[Analytical Perspectives]
[Economic and Accounting Analyses]
[1. Economic Assumptions]
[From the U.S. Government Printing Office, www.gpo.gov]



                                     

========================================================================

                    ECONOMIC AND ACCOUNTING ANALYSES

========================================================================

[[Page 3]] 
   
 
                       1.  ECONOMIC ASSUMPTIONS

                               Introduction
  The economic expansion is about to enter its sixth year. Too often in 
the past when expansions have reached this point, or even sooner, the 
economy has begun to overheat, pushing up inflation and interest rates, 
and ultimately bringing on a recession. In contrast, the policy 
decisions of the last three years have enabled this expansion to attain 
an elusive goal--a ``soft landing'' in which economic growth has slowed 
to a sustainable rate without triggering an increase in unemployment.
  The ``soft landing'' of 1995 is the culmination of three years of very 
successful macroeconomic policy. Over this period, jobs have increased 
and unemployment has fallen, while at the same time, inflation has been 
low and relatively stable. Interest rates have fluctuated, but long-term 
rates are as low as at any time in recent memory. Looking ahead, the 
Administration expects economic growth to continue at a moderate rate 
for the foreseeable future.\1\ Employment is projected to expand 
sufficiently to absorb new workers, keeping the rate of unemployment 
stable. Meanwhile, the Administration expects inflation to continue at a 
low, relatively constant rate, and interest rates to decline further as 
the budget is brought into balance.
  \1\Beyond the next year or two, the Administration does not attempt to 
project the economy's cyclical patterns. The longer term economic 
projections used for the Budget and summarized here are best thought of 
as forecasts of average experience expected to be achieved over a period 
of several years.
---------------------------------------------------------------------------
  The Omnibus Budget Reconciliation Act of 1993 put the Federal budget 
deficit on a downward track that helped to reduce long-term interest 
rates, which in turn helped spark the revival in the economy. The 
Administration's current budget proposals would build on that success 
and cap it with a balanced budget. The Federal Reserve has helped to 
support these needed fiscal actions by pursuing a policy to control 
inflation, while also showing that it is willing to reduce interest 
rates when that is appropriate.
  This chapter begins with a review of recent economic and policy 
developments. With this as background, it then presents the 
Administration's economic assumptions. The assumptions call for a 
continuation of trends already evident in the economy for most of the 
major economic variables. They offer a reasonable and prudent basis for 
making budget projections.
  Two important changes in the statistics on which this forecast is 
based are also described in this chapter. First, real gross domestic 
product (GDP) is now measured on a chain-weighted basis in the National 
Income and Product Accounts. This is reflected in the budget projections 
of real GDP and the aggregate measure of inflation. Second, anticipated 
changes in the calculation of the Consumer Price Index (CPI) will slow 
its growth, and that of related measures of price inflation.
  The chapter compares the Administration's economic assumptions with 
those of the Congressional Budget Office (CBO) prepared at about the 
same time (December 1995). Although there are some differences in the 
underlying policy assumptions on which the two forecasts are based, they 
are quite similar, and the differences between them are well within the 
normal range of forecasting error.
  The chapter also includes an analysis of the impact of changes in the 
economic assumptions since last year's budget on the projected deficit, 
and it concludes with estimates of the sensitivity of the budget to 
changes in economic assumptions.
                           Recent Developments

  1993--Enacting a Responsible Fiscal Policy: The passage of the Omnibus 
Budget Reconciliation Act of 1993 (OBRA93) put fiscal policy on a 
sounder footing and created the preconditions for a healthy expansion. 
The 1992 deficit was $290 billion. Since then, the deficit has fallen 
for three straight years, bringing it down to $164 billion in 1995. That 
is just 2.3 percent of GDP, less than half the level in 1992. The 
improvement in the deficit is traceable to both improvement in the 
economy and to policy changes, of which the President's economic program 
was far and away the most important. The Administration estimated that 
OBRA93 would reduce the deficit during the five years 1994-98 by a 
cumulative total of $505 billion. During the first two years alone, it 
cut deficits by about $130 billion. The economic program has also 
contributed indirectly to the reduction in the deficit by strengthening 
the pace of the economic recovery.
  Stabilizing Inflation: Most previous postwar expansions have ended 
because inflation accelerated, forcing a policy correction. The best way 
to avoid the need for such measures is to act before inflation becomes a 
problem. That is just what monetary policy did during 1994. Entering 
that year, inflation was under control; the CPI had only increased 2.7 
percent over the preceding 12 months. However, 1993 had seen 
unemployment fall by almost a full percentage point as real economic 
growth accelerated, and the economy's momentum was clearly pointing 
towards further large gains in 1994. Those gains were realized, as 1994 
became one of the best years for overall economic performance since the 
end of World War II. During 1994, 3.5 million new jobs were created, and 
the unemployment rate was pulled down by another full percentage point. 
These were welcome developments; but if the economy had continued to 
expand at that rate, shortages of labor
[[Page 4]]
and plant capacity would have 
been sure to emerge, carrying with them a high risk of accelerating 
price increase.
  To avoid that risk, the Federal Reserve raised short-term interest 
rates in several stages during 1994. The intention was to slow growth 
and stabilize unemployment at its new lower levels to avoid the 
inflation risks that faster growth would generate. While the Fed was 
acting to raise short-term rates, investors in the financial markets 
were pushing up long-term rates, anticipating future inflation and the 
possibility of further Fed tightening to choke it off.
  The effect of these developments was seen in 1995. The higher interest 
rates cooled off demand in the economy's interest-sensitive sectors, 
such as housing and consumer durables. In 1995, real GDP rose 1.4 
percent, down from a growth rate of 3.5 percent during the previous 
year.\2\ Although growth slowed, the economy continued to generate new 
jobs at a healthy rate, albeit less rapidly than in 1994; and the 
unemployment rate did not increase. Payroll employment rose by 1.7 
million in 1995 and the unemployment rate averaged 5.6 percent for the 
year, which was its lowest level since 1990.
  \2\These rates are based on the new chain-weighted definition of real 
GDP which is explained more fully below.
---------------------------------------------------------------------------
  The slower growth of economic activity and employment was accompanied 
by continued moderation in wages and prices, exactly what the Fed had 
been hoping to achieve when it tightened policy in 1994. The most 
meaningful measure of overall labor compensation, the Employment Cost 
Index, rose 2.9 percent in 1995--virtually the same increase as in the 
previous year.
  Compensation costs were also held down by a significant deceleration 
in employee benefit costs. Health insurance premiums, which had been 
rising at double-digit rates earlier in the decade, were brought firmly 
under control. The spread of innovations in health care delivery helped 
to bring about this moderation. Although slower growth of employee 
health care costs shows up in the aggregate statistics as a decline in 
the rate of increase in compensation, the long-run effect is likely to 
be an increase in workers' take-home pay. Most studies reveal that 
employee benefits are paid for by workers through lower cash wages. A 
reversal of the trend towards increased benefit costs should strengthen 
cash wages in the long run.
  Moderation in labor markets was mirrored in the product markets. At 
the beginning of 1995, the capacity utilization rate in manufacturing 
had reached nearly 85 percent, a level that in the past had initiated an 
acceleration of price increases. By spring, slower growth caused the 
operating rate to return to a range of around 82 percent, a level 
associated in the past with stable price inflation.
  Reflecting this moderation, the CPI rose only 2.5 percent over the 12 
months of 1995, slightly less even than in 1994. The underlying rate of 
inflation, the CPI excluding food and energy, was also well-behaved, 
rising 3.0 percent during 1995. The inflation rate over the three years 
1993-1995 was the best since the mid-1960s.

  Sustaining the Momentum of the Expansion: As it became clear that 
inflation was under control and likely to remain so for some time, the 
Federal Reserve gradually relaxed its previous tightening. Having 
achieved the desired ``soft landing'', the Federal Reserve took steps to 
make sure the economy would not stall out. It reduced the Federal funds 
rate by one-quarter percentage point in July and in December of 1995, 
and again in January of 1996. Judging from the futures market, the 
financial community anticipates a further reduction of about one-quarter 
percentage point by this summer.
  While the Federal Reserve was lowering short-term rates last year, the 
financial markets were lowering long-term rates even more. The inflation 
fears that had troubled the markets in 1994 were succeeded in 1995 by 
the expectation that inflation would remain subdued. Moreover, 
bipartisan agreement that the budget should be balanced in the coming 
years helped further reduce long-term interest rates. From the end of 
1993 to the beginning of 1996, long-term interest rates fell more than 
two full percentage points. Except for a few months in 1993, the last 
time long-term interest rates were this low was in the 1960s. The drop 
in rates last year is expected to set the stage for a pickup in economic 
activity in 1996.
  Lower interest rates and a healthy economic outlook propelled the 
stock market to record levels. Last year, the Dow-Jones industrial 
average rose 36 percent, and other major indexes were up by similarly 
impressive amounts. In the opening months of this year, stock markets 
set a series of new highs. Financial markets fluctuate, and these gains 
will not continue unabated; but the rise in the stock market last year 
will contribute to the forward momentum in the economy in 1996 by 
lowering the cost of capital to business, which should stimulate 
investment, and by raising household wealth, which will boost consumer 
spending.
                          Economic Projections

                                       Table 1-1.  ECONOMIC ASSUMPTIONS\1\                                      
                                  (Calendar years; dollar amounts in billions)                                  
----------------------------------------------------------------------------------------------------------------
                                                                        Projections                             
                                  Actual -----------------------------------------------------------------------
                                   1994     1995     1996     1997     1998     1999     2000     2001     2002 
----------------------------------------------------------------------------------------------------------------
Gross Domestic Product (GDP):                                                                                   
  Levels, dollar amounts in                                                                                     
   billions:                                                                                                    
    Current dollars............    6,931    7,254    7,621    8,008    8,417    8,848    9,295    9,772   10,268
    Real, chained (1992)                                                                                        
     dollars...................    6,604    6,742    6,888    7,047    7,212    7,380    7,553    7,730    7,911
    Chained price index                                                                                         
     (1992=100), annual average    105.0    107.6    110.6    113.6    116.7    119.9    123.1    126.4    129.8
  Percent change, fourth                                                                                        
   quarter over fourth quarter:                                                                                 
    Current dollars............      5.9      4.1      5.1      5.1      5.1      5.1      5.1      5.1      5.1
    Real, chained (1992)                                                                                        
     dollars...................      3.5      1.5      2.2      2.3      2.3      2.3      2.3      2.3      2.3
    Chained price index                                                                                         
     (1992=100), annual average      2.3      2.5      2.8      2.7      2.7      2.7      2.7      2.7      2.7
  Percent change, year over                                                                                     
   year:                                                                                                        
    Current dollars............      5.8      4.7      5.1      5.1      5.1      5.1      5.1      5.1      5.1
    Real, chained (1992)                                                                                        
     dollars...................      3.5      2.1      2.2      2.3      2.3      2.3      2.3      2.3      2.3
    Chained price index                                                                                         
     (1992=100), annual average      2.3      2.5      2.8      2.7      2.7      2.7      2.7      2.7      2.7
                                                                                                                
Incomes, billions of current                                                                                    
 dollars:                                                                                                       
    Personal income............    5,750    6,104    6,416    6,716    7,025    7,337    7,664    8,031    8,434
    Wages and salaries.........    3,241    3,420    3,607    3,801    3,995    4,193    4,403    4,629    4,864
    Corporate profits before                                                                                    
     tax.......................      528      602      650      702      753      800      843      882      917
                                                                                                                
Consumer Price Index (all                                                                                       
 urban):\2\                                                                                                     
    Level (1982-84=100), annual                                                                                 
     average...................    148.2    152.4    156.6    161.3    165.9    170.5    175.3    180.2    185.2
    Percent change, fourth                                                                                      
     quarter over fourth                                                                                        
     quarter...................      2.6      2.7      3.1      2.9      2.8      2.8      2.8      2.8      2.8
    Percent change, year over                                                                                   
     year......................      2.6      2.8      2.8      3.0      2.8      2.8      2.8      2.8      2.8
                                                                                                                
Unemployment rate, civilian,                                                                                    
 percent:                                                                                                       
    Fourth quarter level.......      5.6      5.6      5.7      5.7      5.7      5.7      5.7      5.7      5.7
    Annual average.............      6.1      5.6      5.7      5.7      5.7      5.7      5.7      5.7      5.7
                                                                                                                
Federal pay raises, January,                                                                                    
 percent:                                                                                                       
    Military...................      2.2      2.2      2.6      3.0      3.1      3.1      3.1      3.1      3.1
    Civilian\3\................  .......      2.0      2.0      3.0       NA       NA       NA       NA       NA
                                                                                                                
Interest rates, percent:                                                                                        
    91-day Treasury bills\4\...      4.3      5.5      4.9      4.5      4.3      4.2      4.0      4.0      4.0
    10-year Treasury notes.....      7.1      6.6      5.6      5.3      5.0      5.0      5.0      5.0      5.0
                                                                                                                
Addendum: GDP and incomes, pre-                                                                                 
 revision basis:\5\                                                                                             
                                                                                                                
Gross Domestic Product (GDP),                                                                                   
 current dollars:                                                                                               
    Levels, dollar amounts in                                                                                   
     billions..................    6,738    7,078    7,428    7,805    8,203    8,623    9,058    9,523   10,005
    Percent change, fourth                                                                                      
     quarter over fourth                                                                                        
     quarter...................      6.5      4.2      5.1      5.1      5.1      5.1      5.1      5.1      5.1
    Percent change, year over                                                                                   
     year......................      6.2      5.0      5.0      5.1      5.1      5.1      5.1      5.1      5.1
                                                                                                                
Incomes, billions of current                                                                                    
 dollars:                                                                                                       
    Personal income............    5,702    6,054    6,357    6,654    6,960    7,270    7,595    7,958    8,358
    Wages and salaries.........    3,279    3,450    3,631    3,826    4,020    4,220    4,431    4,658    4,895
    Corporate profits before                                                                                    
     tax.......................      525      572      608      657      706      749      790      826      859
----------------------------------------------------------------------------------------------------------------
NA=Not Available.                                                                                               
\1\Based on information available as of mid-January 1996.                                                       
\2\CPI for all urban consumers. Two versions of the CPI are published. The index shown here is that currently   
  used, as required by law, in calculating automatic adjustments to individual income tax brackets. Projections 
  reflect scheduled changes in methodology.                                                                     
\3\Percentages for 1994-1996 exclude locality pay adjustments. Percentages to be proposed for years after 1997  
  have not yet been determined.                                                                                 
\4\Average rate (bank discount basis) on new issues within period.                                              
\5\Because the comprehensive revision to the National Income and Product Accounts (which include GDP and        
  incomes) was delayed due to furloughs of Government employees, some budget estimates are based, at least in   
  part, on GDP and incomes data on the pre-revision basis shown in this addendum.                               

  Key assumptions: The economic projections underlying this budget are 
summarized in Table 1-1. They are based on several key assumptions. 
First and foremost, the projections assume that the Administration's 
budget will be adopted. The budget proposals are intended to reduce the 
deficit progressively and achieve a small surplus in 2002, according to 
Congressional Budget Office assumptions, and in 2001 according to 
Administration estimates. Such a policy would foster a continuation of 
the favorable macroeconomic trends that have emerged since 1992. Deficit 
restraint moderates inflationary pressures by restraining demand. It 
enables the Federal Reserve to continue its recent policy of easing 
short-term interest rates. The combination of easier monetary policy and 
fiscal restraint provides an environment in which financial markets can 
keep 
[[Page 5]]
long-term interest rates on a downward path. A policy to balance 
the budget would thus encourage investment, and thereby raise the level 
of productivity and potential output in the long run.
  Real GDP: Economic growth was temporarily restrained in the fourth 
quarter of last year by two shutdowns of the Federal Government, and in 
the first quarter of this year by a record-breaking blizzard. According 
to preliminary estimates, real GDP grew at a 0.9 percent annual rate in 
the fourth quarter; based on partial information, first quarter growth 
may also be relatively weak.
  Growth is expected to pick up as the negative impact of the recent 
disruptions fades. Interest-sensitive sectors, such as consumer durables 
and business equipment spending, are likely to be at the leading edge of 
the acceleration in response to the fall in long-term interest rates 
during 1995 and the surge in the stock market. On average, real GDP is 
forecast to increase 2.2 percent over the four quarters of 1996.
  During 1997-2002, real GDP is projected to rise 2.3 percent annually 
(the Administration's estimate of the economy's potential growth rate). 
Lower interest rates and smaller deficits are projected to increase 
investment and raise the trend growth in output per hour. Productivity 
in the nonfarm business sector had been
[[Page 6]]
growing at 1.1 percent per year 
on average since 1973, but it is projected to increase 1.2 percent 
annually over the next six years.
  Potential GDP growth is also determined by growth of the labor force. 
Labor force participation trends of recent years are assumed to 
continue. The rise in the female participation rate is expected to be 
much less than during the 1970s and 1980s, while the male rate is 
expected to continue to decline. On balance, there is likely to be 
little overall change in labor force participation. During 1997-2002, 
the labor force is projected to grow 1.1 percent per year, about the 
same pace as during the past six years, but noticeably slower than the 
1.7 percent rate during the 1980s when female participation rates rose 
rapidly.

  Unemployment rate: The civilian unemployment rate, which averaged 5.6 
percent during the fourth quarter of 1995, is expected to average 5.7 
percent this year and hold at that level through the end of the 
projection period. With real GDP projected to rise at the rate of growth 
of potential output, the unemployment rate would remain stable.
  Inflation: The chain-weighted GDP price index is projected to rise 2.7 
percent a year over the projection horizon. That is just slightly faster 
than the 2.5 percent estimated for 1995. The Consumer Price Index is 
expected to rise 3.1 percent during 1996, about the same as the 3.0 
percent rise last year in the CPI excluding food and energy. The CPI is 
expected to rise 2.9 percent in 1997 and 2.8 percent per year during 
1998-2002. The deceleration is due to scheduled improvements in the 
methods used to calculate the CPI. These improvements are discussed 
later in this chapter.
  Interest rates: Short- and long-term rates are projected to fall as a 
result of the reduced borrowing needs of the government that result from 
the Administration's budget proposals. The 91-day Treasury bill rate is 
expected to fall to 4.0 percent by 2000 and hold at that level through 
2002; in the fourth quarter of 1995, the rate was 5.3 percent. The yield 
on the 10-year Treasury note is projected to decline to 5.0 percent by 
1998 and hold at that level; in the fourth quarter of last year, the 
yield was 5.9 percent. These projections, in combination with a forecast 
of stable inflation, imply a reduction in real interest rates to levels 
that prevailed when the Federal budget was close to balance. The sharper 
fall in short rates will cause the yield curve to steepen, which is a 
more typical pattern for an expansionary period.
  Incomes: As a result of the drop in interest rates, the share of 
nominal GDP accounted for by personal interest income, a component of 
personal income, is expected to decline. On the other hand, the 
corporate sector is a net borrower, so the profits share and the share 
of dividend income are likely to grow because of the reduction in 
interest costs. The projected share of wages and salaries in GDP is 
expected to remain about unchanged over the projection horizon. After 
adjustment for inflation, real wages and salaries are projected to 
increase 14 percent from 1996 to 2002.
                        Statistical Improvements
  The economic assumptions incorporate two important changes in the way 
economic activity is measured.

  Fixing Biases in Real GDP: For fifty years, the featured measure of 
real GDP was based on a fixed-weight price system, with an update every 
five to ten years to account for shifts in spending patterns. While 
convenient and familiar, that system introduced a ``substitution bias'' 
into the estimate of real GDP and the GDP implicit price deflator. The 
bias was significant whenever relative prices changed rapidly--as for 
example in the 1970s, when oil prices jumped sharply. Until the recent 
revision, 1987 was the base year for the fixed-weight price system. The 
large drop in the quality-adjusted price of computers since 1987 caused 
a growing upward bias in the measurement of real GDP growth.
  To remove these biases, the Bureau of Economic Analysis changed to a 
chain-weighted system for estimating real GDP in January of 1996. The 
weights are now based on nearly contemporaneous spending patterns. Real 
GDP growth for 1993, for example, is calculated using average 
expenditure weights for 1992-1993, and the growth rate for 1994 is 
computed using an average of 1993-1994 spending. Thus, the weights are 
linked year-to-year, hence the term ``chain weights.''
  The substitution bias in the former fixed-weight system distorted the 
picture of real growth and aggregate inflation. The shift to chain 
weights lowered the measured rate of real GDP growth in 1993-94 by about 
\1/2\ percentage point yearly compared with the previous estimate, and 
raised the estimate of aggregate inflation by a similar amount. While 
converting to chain weights provides a more accurate measure of the 
Nation's economic performance, it does have one inconvenience. Real GDP 
no longer exactly equals the sum of real spending by households, 
businesses and governments--the familiar rule that GDP = C+I+G+net 
exports. Now there is a difference, known as ``the residual,'' that 
needs to be added to the components to sum to real GDP.

  Changing the CPI: The CPI is one of the most important statistics 
produced by the Federal Government. It is widely used to measure changes 
in the cost of living.\3\ The CPI's effect on the budget is pervasive; 
it is linked by formula to spending for social security, Federal 
pensions, and many smaller programs, and to the tax brackets and 
exemptions in the individual income tax. It is estimated that a 
reduction of 0.1 percentage point in the average yearly rate of change 
in 
[[Page 7]]
the CPI would reduce the budget deficit by a total of about $20 
billion over the next seven years.
  \3\This is done even though the CPI is explicitly not a cost-of-living 
index. Rather, it measures changes in the average cost of a fixed market 
basket of goods and services. By design, the CPI does not allow for 
those changes in consumption patterns that people make routinely to 
maintain their standard of living when prices are changing.
---------------------------------------------------------------------------
  Given its importance, it is not surprising that the CPI has often been 
criticized. There is no perfect price index, but the Bureau of Labor 
Statistics (BLS), which computes the CPI, strives to eliminate potential 
biases from the index. Over the years, the BLS has been receptive to 
suggestions for improvements. BLS is the main source of the technical 
analysis needed to make such improvements, and it is often the first to 
highlight potential problems.
  Much recent criticism has suggested that the CPI may overstate 
inflation. Various possible causes have been examined. One major problem 
is how to separate quality changes from price increases for goods and 
services. For example, if the price of a visit to the doctor goes up, 
how much of this is due to better service due to improved diagnostic 
equipment and new testing procedures, and how much is a pure price 
increase? Such questions are hard to answer, but critics believe BLS too 
often treats quality improvements as price rises. Another problem area 
is the exclusion of new products or new outlets from the sample used to 
determine the CPI. There are good practical reasons why it takes time to 
incorporate new items into that sample, but the effect may be to miss 
some important price declines that occur as new products and services 
enter the market.
  Finally, there are some technical issues concerning how the CPI is 
measured and put together. BLS has announced that it will introduce two 
methodological improvements in the CPI over the next three years that 
should make the index more accurate. These changes are expected to 
reduce the annual rate of growth of the index by about 0.3 percentage 
points.
  The announced improvements (along with recent revisions to GDP) will 
also narrow the wedge between the rates of change in the CPI, on the one 
hand, and the price indexes for consumer expenditures and for GDP in the 
National Income and Product Accounts on the other. During 1998-2002, the 
annual growth in the CPI is assumed to be 2.8 percent, almost the same 
as the 2.7 percent assumed for the chain-weighted price index for GDP.
  By January 1997, BLS plans to institute new estimation procedures to 
correct what has sometimes been called ``formula bias,'' but which might 
be more accurately described as ``sample rotation bias.'' These new 
procedures are estimated to reduce the growth of the CPI by about 0.2 
percentage point per year. The bias arises because of the need to update 
the sample of items entering the CPI. New brands and varieties of goods 
are continually being introduced in the marketplace, and if the CPI is 
to remain current, it must be based on the current brands of cereals, 
toothpaste, automobiles, et cetera. When new goods are introduced, 
however, the usual BLS procedures can generate inappropriate weights for 
those that are temporarily selling at either abnormally low or 
abnormally high prices. The problem is greatest for items with prices 
that fluctuate around a trend, such as fruits and vegetables. 
Recognizing this, BLS instituted a correction for some components of the 
index in January 1995. One possible course is to apply the same type of 
correction throughout the index.
  Correcting the sample rotation bias in the CPI will also reduce the 
rate of change in the price indexes used to determine real personal 
consumption expenditures in the national income and product accounts, 
which are based on detailed data from the CPI. The effect of a slower 
rise in consumer prices is expected to hold down the growth of the 
overall GDP price index by about 0.1 percentage point yearly. Consumer 
expenditures account for about two-thirds of GDP, and the rest is not 
affected by the change. Measured real GDP growth will, of course, 
increase by a similar magnitude (because total nominal spending growth 
is a datum that is not affected by this change).
  The second scheduled improvement in the CPI is an updating of the 
fixed market basket that is expected to occur in January 1998. 
Currently, the CPI market basket is based on 1982-1984 consumption 
patterns; in 1998, the market basket will be updated to reflect 1993-
1995 spending patterns. This ``rebasing'' of the index occurs about 
every 10 years. Rebasing tends to reduce the measured inflation rate in 
subsequent years by reducing the substitution bias that builds up over 
time as the economy moves away from the base period prices. The new 
weights tend to give more emphasis in the index to goods whose prices 
have been rising relatively less rapidly (because consumers tend to 
shift their consumption toward those items). The budget assumes that the 
change in the CPI market basket will slow the growth of the CPI by about 
0.1 percentage point per year beginning in 1998. This improvement will 
not affect real GDP or the price indexes associated with it.
  These improvements in the CPI will go some way towards correcting its 
apparent tendency to overstate inflation. The largest potential biases--
quality measurement and adjustments for new goods--will not be addressed 
by these changes. Continued research in these areas by BLS and outside 
experts is needed to improve this vital economic statistic.
                           Comparison with CBO
  The Congressional Budget Office (CBO) prepares forecasts of the 
economy that are used by Congress in formulating budget policy. Thus, it 
performs a similar function to that of OMB, the Council of Economic 
Advisers and Treasury for the Executive Branch. While outside observers 
have often compared the CBO forecast with that of the Administration, 
the budget is usually prepared well before the current CBO forecast is 
made public, so a timely forecast comparison is generally impossible.
  Over the past year, however, there has been heightened interest in the 
economic assumptions used for the budget and in the differences between 
Administration and CBO forecasts. That is because the fiscal policy
[[Page 8]] 
objective is now to achieve a balanced budget, rather than a specific 
amount of deficit reduction. Even small differences in economic 
assumptions can matter for the size of policy changes needed to achieve 
budget balance. When the goal is a specific amount of deficit reduction, 
differences in economic assumptions usually have little bearing on the 
size of policy changes needed to achieve a specific amount of budgetary 
savings.

  Post-Policy vs Pre-Policy: One important difference between CBO and 
the Administration concerns the policy assumptions on which the forecast 
is based. The Administration projections always assume that the 
President's budget proposals will be enacted as proposed; the economic 
projections are ``post-policy.'' CBO normally assumes that current law 
will continue; it is a ``pre-policy'' projection.
  This difference often is immaterial in determining the major 
macroeconomic variables. Important as budget policy is, especially in 
the long run, even large dollar changes in programs will often have only 
a modest effect on real GDP or inflation. Therefore, a specific budget 
proposal may make little difference to the macroeconomic outlook. Thus, 
comparisons of CBO and Administration economic projections can be 
meaningful even when the policy assumptions are not identical. Sometimes 
the difference is crucial, however, and that was the case in 1995.

  The Fiscal Bonus: The Administration's policy is to balance the budget 
over the next seven years. The decision to seek a balanced budget has 
major implications for the economic outlook. Such a significant change 
in policy, if enacted, would be likely to cause noticeable changes in 
several macroeconomic variables, especially interest rates and income 
shares. However, CBO's initial forecast for the 1996 budget (and the 
Administration's) assumed that the deficit would not be eliminated over 
this time period.
  In April, CBO presented its estimates of the fiscal bonus that would 
result from balancing the budget following the policies in the 
congressional budget resolution. This bonus took account of the more 
favorable interest rate outlook that would result from a balanced 
budget. It did not, however, reflect the likely shifts in income among 
sectors of the economy that would follow from the lower interest rates 
generated by a balanced budget. This was corrected in December, when a 
revised CBO forecast was prepared that took into account the full range 
of macroeconomic effects that a balanced budget would produce.

  The Treatment of Statistical Biases: The statistical biases in the 
measurement of real GDP and inflation described above posed problems for 
forecasters. Neither CBO nor the Administration was completely 
consistent in dealing with these issues. In some cases, projected 
economic variables reflected the bias that was built into their 
measurement; in other cases, the projections assumed that the bias would 
be corrected somehow during the course of the forecast. In any case, the 
revisions to GDP that were made in January and the planned modifications 
to the CPI go a long way toward removing this source of past difference 
in the forecasts.
  Projection Comparison: The main outlines of the Administration's 
current forecast were determined in December at about the time that CBO 
made public its economic projections. A comparison of the two forecasts 
(including the CBO fiscal bonus to put them on the same policy basis) 
reveals a convergence of views summarized in Table 1-7.