Budget Reconciliation: Issues Concerning the 1990 Reconciliation Act
(Letter Report, 10/07/94, GAO/AIMD-95-3).

When the Omnibus Budget Reconciliation Act of 1990 was passed, overall
savings from the multiyear deficit reduction agreement were pegged at
about $500 billion. However, the deficit has not dropped to the levels
expected at that time. For example, the Congressional Budget Office
(CBO) projections in late 1990 showed that the fiscal year 1993 deficit
would be $170 billion, but the actual deficit was $255 billion. This
report focuses on issues surrounding one component of the deficit
reduction agreement--$98.8 billion in mandatory spending program
reductions and increases in user fee collections. GAO (1) examines the
reasons for shortfalls between deficits projected at the time of the act
and actual deficits experienced since then, (2) ascertains whether
actual results of the 38 individual act provisions GAO examined were
achieved, and (3) describes CBO and agency roles and responsibilities in
the act's reconciliation process as they affected the act's contribution
to deficit reduction.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  AIMD-95-3
     TITLE:  Budget Reconciliation: Issues Concerning the 1990 
             Reconciliation Act
      DATE:  10/07/94
   SUBJECT:  Deficit reduction
             Fiscal policies
             Budget deficit
             Balanced budgets
             Budget outlays
             Budget administration
             Future budget projections
             Macroeconomic analysis
             Program management
IDENTIFIER:  Food Stamp Program
             Unemployment Insurance Program
             Medicaid Program
             FEMA National Flood Insurance Program
             Civil Service Retirement and Disability Fund
             Medicare Prospective Payment System
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed 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.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Report to Congressional Requesters

October 1994

BUDGET PROCESS - ISSUES CONCERNING
THE 1990 RECONCILIATION ACT

GAO/AIMD-95-3

Budget Reconciliation

(935126)


Abbreviations
=============================================================== ABBREV

  ARP - Acreage Reduction Program
  ASCS - Agricultural Stabilization and Conservation Service
  BEA - Budget Enforcement Act of 1990
  BIF - Bank Insurance Fund
  CBO - Congressional Budget Office
  COLA - cost-of-living adjustment
  CSRS - Civil Service Retirement System
  DOD - Department of Defense
  FCIP - Federal Crime Insurance Program
  FDIC - Federal Deposit Insurance Corporation
  FEHB - Federal Employee Health Benefits
  FEMA - Federal Emergency Management Agency
  FIA - Federal Insurance Administration
  FIRREA - Financial Institutions Reform, Recovery, and Enforcement
     Act of 1989
  FmHA - Farmers Home Administration
  FY - fiscal year
  GDP - gross domestic product
  GNP - gross national product
  HCFA - Health Care Financing Administration
  IRS - Internal Revenue Service
  NFIP - National Flood Insurance Program
  NRC - Nuclear Regulatory Commission
  OBRA 90 - Omnibus Budget Reconciliation Act of 1990
  OBRA 93 - Omnibus Budget Reconciliation Act of 1993
  OMB - Office of Management and Budget
  OPM - Office of Personnel Management
  PAYGO - pay-as-you-go
  PPS - Prospective Payment System
  PTO - Patent and Trademark Office
  SSA - Social Security Administration
  TEFRA - Tax Equity and Fiscal Responsibility Act of 1982
  USTTA - United States Travel and Tourism Administration
  VA - Department of Veterans Affairs

Letter
=============================================================== LETTER


B-256700

October 7, 1994

The Honorable Robert H.  Michel
Republican Leader
House of Representatives

The Honorable Newt Gingrich
Republican Whip
House of Representatives

The Honorable Bill Archer
Ranking Republican Member, Committee on Ways and Means
House of Representatives

The Honorable John R.  Kasich
Ranking Republican Member, Committee on the Budget
House of Representatives

The Honorable James A.  Leach
Ranking Republican Member, Committee on Banking, Finance
 and Urban Affairs
House of Representatives

This report responds to your request that we examine the
reconciliation process as it affects deficit reduction.  At the time
of the Omnibus Budget Reconciliation Act of 1990 (OBRA 90) enactment,
overall savings from the multiyear deficit reduction agreement were
estimated to be about $500 billion.  However, the deficit has not
dropped to the levels expected at that time.  For example, the
Congressional Budget Office's (CBO) projections in December 1990
showed that the fiscal year 1993 deficit would be $170 billion, but
the actual deficit was $255 billion. 

The components of the almost $500 billion 5-year (fiscal years 1991
through 1995) deficit reduction agreement enacted as OBRA 90 were (1)
mandatory spending program reductions and increases in user fee
collections ($98.8 billion), (2) revenue increases ($137.2 billion),
and (3) discretionary spending caps which lowered program spending
($182.4 billion).  Additional deficit reduction was to come from
increased Internal Revenue Service (IRS) enforcement ($9.4 billion)
and from the interest savings that result from financing a lower debt
($68.4 billion). 

You asked us to examine certain issues surrounding the first
component:  $98.8 billion in mandatory program spending reductions
and increases in user fee collections.  As agreed with your offices,
this report (1) examines the reasons for the shortfall between
deficits projected at the time of OBRA 90 and actual deficits
experienced since then,\1 (2) ascertains, if possible, whether actual
results of the 38 individual OBRA 90 provisions we examined were
achieved, and (3) describes CBO and agency roles and responsibilities
in the OBRA 90 reconciliation process as they affected OBRA 90's
contribution to deficit reduction. 


--------------------
\1 These are actual deficits for fiscal years 1991 through 1993 and
CBO projected deficits, as of August 1994, for fiscal years 1994 and
1995. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

Looking back at aggregate levels of spending and revenues for fiscal
years 1991 through 1995, worse-than-anticipated economic conditions
was the primary reason that deficits have been greater than projected
at the time of OBRA 90's passage.  In particular,
slower-than-expected economic growth for fiscal years 1991 through
1993 caused federal revenues to decline and spending to increase by
more than the savings projected from OBRA 90's policy changes. 

OBRA 90's policies to reduce mandatory program spending and increase
user fee collections contributed significantly to reducing the
deficit below what it otherwise would have been.  CBO did what
reasonably could be expected to provide the Congress accurate
estimates, and executive branch agencies implemented the provisions
promptly.  Moreover, agency officials' knowledge of their programs
led them to conclude that 32 of 38 OBRA 90 provisions we
examined--constituting 84 percent of the estimated savings of the
programs we examined--did reduce the deficit.  The remaining six
provisions did not, according to agency officials, reduce the deficit
below what it otherwise would have been. 

In most cases, though, agencies could not determine the precise
savings for individual reconciliation provisions.  It is generally
not possible to identify precise savings because isolating the
budgetary impact of individual reconciliation provisions from the
impacts on spending of related OBRA 90 provisions, subsequent
legislation, administrative actions, and behavioral responses cannot
be done accurately.  Nevertheless, we agree that savings have
resulted. 

Despite OBRA 90's contributions to deficit reduction, deficits did
not drop to the levels expected when OBRA 90 was enacted--an outcome
that could frustrate deficit reduction efforts if policymakers come
to believe that the difficult choices they make do not matter. 
However, this does not mean that deficits greater than those
originally anticipated must be accepted.  Under the current process,
the Congress can at any time enact changes in individual mandatory
programs, either alone or as part of reconciliation, to reduce the
deficit.  There is, however, currently no procedure that
automatically triggers a comparison of the current deficit path to
that which was expected when action was taken.  The Congress could
adopt a requirement to look back periodically and compare the current
deficit path to that projected at the time of a prior deficit
reduction agreement and/or the most recent reconciliation
legislation.  If the deficit was greater than that projected, the
Congress would decide explicitly--by voting--whether to accept the
slippage or take action to bring the deficit path closer to the
original goal by recouping some or all of the slippage. 


   BACKGROUND
------------------------------------------------------------ Letter :2

Reconciliation is a budget process tool that the Congress can use to
reduce the deficit by requiring reductions in existing mandatory
spending programs\2 or increases in revenues.  Reconciliation bills
have also been used to enact budget process changes.  For example,
the Budget Enforcement Act of 1990 (BEA) was included as Title XIII
in OBRA 90.  The pay-as-you-go (PAYGO) provisions of BEA limit the
creation of new or the expansion of existing mandatory programs and
reductions in revenues.  However, PAYGO was not designed to restrain
growth in mandatory programs driven by events other than legislative
changes to those programs. 

In reconciliation, the Congress directs authorizing committees to
propose changes in mandatory programs and/or tax laws within the
committees' respective legislative jurisdictions to meet aggregate
spending and revenue targets established in the concurrent resolution
on the budget (referred to as the budget resolution).\3 The budget
resolution may call for reductions in mandatory spending or increases
in revenues.  However, the resolution does not specify the policy
changes committees must make, but instead leaves substantive
decisions on how to meet dollar targets to the committees of
jurisdiction.  Recommendations reported by each committee are
packaged into individual reconciliation bills by the House and Senate
budget committees and voted on in each chamber.  After conference
committee action to reconcile any differences, reconciliation bills
are often passed by the Congress as single, omnibus deficit reduction
measures and sent to the President.  Since 1980, when reconciliation
was first used in this way, 11 reconciliation bills have been
enacted.  OBRA 90 was one of the largest, with an estimated $496.2
billion in deficit reduction savings over 5 years.  (See appendix I.)

To fulfill its responsibilities under the Congressional Budget and
Impoundment Control Act of 1974 (the Budget Act), CBO must estimate
the budgetary impact of all proposed legislation, including
reconciliation bills.  CBO starts with what is called a current
policy baseline.  This baseline assumes no changes in laws governing
revenues and mandatory spending programs.  Estimates for revenue and
mandatory program spending are based on the economic assumptions of
the baseline and reflect such technical factors as changes in the
numbers of beneficiaries.  Prior to OBRA 90, the baseline for
discretionary program spending assumed spending increased at the rate
of inflation assumed in the budget resolution.  Since the enactment
of BEA's caps on discretionary spending, the baseline for
discretionary spending is set at the level of the caps.\4 CBO cost
estimates, including those for reconciliation, show how proposed
legislation would change projected outlays from the baseline.\5 These
estimates are based on the same economic and technical assumptions as
the baseline and as the budget resolution.  The Budget Committees use
these estimates to ensure that committees of jurisdiction have met
their reconciliation targets.  After the President signs a
reconciliation bill, executive branch agencies must implement the
prescribed policy changes. 


--------------------
\2 Mandatory spending is also known as direct spending.  The Budget
Enforcement Act (BEA) defines direct spending as spending for
entitlement authority, the Food Stamp program, and budget authority
provided by law other than appropriations acts.  See Budget Policy: 
Issues In Capping Mandatory Spending (GAO/AIMD-94-155, July 18,
1994). 

\3 The budget resolution is not a law.  It is not signed by the
President. 

\4 Since the discretionary spending caps expire after 1998, CBO
assumes that the discretionary spending baseline increases at the
rate of inflation thereafter.  H.R.  4907, the Full Budget Disclosure
Act, passed by the House on August 12, 1994, would eliminate the use
of the inflation adjustment for discretionary spending and instead
require the use of the most recent actual appropriations.  This bill
would not affect the current policy baseline for mandatory programs. 

\5 Section 201 of the Budget Act requires CBO to use Joint Committee
on Taxation estimates for the impact of changes in revenue
legislation affecting income, estate and gift, excise, and payroll
taxes. 


   SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :3

As agreed with your offices, we limited our examination of OBRA 90
provisions to the act's 11 spending titles, excluding the revenue
titles.  At the time of OBRA 90's enactment, these policy changes in
mandatory spending programs and user fee collections were estimated
to reduce the deficit by almost $100 billion.\6 Appendix II
summarizes the results for the 38 provisions that were part of this
study and provides the provisions' section numbers in the public law,
CBO 5-year savings estimates, and identifies executive branch
agencies responsible for implementation.  Initially, we used the text
of OBRA 90 (Public Law 101-508), House and Senate Budget Committee
documents, CBO documents, and the Congressional Quarterly Almanac to
prepare an inventory of the 270 provisions contained in OBRA 90's 11
spending titles.  Then, we eliminated 192 provisions that were not
directed toward program spending reduction.\7 From the remaining 78,
we selected the 8 highest dollar provisions, which together totaled
about 50 percent of the estimated 5-year savings for mandatory
programs/user fees, and a random sample of 30 more.  Estimated
savings for the 38 provisions totaled $78.9 billion--the 8 highest
were estimated to save $55.8 billion and the 30 were estimated to
save $23.1 billion. 

To answer your question about the reasons for the shortfall between
deficits projected at the time of OBRA 90 and subsequent deficits, we
reviewed CBO publications on baseline projections of the federal
budget and spoke with CBO officials. 

To answer your question about actual savings from individual OBRA 90
reconciliation provisions, we reviewed literature on statistical
methodology to understand the types of problems CBO and agencies
would encounter in isolating the budgetary impact of OBRA 90
provisions on spending from that of other factors.  We also developed
a CBO and an agency interview questionnaire concerning roles and
responsibilities in reconciliation processes and in OBRA 90.  We used
these questions to interview CBO analysts and agency officials
regarding the 38 provisions. 

We asked agency officials with specific legal, programmatic, and/or
budgetary expertise to assess savings in two ways.  First, we asked
agency officials if they were able to identify the specific savings
achieved by individual OBRA 90 provisions and if those savings
matched original CBO estimates.  Second, we asked agency officials if
each of the 38 provisions produced some measure of savings through
increased collections or reduced program spending since either would
help to reduce the deficit. 

We also used our interview questionnaire and information to answer
your questions on what could reasonably be expected of CBO and
executive branch agencies.  At CBO, we spoke with analysts about CBO
estimating practices, including the types of data and models used to
prepare cost estimates.  At agencies, we applied a two-tiered test of
implementation.  The first test, a threshold test, determined whether
the agency issued a formal rule or binding guidance in lieu of a
formal rule when required to do so.  By formal rule, we mean the
process by which an agency publishes regulations in the Federal
Register.  Examples of binding guidance in lieu of a formal rule
include changing agency manuals or issuing agency directives.  The
second test was based upon the timeliness and scope of the formal
rule or binding guidance in lieu of the rule.  By timeliness, we mean
whether an agency issued the formal rule or binding guidance in lieu
of the rule within the first fiscal year of applicability.  By scope,
we mean whether the agency's implementation had obvious omissions
from or additions to the OBRA 90 provision.  For other fiscal years
of applicability, we consulted with agencies to determine if revised
rules or guidance were needed and if so, whether they were issued. 

We did not attempt to reestimate or audit CBO data discussed in this
report.  Further, we did not attempt to audit in detail the
thoroughness or effectiveness of agency implementation efforts and
accuracy of savings measurements. 

We did this work in Washington, D.C., from September 1993 through
July 1994.  We discussed with officials of the executive branch
agencies the facts in the report that describe the OBRA 90 provisions
implemented by their agencies.  We discussed the facts, conclusions,
and matter for congressional consideration with CBO staff, who
generally concurred with our analysis.  We also discussed these items
with OMB staff.  We have incorporated their views, along with those
of the executive branch agencies, where appropriate. 


--------------------
\6 The components of the almost $500 billion 5-year deficit reduction
agreement enacted as OBRA 90 were (1) mandatory spending program
reductions and increases in user fee collections ($98.8 billion), (2)
revenue increases ($137.2 billion), and (3) discretionary spending
caps which lowered program spending ($182.4 billion).  Additional
deficit reduction was to come from increased IRS enforcement ($9.4
billion) and from the interest savings that result from financing a
lower debt ($68.4 billion). 

\7 The 192 provisions break down as follows:  1 was a revenue
provision included in a spending title, 56 increased program
spending, and 135 were changes such as conforming language,
renumbering of law sections, or expressing a sense of the Congress. 


   SAVINGS ANTICIPATED WERE OFFSET
   BY WORSE-THAN- ANTICIPATED
   ECONOMIC CONDITIONS
------------------------------------------------------------ Letter :4

Actual deficits for fiscal years 1991 through 1993 and projected
deficits for fiscal years 1994 and 1995 are higher than those
expected when OBRA 90 was enacted.  CBO estimated that OBRA 90's
policy changes had a cumulative 5-year deficit impact of $496
billion.  However, other factors intervened, offsetting the estimated
deficit reduction at the time of OBRA 90's enactment.  For example,
CBO's December 1990 deficit projections--which included the impact of
OBRA 90--showed a fiscal year 1993 deficit of $170 billion, but the
actual deficit was $255 billion. 

To measure changes in the deficit, CBO divides factors that affect
changes in aggregate outlays and revenues into three
categories--economic factors, policy changes (new legislation), and
technical factors.\8 Table 1 shows changes in CBO's deficit
projections since July 1990. 



                                Table 1
                
                Changes in CBO Deficit Projections Since
                               July 1990

                         (Dollars in billions)

                                    FY      FY      FY      FY      FY
                                  1991    1992    1993    1994    1995
------------------------------  ------  ------  ------  ------  ------
CBO's July 1990 deficit            232     238     196     145     138
 projections
CBO estimates of deficit           -35     -73     -91    -134    -163
 reduction from OBRA 90 policy
 changes
Economic changes (after 7/90)       41      55      60      62      64
Technical changes (after 7/         16      42       5     -17     -10
 90)
CBO December 1990 deficit          254     262     170      56      29
 projections

CBO estimates of changes subsequent to December 1990
----------------------------------------------------------------------
Policy\a                           -19      13      13     -18     -46
Economic                            32      49      47      30      45
Technical
Deposit insurance                  -28    -108     -56      38      11
Medicare and Medicaid                7      19      24      30      46
Major benefit                        8      18      25      25      29
 programs
Other technical                     16      37      32      41      48
Actual deficits and CBO's          270     290     255     202     162
 August 1994 deficit
 projections
----------------------------------------------------------------------
Notes:  (1) Deficit projections reflect the combined or net effect of
revenues and outlays.

   (2) Reestimates with a minus sign (-) decrease the deficit. 

\a This includes the estimated effects of the Omnibus Budget
Reconciliation Act of 1993 (OBRA 93) for fiscal years 1994 and 1995. 

Sources:  CBO and GAO. 

As the table shows, the single largest change--estimated to be $485
billion over 5 years--was due to the economy.  Economic
conditions--notably economic growth and unemployment--turned out to
be significantly worse than forecast either just before OBRA 90's
enactment or in CBO's December 1990 update.  However, the CBO
forecast underlying the OBRA 90 estimates was similar to that of the
consensus of private forecasts and less optimistic than that of OMB. 
(Appendix III provides more details on these economic forecasts.) As
figure 1 shows, CBO, OMB, and private forecasts were all overly
optimistic. 

   Figure 1:  Comparison of CBO,
   Administration, and Private
   Sector Forecasts of Real GNP
   Growth at the Time of OBRA 90

   (See figure in printed
   edition.)

Note:  Blue Chip forecasts are based on a survey of private sector
forecasts. 

Moreover, although CBO was aware that the economy was worsening
during the fall of 1990, CBO's estimates of OBRA 90 policy changes
were based on the economic assumptions underlying its July 1990
baseline.  This is because CBO is required to score
legislation--including reconciliation bills--on the economic
assumptions in place at the time of the most recent budget
resolution.  CBO provided OBRA 90 savings estimates, but the next
complete deficit update was made in December 1990.  These deficit
projections included both the effects of OBRA 90 and economic and
technical changes since July 1990. 

The federal budget deficit is very responsive to the economy,
especially through the economy's effects on revenues.  Differences
between actual economic conditions and those forecast can increase or
decrease the deficit compared to the expected level.  CBO estimates
that a 1 percentage point reduction in real economic growth adds
approximately $30 billion to that fiscal year's deficit--about $26
billion in lower revenues and $4 billion in increased outlays.\9
Conversely, a 1 percentage point increase in real economic growth
decreases the deficit by approximately $30 billion. 

In the period following OBRA 90's passage, the lower-than-
anticipated real economic growth shown in figure 1 (with its adverse
impact on employment) reduced personal income, particularly wages and
salaries, thereby reducing income and payroll tax revenues.  Outlays
for programs such as Unemployment Insurance, Food Stamps, and
Medicaid were also affected because spending in these programs
increases when unemployment rates rise. 

In contrast, in the period since the passage of OBRA 93,
greater-than-expected economic growth and lower unemployment have
accelerated deficit reduction.  For example, immediately following
OBRA 93's enactment, CBO projected that the fiscal year 1995 deficit
would be $196 billion.  Currently however, CBO projects a 1995
deficit of $162 billion. 

Technical factors also drove up spending during the period following
OBRA 90's passage, as table 1 shows.  For example, unanticipated
changes in Medicaid spending resulted from greater-than-anticipated
increases in payments to hospitals. 

Nevertheless, had OBRA 90 not been enacted, deficits would have been
higher.  In each fiscal year from 1991 through 1995 shown in table 1,
actual deficits or CBO deficit projections would have been higher
than without OBRA 90's policy changes. 


--------------------
\8 The technical category is a residual, including all discrepancies
that are neither traceable to legislation nor rigorously linked to
changes in the economy.  These factors include new information about
changes, such as program participants' behavior, projections about
tax collections, or participation rates for entitlement programs. 
For example, unanticipated changes in Medicare utilization rates are
classified as technical changes.  Given the size and complexity of
the budget, technical changes are inevitable. 

\9 Rules of thumb for estimating the effect on the deficit of changes
in various macroeconomic variables are given in CBO, The Economic and
Budget Outlook:  Fiscal Years 1995-1999 (January 1994), pp.75-78. 


   THE MAJORITY OF
   OBRA 90'S NEW POLICIES REDUCED
   THE DEFICIT
------------------------------------------------------------ Letter :5

The deficit is smaller today than it would have been without the
changes enacted in OBRA 90.  As reported by agency officials, 32 of
the 38 provisions we reviewed helped to reduce the deficit either by
reducing program outlays or increasing user fee collections. 
However, as discussed later, officials could not identify the
specific savings for individual provisions. 

Specifically, agency officials' knowledge of their programs led them
to conclude that these 32 provisions (1) directly reduced payments to
program providers, (2) directly eliminated program benefits, (3)
directly reduced payments to beneficiaries, (4) reduced the costs of
programs through efficiencies, (5) increased or levied new fees for
government services, or (6) increased premium rates or penalty fees
paid by program participants.  Two examples follow. 

  -- Office of Personnel Management (OPM) officials reported that the
     Federal Employee Health Benefits (FEHB) reforms provision\10
     (OBRA 90, section 7002) reduced the deficit because OBRA 90's
     policy reduced the costs of providing health benefits to federal
     employees.  Two components of the provision required health care
     carriers to implement cash management controls in the
     administration of FEHB funds and cost-containment measures to
     lower program costs.  For example, carriers must verify the
     medical necessity of treatments or surgeries, substitute
     outpatient for inpatient services when possible, determine the
     appropriate length of stay for inpatient treatment, and
     establish incentives to encourage compliance. 

  -- Officials at the Department of Labor's Mine Safety and Health
     Administration reported that the Mine Safety and Health
     provision
     (OBRA 90, section 3102) reduced the deficit because assessments
     for civil penalties were increased.  Before OBRA 90, the Federal
     Mine Safety and Health Act of 1977 imposed maximum fines of
     $10,000 per violation and an additional fine of $1,000 per
     violation per day if the violation was not corrected.  OBRA 90
     increased these five-fold.  Officials pointed out that, while
     not all fines were increased by the maximum amount, all
     categories of penalties were increased over what they had been
     prior to OBRA 90's passage. 


--------------------
\10 Each OBRA 90 provision described in this report is identified by
its public law section number.  In referring to each provision we
used abbreviated titles.  The full title can be found by referring to
the section number. 


      SOME OBRA 90 PROVISIONS DID
      NOT REDUCE THE DEFICIT
---------------------------------------------------------- Letter :5.1

Agency officials reported that six provisions, accounting for $12.5
billion, or 16 percent of estimated savings in our sample, would not
reduce the deficit.  These new policies did not reduce outlays or
increase user fee or premium collections.  One provision was
inconsistent with an international air agreement (OBRA 90, section
10301) and was not implemented.  Agency officials stated that other
provisions (1) mandated actions that the Federal Deposit Insurance
Corporation (FDIC) and Department of Agriculture had already taken or
were about to take under existing authorities (OBRA 90, sections
2002-2004 and 1103), (2) extended expiring authority for the Flood\11
and Crime Insurance Programs without otherwise changing them (OBRA
90, sections 2302 and 2301), or (3) required payments from an
off-budget entity (Postal Service) to an on-budget entity (Civil
Service Retirement and Disability Fund) (OBRA 90, section 7101).\12
We discuss these six provisions in greater detail in appendix IV,
including (1) the reasons why agency officials believed that the
provisions would not reduce the deficit and (2) the rationale for
scoring these provisions as savings. 

Each of the five provisions implemented was scored appropriately as
savings according to scoring conventions and baseline estimates of
federal outlays in effect at the time.  The baseline provides a
neutral and necessary starting point that CBO uses to hold constant
economic assumptions and technical factors, for example, demography
and behavior.  Without the baseline, the Congress would not have a
way to fairly compare the budgetary impacts of legislative
proposals--including reconciliation provisions--throughout the fiscal
year. 

Changes can and have been made in baseline conventions and scoring
practices.  For example, in OBRA 90 the Congress directed that
mandatory programs with outlays greater than $50 million should be
assumed to continue indefinitely.  This change in baseline
conventions means that future reauthorizations of the flood insurance
program (which typically generates net revenues) will not be scored
as producing savings.\13


--------------------
\11 The Flood Insurance provision has two components: 
reauthorization and premium increase.  CBO estimated their savings at
$594 million and $224 million, respectively.  Agency officials
reported that the reauthorization component did not reduce the
deficit. 

\12 Current law requires that the Social Security trust funds and the
Postal Service be off-budget--excluded from the totals of the
President's budget and the Congress' budget resolutions--even though
these funds are part of total government transactions.  Currently,
the Social Security trust funds and Postal Service are the only
off-budget entities; all other federal and trust funds are on-budget. 
Off-budget totals are added to on-budget totals to produce the
unified budget totals. 

\13 It also means that extensions of expiring mandatory programs will
not be scored as increases in spending. 


   THE SPECIFIC SAVINGS FROM
   INDIVIDUAL OBRA 90 PROVISIONS
   COULD NOT BE DETERMINED
------------------------------------------------------------ Letter :6

While OBRA 90's policies helped reduce the deficit, it is not
possible to determine the specific savings from each provision.  CBO
and executive branch agencies cannot isolate the budgetary impact of
these provisions from that of other influences--related OBRA 90
provisions, subsequent legislation, administrative actions, and
behavioral variables--on spending or collections.  However, because
agencies monitor program spending or user fee collections from year
to year, they can determine changes in spending or increases in
collections in the period since OBRA 90.  These changes represent the
net effect of all legislation, administrative action, and other
intervening variables on programs and therefore cannot be compared to
OBRA 90 savings estimates. 


      SPECIFIC SAVINGS COULD NOT
      BE DETERMINED IN THE
      MAJORITY OF CASES
---------------------------------------------------------- Letter :6.1

For 33 of the 38 provisions, accounting for almost $74.6 billion of
the estimated savings in our sample, agency officials could not
determine the specific savings resulting from OBRA 90 provisions. 
Agency officials cannot accurately isolate OBRA 90's discrete
budgetary impact from that of other factors affecting spending.  They
pointed out that outlays and user fee collections in their programs
are continually affected by factors such as legislation and
behavioral responses.  Agency officials told us that any effort to
isolate the budgetary impact of specific provisions would be
demanding, time-consuming, and subject to challenge.  The following
examples illustrate these problems. 

  -- The Prospective Payment System (PPS) Changes provision (OBRA 90,
     section 4002) reduced payment rates to hospitals for Medicare
     beneficiaries.  The Health Care Financing Administration (HCFA)
     administers the program under the Department of Health and Human
     Services.  HCFA officials told us that legislated and
     administrative changes occur too frequently to isolate the
     budgetary impact of a single change from others.  Also, HCFA
     officials we spoke with pointed out that Medicare outlays are
     affected by changes in behavior on the part of hospital
     administrators, physicians, and program participants, and are
     not necessarily the result of OBRA 90's new policies.\14 For
     example, lowering the payment rates for Medicare inpatient
     services would widen the gap between the actual cost of those
     services and what hospitals may bill for them.  In such a
     situation, hospitals might try to reduce associated financial
     losses by charging multiple or more expensive codes for their
     Medicare patients--for reasons unrelated to these patients'
     medical conditions.  Any resulting increase in Medicare spending
     from such code creep would reduce the amount of savings actually
     achieved from the PPS provision.  Moreover, the budgetary impact
     of such actions to increase hospital revenues cannot be sorted
     out from increased spending due to legitimate changes in medical
     practice such as those based on new information about successful
     techniques.  Both types of increased spending show up as
     increases in utilization.\15 Accordingly, attributing any
     particular dollar effect to OBRA 90 is impossible to do. 

  -- The Determination of Pension for Certain Veterans provision
     (OBRA 90, section 8002) eliminates the total disability
     presumption for veterans turning 65.  Prior to OBRA 90, any
     veteran 65 or older and meeting an income eligibility
     requirement was considered permanently and totally disabled for
     pension purposes.  Now these veterans may be considered
     employable.  Only proven disabilities--not age alone--qualify
     veterans for these benefits.  Department of Veterans Affairs
     (VA) officials reported that this change is likely to have
     affected the number of beneficiaries in two ways.  Some veterans
     apply and are turned down due to the policy change.  Other
     veterans are discouraged from applying for this benefit due to
     the policy change.  The number of veterans in the first group
     can be determined while the number in the second group cannot. 
     Both would be needed to determine savings resulting from this
     provision.  Interactive effects from other OBRA 90 provisions
     also made it impossible to isolate how much of the decline in
     the numbers of pensioners is due exclusively to this provision. 
     For example, the Income Verification Match provision (OBRA 90,
     section 8051) mandated that VA use IRS and Social Security
     Administration (SSA) data to verify eligibility for means-tested
     programs, including VA's health program.  According to VA
     officials, veterans might not have applied for pensions because
     of the income verification program.  Because of these
     interactions, identifying and assigning specific savings to both
     provisions is problematic. 

  -- OBRA 90, section 2005 authorized the FDIC to borrow working
     capital from the Federal Financing Bank.  The estimated savings
     from this provision were based on assumptions about interest
     rates and the amount that FDIC would need to borrow.  FDIC
     officials believed that this provision produced savings by
     allowing borrowing at a lower interest rate.  Two factors make
     it impossible to determine the specific savings.  First, there
     is uncertainty about how much higher the rate would have been if
     FDIC had borrowed from the private sector.  Second, improved
     bank profitability due to lower interest rates decreased FDIC's
     borrowing needs. 

Notwithstanding the inability to isolate the budgetary impact of OBRA
90 provisions, all agency officials we spoke with noted that they
monitored program spending from one fiscal year to the next to
determine how program costs and user fee collections change.  Thus,
agencies could identify changes in the period since OBRA 90. 
However, the results represented the net effect of all legislation,
administrative actions, and other intervening variables that affected
programs.  For this reason, the results cannot be compared with OBRA
90 estimates of individual provisions.  The following example
provides another illustration. 

Medicare is a secondary payor under specified circumstances when
Medicare beneficiaries are covered by other third party payors, that
is, automobile, medical, no-fault, and liability insurance.  The
Medicare Secondary Payor provision (OBRA 90, section 4203) produced
savings by (1) lengthening by 6 months the period during which an
employer sponsored health plan is the primary payor for end-stage
renal disease beneficiaries, (2) extending by 3 years the section
that makes Medicare a secondary payor for disabled beneficiaries, and
(3) extending the authorized period when tax data that improves the
identification and collection of Medicare secondary payor cases can
be shared with HCFA.  A HCFA agency official pointed out that the
Medicare Secondary Payor program was affected by multiple
legislation, for example, the Deficit Reduction Act of 1984,
Consolidated Omnibus Budget Reconciliation Act of 1985, Omnibus
Budget Reconciliation Act of 1989, and OBRA 90.  While it might
appear that OBRA 90 effects could be separated from those of earlier
laws, this is not the case.  HCFA regulations and guidance
implementing these laws as well as OBRA 90 are intermingled, so that
isolating the budgetary impact of one legislative change from another
is virtually impossible.  HCFA's savings measure the net budgetary
impact of all legislation and any other factors that affect savings,
not just the OBRA 90 savings. 


--------------------
\14 Of the 38 provisions in our sample, 9 provisions made changes in
the Medicare program. 

\15 As described later in this report, CBO included behavioral
assumptions about hospital responses in its estimate of the PPS
provision.  However, the fact that CBO could estimate this impact
does not alter agency difficulties in measuring the actual budgetary
impact of hospital behavior on Medicare outlays. 


      SPECIFIC SAVINGS COULD BE
      DETERMINED IN A FEW CASES
---------------------------------------------------------- Letter :6.2

For 5 of the 38 provisions, accounting for $4.3 billion of the
estimated savings, agencies could determine the specific savings
because the budgetary impact of OBRA 90 could be isolated from other
factors.  These provisions involved new user fees or identified
amounts to be deposited in budget or Treasury accounts.  The
following two cases illustrate how agencies determined savings. 

  -- The Railroad Safety User Fee provision (OBRA 90, section 10501)
     requires the Federal Railroad Administration, under the
     Department of Transportation, to cover costs of the railroad
     safety program by assessing a fee on railroads.  Because OBRA 90
     created the new user fee and subsequent legislation and
     administrative action did not supersede OBRA 90, isolating the
     budgetary impact is possible.  Agency officials pointed out that
     billing and accounting systems showed that for fiscal years 1991
     through 1993, fees collected were $14 million, $32 million, and
     $34 million, respectively.  The OBRA 90 savings estimates as
     prepared by CBO were $20 million, $35 million, and $36 million
     for fiscal years 1991 through 1993. 

  -- The Patent and Trademark Office (PTO) surcharge provision (OBRA
     90, section 10101) requires that PTO impose a surcharge on
     certain patent fees and sets in statute the amounts of money
     that PTO should deposit in a U.S.  Treasury account.  Isolating
     the budgetary impact is possible because OBRA 90 set in statute
     the amounts of money that PTO deposits.  PTO imposed a surcharge
     on the patent statutory fees and tracked the amounts collected
     and deposited through a financial reporting system.  In fiscal
     years 1992 and 1993, PTO collected fees sufficient to deposit an
     amount equal to the savings estimates ($95 million and $99
     million, respectively).  However, in fiscal year 1991,
     sufficient fees failed to materialize because of customer
     reaction to the additional surcharge, the Persian Gulf War, the
     recession, and a decline in patent applications from Germany. 
     Thus, in fiscal year 1991 PTO deposited $99.3 million (the total
     amount collected) as compared to the OBRA 90 target deposit of
     $109.8 million for that year. 


   CBO AND EXECUTIVE BRANCH
   AGENCIES DID WHAT REASONABLY
   COULD BE EXPECTED
------------------------------------------------------------ Letter :7

CBO and executive branch agencies had reasonable processes to ensure
that estimates were as accurate as possible and that reconciliation
provisions were implemented promptly.  CBO used the best available
data, appropriate mathematical models, and supervisory review in its
efforts to provide accurate savings estimates as described below. 
The 14 executive branch agencies carried out OBRA 90's policy changes
on time and implemented all but one provision. 


      CBO PRACTICES WERE DESIGNED
      TO MINIMIZE ERROR IN
      ESTIMATES
---------------------------------------------------------- Letter :7.1

CBO analysts used the best available information and professional
techniques to prepare and review the estimates for each of the 38
provisions we selected and reviewed.\16 While we did not assess the
accuracy of individual estimates, we believe that these practices
helped to produce the best estimates possible, within the constraints
CBO faced.  We also found that CBO's semiannual baseline updates have
a useful, if indirect effect, on the quality of estimates. 


--------------------
\16 We were unable to interview the analyst who prepared the OBRA 90
estimate for 12 provisions.  In these cases, we interviewed the
relevant unit chief or the analyst's successor.  All but one of the
estimates in our sample were prepared by analysts in CBO's Budget
Analysis Division.  The remaining estimate, the Mine Safety and
Health provision (OBRA 90, section 3102) was prepared by an analyst
in the Tax Analysis Division. 


         ANALYSTS USED THE BEST
         AVAILABLE INFORMATION IN
         THEIR ESTIMATES
-------------------------------------------------------- Letter :7.1.1

The accuracy of estimates for individual provisions depends heavily
on the quality of the information CBO uses.  While information was in
some cases limited, we believe that analysts used the best available
in the provisions we examined.  For eight provisions, CBO used
multiple information sources.  The type of information from these
sources ranged from raw data to preliminary estimates. 

For 30 provisions, CBO used data from executive branch agencies. 
Because federal agencies routinely collect data to fulfill their
program monitoring and oversight responsibilities, they are CBO's
primary source of up-to-date information.  For a majority of
provisions, analysts stated that federal agencies were the only such
sources of data to develop estimated program costs.  For example, the
Portability of Benefits provision (OBRA 90, section 7202) protects
the pay and benefits of certain Department of Defense (DOD) and Coast
Guard employees who converted to the civil service system.  CBO
relied on DOD's and Coast Guard's data to determine the number of
employees affected by this provision because DOD and the Coast Guard
are apt to be the best sources of data about their own employees.  In
two cases, CBO's estimates relied on preliminary estimates provided
by relevant federal agencies.  This seemed reasonable because these
agencies maintained large, complex databases which were the sole
sources of the necessary information. 

For eight provisions, CBO used data from nonfederal sources. 
Sometimes private sector companies and other entities affected by
changes in federal policy have precise or detailed information
related to program costs.  If CBO is satisfied with the quality of
this information, analysts may use it to supplement agency-provided
data.  For example, one part of the Durable Medical Equipment
provision (OBRA 90, section 4152) limited Medicare payments for
seat-lift chairs to cover only the seat-lift mechanism.  Previously,
Medicare covered the cost of the entire chair.  To prepare the
Durable Medical Equipment estimate, CBO examined many manufacturers'
data on the percentage of total chair costs attributable to the
mechanism. 

For eleven provisions, CBO used account-level information from CBO's
baseline because it was the best source of total program costs for
future fiscal years.  To illustrate, the Nuclear Regulatory
Commission (NRC) user fees provision (OBRA 90, section 6101) requires
NRC to recoup 100 percent of its annual budget authority, less
payments received from the Nuclear Waste Fund.  The analyst used
baseline projections of the NRC's budget authority and payments from
the nuclear waste fund for fiscal years 1991 through 1995 to
calculate the amounts NRC needed to recoup. 


         ESTIMATES WERE PREPARED
         USING MATHEMATICAL
         FORMULAS
-------------------------------------------------------- Letter :7.1.2

Because many factors affect program costs, CBO analysts used
mathematical formulas to prepare estimates for 34 of the 38
provisions.  Some formulas we examined relied entirely on common
arithmetic functions.  Others were sophisticated, interrelated sets
of equations. 

However simple or complex the underlying methodology, CBO considered
and included as appropriate the following factors:


         PROGRAM PARAMETERS
-------------------------------------------------------- Letter :7.1.3

All of the CBO models included variables to account for existing
program costs and changes in costs resulting from the OBRA 90
provision.  For example, CBO's estimate for the Cost-of-Living
Increases in Veteran Compensation Rates provision (OBRA 90, section
8005)\17 included variables to track the number of veterans and
survivors receiving compensation payments, the mortality rates of
veterans, the average costs per veteran, and the amount of the
cost-of-living increase. 


--------------------
\17 This provision established the amount of the calendar year 1991
cost-of-living adjustment (COLA) and shifted the effective date of
the increase by 1 month.  In addition, it requires that COLAs be
rounded down to the nearest dollar. 


         BEHAVIORAL EFFECTS
-------------------------------------------------------- Letter :7.1.4

CBO factored behavioral changes by program participants into 19
estimates.  This was done to avoid overestimating savings.  These
behavioral adjustments generally reflected an assumption--based on
observed responses to prior program changes--that program
participants would try to minimize negative economic effects of
policies on themselves.  CBO reduced savings for the PPS Changes
provision (OBRA 90, section 4002) by 10 percent based on historical
data that documented health care providers' efforts to recoup lost
income from previous reductions in payment rates. 


         INTERACTIVE EFFECTS
-------------------------------------------------------- Letter :7.1.5

CBO included interactive effects in 14 estimates.  Again, this was
done in an effort to avoid overestimating savings.  Two types of
interactions are important.  First, provisions affecting the same
program may interact and produce lower savings than the sum of the
separate estimates.  Second, program reductions in one area may be
offset by increased spending in another.  The Triple Base
provision\18 (OBRA 90, section 1101) and the Calculation of
Deficiency Payments on a 12-Month Average Price provision (OBRA 90,
section 1102), which lowered outlays in agriculture commodity
programs, are examples of intraprogram interactions.  If CBO had not
included the lower production levels resulting from the Triple Base
provision in its calculation of the savings from the deficiency
payment provision, estimated savings would have been overstated.  An
example of interprogram interaction is the Medicare Part B Premium
provision (OBRA 90, section 4301) and the Medicaid program.  Since
Medicaid pays the Medicare Part B premium for poor participants, the
savings from increasing the premium are partially offset by higher
Medicaid payments. 

As discussed earlier, for two provisions, CBO used preliminary
estimates prepared by agencies.  For another two provisions--Postal
Service Funding of COLAs (OBRA 90, section 7101) and Patent and
Trademark Office User Fees (OBRA 90, section 10101)--OBRA 90
identified amounts that agencies should pay or collect for each
fiscal year from 1991 through 1995.  In such cases, CBO's general
rule is to assume that required actions are taken and, therefore,
that savings equal the legislatively identified payment or
collection. 


--------------------
\18 The term Triple Base refers to the three components of base
acreage:  the excluded flexible acreage, acreage idled due to acreage
reduction programs, and the remaining eligible base. 


         CBO'S REVIEW PROCESSES
-------------------------------------------------------- Letter :7.1.6

The assumptions and methodologies used to prepare each of the OBRA 90
estimates we examined received supervisory review before being sent
to the Congress.  The appropriate unit chief\19 initially reviewed
all estimates.  These estimates were then reviewed by the Assistant
Director or Deputy Assistant Director of the Budget Analysis
Division.  Finally, the CBO Director or his deputy personally
reviewed these estimates before signing the transmittal letters. 
Moreover, the unit chief's review of reconciliation estimates is
generally an on-going process. 

Both the more complex and sensitive estimates and those showing
larger savings, received more review.  For example, the FmHA
provision (OBRA 90, section 1202) was sensitive due to shared
jurisdiction between authorizing and appropriations committees.  The
analyst told us that the unit chief and appropriate assistant
director held a series of meetings to discuss this $2 billion
provision.  Another analyst noted that she often involves the CBO
Director because the estimates she prepares are large, complex, and
often controversial. 


--------------------
\19 CBO's Budget Analysis Division prepares all cost estimates for
spending provisions.  Within this division, there are three cost
estimate units (the defense and international affairs, human
resources, and natural and physical resources) a projections unit, a
scorekeeping unit, and two technical support units.  Each of the
units is headed by a unit chief. 


      TECHNICAL AND DATA
      CONSTRAINTS MEAN THAT
      ESTIMATES WILL RARELY BE
      ERROR FREE
---------------------------------------------------------- Letter :7.2

For most provisions, actual results are likely to differ from CBO
estimates due to methodological constraints and data limitations.  In
addition, even with good data, predicting the future is difficult and
requires making assumptions about the behavior of others.\20 The OBRA
90 savings estimates we examined, like all CBO estimates, are
point-in-time measurements which are based on policies and economic
forecasts in effect at the time of the budget resolution.  However,
policies are affected by subsequent legislation and administrative
actions.  The economy and technical factors, too, may turn out to
differ from those assumed.  Together, these factors can push spending
higher or lower than originally predicted.  For example, as we stated
earlier in this report, changes in economic conditions, the single
largest change, pushed deficit projections $485 billion higher.  As
listed in table 1, the changes in technical assumptions--especially
in areas of Medicare and Medicaid-- pushed the deficit $298 billion
higher. 

While CBO relies on the best available data to prepare estimates,
these data may not account for all factors that could affect the
estimates.  For 4 of the 38 provisions, analysts noted that data
limitations may have decreased the accuracy of the estimates.  In the
absence of data, CBO analysts must make assumptions about the factors
that affect estimated savings from the provision.  For example, the
Payment of Group Health Premiums and Cost Sharing Where
Cost-Effective provision (OBRA 90, section 4402) requires states to
pay group health premiums for eligible Medicaid beneficiaries.  The
analyst noted that a national database of the characteristics of the
Medicaid population did not exist.  Thus, CBO was unable to determine
exactly how many beneficiaries might be affected by this provision. 
Accordingly, the analyst relied on other data sources to make a
series of assumptions about the number of affected beneficiaries. 

The experience of CBO analysts helps to compensate for data
limitations and methodological difficulties.  This is because program
knowledge and experience in preparing estimates improve the quality
of assumptions which, in turn, increases the accuracy of estimates. 
The analysts who prepared the OBRA 90 estimates in our sample had an
average of 10 years of CBO or other experience in the relevant
program area and had prepared many cost estimates.  For example,
preparing the estimate for the Copayments for Veterans Medication
provision (OBRA 90, section 8012), which established a copayment for
each 30-day supply of medication on an outpatient basis, required CBO
to make assumptions about the number of veterans receiving
prescriptions and the number of prescriptions written annually.  The
analyst who prepared this estimate had 14 years of CBO experience and
7 additional years of experience with veterans programs. 


--------------------
\20 For a further discussion of how this problem can affect actual
savings see Budget Policy:  Issues In Capping Mandatory Spending
(GAO/AIMD-94-155, July 18, 1994), chapter 4. 


         CBO UPDATES ITS BASELINE
-------------------------------------------------------- Letter :7.2.1

CBO's semiannual updates of its baseline indirectly improve the
quality of its estimates for individual provisions.  The purpose of
these updates is to increase the accuracy of the baseline by ensuring
that baseline assumptions accurately reflect current policies,
economic conditions, and other technical assumptions, such as numbers
of program participants or behavioral responses to program changes. 
Since the assumptions used in the baseline are used to prepare other
estimates, the update process improves the quality of CBO estimates
for legislative proposals.  To update the baseline, analysts examine
developments in policy, the economy, and other variables affecting
program costs that occurred since the previous baseline was prepared. 
Based on this analysis, analysts propose and make changes to the
baseline.  A baseline review panel consisting of the appropriate unit
chiefs and the Budget Analysis Division Director and his deputy,
approves major changes. 


      AGENCIES IMPLEMENTED OBRA 90
      PROVISIONS
---------------------------------------------------------- Letter :7.3

Executive branch agencies executed the changes required by OBRA 90. 
As measured by the first part of our two-tiered implementation test,
we found that agencies issued formal rules or binding guidance in
lieu of the rules when required to do so for all 38 provisions.  As
measured by the second part of our test (timeliness and scope), we
found that agencies: 

  -- Implemented all 38 reconciliation provisions within the first
     fiscal year of applicability. 

  -- Implemented 37 of the 38 reconciliation provisions without
     narrowing or broadening the scope of the OBRA 90 provision.  In
     the one exception, OBRA 90, section 10301, we found that the
     United States Travel and Tourism Administration (USTTA) narrowed
     the scope of the United States Travel and Tourism Facilitation
     Fee provision.  As stated in a final rule published in the
     Federal Register on February 4, 1992, the Under Secretary (of
     Commerce) for Travel and Tourism determined that since
     collections from airlines were suspended for legal reasons,
     charging and collecting fees from passenger cruise ship lines
     would not serve the legislative intent expressed in the statute. 
     OBRA 90 required that the fee be paid on a pro-rata basis by
     commercial airlines and passenger cruise ship lines transporting
     passengers into the United States.\21

Implementation procedures differed for the 38 provisions.  As
described in more detail in appendix V, agencies implemented these
provisions in different ways. 


--------------------
\21 The Under Secretary for Travel and Tourism suspended action to
charge or collect the fees from commercial passenger airlines because
this action was inconsistent with Article 15 of the Chicago
Convention of the International Civil Aviation Organization.  All
collected fees were returned. 


   CONCLUSIONS
------------------------------------------------------------ Letter :8

We share the concern about the gap between expected and actual
deficits experienced subsequent to reconciliation legislation.  As
you have noted, these concerns could frustrate deficit reduction
efforts if policymakers come to believe that the difficult choices
they make do not matter. 

Looking back, the economy was the single largest factor driving
deficits higher than anticipated at the time of OBRA 90's passage. 
In contrast, OBRA 90's policy changes helped reduce the deficit. 
While it is not possible to identify precisely the specific savings
from individual provisions, our findings--that CBO had reasonable
processes to ensure that estimates were as accurate as possible and
that executive branch agencies implemented reconciliation
provisions--and our discussions with knowledgeable budget/program
officials lead us to conclude that, overall, OBRA 90 contributed
significantly to reducing the deficit below what it otherwise would
have been. 

Although it is impossible to determine the specific savings for
individual reconciliation provisions, a deficit vastly greater than
originally forecast need not be accepted.  The Congress can, under
current procedures, at any time enact changes in individual mandatory
programs--either alone or as part of reconciliation--to reduce the
deficit.  Currently, however, there is no procedure that
automatically triggers a comparison between the current and the
expected deficit paths. 

A new process that prompts the Congress periodically to look back and
compare the actual deficit to the deficit projected at the time of a
prior deficit reduction agreement and/or the most recent
reconciliation legislation could be added to current budgetary
controls.  Such a process differs from current budget processes
because--if the deficit were greater than projected--the Congress
would decide explicitly--by voting--whether to accept the slippage
and affirm a higher deficit path or to act to bring the deficit path
closer to the original goal by recouping some or all of this
slippage.  Thus, the new requirement would provide members who make
difficult choices in reconciliation an additional opportunity to
ensure that the deficit path they voted for will, in fact,
materialize. 

This process would be broader than the one we identified to increase
budgetary control of mandatory spending.  In Budget Policy:  Issues
in Capping Mandatory Spending we identified an approach to increase
budgetary control of mandatory spending programs.  Using this
approach, the Congress periodically would vote on whether to make
program changes when mandatory spending exceeds certain targets. 
Under this process, the Congress would consider changes in the
structure and/or benefits in mandatory programs to achieve reductions
in spending if it deemed appropriate.  The process we are suggesting
for consideration in this report would be broader.  It would look
back at the deficit's path--not one component of spending--compared
to that predicted under deficit reduction legislation. 


   MATTER FOR CONGRESSIONAL
   CONSIDERATION
------------------------------------------------------------ Letter :9

The Congress may wish to consider a new process under which it is
periodically prompted to look back at progress in reducing the
deficit.  Such a lookback would compare the current CBO deficit
projections to those projected at the time of a prior deficit
reduction agreement and/or the most recent reconciliation legislation
and analyze the reasons for differences.  For differences exceeding a
predetermined amount, the Congress would decide explicitly--by
voting--whether to accept the slippage or act to bring the deficit
path closer to the original goal by recouping some or all of this
slippage. 

We are sending copies of this report to the Speaker of the House;
House Majority Leader and Majority Whip; Chairmen of the House
Committee on Ways and Means, House Committee on the Budget, and House
Committee on Banking, Finance and Urban Affairs; other interested
congressional committees and Members of Congress; the Directors of
the Congressional Budget Office and the Office of Management and
Budget; appropriate executive branch agencies; and other interested
parties.  We will also make copies available to others upon request. 

Please contact me at (202) 512-9573 if you or your staff have any
questions concerning the report.  Other major contributors are listed
in appendix VI. 

Paul L.  Posner
Director, Budget Issues


RECONCILIATION ACTS
=========================================================== Appendix I

                         (Dollars in billions)

                                                Conferen  Anticipated
                                                ce        fiscal year
                                        Public  report    (FY)
Congress and session      Bill          law     no.       savings\a
------------------------  ------------  ------  --------  ------------
96th, 1979-80,            H.R. 7765     96-     96-1479   $8.2 over
2nd session                             499               FY 1981

97th, 1981-82,            H.R. 3982     97-     97-208    $130.6 over
1st & 2nd                 H.R. 6955     35      97-759    FYs 1982-
sessions                  H.R. 4961     97-     97-760    84
                          (TEFRA)       253               $129.1\b
                                        97-               over
                                        248               FYs 1983-85

98th, 1983-84,            H.R. 4169     98-     None\c    $8.2 over
2nd session                             270               FYs 1984-87

99th, 1985-86,            H.R. 3128\d   99-     99-453    $18.2 over
2nd session               H.R. 5300     272     99-1012   FYs 1986-
                                        99-               88
                                        509               $11.7 over
                                                          FYs 1987-89

100th, 1987-88,           H.R. 3545     100-    100-495   $76 over
1st session                             203               FYs 1988-89

101st, 1989-90,           H.R. 3299     101-    101-      $14.7 over
1st & 2nd                 H.R. 5835     239     386       FY 1990
sessions                  (OBRA 90)     101-    101-964   $496.2 over
                                        508               FYs 1991-95

103rd, 1993-94,           H.R. 2264     103-    103-213   $496 over
1st session               (OBRA 93)     66                FYs 1994-98
----------------------------------------------------------------------
\a The savings estimates reported in this table were taken from
Congressional Quarterly reports which cited conference reports or
other congressional documents. 

\b This reflects $13.3 billion in spending reductions due to H.R. 
6955.  In addition, the Tax Equity and Fiscal Responsibility Act of
1982 (TEFRA) contained $98.3 billion in revenue increases and $17.5
billion in spending reductions. 

\c No conference was held. 

\d H.R.  3128, the Deficit Reduction Amendments of 1985, failed to
clear the first session of Congress before adjournment.  It was
enacted in the second session as a reconciliation bill. 

Source:  Congressional Quarterly Almanac, 1979 through 1993. 


SAMPLE OF OMNIBUS BUDGET
RECONCILIATION ACT OF 1990
PROVISIONS
========================================================== Appendix II



                                    Table II.1
                     
                           8 Highest Dollar Provisions

                              (Dollars in millions)

                                                                      Agency
                                                                      able to
Se                  CBO 5-                                  Agency    track
ct                    year  Agency            Met           says      specific
io                 savings  responsible for   implementati  produced  dollar
n   Title         estimate  implementation    on test       savings   savings
--  ------------  --------  ----------------  ------------  --------  ----------
11  Triple Base      9,085  Department of     yes           yes       no
01  for                     Agriculture
    Deficiency
    Payments

20  Bank                \a  Federal Deposit   yes           no        no
02  Insurance               Insurance
-   Fund (BIF)              Corporation
20  Provisions
04

40  PPS Changes      8,890  Department of     yes           yes       no
02                          Health and Human
                            Services

70  Elimination      7,600  Office of         yes           yes       no
01  of Lump-Sum             Personnel
    Credit With             Management
    Payment
    Shift

43  Part B           7,455  Department of     yes           yes       no
01  Premium                 Health and Human
                            Services

42  Extension of     6,085  Department of     yes           yes       no
03  Secondary               Health and Human
    Payor                   Services
    Provisions

40  Payments for     4,050  Department of     yes           yes       no
01  Capital                 Health and Human
    Related                 Services
    Costs of
    Inpatient
    Hospital
    Care

70  FEHB Reforms     3,646  Office of         yes           yes       no
02                          Personnel
                            Management

Es                55,811\b
ti
ma
te
d
Sa
vi
ng
s
--------------------------------------------------------------------------------
\a CBO scored this provision as having no budgetary impact.  However,
in its official letter CBO also indicated that agency administrative
actions could produce $9 billion in savings.  The Budget Committees
included the $9 billion as part of the $496 billion in total savings. 
Thus, we included this provision in our sample. 

\b Includes $9 billion in BIF savings (sections 2002-2004). 



                                    Table II.2
                     
                     30 Randomly Selected OBRA 90 Provisions

                              (Dollars in millions)

                                                                      Agency
                                                                      able to
Se                  CBO 5-                                  Agency    track
ct                    year  Agency            Met           says      specific
io                 savings  responsible for   implementati  produced  dollar
n   Title         estimate  implementation    on test       savings   savings
--  ------------  --------  ----------------  ------------  --------  ----------
41  Pathology           60  Department of     yes           yes       no
04  Services                Health and Human
                            Services

10  Patent and         495  Department of     yes           yes       yes
10  Trademark               Commerce
1   User Fees

20  Allow FDIC          12  Federal Deposit   yes           yes       no
05  to Borrow               Insurance
    From Federal            Corporation
    Financing
    Bank

80  Use of IRS         743  Department of     yes           yes       no
51  and SSA Data            Veterans Affairs
    for Income
    Verification

23  Extension of       818  Federal           yes           no        no\a
02  the Federal             Emergency
    Flood                   Management
    Insurance               Agency
    Program

80  Cost of             44  Department of     yes           yes       no
05  Living                  Veterans Affairs
    Increases in
    Compensation
    Rates

80  Elimination        313  Department of     yes           yes       no
02  of                      Veterans Affairs
    Presumption
    of Total
    Disability

51  Repeal of          713  Department of     yes           yes       no
16  Retroactive             Health and Human
    Benefits for            Services
    Certain
    Beneficiarie
    s

71  Postal           1,946  Postal Service    yes           no        yes
01  Service
    Funding of
    COLAs

61  NRC User         1,554  Nuclear           yes           yes       yes
01  Fees and                Regulatory
    Annual                  Commission
    Charges

10  Customs User     2,292  Department of     yes           yes       no
00  Fees                    the Treasury
1

44  Prescription     1,930  Department of     yes           yes       no
01  Drug                    Health and Human
    Discounts               Services
    for State
    Medicaid
    Programs

72  Portability        149  Office of         yes           yes       no
02  of Benefits             Personnel
                            Management

10  Railroad           169  Department of     yes           yes       yes
50  Safety User             Transportation
1   Fees

10  NOAA User           12  Department of     yes           yes       no
20  Fees                    Commerce
1

41  Anesthesia         220  Department of     yes           yes       no
03  Services                Health and Human
                            Services

23  Extension of        14  Federal           yes           no        no
01  the Federal             Emergency
    Crime                   Management
    Insurance               Agency
    Program

72  Computer           270  Office of         yes           yes       no
01  Matching and            Management and
    Privacy                 Budget
    Protection

22  Auction of       1,010  Department of     yes           yes       no
01  Insured                 Housing and
    Mortgages               Urban
                            Development

41  Radiology        1,240  Department of     yes           yes       no
02  Services                Health and Human
                            Services

11  Loan             1,151  Department of     yes           yes       no
05  Origination             Agriculture
    Fees and
    Other
    Savings

41  New Health         580  Department of     yes           yes       no
06  Care                    Health and Human
    Practitioner            Services
    s

11  Calculation        755  Department of     yes           yes       no
02  of                      Agriculture
    Deficiency
    Payments
    Based on 12-
    Month
    Average

41  Durable          2,140  Department of     yes           yes       no
52  Medical                 Health and Human
    Equipment               Services

11  Acreage            905  Department of     yes           no        no
03  Reduction               Agriculture
    Program for
    1991 Crop

44  Group Health     1,005  Department of     yes           yes       no
02  Premium and             Health and Human
    Cost Sharing            Services
    Payment
    Where Cost-
    Effective

10  Travel and          78  Department of     yes\b         no        yes\c
30  Tourism Fees            Commerce
1

12  Authorizatio     2,114  Department of     yes           yes       no
02  n Levels for            Agriculture
    FmHA Loans

80  Copayment           85  Department of     yes           yes       no
12  for                     Veterans Affairs
    Medications

31  Mine Safety        247  Department of     yes           yes       no
02  and Health              Labor
    Penalties

Es                  23,064
ti
ma
te
d
Sa
vi
ng
s
--------------------------------------------------------------------------------
\a This provision has two components--reauthorization ($594 million)
and premium surcharge ($224 million).  Surcharge savings can be
tracked. 

\b This provision was implemented within the first fiscal year of
applicability, but the USTTA narrowed the provision's scope. 

\c If action on the provision had not been suspended for legal
reasons, savings would have been tracked. 


COMPARISON OF CBO, ADMINISTRATION,
AND PRIVATE SECTOR ECONOMIC
ASSUMPTIONS TO ACTUALS AND CURRENT
PROJECTIONS
========================================================= Appendix III

                                  1991    1992    1993    1994    1995
--------------  --------------  ------  ------  ------  ------  ------
Real Gross National Product\a (annual percentage change)
----------------------------------------------------------------------
                CBO, 7/90          2.5     2.6     2.6     2.6     2.6
                Administration     2.8     3.2     3.2     3.1     3.0
                 , 6/90
                Blue Chip\b        2.3     2.8     2.7     2.4     2.6

Real Gross Domestic Product\a (annual percentage change)
----------------------------------------------------------------------
                Actual            -0.7     2.6     3.0
                CBO, 8/94                                  4.0     3.0

Consumer Price Index\c (annual percentage change)
----------------------------------------------------------------------
                CBO, 7/90          4.2     4.2     4.0     4.0     4.0
                Administration     4.1     4.0     3.7     3.4     3.0
                 , 6/90
                Blue Chip\b        4.3     4.0     4.1     4.0     4.0
                Actual             4.2     3.0     3.0
                CBO, 8/94                                  2.6     3.1

Unemployment rate\d (percent)
----------------------------------------------------------------------
                CBO, 7/90          5.4     5.4     5.5     5.5     5.5
                Administration     5.6     5.5     5.4     5.3     5.2
                 , 6/90
                Blue Chip\b        5.5     5.6     5.5     5.4     5.3
                Actual             6.7     7.4     6.8
                CBO, 8/94                                  6.2     5.8

Long-term interest rates (10-year Treasury notes)
----------------------------------------------------------------------
                CBO, 7/90          7.8     7.4     7.2     6.9     6.8
                Administration     7.9     7.0     6.1     5.8     5.4
                 , 6/90
                Blue Chip\b,e      8.3     8.0     7.8     7.8     7.8
                Actual             7.9     7.0     5.9
                CBO, 8/94                                  6.8     6.8

Short-term interest rates (3-month Treasury bills)
----------------------------------------------------------------------
                CBO, 7/90          6.9     6.7     6.2     5.6     5.4
                Administration     6.8     5.8     5.1     4.8     4.4
                 , 6/90
                Blue Chip\b        7.5     7.0     7.0     6.9     6.7
                Actual             5.4     3.5     3.0
                CBO, 8/94                                  4.1     5.5
----------------------------------------------------------------------
Note:  Calendar year assumptions. 

\a Beginning in January 1992, CBO's economic assumptions are based on
gross domestic product (GDP) instead of gross national product (GNP). 
Between 1979 through 1991, real GDP growth was about 0.1 percent
higher than real GNP growth.  Real GDP is calculated using base year
1987 dollars. 

\b Blue Chip forecasts for 1991 were based on a survey of private
forecasters published by Eggert Economic Enterprises, Inc.  published
on June 11, 1990.  Blue Chip forecasts for 1992 through 1995 were
based on a survey of private forecasters published by Eggert Economic
Enterprises, Inc.  on March 10, 1990. 

\c CBO and Blue Chip used consumer price index for urban consumers. 
The Administration used the consumer price index for all urban wage
earners and clerical workers.  The actual rate used the consumer
price index for urban consumers. 

\d CBO, Blue Chip, and the actual rate used the average annual
unemployment rate for the civilian labor force.  The Administration
used the total labor force, including armed forces residing in the
United States. 

\e Blue Chip did not project a 10-year note rate.  The Blue Chip
estimates presented in the table were constructed by CBO based on an
estimated relationship between rates on 10-year bonds and the Blue
Chip projected Aaa bond rate. 

Sources:  Congressional Budget Office and Council of Economic
Advisers. 


AGENCY VIEWS ON SAVINGS FROM SIX
OBRA 90 PROVISIONS
========================================================== Appendix IV

As noted in the report, our conclusions about whether OBRA 90
provisions actually reduced the deficit are based on the judgments of
agency experts with specific legal, programmatic, and/or budgetary
expertise in the programs affected by the provisions in our sample. 
The experts we spoke with were unable to determine the specific
savings achieved by individual provisions.  However, they were able
to judge whether provisions produced savings, that is whether
provisions increased collections and/or reduced spending in the
affected program.  When experts reported that a provision did either,
we concluded that the provision helped reduce the deficit. 

All 38 provisions in our sample were scored appropriately as savings
according to scoring conventions and baseline estimates of federal
outlays in effect at the time.  However, agency experts reported that
enacting 6 of the 38 provisions had not directly reduced outlays or
increased user fees or premium collections for their programs.  One
provision was inconsistent with an international air agreement (OBRA
90, section 10301) and was not implemented.  Agency officials stated
that two other provisions merely enacted steps the agency would have
taken anyway and so should not be credited with the savings
attributable to the action (OBRA 90, sections 2002-2004 and 1103). 
They also disagreed that savings should be attributed to provisions
which (1) extended expiring authority for the Flood\1 and Crime
Insurance Programs without otherwise changing them (OBRA 90, sections
2302 and 2301) or (2) required payments from an off-budget entity
(Postal Service) to an on-budget entity (Civil Service Retirement and
Disability Fund) (OBRA 90, section 7101).\2

This appendix describes each of the six provisions in greater detail,
the program affected, how OBRA 90 changed the program, why the
provision was scored as savings, and why agency experts believed that
enacting the provision did not reduce the deficit. 

SECTION 10301:  U.S.  TRAVEL AND
TOURISM ADMINISTRATION
FACILITATION FEE

USTTA in the Department of Commerce was established by the
International Travel Act of 1961.  The purpose of USTTA is to
encourage foreign travel to the United States by providing assistance
to private, state, and local tourism organizations. 

OBRA 90 required, "to the extent not inconsistent with international
treaties or agreements," the Secretary of Commerce to collect a fee
from airlines and cruise ship lines for foreign passengers entering
the United States.  This fee would cover the expenses of USTTA.  The
provision set the fee for 1991 at $1 per passenger.  For 1992 through
1995, the total amount of fees collected should equal USTTA
appropriations.  This provision was estimated to collect $78 million
over 5 years. 

In order to score the savings for 1991, CBO multiplied the
anticipated number of foreign tourists by $1.  Since the 1992 through
1995 collections should equal USTTA's appropriation, the analyst
relied on CBO's baseline estimates of USTTA's budget authority to
determine savings for each year. 

Unlike the other 5 provisions, this provision was not implemented. 
The Secretary of Commerce delegated authority for collecting this fee
to the Under Secretary for Travel and Tourism.  The Under Secretary
determined that the fee on airlines was inconsistent with Article 15
of the Chicago Convention of the International Civil Aviation
Organization and, in accordance with OBRA 90, suspended actions to
charge or collect this fee.  Based on OBRA 90's intent to charge
cruise ships and airlines on a pro-rata basis, the Under Secretary
also suspended collection of the fee on cruise ships.  In addition,
USTTA returned all collected fees.  Thus, this provision did not
reduce the deficit because fees were not collected. 

SECTION 1103:  ACREAGE REDUCTION
PROGRAM FOR THE 1991 CROP

The Agricultural Stabilization and Conservation Service (ASCS) within
the Department of Agriculture is responsible for administering
agricultural commodity programs.  Major agricultural commodities
affected by federal programs include feed grains (corn, sorghum,
oats, and barley), wheat, rice, upland cotton, soybeans, and dairy
products.  In 1990, agricultural price support programs were governed
primarily by the Food Security Act of 1985. 

Farmers who participate in federal support programs are paid a
deficiency payment equal to the deficiency payment rate multiplied by
the number of program acres and the program yield (that is, the
historical amount of per acre production of a commodity).  The
deficiency payment rate is generally the difference between the
market price and a target price established in law. 

Farmers who choose to participate in the federal support programs
must also participate in an Acreage Reduction Program (ARP).  ARPs
require program participants not to plant a crop on a percentage of
their land.  The Secretary of Agriculture annually announces the ARP
percentages for each commodity.  Increasing ARPs affects outlays
primarily in two ways.  First, higher ARPs reduce the number of
payment acres, which decreases the total deficiency payments made to
individual producers.  Second, higher ARPs decrease production of a
commodity, thus increasing the market price.  Increasing the market
price decreases the level of the federal deficiency payments.  Thus,
the higher the ARP, the lower the level of the deficiency payments. 

OBRA 90 mandated a 1991 wheat ARP of not less than 15 percent.  In
addition, OBRA 90 required that the corn ARP be set at no less than
7.5 percent.  Combined, these components were estimated to reduce
outlays by $905 million. 

CBO scored this provision using its baseline models for commodity
programs.  These econometric models are designed to consider the many
factors that affect federal outlays for each commodity program. 
Among the factors included in these models are the program parameters
under current law (such as target prices, acreage reduction
requirements, and loan rates), and variables to determine the supply,
demand, and market price for the commodity.  Estimates from these
models are dependent on assumptions about farm prices, income support
mechanisms, international events, economic conditions, weather, and
crop yields.  The assumptions used to prepare this estimate were the
same ones in effect at the time of the budget resolution. 

CBO's baseline assumed that the wheat and feed grain (including corn)
ARPs for 1991 would be 5 percent.  Since the OBRA 90 provision
increased the ARPs above these levels, it produced savings.  To
calculate the savings from increasing ARPs above the baseline levels,
CBO adjusted variables in the models that would be affected,
beginning with the ARP percentages themselves.  CBO also incorporated
offsets for behavioral effects by farmers--who CBO assumed would seek
to avoid economic losses due to the provisions.  The model was also
modified to consider interactive effects related to certain other
OBRA 90 provisions as well as some provisions of the 1990 farm bill. 
CBO's inclusion of interaction adjustments (offsets) reduced the
savings estimate for the OBRA 90 commodity provisions. 

ASCS officials agreed that higher ARPs saved money but believed that
since this provision did not result in higher ARPs than the Secretary
of Agriculture would have chosen anyway, the savings should not be
attributed to OBRA 90.  In their view, worsening conditions in the
commodity markets meant that the Secretary would have chosen the ARP
levels mandated by OBRA 90, even without the legislation.  For
example, in August 1990, the Secretary announced that the wheat ARP
for 1991 would be between 10 and 20 percent, depending on the final
provisions of the 1990 farm bill. 

SECTIONS 2002-2004:  BANKING
INSURANCE PROVISIONS

FDIC administers BIF.  BIF was created by the Financial Institutions
Reform, Recovery, and Enforcement Act of 1989 (FIRREA).  BIF's
primary purposes are to (1) insure the deposits and protect the
depositors of insured institutions and (2) finance the resolution of
institutions, including managing and liquidating their assets.  The
primary funding source for the BIF is an assessment premium charged
to member institutions.  The assessment premiums charged to
individual BIF members are a percentage of their deposit base.  Under
FIRREA, FDIC could increase assessment rates if the reserve ratio
(that is, the ratio of the value of BIF to the estimated insured
deposits held in BIF member banks) was less than 1.25 percent and the
higher rate would bring the reserve ratio to 1.25 percent within a
reasonable period.  Also, the reserve ratio was capped at 1.50
percent.  The amount of the annual increase was capped at 7.5 cents
per $100 of deposits.  However, FDIC could not raise the rate if the
reserve ratio of the BIF was improving. 

The three BIF provisions in OBRA 90 (1) authorized FDIC to increase
the assessment rates as needed, (2) authorized FDIC to make midyear
adjustments in assessment rates, and (3) removed the cap on the BIF
reserve ratio established by FIRREA.  Together these provisions had
the effect of eliminating restrictions on the amount of annual
assessment increases. 

According to CBO's official cost estimate, this provision had no
budgetary impact.  This was because CBO's baseline in effect at the
time of the budget resolution assumed that FDIC would not use its
existing authority under FIRREA to increase assessments.  Therefore,
increasing FDIC's authority would not result in higher premiums. 

However, in a footnote to this estimate, CBO noted that CBO's
baseline assumed that the 1991 assessment rate would be 15 cents per
$100 of assessable deposits and that FDIC had already increased the
rate to
19.5 cents for 1991.  CBO noted that if the rate were raised to 23
cents for 1992 (in fact, the rate was 23 cents in 1992), the net
decrease in outlays due these administrative actions would be $9
billion over 5 years.  The footnote states that these savings could
be achieved under current law. 

CBO calculated the estimated savings from these administrative
actions by subtracting the baseline level of premium income from the
estimated level resulting from higher assessment rates.  The House
and Senate Budget Committee OBRA 90 conference summary documents
included savings from FDIC's administrative actions and reported that
these provisions produced $9 billion in savings over 5 years. 

According to FDIC officials, these provisions did increase their
flexibility to manage BIF by allowing for more timely increases in
assessments, and they agreed that higher premiums reduced the
deficit.  However, since premium increases were allowable under
pre-OBRA 90 law, they did not believe that OBRA 90 had any direct
impact on premiums--and thus did not reduce outlays and the deficit. 

SECTION 2302:  FLOOD INSURANCE
PROVISION

The National Flood Insurance Program (NFIP), operated by the Federal
Insurance Administration (FIA) within the Federal Emergency
Management Agency (FEMA), was created by the National Flood Insurance
Act of 1968 (Public Law 90-448).  The purposes of NFIP are to (1)
make flood insurance available to property owners, (2) mitigate flood
hazards, and (3) reduce total federal expenditures on disaster
assistance.  Policies are sold only to residents of communities that
choose to join NFIP.  While the federal government is the guarantor
of coverage and bears the risk of loss for all NFIP polices, 85
percent of policies are offered by private insurance companies under
agreements with FIA.  Premium income from policyholders provides the
primary funding source for NFIP. 

OBRA 90 reauthorized NFIP for fiscal years 1992 through 1995.  In
addition, OBRA 90 mandated the creation of a policy service fee to
cover NFIP's administrative costs (that is, salaries, expenses,
mapping activities, flood studies, hazard reduction studies, and
other controllable discretionary activities). 

CBO scored the two components of this provision separately.  CBO
estimated that the policy fee would yield $224 million in savings
over 5 years.  CBO prepared this estimate by multiplying the number
of policies by $25.  CBO scored savings from reauthorization because
reauthorization allowed FIA to continue to sell new policies.  FIA
uses the premium receipts from these policies to offset the costs of
paying out claims.  If the program had not been reauthorized, NFIP
would have continued to pay claims for several years without
receiving any offsetting receipts from new premiums.  CBO estimated
that reauthorizing NFIP would reduce net outlays by $594 million. 

CBO's scoring of savings from reauthorization followed the baseline
rules in effect at the time the estimates were prepared.  However,
BEA mandated that the baseline assume that all expiring programs with
more than $50 million in outlays shall continue.  Thus, under current
baseline practices, CBO would not have scored savings for the
reauthorization component of this provision. 

FIA officials reported that the creation of the policy service fee
did reduce the deficit.  However, they did not believe that the
reauthorization component of this provision reduced the deficit. 
Agency officials stated that it was unlikely that the program would
have been allowed to expire.  In addition, they asserted that if the
federal government were under an accrual accounting system, premium
revenues would be matched against a provision for losses from claims,
and the savings would not exist.  However, as a CBO official pointed
out, the deficit is measured on a cash-flow basis and this provision
did increase the government's cash flow. 

SECTION 2301:  FEDERAL CRIME
INSURANCE PROVISION

The Federal Crime Insurance Program (FCIP), operated by FIA within
FEMA, was created by Title VI of the Housing and Urban Development
Act of 1970 (Public Law 91-609).  The purpose of FCIP is to make
crime insurance available at affordable rates in any state where
there is a critical problem in the availability of crime insurance. 
In 1992, eight states, the District of Columbia, Puerto Rico, and the
Virgin Islands participated in the program.  The FIA uses a service
agent to do processing for FCIP.  Program participants purchase
1-year renewable residential and commercial policies to cover losses
due to robbery and burglary.  Crime insurance policies are sold
either by licensed brokers in participating states or by the FCIP's
service agent.  Premium income from the sales of policies is the
primary funding source for FCIP.  Currently, claims payments and
expenses exceed premium income and FCIP makes up the shortfall by
borrowing from the Treasury. 

OBRA 90 reauthorized FCIP through the end of fiscal year 1995.  CBO
estimated that reauthorizing FCIP through 1995 would reduce net
outlays by $14 million.  As with NFIP, reauthorization allows FCIP to
continue to sell policies and use the income to offset the costs of
paying claims.  CBO also assumed that if the program were
reauthorized, FCIP would continue to pay interest on past borrowing. 

CBO's scoring of savings from reauthorization followed the baseline
rules in effect when estimates were prepared.  As noted earlier, BEA
mandated that the baseline assume that all expiring programs with
more than $50 million in outlays shall continue.  Since NCIP falls
beneath this threshold, CBO would still score savings from
reauthorizing this program. 

Agency officials did not believe that this provision reduced the
deficit.  In their view, reauthorizing FCIP did not produce any
savings because claims and expenses exceed annual premium income. 
However, as a CBO official noted, by allowing FCIP to collect
premiums, the annual losses are reduced. 

SECTION 7101:  FUNDING OF COLAS
FOR POSTAL SERVICE ANNUITANTS AND
SURVIVOR ANNUITANTS

The Postal Reorganization Act of 1970 (Public Law 91-375) created the
U.S.  Postal Service, which replaced the Post Office Department.  The
Postal Service began operations on July 1, 1971.  It was intended to
operate as self-sufficient, independent, business-type entity. 

In 1989, the Postal Service was moved to the off-budget portion of
the federal budget.  Since it is an off-budget entity, the Postal
Service's transactions are excluded from congressional budget
resolutions and totals in the President's Budget.  Currently, only
the Postal Service and Social Security Trust Funds are off-budget. 
Off-budget totals are added to on-budget totals to produce the
unified budget totals for the federal government. 

The Civil Service Retirement System (CSRS), established in 1920, is a
defined benefit plan for federal employees hired before 1984,
including postal workers.  OPM administers this on-budget program. 
Payments to CSRS retirees and their survivors are made by the Civil
Service Retirement and Disability Fund. 

OBRA 90 required the Postal Service to reimburse the Civil Service
Retirement and Disability Fund for COLAs paid to all individuals who
retired between the July 1, 1971, postal reorganization and September
30, 1986, or their survivors.  The provision specified the amounts
the Postal Service would be required to pay to fulfill this
obligation for those fifteen years.  In addition, the Postal Service
must continue to pay COLAs occurring after October 1, 1990, for these
individuals.  The amount of the Postal Service's liability is
prorated to reflect the percentage of an employee's service that
occurred after reorganization.  In essence, this provision directs an
off-budget entity to make payments to an on-budget entity. 

CBO scored this provision based on its general assumption that
agencies will take required actions.  Thus, CBO assumed that the
Postal Service would annually pay the amounts specified in the
statute.  CBO added to these amounts the estimated costs of the CSRS
COLAs for fiscal years 1992 through 1995 to arrive at the total
estimated savings from this provision.  CBO's official estimate for
this provision showed a $1.946 billion decrease in on-budget outlays
and a $1.946 billion increase in off-budget outlays. 

As just noted, off-budget transactions are excluded from budget
resolution totals.  Under this rule, therefore, since this provision
reduces on-budget outlays, it produces savings. 

Postal Service officials did not believe that this provision reduced
the deficit.  Although the Postal Service is making the required
payments, the on-budget savings are offset by increased off-budget
costs.  Thus, this provision has no net impact on the unified budget
deficit.  As a CBO analyst noted, however, the Postal Service is
required to break even over time.  Therefore, if stamp prices were
increased enough to offset the increase in off-budget costs, the
payments to the on-budget entity would reduce the deficit. 


--------------------
\1 The Flood Insurance provision has two components:  reauthorization
and premium increase.  CBO estimated their savings at $594 million
and $224 million, respectively.  Agency officials reported that the
reauthorization component did not reduce the deficit. 

\2 Current law requires that the Social Security trust funds and the
Postal Service be off-budget--excluded from the totals of the
President's budget and the Congress' budget resolutions--even though
these funds are part of total government transactions.  Currently,
the Social Security trust funds and Postal Service are the only
off-budget entities; all other federal and trust funds are on-budget. 
Off-budget totals are added to on-budget totals to produce the
unified budget totals. 


AGENCY IMPLEMENTATION OF OBRA 90
PROVISIONS
=========================================================== Appendix V

Implementation procedures differed for the 38 provisions.  This
appendix describes the different paths used by agency officials to
implement the OBRA 90 provisions. 


      AGENCIES ISSUED A FORMAL
      NOTICE OR BINDING GUIDANCE
      FIRST AND THEN FOLLOWED A
      FORMAL RULEMAKING PROCESS
------------------------------------------------------- Appendix V:0.1

OBRA 90, section 3102, amended the Federal Mine Safety and Health Act
of 1977 by increasing the maximum civil penalty for a violation from
$10,000 to $50,000.  In addition, the maximum penalty for failure to
correct a violation was increased from $1,000 to $5,000 per day.  The
Department of Labor's Mine Safety and Health Administration issued
binding guidance and a notice of proposed rulemaking concurrently in
December 1990.  The binding guidance announced that special
assessments up to the new maximum of $50,000 for violations may be
issued when circumstances warrant while proposed rulemaking was
underway.  The final rule was issued January 1992. 


      AGENCIES FOLLOWED A FORMAL
      RULEMAKING PROCESS
------------------------------------------------------- Appendix V:0.2

OBRA 90, section 4001(b), revises the Medicare payment methodology
for hospital inpatient capital-related costs paid under the
prospective payment system.  During fiscal years 1992 through 1995,
aggregate payments should be made that result in a 10 percent
reduction of what would have otherwise been paid on a reasonable cost
basis.  HCFA administers the program under the Department of Health
and Human Services.  HCFA issued a proposed rule in February 1991 and
a final rule August 1991, effective October 1, 1991. 


      AGENCIES ISSUED A FORMAL
      NOTICE OR BINDING GUIDANCE
      IN LIEU OF A FORMAL
      RULEMAKING PROCESS
------------------------------------------------------- Appendix V:0.3

OBRA 90, section 8012, required VA to assess a $2 copayment for each
30-day supply of medication on an outpatient basis.  The requirement
applied to nonservice disabled veterans and service disabled veterans
with less than 50 percent disability for the treatment of nonservice
related conditions.  VA issued an agency directive in April 1991
announcing the new requirement to all VA offices.  According to a VA
official, a formal notice in the Federal Register was not required. 
VA staff were notified of the new requirement through headquarters
and regional meetings. 


      AGENCIES IMPLEMENTED THE
      PROVISION WITHOUT ISSUING A
      NOTICE, BINDING GUIDANCE, OR
      FORMAL RULE
------------------------------------------------------- Appendix V:0.4

OBRA 90, section 2302, reauthorized the sale of Flood Insurance
Program policies through fiscal year 1995.  It also gave FEMA the
authority to place a surcharge on new and certain existing premiums
to cover administrative costs.  FEMA officials believed that there
was no need to issue guidance because they could not exercise any
discretion regarding implementation of the two components. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix VI

ACCOUNTING AND INFORMATION
MANAGEMENT DIVISION, WASHINGTON,
D.C. 

Margaret T.  Wrightson, Assistant Director
Deborah A.  Colantonio, Evaluator-in-Charge
John W.  Mingus, Evaluator
Warren C.  Underwood, Senior Evaluator
Jeffrey C.  Blaylock, Intern
Andrew K.  Reuter, Intern

*** End of document. ***