Revision of Consumer Price Index is now under way The Bureau of Labor Statistics is in the midst of a 5-year programto update and improve the Consumer Price Index (CPI). The resultingchanges will be introduced in the January 1987 indexes. The 1987revision will use the Consumer Expenditure Survey data from 1982-84 andpopulation distributions from the 1980 census to update the CPI marketbasket. A greatly enhanced housing survey is being developed that willimprove the rental equivalence measure of homeowner costs recentlyintroduced in the CPI. Many of the sampling advances introduced in the1978 revision will be refined for 1987 and other methodologicalenhancements will be made.
This article explains why periodic CPI revisions are needed,briefly reviews previous revisions, and describes the current revisionplans. Why periodic revisions are needed The CPI is a measure of price change for a fixed market basket ofgoods and services of constant quantity and quality purchased forconsumption. It is essential to update that market basket periodicallyso that the CPI reflects price changes of items currently purchased byconsumers. Consumers change their purchasing patterns as a result ofchanges in a number of factors, including relative prices, real income,demographic characteristics, and tastes. Price changes over time may differ among items and thesedifferences can affect consumer demand. This is illustrated by rapidlyrising prices for energy items over the last decade. In the ConsumerPrice Index for All Urban Consumers (CPI-U), energy products (gasoline,motor oil, electricity, natural gas, fuel oil, bottled gas, and coal)rose 218 percent from December 1972 to December 1980, more than twice asfast as the average increase for all items. According to data from theConsumer Expenditure Surveys, urban consumer units1 reported an averageannual expense for energy items of $743 for the 1972-73 period and anannual average of $1,783 for the 1980-81 period.
This 140-percentincrease is substantially smaller than the change that occurred forprices and implies a reduction in consumption of energy items as aresult of higher relative prices. This adjustment was also seen inrelated consumption such as the increased demand for smaller and morefuel-efficient automobiles. Another factor which can influence consumers’ consumptionpatterns is changing real income. If prices paid by consumers and theirmoney income were to increase at the same rate, consumers’ realincome would remain unchanged. Average money income in constant dollarsdeclined about 7 percent between 1972 and 1981, both for households andfor families. However, per capita average real money income increasedby more than 3 percent during the same period.
2 The rise in per capitaincome, in contrast to the decline experienced by families andhouseholds, is a direct result of the average size of families andhouseholds becoming smaller. The impact of rising prices on some families can be offset byhaving additional income from another member joining the labor force.For example, in 1972, 41.5 percent of married women with a husbandpresent were in the labor force. By 1981, the percentage increased to51. The labor force participation rate for married women, husbandpresent, and with children under 6 years of age increased from 30.
1percent in 1972 to a rate of 47.8 in 1981.3 Demographic-related changesof this kind affect expenditure patterns. For example, expenses forsuch items as day care nursery school and babysitting might increase andthere could be additional expenses for eating meals out andtransportation. Still other factors which affect the pattern of consumption overtime are product changes and technological changes which can affectconsumers’ demand for various goods and services. The electronicsindustry in particular has influenced consumer preferences by itsintroduction of such items as the personal computer, video games, andvideo recorders.
Over time, a number of products are modified,expanded, or improved, depending on the demand of consumers, and thesechanges influence subsequent purchasing decisions of consumers. Finally, a more subtle phenomenon which contributes to changes inthe relative importance of items in the market basket is that tastes ofconsumers change. There are a variety of ways in which lifestyles andtastes change, such as the increasing number of persons who are activein some form of physical exercise such as jogging, cycling, or using thefacilities of a physical fitness organization.
These preference shiftsalso change expenditure patterns for items such as sports clothing andequipment, and fees paid for recreational facilities. Population changes. Not only do the consumption patterns ofindividual consumer units change over time, but also the geographicdistribution of the population may change. Between 1970 and 1980, thetotal population of the United States grew 11.5 percent. The populationof the South grew 20 percent and the population of the West increased 25percent.
4 This means that consumer units in the South and Westrepresent approximately 52 percent of the population for which therevised CPI market basket will be based, compared with 48 percent in the1972-73 market basket currently being used. Thus, consumption patternsof consumers in the warmer climates of the South and West will have agreater influence on the CPI than before. Prior revisions The first major activity in prior revisions of the CPI has been theimplementation of a Consumer Expenditure Survey as the basis forselecting and weighting a new market basket of goods and services to bepriced. Until these data are in hand, it is impossible to complete arevision of the CPI. The periods when expenditures were collected thatwere the basis for the last four revisions are as follows: Reference year(s) ; Release of revised CPI 1934-36 1940 1950 1953 1960-61 1964 1972-73 1978 The time between the reference years of the Consumer ExpenditureSurvey and the introduction of the CPI with revised expenditure weightswas typically 3 years, except for the 1978 revision. The 1940 revision introduced the concept of a sample of cities anditems and the principal of imputation, permitting the CPI to representprice change in all cities and all items purchased for consumption.
Prior to 1940, the CPI measured the price change in only the 33 citiesbeing surveyed and for only the items actually priced. Prior to the 1950 Consumer Expenditure Survey (on which the 1953revision was based), BLS conducted experimental surveys and testpricings to improve data collection methods and to establish the basicprocedures for processing these data. The 1953 revision took 3 years toimplement surveys which revised the areas and weights, and updated theitem samples priced.
This effort was primarily a clerical operation. After the 1953 revision, it became apparent that the CPI should berevised every decade. By the late 1950’s, dramatic changes hadoccurred: The composition of the urban population changed, with rapidgrowth of suburban areas, increased use of the automobile affectedlifestyles, and new shopping centers catered to the American consumer.A contributing factor to this growth was the 37-percent increase inpersonal disposable income between 1950 and 1956, with more thantwo-thirds of the rise being reflected in real income. The BLSreceived, in mid-1959, authorization for a revision program, which wascompleted in 1964 with the release of an index with revised weights andoutlet samples which included, for the first time, areas outside thecentral city of metropolitan areas.5 The first year of the 1964 revision was dedicated to pilot surveysfor testing and debugging procedures to be used nationwide.
Afterclerical edits and professional reviews of the data, the computer wasused to process estimates of expenditures and indexes. The 1978 CPI revision took longer than the previous revisionsbecause it included the introduction of new approaches to the collectionof consumer expenditures and a number of complex improvements andinnovations in pricing for the CPI. A thorough examination of the CPI,its concepts and operational processes, was made during the revision.The growth of computer applications during the decade of the 1960’smade it possible to introduce statistical techniques and monthlyoperational processes which were not feasible in earlier efforts ofproducing estimates of monthly price change. Innovations of the 1978 revision Innovations in collecting expenditure data for the 1978 revisioncontributed to a longer time between the Consumer Expenditure Surveyreference data and the introduction of the revised CPI. Prior to the1972-73 Consumer Expenditure Survey, interviewers visited all samplehouseholds during February through June and asked the respondents questions needed to reconstruct their living expenses for the previouscalendar year.
These global estimates of expenditures were used toobtain annual expenditures for most items. Respondents were asked torecall weekly expenditures for food store items and small frequentlypurchased items.6 Several changes in these procedures were made early in 1972. Aquarterly interview survey for a sample of consurmer units wasintroduced. Expenditures for a number of items were collected forpurchases made throughout the preceding 3 months, while other items weresurveyed for varying reference and recall periods. Another separatesample of consumer units was asked to keep two 1-week diaries in whicheach purchase was recorded on the day it was made.
Although this changein methodology was more expensive and took somewhat longer to process,it resulted in a marked improvement in the data used in the estimation of expenditure weights. It reduced the length of recall in thecollection of data, and, therefore, reduced response errors associatedwith either telescoping purchases from an earlier period or forgettingcertain purchases. (Telescoping occurs when the respondent inadvertentlyrecalls and reports a purchase made prior to the survey period.) Another significant innovation in the 1978 revision was theintroduction of the Point-of-Purchase Survey.7 In earlier revisions,the BLS had to rely on secondary data to establish sampling frames usedin selecting outlets in which to price items comprising the marketbasket. These secondary data provided only the broadest classificationof the outlet and provided no detail on the merchandise lines actuallypurchased. For example, it was not possible to identify all the typesof outlets where motor oil was sold, and it was impossible to tellwhether a particular grocery store sold fresh fish.
As a result,despite substantial efforts, it was impossible to obtain a statisticalsample of outlets for the CPI that represented where people shopped.The growth of metropolitan areas and the spread of shopping centersadded to the concern about the quality of outlet samples. In the Point-of-Purchase Survey, consumer units were interviewed ineach local area where prices were to be collected for the CPI.Respondents specified the amount they actually spent in each outlet inwhich they shopped for a category of items. Each category wasstructured to be compatible with a major line of goods or services sold,and so that the category would contain one or more “entry levelitems’8 for which a relatively broad class of products or servicescould be priced to represent that entry level item. (The current CPImarket basket contains 382 such items.) The Point-of-Purchase Surveyrespondents were asked if they purchased an item within a specificcategory during a prescribed reference period.
If a purchase was made,the name and address of the outlet was recorded along with the cost ofeach transaction. Prior to the introduction of the Point-of-Purchase Survey, eachoutlet was selected and weighted without specific regard for therelative sales that the outlet had for the priced item. The onlyexception to this procedure was in the grocery store food index wheresales data were obtained from food chain organizations so thatdifferential weights could be used to weight prices in the food indexfor large food chains. Since 1978, the BLS has used a probabilityprocedure with the value of purchases of each outlet as a measure ofsize to select outlets for each Point-of-Purchase Survey category. Thisensures that the outlet sample has an unbiased representation of largeand small establishments and also allows for the estimation of variancesand sampling error.
In addition, prior to 1978, there was no systematic statisticalprocess for replacement of outlets which closed, moved, or changedmerchandise lines. BLS had to rely primarily on its CPI fieldrepresentatives to locate a comparable establishment to obtain pricequotes for the specific items to be priced. With the composition ofoutlets gradually changing due to the entry of new establishments, itwas difficult to ensure the representativeness of the sample of outlets.In 1978, BLS introduced a new system, based on data from the continuingPoint-of-Purchase Survey, to update CPI outlet samples in each urbanarea on a 5-year cycle. Outlet samples in about 20 percent of the urbanareas priced for the index are updated each year so that the entireoutlet sample is completely updated over a period of 5 years. When substituting a price quote for an outlet item selected fromthe updated sample for a corresponding outlet item quote previouslypriced, it is necessary to factor out of the index measure anydifference between the two prices which results from this substitution.For example, if a man’s 100-percent cotton dress shirt is selectedin the newly selected outlet to replace a cotton blend shirt which waspriced in the outlet to be replaced, the prices of these two items wouldnot be viewed as comparable for an index measuring price change.
Differences that may exist between the new outlet item quote and the oldone are factored out of the measurement of price change by a methoddescribed as linking with an overlap price. This linking method used inoutlet sample updating requires that both the new and old outlet itemquotes be priced in the same month. The price for the item quote or theoutlet being replaced measures any price change from the previous indexmonth up to the link month when both outlets were priced.
The price ofthe item quote from the newly selected outlet is used to measure pricechange from the link month forward. This linking method assures thatthe process of introducing the new item has no effect on the index. Prior to 1978, an item designated for pricing in an outlet wouldhave characteristics specified by commodity analysts in the nationaloffice. The detailed specification was usually the same for all outletsin the country and would generally limit the number of products thatcould be priced to represent the expenditures within the item class.
Inthe 1978 revision, the BLS introduced new sampling procedures to permitall products or services within a respective item classification to beeligible for pricing, thereby increasing the efficiency andrepresentativeness of the index. However, once an outlet item isselected, the field representative records the specific narrowcharacteristics of the item to identify it for continuous pricing aslong as the item is available in the store. The process used in the selection of an item within an outlet iscalled disaggregation. This process gives an opportunity for everyvariety of an item within a store to be selected to represent purchasesfor the whole item class. This disaggregation is an objective andefficient process which results in the selection of a sample ofvarieties that covers the full spectrum of purchased items.9 During the 1978 revision, a great amount of time was spentexamining alternative methods of measuring price change inhomeownership. This effort resulted in the definition of aflow-of-services approach10 which is consistent with the economicconcepts on which the CPI is based.
This approach was not introducedduring the 1978 revision because of difficulties in developing aworkable flow-of-services measure and because of the diversity of viewsheld by various advisory groups.11 Shortly after the revision, concern over the measurement ofhomeownership costs increased because of the major changes that wereoccurring in the financing of homes and the increasing difficulties ofobtaining adequate house price data. Because of these changes and theincreasing impact and importance of the CPI, BLS changed thehomeownership component of the index between revisions.12 A rentalequivalence measure13 was introduced as the measure of homeowner cost inthe January 1983 CPI-U index and in the CIP-W with the January 1985index. The rental equivalence measure estimates the change in sheltercosts as the change in rents which would have to be paid for occupancyof housing units occupied by owners. This new measure replaced theprevious treatment in which homeownership costs were measured by currenthouse prices, mortgage interest, costs of new mortgages, property taxes,property insurance, and maintenance and repair costs. Because itmeasures the cost of consuming shelter services provided by a house(that is, the rent that would be paid), rental equivalence is consistentwith the underlying concept of the CPI as a measure of price change forconsumption.
The old homeownership measure included investment aspectsof homeownership associated with obtaining and maintaining the house asan asset. The 1987 revision The CPI relates to expenditures of the civilian noninstitutionalurban population of the United States. The urban population is definedas persons who live in Metropolitan Statistical Areas as defined by theOffice of Management and Budget (including the rural nonfarm withinthese areas) and urban areas, including places with 2,500 or morepersons outside of the Metropolitan Statistical Areas. Since the 1978 revision, the CPI has been calculated for twopopulations. The All Urban (CPI-U) index is based on expendituresreported by all consumer units in urban areas of the United States withtwo exceptions: consumer units on farms within urban areas and consumerunits receiving a majority of their income from a member who is in themilitary and lives off base with the unit. The CPI-U populationrepresented 81 percent of the total U.S. civilian noninstitutionalpopulation in 1981.
Because a large proportion of the population iscovered, this index is extremely useful in discerning the effect ofchanging prices on consumers. The CPI for Urban Wage Earners and Clerical Workers (CPI-W) isbased on urban consumer units who meet additional requirements relatedto their employment: more than one-half of the consumer unit’sincome has to be earned from clerical or wage occupations, and at leastone of the members had to be employed for 37 weeks or more in aneligible occupation. The CPI-W population was 30 percent of the totalU.
S. population in 1981. Geographic coverage. The first phase of a revision is to make anew selection of the geographic areas, or primary sampling units, inwhich price data collection will be done.
The new area sample for the1987 revision is based on the 1980 Census of Population and uses the newConsolidated Metropolitan Statistical Area definitions.14 The use ofthese definitions resulted in some changes. For instance, thedefinition for the New York area now includes Danbury and other parts ofConnecticut; Philadelphia includes Wilmington and Trenton; Los Angeles includes Riverside-San Bernardino; and San Francisco includes San Jose.The Metropolitan Statistical Areas which are not a part of aConsolidated Metropolitan Statistical Area were defined as individualprimary sampling units. All nonmetropolitan counties were grouped intoprimary sampling units to allow urban places with a population greaterthan 2,500 outside metropolitan areas an opportunity to be selected.
The overall primary sampling units design consisted of 278 metropolitanareas and 810 nonmetropolitan urban areas, which cover all urbanpopulation. Primary sampling units with at least 1.2 million personswere designated “certainty areas.’ This means that each ofthese areas represents itself in the weighting of the estimates to thetotal CPI population. The noncertainty selections have a populationweight that represents the population of all cities including their ownpopulation in their stratum–a collection of similarly sized areas inthe same general geographic region. Twenty-nine largest primarysampling units and two unique areas–Anchorage and Honolulu –weredesignated certainty areas.
15 The remaining primary sampling units were assigned to threecity-size classes–medium-sized cities, small-sized cities, andnonmetropolitan areas–within the Northeast, North Central, South, andWest regions.16 The result of the sampling process was the selection of39 new areas with the retention of 52 primary sampling units from theold sample, of which 30 were certainty selections in the new sample.Overall, the number of primary sampling units to be surveyed for the CPIhas increased by six. A comparison of primary sampling units in the oldand new samples by population size and region is shown in table 1. The South will have eight more primary sampling units than it hadpreviously. Despite the West’s large population growth between the1970 and 1980 censuses, it will still have the same number of primarysampling units; however, it will have two more certainty selections.Two reasons account for the unchanged overall number of primary samplingunits in the West. First, additional primary sampling units wereallocated disproportionately to the West in previous allocations topermit publication of a separate nonmetropolitan urban index for theregion.
Second, use of Consolidated Metropolitan Statistical Areasresulted in two certainty selections, Los Angeles and San Francisco,becoming substantially greater in population. Because of their largerpopulations in the new CPI design, the number of items and outletspriced in each of these two areas will be expanded.17 Allocating samples to produce the most accurate national CPIpossible with the funds available will affect the frequency ofpublishing CPI’s for 13 local areas. Beginning with the January1987 CPI, a monthly index will continue to be published for only thefour largest local areas–New York, Los Angeles, Chicago, andPhiladelphia. The index for Detroit, the smallest of the areas nowpublished monthly, will be compiled on a bimonthly basis only for evenunmbered months. Bimonthly indexes will be published for each of thenext 10 largest areas.
Bimonthly indexes which are now published forthe 12 smaller local areas will be replaced by semiannual averageindexes, and the index for Northeast Pennsylvania (Scranton) will bediscontinued.18 Expenditure weights. The relative weight of each entry level itemin the CPI is tabulated from data obtained by the Consumer ExpenditureSurvey. This survey is actually composed of two separate surveys–aninterview survey and a diary survey–both conducted by the Bureau of theCensus for the BLS. As in the 1972-73 interview survey, Bureau representatives collectdata for expenditures which respondents can remember fairly accuratelyfor periods of approximately 3 months.
Each consumer unit designatedfor sampling is contacted each quarter for five consecutive quarters.The initial contact is used to collect socioeconomic characteristics ofthe unit–an inventory of properties, vehicles, major durable goods, andinsurance policies. In addition, purchases of goods and services made inthe past month are recorded together with a date of purchase and adescription of each item.19 BLS uses only the second, third, fourth, and fifth interviews inestimating a 12-month consumption pattern for the consumer unitssurveyed. The current interview questionnaire differs from that used inthe 1972-73 survey in that it has a uniform reference period of 3 monthsfor each expenditure item, whereas the 1972-73 questionnaire allowed forvariable lengths of the period of recall. The major advantage of auniform reference period is that it permits each interview to be used ina quarterly estimate, even when a consumer unit was not interviewed forthe full 12 months of consumption. All data collected from consumerunits are used, in contrast to 1972-73 when data from consumer units wholater moved were not used. The uniform reference period facilitates rotating the sample.
Eachquarter, one-fifth of the consumer units are interviewed for the firsttime, an additional one-fifth for the second time, and so on. Therotation spreads over the calendar year any bias which may result fromeither conditioning or fatigue on the part of respondents as theyprogress from the first to the fifth interview. Because many expenditureitems are seasonal, it is advantageous to have a mixture of interviewsin each quarterly estimate of consumption patterns.
The purpose of the diary survey is to obtain expenditureinformation for small frequently purchased items which consumers tend toforget. Each selected sample unit is asked to keep 1-week diaries ofexpenditures for 2 consecutive weeks. The diary sample is spread amongthe 52 weeks of the year. However, the sample size is doubled in thelast 6 weeks of the year to obtain better estimates of items purchasedseasonally.
The interviewer, when placing the first-week diary, obtainsthe socioeconomic characteristics of the consumer unit and providesinstructions to the respondent. The respondent records purchases madeby any member of the unit during the week. (This eliminates anyquestions the respondent might have in determining if an item is withinthe scope of the survey.) The diary focuses on the recording ofpurchases made in grocery stores and of meals, snacks, and beveragespurchased in restaurants or other eating places.
Other purchases arealso recorded; therefore, a number of items reported in the interviewsurvey can also be recorded in the diary. A major difference in the twosurveys is that the diary does not record expenses made while out oftown on trips. Both surveys have a sample size of approximately 4,800consumer units per year.
However, in the interview survey, each unitcan potentially provide four quarters of data, whereas in the diary only2 weeks of data can be obtained from the same unit. The BLS staff has to identify from which survey–interview ordiary–estimates should be used in developing expenditure weights andselecting item samples. For many items, the design of each surveypredetermines which data should be used. For example, the diaryestimates are used for all individual food and beverage items becausethe interview survey only collects a total estimate of expenses forthese items. The diary is also used for a number of small andfrequently purchased items in the categories of personal care, householdsupplies, and nonprescription drugs and supplies which are not covered in the interview survey. For other expenses, the interview survey isthe better source as it has an effective sample size of 4,800 units eachquarter and expenses are recalled for a period of 3 months. The diarypanel, in contrast, only has an effective sample of 1,200 units perquarter for a total of approximately 2,400 diary weeks. There are a fewexpenditures that are collected in both surveys for which an evaluationis necessary to determine which estimate is best.
For example, gasolinepurchases are a frequently reported entry in the diary, and the estimateobtained from the interview is based on an average monthly expensepattern. Also, small clothing items such as hosiery and accessoriescould be overlooked in the 3-month recall which is the heart of theinterview survey, but are likely to be recorded in the diary. Each expenditure reported in these two surveys is coded to one ofthe 382 entry level items which constitute the lowest level of the CPIclassification structure. The highest level of the CPI structureconsists of the seven major groups of expenditures: (1) food andbeverages, (2) housing, (3) apparel and upkeep, (4) transportation, (5)medical care, (6) entertainment, and (7) other goods and services.Expenditures within a major group are divided into expenditure classeswhich have been established either by categories of commodities orservices and with some regard to similarity in their characteristics.The CPI structure currently has 68 expenditure classes and a new onewill be established in this revision for electronic products coveringpersonal computers, computer software, calculators, telephones and otherinformation processing equipment.
(See exhibit 1.) Most of the expenditure classes are divided into two or morestrata.20 The stratum is the lowest level for which expenditure weightsare calculated, and thus, the level at which the priced market basket isfixed between revisions. Because the allocation of the sample of quotesand outlets is also done at the item stratum level, the unmber of stratawithin an expenditure class generally has some overall relationship tothe relative importances of expenditures in that class. The variancesof the CPI can be greatly influenced by the way price quotations areallocated among the item strata. In this revision, a paramountconsideration was to maximize the efficiencies that could be achievedthrough sample designs and the allocation of samples. Using data fromthe 1980-81 CES and preliminary data on variances, item strata wererestructured so that, given the available resources for pricing, thevariance of the All Items CPI would be a minimum. A very few selectedstrata were left unchanged because of their individual uses or interest.
The number of strata for which expenditure weights are calculated willdrop from 265 to 203.(21) (See exhibit 1.) Each item stratum has at least one entry level item which isusually structured to facilitate the selection of a unique item to bepriced. If there is much heterogeneity among the goods or serviceswhich comprise an item stratum or in the types of outlets where they arepurchased, the stratum is usually subdivided into two or more entrylevel items. Currently, there are 382 entry level items and althoughthe composition of several will be changed in this revision, the totalnumber will probably not change by much. In the past, there have been a few sample entry level items whichhave not been priced. Sometimes the item was difficult to price becauseits quality changed constantly.
An example would be the pricing ofbooks purchased through book clubs. The book offered variessubstantially over time and various discounts or premiums may be earned.Also, an entry level item may not have been priced because anappropriate outlet sample could not be established. This is the caseparticularly for services provided by household workers and babysitters.If an entry level item or a potential one has a small relativeimportance, the Bureau does not go to a great disproportionate expenseto price it. In the current revision, the Bureau plans to use therelative importances of entry level items reported in the interview anddiary surveys to identify those that have become more significant sincethe last revision. As a result, unpriced strata are expected tocomprise only 1.5 percent of the CPI, compared with 3.
7 percentcurrently. Outlet selection. The 1987 revision will rely primarily on thecontinuing Point-of-Purchase Survey for the selection of outlet samples.When this survey was initially designed in 1974, there was some concernthat it would not be useful in selecting outlets for entry level itemswhich were purchased either infrequently or by a relatively smallpercentage of consumers. In updating outlet samples over recent years,a number of these entry level items have been added to thePoint-of-Purchase Survey. By extending the reference period for suchitems, the continuing Point-of-Purchase Survey has proved effective insecuring a sufficient outlet sample. There are a few entry level items for which outlet samples areobtained from sources other than the Point-of-Purchase Survey.
Generally, these items are found in a relatively small number ofestablishments, and reliable information is readily available forestablishing a measure of size in the sampling frame. Examples of suchentry level items are natural gas, electricity, basic telephone service,casualty insurance premiums, postage rates, and train fares. Theongoing Consumer Expenditure Survey is collecting outlet information(along with the expenditure data) for a small number of these entrylevel items. After these data are evaluated, we will determine if it ispossible to use the Consumer Expenditure Survey for selecting outletsamples for such entry level items as electricity, natural gas, andtuition. Data collected in the Consumer Expenditure Survey with regardto consumption quantities on utility bills will be used for selectingthe consumption amounts to be priced for the CPI. New Strategy: “rolling-in’ samples In previous CPI revisions, a new area sample (primary samplingunit) and new item and outlet samples were introduced at the same time.The 1987 revision will use a concept of rolling-in the new area, item,and outlet samples. That is, the composition of the area and itemsamples will be gradually updated over a period of years, rather thansubstituting the full set of new area and outlet samples at a singletime.
Two innovations of the 1978 revision facilitate this rolling-instrategy: the use of the continuing Point-of-Purchase Survey for asystematic updating of outlet samples, and the broader definition of thecharacteristics of items which define strata. The first stage ofrolling-in the new sample is to initiate pricing in new areas which willbe needed in January 1987 for updating the U.S.
CPI to reflect changesin population distributions. A number of the areas which had beenrepresentative of a specific city-size had sufficient population growthbetween 1970 and 1980 so that they no longer represented that particularcity-size. There are 19 new areas classified as either small- ormedium-sized or nonmetropolitan, and one large-sized area which have tobe surveyed prior to 1987 in order for the U.S. CPI to reflect thedistribution of the U.
S. population as enumerated by the 1980 census. The second aspect of this phased update pertains to the itemsamples in all CPI areas retained in the new design. Any new entrylevel item or any entry level item that is substantially modified indefinition will be initiated in all areas prior to the issuance of therevised CPI for January 1987. The continuing Point-of-Purchase Survey for 1985 will be conductedin the 20 new areas so the item and outlet samples for these areas canbe initiated and results introduced in the January 1987 CPI. (Anadditional 19 new areas will be initiated and introduced over the1987-89 period.
) The new item expenditure weights tabulated from the 1982-84Consumer Expenditure Surveys will replace those tabulated from 1972-73survey data. To make this substitution of expenditure weights withoutcausing a discontinuity in the CPI’s measurement of price change,the index levels using the new expenditure weights will be set equal tothose published for the old series in December 1986. The official CPIfor January 1987, therefore, will reflect the price change betweenDecember and January based on the new expenditure weights.
As in thepast, the Bureau will continue to publish overlap indexes using the oldexpenditure weights for 6 months after the issuance of the revised CPI,for the convenience of users. Outlet samples for entry level items retained from the old primarysampling unit design will be updated through the use of the continuingPoint-of-Purchase Survey and the existing outlet updating procedures. Afew of the retained areas will have their outlet samples updated in 1987when 10 more new areas are rolled-in.
The remaining areas will haveitem and outlet samples rolled-in over the next years. Advantages. By rolling-in the new areas and using the establishedoutlet updating process for areas retained in the CPI design, it ispossible to effect significant time and cost savings. One of the mostcostly activities of past revisions was the initiation of pricing of theitem and the outlet samples in all areas selected in the primarysampling unit redesign. Prior to the introduction of the revised CPI,all of the item and outlet samples had to be initiated and priced in thesame month as the existing samples. Even for the areas retained, areselection of item and outlet samples required substantially more newpricing because the probability of reselecting the same outlet for anentry level item is very small.
Additional field representatives had tobe hird and trained to do this work while pricing was continued toproduce the ongoing CPI. Because the existing CPI is official until the revised index isreleased, the review and processing of data from the new samples must bedone in a framework which does not jeopardize production schedules.Rolling-in the new areas into the CPI estimate over a 3-year periodallows more time to train field representatives and lessens problemsassociated with a rapid expansion and subsequent reduction in staff.More importantly, using the existing updating procedures for introducingnew outlet samples on a systematic basis precludes the need to maintainextended dual operations– one for the existing CPI and one for the datascheduled to supersede it. Over the past 6 years, the Bureau has usedthis technique for updating outlet samples.
A few modifications toaccommodate new areas and entry level items will increase the amount ofdata requiring processing, but by substantially less than the oldprocedure. Expenditure weights for the 203 strata in the CPI market basketwill be tabulated using 3 years of Consumer Expenditure Surveydata–1982, 1983, and 1984. Because the CPI is a base-weighted indexdesigned to reflect price change (and not changes in the quantitiespurchased), these expenditure weights will remain fixed until the nextrevision of the CPI. As in the past, of course, BLS will continue toupdate the outlet sample in one-fifth of the CPI areas each year. Within the CPI fixed-weight constraint, however, BLS intends tomake maximum use of data from the ongoing Consumer Expenditure Survey tokeep the items priced to represent the strata up to date. A number ofCPI strata, for example, are represented by 2 or more entry level items.The sample of entry level items for these multiple-entry level itemstrata have been selected from the Consumer Expenditure Survey.
Beginning in 1987, when the outlet samples are updated forone-fifth of the urban areas and new detailed items are selected forpricing, this sample of entry level items will also be updated based onthe two most recent years of Consumer Expenditure Survey data. Ifrelative shifts of consumption occur among items within a stratum or newproducts appear within the stratum, then entry level item reselectionwill gradually change the composition of the entry level items beingpriced. In other words, the entry level item sample will begin toreflect the changes consumers are making in the variety of productspurchased which make up an item stratum of he index.
The reselection ofthe item samples within each fixed-weight category for one-fifth of thearea sample does not alter the fixed-weight nature of the CPI becausethe population-expenditure weights will remain fixed, as now, at theitem strata level until the next revision. This reselection will notaffect entry level items which have a very large relative importance orare the only ones in the particular strata and, therefore, are certainto be priced in all urban areas. Although the CPI will continue to have its basic fixed-weightcharacter, the existence of annual expenditure data will offer a numberof opportunities for developing experimental indexes with differentcharacteristics. For example, while the expenditure weights for theofficial CPI are updated only about once every 10 years, experimentalindexes could be developed with more frequent weight changes.
Improvements of the 1987 revision Enhanced shelter survey. The adoption in 1983 of rentalequivalence to measure changes in the cost of the shelter component ofowner-occupied homes put the housing component of the CPI on aflow-of-services conceptual footing, and isolated the consumptionelement of owner housing from its investment element. In addition to updating the housing sample based on the 1980census, the 1987 CPI revision program will enhance the rentalequivalence method adopted in 1983.22 The selection of a new housingsample is designed to represent optimally both owners and renters. Amulti-stage sampling procedure was used that stratifies the residentialareas of each primary sampling unit by tenure (percent owner-occupied)and rent level. Smaller areas are then defined and sampled within eachselected area. The housing units of each selected small area arescreened for tenure and sampled at differential rates according totenure.
In heavily owner-occupied areas, for example the renters areselected more frequently than owners in order to find renters who arelike owners, because it is from these rentals that the best estimatescan be made in the implicit rent of owner-occupied dwellings. Enhancement of statistical techniques. Because the ConsumerExpenditure Survey estimates for each of the individual areas of thecountry are based on relatively small samples, BLS has undertakenresearch in statistical techniques to reduce the error on local areasindex weights. In the 1978 revision, a compositing technique was usedin which the local area average expenditures were weighted together withthe expenditure estimates for the same item class for the geographicregion to which the local area belongs. Research done at the BLS during the current revision involved useof the composite estimation of relative importances rather than of meanexpenditures of the item categories.
Relative importances in the CPIare the mean expenditures for each item as a percentage of allexpenditures. BLS statisticians found compositing of relativeimportances to be more effective in reducing the average mean squarederror than compositing of the expenditures themselves.23 Another refinement under consideration is to replace estimates foreach of the four broad geographic regions of the country (Northeast,North Central, South, West) with two sub-area estimates–one for thecertainty areas within each region and one for all other areas withineach region. The relative importances of each certainty area withineach region would be estimated based on composites which use relativeimportances of expenditure patterns from all certainty areas in theregion. The division of the regional estimate between certainty areasand all other areas has also proved effective in reducing the averagemean squared error. Publication of quantitative measures of sampling error for selectedindexes is planned for the 1987 revision. Initially, estimates of theindex variance will be available in the All Items CPI and for some ofthe major group indexes.
Eventually, more indexes will have an estimateof variance published. Enhanced quality. During the implementation of the 1987 revision,the Bureau will add a new dimension to quality assurance and control ofthe CPI program. Throughout the years, the staff has devotedsubstantial time to the inspection of data collection and processingactivities. The goal of the inspections was to identify and correctindividual error. The goal of the new audit process to be instituted inthis revision will be to achieve long-term quality improvement. Thiswill be accomplished, in part, by an independent staff which willsystematically evaluate survey processes empirically.
By having independent audit data for comparison purposes, errorprofiles can be used to identify the type of errors, diagnose theirsources, and prescribe procedural changes to prevent these errors fromoccurring in the first place. The techniques used will include special,detailed evaluation studies of specific processes, ongoing processcontrols and reports, statistical quality control and measurement, and asystem for information feedback and corrective action. The goal is todevelop processes that will result in enhanced estimators of pricechange.
Other concepts to be investigated As part of the revision, BLS will investigate the appropriatetreatment of insurance premiums in the CPI. Currently, premium costsfor health insurance and casualty insurance for vehicles and householdfurnishings are priced for the index. The overriding issue in thepricing is the one of constant quality in the coverage. Quality changesthat affect premium level should be removed before being used in theCPI. Using health insurance as an illustration, there are four factorswhich affect changes in premiums: (1) changes in the costs of medicalprocedures, (2) administrative cost and surplus requirements and theprofit needs of commercial carriers, (3) policy benefit changes, and (4)utilization changes, that is, changes in the frequency of a coveredevent occurring. Changes in the first two factors do not affect policyquality, whereas changes in the latter two will.
For the past 20 years,the Bureau has used an indirect method of pricing health insurancebecause it has been unable to develop an effective methodology forremoving the effect of most changes in the coverage or the utilizationrate.24 The indirect method of pricing health insurance measures changesin medical costs (factor 1) by using the price changes which haveoccurred in physicians’ and hospital fees in the CPI to representthe change in costs that insurance carriers have incurred for theirpolicy holders. Changes in costs for carriers (factor 2) are measuredby the annual changes in the retained earnings (premium revenue lessbenefit payments) of insurance carriers.
Thus, the indirect methodmeasures changes which affect policy premiums while excluding from themeasurement the two factors which affect quality. Direct pricing of a sample of policies was tried during the 1978revision, but was dropped due to the unresolved issue of qualityadjustment. BLS was unable to measure satisfactorily the premium valuefor changes in the coverage of the policies and for the impact ofchanges in the utilization of policies. For the 1987 revision, researchis continuing to determine if a procedure can be adopted that producesadequate direct adjustment for changes in coverage and utilization. Casualty insurance on vehicles and household effects is directlypriced in the CPI. Factors for removing quality changes from thesekinds of policies were developed for changes in deductible provisionsand for mandatory “no fault’ automobile insurance. Generally,other policy coverage changes are treated by not using the policy andits premium in the index calculation for the month of the change.
Withcasualty insurance, however, price changes which result from changes inutilization rates are usually reflected in the index. The difference inthe treatment of utilization changes for health insurance and casualtyinsurance is being reviewed as part of the revision. Evaluating substitute items One of the most difficult problems for those who compile priceindexes is that of quality change. Products and services changeconstantly, and new items replace old ones on the market.
There is alarge body of literature on the effect of quality change on ConsumerPrice Indexes.25 Most of these studies show mixed results. Although itis generally agreed that quality adjustment error exists, the extent ofthe error, and, indeed, even its direction, are not known. A series of practical techniques for handling substitution andquality change issues in an operating environment has been developed.Briefly, if an item and its substitute are comparable, with nosignificant difference in quality, then the prices are directly comparedand used in the index. If the items are judged not comparable, then theprice difference must be broken down into quality change and pricechange.
This process results in one of three actions: (1) a qualityadjustment is made by using the difference in production costs andadding a markup to retail or by some other method of valuing thedifference in characteristics, (2) if both the old and new items can bepriced in the current period, the difference in price in this period isconsidered the value of quality change (this “overlap pricing’is the technique used in outlet updating), or (3) if neither a qualityadjustment nor an overlap price is possible, then the price change ofthe new item is not used in the current estimate, and a current pricefor the old item is imputed using price movements of the quotes withcomparable prices in both periods in the item strata or market basket.This third action (referred to as linking) not only precludes a qualitychange from being reflected in the index, but can also preclude capturing the price change–either positive or negative– which may haveoccurred at the time of the substitution in the specific item. Of the more than 1 million distinct price quotes obtained for itemsother than shelter in the index during 1983, only 3.
8 percent weresubstitutions. But this relatively low frequency of substitution stillhad a major impact on the CPI. Price changes associated with thesubstitutions accounted for more than one-half of the total price changein the year, and quality changes equal to about one-third of the totalprice change were excluded from the index. More than 40 percent ofthese substitutions were comparable, and an additional 8 percent wereadjusted explicitly for quality changes. An additional 45 percent ofthe substitutions were judged noncomparable and linked, while an overlapprice was obtained about 6 percent of the time.
The highestsubstitution rate (17.3 percent) was for apparel and upkeep items.26 In cases where noncomparable substitutes are “linked’ outof the index, there is a danger that the CPI is missing some real pricechange. The converse danger of including some quality change in theindex also occurs when two versions of an item are declared comparable.Because of the significant impact such substitutions have on the index,research is under way to identify methods to reduce the risks associatedwith missing price change by linking and with reflecting quality changeas price change when declaring substitutes comparable. As the revision progresses, detailed reports will be prepared onthe results of specific investigations and research.
FOOTNOTES 1 A consumer unit is comprised of either all members of aparticular household who are related by blood, marriage, adoption, orother legal arrangements such as a foster child; a person living aloneor sharing a household with others or living as a roomer in a privatehome or lodging house or in permanent living quarters in a hotel ormotel, but who is financially independent; or two or more personsliving together who pool their income to make joint expendituredecisions. 2 See Current Population Reports, Consumer Income, Series P-60,No. 142 (Washington, Bureau of the Census, February 1984), tables 3,14, and 38. 3 Elizabeth Waldman, “Labor force statistics from a familyperspective,’ Monthly Labor Review, December 1983, pp.
16-20. 4 Derived from table 52, Persons by Race for Regions: 1980 and1970, United States Summary, General Population Characteristics,PC80-1-81 (Washington, Bureau of the Census, May 1983). 5 For a detailed description of the 1964 CPI revision, see TheConsumer Price Index: History and Techniques, Bulletin 1517 (Bureau ofLabor Statistics, 1966).
6 For a detailed description of the 1960-61 survey, see ConsumerExpenditures and Income: Survey Guidelines, Bulletin 1694 (Bureau ofLabor Statistics, 1971). 7 The Point-of-Purchase Survey is a household survey conducted bythe Bureau of the Census each year in one-fifth of the areas sampled byBLS for the CPI. The survey is designed to periodically update theoutlet sample used for pricing various items. Approximately 4,000households are contacted each year and asked to provide data on names ofretail, wholesale, or service establishments for purchases of 156categories of goods and services. 8 An entry level item is the ultimate sampling unit forexpenditure items selected from the Consumer Expenditure Surveys by theWashington office. Each entry level item establishes the definition tobe used by data collectors in the identification of unique items withinan outlet that can be selected for pricing an entry level item.
9 For further elaborations of the CPI methodologies, see BLSHandbook of Methods, Volume II, The Consumer Price Index, Bulletin 2134(Bureau of Labor Statistics, 1984). 10 The flow-of-services approach measures the cost of consumingshelter services provided by a house. The approach focuses onconsumption and abstracts from the investment aspects of home purchasedecisions. See the following Monthly Labor Review articles: RobertGillingham, “Estimating the user cost of owner-occupiedhousing,’ February 1980, pp.
31-35; and Robert Gillingham andWalter Lane, “Changing the treatment of shelter costs forhomeowners in the CPI,’ June 1982, pp. 9-14. 11 For more information, see “Changing the HomeownershipComponent of the Consumer Price Index to Rental Equivalency,’ CPIDetailed Report, January 1983, pp. 7-13. 12 Janet L. Norwood, “Statement Regarding Changes in theConsumer Price Index,’ USDL News Release, 81-506, Oct. 27, 1981.
This release explains reasons for introducing rental equivalence betweenrevisions. See also “Changing the treatment of shelter costs’and “Changing the Homeownership.’ 13 The rental equivalence approach as incorporated into the CPIattempts to answer the following question: How much rental income dothe owners of housing units forego when they choose to occupy the unitsthemselves instead of renting them out? 14 Consolidated Metropolitan Statistical Area is an area which hasmore than 1 million population and is contiguous to one or more primarymetropolitan statistical areas. 15 Anchorage and Honolulu have been designated certainty areassince 1964, shortly after these territories were legislated tostatehood. They are great distances from the areas comprising the Westregion so it is unlikely that a population market basket of other areaswould provide a good representation of them. 16 In order to allocate the remaining primary sampling units toeach size class of the design as proportionally as possible to its shareof the urban population, the population demarcation between medium-sizedcities and small-sized cities varies by region–from 330,000 in the Westto 500,000 in the Northeast. Further, proportional allocation willpreclude the publication of nonmetropolitan urban areas in the Northeastand West as a minimum of four primary sampling units are required, andthese regions received only two.
When selecting the sample of primary sampling units, majorconsiderations are the costs of hiring and training field staff in newareas as well as the requirements related to the linking of CPI regioncity-size indexes. Thus, the BLS uses a statistical procedure whichmaximizes the probability of retaining primary sampling units from theold design. The goal of this procedure is to increase the number of primarysampling units overlapped between the two designs, compared to anindependent selection of primary sampling units, while at the same timereflecting the shifts in population of primary sampling units betweenthe censuses. The BLS also uses a controlled selection to ensure thatthe representation of the sample by State is directly proportional tothe population of the State. (See Cathryn S.
Dippo and Curtis A.Jacobs, “Area Sample Design for the Consumer Price Index,’1983 Proceedings, American Statistical Association.) 17 The budget of the CPI constrains the number of items andoutlets which can be priced. The item sample design developed in the1978 revision designates the number of price quotes which are requiredfor each item stratum in the CPI market basket.
Some variability in thenumber of quotes obtained occurs because of the greater relativeimportance of some items and differential allocation based on collectioncosts and variances of price change. The basic unit for allocating itemquotes among the primary sampling units selected for pricing is called ahalfsample. Each halfsample has approximately 1,100 quotes and iscalled a halfsample because at least two are required to calculate aprice index for a specific CPI market basket area. The proposed budgetfor maintaining the CPI after the 1987 revision supports 127halfsamples. When allocating the 127 halfsamples among the primary samplingunits of the new design, the primary objective was to make the sample asefficient as possible to minimize the sampling error of the nationalindex.
Each of the 91 primary sampling units was allocated onehalfsample. For the optimization of the design, a primary sampling unitshould only receive an additional halfsample if its population isgreater than 1/127 of the total population. The remaining halfsampleswere allocated among the 15 largest primary sampling units. By doingthis, the efficiency of the national CPI estimate was improved. Withother changes made by establishing population proportionality among theregion according to size of cities, and optimizing the sample allocationbetween major groups, the overall efficiency of the national index willbe improved by approximately 35 percent. However, the policy ofoptimization of the area design did have an impact on the publicationpolicy. 18 Bimonthly indexes will continue to be published for the localareas of Boston, Pittsburgh, Cleveland, St. Louis, Baltimore, Dallas,Houston, Miami, Washington, D.
C., and San Francisco. Semiannualaverages will be published for Buffalo, Cincinnati, Kansas City,Milwaukee, Minneapolis-St.
Paul, Atlanta, Anchorage, Denver, Honolulu,Portland, San Diego, and Seattle. 19 These data are used in a technique described as”bounding’ the reference period of the subsequent interview.Bounding minimizes response errors which may result from the respondentinadvertently duplicating purchases from an earlier period. Byrecording dates and descriptions of purchases for the preceding month ofeach visit, the technique can be repeated in each subsequent interview.
20 Item strata constitute the level of detail for calculating theexpenditure weights of the CPI market basket, and the qualities andimplicit quantities of this market basket are kept fixed betweenrevisions. That is, any change in the CPI from one month to another isthe effect of price changes of the item strata comprising the marketbasket. 21 The reduction in the number of strata will affect the number ofindexes that are currently published. The BLS will, however, produce anumber of substratum (entry level items) indexes for old item stratathat are now published and that have a significant number of pricequotations collected. 22 See Walter F. Lane and John P.
Sommers, “Improved Measuresof Shelter Costs,’ 1984 Proceedings, American StatisticalAssociation. 23 Michael P. Cohen and John P. Sommers, “Evaluation ofMethods of Composite Estimation of Cost Weights for the CPI,’ 1984Proceedings, American Statistical Association. 24 See Daniel H. Ginsburg, “Medical care services in theConsumer Price Index,’ Monthly Labor Review, August 1978, pp.35-39.
25 For a detailed discussion of this and other problems, see JanetL. Norwood, Problems in Measuring Consumer Prices, Report 697 (Bureau ofLabor Statistics, 1983) and Jack E. Triplett, “Quality Bias inPrice Indexes and New Methods of Quality Measurement,’ in ZviGriliches, ed., Price Indexes and Quality Change (Cambridge, MA, HarvardUniversity Press, 1971).
26 Paul A. Armknecht, “Quality Adjustments in the CPI andMethods to Improve It,’ 1984 Proceedings, American StatisticalAssociation. Table: 1. Current and new primary sampling units, by city-size andregion Table: Exhibit 1. Strata titles for revised Consumer Price Index