Skip Navigation
Introduction

William J. Fowler, Jr.
National Center for Education Statistics


About the Editor



The National Center for Education Statistics (NCES) commissioned the papers in this publication to address advances in measuring education inflation and adjusting for it, as well as to examine the emergence of a new focus on school spending, rather than school district spending, as well as new, private sources of funding for public education, and a review of the status of assessing educational productivity. The first two papers continue the NCES tradition of commissioning papers to address the measurement problems of the education finance research community. The other papers examine the relationship between school district and school spending, and private sources of funding public education, of which surprisingly little is known. The final paper examines the existing attempts to estimate the cost of educational outcomes, and the implications for policymakers and researchers. Before proceeding to precis these works, let us turn to exciting additions to the NCES web page in school finance.

What is new at NCES in education finance?

A primary concern of NCES is to report education finance data that address the needs of policy analysts and policymakers, as well as the needs of the education finance research community. Many persons wish to be notified of free NCES publications, CD-ROMs, or data sets when they become available. NCES has established an e-mail notification system that persons having access to the Internet may use to sign up for announcements of interest. Persons who have signed up for the service can cancel it at any time. Figure 1 shows the example of this service located at NewsFlash.

figure1

In another such effort, NCES has added a "Peer Search" Internet tool to the education finance web page [http://nces.ed.gov/edfin]. Once the "Peer Search" button is selected, located in the left frame of the web page (figure 2), type in the school district name inside the box provided (figure 3). Once you have chosen a school district, the "Peer Search" tool goes to NCES' Common Core of Data (CCD) school district database and compares the spending of the chosen school district with others that are similar in terms of size, wealth, pupil-teacher ratio, urbanicity, and school district type. Bar charts appear for each of the spending variables that are compared (figure 4). However, should you wish to see the actual values (figure 5), you may click on "Summary", located at the top or "Group Details" located at the bottom of the web page (figure 4).

figure2

figure3

figure4

figure5

The "advanced" button, shown when you select the school district (see figure 3), permits the user to modify the characteristics that are used to select school districts. For example, some users believe that school district spending should not be compared unless the characteristics of the students are included. The "advanced" function of the "Peer Search" tool permits the addition of such student characteristics as the percent minority or the percent in poverty to be added to the comparison. In this way, only districts with these student characteristics will be compared. Users may also click "help" at any time for assistance.

NCES wishes to solicit user feedback about the "Peer Search" Internet tool, and is constantly modifying it, based upon comments received. For example, one feature might be the ability to download the selected peer school district data.

A third innovation for NCES is that there is a one-stop place to obtain, free of charge, individual copies of any U.S. Department of Education publication (including CCD CD-ROMs). Called Ed Pubs, the service can be reached by either calling (877) 433-7827 or by e-mailing edpubs@inet.ed.gov. It is helpful if you know the title and publication number of the publication you need. The EDFIN home page has a button to process a list of education finance publications; from this list you can view the title and NCES publication number as well as download a publication. However, printed copies are usually superior to copies printed from the EDFIN web site, and downloading a publication from the web site can be very time-consuming, depending upon the speed of your internet connection.

NCES anticipates revising the publication "Financial Accounting for Local and State School Systems, 1990" to modernize the accounting procedures, to incorporate principles of public school accounting, and to reflect such new procedures as programmatic and school-level accounting. NCES hopes that the new volume will be available for the year 2000.

A precis of the articles in this publication

The first paper, Adjusting for Differences in the Costs of Educational Inputs, by Eric A. Hanushek, the University of Rochester, discusses complexities in deflating educational revenues, desirable if an assessment in the productivity of the education sector is to be made. Although total educational spending has been rising, it may be the result of inflation. In order to understand whether real resources for education are increasing, inflation must be removed from the increase. Another use is to compare spending across states or districts, corrected for purchasing power. The question, of course, is how to make these adjustments.

 

The idea behind price indices is that they should provide an indication of how much more it costs today than yesterday to purchase the same amount of a given commodity.

Complications arise when the purchase of a commodity changes relative to other commodities, for example, if more pens are purchased than pencils for "writing instruments." Also, commodities change over time. In addition, if the "writing instruments" are purchased where competition does not flourish, the price may be excessive (think of military hammer prices). Finally, services are more difficult to measure than commodities, such as hammers. As Hanushek asserts, these combined problems suggest developing reliable price indices for education will be difficult.

 

Hanushek reviews alternative proposals for inflation deflators, including the Net Services Index (NSI) and the Hedonic Price Index (hedonics, in this case, refers to the amenities in a school district). Hanushek explains that the NSI is designed to compare education prices with those in other service sectors expected to be similar to education. Hanushek has previously argued that the authors of the NSI have inadvertently provided evidence for a productivity collapse in education. The hedonic wage index of Chambers makes two advances, incorporating labor market factors (such as working in a high-crime area), and the discretionary choice of school districts to hire higher-quality staff. However, the Chamber's technique relies on a large NCES data set (Schools and Staffing Survey (SASS)) that samples teachers and principles periodically (currently, every five years). Thus, there are extensive analysis costs and only certain occasions to conduct such analyses, rather than a yearly measure. In addition, if there are unmeasured quality differences, they could change over time, and the index would be inaccurate. Since Chamber's results show instability over time, it is more difficult to determine how costs have changed between any survey years.

These conclusions lead Hanushek to propose a new approach, using either the Gross Domestic Product (GDP), or modifying the Chamber's approach by creating a generalized hedonic approach, utilizing the Current Population Survey (CPS), which, although it could not be used at the school district level, would be applicable at state, regional, and national levels. Dan Goldhaber, The Urban Institute, attempts to apply Hanushek's proposal in the next paper.

Goldhaber's paper, An Alternative Measure of Inflation in Teacher Salaries," develops a cost index using data drawn from an annual survey of individuals from the labor market, the Current Population Survey (CPS). Using CPS data for all college graduates in 1987-88, 1990-91, and 1993-94, Goldhaber was able to compare his results to Chamber's. He combined that data with the U.S. Geological Survey, the National Weather Service, and the County and City Data Book. The hedonic methodology permits wages to be decomposed into the part attributable to individual characteristics (i.e., education, experience, sex) and that attributable to community characteristics (i.e., housing values, climate). In competitive labor markets, higher wages have to be paid for an absence of amenities. Goldhaber conducts his analysis, and finds that among discretionary factors (that is, those employers can choose among) wages rise at a decreasing rate with age, and are higher for those with higher educational attainment. White males earn 12 to 15 percent more than any other group. Married workers and union members also receive between 9 and 12 percent more in wages.

There are also community characteristics that cannot be changed or over which the school district has control. For example, a 10 percent increase in housing values drove wages up 1 percent. A 10 degree lower difference in annual temperature translated into 2 to 6 percent higher wages. Most importantly, wages for individuals with the same set of characteristics varied across states and over time, holding constant some discretionary factors. Goldhaber uses the following example:

 

If the 1987 average starting salary for a teacher in Michigan was $25,000, it would only cost about $19,300 to hire a teacher with comparable skills in South Dakota but would cost about $31,200 to hire an equivalent teacher in Alaska.

 

Goldhaber also compares a variety of inflation measures and finds that they are within a few percent of each other. The two that differ the most are the NSI and the Chamber's hedonic index. Goldhaber compares his General Wage Index (GWI) with Chamber's, and he thinks one explanation might be the uncompetitive nature of teacher labor markets. He finds some evidence that teacher costs are higher in states with significant teacher bargaining power. Certainly, more research than this cursory evidence is needed, particularly in light of small sample sizes in some states.

This approach yields similar state wage rankings with the Chamber's approach. Since it is an annual survey, it permits annual updates, allowing researchers to see how teachers' salaries change over time. Its major drawback is that it cannot yield a school-district-level adjustment. However, CPS included county-level identifiers in 1996, which may permit us to revisit this issue.

Surprisingly, there has been little examination of spending in schools, particularly in relationship to their school district, perhaps because of the relative paucity of school-level finance data. What little research has been conducted has emphasized intra-district equity, rather than the causal factors explaining differences in spending between schools within a school district. Amy Ellen Schwartz New York University, examines in School Districts and Spending in the Schools, not only the distribution of spending across schools using 3,284 schools' and 586 districts' data from Ohio, but also the mechanisms for these differences. For example, can the differences be explained by size or other school characteristics, and to what degree is there agreement between districts on what should drive school spending? Schwartz also examines the largest nine school districts reporting data in Ohio, representing 17 percent of the children, since almost 56 percent of the school districts in Ohio (327) have four schools or fewer, and another 36 percent (212) have nine schools or fewer. Thus, only 8 percent, or 47 school districts, have 10 or more schools.

Unfortunately, the Ohio data set has few contextual variables for their schools, such as enrollment, free lunch and student ethnicity, and elementary, middle or secondary. Schwartz finds these few descriptive data only explain less than 30 percent of the variance in per pupil total spending or instructional spending. She finds that Ohio elementary schools receive less per pupil than high schools, and that per pupil spending declines with the size of the student body. She concludes that school spending is largely unexplained by school characteristics. When she controls for differences between districts, greater spending is directed at schools with more poor children (although the magnitude is small, less than $556 per one percentage point increase in poor students).

Schwartz finds that overall spending is higher in larger districts, but the disparities between grade-level schools also grows. High schools in her nine large districts (with more than 20 schools) receive at least $3,000 per student more than elementary schools. Total spending is better explained than instructional spending, and some school districts, such as Columbus, with 130 schools, and Toledo with 60 schools, seem to have a de facto funding formula. Although all district types direct greater spending to schools with a higher percentage of non-white students, the increment is greatest in the small districts. Schwartz concludes that a move to any statewide formula based upon the school characteristics currently in the state database would produce significant changes in the pattern of spending across Ohio public schools. For example, 65 schools (those currently spending the most) would be allocated over 30 percent less money than they currently spend.

A little-noticed change in school funding began in 1990 when a century-long growth in real resources came to an end. Local school districts, faced with revenue restrictions, turned to non-traditional sources of revenue, such as user fees; partnerships with postsecondary institutions; donations; volunteer services; interest earnings on investments; and the creation of educational foundations to promote giving from individuals and businesses. These nontax sources of revenue are not consistently reported by local school districts in their comprehensive annual financial reports. Michael F. Addonizio, Wayne State University, examines these nontraditional revenues, particularly in their impact in Michigan school districts in New Revenues for Public Schools: Alternatives to Broad-Based Taxes.

For the past century, public elementary and secondary education in the United States has enjoyed remarkably steady revenue growth, notes Addonizio. From 1890 to 1990, real expenditure per pupil increased at 3.5 percent per year, more than triple the growth of the Gross National Product (GNP) over this period, resulting in K-12 public school expenditures increasing from less than 1 percent of GNP in 1890 to 3.4 percent in 1990. This increase resulted from a combination of falling pupil-staff ratios, increasing real wages paid to teachers, the expansion of educational services for handicapped students, and rising expenditures outside the classroom. From 1990 to 1993, real spending grew only 0.6 percent. In part, this was due to increasing enrollments, the rapid growth of special education enrollments, and the passage of stringent tax and spending limits enacted by some 43 states. Traditionally, the "nontraditional revenues" have been of relatively small magnitude, consisting of only 7 to 9 percent of total revenues, although there has been evidence that it is the relatively wealthy school districts that enjoy this revenue.

Addonizio classifies the sources of non- traditional revenue. Under "Donor Activities" are direct donations, such as from corporations. An example might be the Safeway program to donate computers to schools. Perhaps the leading indirect donation is in school district foundations that are growing rapidly. According to the National Association of Educational Foundations (NAEF), by the year 2000, there will be 4,000 public school foundations throughout the country. Booster Clubs are an indirect donation that support programs, such as athletics, band, orchestra, and the like, and often donate equipment and uniforms. Enterprise activities have also always been present in schools, and consist of user fees (such as food service, student parking, pupil transportation, tuition fees for electives, textbooks, and extracurricular activities). Sale of school access and leasing of facilities and services are also well acknowledged.

As noted by Addonizio, although the Governmental Accounting Standards Board (GASB) has a draft to recognize the financial contributions of these "affiliated organizations," the statement has not become a "pronouncement," under which those school districts which follow Generally Accepted Accounting Principles (GAAP) would have to report "material" amounts.

Perhaps most interesting is Addonizio's analysis of the rise of educational foundations in Michigan after the passage of the Michigan school finance reform in 1994, and particularly in 1997, as the constraints on traditional revenue sources became binding on school districts. Although revenues have been quite modest, districts with foundations enjoy higher household income, higher achievement, and are larger than their nonfoundation counterparts, as well as largely nonminority districts. This raises concerns about school finance equity for poor and heavily minority school districts.

David H. Monk, Cornell University, and Jennifer King Rice, the University of Maryland, explore the current state of modern education productivity research, and its emerging implications for the financing of education in Modern Education Productivity Research: Emerging Implications for the Financing of Education. Their premise is that the education production function is a useful device for those striving to improve the performance of school systems, and closely related to the education cost function. Since it is necessary to understand the relationship between productivity and cost, that is where they begin. They assert:

 

One of the dilemmas facing policymakers is the design of appropriate responses to evidence of inefficiency with the educational system.

 

Monk and King believe that much of the policymaking significance of resources lies in the potential ability of resources to shape and define desired outcomes. "Some resource combinations simply have higher productivity potentials than do others." The choice of the resource combination may be externally imposed, or arise out of "complicit behavior" by those associated with school districts. Thus, discrepancies can arise between the ideal and actual resource allocation practices.

Since the resources required for one student to learn can be affected by the characteristics of fellow learners, Monk and King first devise an ideal resource and cost distribution, and then contrast that to actual resource allocation practice. Their idea is to ask how much of the service in use will be required to overcome whatever lack of motivation there might be on the part of a student, a teacher, or both. If the wrong service delivery configuration is deployed, the cost of realizing the desirable outcome could become very large. As one example of the problem, they assert it would be inappropriate to hold a building-level administrator accountable for a school that is too small, and costly.

When Monk and King turn to existing attempts to estimate the cost of educational outcomes, they create a continuum from least dependent upon economics to the most dependant upon economics. They begin with "educator judgements" proceed through "hedonic wage models," the "cost of observed best practices," to "cost functions," which are the province of the most sophisticated econometric modeling. Monk and King believe there are good reasons to exercise caution in applying the most sophisticated techniques, because it is not clear if they generate trustworthy efficiency levels, particularly in the face of differences in real costs.

They conclude that education policymakers face contentious choices in a climate of limited resources, being responsible for making parsimonious resource decisions. It is in this climate that the distinction between actual practice and realistic best practice is most important.

The greater the discrepancy in the cost associated with realistic best practice and actual best practice, the more productive the system can become.



Return to the Table of Contents