[Congressional Record Volume 141, Number 164 (Monday, October 23, 1995)]
[Senate]
[Pages S15445-S15463]
From the Congressional Record Online through the Government Publishing Office [www.gpo.gov]




                        THE CONSUMER PRICE INDEX

  Mr. MOYNIHAN. Mr. President, some 32 years ago, in the administration 
of John F. Kennedy, I became Assistant Secretary of Labor for Policy 
Planning and Research. This was a new position. In this new position, I 
was nominally responsible for the Bureau of Labor Statistics. I say 
nominally out of respect for the independence of that venerable 
institution which long predated the Department of Labor itself. The 
then-commissioner, Ewan Clague, could not have been more friendly and 
supportive and in time I grew to know more of the field. At that time 
the monthly report of the unemployment rate was closely watched by 
capital and labor, as we would have said, and was frequently 
challenged. Committees regularly assembled to examine and debate the 
data. Published unemployment rates, based on current monthly survey 
methodology appeared, if memory serves, in 1948 and so the series was 
at most 14 years in place at this time. By contrast, the Consumer Price 
Index dated back to 1919. And yet, while the statisticians were 
increasingly confident of the accuracy by which they measured 
unemployment, they were never entirely happy about the CPI. Its 
computation was, and remains, a difficult and ever-changing effort. In 
particular, the statisticians worried that the Consumer Price Index was 
increasingly used as a surrogate for the cost-of-living index. They 
felt this would lead to great troubles as surely the CPI overstated 
inflation. I think they would have been even more alarmed to know that 
in the two decades that followed we would use the CPI to index some 30 
percent of Government outlays and 45 percent of Government revenues.
  This problem inevitably grew more salient at times of true inflation. 
Thus, on October 26, 1980, an article in the Business and Finance 
section of the Washington Post described the election difficulties 
President Carter was facing owing to double-digit inflation. The story 
noted ``The consumer price index overstates the impact of inflation, 
the White House contends.'' As we know, it contended to no avail, but 
the difficulties with the CPI as a proxy for the cost of living 
continued.
  In the spring 1981 issue of the Public Interest, Dr. Robert J. 
Gordon, now chairman of the department of economics at Northwestern 
University, wrote:

       . . . the [United States] CPI is probably the single most 
     quoted economic statistic in the world.

  We are now slowly waking up to the further fact, well known in the 
economics and statistics communities, that the Consumer Price Index is 
not a measure of the change in the cost of living. It is so stated in a 
pamphlet published by the Bureau of Labor Statistics entitled 
``Understanding the Consumer Price Index: Answers to Some Questions'':

       Is the CPI a cost-of-living index?
       No, although it frequently and mistakenly is called a cost-
     of-living index. The CPI is an index of price change only. It 
     does not reflect the changes in buying or consumption 
     patterns that consumers probably would make to adjust to 
     relative price changes. For example, if the price of beef 
     increases more rapidly than other meats, shoppers may shift 
     their purchases away from beef to pork, poultry, or fish. If 
     the charges for household energy increase more rapidly than 
     for other items, households may buy more insulation and 
     consume less fuel. The CPI does not reflect this substitution 
     among items as cost-of-living index would. Rather, the CPI 
     assumes the purchase of the same market basket, in the same 
     fixed proportion (or weight) month after month.

  Despite this caution from the agency that compiles the CPI, the index 
is used as a yardstick for adjusting Government benefits, including 
Social Security, and provisions of the Internal Revenue Code.
  And yet, it is now well recognized that changes in the CPI overstate 
the change in the cost of living.
  The administration recognizes this fact.
  Congress recognizes this fact.
  And a Commission of eminent economists appointed by the Senate 
Finance Committee recognizes this fact.
  In an October 3, 1994, memorandum entitled ``Big Choices,'' Dr. Alice 


[[Page S 15446]]
Rivlin, then Acting Director of OMB and now Director--and a 
distinguished economist who has served as the president of the American 
Economic Association--noted that among the options available to reduce 
the budget deficit were several COLA proposals including, and I quote:

       CPI minus 0.5 ``technical'' reform (CPI may be overstated 
     by 0.4% to 1.5%).
       CPI minus 2 for five years.

  The budget resolutions passed by the Senate and House built into 
their baseline lower CPI assumptions than were projected by CBO in 
January. The lower assumptions reflect the expectation that scheduled 
BLS revisions of the CPI will lower the reported CPI. The Senate 
assumed a two-tenths of a percentage point adjustment; the House 
assumed a six-tenths of a percentage point adjustment. The conference 
report adopted the Senate version.
  In their report--Senate Report 104-82--the Senate Budget Committee 
noted:

       In January, CBO projected CPI inflation would remain at 3.4 
     percent for 1998 and thereafter. The downward revision 
     reported here relative to the January figures reflects CBO's 
     new appraisal that the 1998 benchmark revision to the CPI 
     planned by the Bureau of Labor Statistics will likely reduce 
     the rise in the computed measure of the CPI by 0.2 percentage 
     points a year. Federal Reserve Chairman Greenspan and CPI 
     experts have recently testified before the Senate that 
     incomplete evidence suggests CPI inflation may be overstated 
     by as much as 1.0 to 1.5 percentage points a year. However, 
     in advance of further, more conclusive analysis, CPI biases 
     remain speculative and have not been incorporated into the 
     Committee assumptions.

  And the budget resolution, adopted by the Senate on May 25, 1995, 
contained this language:

     SEC. 304. NONPARTISAN ADVISORY COMMISSION ON THE CPI.

       (a) Findings.--The Congress finds that--
       (1) Congress intended to insulate certain government 
     beneficiaries and taxpayers from the effects of inflation by 
     indexing payments and tax brackets to the Consumer Price 
     Index (CPI);
       (2) approximately 30 percent of total Federal outlays and 
     45 percent of Federal revenues are indexed to reflect changes 
     in the CPI; and
       (3) the overwhelming consensus among experts is that the 
     method used to construct the CPI and the current calculation 
     of the CPI both overstate the estimate of the true cost of 
     living.
       (b) Sense of the Senate.--It is the sense of the Senate 
     that--
       (1) a temporary advisory commission should be established 
     to make objective and nonpartisan recommendations concerning 
     the appropriateness and accuracy of the methodology and 
     calculations that determine the CPI;
       (2) the Commission should be appointed on a nonpartisan 
     basis, and should be composed of experts in the fields of 
     economics, statistics, or other related professions; and
       (3) the Commission should report its recommendations to the 
     Bureau of Labor Statistics and to Congress at the earliest 
     possible date.

  The conference agreement on the concurrent budget resolution for 
fiscal year 1996 passed the Senate on June 29, 1995. The conference 
report included the following:

     SEC. 309. SENSE OF THE SENATE ON THE ASSUMPTIONS.

       It is the sense of the Senate that the aggregates and 
     functional levels included in this budget resolution assume 
     that--

                           *   *   *   *   *

       . . . (6) a temporary nonpartisan commission should be 
     established to make recommendations concerning the 
     appropriateness and accuracy of the methodology and 
     calculations that determine the Consumer Price Index (CPI) 
     and those recommendations should be submitted to the Bureau 
     of Labor Statistics at the earliest possible date.

  Earlier, on March 13, April 6, and June 6, the Finance Committee held 
hearings on this subject. Testimony was received from 13 established 
economists who collectively represented virtually all the expertise 
that exists on this issue.
  A remarkable consensus emerged at those hearings.
  I ask unanimous consent that a list of the witnesses, along with 
their affiliations, and their estimates of the degree to which changes 
in the CPI overstate changes in the cost of living be printed in the 
Record.
  There being no objection, the list was ordered to be printed in the 
Record, as follows:

                     Estimates of CPI Overstatement

                 (In order of appearance of witnesses)

       March 13, 1995 Hearing:
       Chairman Alan Greenspan, Federal Reserve: 0.5 to 1.5 
     percentage points.
       Cmsr. Katharine Abraham, Bureau of Labor Statistics (BLS): 
     No estimate offered.
       Dr. Robert Gordon,\1\ Northwestern University Dept. of 
     Economics: Minimum of 1.7 percentage points.
       Director June O'Neill, Congressional Budget Office: 0.2-0.8 
     of a percentage point (based on CBO report 10/94).
       April 6, 1995 Hearing:
       Dr. Dale Jorgenson,\1\ Harvard University Dept. of 
     Economics: Around 1 percentage point.
       Dr. W. Erwin Diewert, Univ. of British Columbia/Dept. of 
     Economics: 1.3 to 1.7 percentage points.
       Dr. Ariel Pakes, Yale University Dept. of Economics: 0.8 of 
     a percentage point.
       Dr. Joel Popkin, Popkin & Co. (former Assistant 
     Commissioner for Prices and Living Conditions at BLS): No 
     estimate offered.
       June 6, 1995 Hearing:
       Dr. Michael Boskin,\1\ Senior Fellow, Hoover Institute, 
     Stanford Univ.: At least 1.0 percentage point, maybe 2.0 
     percentage points.
       Dr. Ellen Dulberger,\1\ Director, Strategy and Economic 
     Analysis IBM: CPI overstatement is greater than others have 
     stated and likely to grow.
       Dr. Zvi Griliches,\1\ Harvard University Dept. of 
     Economics: 0.4 to 1.6 percentage points.
       Dr. Janet Norwood, Senior Fellow, Urban Inst. (former BLS 
     Commissioner): No estimate offered.
       Dr. Robert Pollak, University of Washington Department of 
     Economics: No estimate offered.

     \1\ CPI Commission members.

       Average of Mid-Point Estimates by CPI Commission Members: 
     1.3 percentage points at a minimum (assumes Dulberger's 
     minimum is 1.3 points, the average of other four members).

  Mr. MOYNIHAN. Mr. President, again: Dr. Alan Greenspan, Chairman of 
the Federal Reserve Board--0.5 to 1.5 percentage points.
  Dr. Dale Jorgenson, chairman of the department of economics at 
Harvard University--around 1 percentage point.
  Dr. Robert Gordon, chairman of the economics department at 
Northwestern University--at least 1.7 percentage points. Note that in 
1981 Professor Gordon wrote the Public Interest article, cited earlier, 
in which he laid out many of the issues related to the accurate 
measurement of changes in the cost of living.
  Dr. Michael Boskin, professor of economics at Stanford University and 
Chairman of the Council of Economic Advisers in the Bush 
administration--at least 1 percentage point, maybe 2 percentage points.
  In all, 9 of the 13 witnesses provided numerical estimates of the 
overstatement. The average of the estimates: about 1.1 percentage 
points. The calculation is based on a minimum estimate for some 
witnesses. Even if we assume a zero estimate of the overstatement for 
those who provided no estimate--and few, if any, would so contend--the 
average for all the witnesses would be 0.8 of a percentage point.
  Not too different from the 0.4 to 1.5 percentage points noted by OMB 
Director Rivlin in her memo last October.
  The complete record of these hearings is printed as Senate Hearing 
104-69--Consumer Price Index. I hope Senators will obtain copies and 
review the hearing record.
  Following the hearings, then Finance Committee Chairman Packwood and 
I, as ranking member, announced on June 26, 1995, the appointment of a 
nonpartisan Commission to:

       . . . study the methodology used to calculate the Consumer 
     Price Index (CPI) and to advise Congress on whether this 
     methodology provides an accurate measure of the cost of 
     living.

  At that time I stated:

       . . . Current law makes it clear that certain federal 
     programs should be adjusted for changes in the cost of 
     living. What is not clear is whether changes in the CPI, 
     which is used as a proxy for changes in the cost of living, 
     accurately measures these changes. A study by a non-partisan 
     commission will provide invaluable advice to Congress on this 
     important issue.

  The Commission, chaired by Dr. Michael Boskin, issued its interim 
report on September 15, 1995.
  The report, ``Toward a More Accurate Measure of the Cost of Living,'' 
included the following observations and conclusions in the executive 
summary:

       . . . While the CPI is the best measure currently 
     available, it is not a true cost of living index (this has 
     been recognized by the Bureau of Labor Statistics for many 
     years). Despite important BLS updates and improvements in the 
     CPI, changes in the CPI have substantially overstated the 
     actual rate of price inflation, by about 1.5% per annual 
     recently. It is likely that a large bias also occurred 
     looking back over at least the last couple of decades, 
     perhaps longer, but we make no attempt to estimate its size.
     
[[Page S 15447]]

       . . . Changes in the CPI will overstate changes in the true 
     cost of living for the next few years. The Commission's 
     interim best estimate of the size of the upward bias looking 
     forward is 1.0% per year. The range of plausible values is 
     0.7% to 2.0%. The range of uncertainty is not symmetric. It 
     is more likely that changes in the CPI have a larger than a 
     smaller bias.
       . . . The upward bias programs into the federal budget an 
     annual automatic real increase in indexed benefits and real 
     tax cut.

  Let me now elaborate on the implications of these points made by the 
Commission.
  Current law requires the Government to adjust some benefits and tax 
provisions for changes in the cost of living.
  The 1972 Amendments to the Social Security Act included this 
language:

       Section 202. (a) 1 Section 215 of the Social Security Act 
     is amended by adding at the end thereof the following new 
     subsection: Cost-of-Living Increases in Benefits.

  Similarly, section 104(f)(3) of the Economic Recovery Tax Act of 1981 
states:

       . . . the cost of living adjustment for any calendar year 
     is the percentage . . .

  The objective of these statutes is clear: Benefits and Tax Code 
provisions should be adjusted for changes in the cost of living. 
However, the law stipulates that the adjustments should be based on 
changes in the CPI as a proxy for changes in the cost of living. But 
with mounting evidence that changes in the CPI overstate changes in the 
cost of living, implementation of the policy is thwarted. The law is 
being thwarted.
  What can be done to ensure that the policies Congress has adopted are 
faithfully executed? That is, how can we ensure that adjustments in 
benefits and Tax Code provisions more accurately reflect changes in the 
cost of living? Two things.
  First, continue to support ongoing efforts by the BLS in its routine 
updating and rebenchmarking of consumer expenditure patterns, and in 
its research activities. Talented and dedicated BLS researchers have 
identified many of the complex measurement issues that must be 
addressed when compiling a CPI in a world in which the quality of 
products changes and new goods are introduced with resolute regularity.
  Second, Congress must recognize that, despite the best intentions of 
the BLS as it continues with its updates and research, the CPI is not, 
as the BLS readily acknowledges, a cost-of-living index. To achieve its 
policy objectives--so clearly stated in the law--Congress must 
implement legislative corrections that, when combined with the most 
accurate CPI that the BLS can produce, will result in changes in 
benefits and Tax Code provisions that accurately reflect changes in the 
cost of living.
  As noted earlier, the Boskin commission on the CPI suggests that for 
now, the correction Congress should adopt is 1 percentage point.
  The Commission's report also highlights the budget implications of 
failing to correctly implement policies designed to adjust for changes 
in the cost of living. We should not harbor any misgivings merely 
because these changes will dramatically improve the budget outlook. The 
error is there and should be corrected without regard to budget 
implications.
  Even so, it must be acknowledged that the budget implications are 
enormous. One could say awesome.
  CBO estimates a cumulative 10-year reduction in the deficit of $634 
billion from a 1 percentage point downward adjustment in automatic 
changes of benefits and tax provisions. By the 10th year the annual 
reduction in the deficit is almost $140 billion. Extrapolating from 
these CBO projections, my staff estimates the 12-year cumulative 
reduction in the deficit at almost $1 trillion.
  And the corrections affect both sides of the budget ledger. About 
one-half of the cumulative reduction in the deficit is due to lower 
outlays; one-third due to higher revenues, and the remainder results 
from reductions in interest payments.
  And while we are thinking about saving the Social Security trust 
fund, consider this fact. Harry Ballantyne, Chief Actuary of the Social 
Security Administration, estimates that the date of exhaustion of the 
OASDI fund is extended by 19 years from 2030 to 2049 by a 1 percentage 
point downward adjustment in the CPI.
  Exhaustion is defined as the year in which the trust fund has used up 
all its reserves of Treasury securities with the expectation that 
annual outlays will continue to exceed annual income.
  This is a real fiscal dividend. We can get things right and save the 
trust fund.
  Mr. President, I ask unanimous consent that the following reports and 
documents cited in my remarks be printed in the Record after my 
statement.
  First, ``The Consumer Price Index: Measuring Inflation and Causing 
It'' by R.J. Gordon, 1981, in the Public Interest 63: Spring.
  Second, ``Understanding the Consumer Price Index: Answers to Some 
Questions'' by the U.S. Department of Labor, Bureau of Labor 
Statistics, May 1994.
  Third, ``Toward a More Accurate Measure of the Cost of Living'' by 
the Advisory Commission to Study the Consumer Price Index, September 
15, 1995.
  Fourth, table on the change in deficit from a downward adjustment in 
the CPI of 1 percentage point by the Congressional Budget Office, March 
15, 1995.
  Fifth, memorandum prepared by Harry C. Ballantyne, September 28, 
1995, on: Estimated Long-Range Effects of Alternative Reductions in 
Automatic Benefit Increases.
  There being no objection, the material was ordered to be printed in 
the Record, as follows:

                [From the Public Interest, Spring 1981]

      The Consumer Price Index: Measuring Inflation and Causing it

                         (By Robert J. Gordon)

       Inflation is widely believed to be the most important 
     economic problem facing the United States and most other 
     countries in the world. Thus it is not surprising that the 
     monthly publication of the U.S. Consumer Price Index (CPI) is 
     so closely watched both inside and outside of government. 
     Large increases in the CPI are bad news for Administration 
     officials, particularly in election years, and may lead to 
     sudden policy reversals such as the introduction of the 
     Carter Administration's ill-fated credit controls in March 
     1980. Large increases in the CPI, however, are good news for 
     millions of recipients of social security benefits, 
     government retirement pay, and other payments that by law or 
     contract must be escalated in step with the CPI. Also, since 
     foreigners watch the CPI closely for clues to the future 
     course of U.S. interest rates and the exchange value of the 
     dollar, the CPI is probably the single most quoted economic 
     statistic in the world.
       Imagine that someone pushes the wrong button on a computer 
     at the Bureau of Labor Statistics (BLS), the division of the 
     Department of Labor that is responsible for the CPI, and 
     records that the increase in the CPI over a particular year 
     is 15 percent instead of the true rate of 10 percent. 
     Government officials would probably react with restrictive 
     policy measures--some combination of expenditure reductions, 
     tax increases, and higher interest rates. Thousands, 
     perhaps millions, of Americans might be thrown out of 
     work. Millions of others receiving social security 
     benefits or union wages escalated by the CPI would enjoy a 
     windfall gain, since their payments would go up by more 
     than the true inflation rate. The unnecessary extra 
     benefit payments would cause the government deficit to 
     balloon, putting extra pressure on the Federal Reserve to 
     print more money and finance still more inflation, while 
     the higher union wage payments would put pressure on firms 
     to raise prices faster than otherwise.
       Exactly this chain of events occurred in the United States 
     in 1979 and 1980, but not because of an easily correctable 
     slip by BLS. Instead, a serious overstatement of inflation by 
     the CPI was caused by built-in design flaws. These defects 
     have come to light not through the snooping of some 
     measurement-minded Woodward or Bernstein, but rather as a 
     result of a growing discrepancy between the CPI and a 
     competing government measure of consumer prices called the 
     ``Personal Consumption Expenditures deflator,'' published by 
     a division of the Department of Commerce, and usually called 
     the ``PCE deflator'' for short. Table I shows that after 
     registering only a small difference in early 1978 and most 
     earlier years, the inflation rate recorded by the two indexes 
     grew apart by an amount that reached an annual rate of 5 
     percent in the first half of 1980.

   TABLE I.--INFLATION RATES AS ESTIMATED BY THE CPI AND PCE DEFLATOR   
                [Percentage changes at annual rates] \1\                
------------------------------------------------------------------------
                                                       PCE              
                                            CPI     Deflator  Difference
------------------------------------------------------------------------
1947-77........................     3.4        3.3       0.1            
1978-80 by half year...........  ......  .........  ........            
1978, first half...............     8.9        8.3       0.6            
1978, last half................     9.0        6.8       2.2            
1979, first half...............    12.6       10.0       2.6            
1979, first half...............    13.0        9.8       3.2            

[[Page S 15448]]
                                                                        
1980, first half...............    16.2       11.2       5.0            
------------------------------------------------------------------------
\1\ Source: CPI from Bureau of Labor Statistics, PCE Deflator from      
  Survey of Current Business, various issues. These figures do not      
  reflect the data revisions announced in December 1980 for the PCE     
  deflator. A preliminary inspection suggests that the inflation rate of
  the PCE deflator in the new data is between 0.5 and 1.0 percentage    
  points lower for each period shown since 1977. Because the CPI has not
  been revised, the difference between the two indexes has been further 
  enlarged by the revisions.                                            


       The story of the two inflation indexes is a fascinating 
     one, even for those whose eyes glaze over at talk of 
     measurement procedures and who prefer to treat government 
     economic data as unchallenged gospel. Since the CPI and PCE 
     deflator are compiled from a common set of underlying price 
     data by two different sets of rules, part of the tale 
     involves the rules themselves, why the lead to 
     different results, and why the CPI rules are widely 
     believed to be inferior to those used in the PCE deflator. 
     Another aspect involves the internal workings of the BLS, 
     where staff bureaucrats have long urged the replacement of 
     obsolete rules for the measurement of housing prices but 
     were forced by political pressure to retain the old rules 
     in the new version of the CPI introduced in 1978. A final 
     and less-reported chapter involves the adequacy of the 
     underlying price data that both the CPI and PCE deflator 
     share in common. These form the basis for all economic 
     measures of real economic progress, or the lack of it, 
     including those that show a drastic slowdown in the growth 
     of U.S. productivity in the last decade. Howe effectively 
     do official procedures handle innumerable situations when 
     a new model or product costs more than the item it 
     replaces, but differs in quality as well? New radial tires 
     last longer than the old bias-ply type, and recent-vintage 
     television sets both perform better and need fewer repairs 
     than their predecessors. But if price indexes are not 
     adjusted adequately for these quality improvements, 
     inflation is overstated and the improvement in our 
     productivity and standard of living is understated.


                          a two-class society?

       The CPI was first published by the BLS in 1919 to help set 
     wage levels for workers in shipbuilding yards, and its use as 
     a standard for wage increases has always been one of its main 
     purposes. Currently about 8 million workers are covered by 
     collective bargaining contracts that provide for increases in 
     wage rates based on increases in the CPI, and these wages set 
     a pattern that millions of other workers try to emulate. More 
     recently, many types of government payments have been linked 
     to the CPI. Among those who reap a windfall if the annual CPI 
     increase is overstated are 31 million social security 
     beneficiaries and 2.5 million retired military and Federal 
     Civil Service employees and survivors. Others receive 
     payments geared to a particular component of the CPI, 
     especially 20 million food stamp recipients and 25 million 
     children who eat federally subsidized school lunches. In all 
     about half the population, including dependents is affected 
     by changes in the CPI.
       The use of escalator clauses has created a two-class 
     society, separating those who are protected against 
     inflation, legally or by contract, from those who are not. 
     Steelworkers, Chicago bus drivers, and other union members 
     enjoying generous escalator clauses have moved several steps 
     up the relative income ladder at the expense of white-collar 
     workers and others whose wages are not escalated. Social 
     security recipients enjoyed a 14.4 percent boost in benefits 
     in July 1980, as compared to an increase in the government's 
     average hourly earnings index of only 9.2 percent in the year 
     ending that month. Use of that earnings index rather than the 
     CPI for escalation in 1980 would have reduced the federal 
     deficit by about 8 billion. Use of the PCE deflator would 
     have been almost as desirable, saving about $6 billion.\1\ 
     Thus some of the much-discussed financial crisis of the 
     Social Security System results from the use of the CPI for 
     escalation purposes.
     \1\ The actual social security increase was based on the CPI 
     change in the twelve months ending in March, 1980.
---------------------------------------------------------------------------
       While adjustment of payments is the most tangible function 
     of the CPI, there are two other uses which figure prominently 
     in discussions of economic performance and policy. The first 
     and most obvious is that the CPI itself is a readily 
     available measure of inflation and serves as a widely-quoted 
     verdict on the success or failure of economic policy. The 
     second is that the individual CPI item indexes for pork 
     gasoline, and other products are the sources of other price 
     indexes. The CPI and PCE deflator displayed in Table I are 
     both based on the same price-change data for pork and 
     gasoline, but they combined these individual item indexes 
     with different weights. Because the Commerce Department 
     procedures put less weight on energy prices, which rose 
     rapidly during the 1978-80 period (as well as no weight at 
     all on mortgage interest rates), they yield a slower overall 
     increase when the PCE deflator is added up. It is the PCE 
     deflator, and the broader ``GNP deflator'' of which it is a 
     major component, that allow the Commerce Department to 
     translate data a current-dollar sales and personal income 
     into quarterly estimates of real Gross National Product. The 
     basic measure of the economy's productive performance.\2\ 
     Real GNP, in turn, is divided by BLS data on hours spent at 
     work to yield data on the nation's hourly productivity.
     \2\ About two-thirds of Gross National Product consists of 
     Personal Consumption Expenditures deflated by the PCE 
     deflator. The other third consists of construction spending, 
     business equipment purchases, government wages and purchases 
     of goods, and the excess of exports over imports. Each of 
     these other components has its own deflator based on a wide 
     variety of data sources.
---------------------------------------------------------------------------


                    the ever-changing market-basket

       The CPI reports the price in any given month of a so-called 
     ``fixed market-basket'' of commonly purchased items. Today's 
     price of the market-basket is expressed relative to what the 
     same items would have cost in 1967, the arbitrary ``base 
     year'' of the index. As shown on the top line of Table II, 
     the CPI was at a level of 251.7 in September 1980, 
     indicating that items costing $10,000 in 1967 would have 
     cost $25,170 if purchased in September 1980. Public 
     attention tends to focus on recent changes in the CPI 
     rather than on the cumulative change since 1967. Thus, 
     newspaper reports do not highlight the index level of 
     251.7, but rather the change over the past year and month. 
     In September 1980, the change in the CPI over the previous 
     year registered 12.7 percent, and the change from August 
     to September was 1.0 percent, usually expressed at an 
     annual rate. The sense of panic that surrounded the Carter 
     Administration's economic policy in March and April of 
     1980 was directly set off by three consecutive monthly CPI 
     increases of 1.4 percent, or 18.2 percent when expressed 
     as an annual rate.

       TABLE II.--A SAMPLE OF CPI ITEM INDEXES, SEPTEMBER 1980.\1\      
------------------------------------------------------------------------
                                                               Percent  
                                               Index Level   change from
                                               (1967=100)     September 
                                                                1979    
------------------------------------------------------------------------
All items...................................         251.7          12.7
White bread.................................         219.6           9.4
Sirloin Steak...............................         280.9          11.9
Eggs........................................         179.9           5.4
Potatoes....................................         313.2          57.2
Roasted coffee..............................         426.1           0.0
Whiskey.....................................         137.6           6.7
Residential rent............................         195.1           9.0
Contracted mortgage interest................         500.9          26.3
Fuel oil....................................         585.4          21.3
Telephone services..........................         137.0           3.5
Television..................................         105.0           2.0
Women's dresses.............................         168.5          -1.5
New cars....................................         181.7           9.4
Airline fares...............................         310.3          44.9
Hospital room...............................         428.4          13.8
School books and supplies...................         221.0           9.7
------------------------------------------------------------------------
\1\ Source: Consumer Price Index Detailed Report, September 1980.       

       The task of constructing the CPI involves (1) determining 
     what people buy, (2) determining where they buy, and (3) 
     determining what they pay for what they buy. The first task 
     was carried out by the BLS and Census Bureau in 1972-74 and 
     involved quarterly interviews with about 20,000 families and 
     a survey of another 20,000 families who were asked to keep 
     diaries of small, frequent purchases for two weeks. Because 
     this effort of carrying out the Consumer Expenditure Survey 
     is so complex and expensive, Congress is only willing to 
     allocate funds for such a survey every decade. The 
     previous Consumer Expenditure Survey had been carried out 
     in 1960-61 and was the basis of the CPI until 1977. Thus 
     in late 1977 the ``old CPI'' was based on expenditure data 
     that were sixteen years out of date, and the ``new CPI'' 
     introduced in 1978 was based on an expenditure survey that 
     was already five years out of date.
       Determining where people buy, so that the right amount of 
     information might be collected from particular retail 
     outlets, discount stores, and mail-order houses, was 
     accomplished by a ``point-of-purchase'' survey of another 
     23,000 families in the early 1970s. This scientific basis for 
     the collection of price data represents a substantial 
     improvement on the arbitrary choices of outlets in the CPI 
     for earlier years. With the allocation of individual items 
     and retail outlets established by these various surveys, the 
     month-to-month job of collecting the actual price quotations 
     is carried out by BLS data collectors who have considerable 
     latitude to choose the specific brands and types of goods to 
     be priced each month within the general item definitions laid 
     down by the central BLS office. An incredible total of one 
     and a half million individual price quotations are obtained 
     each year, of which 700,000 are for food, 100,000 are for 
     rent and property taxes, and the remainder are for other 
     items. Data sources, called ``reporters,'' include about 
     2,300 food store outlets, 18,000 rental units, 18,000 housing 
     units, and 22,300 other sources.


                 the importance of weighting procedures

       Every month the CPI publishes an overall index, summary 
     indexes for major groups of items like food and apparel, and 
     about 250 item indexes, a few of which are shown as examples 
     in Table II. What is striking here is the wide variety of 
     price increases registered by different items since 1967, 
     ranging from 5 percent for television sets to 485 percent for 
     fuel oil. Clearly the overall inflation rate registered by 
     the CPI depends on how much weight is attached to each item. 
     Someone who spends equal shares of his income on rent, TV 
     sets, telephone calls, eggs, and whiskey, would have 
     experienced a price increase since 1967 of only 51 percent, 
     or a compounded rate of only 3.2 percent per 

[[Page S 15449]]
     year. Someone else who spends equal shares on steak, potatoes, coffee, 
     fuel oil, and mortgage interest, would have experienced an 
     increase since 1967 of 321.3 percent, or a compounded rate of 
     11.7 percent per year. Since average hourly earnings 
     increased by 7.5 percent over the same period, the first 
     spending pattern would have allowed a substantial increase in 
     real income, whereas the second pattern would have resulted 
     in a drastic drop in real income.
       Consumers are under constant pressure to shift their 
     spending patterns to avoid goods that have unusually high 
     price increases--for example, to reduce fuel usage in favor 
     of wool sweaters, or to shift from coffee to whiskey. Any 
     index like the CPI that uses fixed expenditure weights must 
     exaggerate the inflation rate as compared to an index like 
     the PCE deflator that uses current weights, since the CPI 
     assigns relatively large weights to high-inflation items like 
     fuel oil and coffee based on their shares in consumer 
     expenditure in the ``good old days'' of 1972-73, before the 
     consumer reaction against their increase in price. The fixed 
     weights used in the CPI would not be an important defect if 
     all products changed in price by roughly the same amount over 
     long periods of time. But the large variety of price changes 
     between 1967 and 1980 displayed by the index numbers for 
     individual items in Table II has made the fixed-weight 
     problem a source of upward bias in the CPI during the past 
     three years, as obsolete weights magnify the high inflation 
     rates of products like fuel oil.
       How much of an exaggeration in the CPI's measured inflation 
     rate is caused by this so-called ``substitution bias''? We do 
     not learn the answer to this question by examining the 
     massive differences between the CPI and PCE deflator 
     displayed in Table I, since these are largely caused by other 
     factors besides substitution. Instead, we can determine the 
     contribution of consumer substitution away from high-
     inflation items by examining the effect of three different 
     weighting schemes for the data used in the PCE deflator. The 
     first is the scheme used in the published ``implicit PCE 
     deflator'' itself. Table III shows an example of how the 
     implicit PCE deflator would be calculated for a simple 
     economy consisting only of spending on coffee and whiskey. 
     Sections 1 and 2 exhibit prices and quantities in three 
     different periods: the 1972 base period and two successive 
     quarters in 1980. Section 3 multiplies price times quantity 
     in each period to obtain actual expenditures. Section 4 then 
     computes ``real'' expenditures in constant 1972 prices by 
     multiplying the actual quantities purchased in each period by 
     the constant prices of 1972.


                 the consumer price index and inflation

  Table III.--METHODS OF CALCULATING PRICE INDEXES (FOR A HYPOTHETICAL  
                              ECONOMY) \1\                              
------------------------------------------------------------------------
                                                             1980       
                                                     -------------------
                                              1972      First    Second 
                                                       quarter   quarter
------------------------------------------------------------------------
                        THE HYPOTHETICAL ECONOMY                        
1. Prices:                                                              
    Coffee per pound......................        $1        $4        $5
    Whiskey per bottle....................        $5        $5        $5
2. Units sold:                                                          
    Pounds of coffee......................         5         3         2
    Bottles of whiskey....................         1         2         3
                                           =============================
3. Actual expenditures:                                                 
    Coffee................................        $5       $12       $10
    Whiskey...............................        $5       $10       $15
                                           -----------------------------
      Total...............................       $10       $22       $25
                                           =============================
4. Real expenditures in 1972 prices:                                    
    Coffee................................        $5        $3        $2
    Whiskey...............................        $5       $10       $15
                                           -----------------------------
      Total...............................       $10       $13       $17
                                                                        
                 THE EFFECTS ACCORDING TO THREE INDEXES                 
                                                                        
5. Implicit PCE deflator..................       100       169       147
6. Chain index of 1980 change.............  ........  ........      11.7
7. 1972 fixed-weight index................       100       250       300
------------------------------------------------------------------------
\1\ Notes: The implicit PCE deflator in section 5 is 100 times the ratio
  of total actual expenditures (section 3) to real expenditures (section
  4).                                                                   
The Chain Index in section 6 multiplies the price change for the second 
  quarter of 1980 for each item (25 percent for coffee, zero for        
  whiskey) by the average expenditures share of each product in both    
  quarters of 1980 (22/47 and 24/47, respectively).                     
The Fixed-weight Index in line 7 multiplies the level of the item index 
  for each period (100, 400, and 500 for coffee; 100 each period for    
  whiskey) by that item's share in 1972 expenditures (50 percent for    
  each product in this case).                                           

       The PCE deflator is simply defined as the ratio of actual 
     expenditures to real expenditures, and this is written in 
     section 5, along with the percentage change between periods. 
     This extreme example reveals a defect of the PCE deflator, 
     which uses weights that shift each period. The alteration in 
     weights in successive periods causes the deflator to mix up 
     the measurement of price changes with the effect of shifting 
     weights. Thus, in the second quarter of 1980 the price of 
     coffee increases by 25 percent, and the price of whiskey 
     stays constant, but the PCE deflator registers a 13 percent 
     decline in spite of the fact that no single price has 
     dropped! Why? Expenditures in that quarter have shifted 
     toward whiskey, which has had no price increase at all since 
     the base year of 1972; thus the higher weight increases the 
     influence of whiskey's cumulative absence of price change 
     since 1972, which has nothing to do with actual inflation in 
     1980.\3\
     \3\ If the same example were recalculated for a deflator 
     using a base of 1980, second quarter (rather than 1972), the 
     result would be an increase in the deflator of 14 percent 
     rather than a decline of 13 percent.
---------------------------------------------------------------------------
  

       How can we obtain the advantage of the up-to-date weights 
     used in the PCE deflator without the deflator's disadvantage 
     of mixing together price changes and weight changes? This is 
     accomplished by the ``chain index,'' which is calculated by 
     averaging together the changes in individual prices between 
     and periods rather than by computing an index level as in the 
     case of the implicit deflator. These individual changes are 
     weighted by the average share of expenditures of each 
     category in the two adjacent quarters taken together. In our 
     example the increase in the chain index is 11.7 percent 
     (shown in section 6), which makes intuitive sense as an 
     average of the 25 percent increase in the price of coffee and 
     the zero percent increase in the price of whiskey. (Since the 
     share of expenditures on constant-price whiskey is a bit more 
     than half in the two quarters, $25/$47, the chain index comes 
     out showing a bit less of an increase than a simple 
     unweighted average of 25 and zero).
       Finally, the third alternatives is to combine the coffee 
     and whiskey prices with fixed 1972 expenditure weights. This 
     creates an index analogous to the CPI. As shown in section 7, 
     the fixed-weight index yields a 20 percent price increase for 
     the second quarter of 1980, reflecting the higher weight of 
     coffee in 1972 spending patterns. In this extreme case the 
     bias in the fixed-weight index stemming from consumer 
     substitution is represented by the difference between the 20 
     percent increase in the index compared to the 11.7 percent 
     increase in the chain index.
       While real-world price changes vary all over the map, the 
     relatively large share in spending of items experiencing 
     roughly average price increases makes the problem of consumer 
     substitution in the actual CPI less important than in our 
     extreme example. This is shown in Table IV, which displays an 
     array of price change indexes, ranging in order from the 
     implicit PCE deflator in section 1 to the CPI itself in 
     section 5. The five indexes here allow us to decompose the 
     difference between the implicit PCE deflator and the CPI into 
     three main factors. The chain index in section 2 differs from 
     the implicit deflator in section 1 by eliminating the 
     undesirable impact of changing weights, thus the difference 
     between section 2 and section 1 shows the modest quantitative 
     impact of shifting weights. Next, section 3 lists the PCE 
     deflator recalculated with fixed 1972 weights. The difference 
     between this fixed-weight version of the PCE deflator and the 
     chain index in the section above shows the effect of consumer 
     substitution away from items with rapidly rising prices. The 
     difference is negligible in 1977 and 1978 but became 
     magnified in 1979 and 1980, largely due to the over-weighting 
     of energy prices in the fixed-weight index. Nevertheless, in 
     the first half of 1980 shifting weights and the substitution 
     effect together contributed only 0.8 out of the 4.4 
     percentage point difference between the Consumer Price Index 
     and the implicit PCE deflator.

           Table IV.--FIVE MEASURES OF INFLATION, 1977-80 \1\           
                              [In percent]                              
------------------------------------------------------------------------
                                                               Late 1979-
                                 Late       Late       Late        mid  
                               1976-77    1977-78    1978-79      1980  
------------------------------------------------------------------------
1. PCE deflator.............        5.6        7.4        9.9       11.6
2. PCE deflator with ``chain                                            
 weights''..................        6.0        7.8       10.3       11.9
3. PCE deflator with ``fixed                                            
 weights''..................        5.9        7.9       10.7       12.4
4. CPI with PCE treatment of                                            
 home ownership.............        6.3        7.9       10.8       12.2
5. CPI......................        6.8        9.0       13.3      16.0 
------------------------------------------------------------------------
\1\ Source: Alan S. Blinder, ``The Consumer Price Index and the         
  Measurement of Recent Inflation,'' Brookings Papers on Economic       
  Activity, vol. 11 (1980, no. 2), Tables II, IV and VI.                
                                                                        
Note.--CPI figures are for December through December, or December       
  through June in the last column. PCE deflator figures are for fourth  
  quarter through fourth quarter, or fourth quarter through second      
  quarter in the last column.                                           

                     accounting for home ownership

       The bulk of the excessive inflation rate measured by the 
     CPI can be explained by its bizarre treatment of home 
     ownership. Section 4 displays a special version of the CPI 
     that replaces the actual home ownership component by the PCE 
     measure and weighting of home ownership cost. The difference 
     between the actual CPI in section 5 and the special version 
     in section 4 shows that the choice of home ownership 
     treatment makes an enormous difference, a full 3.8 percentage 
     points in the first half of 1980.
       Far from being a source of higher prices, squeezed budgets, 
     and falling living standards, most Americans have found home 
     ownership to be a source of wealth creation and one of the 
     few spots in the family budget that is largely insulated from 
     inflation. The treatment of homeownership in the CPI makes 
     the fatal error of treating the whole population as if it 
     were in the predicament of a newlywed couple buying its first 
     house. This unlucky pair, late arrivals on the housing 
     inflation merry-go-round, over the past several years has 
     indeed faced a substantial increase in the monthly payment 
     required to own its first house. But the vast majority of 
     home owners has been protected from these higher costs. 
     Increases in home purchase prices for existing home owners 
     are a source of higher wealth, and ``leverage'' (the small 
     initial share of their down-payment equity) makes the value 
     of their equity increase by a multiple of the percentage 
     annual increase in house prices. Because income is 
     properly defined as consumption plus the change in 

[[Page S 15450]]
     one's wealth, higher home prices by this definition also raise 
     individual incomes. Increases in mortgage interest rates 
     do not represent a higher cost for holders of existing 
     mortgages, since most of these were negotiated at fixed 
     interest rates. The monthly payment to the local savings 
     bank is the same today as it was in the month of the first 
     payment when the house was purchased two or five or 
     fifteen years ago, and thus is a steadily falling 
     proportion of annual earnings that allows the paycheck to 
     be diverted to other needs. Home ownership has been a 
     blessing--a source of wealth and six-figure balance sheets 
     for many Americans--rather than the curse that the CPI's 
     treatment would imply.
       In Table V the housing component of the PCE deflator is 
     compared with the various parts of the rent and home 
     ownership component of the CPI. It is evident that the 
     difference between the PCE and CPI treatments involves both 
     the weights and the actual price increases registered by the 
     individual components. The housing component represents 17.4 
     percent of the weight in the PCE deflator, as contrasted with 
     the 30.2 percent weight for rent and home ownership together 
     in the CPI. The increase in the PCE component in the year to 
     September 1980 was only 9.0 percent, as compared to a 
     weighted average of 15.4 percent for rent and home ownership 
     together in the CPI. There are numerous weak points, both 
     major and minor, in the CPI treatment of housing. The most 
     important are (1) the overweighting of the home-purchase and 
     mortgage-interest-rate components, (2) the treatment of 
     existing mortgage contracts as involving variable rather than 
     fixed rates, and (3) the failure to subtract from the higher 
     home prices and mortgage rate the benefits that consumers 
     receive from interest tax deductions and from the capital 
     gains due to higher house prices.

Table V.--RENT AND HOME OWNERSHIP COSTS: CPI WEIGHTS AND PRICE INCREASES
                                   \1\                                  
                              [In percent]                              
------------------------------------------------------------------------
                                              Weight in   Annual rate of
                                                total         change    
                    Item                        index,    September 1979-
                                               December   September 1980
                                                 1979                   
------------------------------------------------------------------------
A. PCE deflator housing component..........         17.4            9.0 
B. CPI components:                                                      
    1. Residential rent....................          5.3            9.0 
    2. Home ownership......................         24.9           16.8 
        Home purchase......................         10.4           13.8 
        Contractual mortgage interest cost.          8.7           21.8 
        Property taxes.....................          1.7            3.5 
        Property insurance.................          0.6           13.6 
        Maintenance and repairs............          3.4            9.0 
------------------------------------------------------------------------
\1\ Sources: CPI: Same as Table II. PCE Deflator: Survey of Current     
  Business, October 1980. PCE data refer to the quarter in which        
  indicated month occurred.                                             

       1. Overweighting of home purchase prices and mortgage 
     interest rates. Table V shows that the weight attached to 
     mortgage interest is almost as large as that attached to home 
     purchase. The CPI makes the incredible error of treating home 
     purchase and mortgage interest payments as separate unrelated 
     transactions; it counts the house price once as the weight 
     for home price changes and then counts most of it again as 
     the weight for changes in mortgage interest rates. This 
     double-counting can be appreciated in an example involving a 
     new home purchased for $40,000 in 1972, financed by a 20 
     percent down payment ($8,000) and a twenty-five-year $32,000 
     mortgage taken out at a typical 1972 interest rate of 7.5 
     percent.\4\ The BLS procedure computes the weight for the 
     purchase price component from the 1972-73 consumer 
     expenditure survey based on purchases of newly constructed 
     houses; if every survey respondent had annual consumption 
     expenditures of $20,000, and 5 percent of them purchased a 
     new $40,000 house, this would yield a weight for a home 
     purchase of 10 percent. But that is not all. Fully half of 
     the mortgage payments over the 25 year term ($26,429, in this 
     case) is included as an additional expenditure, so that 
     mortgage interest costs receive a weight of 6.6 percent in 
     this example. A minimum requirement for consistency in the 
     CPI should be that the weight on housing reflects the amount 
     actually spent--$40,000 in this case. People do not buy 
     houses and mortgages separately; they obtain mortgages so 
     that they do not actually have to lay down $40,000 in cash!
     \4\ This example is taken from the article by Alan Blinder 
     cited in the note to Table IV.
---------------------------------------------------------------------------
       2. Assumption of variable rates on all existing contracts. 
     The CPI does not describe the housing-cost experience of 
     actual U.S. homeowners but rather of a fictitious society in 
     which the interest rate on all outstanding mortgages is 
     renegotiated every month. Imagine that the average mortgage 
     lasts 10 years, and that the mortgage rate has risen in the 
     past decade from 5 to 15 percent at a pace of exactly one-
     twelfth of a percentage point every month. Then the average 
     rate paid on outstanding contracts would be 10 percent. Now 
     imagine that on January 1, 1981, the rate on mortgage 
     closings suddenly jumps from 15 to 17 percent. The CPI uses 
     the mortgage closing rate for the first five days of the 
     previous month, and so in this example the mortgage 
     component of the February 1981 CPI would show an increase 
     of 13.3 percent. If all other items were increasing at an 
     average of 1 percent per month, or 12.7 percent per year, 
     this treatment of the mortgage interest rate would be 
     enough to cause scare headlines, since the annual rate of 
     increase of the all-items CPI in February would be 27.9 
     percent. But in truth, since a single month is initially 
     involved and the average mortgage lasts for ten years, 
     less than one percent of total mortgage payments are 
     affected by the new rate. The average mortgage interest 
     rate paid would change from 10.0 to 10.1 percent, for an 
     increase of just one percent, exactly the same as the 
     assumed increase in all other items. Scare headlines would 
     be avoided, and the February announcement of the CPI would 
     report an annual rate of increase of 12.7 rather than 27.9 
     percent.
       3. Use of actual rather than real after-tax interest rate. 
     Does a higher mortgage interest rate actually raise the true 
     cost of borrowing, as assumed by the CPI? Not necessarily, 
     because borrowing cost consists of the actual interest rate 
     paid, less the percentage increase in the price of the item 
     purchased with the borrowed funds, less any tax deductions 
     for interest paid. Sensible home owners and business 
     borrowers know that a 15 percent interest rate is not a 
     suffocating burden if borrowing allows them to buy cheap now 
     and sell dear later. In fact it is easy to show how an 
     increase over a decade from a 5 to 15 percent mortgage rate 
     actually could have reduced real borrowing costs. Imagine 
     that over the same period the inflation increased from zero 
     to 10 percent, and that the income tax rate remained fixed at 
     20 percent. Since all interest paid (not just the net-of-
     inflation part) is deductible, the real cost of borrowing can 
     decline if inflation is high enough.


            The home-ownership blunder, and how to right it

       There are no defenders of the present treatment of home 
     ownership costs in the CPI, which has remained essentially 
     unchanged since 1953.\5\ Yet year after year between 1977 and 
     1980 its damage grew as escalated union wages, government 
     transfer payments, and the government deficit were pushed up. 
     During the deliberations that led to the 1978 CPI revision, 
     there was unanimous staff support in BLS for killing the 
     present procedure. Yet the staff was overruled by the late 
     Julius Shiskin, then Commissioner, who wrote that ``I have 
     decided that the present treatment will be continued . . . 
     This decision is based on the fact that there is widespread 
     disagreement among the business, labor, and Government 
     advisers to the Bureau of Labor Statistics concerning the 
     approaches to the cost of shelter proposal by the Office of 
     Prices and Living Conditions.'' \6\ One interpretation of 
     this remark is that the last refuge of a bureaucrat faced 
     with controversy is to retain the status quo. Another 
     possibility is that the key word in Shiskin's letter is 
     ``labor,'' and that labor unions were unwilling to accept any 
     tampering with the CPI that might jeopardize the privileged 
     position that they had enjoyed during the 1973-74 high-
     inflation period thanks to their CPI-escalated contracts. In 
     light of the fact that the Carter Administration bowed to 
     union pressure on the issue of the minimum wage, it is not 
     implausible that union pressure was behind Shiskin's 
     decision. In any case there is no doubt that labor unions 
     have been among the main beneficiaries of his vote for the 
     status quo.
     \5\ In January 1981 the BLS announced that ``the much-
     criticized home-purchase component of the consumer price 
     index will be deleted and will probably be replaced with an 
     estimate for rents'' (New York Times, January 29, 1981, p. 
     1). This announcement thus endorses the conclusion of this 
     section (written before the announcement) that the ``rental 
     equivalence'' method should have been used all along. 
     Unfortunately, the change will not be made until 1985, so 
     this section of the text remains relevant for the first half 
     of this decade.
     \6\ Letter from Julius Shiskin to Lyle Gramley of the Council 
     of Economic Advisers, April 15, 1977.
---------------------------------------------------------------------------
       The two main candidates suggested by economists to replace 
     the present treatment are the same as those proposed by the 
     BLS staff during the 1972-77 deliberations on the CPI 
     revision--the ``user cost'' and ``rental equivalence'' 
     approaches. In fact, in an end run around its own index, the 
     BLS now publishes five alternative versions of the CPI using 
     different measures of home ownership cost. Of the five 
     alternatives, four represent different ways of treating user-
     cost, and the fifth is based on the rental equivalence 
     method. (It is the fifth alternative that is displayed on 
     line 4 of Table IV.)
       1. The user-cost of housing. Economists love to dazzle 
     their students with ``user cost'' formulas of the type 
     developed in the early 1960's by Harvard's Dale Jorgenson for 
     the purpose of explaining business investment behavior. The 
     aim is to come up with a figure to represent the amount for 
     which a capital good could be rented. Unlike the present CPI 
     approach, which is based on the current price paid for new 
     houses by the small fraction of people who actually purchase 
     them in a given year, the user-cost approach measures the 
     current annual capital and operating cost of home ownership 
     for everyone. User-cost formulas typically sum up the annual 
     mortgage interest costs, plus the interest that would have 
     been earned on the down payment if it had been invested in 
     a financial asset, plus operating costs like taxes, 
     insurance, and repairs, minus capital gains due to higher 
     house prices, and minus tax deductions made possible by 
     the payment of mortgage interest.
       The basic problem with the user-cost approach is that there 
     are several alternative ways of measuring the ingredients in 
     the formula, especially interest rates, tax rates, and 
     capital gains. Are capital gains to be 

[[Page S 15451]]
     counted as those expected when the mortgage was taken out or those 
     actually realized? Is the mortgage interest rate to be the 
     current rate or an average of past rates? How is the personal 
     tax rate relevant for mortgage interest deductions to be 
     determined? The BLS provides four different measures of user 
     cost to provide a menu of outcomes, and all of them display 
     much more volatility than actual rent. If an economist's 
     approximation of how much a house should rent for does not 
     behave at all like actual observed rents, then that ought to 
     be telling him something.
       2. Rental equivalence. The idea of rental equivalence is 
     simple and in fact is already used in the PCE deflator: 
     Simply assume that the costs of home ownership moves in 
     proportion to actual rents as measured by the CPI rent index, 
     and apply a weight based on the estimated rental value of 
     owner-occupied homes. Residential rent has increased more 
     slowly than the average for other CPI items, and much more 
     slowly than the present CPI home ownership component. 
     Objections to the rental equivalence approach center around 
     the fact that most single-family homes are not rented, and so 
     the rental information collected by the CPI may not reflect 
     hypothetical rents of single-family homes. Nevertheless 
     landlords face the same interest costs as home owners and 
     enjoy roughly the same tax deductions and capital gains. The 
     fact that actual rents exhibit more gradual changes than 
     hypothetical user-cost measures does not necessarily imply an 
     error but rather reflects the tendency for prices of physical 
     goods and services to adjust more slowly to changing 
     conditions than prices of financial assets. Just as a 
     company's stock price typically jumps around much more than 
     the prices of the things it sells, so housing prices and 
     interest rates jump around more than the rental value of 
     houses. This makes sense in the case of rent, since changes 
     in current mortgage interest rates do not actually affect 
     landlords who have long-term fixed-rate mortgages, and 
     changes in current capital gains have no impact (except on 
     paper wealth) if the building is going to be held over a long 
     period rather than sold at today's price.
       Since the rental equivalence method is appealing, why not 
     just adopt it? Use of rent data for the CPI home ownership 
     component would justify expanding the sample of rent 
     information to include more single-family houses. I suspect 
     that much of the resistance to the rent approach stems from a 
     belief that rent data are tainted, since rents have been 
     rising so much less rapidly than the cost of construction (95 
     percent vs. 192 percent, respectively, between 1967 and 
     1980). But there is an economic reason for this divergence. 
     My parents recall renting a house in Berkeley, California, in 
     1938 for $65 per month that was also for sale at the same 
     time for $7,500. The house now would sell for $250,000 but 
     could not rent for $2,167 a month (an equivalent percentage 
     of sale price). In fact, a rent below $1,000 would be typical 
     for the kind of house in the current Berkeley rental market. 
     Why? Landlords and home owners renting out their homes no 
     longer have to recoup all of their cash mortgage interest and 
     operating expenses from rent, since likely taxed capital 
     gains and tax deductions on mortgage interest now pay part of 
     the bill. Thus the slow increase in rents is not a fiction, 
     but reflects economic reality.


                    Accounting for changing quality

       Up to this point all of the issues have involved 
     differences between the CPI and PCE deflator. But now we turn 
     to the question of the changing quality of products, where 
     both indexes are on the same footing because they use the 
     same underlying price figures obtained by the BLS data 
     collectors. When a new model of a product is introduced that 
     contains one or more extra features, part of its higher price 
     may be explained by its higher quality. The gradual 
     acquisition of higher quality goods has been an important 
     source of a rising standard of living for Americans, and so 
     we must make sure that adequate adjustments are made for the 
     fraction of price increases that actually represent higher 
     quality.
       Quality change poses a problem for the CPI, which attempts 
     to measure changes in the price of goods and services in a 
     fixed market basket. The apparently straightforward task of 
     collecting information on the price of a fixed set of goods 
     is continually complicated by the fact that some goods go out 
     of existence to be replaced by new models or new products. 
     The issue of quality adjustments involves precisely how and 
     when the new models are introduced into the overall index.
       Over its history the CPI market basket has continually 
     changed, providing an interesting--though usually out-of-
     date--commentary on social history. From 1918 to 1940, the 
     CPI index that covered shaving was the price of a barber 
     shave, and then switched in 1940 to the safety-razor 
     blade, despite the fact that safety razors had largely 
     replaced other barber shaves in the 1920's. From 1940 to 
     1952 the index item was the blade, joined from 1952 to 
     1964 by shaving cream, followed from 1964 to 1977 by the 
     shaving cream alone, followed since 1977 by a combination 
     of dental and shaving toiletry products. Since 1964 there 
     has been no blade in the CPI, and thus no consideration of 
     the new world opened up for most men by the invention of 
     the double-edged blade in the early 1970's.
       Other products have come and gone as well. In 1940 the 
     index dropped not only barbershop shaves, but also high 
     button shoes, men's nightshirts, and girls' cotton bloomers. 
     The 1953 revision eliminated salt pork and laundry bar soap 
     but added televisions, frozen foods, Coca-Cola, and whiskey. 
     Pajamas, which had replaced nightshirts in 1940, themselves 
     disappeared in 1964, leaving only sheets and blankets to 
     cover the sleeping American male. Appendectomies also 
     disappeared in 1964, the year funeral services were added. 
     Among the new product categories introduced in the 1978 
     revision were pet supplies and expenses, indoor sports 
     equipment, tranquilizers, and electronic pocket calculators.
       How are new models and products introduced into the CPI? 
     There are three main methods.
       1. Direct comparison. When a quality change is considered 
     to be ``small,'' in the judgment of BLS staff members, it is 
     neglected. All of the observed price change would be recorded 
     as a change in the CPI item index, with no adjustment for 
     quality change. If we assume that most model change-overs 
     involve quality improvements, the direct comparison method 
     imparts an upward bias to the CPI--that is, causes it to 
     register too much inflation.
       2. Linking. When the BLS staff members assess the quality 
     change as too important to be ignored, then they introduce a 
     linking procedure. This effectively imputes to the product 
     whose quality changed the price movement of similar goods 
     whose quality did not change. Let us imagine that an old-
     fashioned cotton sheet selling for $5.00 is replaced by a 
     polyester permanent press sheet selling for $8.00 which lasts 
     twice as long. The CPI linking procedure pays no attention to 
     increased durability, but simply replaces the observed price 
     increase by the actual price increase of other unchanged 
     items in the same ``household linens category.''
       3. Cost data. In some cases the BLS obtains the cost of the 
     quality change directly from the manufacturer. First, staff 
     members must determine whether a change claimed by the 
     manufacturer to improve quality actually does so. The 
     criterion for the judgment is whether the change improves the 
     value of the product for the user. (Several years ago the BLS 
     would not include a change by an auto manufacturer from a 
     dial to digital clock on the grounds that this change did not 
     increase the ``user value'' of the automobile.) The value of 
     those quality changes that are not disallowed is based on the 
     manufacturer's estimate of the extra cost involved in making 
     the higher-quality item. This procedure is obviously subject 
     to the flaw that the manufacturer may overstate the cost of 
     the quality improvement in order to disguise a portion of 
     actual price increases, particularly in a period in which 
     government price controls or guidelines are attempting to 
     hold a lid on prices. This source of error would tend to bias 
     the CPI downward and cause it to register too little 
     inflation.
       The automobile is the only product which is given the full-
     blown cost-adjustment treatment. Every September several BLS 
     officials travel to Detroit to consult with the major 
     manufacturers in order to identify those specification 
     changes on new models for which adjustments must be made. If 
     a producer has introduced a new, heavier bumper, whether on 
     its own initiative or to comply with federal safety 
     regulations, the firm is asked to supply an estimate of the 
     difference in the cost of producing the new bumper as 
     compared to the old bumper. This difference in cost is then 
     subtracted from the reported price increase of the new model 
     automobile.
       Because the BLS devotes so much more attention to 
     automobiles than to other products, there is a chance that 
     the recorded differences between the inflation rates 
     registered by autos and other products may reflect differing 
     quality-adjustment procedures rather than a true difference 
     in price behavior. For instance, between 1972 and 1978 the 
     measured price of automobiles went up 27 percent, but the 
     price indexes for other types of moving mechanical equipment 
     like tractors and construction machinery (part of the 
     Producers' Price Index compiled by the BLS) increased by 
     about 80 percent.


             product price cycles and increased performance

       The typical product, whether automobiles in the 1920's, TV 
     sets in the 1950's, or electronic calculators in the 1970's, 
     experiences after its invention an initial period of 
     declining price, as its manufacturers spread the fixed cost 
     of its development over more and more units sold. Then, as 
     a product becomes ``mature,'' there is less opportunity 
     for efficiency gains to cancel out increased wages and 
     other costs, so prices begin to rise. Three aspects of CPI 
     procedures cause it to understate quality improvements and 
     to overstate price change. First, the use of obsolete 
     weights from decade-old expenditure surveys tends to place 
     too little weight on modern products where price increases 
     are relatively slow--this ``consumer substitution'' 
     problem was examined above. Second, new models and 
     products are typically introduced into the index much 
     later than the date when their sales volume becomes 
     important. And finally, the linking procedure, by far the 
     most common quality-adjustment technique used by the BLS, 
     tends both to treat new products as if they were mature 
     products and to ignore performance improvements.
       The long intervals between CPI revisions, and the 
     officially sanctioned tendency for data collectors to cling 
     to existing models 

[[Page S 15452]]
     until they disappear from the marketplace, imply that items with 
     declining prices are typically absent from the index. Albert 
     Rees, who in 1960 performed a fascinating comparison of BLS 
     item indexes with price data for the same products from mail-
     order catalogues, recalls with amusement a visit with a store 
     owner to identify the particular model cooking pot that was 
     then being priced by a BLS field representative. ``Oh, you 
     mean this old model up here on the top shelf. We never sell 
     these any more,'' answered the store owner, ``but that BLS 
     field representative keeps asking us for its price.''
       More important are the new products that enter the CPI late 
     in the product price cycle. The United States became a 
     motorized society in the 1920's and 1930's, when there was an 
     enormous improvement in the performance of automobiles along 
     with a decline in their price--but the automobile was not 
     included in the CPI until 1940. Penicillin entered the CPI in 
     1951, after it had already experienced a 99 percent decline 
     from its initial price. The pocket calculator entered the CPI 
     in 1978, after it had declined in price about 90 percent from 
     early 1970-71 models and about 98 percent from the price of a 
     comparable electromechanical desk calculator of the 1960's.
       The linking procedure misses quality improvements for two 
     reasons. First, as in the cotton sheet example, the price 
     change is taken to be identical to other items in the sample 
     product group that remain unchanged in quality. But these are 
     likely to be mature products experiencing price increases, 
     whereas the item that is improved in quality is more likely 
     to be in the early stage of its product cycle. Perhaps more 
     important, the CPI ignores changes in performance that 
     tend to accompany model changes. In the cotton sheet 
     example, the new sheet lasts twice as long. Since 
     consumers presumably are buying years of service from 
     long-lasting items like sheets, the CPI treatment ignores 
     the lower price of a ``sheet-year,'' since the service 
     life in the example is assumed to double while the price 
     only increases by 60 percent. (It is a sign of the times 
     that many goods like sheets and draperies are officially 
     classified as ``nondurable'' yet actually last longer than 
     many ``durable'' goods.)
       The most striking fact about the treatment of quality 
     change in the CPI is that it is inconsistent with its own 
     stated objective, which is to adjust for changes in quality 
     when they improve the value of a product to the user. In the 
     sheet example and in many others there is no attempt to 
     measure the change in product performance. Consumers value 
     sheet-years, motor-oil-miles, and tire-miles, rather than 
     sheets, quarts of motor oil, and tires independent of their 
     durability. F. Lee Moore has calculated that between 1935 and 
     1978 the price of tires per mile of tire-life declined by 9 
     percent, in contrast to an increase in the CPI tire index of 
     140 percent. Over the same period, the price of motor oil per 
     mile declined by 52 percent as compared to an increase in the 
     CPI of 234 percent.\7\ There are other examples of improved 
     performance that are missed by the CPI's attention to ``price 
     per item'' instead of ``price per service desired by the 
     user.'' Among these are the increased service life of light 
     bulbs, spark plugs, and appliances.
     \7\ F. Lee Moore, ``Index Mischief: Price versus Cost,'' 
     Electric Perspectives, 1978, no. 5, pp. 8-27.
---------------------------------------------------------------------------
       Our previous discussion of the user cost of housing can be 
     applied more broadly to any good which lasts a significant 
     length of time. Consumers care about the total annual 
     operating costs of automobiles and appliances having a given 
     level of performance, not purchase price alone. Auto 
     manufacturers have diverted development efforts from the old 
     concentration on styling and tailfins to a new obsession with 
     increased fuel efficiency. Yet there is no procedure in the 
     CPI to adjust for improvements in automobile fuel 
     efficiency.\8\ A lab at M.I.T. several years ago studied the 
     repair records of appliances and found that the frequency of 
     refrigerator repairs had dropped by a factor of two, and TV 
     repairs by a factor of four, between the mid 1950's and early 
     1970's.
     \8\ In the case of automobiles the BLS has measured the price 
     change on new downsized models as equal to models that are 
     unchanged in size. This is the correct procedure if the fuel 
     savings on the new models just balance the consumer value of 
     the loss in comfort and performance, but not otherwise.
---------------------------------------------------------------------------
       In a study that makes al lowances for improved electricity 
     efficiency and other characteristics, I have estimated 
     that the quality-adjusted prices of refrigerators, washing 
     machines, and air conditioners declined at about twice the 
     rate registered by the CPI between 1950 and the mid 
     1960's.
       Performance improvements are not just limited to goods, but 
     also extend to services. That vanishing breed, the domestic 
     household worker, now accomplishes more per hour with modern 
     appliances and fabrics than her 1925 counterpart, yet her 
     ``price'' is a straight hourly wage. The apparently 
     outrageous increases in hospital room charges exhibited in 
     Table II disguise improvements in the quality of medical care 
     provided to the typical patient, and today's guest at a 
     Holiday Inn or other medium-priced hotel enjoys telephone and 
     television service that was unavailable to his luxury-hotel 
     counterpart of 50 years ago. An airline passenger mile is a 
     more comfortable, faster, and safer, commodity than it was in 
     1955, and yet the CPI prices a homogeneous passenger mile. 
     There is no doubt that train service has deteriorated, but 
     this is of minor importance in an index that keeps its weight 
     up to date.
       Of all products in the U.S. economy, the one displaying the 
     faster rate of price decline throughout the entire postwar 
     era has been the electronic computer. Yet the U.S. government 
     does not compile a price index for computers, so that the 
     output and productivity gains achieved by companies like IBM 
     and the office machinery industry as a whole are not captured 
     by aggregate indexes of output and productivity. This does 
     not involve the CPI directly, because until recently few 
     computers were sold directly to consumers. Government 
     officials are quick to admit that IBM's output and 
     productivity achievements are missed in official data in the 
     year the computers are manufactured, but they claim that the 
     higher efficiency made possible by computers is accurately 
     captured when they are used in subsequent years in the 
     production of consumer goods. This position is partly true, 
     since the use of computers to replace workers in consumer-
     goods factories has contributed to measured productivity 
     advances.
       Yet for a wide variety of consumer services the CPI is not 
     capturing the improvements that the computer has provided. On 
     many airlines computers make possible pre-reserved seats and 
     one-stop check-in, and airline managements were willing to 
     invest in computerized equipment in the belief that consumers 
     should value the extra services provided. Yet the CPI does 
     not value the extra services, treats an airline passenger-
     mile as an unchanged commodity, and leaves the impression in 
     our national data that the investment in the extra computer 
     has produced nothing. The same point applies to 24-hour 
     money machines provided on street corners by banks, and 
     other financial services. It is doubtful that the world-
     wide convenience made possible by major credit cards would 
     have occurred without the computer, yet the CPI ignores 
     the saving of time and fees by consumers who no longer 
     have to purchase so many travelers checks and letters of 
     credit.
       Even the much-criticized U.S. government has been a source 
     of an unmeasured improvement in our standard of living. For 
     25 years we paid an increased gasoline excise tax, treated by 
     the CPI as an increase in the price of gasoline, in order to 
     finance construction of the interstate highway system. 
     Automobile travel is now faster and safer, but this 
     government activity is treated as having only costs, with no 
     benefits.
       The interstate highway example is interesting because it 
     conflicts with a controversial decision that treats anti-
     pollution and safety devices on automobiles in the CPI as an 
     increase in quality rather than an increase in price. 
     Government environmental and safety legislation is treated as 
     having wisely balanced the cost of the devices against the 
     benefits received by the nation as a whole in reduced 
     pollution and greater safety, in contrast to the interstate 
     highway case where benefits are ignored. If government 
     regulatory efforts, like most economic activities, are 
     subject to increasing costs and diminishing benefits as more 
     and more of the pollution is eliminated, then the CPI 
     treatment may have been conservative a decade ago, in the 
     early stages of regulation, but overly generous recently. The 
     growing consensus that many recent government regulations do 
     not provide benefits to balance their costs would imply that, 
     at least for this one reason, the Consumer Price Index 
     understates inflation.
       As we plunge further into the murky depths of index-making, 
     at some point we leave the realm of the statistician and 
     enter the realm of the philosopher. Where do we draw the line 
     between a new model of an old product and an entirely new 
     product? The CPI states that the price of admission to movies 
     increased 330 percent between 1948 and 1978. Yet the 
     invention of television allowed the price of two hours of 
     movie-like entertainment to decline substantially, even if we 
     cancel out the agony of commercials against the saving in 
     baby sitters, parking fees, and transportation expenses. A 
     long list of such broadly conceived substitutions could be 
     complied--permanent press clothing for commercial laundries, 
     phone for mail, appliance for domestic servants.


                             A better index

       The CPI is a severely flawed index, as shown both by our 
     comparison with the PCE deflator and our examination of the 
     pervasive nature of unmeasured quality change. Yet it is 
     striking that the BLS spent $50 million during 1972-77 to 
     revise the CPI without curing any of its major defects. In a 
     six-month overlap period in early 1978, the expensively 
     revised ``new CPI'' registered an increase that differed from 
     the ``old CPI'' by only 0.1 percentage point.
       It seems clear in retrospect that the BLS spent its 
     revision money on the wrong things, improving the number of 
     outlets covered or the number of consumers surveyed rather 
     than investing money in more rent data on single-family homes 
     or on performance data for newly introduced models and 
     products. What the CPI needs, in addition to the use of more 
     up-to-date weights and a rental equivalence approach to the 
     measurement of home ownership costs, is a vastly improved 
     effort to measure the improved performance and efficiency of 
     consumer goods and services, as well as the occasional 
     decline in product quality. Much can be done with existing 
     performance and efficiency data available from the published 
     test reports of Consumers Union and other organizations, and 
     in selective cases the BLS could 

[[Page S 15453]]
     institute its own testing program or contract for tests from private 
     organizations.
       It is now 20 years since a committee headed by George 
     Stigler recommended many of the same improvements in the CPI. 
     It is discouraging that so little has been done by so many 
     for so long. BLS officials tend to reject suggestions for a 
     more imaginative approach to quality measurement as too 
     ``subjective,'' when what is needed is a more frequent 
     application of simple common sense. In the now-classic words 
     of Martin Bronfenbrenner, addressed to the Stigler Committee 
     in 1960, ``it is better to be imprecisely right than 
     precisely wrong.'' And in an era in which each change in the 
     CPI sets off a wave of redistributional adjustments, that 
     observation is precisely right.
                                                                    ____


   Understanding the Consumer Price Index: Answers to Some Questions


                                preface

       The continually growing uses and users of the Consumer 
     Price Index (CPI) have generated an increasing number of 
     questions about the CPI. Although the Bureau of Labor 
     Statistics (BLS) has provided extensive material to the 
     public describing the CPI since its 1987 revision, much of 
     this material has been quite technical. BLS has developed 
     this pamphlet, therefore, to (1) answer frequently asked 
     questions about the CPI, (2) familiarize users of the CPI 
     with some of the most important of the new procedures 
     introduced with the 1987 CPI Revision, and (3) help users of 
     the CPI better understand and use it.
       Material in this publication is in the public domain and, 
     with the appropriate credit, may be reproduced without 
     permission.
       Information in this publication will be made available to 
     sensory impaired individuals upon request. Voice phone: (202) 
     606-STAT; TDD phone: (202) 606-5897; TDD Message Referral 
     phone: 1-800-326-2577.


                            what is the cpi?

       The Consumer Price Index (CPI) is a measure of the average 
     change over time in the prices paid by urban consumers for a 
     fixed market basket of consumer goods and services from A to 
     Z. The CPI provides a way for consumers to compare what the 
     market basket of goods and services costs this month with 
     what the same market basket cost a month or a year ago.


                          how is the cpi used?

       The Consumer Price Index affects nearly all Americans 
     because of the many ways it is used. Three major uses are:
       As an economic indicator: The CPI is the most widely used 
     measure of inflation and is sometimes viewed as an indicator 
     of the effectiveness of government economic policy. It 
     provides information about price changes in the Nation's 
     economy to government, business, labor, and other private 
     citizens and is used by them as a guide to making economic 
     decisions. In addition, the President, Congress, and the 
     Federal Reserve Board use trends in the CPI to aid in 
     formulating fiscal and monetary policies.
       As a deflator of other economic series: The CPI and its 
     components are used to adjust other economic series for price 
     changes and to translate these series into inflation-free 
     dollars. Examples of series adjusted by the CPI include 
     retail sales, hourly and weekly earnings, and components of 
     the national income and product accounts.
       An interesting example of this is the use of the CPI as a 
     deflator of the value of the consumer's dollar to find its 
     purchasing power. The purchasing power of the consumer's 
     dollar measures the change in the quantity of goods and 
     services a dollar will buy at different dates. In other 
     words, as prices increase, the purchasing power of the 
     consumer's dollar declines.
       As a means of adjusting dollar values: As inflation erodes 
     consumers' purchasing power, the CPI is often used to adjust 
     consumers' income payments, for example, Social Security; to 
     adjust income eligibility levels for government assistance; 
     and to automatically provide cost-off-living wage adjustments 
     to millions of American workers.
       The CPI affects the income of almost 70 million persons as 
     a result of statutory action: 43.1 million Social Security 
     beneficiaries, about 22.6 million food stamp recipients, and 
     about 3.9 million military and Federal Civil Service retirees 
     and survivors. Changes in the CPI also affect the cost of 
     lunches for 24.2 million children who eat lunch at school, 
     while collective bargaining agreements that tie wages to the 
     CPI cover about 2.8 million workers.
       Another example of how dollar values may be adjusted is the 
     use of the CPI to adjust the Federal income tax structure. 
     These adjustments prevent inflation-induced increases in tax 
     rates, an effect called ``bracket creep.''


                   is the cpi a cost-of-living index?

       No, although it frequently and mistakenly is called a cost-
     of-living index. The CPI is an index of price change only. It 
     does not reflect the changes in buying or consumption 
     patterns that consumers probably would make to adjust to 
     relative price changes. For example, if the price of beef 
     increases more rapidly than other meats, shoppers may 
     shift their purchases away from beef to pork, poultry, or 
     fish. If the charges for household energy increase more 
     rapidly than for other items, households may buy more 
     insulation and consume less fuel. The CPI does not reflect 
     this substitution among items as a cost-of-living index 
     would. Rather, the CPI assumes the purchase of the same 
     market basket, in the same fixed proportion (or weight) 
     month after month. About every 10 years the market basket 
     is thoroughly updated to allow for the introduction of new 
     products and services and to reflect more current spending 
     patterns. (See question 6.) In addition, the CPI does not 
     reflect taxes that are not directly associated with the 
     purchase of specific goods and services. In other words, 
     the CPI excludes taxes such as income and Social Security 
     taxes.
       It is important to note that local area CPI's cannot be 
     used to compare levels of living costs or prices between 
     areas. (See answer to question 17: ``Can the CPI's for 
     individual areas be used to compare living costs among the 
     areas?'')


               Whose buying habits does the CPI reflect?

       The CPI reflects spending patterns for each of two 
     population groups: All urban Consumers (CPI-U) and Urban Wage 
     Earners and Clerical Workers (CPI-W). The CPI-U represents 
     about 80 percent of the total U.S. population. It is based on 
     the expenditures reported by almost all urban residents, 
     including professional employees, the self-employed, the 
     poor, the unemployed, and retired persons as well as urban 
     wage earners and clerical workers. Not included in the index 
     are the spending patterns of persons living outside urban 
     areas, farm families, persons in the Armed Forces, and those 
     in institutions (such as prisons and mental hospitals).
       The CPI-W is based on the expenditures of urban households 
     that meet additional requirements: More than one-half of the 
     household's income must come from clerical or wage 
     occupations and at least one of the household's earners must 
     have been employed for at least 37 weeks during the previous 
     12 months. The CPI-W's population represents about 32 percent 
     of the total U.S. population and is a subset, or part, of the 
     CPI-U's populations.


         Does the CPI measure my experience with price change?

       Not necessarily. It is important to understand that BLS 
     bases the market baskets and pricing procedures for the CPI-U 
     and CPI-W on the experience of the relevant average 
     household, not on any specific family or individual. It is 
     unlikely that your experience will correspond precisely with 
     either the national indexes or those for specific cities or 
     regions.
       For example, if you or your family spend a larger than 
     average share of your budget on medical expenses, and medical 
     care costs are increasing more rapidly than other items in 
     the CPI market basket, your personal rate of inflation (or 
     experience with price change) may exceed the CPI. Conversely, 
     if you heat your home with solar energy, and fuel prices are 
     rising more rapidly than other items, you may experience less 
     inflation than the general population.
       This phenomenon explains why people sometimes question the 
     accuracy of the published indexes. A national average 
     reflects all the ups and downs of millions of individual 
     price experiences. It seldom mirrors a particular consumer's 
     experience.


                  How is the CPI market basket chosen?

       The CPI market basket is developed from detailed 
     expenditure information provided by families and 
     individuals on what they actually bought. For the current 
     CPI, this information was collected from the Consumer 
     Expenditure Survey over the 3 years 1982, 1983, and 1984. 
     In each of the 3 years, about 4,800 families, from around 
     the country, provided information on their spending habits 
     in a series of quarterly interviews. To collect 
     information on frequently purchased items, such as food 
     and personal care products, another 4,800 families in each 
     of the 3 years kept diaries listing everything they bought 
     during a 2-week period.
       Altogether, about 29,000 individuals and families provided 
     expenditure information for use in determining the 
     importance, or weight, of each item in the index structure.
       Due to time constraints, we used data from only the first 2 
     years of the Consumer Expenditure Survey to select the items 
     to be priced. In addition, we update the sample of stores and 
     service outlets in roughly 20 percent of the urban areas 
     priced for the CPI each year. New items are introduced with 
     these new samples.


              what goods and services does the cpi cover?

       The CPI represents all goods and services purchased for 
     consumption by urban households. We have classified all 
     expenditure items into over 200 categories, arranged into 7 
     major groups. Major groups and examples of categories in each 
     are as follows:
       Food and beverages (cookies, cereals, cheese, coffee, 
     chicken, beer and ale, restaurant meals); housing 
     (residential rent, homeowners' costs, fuel oil, soaps and 
     detergents, televisions, local telephone service); apparel 
     and its upkeep (men's shirts, women's dresses, jewelry); 
     transportation (airline fares, new and used cars, gasoline, 
     car insurance); medical care (prescription drugs, eye care, 
     physicians' services, hospital rooms); entertainment 
     (newspapers, toys, musical instruments, admissions); and 
     other goods and services (haircuts, college tuition, bank 
     fees).
       In addition, the CPI includes various user fees such as 
     water and sewerage charges, auto registration fees, vehicle 
     tolls, and so forth. Taxes that are directly associated with 
     the prices or specific goods and services (such as sales and 
     excise taxes) are also included. But, the CPI excludes taxes 
     not directly associated with the purchase of 

[[Page S 15454]]
     consumer goods and services (such as income and Social Security taxes).
       The CPI does not include investment items (such as stocks, 
     bonds, real estate, and life insurance). These items relate 
     to savings and not day-to-day living expenses.
       For each of the over 200 item categories, the Bureau has 
     chosen samples of several hundred specific items within 
     selected business establishments, using scientific 
     statistical procedures, to represent the thousands of 
     varieties available in the marketplace. For example, in a 
     given supermarket, the Bureau may choose a plastic bag of 
     golden delicious apples, U.S. extra fancy grade, weighing 4.4 
     pounds to represent the ``Apples'' category.


               how are cpi prices collected and reviewed?

       Each month, Bureau of Labor Statistics (BLS) field 
     representatives visit or call thousands of retail stores, 
     service establishments, rental units, and doctors' offices, 
     all over the United States to obtain price information on 
     thousands of items in the CPI market basket. For the entire 
     month they record the prices of about 90,000 items. These 
     90,000 prices represent a scientifically-selected sample of 
     the prices of goods and services sold to urban consumers 
     throughout the country.
       During each call or visit, the field representative 
     collects price data on a specific good or service that was 
     precisely defined during an earlier visit. If the selected 
     item is available, the field representative records its 
     price. If the selected item is no longer available or if 
     there have been changes in the quality or quantity (for 
     example, eggs sold in packages of 8 when previously they had 
     been sold by the dozen) of the good or service since the last 
     time prices had been collected, the field representative 
     selects a new item or records the quality change in the 
     current item.
       The recorded information is sent to the national office of 
     BLS where commodity specialists who have detailed knowledge 
     about the particular goods or services priced, review the 
     data. The specialists check the data for accuracy and 
     consistency and make any necessary corrections or 
     adjustments. These can range from an adjustment for a change 
     in the size or quantity of a packaged item to more complex 
     adjustments based upon statistical analysis of the value of 
     an item's features or quality. Thus, the commodity 
     specialists strive to keep changes in the quality of items 
     from affecting the CPI's measurement of price change.


                       how is the cpi calculated?

       The CPI is a product of a series of interrelated samples. 
     First, using data from the 1980 Census of Population, BLS 
     selects the urban areas from which prices are to be collected 
     and chooses the housing units within each area that are 
     eligible for use in the shelter component of the CPI. The 
     Census of Population also provides the data which allows the 
     assigning of the number of consumers represented by each area 
     priced for the CPI. Next, another sample of about 24,000 
     families serves as the basis for a Point-of-Purchase survey 
     that identifies the places where households purchase various 
     types of goods and services.
       Data from the Consumer Expenditure Survey conducted from 
     1982 through 1984, involving a national sample of almost 
     29,000 families, provided detailed information on their 
     spending habits. This enabled BLS to construct the CPI market 
     basket of goods and services and to assign each item in the 
     market basket a weight or importance based on total family 
     expenditures. The final stage in the sampling process is the 
     selection of the specific detailed item to be priced in each 
     outlet. This is done using a method called 
     ``disaggregation.'' For example, BLS field representatives 
     may be directed to price ``fresh whole milk.'' Through the 
     disaggregation process, the field representative selects the 
     specific kind of fresh whole milk that will be priced over 
     time in the outlet. By this process, each kind of whole milk 
     is assigned a probability, or weight, based on the quantity 
     of it the store sells. If, for example, Vitamin D, 
     homogenized milk in half-gallon containers makes up 70 
     percent of the sales of fresh whole milk, and the same milk 
     in quart containers accounts for 10 percent of all whole milk 
     sales, then the half-gallon container will be seven times 
     more likely to be chosen than the quart container. After 
     probabilities are assigned, one kind of milk is chosen by 
     an objective selection process based on the theory of 
     random sampling. The particular kind of milk that is 
     selected by disaggregation will continue to be priced each 
     month in the outlet.
       To sum up, the price movement measurement (see question 8) 
     is weighted by the importance of the item in the spending 
     patterns of the appropriate population group. The combination 
     of all these factors gives a weighted measurement of price 
     change for all the items in all the outlets, in all the areas 
     priced for the CPI.


                  how do i read or interpret an index?

       An index is a tool that simplifies the measurement of 
     movements in a numerical series. Most of the specific CPI 
     indexes have a 1982-84 reference base. That is, we set the 
     average index level (representing the average price level)--
     for the 36-month period covering the years 1982, 1983, and 
     1984--equal to 100. We measure changes in relation to that 
     figure. An index of 110, for example, means there has been a 
     10-percent increase in price since the base period; similarly 
     an index of 90 means a 10-percent decrease. Movements of the 
     index from one date to another can be expressed as changes in 
     index points (simply, the difference between index levels), 
     but it is more useful to express the movements as percent 
     changes. This is because index points are affected by the 
     level of the index in relation to its base period, while 
     percent changes are not.
       In the following table, item A increased by half as many 
     index points as item B. Yet, because of the different 
     starting figures, both had the same percent change; that is, 
     prices advanced at the same rate. On the other hand, items B 
     and C show the same change in index points, but the percent 
     change is much greater for item C because of its lower 
     starting value.
       We usually update reference base periods every 10 years or 
     so to make it easier for people to relate changes in the CPI 
     to other economic and cultural changes. We chose the 1982-84 
     period because it coincided with the time period of the CPI's 
     expenditure weights.

------------------------------------------------------------------------
                                             Item A    Item B    Item C 
------------------------------------------------------------------------
Year I....................................     112.5     225.0     110.0
Year II...................................     121.5     243.0     128.0
Change in index points....................       9.0      18.0      18.0
Percent change............................     (\1\)     (\2\)     (\3\)
------------------------------------------------------------------------
\1\ Item A: 9.0/112.5100=8.0                                            
\2\ Item B: 18.0/225.0100=8.0                                           
\3\ Item C: 18.0/110.0100=16.4                                          

               is the cpi the best measure of inflation?

       Inflation is the widespread and persistent increase in 
     costs and prices over the Nation's entire price and cost 
     structure, with expectations that the increase will continue 
     to occur in the future.
       Various techniques have been devised to measure different 
     aspects of inflation. The CPI measures inflation as 
     experienced by consumers in their day-to-day living expenses; 
     the Producer Price Index (PPI) captures it at earlier stages 
     of the production and marketing process; the Employment Cost 
     Index (ECI) measures it in the labor market; the BLS' 
     International Price Program measures it for imports and 
     exports; and the Gross Domestic Product Deflator (GDP-
     Deflator) measures combine the experience with inflation of 
     governments (Federal, State and local), businesses, and 
     consumers. Finally, there are more specialized measures, such 
     as measures of interest rates and measures of consumers' 
     and business executives' expectations.
       The ``best'' measure of inflation for a given application 
     depends on the intended use of the data. The CPI is generally 
     the best measure for adjusting payments to consumers when the 
     intent is to allow them to purchase, at today's prices, the 
     same market basket of consumer goods and services that they 
     could purchase in an earlier reference period. It is also the 
     best measure to use to translate retail sales and hourly or 
     weekly earnings into real or inflation-free dollars.


       which index is the ``official cpi'' reported in the media?

       Each month, BLS releases thousands of detailed CPI numbers 
     to the press. However the press generally focuses on the 
     broadest, most comprehensive CPI. This is known as ``the 
     Consumer Price Index for All Urban Consumers (CPI-U) for the 
     U.S. City Average for all Items, 1982-84 = 100.'' Often, the 
     media will report some or all of the following:
       a. the index level (for example, July 1992 = 140.5)
       b. the 12-month percent change (for example, July 1991 to 
     July 1992 = 3.2 percent).
       c. the 1-month percent change on a seasonally adjusted 
     basis (for example, from June 1992 to July 1992 = 0.1 
     percent).
       d. the annual rate of percent change so far this year (for 
     example, from December 1991 to July 1992 if the rate of 
     increase over the first 7 months of the year continued for 
     the full year, after the removal of seasonal influences, the 
     rise would be 2.9 percent).
       e. the annual rate based on the latest seasonally adjusted 
     1-month change. For example, if the June 1992 to July 1992 
     rate continued for a full 12 months, the rise, compounded, 
     would be 1.7 percent.


                what index should i use for escalation?

       The decision to employ an escalation mechanism, as well as 
     the choice of the most suitable index, is up to the user. 
     When drafting the terms of an escalation provision for use in 
     a contract to adjust future payments, both legal and 
     statistical questions can arise. While BLS cannot help in any 
     matters relating to legal questions, it does provide basic 
     technical and statistical assistance to users who are 
     developing indexing procedures.
       Some examples of technical or statistical guidelines from 
     BLS follow:
       BLS strongly recommends using indexes unadjusted for 
     seasonal variation (i.e., not seasonally adjusted indexes) 
     for escalation. (See answer to question 14 for a further 
     explanation of seasonally adjusted indexes and why we do not 
     recommend seasonally adjusted indexes for use in escalation.)
       BLS recommends using national or regional indexes for 
     escalation due to the volatility of the local indexes. (See 
     answer to question 15 for an explanation of this point).
       If you have further questions, the Bureau has prepared a 
     detailed report, Using the Consumer Price Index for 
     Escalation. For copies write or call the nearest BLS regional 
     office listed at the end of this report, or call (202)--606-
     7000.


              When should I use seasonally adjusted data?

       By using seasonally adjusted data, economic analysts and 
     the media find it easier 

[[Page S 15455]]
     to see the underlying trend in short-term price change. It is often 
     difficult to tell from raw (unadjusted) statistics whether 
     developments between any 2 months reflect changing economic 
     conditions or only normal seasonal patterns. Therefore, many 
     economic series, including the CPI, are seasonally adjusted 
     to remove the effect of seasonal influences on the changes, 
     thereby revealing the underlying trend. Seasonal influences 
     are those that normally occur at the same time and in about 
     the same magnitude every year. They include price movements 
     resulting from changing climatic conditions, production 
     cycles, model changeovers, and holidays. We re-estimate or 
     revise seasonally adjusted indexes annually.
       The unadjusted data reflect the actual prices consumers 
     pay. Therefore, unadjusted data are appropriate for 
     escalation purposes.


            What area indexes are published, and how often?

       Besides monthly publication of the national (or U.S. City 
     Average) CPI-U and CPI-W, monthly indexes are also published 
     for the four regions--Northeast, North Central, South, and 
     West. Monthly indexes are also published for urban areas 
     classified by population size--all metropolitan areas over 
     1.2 million, mid-sized metropolitan areas, small metropolitan 
     areas, and all nonmetropolitan urban areas. Indexes also are 
     available within each region cross-classified by area size. 
     For the Northeast and West, however, some of the population-
     size classes are not available. BLS also publishes indexes 
     for 29 local areas. These local area indexes are byproducts 
     of the national CPI program. Each local index has a much 
     smaller sample size than the national or regional indexes and 
     is, therefore, subject to substantially more sampling and 
     other measurement error. As a result, local area indexes are 
     more volatile than the national or regional indexes, even 
     though their long-term trends are similar. Therefore, BLS 
     strongly urges users to consider adopting the national 
     average (or regional) CPI's for use in their escalator 
     clauses. If used with caution, local area CPI data can be 
     used to illustrate and explain the impact of local economic 
     conditions on consumers' experience with price change. Local 
     area data are available on the following schedule:
       We publsh five major metropolitan areas monthly: Chicago-
     Gary-Lake County. IL-IN-WI; Los Angeles-Anaheim-Riverside, 
     CA; New York-Northern NJ-Long Island, NY-NJ-CT; Philadelphia-
     Wilmington-Trenton, PA-NJ-DE-MD; San Francisco-Oakland-San 
     Jose. CA.
       Data for an addition 10 metropolitan areas are published 
     every other month [on an odd (January, March, etc.) or even 
     (February, April, etc.) month schedule] for the following 
     areas:
       Baltimore, MD--odd.
       Houston, TX--even.
       Boston-Lawrence-Salem, MA-NH--odd.
       Miami-Fort Lauderdale, FL--odd.
       Cleveland-Akron-Lorain, OH--odd.
       Pittsburgh-Beaver Valley, PA--even.
       Dallas-Fort Worth, TX--even.
       St. Louis-East St. Louis, MO-IL--odd.
       Detroit-Ann Arbor, MI--even.
       Washington, DC-MD-VA--odd.
       (Note: The designation even or odd refers to the month 
     during which the area's price change is measured. Due to the 
     time needed for processing, data are released 2 to 3 weeks 
     into the following month.)
       Data are published for another group of 12 metropolitan 
     areas on a semiannual basis. These indexes, which refer to 
     the arithmetic average for the 6-month periods from January 
     through June and July through December, are published with 
     release of the CPI for July and January, respectively, in 
     August and February for: Anchorage, AK, Kansas City, MO-
     KS, Atlanta, GA, Milwaukee, WI, Buffalo-Niagara Falls, NY, 
     Minneapolis-St. Paul, MN-WI, Cincinnati-Hamilton, OH-KY-
     IN, Portland-Vancouver, OR-WA, Denver-Boulder, CO, San 
     Diego, CA, Honolulu, HI, Seattle-Tacoma, WA.
       Finally, BLS recently began publication of CPI's for two 
     metropolitan areas on an annual basis. These indexes 
     represent the arithmetic averages for the 12-month period 
     from January through December of each year. They are 
     published with the release of the CPI for January, i.e., in 
     February. These areas are: New Orleans, LA; Tampa-St. 
     Petersburg-Clearwater, FL.


 what area cpi should i use if there is no cpi for the area i live in?

       Although the BLS can provide some guidance on this 
     question, users must make the final decision.
       As noted in the answers to Questions 13 and 15, BLS 
     strongly urges the use of national or at least regional CPI's 
     for use in escalator clauses. These indexes are more stable 
     and subject to less sampling and other measurement error than 
     local area indexes. They are, therefore, more statistically 
     reliable.


  can the cpi's for individual areas be used to compare living costs 
                            among the areas?

       No, an individual area index measures how much prices have 
     changed in that particular area over a specific time period. 
     It does not show whether prices or living costs are higher or 
     lower in that area relative to another. In general, both the 
     market basket and relative prices of goods and services in 
     the base period vary substantially across areas.
       The following illustration shows that while Area B has 
     higher prices than Area A, the price change in Area A has 
     been greater than in Area B. The CPI measures the rates of 
     change in prices rather than the level of prices.

------------------------------------------------------------------------
                                         Base period     Current period 
                                     -----------------------------------
                                       Price    Index    Price    Index 
------------------------------------------------------------------------
Area A..............................    $0.30      100    $0.55      183
Area B..............................     0.60      100     0.90      150
------------------------------------------------------------------------

                   what types of data are published?

       These are many types of data published as outputs from the 
     CPI program. The most popular are indexes and percent 
     changes. Requested less often are relative importance data 
     (or relative expenditure weights), base conversion factors 
     (to convert from one CPI reference base to another), seasonal 
     factors (the monthly factors used to convert unadjusted 
     indexes into seasonally adjusted indexes), and average food 
     and energy prices. Index and price change data are available 
     for the U.S. City Average (or national average), for various 
     geographic areas (regions and metropolitan areas), for size 
     classes of urban areas, and for cross-classifications of 
     regions and size classes. Indexes for various groupings of 
     items are available for all geographic areas and size 
     classes.
       There are individual indexes available for over 200 items 
     (e.g., apples, men's shirts, airline fares), and over 120 
     different combinations of items (e.g., fruits and vegetables, 
     food at home, food and beverages, and All items), at the 
     national or U.S. City Average level. BLS classifies consumer 
     items into seven major groups: food and beverages, housing, 
     apparel and upkeep, transportation, medical care, 
     entertainment, and other goods and services. Some indexes 
     are available as far back as 1913.
       Each month, indexes are published along with short-term 
     percent changes, the latest 12-month change and, at the 
     national item and group level, unadjusted and (where 
     appropriate) seasonally adjusted percent changes (and 
     seasonal factors), together with annualized rates of change. 
     These annualized rates indicate what the rate of change would 
     be for a 12-month period, if a price change measured for a 
     shorter period continued for a full 12-months.
       The answer to question 15 provides information about the 
     areas and size classes for which indexes are published. For 
     areas, we publish less detailed groupings of items than we do 
     for the national level. The following table illustrates this 
     point:

                                ALL ITEMS                               
------------------------------------------------------------------------
               Baltimore, MD                      U.S. city average     
------------------------------------------------------------------------
Food and beverages........................  Food and beverages.         
Food......................................  Food.                       
Food at home..............................  Food at home.               
Cereals and bakery products...............  Cereals and bakery products.
                                            Cereals and cereal products.
                                            Flour and prepared flour    
                                             mixes.                     
                                            Cereal.                     
                                            Rice, pasta, and corn meal. 
                                            Bakery products.            
                                            White bread.                
                                            Fresh other bread, biscuits,
                                             rolls, and muffins.        
                                            Cookies, fresh cake and     
                                             cupcakes.                  
                                            Other bakery products.      
------------------------------------------------------------------------

       Annual average indexes and percent changes for these 
     groupings are published at the national and local levels.
       Semiannual average indexes and percent changes for some of 
     these groupings are also published.
       Each month, we publish average price data for some food 
     items items (for the U.S. and 4 regions) and for some energy 
     items (for the U.S., 4 regions, 4 size-classes, 13 cross-
     classifications of regions and size-classes, and for 15 
     metropolitan areas).


                What are some limitations of the index?

       The CPI is subject to both limitations in application and 
     limitations in measurement.

                       Limitations of application

       The CPI may not be applicable to all population groups. For 
     example, it is designed to measure the experience with 
     average price change of the U.S. urban population and, thus, 
     may not accurately reflect the experience of rural residents. 
     Also, the CPI does not provide data separately for the rate 
     of inflation experienced by subgroups of the population, such 
     as the elderly or the poor.
       As noted in the answer to question 17, the CPI cannot be 
     used to measure differences in price levels or living costs 
     between one place and another; it measures only time-to-time 
     changes in each place. A higher index for one area does not 
     necessarily mean that prices are higher there than in another 
     area with a lower index, it merely means that they have risen 
     faster since their common base period.
       The CPI cannot be used as a measure of total change in 
     living costs, because changes in these costs are affected by 
     such factors as changes in consumers' market baskets, social 
     and environmental changes, and changes in income taxes, which 
     the CPI does not include.

                       Limitations in measurement

       Limitations in measurement can be grouped into two basic 
     types, sampling errors and non-sampling errors.
       Sampling errors: Since the CPI measures price change based 
     on only a sample of items, the published indexes differ 
     somewhat from what the results would be if actual records of 
     all retail purchases by everyone in the index population 
     could be used to compile the index. These estimating or 
     sampling errors are limitations on the precise accuracy of 
     the index, not mistakes in index calculation. The accuracy 
     could be increased by using much larger samples, but the cost 


[[Page S 15456]]
     would be multiplied. Most of those who have examined the index have 
     found it to be sufficiently accurate for most of the 
     practical uses made of it. The CPI program has developed 
     measurements of sampling error.
       Nonsampling errors: These errors occur from a variety of 
     sources. Unlike sampling errors, they can cause persistent 
     bias in the index measurement. They are caused by problems of 
     price data collection, logistical lags in conducting surveys, 
     difficulties in defining basic concepts and their operational 
     implementation, and difficulties in handling the problems of 
     quality change. Nonsampling errors can be far more hazardous 
     to the accuracy of a price index than sampling error, per se. 
     BLS expands much effort to minimize these errors. Highly 
     trained personnel are relied on to insure comparability of 
     quality of items compared from period to period (see answer 
     to question 8.); collection procedures are extensively 
     documented and recurring audits are conducted. The CPI 
     program has started a program of continuous evaluation to 
     identify needed improvements and has introduced improvements 
     as their benefits were proven and as our budget permitted.


           will the cpi be updated or revised in the future?

       Yes. The CPI will need revisions as long as there are 
     significant changes in consumer buying habits or shifts in 
     population distribution or demographics. The Bureau, by 
     developing annual Consumer Expenditure Surveys and Point-of-
     Purchase Surveys, has the flexibility to monitor changing 
     buying habits in a timely and cost-efficient manner. In 
     addition, the censuses conducted by the Department of 
     Commerce provide information that permits us to adapt to 
     shifts in the population distribution and other demographic 
     factors at 10-year intervals.
       As a matter of policy, BLS is continually researching 
     improved statistical methods. Thus, even between major 
     revisions, we are making further improvements to the CPI. For 
     example, changes in children's day care and nursery school 
     expenses, until recently, had been represented by changes in 
     State and local minimum wages. The development of an adequate 
     sample of day care providers and nursery school reporters 
     enabled us to obtain prices for day care and nursery school 
     services directly.


                     how can i get cpi information?

       BLS furnishes CPI data to the public in a variety of 
     methods and formats.
       The Electronic News Release: This is the quickest. It is 
     reachable electronically immediately at release time (which 
     is approximately 2 weeks after the reference month) through 
     the BLS News Release Service. A fee is charged for this 
     service. Write to the Office of Publications and Special 
     Studies, Bureau of Labor Statistics, 2 Massachusetts Avenue, 
     NE, Washington, DC 20212-0001, or call (202) 606-5888.
       Telephone: A wide range of summary CPI data are provided on 
     a 24-hour recorded message, including key CPI numbers plus 
     the next release date. Call (202) 606-STAT. Another recorded 
     message, of less than 3 minutes, provides information about 
     the U.S. and Washington All Items CPI's and the next release 
     date. Call (202) 606-6994. Technical information is 
     available, between 8:15 and 4:45 Eastern time, Monday 
     through Friday, at (202) 606-7000. BLS Regional Offices 
     also provide CPI information by telephone.
       Mailgram: This arrives overnight. It is provided through 
     the National Technical Information Service, U.S. Department 
     of Commerce, 5285 Port Royal Road, Springfield, VA 22151. It 
     costs $190 per year in the contiguous United States. It 
     provides selected U.S. City Average CPI data.
       Machine-readable form: A single magnetic tape which 
     contains all current and historical CPI data is $95. Data 
     diskettes are also available. These offer CPI-U and CPI-W 
     indexes for the U.S. city average for 104 selected items, and 
     All items indexes for 54 selected areas, for all months of 
     the current year and the previous year. A single copy costs 
     $38 and a 12-month subscription $290. These arrive about a 
     week after the data are released. For information, write to 
     the Office of Publications and Special Studies, Bureau of 
     Labor Statistics, 2 Massachusetts Avenue, NE, Washington, DC 
     20212-0001 or call (202) 606-5886. Custom diskettes providing 
     data requested by the user are also available. Call (202) 
     606-6968.
       Free CPI Summary News Release: This 2-page release provides 
     CPI-U and CPI-W indexes, 1-month and 12-month percent changes 
     for 104 selected items for the U.S. city average, a brief 
     analysis of recent CPI movement, and All items indexes for 36 
     selected areas and groupings of areas for available periods 
     within the past 3 months, with their latest 12-month percent 
     change. It arrives about 3 weeks after the release of the 
     CPI. You can request that we add your name to this free 
     mailing list by writing to the Office of Publications and 
     Special Studies, Bureau of Labor Statistics, 2 Massachusetts 
     Avenue, NE Washington, DC 20212-0001 or by calling (202) 606-
     STAT. BLS Regional Offices (see end of this brochure) also 
     maintain free mailing lists for local and regional CPI 
     information.
       CPI Detailed Report: This is the Bureau's most 
     comprehensive report on consumer prices. It is published 
     monthly and costs $26 a year, $7 for a single copy. It can be 
     ordered from: New Orders, Superintendent of Documents, P.O. 
     Box 371954, Pittsburgh, PA 15250-7954. It includes text, 
     statistical tables, graphs, and technical notes. Besides 
     index data, the Detailed Report includes average prices for 
     some food and energy items. It arrives 3-4 weeks after the 
     release date.
       Monthly Labor Review: The MLR provides selected CPI data 
     included in a monthly summary of BLS data and occasional 
     analytical articles and methodological descriptions too 
     extensive for inclusion in the CPI Detailed Report. It can be 
     ordered from: New Orders, Superintendent of Documents, P.O. 
     Box 371954, Pittsburgh, PA 15250-7954. It costs $25 a year, 
     $7 for a single copy.
       Historical tables: These show all of the published indexes 
     for each of the detailed CPI components listed in the CPI 
     Detailed Report. They are available upon request. We impose 
     fees for large requests. For information call (202) 606-7000.
       Special publications: Various special publications are 
     available upon request. Examples of these are: Relative 
     Importance of Components in the Consumer Price Index, Using 
     the CPI for Escalation, fact sheets like ``Rebasing the 
     Consumer Price Index'' and associated conversion factors, and 
     assorted checklists which describe the items eligible for 
     pricing. For information call (202) 606-7000.

          Toward a More Accurate Measure of the Cost of Living

   (Interim report to the Senate Finance Committee from the Advisory 
   Commission To Study the Consumer Price Index, September 15, 1995)

                                               September 15, 1995.
     Hon. William V. Roth, Jr., Chairman,
     Hon. Daniel P. Moynihan, Ranking Minority Member,
     Committee on Finance, U.S. Senate, 211 Dirksen Senate Office 
         Building, Washington, DC.
       Dear Senators Roth and Moynihan: The Advisory Commission to 
     Study the Consumer Price Index herewith submits its Interim 
     Report in accordance with its charter based on Senate 
     Resolution 73, Section 11b.
           Sincerely,
     Michael J. Boskin,
       Chairman.
     Ellen Dulberger,
       Member.
     Zvi Griliches,
       Member.
     Robert J. Gordon,
       Member.
     Dale Jorgenson,
       Member.


                           executive summary

       1. The American economy is flexible and dynamic. New 
     products are being introduced all the time and existing ones 
     improved, while others leave the market. The relative prices 
     of different goods and services changes frequently, in 
     response to change in consumer tastes and income, and 
     technological and other factors affecting cost. This makes 
     constructing an accurate cost of living index more difficult 
     than in a static economy.
       2. Estimating a cost of living index requires assumptions, 
     methodology, data gathering and index number construction. 
     Biases can come from any of these areas.
       3. The strength of the CPI is in the underlying simplicity 
     of its concept: pricing a fixed (but representative) market 
     basket of goods and services over time. Its weakness follows 
     from the same conception: the ``fixed basket'' becomes less 
     and less representative over time as consumers respond to 
     price changes and new choices.
       4. There are five categories of potential bias in using 
     changes in the CPI as a measure of the change in the cost of 
     living. 1) Substitution bias occurs because a fixed market 
     basket fails to reflect the fact that consumers substitute 
     relatively less for more expensive goods when relative prices 
     change. 2) Outlet substitution bias occurs when shifts to 
     lower price outlets are not properly handled. 3) Quality 
     change bias occurs when improvements in the quality of 
     products, such as greater energy efficiency or less need for 
     repair, are measured inaccurately or not at all. 4) New 
     product bias occurs when new products are not included in the 
     market basket, or included only with a long lag. 5) Formula 
     bias occurs when the method of aggregating from the many 
     thousands of elementary products for which price quotations 
     are obtained to a modest number of groups of goods is 
     inappropriate. The report discusses and estimates the size of 
     each of the potential sources of bias.
       5. While the CPI is the best measure currently available, 
     it is not a true cost of living index (this has been 
     recognized by the Bureau of Labor Statistics for many years). 
     Despite important BLS updates and improvements in the CPI, 
     changes in the CPI have substantially overstated the actual 
     rate of price inflation, by about 1.5% per annum recently. It 
     is likely that a large bias also occurred looking back over 
     at least the last couple of decades, perhaps longer, but we 
     make no attempt to estimate its size.
       6. Changes in the CPI will overstate changes in the true 
     cost of living for the next few years. The Commission's 
     interim best estimate of the size of the upward bias looking 
     forward is 1.0% per year. The range of plausible values is 
     0.7% to 2.0%. The range of uncertainty is not symmetric. It 
     is more likely that changes in the CPI have a larger than a 
     smaller bias.
       7. The upward bias programs into the federal budget an 
     annual automatic real increase in indexed benefits and real 
     tax cut.
       8. CBO estimates that if the change in the CPI overstated 
     the change in the cost of living by an average of 1% per year 
     over the 

[[Page S 15457]]
     next decade, this bias would contribute almost $140 billion to the 
     deficit in 2005 and $634 billion to the national debt by 
     then. The bias alone would be the fourth largest federal 
     program, after social security, health care and defense.
       9. Some have suggested that different groups in the 
     population are likely to have faster or slower growth in 
     their cost of living than recorded by changes in the CPI. We 
     find no compelling evidence of this to date, in fact just the 
     opposite, but further exploration of this issue is desirable.
       10. In our final report we expect to have a more complete 
     analysis and evaluation together with specific 
     recommendations for procedures to improve and/or complement 
     the CPI.


                          i. introduction \1\

       Accurate measures of changes in the cost of living are 
     among the most useful and important data necessary to 
     evaluate economic performance. The change in the cost of 
     living between two periods, for example 1975 and 1995, tells 
     us how much income people would have needed in 1975, given 
     the prices of goods and services available in that year, to 
     be at least as well off as they are in 1995 given their 
     income and the prices of goods and services available then. 
     For example, if a family with a $45,000 income in 1995 would 
     have needed $15,000 in 1975, the cost of living has tripled 
     in the interim.
     Footnotes at end of article.
---------------------------------------------------------------------------
       If the American economy was quite static, with very few new 
     products introduced, very little quality improvement in 
     existing products, little change in consumers' tastes, and 
     very small and infrequent change in the relative prices of 
     goods and services, measuring changes in the cost of living 
     would be conceptually quite easy and its implementation a 
     matter of technical detail and appropriate execution. 
     Fortunately for the overwhelming majority of Americans, our 
     economy is far more dynamic and flexible than that. New 
     products are being introduced all the time and existing ones 
     improved, while others leave the market. The relative prices 
     of different goods and services change frequently, in 
     response to changes in consumer taste and income, and 
     technological and other factors affecting costs. Consumers in 
     America have the benefit of a vast and growing array of goods 
     and services from which to choose, unlike consumers in some 
     other countries.
       But because the economy is complex and dynamic is no reason 
     to bemoan the greater difficulty in constructing an accurate 
     cost of living index. Major improvements can and should be 
     made to the various official statistics that are currently 
     used as proxies for changes in the cost of living, such as 
     the well-known Consumer Price Index (CPI).
       The Consumer Price Index measures the cost of purchasing a 
     fixed market basket of goods and services. Based on surveys 
     of households from some base period, the index sets weights 
     (expenditure shares) for different goods and services. The 
     weights reflect average or representative shares for the 
     groups surveyed.\2\ Keeping these weights fixed through time, 
     the CPI is then calculated by attempting to measure changes 
     from one month to the next in prices of the same, or quite 
     closely related, goods and services.
       But through time consumption baskets change, in part 
     because of changes in the relative prices of goods and 
     services, and therefore the weights from the base period no 
     longer reflect what consumers are actually purchasing. This 
     failure to adjust for the changes in consumer behavior in 
     response to relative price changes is called substitution 
     bias. It is a necessary result of keeping the market basket 
     fixed. Because the market basket is updated only every decade 
     or so, as we get further away from the base period, there is 
     more opportunity for relative prices to diverge from what 
     they were in the base period, and for consumption baskets to 
     change substantially.
       Just as there are changes in what consumers purchase, there 
     are also trends and changes in where purchases are made. In 
     recent years, there has been a transformation of retailing. 
     Superstores, discount stores, and the like now comprise a 
     large and growing fraction of sales relative to a decade or 
     two ago. As important as keeping up with the basket of goods 
     that consumers actually purchase is keeping up with the 
     outlets where they actually purchase them, so that the prices 
     paid are accurately recorded. The current methodology suffers 
     from an outlet substitution bias, which insufficiently takes 
     into account the shift to discount outlets.
       Many of the products sold today are dramatic improvements 
     over their counterparts from years ago. They may be more 
     durable and subject to less need for repair, more energy 
     efficient; lighter; safer; etc. Sometimes, at least 
     initially, a better quality product replacing its counterpart 
     may cost more. Separating out how much of the price increase 
     is due to quality change rather than actual inflation in the 
     price of a standardized product is far from simple, but is 
     necessary to obtain an accurate measure of the true increase 
     in the cost of living. To the extent quality change is 
     measured inaccurately or not at all, there is a quality 
     change bias in the CPI.
       The same is true with the introduction of new products, 
     which have substantial value in and of themselves--not many 
     of us would like to surrender our microwave ovens, radial 
     tires, and VCR's--as well as the value of greater choice and 
     opportunities opened up by the new products. To the extent 
     new products are not included in the market basket, or 
     included only with a long lag, there is a new product bias in 
     the CPI.
       Finally, in a dynamic, complex economy like the 
     contemporary United States, there are literally many 
     thousands of goods and services consumed. Price data are 
     collected at a considerable level of disaggregation and how 
     the price changes are aggregated into an overall index 
     involves quite technical issues that can lead to a formula 
     bias in the CPI.
       Even if not federal program on either the outlay or revenue 
     side of the budget was indexed, it would still be desirable 
     to improve the quality of measures of the cost of living from 
     the standpoint of providing citizens a better and more 
     accurate estimate of what was actually going on in the 
     economy, a way to compare current performance to our 
     historical performance or to that of other countries. For 
     example, the most commonly used measure of the standard of 
     living is real income or output per person. To measure 
     changes in real income requires the separation of nominal 
     income changes from price changes. Obviously, that requires 
     an accurate measure of price changes.
       But numerous federal, state and local government programs 
     and tax features are ``indexed'' for changes in the cost of 
     living by the changes in the Consumer Price Index. The CPI is 
     also used to index a large number of private sector 
     contracts, including wages in collective bargaining 
     agreements and rents, to name obvious examples that affect 
     millions of Americans. Currently, slightly under one-third of 
     total federal outlays, mostly in retirement programs, are 
     directly indexed to changes in consumer prices. Several 
     features of the individual income tax, including the tax 
     brackets, are indexed; the individual income tax accounts for 
     a little under half of federal revenues.
       Congress indexed these outlay programs and tax rules in 
     order to help insulate or protect the affected individuals 
     from bearing the brunt of increases in the cost of living. 
     Yet the Bureau of Labor Statistics, the agency responsible 
     for compiling and presenting the Consumer Price Index, has 
     explicitly stated for years that the CPI is not a cost of 
     living index, presumably for some of the reasons mentioned 
     above. If the Consumer Price Index as currently produced, and 
     as likely to be produced over the next few years, is not an 
     appropriate cost of living index for the task Congress had in 
     mind, then it is desirable to consider alternative measures.
       The consequences of changes in the Consumer Price Index 
     overstating changes in the cost of living can be dramatic. 
     For example, if use of the CPI is expected to overstate the 
     increase in the cost of living by one percentage point per 
     year over the next seven years, the national debt would be 
     almost $300 billion greater in 2002 than if a corresponding 
     correction were made in the indexing of outlays and revenues.
       This interim report proceeds as follows: Section II 
     discusses the historical and prospective budgetary 
     implications of changes in the CPI overstating changes in the 
     cost of living. Section III details why the CPI is not a true 
     cost of living. Section III details why the CPI is not a true 
     cost of living index and discusses several sources of bias. 
     Section IV describes in greater detail the bias from quality 
     change and new products. Section V introduces the issue of 
     separate price indexes for different groups. The Conclusion 
     summarizes the interim findings of the Commission.


                    II. Indexing the Federal Budget

       The issue proposed for fiscal policy makers by an upward 
     bias in the CPI has been stated with admirable clarity by the 
     Congressional Budget Office (1994): The budgetary effect of 
     any overestimate of changes in the cost of living highlights 
     the possibility of a shift in the distribution of wealth. If 
     the CPI has an upward bias, some federal programs would 
     overcompensate for the effect of price changes on living 
     standards, and wealth would be transferred from younger and 
     future generations to current recipients of indexed federal 
     programs--an effect that legislators may not have 
     intended.\3\
       Social Security is by far the most important of the federal 
     outlays that are indexed to the CPI. However, Supplemental 
     Security Income, Military Retirement, and Civil Service 
     Retirement are significant programs that are similarly 
     indexed. Other federal retirement programs, Railroad 
     Retirement, veterans' compensation and pensions, and the 
     Federal Employees' Compensation Act also contain provisions 
     for indexing. The Economic Recovery Tax Act of 1981 indexed 
     individual income tax brackets and the personal exemption to 
     the CPI.
       How important have the budgetary consequences of upward 
     bias in the CPI been historically? Obviously, a precise 
     answer to this question would require extended study, taking 
     into account the timing of the bias, the parallel development 
     of indexing provisions in specific federal outlays and 
     revenues, and interest on the accumulation of debt that has 
     resulted. An indication of the potential size of these 
     effects can be inferred from one important historical example 
     of one clearly identified source of bias. A careful study of 
     this type, which focuses on the most important federal 
     program affected by indexing, namely, social security 
     benefits, has been conducted by the Office of Economic Policy 
     (OEP) of the Department of the Treasury.
       On February 25, 1983, the Bureau of Labor Statistics (BLS) 
     introduced an important technical modification in the 
     Consumer Price Index for All Urban Consumers (CPI-

[[Page S 15458]]
     U). This altered the treatment of housing costs by shifting the costs 
     for homeowners to a rental equivalent basis. The new 
     treatment of housing costs was incorporated into the Consumer 
     Price Index for Urban Wage Earners and Clerical Workers (CPI-
     W), used to index social security benefits, in 1985.
       The rental equivalent measure of housing costs was a 
     conceptual improvement and has been retained in subsequent 
     official publications. However, housing costs in preceding 
     years employed a ``homeownership'' measure ``. . . based on 
     house prices, mortgage interest rates, property taxes and 
     insurance, and maintenance costs.'' \4\ The treatment of 
     housing costs prior to 1983 was not modified in publishing 
     the revised CPI-U, so that the new treatment of housing 
     introduced a discrepancy in the conceptual basis for the CPI-
     U before and after 1983. Similarly, housing cots in the CPI-W 
     prior to 1985 have not been modified.
       BLS developed an ``experimental'' price index, CPI-U X1, 
     based on a rental equivalent treatment of housing extending 
     back to 1967. This provides the basis for the OEP assessment 
     of bias in the CPI-W. The bias for 1975, the first year that 
     social security was indexed to the CPI-W, was 1.1 percent. 
     This bias mounted over subsequent years, reaching 6.5 percent 
     by 1982 and then declining to 4.7 percent in 1984.\5\
       Overpayments of social security benefits resulting from the 
     bias in the CPI-W mounted through 1983, reaching a total of 
     $7.1 billion or 7.1 percent of benefits paid in that year. 
     These overpayments have resulted in a lower balance in the 
     OASI trust fund and a larger federal deficit and debt. OEP 
     estimates interest costs associated with these deficits at 
     the rate of interest paid or projected to be paid on the OASI 
     trust fund. Beginning in 1985 interest costs predominate in 
     the total. In the current fiscal year the total cost is $16.7 
     billion, of which $12.6 billion is interest. The cumulative 
     effect of just this one source of bias in the CPI-W via this 
     one program on the federal debt amounts to $213.2 billion, as 
     of 1995.
       In summary, the BLS made two decisions in revising the 
     treatment of housing costs in the CPI-W in 1985. The first 
     decision was to change the treatment of housing costs to a 
     rental equivalent basis beginning in January 1985. The second 
     was not to revise the treatment of housing costs for 1984 and 
     earlier years. As a consequence of these two decisions the 
     level of the CPI-W is 4.7 percent above the CPI-U X1, a 
     measure of the cost of living based on the same primary data 
     sources and similar methodology, but with a consistent 
     treatment of housing costs.
       The increases in federal outlays resulting from the bias in 
     the CPI-W cannot be justified as cost of living adjustments. 
     These increases are the consequence of an inappropriate 
     treatment of housing costs before 1985 and have resulted in 
     large transfers to beneficiaries of the OASI program that are 
     devoid of any economic rationale. The overpayments have 
     continued up to the present, but are declining in importance. 
     However, the resulting decline in the OASI trust fund 
     continues to mount due to rising interest costs and now 
     contributes more than two hundred billion dollars to the 
     federal debt!
       Of course, nobody would suggest retroactively undoing the 
     overindexing due to this or any other source of bias. The 
     point of this discussion is to demonstrate how important it 
     is to correct biases in the CPI (in either direction) as 
     quickly and fully as possible before their consequences 
     mount, indeed compound.
       What would be the effect of an upward bias in the CPI on 
     future deficits? More than half of federal spending of $1.5 
     trillion is now attributable to entitlements and mandatory 
     spending programs. In January 1995 the annual Congressional 
     Budget Office (CBO) outlook for the economy and the federal 
     budget showed that this proportion is projected to rise to 
     almost two-thirds of federal spending during fiscal year 
     1998. Cost-of-living adjustments at a projected rate of 3.0 
     percent will contribute $43 billion to total spending on 
     mandatory programs in that year and $80 billion in fiscal 
     year 2000.\6\ This is 6.8 percent of projected spending on 
     mandatory programs in fiscal year 2000.
       Testimony presented by the CBO to the Committee on Finance 
     shows the impact of a hypothetical correction (reduction) of 
     0.5 percent in cost of living adjustments for fiscal years 
     1996-2000.\7\ Federal outlays would decline by $13.3 billion 
     in fiscal year 2000, while revenues would rise by $9.6 
     billion. The decline in debt service resulting from reduced 
     deficits in fiscal years 1996-2000 would be $3.3 billion, 
     yielding a total contribution to deficit reduction of $26.2 
     billion in fiscal year 2000.\8\ This is more than ten percent 
     of the deficit projected by CBO in that year.
       The CBO has provided the Commission with projections of the 
     impact of hypothetical corrections (reductions) of 0.5 and 
     1.0 percent in cost of living adjustments for fiscal years 
     1996-2005. With a reduction of 0.5 percent the total 
     contribution to deficit reduction rises to $71.9 billion in 
     2005. Of this amount, an increase in revenue accounts for 
     $21.9 billion and reductions in outlays, including debt 
     service, amounts to $32.7 billion (of which debt service is 
     $17.3 billion). The total reduction is almost seventeen 
     percent of the projected deficit in 2005. The cumulative 
     reduction in debt held by the public in 2005 is $319.6 
     billion or about 2.7 percent of the GDP projected for that 
     year.
       CBO projections for the impact of a hypothetical correction 
     (reduction) in cost of living adjustments of 1.0 percent are, 
     of course, even more dramatic. The total change in the 
     deficit in the year 2005 is $139.1 billion. Federal revenues 
     would be increased by $40.8 billion and federal outlays 
     reduced by $98.3 billion, of the reduction in outlays $34.4 
     billion can be attributed to lower debt service and $63.9 
     billion to lower outlays on indexed programs. (See Appendix 
     Figure A-1 for detail not reproducible in Record). The 
     cumulative reduction in outstanding federal debt by 2005 is 
     $634.3 billion. (See Appendix Figure A-2 for detail not 
     reproducible in Record). This is almost 9.4 percent of the 
     debt projected for that year and almost 5.5 percent of the 
     GDP!
       Stated differently, if the change in the CPI overstated the 
     change in the cost of living by an average of 1% per year 
     over this period, this bias alone would contribute almost 
     $140 billion to the deficit in the year 2005. That is one-
     third the projected baseline deficit (which assumes no policy 
     changes such as the current balanced budget proposals). More 
     remarkably, the upward bias by itself would constitute the 
     fourth largest federal outlay program, behind only social 
     security, health care and defense!
       In summary, an upward bias in the CPI would result in 
     substantial overpayments to the beneficiaries of federal 
     entitlements and mandatory spending programs. In addition, 
     such a bias would reduce federal revenues by overindexing the 
     individual income tax. In short, the upward bias programs 
     into the federal budget every year an automatic, real 
     increase in indexed benefits and a real tax cut. Correction 
     of biases in the CPI, while designed to more accurately 
     adjust benefits and taxes for true changes in the cost of 
     living, would also contribute importantly to reductions in 
     future federal budget deficits and the national debt. These 
     reductions can be attributed to higher revenues, lower 
     outlays, and less debt service. Lower outlays-cuts in indexed 
     federal spending programs and reduced interest payments-
     account for over two-thirds of the long-run deficit 
     reduction, while higher revenues account for the rest.


 iii. the consumer price index and a cost of living index: measurement 
                                 issues

       A cost of living index is a comparison of the minimum 
     expenditure required to achieve the same level of well-being 
     (also known as welfare, utility, standard-of-living) across 
     two different sets of prices. Most often it is thought of as 
     a comparison between two points of time. As with any 
     practical application of theory to index number production, 
     estimating a cost of living index requires assumptions, a 
     methodology, data gathering processes and index number 
     construction.
       There are two sets of potential biases in the CPI: biases 
     relative to an ``ideal'' cost of living index and biases 
     which arise within its own terms of reference. The strength 
     of the CPI is in the underlying simplicity of its concept: 
     pricing a fixed (but representative) market basket of goods 
     and services over time. Its weakness follows from the same 
     conception: the ``fixed basket'' becomes less and less 
     representative over time as consumers respond to price 
     changes and new choices.
       Consumers respond to price changes by substituting away 
     from products that have become more expensive and toward 
     goods whose prices have declined relatively. As the world 
     changes, they are faced with new choices in shopping outlets, 
     varieties, and entirely new goods and services, and respond 
     to these as well. These changes make the previous ``fixed 
     basket'' increasingly irrelevant.
       In trying to keep true to its concept in a rapidly changing 
     world, the current CPI procedures encounter difficulties. 
     Biases result when they ignore some of these changes such as 
     the appearance of discounters, and also when they try to do 
     something about them such as when items are rotated out of 
     the sample and replaced with new items. Attempting to capture 
     the changes in a way that tries to mimic the pricing of a 
     ``fixed basket'' within a rather patchwork framework just 
     cannot be done without introducing other problems into the 
     resulting index. These different biases overlap and have been 
     discussed under a number of headings: substitution bias; 
     formula bias; outlet substitution bias; quality change; and 
     new product bias.
       The ``pure'' substitution bias is the easiest to 
     illustrate. Consider a very stylized example, where we would 
     like to compare an initial ``base'' period 1 and a subsequent 
     period 2. For simplicity, consider a hypothetical situation 
     where there are only two commodities: beef and chicken. In 
     period 1, the prices per pound of beef and chicken are equal, 
     at $1, and so are the quantities consumed, at 1 lb. Total 
     expenditure is therefore $2. In period 2, beef is twice as 
     expensive as chicken ($1.60 vs. $0.80 per pound), and much 
     more chicken (2 lb.) than beef (0.8 lb.) is consumed, as the 
     consumer substitutes the relatively less expensive chicken 
     for beef. Total expenditure in period 2 is $2.88. The 
     relevant data are presented in Table 1. How can we compare 
     the two situations? Actually, there are several methods, each 
     asking slightly different questions and therefore, not 
     surprisingly, giving different answers.\9\

                                                                        

[[Page S 15459]]
                                                                       TABLE 1.--HYPOTHETICAL EXAMPLE OF SUBSTITUTION BIAS                                                                      
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                                                                       Price relatives                  Relative weights        
                                                          Price in period    Quantity in    Price in period    Quantity in   -------------------------------------------------------------------
                                                                 1             period 1            2             period 2          P2/P1            P1/P2              1                2       
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Beef....................................................               1                1              1.6              0.8              1.6             0.63              0.5             0.43 
Chicken.................................................               1                1              0.8              2.0              0.8             1.25              0.5             0.57 
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------



       The simplest comparison is to ask ``How much more must I 
     spend in my current situation (period 2) to purchase the same 
     quantities that I purchased initially (in period 1)?'' \10\ 
     This is the question asked by the CPI. The price index for 
     period 2 relative to period 1 uses the initial period 1 
     basket of consumption as the weights in the computation. To 
     buy 1 lb. of beef and 1 lb. of chicken in period 2 costs 
     $2.40. The price index for period 2 relative to period 1 is 
     1.20 (2.40/2.00), that is a 20% increase.
       Intuitively, it is easy to understand why such a 
     computation imparts an upward (substitution) bias to the 
     measure of the change in the true cost of living. It assumes 
     the consumer does not substitute (cheaper) chicken for beef. 
     In the real world, as in the hypothetical example, consumers 
     change their spending patterns in response to changes in 
     relative prices and, hence, partially insulate themselves 
     from price movements.
       An alternative approach would be to ask the question ``How 
     much more am I spending in my current situation (period 2) 
     than I would have spent for the same goods and services at 
     the prices that prevailed initially (in period 1)?'' \11\ 
     This price index compares expenditures in period 2 ($2.88) 
     with what it would cost to buy the current (period 2) market 
     basket at the initial prices ($0.80 for the beef plus $2.00 
     for the chicken equals $2.80). This price index is 1.03, that 
     is only a 3% increase. This approach understates the rise in 
     the true cost of living as it overstates substitution.
       The idea of a cost of living index is not to keep the 
     consumption basket fixed, but to allow for the substitution 
     that follows relative price changes. The question answered by 
     a true cost of living index is instead ``How much would we 
     need to increase (or decrease) the initial (period 1) 
     expenditure in order to keep the consumer just as well off in 
     period 2?'' Such a question cannot be answered without 
     knowing the consumer's preferences in more detail, but a very 
     good approximation may be obtained by interpolating between 
     the two answers (that arise from the different base periods). 
     There are alternative ways of doing so, each involving a 
     different mathematical formula. A commonly accepted approach 
     is to use the geometric mean (the square root of the product) 
     of the two answers.\12\ In our example, this comes to 1.11, 
     an 11% increase. By comparison, the CPI-type fixed base index 
     contains an upward bias of 0.09 (1.20-1.11); thus, almost 
     half of the increase in the CPI-type calculation is 
     substitution bias.
       How large are such substitution biases in the real world? 
     That depends on how out of date the base period weights used 
     in constructing the index are and on how much relative prices 
     have changed in a consistent and permanent direction. If 
     relative prices diverge over time and do not just fluctuate, 
     there is a permanent bias in the standard fixed base formula. 
     Since we have been experiencing various consistent price 
     trends, the further one gets away from the base period (for 
     which the weights are approximately correct), the larger the 
     bias.
       Most of the computations done for large groupings of 
     commodities (relatively aggregated commodity levels) show 
     small biases in the growth rates of the CPI, rising from 
     about 0.15 percent per year in the first five years after new 
     expenditure weights are introduced, to about 0.30 percent per 
     year in the subsequent five years. These estimates are based 
     on research covering the period 1982-91 and updated to 
     1993.13 14 The bias increases as average consumption 
     patterns drift further away from what they were in the base 
     period. Therefore, this bias may be expected to increase 
     further in the next few years, perhaps to 0.40 percent per 
     year, until the newly revised CPI is released in 1998. At 
     that point, the weights will be shifted to reflect average 
     consumer expenditures in 1993-5, (and will already be four 
     years out of date!). Although the substitution bias will 
     then decline for awhile, it will grow subsequently as 
     prices and consumptions patterns drift away again from 
     those in the new base period unless the BLS changes its 
     procedures and moves toward some different index number 
     formula with shifting weights.
       These estimates may be low. They are based on computations 
     using rather high level groupings (200 commodity subindexes) 
     of the many underlying varieties and models of specific 
     products and services and may miss some of the large 
     substitutions that occur at the more detailed level. Indeed, 
     one may interpret as additional evidence on this point, the 
     results of a simulation experiment by BLS researchers which 
     applied different index number formulae at the item, or 
     ``elemental,'' level, for price changes in 1991-2 and yielded 
     an estimate of the bias equal to 0.50 percent.\15\
       Recognizing the continuously changing assortment of 
     commodities in the market, the BLS improved its price 
     measurement procedures in 1978. The improved procedure 
     chooses items to be priced based on a probability sample and 
     rotates these items on a staggered, five year cycle. The idea 
     was laudable, but embedding it in a conceptually ``fixed-
     weight,'' ``fixed-basket'' index created unanticipated 
     problems which have become known as ``formula'' bias.
       In essence, the problem arises as the procedure exaggerates 
     (gives too much importance to) the effect of short run 
     variability of prices (such as items on sale). This bias was 
     discovered and evaluated by BLS researchers and appears to be 
     most important in seasonal items such as fruit and 
     vegetables, but has apparently also affected the residential 
     housing component of the index.\16\ \17\ The overall bias 
     from this source has been estimated to be on the order of 
     0.50 percent per year. However, now that this formula bias is 
     understood, procedures are being developed which will largely 
     eliminate it when implemented.
       While the formula bias in the CPI can, should and hopefully 
     will be eliminated in the future, the problems of outlet and 
     variety substitution are unlikely to diminish soon. Just as 
     consumers change the goods they purchase in response to 
     changes in relative prices as in the beef and chicken 
     example, so do they change the location of where they make 
     their purchases. The opening of a new discount store outlet 
     may give consumers the opportunity to purchase a given good 
     at a lower price than before. At present, the CPI procedures 
     ignore such reductions that occur when consumers change 
     outlets. However, if consumers cared only about obtaining 
     goods at the lowest price, then we would observe all goods 
     sold at the same price at all outlets. Instead, we observe 
     low prices at discount stores and warehouse clubs at the same 
     time as medium prices at supermarkets and higher prices at 
     convenience stores. Evidently, consumers care not only about 
     prices, but the level of services such as availability of 
     clerks, wrapping services, and the distance between home and 
     alternative outlets.
       Current procedures in the CPI ignore price changes when 
     consumers switch outlets. This incorporates into the CPI the 
     implicit assumption that price differentials among outlets 
     entirely reflect the differences in service quality. This 
     approach would be legitimate if the economy stood still with 
     a stable set of outlets providing alternative levels of 
     service quality. However, there has been a continuous 
     increase in the market share of discount stores as more 
     efficient technologies of distribution allow low price 
     outlets to expand while older, higher priced outlets have 
     contracted and in some cases gone out of business. This 
     shift in market share indicated that many consumers 
     respond to price differentials and do not consider them to 
     be fully offset by difference in service quality. 
     Completely ignoring all differences in service quality by 
     incorporating all such price reductions into the CPI would 
     err in the opposite direction. Further research is 
     required to disentangle true changes in prices from 
     changes in service quality. This problem is analogous to 
     the need to disentangle the changes in prices from changes 
     in product quality.
       Quality change and new goods present the most difficult 
     problems for measurement. They include capturing the 
     introduction of new products in a timely manner; making 
     direct quality comparisons of new products with existing 
     ones; making direct quality comparisons of new products with 
     other products against which they compete (in other 
     classification groupings such as a new drug and the surgical 
     treatment it replaces); and capturing the combined impact of 
     quality and substitution as these new products displace 
     others within and across their classification grouping.
       A full treatment of these issues reinforces the problem of 
     focusing on the ``average'' or ``representative'' consumer. 
     Different consumers have different tastes and time costs, and 
     hence value the appearance of new outlets and new products 
     differentially, with some (the majority) becoming better off 
     with supermarkets and others losing out as the corner grocery 
     store disappears. The CPI is not equipped to account for 
     special characteristics of different consumers or groups of 
     consumers.\18\ \19\ The following sections explore some of 
     these problems.
       There are still other issues that would in principle apply 
     to obtaining a true cost of living index (COLI). Consider two 
     examples: the negative effects of higher crime rates and the 
     concommitant purchases of security devices and higher 
     insurance premiums and the positive effects of improvements 
     in information technology that permit a parent to work at 
     home when a child is ill. Surely these would enter a 
     calculation of ``the minimum expenditure necessary to be at 
     least as well off.'' The Commission notes these 
     considerations but is not prepared to quantify them at this 
     time.


                           IV. QUALITY CHANGE

       The difficulty created by quality change in existing 
     products, and by the introduction of 

[[Page S 15460]]
     new products, is highlighted by returning to the definition of a cost 
     of living index--a comparison between two time periods of the 
     minimum expenditure required to achieve the same level of 
     well-being. What does the ``same level'' mean when entirely 
     new products are introduced that were unavailable in the 
     first time period?
       A pervasive phenomenon called the ``product cycle'' is 
     critical in assessing the issue of new product bias in the 
     CPI and applies as well to new models of existing products. A 
     typical new product is introduced at a relatively high price 
     with sales at a low volume. Soon improvements in 
     manufacturing techniques and increasing sales allow prices to 
     be reduced and quality to be improved. For instance, the VCR 
     was introduced in the late 1970s at a price of $1,000 and 
     with clumsy electromechanical controls; by the mid 1980s the 
     price had fallen to $200 and the controls were electronic, 
     with extensive preprogramming capabilities. Later on in the 
     product cycle, the product will mature and eventually will 
     increase in price more rapidly than the average product of 
     its class. The sequence is easily visualized as a ``U''-
     shaped curve--the price of any given product relative to 
     the consumer market basket starts high, then goes down, is 
     flat for a while, and then goes back up. To the extent 
     that the CPI over-weights mature products and underweights 
     new products, it will tend to have an upward bias.
       Our discussion of quality change and new product bias 
     begins with a review of the methods used by the CPI to handle 
     quality changes in existing products and then turns to 
     problems posed by new products. The BLS has four different 
     methods to cope with a model change for an existing product.
       The ``direct comparison'' method treats all of the observed 
     price change between the old model and the new model as a 
     change in price and none as a change in quality. There is no 
     necessary bias, because quality can decrease as well as 
     increase. But in practice goods tend to undergo steady 
     improvement, and often a better model is introduced with no 
     change in price, causing the quality change to be missed 
     entirely.
       The ``deletion'' method makes no comparison at all between 
     the prices of the old and new model. Instead, the weight 
     attributable to this product is applied to the average price 
     change of other products in the same commodity 
     classification. To the extent that the deletion method is 
     used, the CPI consists disproportionately of commodities of 
     constant quality which may be further along in the product 
     cycle.
       The ``linking'' method can be used if the new and old model 
     are sold simultaneously. In this case the price differential 
     between the two models at the time of introduction of the new 
     model can be used as an estimate of the value of the quality 
     differential between the two models. Unfortunately, new 
     models usually replace old models entirely, and the link 
     prices are not observed. Also, a quality improvement in the 
     new model can occur even if it costs less or the same as the 
     old model, as in the case of the VCR where the price fell 
     continuously while programming capability and reproduction 
     quality improved.
       The ``cost estimation'' method attempts to establish the 
     cost of the extra attributes of the new model. Problems in 
     practice with the costing method have been its infrequency of 
     use, and the fact that it has been applied disproportionately 
     in the case of automobiles relative to other products. This 
     raises the possibility that there is a spurious upward 
     ``drift'' in the price of other products relative to 
     automobiles due to an uneven application of the costing 
     method.
       This list of method reveals at least two potential sources 
     of upward bias, the use of the direct comparison method that 
     does not address the quality issue at all, and the use of the 
     deletion method that bases price change on models that are 
     unchanged in quality and may be further along in the product 
     cycle. A greater difficulty is that the CPI makes no attempt 
     to create systematic estimates of the value of quality 
     improvements which increase consumer welfare without raising 
     the price of products. For instance, many consumer electronic 
     products and household appliances have experienced a 
     reduction in the incidence of repairs and in electricity use, 
     and few if any of these improvements have been taken into 
     account by the CPI.
       The CPI uses only rarely an alternative methodology called 
     the ``hedonic regression method'' for estimating the value of 
     quality change. The hedonic approach can be viewed as an 
     alternative method to manufacturers' cost estimates in making 
     quality change adjustments. It assumes that the price of a 
     product observed at a given time is a function of its quality 
     characteristics, and it estimates the imputed prices of such 
     characteristics by regressing the prices of different models 
     of the product on their differing embodied quantities of 
     characteristics. Thus the hedonic approach is less a new 
     method than an alternative to cost estimates to be used 
     when practical factors make it more suitable than the 
     conventional method.
       By their very nature hedonic indexes require large amounts 
     of data. Given the thousands of separate products that are 
     produced in any modern industrial society, the need to 
     collect a full cross-section of data on each product presents 
     an insurmountable obstacle to the full-blown adoption of the 
     hedonic technique. Further, it is impossible to construct a 
     hedonic index in the timely fashion required by the CPI, with 
     its orientation to producing within a few weeks an estimate 
     of month-to-month price changes that can never be revised. 
     Accordingly, most hedonic studies have been retrospective and 
     can be used to gauge the accuracy of individual components of 
     the CPI rather than being used in the actual month-to-month 
     construction of the CPI. This is one important reason to 
     consider broadening the concept of the CPI to include both 
     the current index dedicated to timely measures of month-to-
     month price changes, and a second supplementary index 
     produced with a greater time lag, and subject to periodic 
     revision, dedicated to accurate measurement of price changes 
     over years and decades.
       We turn now to the issue of new product bias. There is no 
     debate regarding the reality of the product cycle, and nobody 
     debates the fact that the CPI introduces products late, thus 
     missing much of the price decline that typically happens in 
     the first phase of the product cycle. For example, the 
     microwave oven was introduced into the CPI in 1978 and the 
     VCR and personal computer in 1987, years after they were 
     first sold in the marketplace.
       A second aspect of new product bias results from a narrow 
     definition of a commodity. When a new product is finally 
     introduced into the CPI, no comparison is made of the price 
     and quality of the new product with the price and quality of 
     an old product that performed the same function. For 
     instance, people flock to rent videos, but the declining 
     price of seeing a movie at home, as compared to going out to 
     a theater, is not taken into account in the CPI. Similarly, 
     the CPI missed the replacement of electric typewriters by 
     electronic typewriters and then PCs with word-processing and 
     spell-checking capability, or CD-ROM encyclopedias that cost 
     far less than old-fashioned bound-book versions and eliminate 
     many trips to the library. Inevitably, however, many new 
     products embody genuinely new characteristics that have no 
     previous counterpart. How does one value electronic mail that 
     provides a new set of bonds and communication between parents 
     and their children who are off at college?
       This discussion of new products leads inevitably to deeper 
     questions about changes in the standard of living of the 
     average American. Positive changes made possible by consumer 
     electronics need to be weighed against increasing crime rates 
     that have forced some families to divert expenditures to 
     burglar systems and security guards. The industrial 
     revolution caused widespread air and water pollution, while 
     numerous factors since the mid-1960s have caused a major 
     decline in the presence of many types of contaminants in the 
     air and water.\20\
       How large is the bias in the CPI introduced by inadequate 
     treatment of quality change, and by the problems created by 
     new products? Estimates of bias vary widely by product, and 
     there are examples of both positive and negative bias. For 
     instance, one study found an upward bias in the CPI index of 
     TV sets of six percent per year, of which almost half was due 
     to the failure of the CPI to place a value on reduced repair 
     incidence and electricity use. Most other studies of consumer 
     durable have found an upward bias in the CPI, except in 
     the case of new automobiles for the period since the late 
     1960s. As stated above, the automobile is a complex 
     product in which many small improvements have been made 
     over the years. Evaluating the negative quality change in 
     the shift to smaller cars as against the substantial 
     improvements in fuel economy (which are worth different 
     amounts in different periods, depending on gasoline 
     prices) is a complex task. However, there seems to be 
     little doubt that the CPI index for used autos has been 
     upward biased, as few if any adjustments for quality 
     change were made to this index during much of the postwar 
     period, and the price index for used autos drifts upward 
     relative to new autos by an implausible amount.
       Studies have found a downward bias in the CPI in two 
     important areas. Prior to 1988, the CPI index for rental 
     housing (which since 1983 has also been used for owner-
     occupied housing) did not take into account the deterioration 
     in housing stock quality as a result of aging and 
     depreciation. Clothing is another problematic area, where the 
     difficult task of separating taste or fashion changes from 
     quality changes, as well as a strong seasonal pattern in 
     clothing prices, may have created a substantial downward bias 
     in apparel prices.
       Thus we find that studies point to substantial upward bias 
     for some products, mainly consumer durables, but 
     countervailing downward bias for several important 
     categories, namely home rent and apparel. Further, the 
     sources of bias shift over time. Since 1987 the BLS has made 
     an attempt to adjust the prices of used cars for quality 
     change, reducing or eliminating that previous source of 
     upward bias. Going in the opposite direction, since 1988 the 
     BLS has eliminated the downward bias due to the failure to 
     take account of aging and depreciation in rental housing.
       Nevertheless, it is likely that there is a substantial 
     upward bias in the CPI, however hard it may be to measure, 
     and much of this is likely to come from new products. 
     Whatever invention we take--whether the automobile that 
     allowed limitless flexibility in the time and destination of 
     rapid transportation, or the jet plane and communications 
     satellite that tied together people in far-flung nations, or 
     the television and VCR that allowed almost any motion picture 
     to enter the home, or the PC with CD-ROM that 

[[Page S 15461]]
     promises ultimately to bring the Library of Congress into every home--
     these new developments have made human life better on a large 
     scale.
       In the concluding section of the interim report, we put 
     forth estimates for the main categories of CPI bias, stated 
     in the form of a ``point estimate'' and a range of 
     uncertainty. In the category of quality change bias 
     (excluding new product bias), we have chosen a relatively 
     conservative point estimate of 0.2 percent per year. Existing 
     studies of consumer durables, weighted by the share of 
     consumer durables in total consumption, point to a bias of at 
     least 0.3 percent per year. Our choice of 0.3 balances the 
     effect of a possible downward bias in apparel against the 
     likelihood that substantial quality change is missed in many 
     areas of nondurable goods and services. Because we are more 
     uncertain in the direction of a higher upward bias, our range 
     of uncertainty for quality change is asymmetric, going from 
     0.2 to 0.6.
       The most difficult question of all is to place a point 
     estimate on new product bias. We have approached this 
     question by carrying out the following thought experiment. 
     Take the market basket of goods and services available in 
     1970 and labeled with 1970 prices. Take the market basket 
     available in 1995 and labeled with today's prices. Ask the 
     consumer, how much more income would you require to be as 
     satisfied with the 1995 basket and prices as with the 1970 
     basket and prices? The CPI says 4 times as much income would 
     be necessary, because the CPI has quadrupled since 1970. But 
     that 1970 market basket has no VCRs, microwave ovens, or 
     modern anti-ulcer drugs; its color TV sets break down all 
     the time; and it refrigerators use a lot of electricity. 
     Consumers forced to answer this question are going to miss 
     many benefits of modern life and are not going to say that 
     four times as much income would be necessary--maybe 3 
     times, maybe 3.5 times, but not 4 times. That is the 
     ultimate test of new product bias in the CPI.
       To translate this approach into an annual rate of change, 
     an answer of ``3.5 times'' would imply an upward bias of 0.54 
     percent a year.\21\ The commission has chosen to take a 
     lower, more conservative point estimate of a new product bias 
     of 0.3 percent per year, but to extend the range of 
     uncertainly from 0.2 to 0.7 percent per year. We will attempt 
     in our final report to assemble new evidence on this issue 
     and to narrow the range of uncertainty.


                       v. separate price indexes?

       In principle, if not practice, a separate cost of living 
     index could be developed for each and every household based 
     upon their actual consumption basket and prices paid. As 
     noted above, the aggregate indexes use data reflecting 
     representative consumers. Some have suggested that different 
     groups in the population are likely to have faster or slower 
     growth in their cost of living than recorded by changes in 
     the CPI. We find no compelling evidence of this to date, and 
     in fact two studies suggest that disaggregating by population 
     group, for example by region or by age, would have little 
     effect on measured changes in the cost of living.\22\ 
     Further, work on this subject remains to be done.
       Beyond the different consumption baskets, it is important 
     to understand our analysis of the sources of bias are applied 
     to representative or average consumers. Some consumers will 
     substitute more than others, and the substitution bias may be 
     larger for some, smaller for others. Likewise, some are more 
     likely to take advantage of discount outlets; others less so. 
     Perhaps more importantly, the benefits of quality change and 
     the introduction of new products may diffuse unevenly 
     throughout the population. Some will quickly gain the 
     benefits of cellular telephones, for example, while others 
     may wait many years or decades or never use them. This is yet 
     another reason why we have been very cautious in our point 
     estimates for these particular sources of bias.


                             vi. conclusion

       While the CPI is the best measure currently available, it 
     is not a true cost of living index. It suffers, as do all 
     price indices, from a variety of conceptual and practical 
     problems as the vehicle for measuring changes in the cost of 
     living. Despite important BLS updates and improvements in the 
     Consumer Price Index, it is likely that changes in the CPI 
     have substantially overstated the actual rate of price 
     inflation. Moreover, revisions have not been carried out in a 
     way that can provide an internally consistent series on the 
     cost of living over an extended span of time. More 
     importantly, changes in the Consumer Price Index are likely 
     to continue to overstate the change in the true cost of 
     living for the next few years. This overstatement will 
     have important unintended consequences, including 
     overindexing government outlays and tax rules and 
     increasing the federal deficit and debt. If the intent of 
     such indexing is to insulate recipients and taxpayers from 
     changes in the cost of living, use of the Consumer Price 
     Index has in the past, and will in the future, 
     overcompensate (on average) for changes in the true cost 
     of living.
       Table 2 presents the Commission's evaluation of the biases 
     in using changes in the Consumer Price Index as a measure of 
     changes in the cost of living for the recent historical past 
     (the last few years). It presents point estimates, and 
     plausible ranges of values, for each of the five sources of 
     potential bias as well as the overall bias. Our best judgment 
     of the overstatement of the change in the cost of living 
     embedded in changes in the CPI for this historical period is 
     1.5% per annum. It is likely that a large bias also occurred 
     looking back over at least the last couple of decades, 
     perhaps longer, but we make no attempt to estimate its size.

       TABLE 2.--ESTIMATES OF RECENT HISTORICAL BIASES IN THE CPI       
                           [Percent per annum]                          
------------------------------------------------------------------------
                  Source of bias                     Estimate    Range  
------------------------------------------------------------------------
Substitution bias.................................        0.3    0.2-0.4
Outlet bias.......................................        0.2    0.1-0.3
Formula bias......................................        0.5    0.3-0.7
Quality change....................................        0.2    0.2-0.6
New products......................................        0.3    0.2-0.7
                                                   ---------------------
      Total.......................................        1.5    1.0-2.7
------------------------------------------------------------------------
NB: Total bias assumed to be additive across types and independent of   
  the level of inflation. See text.                                     

       A plausible range of values is 1.0% to 2.7% per annum. The 
     point estimate of 1.5% includes 0.5% for formula bias, which 
     is the technical problem in using methods that impart an 
     upward bias in the movement from elementary or extremely 
     disaggregated price quotations to broader commodity groups. 
     The BLS is aware of this problem, and is moving to correct 
     it. Hopefully, it will be eliminated quickly.
       Excluding formula bias, the point estimate is 1.0% per 
     annum, and the range is 0.7% to 2.0% per annum. Note that the 
     range of uncertainty is not symmetric around our point 
     estimate. It is far more likely that changes in the CPI have 
     embedded a larger than a smaller bias. The range of potential 
     upward bias is significantly larger because we have been 
     conservative in our point estimates of the biases from the 
     sources of quality change and new products. The conceptual 
     issues involved in measuring these two sources of bias are 
     even more difficult than the other sources, and the range of 
     studies upon which to base such conclusions at this point is 
     insufficient to support our ``best judgment'' as strongly as 
     those for the other sources of bias. Hence, we have been 
     especially cautious in these two areas.
       Past is not necessarily prologue. What can we say about the 
     likely sources of bias moving forward, as opposed to 
     estimates of the biases looking back at recent history? We 
     believe the substitution bias is likely to be as large or 
     larger as in the recent past. It is likely that the 
     substitution bias will drift up a little bit, perhaps to 0.4 
     %, until 1998 when the CPI will incorporate the new 
     expenditure weights from the 1993-95 expenditure survey. Note 
     that at that time the expenditure weights will still be four 
     years out of date and thus much substitution may have already 
     occurred. However, at that time it is likely that the 
     substitution bias will decrease considerably, to no more than 
     0.2%. As time moves on, it will likely drift up again. So, 
     even though the base year will be updated in 1998, it is 
     likely that for several years the substitution bias will 
     continue to be large then shrink for a short period before 
     gradually drifting back up again by the turn of the century. 
     Thus, a substitution bias on the order of 0.3% is likely to 
     be a good approximation on average for the next decade, 
     although not year by year.
       Until and if procedures are changed, we expect the outlet 
     substitution bias to be approximately 0.2% per year. As noted 
     above, we believe the BLS has discovered, and is developing 
     procedures to eliminate, the formula bias. Our estimate for 
     the future of 0.0% assumes that the BLS will quickly and 
     completely remove the formula bias. To the extent that 
     methods are changed slowly or incompletely, a sizable formula 
     bias will remain. Thus, again, the 0.0% is perhaps 
     conservative, especially for the very short-run. Finally, our 
     estimates for quality change and new products of 0.2% and 
     0.3%, which, as discussed above, we believe to be quite 
     conservative, are likely to apply in the future as well.

          TABLE 3.--ESTIMATES OF LIKELY FUTURE BIAS IN THE CPI          
                           [Percent per annum]                          
------------------------------------------------------------------------
                  Source of bias                     Estimate    Range  
------------------------------------------------------------------------
Substitution bias.................................        0.3    0.2-0.4
Outlet bias.......................................        0.2    0.1-0.3
Formula bias......................................        0.0  .........
Quality change....................................        0.2    0.2-0.6
New products......................................        0.3    0.2-0.7
                                                   ---------------------
      Total.......................................        1.0    0.7-2.0
------------------------------------------------------------------------
Assumes BLS quickly and completely fixes the problem. Will continue to  
  be substantial until this occurs.                                     

       This brings our estimate of the upward bias of changes in 
     the CPI as a measure of the change in the cost of living to 
     1.0% per year. However, the certainty that the Commission 
     ascribes to alternative estimates clearly is greater the 
     lower the estimate within the plausible range. For example, 
     while 1.0% is our interim best estimate and likely to be 
     conservative, we are even more certain that the lower end of 
     our plausible range does not overstate the upward bias in the 
     CPI.
       These separate biases are approximately additive and likely 
     to be independent of modest swings in the true inflation 
     rate. Thus, a bias of 1% implies that when changes in the CPI 
     show inflation rising from 3% to 5%, it is likely actually to 
     be rising from 2% to 4%. Note the bias primarily affects the 
     level, not the change, in the inflation rate. At very high 
     rates of inflation, the bias may increase (one might assume 
     greater outlet 

[[Page S 15462]]
     and commodity substitution), but we currently have no evidence 
     regarding this issue.
       Figure 2 shows the compounding effect over time of such a 
     bias on the index. While 1.0% may seem to be a small amount 
     in any given year, cumulatively year after year it adds up to 
     a sizable difference. [Figure 2 not reproducible in Record]
       An additional word of caution is in order. This Commission 
     has thus far relied primarily on studies already produced 
     prior to the convening of the Commission, with a small amount 
     of additional work that we have been able to commission in 
     the two months since our inception. Thus, our judgments 
     reported above are not much advanced beyond what was 
     available in the three rounds of Senate Finance Committee 
     Hearings earlier this year. Given the short time available to 
     this Commission, there are many issues which we have not yet 
     been able to explore adequately. While we expect the interim 
     conclusions to hold up under further examination, they will 
     also be subject to amendment as we proceed with our 
     investigation.
       In our final report we expect to have a more complete 
     analysis and evaluation and will certainly have specific 
     recommendations for procedures to improve and/or complement 
     the CPI. It may be possible to implement some of these 
     suggestions quickly, others may take considerable time and 
     additional resources.


  members of the advisory commission to study the consumer price index

       Michael J. Boskin, Ph.D., Chairman, Tully M. Friedman 
     Professor of Economics and Senior Fellow, Hoover Institution, 
     Stanford University, Stanford, California.
       Ellen R. Dulberger, Ph.D., Program Director IBM Global I/T 
     Services Strategy and Economic Analysis, White Plains, New 
     York.
       Robert J. Gordon, Ph.D., Chairman, Department of Economics; 
     and Stanley G. Harris, Professor in the Social Sciences, 
     Northwestern University, Evanston, Illinois.
       Zvi Griliches, Ph.D., Paul M. Warburg, Professor of 
     Economics, Harvard University, Cambridge, Massachusetts.
       Dale Jorgenson, Ph.D., Chairman, Department of Economics, 
     and Frederic Eaton Abbe, Professor of Economics, Harvard 
     University, Cambridge, Massachusetts.


                               Footnotes

     \1\ We would like to thank the staffs of the Bureau of Labor 
     Statistics, Congressional Research Staff and the 
     Congressional Budget Office for valuable assistance and 
     cooperation in the early stages of the Commission's work.
     \2\ The two most commonly used measures are the CPI-U and 
     CPI-W. The former is for all urban consumers, roughly 80% of 
     the population; the latter is for urban wage and clerical 
     workers, about 32% of the population. Note that the 
     expenditure shares may be quite different than the average 
     for any particular household, and also on average for 
     subgroups of the population. Also, the prices paid for some 
     products may differ for some households from the prices 
     actually sampled. In principle, if not practice, a separate 
     cost of living index could be developed for each and every 
     household based on their actual consumption basket and prices 
     paid. The overall index is used to approximate this with the 
     data reflecting representative consumers. Whether this is 
     itself sufficiently misleading as to warrant separate price 
     indexes for different population subgroups is discussed 
     below.
     \3\ Congressional Budget Office (1994), ``Is the Growth of 
     the CPI a Biased Measure of Changes in the Cost of Living?'' 
     CBO Papers, Washington. Congress of the United States, 
     October, p. 32.
     \4\ Robert Gillingham and Walter Lane, ``Changing the 
     Treatment of Shelter Costs for Homeowners in the CPI,'' 
     Monthly Labor Review, June 1982, p. 9.
     \5\ James Duggan, Robert Gillingham, and John Greenlees, 
     ``Housing Bias in the CPI and its Effect on the Budget 
     Deficit and the Social Security Trust Fund,'' Office of 
     Economic Policy, U.S. Department of the Treasury, June 30, 
     1995, page 6.
     \6\ Congressional Budget Office, ``The Economic and Budget 
     Outlook: Fiscal Years 1996-2000,'' Report to the Senate and 
     House Committees on the Budget, Washington, Congress of the 
     United States, January 1995, Table 2-8, p. 43.
     \7\ June O'Neill (1995), ``Prepared Statement,'' Consumer 
     Price Index, Hearings before the Committee on Finance, U.S. 
     Senate, 194th Congress, First Session, Washington, 
     U.S.G.P.O., Table 1, p. 146.
     \8\ All CBO budget estimates are relative to CBO's January 
     1995 baseline and do not include the small adjust assumed in 
     the out-years of the budget resolution.
     \9\ Each method has come to be named for its inventor. See 
     below.
     \10\ This index is called the Laspeyres index.
     \11\ This index is called the Paasche index.
     \12\ This index is called the Fisher, or Fisher Ideal, index.
     \13\ A.M. Aizcorbe and P.C. Jackman, ``The Commodity 
     Substitution Effect in CPI Data, 1982-91.'' Monthly Labor 
     Review, U.S. Bureau of Labor Statistics, pp. 25-33 (December 
     1993).
     \14\ Updated by BLS for the Commission.
     \15\ B.R. Moulton, ``Basic components of the CPI: Estimation 
     of Price Changes.'' Monthly Labor Review, U.S. Bureau of 
     Labor Statistics, pp. 13-24 (December 1993).
     \16\ M. Reinsdorf, ``The Effect of Outlet Price Differentials 
     in the U.S. Consumer Price Index,'' in Price Measurements and 
     Their Uses, M.F. Foss, M.E. Manser and A.H. Young (eds), NBER 
     Studies in Income and Wealth, Vol. 57, pp. 227-254 (1993).
     \17\ B.R. Moulton, ``Basic components of the CPI: Estimation 
     of Price Changes,'' Monthly Labor Review, U.S. Bureau of 
     Labor Statistics, pp. 13-24 (December 1993).
     \18\ Dale W. Jorgenson and Daniel T. Slesnick. ``Individual 
     and Social Cost-of-Living Indexes.'' Price Level Measurement, 
     W.E. Diewert and C. Montmarquette (eds.), Ottawa, Statistic 
     Canada, pp. 241-336 (1983).
     \19\ F.M. Fisher and Zvi Griliches, ``Aggregate Price 
     Indices, New Goods, and Generics,'' Quarterly Journal of 
     Econometrics (1995).
     \20\ The CPI implicitly values the improvement in air quality 
     made possible by mandated anti-pollution devices in 
     automobiles, since it treats the cost of mandated anti-
     pollution and safety devices as an improvement in quality 
     rather than an increase in price. However, the CPI is 
     inconsistent, since a portion of the higher cost of 
     electricity, steel, and other products is also due to 
     environmental regulation, and the benefits of higher air and 
     water quality made possible by regulation of products other 
     than automobiles is not taken into account.
     \21\ An index that rises from 1.0 to 4.0 over 25 years 
     exhibits a compound growth rate of 5.55 percent per year. An 
     index that rises from 1.0 to 3.5 over 25 years exhibits a 
     compound growth rate of 5.01 percent per year. The difference 
     is 0.54 percent per year.
     \22\ See M. Boskin and M. Hurd, ``Indexing Social Security 
     Benefits: A Separate Price Index for the Elderly,'' Public 
     Finance Quarterly, Volume 13, Number 4, pp. 436-449 (October 
     1985); Dale W. Jorgenson and Daniel T. Slesnick, ``Individual 
     and Social Cost-of-Living Indexes,'' Price Level Measurement, 
     W.E. Diewert and C. Montmarquette (eds.), Ottawa, Statistics 
     Canada, pp. 241-336 (1983). However, very preliminary 
     unpublished work suggests that for the period 1982-91 the 
     larger fraction of expenditures on out-of-pocket healthcare 
     by the elderly combined with the more rapid rise in 
     healthcare prices than overall prices for this period might 
     lead to a slightly faster rise in a price index for the 
     elderly. The rate of healthcare price inflation has slowed 
     substantially of late, so it is unlikely this result will be 
     reproduced for the mid-1990s.
     \23\ The bias is currently running at 1.5% per annum or more, 
     in our best judgment. We do not estimate it year by year for 
     this period but believe this estimate is close on average for 
     the period. Figure 1 is for illustrative purposes only.
                                                                    ____


                                  CHANGE IN DEFICIT IF ADJUSTMENT MADE FOR CPI OVERSTATEMENT (1 PERCENTAGE POINT LESS)                                  
                                                                [In billions of dollars]                                                                
--------------------------------------------------------------------------------------------------------------------------------------------------------
                                               1996       1997       1998       1999       2000       2001       2002       2003       2004       2005  
--------------------------------------------------------------------------------------------------------------------------------------------------------
Change in Revenues a......................       -1.8       -5.5       -9.8      -13.1      -17.7      -23.0      -27.1      -31.8      -36.2      -40.8
                                           =============================================================================================================
Change in Outlays:........................                                                                                                              
    Social Security/RR Retire.............       -2.6       -6.2      -10.1      -14.1      -18.4      -22.8      -27.4      -29.2      -37.8      -43.6
    SSI...................................       -0.2       -0.5       -0.8       -1.2       -1.9       -2.1       -2.9       -3.6       -4.3       -5.1
    Civil Service Retirement..............       -0.2       -0.7       -1.1       -1.5       -1.9       -2.4       -2.9       -3.4       -3.9       -4.5
    Military Retirement...................          d       -0.3       -0.6       -1.2       -1.6       -2.0       -2.4       -2.9       -3.4       -3.9
    Vets Comp & Pensions..................       -0.1       -0.3       -0.5       -0.6       -0.9       -1.3       -1.6       -2.1       -2.5       -3.1
    EITC a................................        (d)       -0.5       -1.1       -1.8       -2.4       -3.1       -3.9       -4.7       -5.4       -6.2
    Other b...............................        (d)        (d)       -0.1       -0.1       -0.1       -0.1       -0.1       -0.1       -0.1       -0.1
    Offsets c.............................        (d)        0.1        0.2        0.4        0.7        1.0        1.4        1.8        2.3        2.7
                                           -------------------------------------------------------------------------------------------------------------
      Total Outlay Change.................       -3.1       -8.4      -14.1      -20.2      -26.5      -32.7      -39.8      -44.1      -55.2      -63.9
    Debt Service..........................       -0.2       -0.8       -2.0       -4.0       -6.7      -10.2      -14.7      -20.1      -26.6      -34.4
                                           -------------------------------------------------------------------------------------------------------------
      Change in Deficit...................       -5.0      -14.7      -25.9      -37.3      -50.9      -65.9      -81.6      -96.0     -117.9     -139.1
--------------------------------------------------------------------------------------------------------------------------------------------------------
(a) Estimates for 1996-2000 prepared by the Joint Committee on Taxation. CBO, based on the JCT model, has extrapolated projections for 2001-2005.       
(b) FECA, foreign service retirement, PHS retirement, and Coast Guard retirement.                                                                       
(c) Includes Medicare, Medicaid, and Food Stamp offsets to cuts in the Social Security COLA.                                                            
(d) Less than $50 million.                                                                                                                              
                                                                                                                                                        
Notes: CBO estimates that the CPI has probably grown faster than the cost of living by between 0.2 and 0.8 of a percentage point in recent years. For   
  purposes of these calculations, though, CBO has assumed an adjustment of a full percentage point. Revenue increases are shown with a negative sign    
  because they reduce the deficit.                                                                                                                      
Source: Congressional Budget Office.                                                                                                                    

                 [Memorandum as of September 28, 1995]

     From: Harry C. Ballantyne
     Subject: Estimated Long-Range Effects of Alternative 
         Reductions in Automatic Benefit Increases--Information

       The following table shows our estimates of the long-range 
     effects of modifying the present-law calculation of all 
     future automatic benefit increases by reducing each increase 
     by one percentage point (or alternatively one-half of one 
     percentage point) from the present-law increase, which is 
     equal to the percentage increase in the CPI-W. The estimates 
     are based on the assumption that the reduction would first be 
     reflected in the next automatic benefit increase, for 
     December 1995, or, alternatively, that the reduction would 
     first be reflected in the automatic benefit increase for 
     December 1996. The estimates are based on the intermediate 
     assumptions in the 1995 Trustees Report and are shown for the 
     combined OASI and DI Trust Funds.

                                                                                                                                                        

[[Page S 15463]]
----------------------------------------------------------------------------------------------------------------
                                                                               Reduction of 1%    Reduction of  
                                                                                  effective      0.5% effective 
                                                                     Present     December--        December--   
                                                                       law   -----------------------------------
                                                                                1995     1996     1995     1996 
----------------------------------------------------------------------------------------------------------------
Change in actuarial balance over next 75 years (percent)...........  .......     1.44     1.41     0.74     0.73
Actuarial balance (percent)........................................    -2.17    -0.74    -0.76    -1.43    -1.44
Year of exhaustion.................................................     2030     2049     2048     2036     2036
First year in which outgo exceeds tax income.......................     2013     2018     2018     2015     2015
Maximum trust fund ratio (percent).................................      269      408      397      332      327
Year Maximum ratio is reached......................................     2011     2015     2015     2014     2014
----------------------------------------------------------------------------------------------------------------



                                              Harry C. Ballantyne,
     Chief Actuary.

                          ____________________