts,
as mandatory spending programs are created, reexamined, or reauthorized,
Congress should consider incorporating budget triggers that would signal
the need for action. Further, it should determine whether in some cases it
might be appropriate to consider automatically causing some action to be
taken when the trigger is exceeded. Once a trigger is tripped, Congress
could either accept or reject all or a portion of a proposed response to
the spending growth. The Office of Management and Budget and agencies
responsible for the seven case study accounts either did not have comments
or provided comments that were clarifying and/or technical in nature,
which were incorporated as appropriate.
One idea to constrain growth in mandatory programs is to develop
program-specific triggers that, when tripped, prompt a response. A trigger
could result in a "hard" or automatic response, unless Congress and the
President acted to override or alter it. Alternatively, reaching a trigger
could require a "soft" response, such as a report on the causes of the
overage, development of a plan to address it, or an explicit and formal
decision to accept or reject a proposed action or increase. By identifying
significant increases in the spending path of a mandatory program
relatively early and acting to constrain it, Congress may avert larger
financial challenges in the future. However, both in establishing triggers
and in designing the subsequent responses, the integrity of program goals
needs to be preserved. In addition, tax expenditures operate like
mandatory programs but do not compete in the annual appropriations
process. The analysis GAO applied to spending in this report would also be
useful in examining tax expenditures.
The budget experts GAO consulted had mixed views of triggers. Proponents
of triggers noted that mandatory spending is currently unconstrained and a
mechanism that causes decision makers to at least periodically reevaluate
spending is better than allowing spending to rise unchecked. Others,
however, expressed considerable skepticism about the effectiveness of
triggers; many felt they would either be circumvented or ignored. While
GAO appreciates the views expressed by budget experts, in our opinion
establishing budget triggers warrants consideration in efforts to
constrain significant and largely unchecked growth in mandatory programs.
However, recognizing the natural tension in balancing both long-term
fiscal challenges and other public policy goals, each program needs to be
considered individually to ensure that any responses triggered strike the
appropriate balance between the long-term fiscal challenge and the program
goals.
To better understand growth in mandatory spending and thus inform GAO's
thinking on triggers, for seven case study accounts GAO categorized the
reasons provided by agencies for differences between estimated and actual
outlays during a 5-year period as the result of legislative, economic, or
technical changes. Out of 40 differences, subsequent legislation was the
primary reason for 19, economic changes for 7, and technical changes for
13. In many cases, a combination of these factors caused the differences.
Conceptual Differences between Hard and Soft Responses
*** End of document. ***