The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
May 30,1 980
MeasuringInflation
The consumer price index (CPI) rose at a
sharply decelerated pace in April, its
0.9-percent increase contrasting sharply \,,-dth
the successive l A-percent increases in each
of the first three months of the year. Indeed,
the April 1 980 increase was the smallest
monthly increase since January 1979. But in
periods of deceleration as well as periods of
acceleration, questions arise about the CPl's
value as an indicator of the actual extent of
the nation's inflation problem.
The CPl's role is magnified by its use as an
indexing mechanism to compensate private
and public income recipients for past losses
to inflation. Most major labor contracts
contain escalator provisions based on the
CPI, while roughly half of all Federal
spending is now indexed for inflation. (At
midyear, for example, all social-security
recipients will receive a 13-percent benefit
increase based on the CPI increase of the past
year.) Because of the stakes involved, we
would do well to examine the criticisms
leveled against the consumer-price index, to
determine whether that index can be
improved or whether some other measure
would provide a better yardstick for inflation.
with items whose prices have risen most
rapidly, and in this way overstates the overall
rate of inflation.
Some critics also contend that the CPI
overstates inflation because it fails to take
quality changes adequately into account.
Everyone agrees that quality changes inan
item should not be reflected as a price
change, since the index is supposed to
measure the cost of purchasing a constant
market basket of goods and
The
Bureau of Labor Statistics (BLS) attempts to
handle the problem by adjusting for the cost
associated with producing specific quality
changes. But at times BLS may fail to adjust
enough for such changes, especially in those
areas where new technology creates better
quality at the same or reduced cost. On the
other hand, BLSmay err in the other direction
. by failing to adjust prices upward for quality
deterioration -by fai ling to adjust for
products which break down faster than they_
formerly did because of the use of lowerquality materials in their production. On
balance, according to key studies, these
quality factors may offset each other in the
construction of the overall index.
Weighting, quality problems
Some critics contend that the CPI overstates
the cost of consumer purchases because it is a
fixed-weight index-because it measures
price changes of items in a market basket
whose composition has remained
unchanged since the 1 972-73 base period.
But some price changes induce consumers
who wish to maintain their standard of living
to switch to products not purchased
previously, or else to change the proportions
in which items are purchased. Many
consumers buy proportionately less of those
items whose prices have risen most rapidly,
while buying more of those items whose
prices have risen at a slower pace.
Consequently, when prices rise, a
fixed-weight index becomes overly weighted
problem
In recent years, the strongest criticism of the
CPI has arisen because of its treatment of
owner-occupied housing-an increasingly
expensive item whose treatment has come to
account for most of the recent difference
between the CPI and alternative inflation
measures. BLS now values all goods and
services in the 1 972-73 market basket at the
prices the consumer pays wheil purchasing
those items. Consequently, BLS includes in
the index the transaction price of a newly
pu rchased home as well as the cost of
contracted mortgage-interest paymentsalong with the purchase prices of toothpaste,
milk, TV sets, and a host of other goods and
services..
This does not mean, as frequently alleged,
that BLStreats housing prices and mortgage
costs as if all home-owners purchased their
houses in the current survey month. In fact,
only 6 percent of all families are counted as
current home purchasers, because that was
the percentage when the composition of the
CPt market basket was determined in
1972-73. Even so, we may question whether
the total transaction price of a house shou Id
be included in an index of consumption
items. The transaction price of a house
implicitly includes both whatthe consumer is
willing to pay for shelter and what he is
willing to pay for housing as an investment.
Including the investment return as a
home-ownership cost necessarily
exaggerates the cost of shelter in the CPI, and
thus leads to an overstatement of inflation,
during the recent period of soaring
home-appreciation values.
This problem can be corrected through
the use of a flow-of-services approachthrough the estimation of the major cost
components that owners incur when
providing housing for themselves. The flowof-services approach, unlike the current
official approach, does not assume that the
entire value of the house is "consumed" or
used in the month it is purchased.
The flow-of-services approach can
be measured on a user-cost basis, by
estimating the major costs that owners incur,
or on a rental-equivalence basis, by
estimating what a home-owner would pay if
renting the house. The two approaches are
theoretically equivalent but sometimes yield
different resuIts.
The major components of the user-cost
alternative include mortgage costs,
maintenance and repairs, property taxes,
insurance, capital gains which arise from
changing home values, and home-equity
costs that represent the foregone retu rn
on money invested in owning a house. The
latter cost is not clearly defined, and analysts
debate whether it should be represented
by the rate of return on some set of financial
assets, by the mortgage rate itsel( or some
other opportunity cost. Unfortunately, the
estimate of home-ownership costs depends
critically on which alternative is chosen,
and at least until this issue is resolved,
a user-cost estimate of housing services
cannot be incorporated in an official CPt
estimate. On the other hand, the rentalequivalence method suffers from the
difficulty of obtaining a good sample of rental
housing during certain periods, such as
rent-control periods, or in certain areas, such
as suburban developments which lack
sufficient rental housing.
In spite of these problems, a flow-of-services
approach represents a conceptually more
accurate measure of home-ownership costs
than the price-transactions approach of the
official CPI, and the BLS thus is pursuing
further studies of the flow-of-services
approach. In 1 979, both flow-of-services
approaches led to lower estimates of inflation
than the official CPt which averaged
11 .2 percent. The rental-equivalence method
produced a 9.6-percent estimate of over-all
consumer inflation, while a user-cost method
produced a 10.1 -percent estimate.
Best inflation measure?
In view of all the problems with the CPI, some
observers have proposed concentrating
instead on the price index (deflator) for
personal consumption expenditures in the
GN P accounts. This PCEdeflator uses the
same prices for consumption items as does
the CPI. It differs in one important respect,
however, in its treatment of home-ownership
costs. By using a rental-equivalence basis
instead of a price-transaction basis, the PCE
deflator avoids the overstatement of housing
costs which mars the official CPI. Largely for
that reason, the deflator was 2.3 percentage
points below the official CPI during 1 979,
after averaging 0.6 percentage points lower
during the preceding decade (see chart).
The PCEdeflator also differs in its treatment of
weights, by using variable weights rather than
fixed weights in the calculation of the index.
This index compares the prices of today's
market basket with what that basket wou Id
have cost in 1972 prices. Therefore, every
quarter the PCEdeflator prices a different
market basket. It does not attempt to keep
track of the changing costs of maintaining a
standard of living, or of buying a given market
basket of goods and services. In addition, the
PCEdeflator has other difficulties as a
cost-of-living-index. Its market basket
excludes purchases of used cars and all used
items, and takes no account of consumer
financing costs.
avoids the CPI overstatement of housing
costs, while avoiding the PCEdeflator's
exclusion of items in the consumer's market
basket. These alternative (PI estimates
generally lie between the official CPI and the
PCEdeflator; in 1979, for example, the
user-cost alternative yielded a lO. l-percent
estimate of inflation, compared with
11.2 percent for the official CPI and
8.9 percent forthe PCEdeflator.
These alternative approaches have gai ned
increasing support as betteralternatives for
measuring inflation. Both the labor and
business advisory panels to BL5, while
recognizing the measurement diffkulties,
have recommended that BL5 move toward
adoption of these more conceptually correct
methods of pricing the flow of services from
owner-occupied housing in the CPI.
These considerations suggest the need for a
measure which avoids the drawbacks of both
the official (PI and the PCEdeflator in
estimating inflation. Either the rentalequ ivalence or the user-cost alternative to the
official CPI might provide just such a
measure-because as noted above, each
Ro§eMcElhattan
Change (%)
1:2
CPI (U ser
8
4
Pri ce I n dexes
1 968
1979
3
--------_._---
..
__ _ ---_.---_._
.. ..
_----_.
.. -._...- ... _----._--.. _..
...
_-----_.
-. _---_.
_-----------
.. uogaJo .. E'pE'AaN oLjE'PI
E'!UJ01!1
E':) .. E'UOZ!J\! .. E')ISE'IV
Q
!!E'ME'H ..
·j!Ii!'J
CGJ
Ui!'S
lSL · O N
OI Vd
:J9VIS Od ·s·n
llVW SSV1 J
:{1\illC?i>Will:{1
JJ\2?(([[C?i>Clli
BANKING DATA-TWELfTH fEDERALRESERVE
DISTRICT
(Dollar amounts in millions)
SelectedAssetsandliabilities
large CommercialSanies
Loans (gross, adjusted) and investments*
Loans (gross, adjusted) - total#
Commercial and industrial
Real estate
Loans to individuals
Securities Iqans
U.s. Treasury securities*
Other securities*
Demand deposits - total #
Demand deposits - adjusted
Savings deposits - total
Time deposits - total #
Individuals, part. & corp.
(Large negotiable CD's)
Averages
of Daily figures
Member Bank ReservePosition
Excess Reserves (+ )/Deficiency (-)
Borrowings
Net free reserves ( + )/Net borrowed( - )
Amount
Outstanding
Change
from
5/14/80
5/7/80
137,327
115,664
33,148
46,243
24,019
1,173
6,343
15,320
42,220
30,454
26,191
64,474
55,565
23,018
-
Change from
year ago
Dollar
.Percent
+
+
372
377
242
+ 111
114
+ 34
14
+ 19
684
2
+ 125
469
- 354
351
Weekended
Weekended
5/14/80
5/7/80
282
4
278
+ 8.7
+ 11.9
+ 7.8
+ 23.8
+ 10.0
10,952
12,268
+ 2,385
+ 8,889
+ 2,174
525
1,448
132
+
132
252
+
3,570
+ 14,316
+ 14,734
+ 5,891
291
34
257
-
+
-
+
-
+
+
+
30.9
18.6
0.9
0.3
0.8
12.0
28.5
36.1
34.4
year-ago p!=riod
-
18
129
111
* Excludes trading account
#
securities.
Includes items not shown separately.
Editorial commentsmayhe addressedto the editor (William Burke)or to the author .... Freecopiesof this
andother federal Reservepublicationscanbeobtainedby callingor writing the PublicInformationSection,
federal ReserveBankof SanFrancisco,P.O. Box 7702,SanFrancisco94120.Phone(415)544-2184.
@FedFRASER