Skip Navigation
small NCES header image

Richard Rothstein
Economic Policy Institute

About the Author

Richard Rothstein is a research associate of the Economic Policy Institute, adjunct professor of public policy at Occidental College in Los Angeles, and editorial board member of The American Prospect. He writes on public policy for a variety of publications.

Rothstein's article in this journal summarizes highlights from his full report, Where's the Money Gone? Changes in the Level and Composition of Education Spending, 1967-1991, by Richard Rothstein with Karen Hawley Miles, a school district consultant and doctoral candidate at Harvard's Graduate School of Education. The report was published by the Economic Policy Institute in November, 1995, based on a two-year data gathering project in nine typical U.S. school districts. Funding for the report and data collection was provided by the Metropolitan Life Foundation.

Rothstein was co-editor, with Edith Rasell, of School Choice: Examining the Evidence, also published by the Economic Policy Institute (1993). In addition to education, he also researches international trade and labor standards. Prior to his work at the Economic Policy Institute, Rothstein was a program analyst for the Los Angeles school board. He was previously a high school teacher, college instructor, and union officer and representative.

Rothstein is currently revising the report summarized here, to include data through school year 1995-96. He plans, in spring, 1997, to publish a follow-up report, covering programmatic changes in school spending from 1966-67 to 1995-96, and from 1990-91 to 1995-96.

Rothstein welcomes comments and inquiries about this work, via e-mail to <rothstei@oxy.edu>.

Summary of Where's the Money Gone: Changes in the Level and Composition of Education Spending (1967-1991)
Richard Rothstein
Economic Policy Institute

When Benno Schmidt resigned Yale's presidency to head a private school network, he explained why he had given up on public schools: "We have roughly doubled per-pupil spending (after inflation) in public schools since 1965," but the "nation's investment in educational improvement has produced very little return."¹

This is a conventional claim of public school supporters and critics alike. School finance expert Allan Odden notes that "real" education expenditures increased by 58 percent in the 1960s, 27 percent in the 1970s, and 30 percent in the 1980s, "but student performance--and thus education productivity--have not improved that much." According to a Brookings Institution report by John Chubb and Eric Hanushek, "since the Soviets launched...Sputnik,...real expenditures per student rose at an annual rate of 3 3/4 percent, nearly tripling between 1960 and 1988... Spending has nearly tripled and performance has dropped."

The declining "productivity" claim is so well established that few analysts have sought empirical verification. Rather, the notion is a prelude to reform prescriptions; if, after all, growth in public education spending has outpaced any rise in school achievement, the challenge is to design systems that use money more effectively, with no need to consider proposals for additional funds.

While education spending has risen substantially, the increase is both smaller and more complex than most assume: real school spending increased by 61 percent between 1967 and 1991, about half the expected real growth. Barely one-quarter of this increase occurred in "regular education," the traditional school activities whose outcomes can be measured in test scores, graduation rates, etc.

In this analysis, we did not adjust the 1967 school expenditure data with the Consumer Price Index for all urban consumers (CPI-U). Instead, the Economic Policy Institute (EPI), with assistance from the Bureau of Labor Statistics (BLS), developed a modified index (we call it the "net services" index (NSI)) to measure inflation in service industries like education. When we use the NSI to adjust schools' average per-pupil spending of $687 in 1966-67, we find that "real" spending in 1967 was $3,456. Real school spending growth is the amount of new money spent in excess of inflationary changes measured by the NSI--or 61 percent real growth.

Assigning Expenditures to Programs

States and districts do not normally report school spending data by program type, although education researchers increasingly ask them to do so. To understand which programs received which funds, we first defined the programs and then categorized each expenditure of the nine sample districts by program and employee type (i.e., teachers, aides, other professionals, or other employees).

When classifying programs into categories, we found that distortions are inevitably implied by any taxonomy of school programs. Former Education Secretary William Bennett's accusation that schools waste funds in an "administrative blob" has led many researchers to focus on distinguishing administrative from classroom costs. They generally find administrative costs (central and school-level) of 8 percent to 17 percent. We, on the other hand, assigned many expenditures, often termed "administrative costs," to programs.

The conventional focus on the distinction between "classroom" and "administrative" expenditures implicitly posits an industrial model of schools: classrooms are like factory floors where "direct" teaching labor carries out production, while other functions provide indirect support. But, as in manufacturing, schools do not succeed as "direct to indirect" ratios increase. Success depends on the intelligence with which the enterprise is planned and coordinated, as well as on the product mix created. The implicit notion in educational debate that classrooms are "profit centers," while curriculum libraries or school buses are "cost centers," prevents thoughtful analysis of programmatic productivity. Central office development of curriculum guides, for example, and teachers' transmission of this curriculum, are equally necessary to instruction-- either may be conducted effectively or wastefully. By calculating total costs for each program, including administrative costs, we do not suggest that leadership funds are well spent, any more than we believe that a separation of classroom expenditures would identify the most effective teaching techniques. This cannot be determined by finance analyses alone and must be addressed in separate inquiries.

Our approach creates categories that are not comparable to those studies that specifically segregate administrative costs, because we distinguish administrators who are associated with a particular program from those responsible for the overall direction of the school enterprise. We assign the former exclusively to the program with which they are identified; we include the latter in a general administrative "overhead" or indirect category that is ultimately allocated to programs in proportion to each program's "direct" expenditures.

Categorizing expenditures by program creates two unfamiliar results. First, those who specialize in understanding school finance will be unaccustomed to seeing programmatic expenditures. Second, because we include pro-rata shares of state government expenditures (i.e., textbook selection, testing, special schools, teacher certification, and retirement fund contributions), total expenditure figures that we obtain will differ from those in district reports. Because our categories may not easily be comparable to those used in each school district's report, we recognize that our data must be subject to verification by other researchers. Therefore, we do not offer anonymity to districts in the sample, and EPI will make arrangements for qualified researchers who wish to confirm these calculations.

Findings

Table 1 describes changes in the shares of total spending for each program for the nine districts included in this analysis. The table shows that special education's share increased the most, increasing from less than 4 percent of all spending in 1967 to 17 percent in 1991. In contrast, the share of funds allocated to regular education declined. In 1967, regular education consumed almost all of the available elementary and secondary education dollars--80 percent went to regular classroom teaching, school libraries, textbooks, curriculum development, and teacher training. (This calculation includes a pro-rata, 80 percent share of maintenance and general administrative costs). The share of elementary and secondary spending for regular education fell to 59 percent in 1991. Table 2 describes the distribution of net new money in 1991 in the nine districts. Special education took the largest share (38 percent) of net new money, while regular education received only 26 percent of net new money.

Table 1.-Shares of total per-pupil spending for each program, average of nine districts: 1967 and 1991 (Programs listed in order of 1991 share of total per-pupil spending)

----------------------------------------------------------------------------------------------
                                          Share of total     
                                        per-pupil spending   Change in  
                                        ------------------     Share
Program                                   1967     1991       1967-91
Regular education                         79.6 %   58.8 %      (20.9)
Special education                         3.7      17.0         13.3
Compensatory education                    5.4       4.3         (1.1)
Attendance, counseling, dropout
  prevention, alternative education       2.1       4.1          2.0
Food services                             2.0       4.1          2.1
Regular student transportation            3.9       3.4         (0.5)
Vocational education                      1.4       3.0          1.6
Bilingual education                       0.3       1.8          1.5
Desegregation                             0.0       1.6          1.6
Regular health & psychological services   1.3       0.9         (0.3)
After-school athletics                    0.4       0.7          0.3
Security and violence prevention          0.1       0.4          0.3
All programs                            100.0     100.0
Overhead allocated to above programs:
  General and school administration       9.4       9.7          0.2
  Operations and maintenance             15.7      14.3         (1.4)
----------------------------------------------------------------------------------------------

SOURCE: Where's the Money Gone? Changes in the level and Composition of Education Spending, 1967-1991, by Richard Rothstein with Karen Hawley Miles. Washington, DC.: Economic Policy Institute, 1995. (Table 5).


Table 2.-Shares of net new per-pupil spending, by program: 1967-91 (Programs listed in order of net new per-pupil spending)

---------------------------------------------------------------------------
                                   Share of net new
                                   per-pupil spending,
                                   1991 nine district
Program                              average
---------------------------------------------------------------------------
Special education                         38.0 %
Regular education                         25.9
Food services                              7.5
Attendance, counseling, dropout
  prevention, alternative education        7.4
Vocational education                       5.2
Desegregation                              4.1
Bilingual education                        3.9
Compensatory education                     2.9
Regular student transportation             2.8
After-school athletics                     1.1
Security and violence prevention           0.9
Regular health & psychological services    0.5
All programs                             100.0
Overhead allocated to above programs:
  General and school administration        9.8
  Operations and maintenance              12.3
---------------------------------------------------------------------------

SOURCE: Where's the Money Gone? Changes in the level and Composition of Education Spending, 1967-1991, by Richard Rothstein with Karen Hawley Miles. Washington, DC.: Economic Policy Institute, 1995. (Table 6).

Tables 3 and 4 compare changing shares of total spending and amounts of net new money allocated to regular and special education, the two largest spending categories for each district. These tables show that each district in the sample increased its share spent on special education and decreased the share of money spent on regular education.

Table 3.-Changes in regular education share of per-pupil spending: 1967-91 (Listed in order of 1991 share for regular education)
-----------------------------------------------------------------------------
                                Regular education
                                  share of per  Regular education
                                  spending cha  share of net new
                                    in share    per-pupil spending:
District            1967     1991    1967-91    1967-91
-------------------------------------------------------------------------------
Bettendorf           92 %     72 %     (20)%       36 %
Boulder              84       64       (20)         9
Anne Arundel         83       62       (21)        41
Spring Branch        83       60       (23)        44
Middletown           78       59       (19)        35
East Baton Rouge     76       57       (18)        22
Clairborne           57       53        (4)        46
Fall River           76       51       (25)         4
Los Angeles          87       51       (36)        (5)
Average              80       59       (21)        26
-------------------------------------------------------------------------------
SOURCE: Where's the Money Gone? Changes in the level and Composition
of Education Spending, 1967-1991, by Richard Rothstein with Karen Hawley Miles.
Washington, DC.: Economic Policy Institute, 1995. (Table 7).

Table 4.-Changes in special education share of per-pupil spending: 1967-91 (Listed in order of 1991 share for special education)

--------------------------------------------------------------------------------
                                                   
                                  Special education  Special education
                                  share of per-pupil Share of net new
                                  spending change    per-pupil spending:
District             1967     1991   1967-91         1967-91
--------------------------------------------------------------------------------
Fall River            8 %     22 %     14 %            49 %
Middletown            2       22       20              46
Los Angeles           2       18       16              42
Anne Arundel          3       18       15              33
Boulder               4       16       12              47
East Baton Rouge      4       16       12              40
Bettendorf            3       16       12              39
Spring Branch         3       13       10              19
Clairborne            4       12        9              26
Average               4       17       13              38
--------------------------------------------------------------------------------

SOURCE: Where's the Money Gone? Changes in the level and Composition of Education Spending, 1967-1991, by Richard Rothstein with Karen Hawley Miles. Washington, DC.: Economic Policy Institute, 1995. (Table 8).


Regular Education Spending

The decline in the share of total spending for regular education does not mean that real per-pupil spending for regular education decreased. A smaller share of a larger budget can still provide increases. Because total per-pupil spending grew by 73 percent in these nine districts, real regular education per-pupil spending increased despite regular education's reduced relative priority.

Table 5 shows the real per-pupil growth of regular education spending in the nine districts. The range of percentage change in spending between 1967 and 1991 is broad--from a percentage change of 77.9 percent in the Spring Branch school district to a percentage change of 3.5 percent in the Los Angeles school district. The average percentage change was a 28 percent increase in spending per-pupil. Table 6 distinguishes direct from indirect (general administration, operations, and maintenance) expenses in regular education. This table shows that the 3.5 percent decline in the Los Angeles school districts' per-pupil spending on regular education (an average annual decrease of 0.1 percent) partly reflects a much larger decline (an annual average of 2.7 percent) in indirect operations and maintenance expenditures. However, even without this indirect overhead allocation, the direct per-pupil expenditures for regular education in Los Angeles increased only by 9 percent between 1967 and 1991, an average annual rate of only 0.3 percent. This rate is significantly less than the average for the nine districts sampled (where direct per-pupil expenditures for regular education, exclusive of administrative, operations, and maintenance overhead, increased by an annual average of 1.1 percent).

Table 5.-Growth in real regular education per-pupil spending

-------------------------------------------------------------------------------
                                                         Average total
                            Regular education  Total       annual
                            spending per-pupil percent     percent
                            (in 1991 dollars)  Change       change
District                      1967    1991    1967-91      196791
-------------------------------------------------------------------------------
Spring Branch                $1,825  $3,247    77.9 %        2.4 %
Anne Arundel                  2,513   3,780    50.4          1.7
Clairborne                    1,027   1,524    48.4          1.7
Middletown                    3,424   4,691    37.0          1.3
Bettendorf                    2,674   3,229    20.8          0.8
East Baton Rouge              2,096   2,424    15.7          0.6
Boulder                       3,189   3,317     4.0          0.2
Fall River                    2,279   2,345     2.9          0.1
Los Angeles                   3,118   3,010    (3.5)        (0.1)
Average change                                 28.2          1.0
-------------------------------------------------------------------------------

SOURCE: Where's the Money Gone? Changes in the level and Composition of Education Spending, 1967-1991, by Richard Rothstein with Karen Hawley Miles. Washington, DC.: Economic Policy Institute, 1995. (Table 9).


Table 5.-Growth in real regular education per-pupil spending

-------------------------------------------------------------------------------
                                                         Average total
                            Regular education  Total       annual
                            spending per-pupil percent     percent
                            (in 1991 dollars)  Change       change
District                      1967    1991    1967-91      196791
-------------------------------------------------------------------------------
Spring Branch                $1,825  $3,247    77.9 %        2.4 %
Anne Arundel                  2,513   3,780    50.4          1.7
Clairborne                    1,027   1,524    48.4          1.7
Middletown                    3,424   4,691    37.0          1.3
Bettendorf                    2,674   3,229    20.8          0.8
East Baton Rouge              2,096   2,424    15.7          0.6
Boulder                       3,189   3,317     4.0          0.2
Fall River                    2,279   2,345     2.9          0.1
Los Angeles                   3,118   3,010    (3.5)        (0.1)
Average change                                 28.2          1.0
-------------------------------------------------------------------------------

SOURCE: Where's the Money Gone? Changes in the level and Composition of Education Spending, 1967-1991, by Richard Rothstein with Karen Hawley Miles. Washington, DC.: Economic Policy Institute, 1995. (Table 9).


If the changes discussed above are typical of other urban mega-districts, it might help to explain why there is great concern about the academic outcomes of these districts. Other urban districts in the sample, (i.e., Fall River and East Baton Rouge), also saw real direct per-pupil regular education costs increase by less than the sample average, and this increase was partly a result of drastic reductions in maintenance expenses (and, in Fall River, administrative expenditures as well) attributed to regular education. In contrast, suburban districts such as Spring Branch, Anne Arundel, and Middletown saw direct regular education per-pupil spending increase at an average annual rate of 2.4 percent, 1.7 percent and 1.4 percent, respectively.

The Components of Regular Education Spending and Sources of Change

The 28 percent real growth of regular education spending raises three issues:

  • Did regular education spending components change over time?
  • How much average teacher salary growth is attributable to higher salary scales, and how much resulted from teachers ascending pay scales by gaining experience and education?
  • Have staffing patterns (class size, for example) changed?

To answer these questions Karen Hawley Miles analyzed staffing and enrollment detail for three districts of different sizes: Middletown (small), Boulder (medium-sized), and East Baton Rouge (large). While the regular education costs grew by 28 percent for the nine-districts in the sample, Miles found that in this subsample regular education costs increased 19 percent. The EPI report uses her subsample analysis, in combination with expenditure data from the nine districts, to assess basic patterns, understand variation, and highlight important questions for further investigation. Our analysis suggests that:

  • Dollar allocations between teaching, administration, instructional materials, and other regular education functions changed little between 1967 and 1991. But, this continuity masks important trends.
  • In the three districts Miles examined in greater detail, per-pupil spending on compensation for regular education teachers grew by 23 percent.
  • Salary scales barely kept pace with inflation: teachers' age (experience) and education almost fully account for the increase in average salaries.
  • Spending on instructional aides grew, but still accounts for a relatively small portion of instructional spending.

Further Research

EPI's study, Where's the Money Gone, takes preliminary steps in research areas previously ignored by the education research community. Elaboration and refinement of the "net services index" (NSI), could provide more realistic pictures of education spending changes in other districts and time periods, as well as permit more realistic comparisons of resource growth in different districts and regions. The classification of expenditures by program for other districts and other time periods could help confirm whether the initial conclusions of the EPI report are valid. A uniform programmatic accounting system, if adopted by districts, would help policymakers learn where education money is spent. Development of outcome measures for many school programs is needed before programmatic expenditure data can be used in future discussions of school productivity.

¹ Schmidt, Benno C., Jr. 1992. "Educational innovation for profit." The Wall Street Journal , June 5.



[Administrative Expenditure Limits for Texas Public Schools] Previous Table of Contents Next[Improving School Preformance While Controlling Costs]
Would you like to help us improve our products and website by taking a short survey?

YES, I would like to take the survey

or

No Thanks

The survey consists of a few short questions and takes less than one minute to complete.
National Center for Education Statistics - http://nces.ed.gov
U.S. Department of Education