The School Staffing Surge, Hiring Outpaced Enrollment for Decades

Walk into almost any school office in America and count the desks. Then walk down the hall and count the classrooms. If the ratio surprises you, you have just done what this post does with numbers.

My claim is a plain one. Over the last several decades, American public schools have hired non-teaching staff far faster than they have enrolled students, and that was a choice, not a law of nature, and the money it consumed is money that could have gone to the people who stand in front of the children. I will build it from federal staffing and enrollment data and the work of an economist, Benjamin Scafidi, who spent years assembling the figures. I will show every piece of arithmetic, separate the hiring that the law requires from the hiring that was discretionary, and then say where a district could redirect dollars.

The terms

An FTE is a full-time equivalent, a way of counting staff so that two half-time employees count as one. The National Center for Education Statistics publishes counts of public school staff in FTEs, alongside enrollment, in its Digest of Education Statistics.

Teachers are the instructional staff who teach classes. Non-teaching staff is a broad category: administrators and administrative support, instructional aides, guidance and student-support personnel, and the rest. It is not a synonym for waste. Many of these jobs are useful and some are legally required. The question is not whether any of them should exist. It is how many, and in proportion to what.

Growth multiplier is my shorthand for the ratio of a later figure to an earlier one. If a number rises by 100 percent, the multiplier is 2.0. I will use it because percentages above 100 confuse people, and multipliers do not.

The numbers

In 2012, the Friedman Foundation for Educational Choice (now EdChoice) published Benjamin Scafidi's report, The School Staffing Surge: Decades of Employment Growth in America's Public Schools. Scafidi was then an economist at Georgia College & State University. His method was unglamorous and valuable: he took the NCES series on students and staff from 1950 to 2009 and compared their growth.

The figures he reported:

• Student enrollment grew by about 96 percent.
• The number of teachers grew by about 252 percent.
• The number of administrators and other non-teaching staff grew by about 702 percent.

Convert those to multipliers. Students: 1 + 0.96 = 1.96. Teachers: 1 + 2.52 = 3.52. Administrators and other staff: 1 + 7.02 = 8.02.

Now the ratios, which are the part that matters.

• Teachers per student. The teacher multiplier divided by the student multiplier is 3.52 / 1.96 = 1.80. There are about 1.8 times as many teachers per student as in 1950. In other words, classes are smaller. That is a real change and, I think, a good one, within limits.
• Non-teaching staff per student. The multiplier for non-teaching staff divided by the student multiplier is 8.02 / 1.96 = 4.09. There are about four times as many non-teaching staff per student as in 1950.

Read those two lines again. Staffing grew faster than enrollment on both counts, but the non-teaching category grew more than twice as fast as the teaching category: 8.02 / 3.52 = 2.28. If you are a taxpayer asking where the additional money went, that last ratio is the place to start.

Scafidi later updated his analysis in follow-up reports that looked at more recent decades. I will not quote the later figures here from memory. Look them up, because the pattern was similar: staffing growth outpaced enrollment growth, and non-teaching positions grew fastest. The point is robust to the choice of end date, and I have been deliberate about showing you the 1950-to-2009 comparison, which is the one most widely cited.

What those ratios do not say

Honest analysis owes the reader the limits of its evidence, so let me state them.

First, these are national averages, and they hide large differences between districts. A rural district with a small enrollment and a big geographic footprint has different needs from a city.

Second, a ratio is not a verdict. In 1950 many functions that now require staff either did not exist or were done by teachers, parents and volunteers. A school today may employ a nurse, a counselor, a technology specialist and a bus-route coordinator where a school in 1950 had none of them. Some of that is progress.

Third, the categories overlap in practice. Instructional aides, for example, work in classrooms alongside teachers, so counting them as "non-teaching" is accurate but incomplete.

None of this saves the conclusion that the system grew its overhead faster than its core. It only means we should ask, position by position, what each addition was for.

The steelman: it was the law

The strongest reply to the staffing argument is that Congress and the courts required much of the hiring. This reply is true in part, and it deserves a fair hearing.

In 1975, Congress passed the Education for All Handicapped Children Act, later renamed the Individuals with Disabilities Education Act, or IDEA. It required public schools to provide a free appropriate public education to children with disabilities, with individualized education programs, related services and, where appropriate, instruction alongside non-disabled peers. Before that law, many children with disabilities were excluded from public schools or poorly served. The law corrected a real wrong, and it cost real money, in aides, therapists, psychologists and specialists.

Other requirements followed: services for students learning English, health-related mandates, reporting and compliance duties tied to federal and state programs, and the paperwork that comes with accepting federal dollars. A district that wants to keep its funding must employ people who understand the rules.

I accept the point and credit it. It explains a significant part of the growth in non-teaching positions, and anyone who tells you that every non-teaching hire was frivolous is not being serious.

The answer: mandates explain some of the surge, not all of it

Here is how I would answer, in four points.

1. The timing does not fit a single cause. The special education mandates began in 1975, but the series Scafidi used starts in 1950 and covers six decades. A single legal change does not obviously explain growth that spans that long.
2. Special education is not the whole non-teaching category. The non-teaching category includes central-office administration and general support positions that no special education mandate requires, and Scafidi has argued that if mandates were the whole cause, the growth would not have been so widespread across states and districts.
3. Discretion exists even inside a mandate. The law requires services for each child's needs. It does not dictate that a district must add a layer of central administrators to coordinate them. The mandate fixes the obligation, but the district chooses how to meet it, and the options differ widely in cost.
4. The money has an opportunity cost. Every dollar spent on a position is a dollar not spent on something else. If that dollar had gone to teacher pay, it would have shown up in the paycheck of the person the research tells us matters most.

The right way to say this is: some of the non-teaching hiring was required, some was wise, and some was discretionary. The honest policy question is how to tell them apart.

Where the money would have gone

Now the opportunity cost, with a labelled hypothetical, since I will not pretend to know your district's budget.

Picture a district, which I will call hypothetical because it is invented for the arithmetic, with 10,000 students, 500 teachers and 600 non-teaching staff. Suppose the average non-teaching employee costs the district $90,000 a year in salary and benefits. These are made-up round numbers.

• Suppose the district trims non-teaching positions by 10 percent through attrition, not layoffs: 10 percent of 600 is 60 positions.
• Savings: 60 x $90,000 = $5,400,000 a year.
• If every dollar went to the 500 teachers, each would receive $5,400,000 / 500 = $10,800 a year.

That is a raise of the size most teachers have never seen, funded without a tax increase. I want to be clear that this is an illustration of method, not a forecast for any real district. Real districts have contracts, pension obligations and legal duties, and a 10 percent trim might be too much or too little. But the structure of the calculation is the thing to notice: when overhead grows faster than the classroom, there is room in the budget to pay teachers more without asking the taxpayer for a cent.

This is the heart of the conservative argument. Teachers are underpaid in the sense that their salary schedule gives them little reward for excellence and much for tenure. At the same time, the system carries overhead that would amaze a visiting auditor. The two facts are connected by the budget.

What a district could actually do

Here is where I think a board or a state legislature could act, in rough order of how little controversy it causes.

• Publish the ratios. Require every district to report, on one page of its website, students per teacher, students per non-teaching staff member and the trend over ten years. Transparency costs nothing and invites the right questions.
• Fill vacancies in the classroom first. When a position opens, require the superintendent to justify why it is not a teaching position.
• Audit the central office. Hire an outside reviewer to compare the district's administrative staffing to that of similar districts with similar results.
• Separate the mandated from the discretionary. Make the distinction visible in the budget, so that the board and the public can tell which positions the law requires and which the district chose.
• Pay for results as well as seniority. Direct the savings toward raises for teachers in hard-to-staff subjects and schools, and for teachers whose students grow the most. We look at why the salary schedule makes this hard in What a Teachers Union Contract Says and Why Every Parent Should Read One.
• Consider competition. When families can take their child's funding to another school, a district has a powerful reason to keep its costs in line. See Tax-Credit Scholarships, the Choice Program Built on Private Donations for one version of that idea.

Where this fits in the larger argument

This post is one half of an argument. The other half is in Real Spending Per Pupil More Than Doubled Since 1970, So Where Are the Gains: spending per pupil rose steeply, and outcomes did not. The staffing numbers are an explanation of part of where the money went. Together they say the same thing: the system's resources grew, but the growth did not land in the classroom.

There is also a constitutional point, one that does not concern the Constitution's text directly but the principle behind it. Madison wrote in Federalist 51 that if men were angels, no government would be necessary, and that in framing a government to be administered by men over men, the great difficulty is first to enable the government to control the governed and in the next place to oblige it to control itself. A school district is a government. It controls itself only if the people who pay for it can see what it does. The first step in that process is to count the desks.

You can see how the principle applies at the national level on the national numbers page, and the logic of bureaucracies that grow on their own is the subject of the regulation bill a ten-person company pays that a giant can shrug off in a different setting. For a broader look at how school budgets divide, Where Tuition Dollars Go, From Administrators to the Lazy River runs the same test on colleges, and Who Actually Runs Your School District, Board, Superintendent or State explains who has the authority to change any of it.

The objection from the other side, once more

Someone will say that I am attacking the people who staff the schools, the counselors and aides who work with children every day. I am not. A counselor who helps a struggling teenager is doing something valuable, and an aide who sits beside a child with a disability is doing something the law rightly requires. My argument is with the structure of incentives that adds layers of overhead without anyone having to prove they are needed, and with a budget process that treats the classroom as the last place to spend. Defend the individual jobs if you wish. But a budget is a set of choices, and you cannot defend every choice by pointing to the person it employs.

The case in brief

1. Between 1950 and 2009, by Benjamin Scafidi's analysis of federal data, student enrollment grew about 96 percent, teachers about 252 percent and non-teaching staff about 702 percent.
2. Per student, that is about 1.8 times as many teachers and about 4.1 times as many non-teaching staff, so overhead grew more than twice as fast as teaching.
3. Special education mandates beginning in 1975 explain part of the growth, but the length of the trend and the mix of positions suggest that a good share was discretionary.
4. Illustrative arithmetic shows how a modest trim in overhead could fund teacher raises without new taxes.
5. The fix is transparency, classroom-first hiring and pay that rewards results, which are choices a board can make.

What you can do: ask your school board to post the district's students-per-teacher and students-per-non-teaching-staff ratios for the last ten years, and read them before the next budget vote.