In 1969 a company was paid eighty dollars every time a child gained a grade level in reading. It taught them the test.
* * * * *
“If a nation expects to be ignorant and free, in a state of civilisation, it expects what never was and never will be.”
— Thomas Jefferson to Charles Yancey, January 6, 1816.
* * * * *
This essay is Part III of VI. Links to Parts I and II are right here. Click to go to the part you want. Part I Part II Part III
* * * * *
In October 1969, in a school district on the Texas–Arkansas line, a company was paid eighty dollars every time a child gained one grade level in reading.
The man who wrote that contract had come from the Office of the Secretary of Defense. I spent four months tracing him, because this series has been arguing that a method built at the Pentagon under Robert McNamara escaped and colonized the institutions that came after, and here at last was a man carrying it out of the building by hand.
He was carrying it home.
American schools did not learn quantified management from the Defense Department. They had been doing it for half a century before McNamara arrived.
Raymond Callahan set this out in 1962, in a book called Education and the Cult of Efficiency, and it has never been comfortably answered. His subject is what American school administrators did between 1910 and 1929, and what they did was adopt the values and the techniques of the business-industrial world wholesale.
The instrument was the school survey.
Between 1911 and 1925, Callahan writes, so many surveys of schools were made that there was hardly a state or local system in America that was not surveyed. Boise was done in a single week in 1913 by three university men, and its superintendent took the results to the national meeting the following year. The criterion those surveys applied was not the quality of the product. Callahan’s summary of the record is that the emphasis was on the lowest cost.
Superintendents learned to defend themselves in the language of business because business was what the school boards and the newspapers understood, and the pressure was financial and relentless. His verdict is that the men who could have resisted were the university leaders — Strayer, Cubberley, Bobbitt — and that they moved with the tide.
He calls what followed a fourfold tragedy:
- educational questions subordinated to business considerations;
- administrators produced who were not in any true sense educators;
- a scientific label put on unscientific methods; and
- an anti-intellectual climate strengthened.
Three school districts serve Texarkana, which straddles a state line: Texarkana Arkansas, Texarkana Texas, and Liberty-Eylau.
In 1968 they shared one problem. Children were leaving. The federal remedial money had been arriving since 1965 and nobody could say what it had bought, because nobody was measuring anything except whether the money had been spent. Desegregation guidelines were arriving too, and the Texas district would withdraw from what followed rather than deal with them.
In December of that year, two districts began talking to a consultant named Charles Blaschke.
He had taken economics at Texas A&M and a master’s in public administration at Harvard, and had spent the years since moving between offices that were all, in one way or another, about whether government could tell what it was buying. Two of those years, from 1966 to 1968, he served in the Army attached to the Office of the Secretary of Defense, and what he did there was study how the Department of Defense used and managed educational technology — how the largest training organization in the world taught people things, and whether it could tell.
He was not a systems analyst, and the record does not let anyone call him one.
What he proposed was that the districts stop buying teaching and start buying results.
The contract went to Dorsett Educational Systems of Norman, Oklahoma, a maker of teaching machines. Six Rapid Learning Centers, two class periods a day, remedial reading and mathematics. Dorsett guaranteed one grade level of measured gain per subject within eighty hours of instruction, for eighty dollars — or proportionally for each fraction thereof, on a sliding scale that paid more for faster. The ceiling was a hundred and thirty-five thousand dollars.
The first center opened on October 15, 1969. By the end of the year, three hundred and fifty-one children had passed through them.
Here is what it is like to stand in one. The desks face a machine that holds a filmstrip, a small speaker, and a row of buttons. A child presses a button and a counter increments. There is no lesson plan on the wall, because there is no lesson — there is a program. Edward Trice, superintendent of the Arkansas district, described what had happened to the adult in the room: the role was altogether different from the role of the teacher in a traditional classroom, and she could be called an instructional manager.
● ● ●
Results were announced in February 1970, four months in, and they were reported as gains of a grade level and more. Educators came from across the country to look.
On March 3, 1970, President Nixon delivered an education message that took up accountability as a national watchword, and the word entered American schooling as a term of art. Leon Lessinger, who administered the federal program at the U.S. Office of Education that had funded Texarkana, became its most visible champion. Writing in 1974, a religious educator named Gerald Teller traced the whole thing back in a sentence: during the Kennedy years, through the efforts of Robert McNamara, the idea of business efficiency was brought to government, and Lessinger was largely influenced by that movement.
Teller had it half right, which is the interesting half — and he knew it, because the same short article quotes Callahan’s preface on the weakness and vulnerability of school men before the business ideology. He had both halves in front of him in 1974 and joined them the wrong way round. The efficiency idea had been in American schools since before the First World War. What McNamara did was give it a method and a prestige it had never had, and it was the prestige that travelled.
On April 17, 1970, a letter went out from an office on K Street to a man at the World Bank. It began, “Dear Bob.”
Arthur W. Barber was president of the Institute for Politics and Planning, the consultancy that had carried Blaschke into Texarkana. From 1962 to 1967 he had been Deputy Assistant Secretary of Defense for International Security Affairs. He was writing to Robert McNamara, who had been his Secretary of Defense for all of it. Two of the Institute’s eleven trustees had run pieces of McNamara’s Pentagon; the other nine included a senator, a governor, two union presidents, and a theologian.
Barber had read McNamara’s Columbia University address of February 20th and wanted to offer criticism. McNamara had said that as the Seventies opened, in field after field there were more questions than answers, and the urgent need was for new instruments of research and analysis. Barber told him that what was needed was not more research instruments but performance criteria and a measurement of achievement — that the almost complete absence of measurement of educational achievement and post-education employment or careers was a world-wide scandal, and that education had to be broken into functional objectives with a framework of incentives to achieve them at the lowest possible cost.
Then: We have done this in reading and math. Behind the letter in the folder sits a description of the Texarkana project.
I want to be careful here, because my first draft of this passage was wrong and the way it was wrong is worth more than the passage was.
I had Barber correcting McNamara — one man wanting instruments that describe, the other wanting instruments that pay.
It is a tidy opposition and it does not survive the record. Six weeks into the job, in a private list of things to do, McNamara had written that the Bank should replace its practice of considering loans one at a time with a five-year program based on systems analysis and overall development strategy, directed from the top. He then did it. He announced in September 1968 that Bank lending would double over five years. He created a Programming and Budgeting Department that reported to him. In 1969 he introduced the Country Program Paper, drafted to conform to the lending targets and used as the basis for lending decisions. Sharma, his best historian at the Bank, says plainly what the model was:
McNamara had built a planning and budgeting system at the Pentagon to tame the generals, and he built its equivalent in Washington to do the same to the Bank.
He was already running exactly the kind of regime Barber was selling. Within a few years his staff would be hurrying loans through at the end of reporting periods to make the numbers, and would have a word for it.
So the letter is not a criticism. It is an escalation, written by a man who knew his correspondent would agree.
And I had made a smaller mistake inside the larger one, and the correction belongs in the article rather than in a footnote. I wrote that Barber quoted McNamara accurately and stopped one clause early. He did. The sentence in the speech runs on: with which to dispel our ignorance of the social dimensions of economic change, and help us formulate a more comprehensive strategy for the decade ahead. Barber stopped one clause early. I stopped two. In the paragraph where I was accusing him of trimming a quotation to make it mean what he needed.
There is no record that McNamara replied. The letter was received, routed, and filed.
Three months later the program came apart.
A junior high student taking the May post test remarked, within hearing of the project director, that he was tired of answering the same old question. Dorsett’s instructional materials, written that spring by the company’s chief programmer, contained items from the test that determined whether Dorsett got paid. Dorsett said 6.5 percent of items were affected and that the results were not greatly affected. The district’s evaluator, applying rules supplied by Educational Testing Service, ruled the posttests invalid on July 20, 1970. Thirty thousand dollars of the hundred and thirty-five was withheld. Dorsett sued for it in 1972.
Trice defended the program to the end, and his defence deserves to be heard: eight of eight hundred potential dropouts had left in two years, against what he calculated as a normal rate of two hundred. He called the testing incident minor but unfortunate.
He may have been right about the dropouts.
What could no longer be said was whether the children could read.
● ● ●
The idea outlived the contract, as ideas do when the thing that fails is only the mechanism.
Performance contracting spread — perhaps as many as fifty projects in 1970-71, though the counts differ and none is firm. The Office of Economic Opportunity ran a national experiment across twenty sites at a cost somewhere near six million dollars and announced at the end of January 1972 that it had not worked. That verdict is where I overreached in my first draft: the General Accounting Office expressly declined to endorse it, holding that shortcomings in the design and implementation left the merits of performance contracting unanswered. Educational Testing Service’s own Edward O’Connor and UCLA’s Stephen Klein argued in 1973 that contractors had been given lower-achieving students and that the design was pre-experimental.
So the honest sentence is that the proposition was tested inconclusively, the mechanism was dropped, and the assumption underneath it was never examined at all.
One thing in this record will not resolve, and it is the thing I most wanted to resolve. I do not know whether performance contracting collapsed because it was wrong or because it was crude — because paying a vendor by the grade level is a blunt way to do something that might be done well.
What I can see is that nobody went back to check. The apparatus moved on to statutes.
It moved through the states first. Florida authorized a required graduation test in 1976 and gave it in October 1977, though the superlative usually attached to it is Florida’s own and Oregon has a claim. Goals 2000 and the Improving America’s Schools Act built federal scaffolding in 1994. Texas Senate Bill 7 in 1993 established subgroup reporting, state-set standards, incremental targets, and a ladder of sanctions, and No Child Left Behind carried a version of that architecture nationally in 2002 — drawing, its own architects said, on Texas and North Carolina and Massachusetts, and on a decade of argument from civil-rights organizations and business coalitions and congressional Democrats.
It is a wider lineage than Austin alone.
What it produced in the school day is countable.
In its fifth-year survey the Center on Education Policy asked 349 districts what had changed. Thirty-six percent had cut social studies, by an average of seventy-six minutes a week. Twenty-eight percent had cut science. Twenty percent had cut recess. Among districts with a school identified for improvement, fifty-one percent had cut social studies, by an average of ninety minutes. And among the forty-four percent of districts that reported both adding time to reading or mathematics and cutting it somewhere else, the total came to a hundred and forty-five minutes a week — about a third of the time those subjects had held before.
That last figure describes a subset of districts, and I have seen it quoted as though it described every child.
What it describes is a large minority of American school districts, and it was produced by administrators who broke no rules.
Now the case against everything I have written, at its strongest, and it does not come from one witness.
It comes from a coalition. Through the 1990s the Education Trust, the Citizens’ Commission on Civil Rights, the Center for Law and Education, and the Leadership Conference on Civil Rights argued that state and local discretion had produced low expectations, weak tests, and aggregate averages that hid the failure of poor and minority children inside respectable district means. Their argument went past publishing the data. Publication without consequence had already been tried and had already failed — that visibility alone would not overcome institutional resistance, and that the federal government would have to attach something to it.
John Cole, president of the Texas Federation of Teachers, had watched the thing they were describing. He grew up in Corpus Christi and taught in its schools, segregated until 1976, in the Hispanic junior high, where he found no standards for coursework at all. What he describes among school board members and superintendents is a conspiracy of silence about which schools and which children were failing.
And the accountability regime those arguments produced did some of what they hoped. David Deming and his colleagues found that Texas schools under pressure at the low-performing threshold raised scores, and that the students affected were later more likely to attend college, finish four-year degrees, and earn more at twenty-five.
The same study found harm to the lowest-scoring students. Thomas Dee and Brian Jacob found NCLB raised fourth-grade mathematics on the NAEP — an external test with nothing attached to it — by about a fifth of a standard deviation.
Those gains are real and I am not going to write around them. Children learned things they would not otherwise have learned, and some of them are better off in measurable ways decades later. Any account of this period that cannot hold that alongside everything else is propaganda.
● ● ●
Donald Campbell had described the whole mechanism before Texarkana finished failing.
In 1975, in a chapter for a volume out of a joint Dartmouth and OECD conference, Campbell set out what he called his pessimistic laws: the more any quantitative social indicator is used for social decision-making, the more subject it will be to corruption pressures, and the more apt it will be to distort and corrupt the social processes it is intended to monitor.
He illustrated it seven times. Chicago’s voting statistics. Police clearance rates. The Nixon crime crackdown. Employment-office creaming. Then achievement testing, and Texarkana by name. Then Soviet production quotas. And then, introduced as a final example, the Vietnam body count.
The first essay in this series as well as this one are in that list, written down by a man who saw in 1975 that they were the same thing happening twice.
His claim was comparative — the more used, the more subject, the more apt — and he said his evidence was predominantly anecdotal, and he proposed remedies. On achievement tests specifically he was blunter: when test scores become the goal of the teaching process, they lose their value as indicators and distort the educational process in undesirable ways.
The warning was taken up by exactly the people who should have taken it up. Daniel Koretz built a working account of score inflation on it and designed responses. The National Research Council turned it into recommendations — evaluate high-stakes systems against independent low-stakes tests, use multiple measures, follow students into later outcomes. The warning was heard, understood, operationalized, and made actionable by serious people who knew what they were doing.
And in 2002 the Congress of the United States passed a law with a single proficiency measure, a ladder of sanctions, and a requirement that every child in the country be proficient by 2014.
That is the finding, and it is worse than being ignored. The warning worked as scholarship and lost as politics, and it lost to people who had it on their shelves.
Kuznets was quoted and not followed. Campbell was read, operationalized, and overruled. There are two ways to lose an argument you have already won, and this series has now found both.
Here is what the instrument does not reach.
Chapter One of a book I wrote last year argues that the founders meant this country to be different, and that the difference was education — not literacy alone but discernment. Citizens who could read the laws and question them, spot a lie and name it, hear a demagogue and turn away.
I had it in my head that none of those is a test item, and that is wrong. The NAEP civics assessment devotes between a third and a fifth of its time to evaluating, taking, and defending a position, with constructed responses as well as multiple choice. The capacities can be assessed. Somebody built an instrument to do it.
The problem is not that discernment cannot be measured. It is that when the day got shorter, it was not the thing being measured.
December 20, 2004. An assistant principal at a Philadelphia elementary school writes to a group of fifth- and sixth-grade teachers. Please suspend the teaching of social studies until April, after the testing period. A word of caution — our fifth grade math scores on the benchmark tests are dismal, so we will double our math time to help our students become proficient in math. The email ends, as emails do, with a sign-off.
Nobody in that building did anything wrong. A test was coming, the scores were bad, and the school day is a fixed quantity.
Named people said the same and defended it. Samuel Harris, a principal in Sacramento, restricted about a hundred and twenty-five of his lowest-performing students to reading, mathematics, and gym, and said that when you look at a kid and you know he can’t read, that’s a tough call you’ve got to make. Henry Lind, a superintendent in Cuero, Texas, doubled reading and mathematics for his sixth graders and reduced other subjects to do it: if you can’t make layups, then you’ve got to work on layups.
Both men are describing triage, and triage is a defensible thing to do when you can see the bleeding.
I built one of these. Field Studies came out of studying what corporations are looking for in terms of critical skills, and what I was doing was making legible the qualities employers could describe perfectly well and could not count. What I am less sure of, at eighty-five, is that visible and countable are the same operation, and the reason I am less sure is that I have spent four months watching what happened to a school district that assumed they were.
The full accounting is the last essay in this series. I am not looking forward to writing it.
● ● ●
You are inside this. There is a number you answer to, you know what it misses, and you know the parts it misses are not the small ones, and you have probably explained that to someone who nodded and then asked what the number was.
If you have children in an American school, you may be inside it twice. Somewhere in a large minority of American districts, thirty minutes came out of a day, correctly, under the rules, and what went was the part with no number attached. Whether that happened to your child is a question you would have to ask, and the honest answer is that most parents never did, because nothing arrived in the post to tell them.
Now the correction I began in the last essay and have to finish here.
I have been arguing that a method built at the Pentagon in 1961 escaped into the institutions that followed.
In education that is backwards. American schools were running cost accounting and efficiency surveys in 1915, half a century before McNamara took office. What arrived in Texarkana in 1969 was not an infection from the Defense Department. It was the same idea coming home with better credentials — carried by men who had spent the sixties watching the most prestigious organization in American government do arithmetic, and who brought back not a technique but a warrant.
That is a smaller claim about transmission and a larger one about everything else. If education invented this and re-imported it, the Pentagon is not the source of the disease in this series; it is the place the disease acquired its authority.
And that is worse, because a method that spreads by contagion can be quarantined — you find the carrier and you stop him — while a method that keeps being reinvented, in one institution after another, by capable people solving real problems with the best tools on the shelf, cannot be quarantined at all. Callahan’s superintendents in 1915 were serious men. So was Blaschke. So were the civil-rights lawyers who wanted consequences attached to the numbers. Each of them faced an institution that could not say what it was doing, and each reached for the only instrument there was.
A hostile reader will say the Texarkana instrument was never built to describe — that the contract wired the test to a payment from the first day, so nothing was promoted.
Look again at what got wired. The contract was built to decide. The test was an achievement test of the ordinary kind, built to tell a teacher what a child could do — and the moment eighty dollars depended on the number it produced, it stopped being able to tell anyone anything. That is not a metaphor about instruments in general. It is what happened, in one room, in May 1970, when a boy said he was tired of answering the same old question.
The instrument was built to describe.
It was promoted to decide.
* * * * *
“I know no safe depository of the ultimate powers of the society but the people themselves; and if we think them not enlightened enough to exercise their control with a wholesome discretion, the remedy is, not to take it from them, but to inform their discretion by education. This is the true corrective of abuses of constitutional power.”
— Thomas Jefferson to William Charles Jarvis, September 28, 1820.
Next: the examining room, where the thing being counted is a body.
* * * * *
Charles C. Jett is an author, civic educator, and Professional Certified Coach based in Chicago. A graduate of the U.S. Naval Academy (Class of 1964) and Harvard Business School, he served during the Cold War aboard the nuclear submarine USS Ray (SSN 653), where he developed the Geographic Plot, a contact-tracking method for submarine operations. He is the author of six books, including Super Nuke!, hosts four podcasts, and writes across his Critical Skills Blog platform on history, leadership, and the health of the American republic. He and his wife, Dr. Nancy Church, live and co-host the Chicago Salons at Water Tower Residences.