Testimony of Katharine G. Abraham Commissioner of Labor Statistics Before the Subcommittee on Human Resources House Committee on Government Reform and Oversight April 30, 1997

I appreciate the opportunity to testify today in response to the Subcommittee’s questions about the actions the Bureau of Labor Statistics (BLS) has taken and will be taking to improve the accuracy of the Consumer Price Index (CPI). To provide some context for the activities I will describe this morning, I also will discuss some of the measurement issues raised in the December 1996 report of the Senate Finance Committee’s Advisory Commission to Study the Consumer Price Index.

It is important to note at the outset that the BLS has a long tradition of being in the forefront of price measurement research and operational innovation. Attached to my statement is a document listing many of the improvements to the procedures used in constructing the CPI that the BLS has made over the years (Attachment A). I would particularly draw the Subcommittee’s attention to those improvements introduced over just the past two years, including steps taken in 1995 and 1996 to identify and resolve the so-called "formula bias" problem, and the introduction earlier this year of a new approach to the measurement of prices for hospital services. These refinements are indicative of the Bureau’s ongoing commitment to keeping its measures as accurate and up-to-date as possible.

The Bureau, in fact, has efforts currently underway and future activities planned that will further improve the accuracy of the index. Just this month we have begun issuing a new experimental measure that is designed to help us in identifying the best way to address the so-called "lower-level" substitution bias in the CPI (which I will describe at greater length a bit later in my testimony). I also am pleased that the President’s Fiscal Year (FY) 1998 budget includes a program increment that will allow us to take several additional steps toward increasing the accuracy of the CPI. The BLS will be requesting resources to speed up the process of updating the CPI market basket in future revisions. Resources to expand the collection of information on the prices and characteristics of certain goods and services, together with resources to be devoted to the early identification of new goods as they become available in the marketplace, also will be requested. This information will enable us to improve the methods we use to adjust for quality change and to insure that new items are brought into the index in a more timely fashion. Finally, the request provides for the production of supplementary measures of change in consumer prices that we believe would provide closer approximations to the change in the cost of living than the currently published CPI. At the appropriate points in my testimony, I will indicate the relationship between these current and planned activities and the issues raised in the Commission’s report.

The Advisory Commission’s report begins with one overarching recommendation: "The BLS should establish a cost of living index (COLI) as its objective in measuring consumer prices." This seems basically right to me. Indeed, the BLS long has said that it operates within a cost-of-living framework in producing the CPI. That framework has guided, and will continue to guide, operational decisions about the construction of the index. Putting things slightly differently, if the BLS staff or other technical experts knew how to produce a true cost of living index on a monthly production schedule, that would be what we would produce. I therefore have no fundamental disagreement with the Commission about what the objective of our CPI program ought to be, though we disagree to some extent about what changes to the index would be feasible and prudent and about the timetable on which those changes could be implemented.

More specifically, the Commission’s report focuses on two broad issues concerning the CPI as a proxy measuring changes in the cost of living of the U.S. consumer. The first is substitution bias, comprising what the Commission terms lower-level and upper-level components. The Commission believes that these components together impart an upward bias in the CPI of 0.4 percentage point per year. The second broad issue involves how best to treat changes in the quality of goods and services that consumers buy, changes in how and where those goods and services are sold, and the emergence of new goods and services. The Commission believes that failure to adjust adequately for these effects imparts a 0.7 percentage point per year upward bias to the CPI. The total overstatement of the change in the cost of living due to substitution bias and other problems together is judged by the Commission to amount to 1.1 percentage points per year.

Let me talk first about substitution bias. Like the Commission members, I also am an economist. Almost any economist would agree that an index such as the CPI that tracks the cost of purchasing a fixed market basket of goods and services represents an upper bound on the change in the cost of living. Indeed, for many years, the BLS has attempted to explain exactly this point.

Operationally, as the Commission suggests, substitution bias may show up at two levels. By way of background, the CPI is constructed by first aggregating the roughly 90,000 price quotations collected each month to form a series of subindexes for categories of items such as "Apples," "Men’s Shirts," and "Prescription Drugs," and then aggregating those subindexes to form the overall CPI. The formula used to aggregate the individual price quotations to form the subindexes does not account for consumers’ ability to substitute across items within item categories when the relative prices of those items change - for example, when the price of Delicious apples increases and the price of Granny Smith apples falls. Similarly, the formula used to aggregate the subindexes to form the overall CPI does not reflect the substitution across item categories that takes place when the relative prices of items in different categories change - for example, when the price of apples falls relative to the price of oranges. Were such substitution taken into account, the CPI undoubtedly would rise less rapidly.

To address the so-called lower-level substitution problem, the Commission has suggested adoption of an alternative formula for aggregating price quotations, one that has been under investigation by the BLS over the past several years. As noted above, the current CPI formula does not allow for the potential substitution among items within a category, such as between different varieties of apples, when the relative prices of those items change. The proposed alternative formula, termed the geometric mean formula, is based on a different assumption about consumers’ substitution behavior, namely that consumers substitute among items in such a way as to hold the share of their expenditures devoted to each item constant. Neither the assumption of no substitution underlying our current practice nor the assumption underlying the geometric mean formula is likely to provide a close approximation in all cases. It may be more plausible to assume that consumers substitute freely between types of apples or between brands of television sets when their relative prices change than to assume similar substitutability between types of prescription drugs or between electric power companies in an area. As I indicated earlier, the BLS has begun issuing a monthly experimental measure that is constructed using the geometric mean formula in all index components, and will make a decision by the end of this year as to which components of the official CPI should employ the geometric mean formula..

Upper-level substitution bias occurs because the formula currently used to aggregate CPI subindexes ignores the fact that consumers substitute across item categories when relative prices change. Here, however, the nature of the operational problem faced by the BLS is a bit different than that at the lower level of item aggregation. The detailed data needed to account for lower-level substitution in the calculation of CPI subindexes are simply not available. In contrast, at the upper level of item aggregation, the BLS does collect information on consumer expenditures across item categories, like apples, men’s shirts, and prescription drugs. Therefore, it is possible to construct a measure that accounts for substitution across those item categories in response to relative price changes, though not on the same schedule as the current CPI. The expenditure information required to construct such a measure -- one of the so-called superlative indexes -- is available only with a lag, so that the index cannot be produced until the fall following the year to which it applies. The BLS currently produces these measures on an experimental basis, and would be happy to produce them to a higher standard of precision and reliability. Thus, we are receptive to the spirit of the Commission’s recommendation that we produce an annual superlative index as a supplement to the official monthly CPI, and will be able to make substantial headway in this regard if we receive the FY 1998 program increase we are requesting.

Recognizing the unavoidable time lag in producing a true superlative measure, the Commission recommends that the BLS explore steps that might make the monthly CPI a better approximation to such an index. The Commission has suggested, for example, that updating the index’s expenditure weights on a continuous rather than a periodic basis and changing the formula for aggregating subindexes might make the CPI behave more like a superlative index. The BLS is, of course, open to exploring this sort of option, as can be seen in the variety of experimental indexes we have published for some time, and we will continue our work in this area. Adopting any option that has neither a sound theoretical foundation nor a clear empirical justification, however, would be a mistake. We can produce superlative measures, albeit with a lag, and thus convincingly deal with the "substitution bias" problem. I believe we would gain little, and possibly do much damage to the credibility of our statistical system, if we were to move hastily to adopt untested techniques for producing the official CPI.

Because it has received so much attention, I have spent a good deal of time talking about substitution bias. The largest share of the bias in the CPI that the Commission concludes exists -- 0.7 percentage point per year, or nearly two-thirds of the total of 1.1 percentage points per year -- arises from other sources. The Commission believes that the failure to make adequate adjustment for changes in the quality of the goods and services people buy and to account properly for the value to consumers of newly available goods, together with deficiencies in the way the CPI treats differences in the prices charged at different retail outlets, constitute a serious problem.

Before commenting on the evidence marshaled by the Commission in support of its conclusions in the quality/new goods area, I would like to note that the BLS already has procedures in place designed to account for changes in the quality of the items being priced. (It often mistakenly has been assumed, though not by the Commission, that BLS makes few or no such adjustments.) Although I would readily acknowledge that our adjustment procedures are not perfect, they do have a very important effect on the rate of price change the BLS reports. The best available information on this point applies to a CPI subindex covering roughly the commodities and services component of the market basket (about 70 percent of the total, with shelter the largest exclusion). During 1995, this subindex would have risen by 3.9 percentage points had these procedures not been applied. Because of their application, however, the subindex actually rose by only 2.2 percentage points over the year. Roughly speaking, these figures imply that the adjustments made by the BLS for changes in the quality of these goods and services amounted to 1.7 percentage points over the course of a single year. (These figures are somewhat different from those I reported in other testimony earlier this year. We have refined our prior estimates principally to exclude some "quality adjustments" that are made to account for changes in units of measurement or package size that might not normally be thought of as quality change.) I would add that the BLS also has established procedures for bringing new items and new outlets into the index. The expenditure share information used to aggregate the CPI subindexes is updated only once every ten years or so, but the specific stores in which prices are collected and the specific items priced are reselected on a five-year cycle. Although more frequent sample rotations undoubtedly would be desirable, it is a fact that a considerable share of the resources available for producing the CPI are devoted to ensuring that the sample of items priced is representative of what consumers actually are purchasing.

The Commission does not argue, of course, that the BLS is not making a good effort to address quality/new goods biases, but rather that, in spite of a good effort, residual bias remains. The report’s approach to assessing this residual bias is to divide the index into 27 categories, and then to make a judgment about the magnitude of the bias in each case. Unfortunately, the evidence applicable to many of these categories is rather sparse.

In some of the categories, absent evidence, the Commission is forced to fall back on its best judgment. The food and beverages categories are perhaps the best examples; the Commission’s estimates of upward biases in these categories rest exclusively on not implausible, but unsubstantiated, judgments regarding the value to consumers of increased variety on grocery and liquor store shelves, together with the value of greater choice in restaurants.

In other cases, members of the Commission have produced evidence that bears on the trend in prices for particular sorts of items. I cannot say, however, that this evidence always leads me to the same conclusions as those reached by the Commission. The Commission’s estimate that the growth in prices of new and used cars has been overstated by 0.6 percentage point per year in the recent past, for example, rests on data showing that the average age of cars on the road has risen, together with an assumption that current CPI procedures do not capture any of the increases in automobile durability that may have occurred. This latter assumption, however, is incorrect; attached to my testimony is a document listing some of the many durability-related model changes for which adjustments have been made in the CPI over the past few years (Attachment B).

The Commission’s estimate that the CPI has overstated the rate of growth of apparel prices by 1.0 percentage point per year since 1985, to take another example, rests on a comparison of the official CPI data with price indexes constructed using Sears catalogue prices for items remaining unchanged from one year to the next. Even beyond the reservations I have about drawing any general conclusions based upon the prices charged by a single catalogue merchant, I am skeptical of any index based only on the prices of unchanging items, particularly in a market segment where changing fashion is as important as it is in apparel.

On another note, I also would have found the report more persuasive had the Commission made a more systematic effort to explore the possible existence of negative biases in the CPI. Other analysts have hypothesized reduced convenience and comfort of air travel, and deteriorating quality of higher education, as examples of quality decreases that are ignored in the CPI. More generally, whereas the Commission notes some service quality improvements, such as the introduction of automatic credit-card readers at gasoline pumps, the BLS often hears complaints about broad-ranging declines in the quality of customer service, which are equally difficult to incorporate in the CPI.

A more subtle issue is that price increases for many goods occur intermittently and often are timed to coincide with model replacements or other quality improvements. The BLS commonly adjusts for quality differences between successive models by, in effect, treating the difference in price between them as wholly attributable to a difference in quality. There is a risk that this procedure over-adjusts for quality change, imparting a downward bias to the index. Methods have been introduced to try to minimize that possibility, but the Commission paid little attention to this potential problem.

Close to half of the quality/new goods bias the Commission believes exists in the overall CPI is judged to occur in just two areas of the index: medical care and high-tech consumer goods. These clearly are components of the index in which the BLS faces particularly difficult measurement problems, though I cannot say what the magnitude of any bias in these index components might be.

From a BLS perspective, the most important question about possible quality/new goods problems is what we might do to improve our procedures and ameliorate those problems. Recognizing the particular difficulties associated with measuring medical care prices and high-tech consumer goods prices, the BLS has devised and announced important improvements in our methods. These include changes in our hospital price measurement procedures, effective with the data for January of this year, and prospective changes in our sample rotation procedures that will allow us to update item samples in rapidly changing market segments more frequently than once every five years (at the cost of less frequent updates in more static market segments). In addition, the FY 1998 budget we are submitting would allow us to make important progress in the quality/new goods area, by supporting greater use of techniques that explicitly account for changes in the characteristics of items being purchased and implementation of more aggressive procedures for identifying and beginning to price new goods promptly once they appear in the marketplace.

The Commission’s report also discusses the question of new outlet bias, namely, how changes in the mix of retail outlets at which consumers shop ought to be treated. Current CPI procedures treat purchases of a particular item at different retail outlets as distinct transactions; the prices at the different stores are never directly compared. This could impart an upward bias to the CPI if, for example, stores offering lower prices but comparable service gained in market share. As a practical matter, however, measurement of any such bias is complicated by the fact that different types of outlets commonly offer quite different shopping environments. Research on the factors affecting consumers’ choices about where to shop ultimately may be helpful in devising appropriate procedures for dealing with changes in outlet mix.

All of this, however, leaves us a long way from having a complete solution to the quality/new goods and new outlet problems the Commission believes exist with the CPI. There is much of what the Commission discusses that we do not know how to measure -- or, to put it another way, for which economists simply do not have operational procedures to correct the problems cited. Let me try to illustrate what I mean.

Has the variety of goods and services available to consumers grown? I am certain that it has. Is this variety of value to consumers? Again, I would answer yes. We are, however, a very long way from being able to measure the value of that variety, and thus a very long way from being able to reflect the value of increased variety in the monthly CPI. We have been actively working on potential uses for scanner data in the CPI, one of which might be to allow us to identify new product introductions soon after they occur. Unfortunately, the techniques available for measuring the gains in consumer welfare from those new products (and the losses from product disappearances) are in their infancy, and may never be adaptable for implementation in a large, ongoing price measurement program like the CPI.

To take another example, I would readily acknowledge that there have been major improvements in the medical treatment available for many serious health problems -- improvements that have been of indubitable value to those suffering from the afflictions in question. Unfortunately, as a general matter, the BLS has no good way to measure the value of these improvements. Consider, to take just one example, a hypothetical improvement in knee surgery techniques that gives patients greater mobility following surgery than they previously could have expected. This improved mobility undoubtedly would be of value to those who benefit from the improvement in technique, but there is no obvious or clearly objective way to quantify that value. This is, I believe, an important point about which the Commission and the BLS are in agreement.

The BLS is committed to producing the very best CPI it can. Indeed, as I’ve noted, our Fiscal Year 1998 budget request proposes an increase in funding that would enable us to make significant progress on a number of the issues we have discussed here today. Although I believe that we can make important improvements in the CPI, I do not believe it to be possible to produce a perfect cost-of-living measure. This means that those who use the data we are able to produce should recognize the limitations of those data and exercise judgment accordingly concerning whether and how the data ought to be used.

Attachment A: Improvements to the CPI

Attachment B: Examples Of New Car Reliability/Durability Quality Adjustments In The CPI Since 1992

• Improved corrosion protection - body, electrical system, fuel tank, pump, shocks, brakes and cables
• Increased warranties
• Body side cladding
• Sealing improvements
• Stainless steel exhaust
• Longer life spark plugs - 100,000 mile life
• Improved steering gears
• Powertrain improvements
• Dextron III transmission fluid - 100,000 mile life
• Water pump front face - 150,000 mile life
• Battery saver
• Increased catalyst load - 100,000 mile life
• Rust resistant fuel injection -100,000 mile life
• Clearcoat paint
• Sided galvanized steel body panels
• Serpentine drive belt