USING THE CONSUMER PRICE INDEX (CPI) TO ADJUST FOR INFLATION

The Consumer Price Indexes (CPIs) represent changes in the prices of all goods and services purchased for consumption by households. Indexes vary for specific areas or regions, periods of time, major groups of consumer expenditures, and population groups. Indicators 20, 34, 35, 36, 37, and 42 in The Condition of Education use the U.S. All Items CPI for All Urban Consumers (CPI-U).

The CPI-U is the basis for both the calendar year CPI and the school year CPI. The calendar year CPI is the same as the annual CPI-U. The school year CPI is calculated by adding the monthly CPI-U figures, beginning with July of the first year and ending with June of the following year, and then dividing that figure by 12. The school year CPI is rounded to three decimal places. Data for the CPI-U are available on the Bureau of Labor Statistics (BLS) website (see below). Also, figures for both the calendar year CPI and the school year CPI can be obtained from the Digest of Education Statistics, 2007 (NCES 2008-022), an annual publication of the National Center for Education Statistics (NCES).

Although the CPI has many uses, its principal function in The Condition of Education is to convert monetary figures (salaries, expenditures, income, etc.) into inflation-free dollars to allow comparisons over time. For example, due to inflation, the buying power of a teacher’s salary in 1998 is not comparable to that of a teacher’s salary in 2002. In order to make such a comparison, the 1998 salary must be converted into 2002 constant dollars by multiplying the 1998 salary by a ratio of the 2002 CPI over the 1998 CPI. As a formula, this is expressed as

The reader should be aware that there are alternative price indexes to the CPI that could be used to make these adjustments. These alternative adjustments might produce findings that differ from the ones presented here. For more detailed information on how the CPI is calculated or the other types of CPI indexes, go to the BLS website (http://www.bls.gov/cpi/).

Using the Comparable Wage Index (CWI) to Adjust for Geographic Cost Differences

The Comparable Wage Index (CWI) reflects systematic, regional variations in the salaries of college graduates who are not educators. Provided that these noneducators are similar to educators in terms of age, educational background, and tastes for local amenities, a CWI can be used to measure the uncontrollable component of variations in the wages paid to educators. Intuitively, if accountants in the Atlanta metro area are paid 5 percent more than the national average accounting wage, Atlanta engineers are paid 5 percent more than the national average engineering wage, Atlanta nurses are paid 5 percent more than the national average nursing wage, and so on, then the CWI predicts that Atlanta teachers should also be paid 5 percent more than the national average teacher wage.

The CWI was calculated by combining baseline estimates of these salaries from the 2000 U.S. Census with annual data from the Bureau of Labor Statistics’ Occupational Employment Statistics (OES) Survey. Different sets of CWIs are available for adjusting finance data at different levels of aggregation: the region, state, labor market, and school district. The school district CWI can also be used to adjust for other geographic levels as well.

In indicator 37, for each year under study, an index number was developed for each of the five district poverty levels used in the indicator. These district poverty-level index numbers were calculated by (1) summing, within each poverty level, each district’s index number multiplied by the district’s enrollment, and then (2) dividing that sum by the total enrollment in that district poverty level. The same method was used to develop the 20 index numbers for the different combinations of community type and district poverty level for 2004–05 and the national index numbers for each year.

When a series of annual CWIs are used for adjusting trend data, an adjustment for inflation is inherently introduced along with the CWI’s adjustment for geographic cost differences. In order to maintain consistency in adjustments for inflation across indicators in The Condition of Education, CWI adjustments for trend data are further adjusted to eliminate the inherent CWI adjustment for inflation. This is done for each year by dividing each district’s CWI by the national CWI, which leaves only an adjustment based on geographic cost differences for each year. Adjustments for the effect of inflation are then made using the Consumer Price Index (CPI).

Indicator 36 presents two Theil coefficients for instruction expenditures: one that is adjusted for cost differences and one that is unadjusted. The adjusted Theil coefficient in this indicator is calculated in the same way as the unadjusted Theil coefficient, except that each district’s instruction expenditures that are used in the calculation have been adjusted first by dividing the district’s instruction expenditures by the district’s CWI. (For details on how the Theil coefficient is calculated, see The Variation in Expenditures per Student and the Theil Coefficient section later in this supplemental note.)

Further information about the CWI is available at http://nces.ed.gov/edfin/adjustments.asp.

Classifications of Expenditures for Elementary and Secondary Education

Indicators 35, 36, and 37 examine expenditures for public elementary and secondary education. Indicator 35 uses total expenditures as a whole together with the four major functions (categories) of total expenditures: current expenditures, capital expenditures, interest on school debt, and other expenditures. Current expenditures in turn is broken into seven subfunctions (subcategories): expenditures for instruction, administration, student and staff support, operation and maintenance, transportation, food services, and enterprise operations. Indicator 36 uses expenditures for instruction (referred to usually as instruction expenditures) in its analysis. Indicator 37 uses two categories of expenditures in its analysis: current expenditures and instruction expenditures.

Total expenditures for elementary and secondary education include all expenditures allocable to per student costs: these are all current expenditures for regular school programs, capital outlay, interest on school debt, and other expenditures. Expenditures on education by other agencies or equivalent institutions (e.g., the Department of Health and Human Services and the Department of Agriculture) are included.

Current expenditures include expenditures for instruction, administration, student and staff support, operation and maintenance, transportation, food services, and enterprise operations. Thus, current expenditures include such items as salaries for school personnel, fixed charges, student transportation, school books and materials, and energy costs. Current expenditures and each of its seven subfunctions can be further broken down by the object of the expenditure: salaries, employee benefits, purchased services, supplies, tuition, and other.

Instruction expenditures include salaries and benefits for teachers and instructional aides, supplies, and purchased services such as instruction via television. Also included are tuition expenditures to other local education agencies.

Administration expenditures include expenditures for general administration (boards of education staff and executive administration) and school administration (i.e., the office of the principal, full-time department chairpersons, and graduation expenses).

Student and staff support expenditures include expenditures for student support (attendance and social work, guidance, health, psychological services, speech pathology, audiology and other student support services); instructional staff services (instructional staff training, educational media [libraries and audiovisual], and other instructional staff support services); and other support services (business support services, central support services, and other support services not reported elsewhere).

Operation and maintenance expenditures include expenditures for supervision of operations and maintenance; operating buildings (heating, lighting, ventilating, repair, and replacement); care and upkeep of grounds and equipment; vehicle operations and maintenance (other than student transportation); security; and other operations and maintenance services.

Transportation includes expenditures for vehicle operation, monitoring, and vehicle servicing and maintenance.

Food services include all expenditures associated with providing food to students and staff in a school or school district. The services include preparing and serving regular and incidental meals or snacks in connection with school activities as well as the delivery of food to schools.

Enterprise operations include expenditures for operations funded by sales of products or services together with amounts for direct program support made by state education agencies for local school districts.

Capital outlays include funds for the acquisition of land and buildings; building construction, remodeling, and additions; the initial installation or extension of service systems and other built-in equipment; and site improvement. The category also encompasses architectural and engineering services including the development of blueprints.

Interest on debt includes expenditures for long-term debt service interest payments (i.e., those longer than 1 year).

Other expenditures include expenditures for adult education, community colleges, and private school programs, which are funded by local and state education agencies and by community services.

CLASSIFICATIONS OF REVENUE

In indicator 34, revenue is classified by source (federal, state, or local). Revenue from federal sources includes direct grants-in-aid to schools or agencies, funds distributed through a state or intermediate agency, and revenue in lieu of taxes to compensate a school district for non-taxable federal institutions within a district’s boundary. Revenue from state sources includes both direct funds from state governments and revenue in lieu of taxation. Revenue from local sources includes revenue from such sources as local property and nonproperty taxes; investments; and revenue from student activities, textbook sales, transportation and tuition fees, and food services. Intermediate revenue comes from sources that are not local or state education agencies, but operate at an intermediate level between local and state education agencies and possess independent fundraising capability—for example, county or municipal agencies. Intermediate revenue is included in local revenue totals. In indicator 34, local revenue is classified as either local property tax revenue or other local revenue.

In indicator 34, alternative local government revenue numbers for Texas were used in the calculation of the percentage distribution for the South in 1992–93 because, for that state, much of the revenue that was classified as local government property taxes was classified as revenue from intermediate sources. The alternative Texas local government property tax revenue for 1992–93 was calculated by applying the average of the proportions of the 1991–92 and 1993–94 local government property tax revenue to all local government revenue to the 1992–93 total for all local government revenue. Other local government revenue was calculated in a similar fashion.

THE VARIATION IN EXPENDITURES PER STUDENT AND THE THEIL COEFFICIENT

Indicator 36 uses the Theil coefficient to measure the variation in expenditures per pupil in regular public school elementary and secondary schools in the United States.

The Theil coefficient was developed by Henri Theil to measure the amount of information conveyed by a single message that an event has occurred. It was derived from the study of what Theil called the “information concept.” If we know an event is likely (i.e., the probability of the event is close to 1.0), then the amount of information conveyed is low (i.e., it is no surprise that the event occurred). But if the probability is low (i.e., near zero), a message saying it occurred provides a significant amount of information. Intuitively, and later rigorously proven by Theil and others, the function of the amount of information conveyed is logarithmic (i.e., h(z) = ln(1/z), where h = information function and z = probability of event).

Having developed the information function as a measure of the amount of information conveyed, Theil then suggested that this information function could also be used as a measure of dispersion. For example, if instructional expenditures per pupil in the nation are relatively close together (i.e., low disparity), then relatively little information would be provided by random draws of the districts (i.e., the 1/z_i —the probabilities—are high, but the value of the information function—the sum of the logarithms—is low). In contrast, if instructional expenditures per pupil are very dissimilar, then probabilities for drawing a given level of expenditures are lower, and the information gained from a random draw will be high. Thus, the information function can be a measure of dispersion, and a comparison of the values of Theil coefficients for groups within a set (i.e., districts within the nation) will indicate relative dispersion and any variations that may exist among them. The Theil coefficient was subsequently used to measure the trends in variation of a number of items, including expenditures per student (see NCES 2000-020 and Murray, Evans, and Schwab 1998).

The Theil coefficient has a convenient property when the individual units of observation (e.g., school districts) can be aggregated into subgroups (e.g., states): the Theil coefficient for the aggregation of all the individual units of observation can be decomposed into a measure of the variation within the subgroups and a measure of the variation between the subgroups. Hence, in the examination of the variation in instructional expenditures in the United States, the national variation can be decomposed into measures of between-state and within-state variation.

The between-state Theil coefficient, T_B, equals:

where P_k is the enrollment in state k, X{bar}k is the student-weighted mean expenditure per student in state k, and X{bar} is the student-weighted mean expenditure per student for the country.

The within-state Theil coefficient, T_W, equals:

where T_k is the Theil coefficient for state k.

T_k equals:

where P_jk is the enrollment of district j in state k and X_jk is the mean expenditure per student of district j in state k.

The national Theil coefficient, T, is:

CLASSIFICATIONS OF EXPENDITURES FOR INTERNATIONAL COMPARISONS

Indicator 38 presents international data on public and private expenditures for instructional and noninstructional educational institutions. Instructional educational institutions are educational institutions that directly provide instructional programs (i.e., teaching) to individuals in an organized group setting or through distance education. Business enterprises or other institutions providing short-term courses of training or instruction to individuals on a “one-to-one” basis are not included. Noninstructional educational institutions are educational institutions that provide administrative, advisory, or professional services to other educational institutions, although they do not enroll students themselves. Examples include national, state, and provincial bodies in the private sector; organizations that provide education-related services such as vocational and psychological counseling; and educational research.

Public expenditures refer to the spending of public authorities at all levels. Total public expenditures used for the calculation in indicator 38 correspond to the nonrepayable current and capital expenditures of all levels of the government directly related to education. Expenditures that are not directly related to education (e.g., culture, sports, youth activities, etc.) are, in principle, not included. Expenditures on education by other ministries or equivalent institutions (e.g., Health and Agriculture) are included. Public subsidies for students’ living expenses are excluded to ensure international comparability of the data.

Private expenditures refer to expenditures funded by private sources (i.e., households and other private entities). “Households” mean students and their families. “Other private entities” include private business firms and nonprofit organizations, including religious organizations, charitable organizations, and business and labor associations. Private expenditures comprise school fees; the cost of materials such as textbooks and teaching equipment; transportation costs (if organized by the school); the cost of meals (if provided by the school); boarding fees; and expenditures by employers on initial vocational training. Private educational institutions are considered to be service providers and do not include sources of private funding.

Current expenditures include final consumption expenditures (e.g., compensation of employees, consumption of intermediate goods and services, consumption of fixed capital, and military expenditures); property income paid; subsidies; and other current transfers paid. Capital expenditures include spending to acquire and improve fixed capital assets, land, intangible assets, government stocks, and non-military, nonfinancial assets, as well as spending to finance net capital transfers.