During the 1994-95 school year, Americans spent $278,966,000,000 on public education. The federal government accounted for 6.8 percent of this spending with the rest divided about equally between state and local government agencies. Over 44.1 million enrolled students were supervised by 3.3 million teachers, principals and other instructional staff. That year, every man, woman and child in the United States contributed $1,071 to the support of public schools. Public education is obviously a large-scale government enterprise that requires an equally large commitment of resources.

The most common statistic in reference to school finance is per-pupil spending. There are many different ways to express per-pupil spending and, unfortunately, commentators often fail to emphasize these subtle distinctions. Generally, the money factor used for computation is either current expenditures or total expenditures. Total expenditures are current expenditures plus capital outlay and interest on school debt. The pupil factor used for computation is either total enrollment or average daily attendance (ADA). The simplest way to express this distinction is to say that ADA tells us how many students attended school on an average day, while enrollment tells us how many should have attended.

So we have four different ways of figuring per-pupil spending for any specific year. Each table in this report specifies which formulation it is using, but use caution when comparing per-pupil spending figures from one table to those of another.

We generally see per-pupil spending statistics set out in a ranking of states, so that Connecticut's spending can be compared to New York's. However, per-pupil spending may vary widely within a state, with neighboring districts spending significantly different amounts of money per pupil. Perhaps it would be useful to compare district to district, rather than state to state.

Tables 1 and 2 are a top 25 and bottom 25  per-pupil spending ranking of districts with more than 20,000 students. The tables use current expenditures for 1993-94 and fall enrollments based on U.S. Census figures. The average for all U.S. districts with more than 20,000 students was $5,249.

Geography appears to account for the majority of the difference in spending. Districts in the Northeast make up 13 of the top 14 in spending. But why is Newark spending 42 percent more per student than New York City C a short drive away? Why do Michigan and Wisconsin districts spend more than Ohio and Illinois districts? Why do mostly white Utah districts spend less than mostly black Alabama districts? How can the Fairfield-Suisun Joint Unified School District C 50 miles from San Francisco C spend only $3,722 per student?

These tables also bring into question the widely held belief that urban districts are starved for resources compared to nonurban districts. While it is likely that suburban districts with enrollments under 20,000 might outspend large urban districts, these statistics indicate that, in general, the large urban districts are spending more per-pupil than the large suburban districts. Why?

Perhaps if we examined where the money is spent we could answer those questions. How are current expenditures determined and how are they categorized? The standard approach to current expenditures is to divide them into functions: instruction, student services, instructional services, general administration, school administration, operation and maintenance, student transportation, student support services, food services, and enterprise operations.

Instruction is classroom spending, including the cost of the teacher and classroom supplies.

Student services include guidance, health, attendance and speech pathology services.

Instructional services include curriculum development, staff training, libraries and media centers.

General administration is the off-site bureaucracy at district, county, state and federal levels.

School administration includes principals, vice principals and other on-site administrators.

Operation and maintenance are costs associated with the buildings and structures.

Student transportation is self-explanatory.

Student support services is a category for miscellaneous expenditures for various special programs.

Food services includes the cost of providing breakfasts, lunches and snacks to students.

Enterprise operations include the cost of operating school bookstores or computer centers.

It bears noting that the percentages hardly vary by district. That is, wealthy and poor school districts allocate their funds in approximately these same percentages no matter where they are located.

Expressed in such a way, these percentages are designed to solicit a particular response, which would be: "Well, more than 61 percent spent in the classroom, about another 16 percent that directly benefits students, 10 percent on maintenance, and only 8 percent on administration. Not bad."  That is a satisfactory response for most school officials, but it doesn't go far enough. We've determined where we spent the cash, but who received it? Obviously not the students, so it must be the employees.
Education is without a doubt a labor-intensive enterprise. But just how much education spending goes to labor might come as a shock to some people. Let's re-categorize the current expenditures for 1994-95, this time dividing them more simply. Table 4 breaks spending into salaries, benefits, purchased services, supplies, and other.

Expressed this way, we might expect a different type of response, which is: "Almost 83 percent goes toward salaries and benefits?!" Suddenly, concerns about teachers spending their own money on classroom supplies don't seem so cut-and-dried anymore. Are they spending money that should have been allocated to supplies rather than to them in the first place? It raises huge unanswered questions about the education labor pool. Is better compensation buying us better teachers, administrators and principals? Are more highly qualified people entering the teaching profession because of the pay and benefits? Have we tapped out the education labor market, meaning higher wages will only benefit those already in education? Or do we need to make a quantum leap in salaries, increasing them by as much as 50 percent to attract those who would otherwise become accountants, lawyers, doctors and engineers? It bears mentioning that the percentage of spending that has been going to salaries and benefits has been slowly increasing over the last five years. So the question is no longer: Should we spend more on education? With 83 cents of every dollar going to labor, the question is: Should we pay education employees more money?

Of course, the term "education employees" covers a lot of ground. It includes everyone from superintendents to custodians, librarians and secretaries. The number of non-teaching school employees has become a bone of contention in the education policy debate. It enters into debates about administrative costs and decentralization of authority and responsibility for decision-making. Once again, advocates for one side or the other can manipulate the figures, usually by altering categories, in favor of their position. In order to avoid problems over who is an administrator, who is primary staff and who is support staff, etc., Table 5 classifies school employees by physical proximity to students. In other words, they are placed in categories determined by how far removed from the classroom they are. First we have teachers, whose primary work is done in the classroom. Then we have school staff, such as principals, assistant principals, support staff, library staff, instructional aides and guidance counselors, who are those people outside of the classroom but inside the school. Then we have district staff, who are officials, administrators and support staff who work away from the school but inside the region where the school is located. Finally, we have other staff, who are those education employees who work outside - or above - the district level.

Table 5 ranks each state and the District of Columbia in percentage terms. The state with the highest percentage of teachers among total staff is ranked number one. By adding the numbers to the right, the reader may draw ever larger concentric circles around the students. The percentages are based on full-time equivalent employment figures reported for Fall 1995. Some state's totals may not equal exactly 100% due to rounding. The percentages for the United States as a whole are: teachers - 52.0%, school staff - 19.8%, district staff - 4.6%, other staff - 23.6%.

Table 5 provides a good picture of exactly where each state's school system is applying its labor force. Though Vermont (at 49.1%) and Michigan (at 46.9%) both have relatively low ratios of teachers as a percentage of their total number of education employees, there are important differences between them. Vermont employs a higher than average percentage of staff at the school site. Michigan, on the other hand, has the highest percentage of employees outside the district level of any state.

California and Texas have an identical percentage of teachers (52%), but California has 6% of its employees at the district level compared to 1.3% of Texas employees. Texas leads California in employees outside the district - 28.3% compared to 19.1%. Why?

The purpose of examining these numbers is to determine whether the labor pool is applied in the most efficient place. Is a large percentage of employees outside the district reducing redundancy at the district level, or is it a bloated bureaucracy? Are large school staffs needed to free teachers to instruct students, or are too many specialists and administrators reducing the funds available to hire more teachers? There may be very good reasons why 11 states have more non-teaching education employees than they have teachers, but state officials should be made to defend them.

But even in states where teachers are outnumbered, the primary budgetary factors for schools are teachers' pay and benefits. No other line-item comes close. Support for increased education spending is a de facto call for increased spending on teachers. It is a reality well understood by the teachers' unions, but barely appreciated by the taxpaying public and its representatives.