• Two Approaches to Developing an Index of Inflation for Public Libraries
• Comparing the Fixed-Market-Basket Approach and the Cost of Services Model
• A Comparison of Public Library Inflation Rates Using Each Approach
• Implications for Further Research
• References

One source of information on public library expenditures is the Public Libraries Survey, conducted annually by the National Center for Education Statistics (NCES). This survey utilizes data collected from each state through the Federal-State Cooperative System (FSCS) for Public Library Data. However, because these data are not indexed for inflation, the true impact of inflation on public libraries cannot be assessed. For example, one cannot determine whether the increases in total library revenue that have been shown by FSCS data in recent years led to increases in services or were consumed by inflation.

This report presents two approaches to measuring inflation faced by public libraries:

• an approach based on a fixed market basket (FMC) of the prices of library inputs (i.e., prices of goods and services purchased by libraries, including personnel), which yields a public library input cost index (PLICI); and
• an approach based on an econometric model of library services and costs, which yields a public library cost of services index (PLCSI).

The FMB approach produces an index that is a weighted average of the indexes of the prices of library inputs. This approach uses a methodology similar to that employed in the development of the standard CPI. The standard CPI is essentially an index of the differences in the prices of consumer goods and services between two points in time, weighted by the typical basket of goods and services consumed by households during a base time period. Similarly, the input cost index developed in this report, using the FMB approach, is an index of the differences in the price of library inputs between two points in time, weighted by the typical basket of inputs purchased by libraries. This approach relies on a variety of data sources for the various price data that make up the PLICI. Using this methodology, one can determine a weighted average rate of inflation in the prices of these library inputs, where the weights used to aggregate these individual inputs are the average proportions of public library budgets (i.e., the budget shares) allocated to each input category. These weights or budget shares simply measure the importance of each input in the overall budget for public library operations. This report refers to the inflation index derived using the FMB as the PLICI.

This approach is based on a model of public library services similar to models used by economists to analyze the costs of production in any goods or service industry.¹It is represented by an econometric model of the systematic patterns of variation in library expenditures over time. In addition, the model controls for cost variations associated with changes in the level of library services such as circulation, reference transactions, and library visits, as well as differences in geographic location. By controlling for variations in various types and levels of services rather than holding input levels fixed, this econometric model permits the inflation rates to take into account the effects of input substitutions and technological changes in the cost of doing business for libraries. The phrase "input substitutions" refers to the notion that those in charge of library operations will substitute away from utilizing relatively more expensive inputs toward the use of less expensive inputs over time to maintain service levels at the minimum possible cost. The phrase "technological changes" involves improvements in service levels (or reductions in costs with no diminution in services) that may arise, for example, from the use of computer technology or other time-saving procedures or devices. This cost of services model primarily uses a single data source—the NCES FSCS data on public libraries. This report refers to the inflation index derived using this cost of services model as the PLCSI.

Each approach involves certain assumptions about the way public libraries operate, and each contains limitations in the way cost data maybe interpreted. In addition, the data requirements for using each model differ significantly, and the quality of the data used in calculating each varies considerably. A major difference between the two approaches is the clarity of what underlies the two indexes. Using the FMB model to derive the PLICI, one can see and more easily understand the data components, such as the cost indexes of the various inputs and the budget shares used to aggregate them into a single index. Moreover, this methodology may be familiar to those who are aware of the CPI, which has been published by the U.S. Bureau of Labor Statistics for decades. In contrast, deriving the PLCSI relies on the analytical tools of the economist, which may appear to the non economist as a bit of a black box. Yet economists have used the cost model for decades to analyze production and costs in many industries, including library services (e.g., see Chressanthis 1995 and DeBoer 1992).

Another major difference between the two indexes is that the PLICI represents essentially a weighted average of the series of public library input prices, while the PLCSI places emphasis on the cost of producing library services. As such, the PLCSI attempts to account for the patterns of variation in changes (e.g., improvements) in the level of library services, as well as differences in geographic location. By focusing on the types and levels of library services, the inflation rates produced by the PLCSI reflect input substitutions in response to relative price changes or changes in technology over time, which affect the way library inputs are combined to produce services. The inflation rates produced by the PLICI do not account for these factors.

It is worth noting that the PLCSI, by controlling for various types and levels of services in the way that it does, addresses at least some of the problems that economists have contended create bias in the CPI and other fixed-basket price indexes. A recent paper by Moulton (1996) addresses some of these problems with regard to the construction of the CPI.

During the period from 1989-90 to 1992-93, the PLICI created by the American Institutes for Research (PLICIa)²shows an average annual rate of inflation of 4.3 percent in the prices of library inputs. In marked contrast, the cost-based PLCSI exhibits an average annual inflation rate of 3.9 percent during that same period. For comparison purposes, household consumer prices rose at an average annual rate of 3.9 percent, while producer prices rose at 2.4 percent over this same period.³

The PLICIa estimates of annual inflation rates based upon the FMB approach show roughly similar patterns of decline from 1989-90 to 1992-93 as annual inflation rates based upon the CPI. This is not surprising since several components of the CPI were used to calculate the input cost index of various library expenditure categories using the FMB approach. For example, the input cost index of the major library expenditure category, books and periodicals, is based upon the CPI data.

Inflation rates derived from the cost of services model show lower rates of inflation than those derived using the FMB approach. This is consistent with the expectation that the cost of services model should control better for increases in the costs of library services due to improvements in the level of services or technological change.⁴

This report provides suggestions about further data collection and research that would be useful in exploring alternative ways of developing a PLICI. The kinds of econometric models used in the development of the PLCSI have the potential to address the factors underlying differences in available library services. This can be accomplished by examining the systematic relationship between library outcomes or services in local communities in relation to variations in local community characteristics (e.g., income and education levels of the local community) and the federal and state grants on library spending and service levels.

Data sources: The NCES Public Libraries Survey (PLS), FY 1989-93; the American Library Association (ALA) Survey of Librarian Salaries, 1988-94; the U.S. Census Bureau's County and City Data Book, 1990 (1989 data); and other previously published data, as cited in the text. For technical information, see the complete report: Chambers, J.C., and Vergun, R. (1999). Measuring Inflation in Public Libraries: A Comparison of Two Approaches, the Input Cost Index and the Cost of Services Index (NCES 1999-326). Author affiliations: J.C. Chambers and R. Vergun, American Institutes for Research. For questions about content, contact Adrienne Chute (Adrienne.Chute@ed.gov). To obtain the complete report (NCES 1999-326), call the toll-free ED Pubs number (877-433-7827), visit the NCES Web Site (http://nces.ed.gov), or contact GPO (202-512-1800).