United States
Official title
Consumer Price Index for All Urban Consumers (CPI-U).
Scope
The index is computed monthly and covers about 80 per cent of the
total noninstitutional population. Rural residents outside
metropolitan areas, all farm residents, the armed forces and
persons in institutions are excluded from the index.
Official base
1982-84 = 100.
Source of weights
The weights were derived from the consumer expenditure survey
carried out during 1982-84, consisting of
an interview survey and a diary survey. Approximately
5,000 consumer units were contacted each year for each type of
survey. The consumer unit concept is based on the economic
interdependencies within a housing unit and thus differs from
the concepts of the family and the household. The expenditure
weight for each item stratum is an estimate of total the expenditure
by the index population for that item, calculated as the
product of estimates of mean expenditures of consumer units and
the number of consumer units.
Weights and composition
| Major groups
| Number of items
| Weights
| Approximate number of price quotations
|
| Food and beverages
| 100 | 17.627 | 44102
|
| Housing
| 97 | 41.544 | (c) 18864
|
| Clothing, footwear and upkeep
| 59 | 6.097 | 11059
|
| Transport
| 38 | 17.013 | 6564
|
| Medical care
| 15 | 6.689 | 4483
|
| Entertainment
| 30 | 4.357 | 2800
|
| Other goods and services
| 31 | 6.674 | 2080
|
| Total
| (a) 370 | (b) 100.000 | 89952
|
Note: (a) Includes 23 unpriced items.
(b) Relative importance for December 1991.
(c) Includes 9,000 quotations for shelter.
Household consumption expenditure
Household consumption expenditure refers to all goods and
services that people buy for their day-to-day living.
Method of data collection
The outlet sample is selected from a point-of-purchase survey
conducted each year in about one-fifth of the urban areas
included in the CPI. Consumer units are interviewed in each of
the areas in which prices are collected. Respondents are asked
for information on purchases of items within specific categories
during a prescribed reference period. If a purchase has been
made, the name and address of the outlet is recorded along with
the cost for each transaction. The Bureau of Labor Statistics
then selects a probability sample from these outlets for each
expenditure category, using the expenditure at each outlet as a measure
of size. This ensures an unbiased outlet sample with all types of
establishments represented; the system also permits the estimation of
variances and sampling error.
Prices of most goods and services are obtained through personal visit
by the Bureau's trained representatives to approximately 21,000 retail
establishments and 6,000 housing units. Mail questionnaires are used to
obtain public utility rates, some fuel prices and prices for certain
other items. Prices are collected in 85 urban areas across the country
from retail outlets and service establishments. Prices of food, fuel
and a few other items are obtained each month in all 85 locations.
Other items and services are priced each month in the five
largest metropolitan areas and every two months in the remaining areas.
Food is priced throughout the month. The prices used to calculate the
index are the regular cash prices in effect, including all taxes
directly associated with the purchase and use of the items.
Housing
Information on rent is collected via the rent survey, covering a sample
of about 60,000 housing units divided into six panels. Every month, one
panel is priced and the panels are rotated so that the same units are
priced every six months. The rental equivalence method is used to
measure the cost of owner-occupied housing. The price of home repair
materials purchased by tenants is included, as well as the rental values
of owner-used vacation property. The shelter index is adjusted for the
change in quality resulting from the effect of aging on rental housing,
using the hedonic regression method.
Specification of varieties
Items for pricing are selected by the price collector within each sample
outlet in proportion to their dollar volume of sales in the outlet, i.e.
with probability proportional to size (the percentage of sales is the
measure of size). The price collector prepares a detailed description
of the selected item using a check-list developed by commoditiy
specialists for each entry-level item. This outlet-specific
specification is used to identify the item for pricing.
Substitution, quality change, etc.
Quality adjustments are made by the commodity specialist when
substitution occurs. If quality adjustment is not possible,
a linking method is used.
Seasonal items
Prices for seasonal items not currently available are imputed
using standard imputation procedures.
Computation
The index is computed according to the Laspeyres formula as a
weighted arithmetic average with fixed base, using weights
corresponding to the base period.
Weighted price relatives for the current and previous periods are
used to compute the item indices. The weights are based on the
importance of the item, the outlet and the area, and on the number of
other sampled price quotations.
The index for All Urban Consumers is an aggregation of data for the
85 primary sampling units where prices are collected. Weights
reflecting estimated current expenditure for the base period market
basket are determined for each item in each primary sampling unit. The
market baskets are then weighted using the population represented by
each primary sampling unit to obtain the All Urban item weights.
Organisation and publication
US Department of Labor, Bureau of Labor Statistics: The CPI
Detailed Report
(Washington, DC).
Idem: The Consumer Price Index: The 1987 Revision, Report
736
.
Idem: The Consumer Price Index: BLS Handbook of
Methods
Bulletin No. 2285, Chapter 19.