The Most Expensive Failure How the American Education Establishment Spent Three Generations of Children
Updated: May 18
THE OUTLAW ARMORY
A Forensic Civic Accountability Publication
R. Bryant Starnes
Outlaw Livin' LLC
R. Bryant Starnes
Outlaw Livin' LLC
I. THE CASUALTY
An eleven-year-old American child, in the spring of 2026, can type. He cannot write. He cannot form letters with a pencil and produce a coherent paragraph by hand. His mother — herself a public-school teacher — discovered this at her own kitchen table. She is not the only mother who has made that discovery. She is one of many.
The child is not unusual. The 2024 National Assessment of Educational Progress, administered by the U.S. Department of Education, reports that forty percent of American fourth-grade students score below the NAEP Basic level in reading. A student below NAEP Basic, by the test administrator's own published definition, likely cannot identify a reason for a character's action that is implied in a short story. One-third of eighth-grade students score below NAEP Basic in reading — the largest percentage in the history of the assessment. A third of American thirteen-year-olds, in other words, cannot reliably identify the main idea, the order of events, or the traits of a character in a written text.
These are not opinions. These are findings of the federally chartered National Assessment Governing Board, released January 29, 2025, as official results of the 2024 NAEP.
This document examines how the American education establishment — federal, state, and local — produced this outcome while spending more public money on the project than any society in human history has ever spent. It examines who made the decisions, where the money went, and why the harm should be understood as a structural failure, not a budget shortfall.
The thesis is direct. The educated casualty was not produced by underinvestment. The casualty was produced by the deliberate misallocation of an unprecedented investment, by the institutional preference for adult employment over child outcomes, and by a thirty-year refusal to apply cognitive science already settled in the peer-reviewed literature.
The decision-makers responsible remain employed. The children responsible for nothing cannot read this sentence.
II. THE MONEY ARRIVED
The first myth that must be dispatched is the myth of underfunding. Every public discussion of educational outcomes opens with the claim that schools need more money. The historical record contradicts the claim at every measurement.
Per-Pupil Spending, 1970–2023
Per-pupil spending in American public elementary and secondary schools, expressed in inflation-adjusted constant dollars, has not merely grown since 1970. It has tripled.
1970: approximately $5,037 per pupil (2017 constant dollars), per NCES.
2017: approximately $12,794 per pupil — a 154 percent real increase.
2020–21: $16,280 per pupil (2022–23 constant dollars), per NCES Digest of Education Statistics, Table 236.55.
FY 2023: $16,526 average per pupil, per Education Data Initiative aggregation of NCES F-33 Public Education Financial Survey data.
The Reason Foundation, analyzing NCES historical data, calculates that inflation-adjusted K–12 spending per student has increased by 280 percent since 1960. Even at the most conservative inflation adjustment, the real cost of educating one American child has more than doubled in fifty years.
The ESSER Appropriation: $190 Billion
Between March 2020 and March 2021, the United States Congress appropriated, in three tranches, approximately $189.5 billion through the Elementary and Secondary School Emergency Relief fund:
ESSER I (CARES Act, March 2020): $13.2 billion.
ESSER II (CRRSA Act, December 2020): $54.3 billion.
ESSER III (American Rescue Plan, March 2021): $121.9 billion.
This single emergency allocation represented approximately three times the federal government's normal annual K–12 contribution. By the Education Commission of the States' own accounting, ESSER totaled roughly one thousand dollars per student per year for five consecutive school years on top of all other federal, state, and local spending.
The money was real. It arrived. It was disbursed. The thesis of underfunding cannot survive the appropriation record.
III. THE BUREAUCRACY ABSORBED THE MONEY
The money arrived. The classroom did not receive it. This is not opinion. It is staffing arithmetic.
Administrative Growth vs. Student Growth, 1950–2019
The Friedman Foundation for Educational Choice, using NCES data, calculated that between 1950 and 2015, in American public schools:
Student enrollment grew approximately ninety-six percent.
Teacher headcount grew approximately two hundred fifty-two percent.
Administrative and non-teaching staff grew approximately seven hundred two percent.
The administrative class grew at roughly seven times the rate at which the population it was hired to serve grew. The growth was not driven by enrollment pressure. It was driven by institutional preference.
The pattern was not historical. It accelerated. Center for Education Statistics data, analyzed by Hillsdale College and other independent reviewers, shows that between 2000 and 2019:
Student enrollment grew approximately 7.6 percent.
Teacher headcount grew approximately 8.7 percent.
District-level administrator headcount grew approximately 87.6 percent.
District administrators were hired at approximately ten times the rate at which students enrolled. The institution grew its own management layer faster than it grew the classrooms those managers were ostensibly hired to support.
Comparator: Public vs. Religious Systems
The American Enterprise Institute, citing NCES data, has documented that the New York City public school system employed approximately six thousand administrators while the New York Catholic school system employed approximately twenty-four. The public system had roughly four times the student enrollment of the Catholic system. The administrative ratio was approximately two hundred fifty to one.
Catholic school students, by every public measure of literacy and graduation, outperform NYC public school students. The system with one two-hundred-fiftieth the administrative overhead produces the better-educated child. The comparison forecloses any argument that administrative density is necessary to outcomes. The comparison supports the contrary inference.
Teacher Compensation
If administrative growth were funded by surplus, teacher pay would have grown alongside it. The historical record shows the opposite.
The Economic Policy Institute's annual teacher pay study, using Current Population Survey data, finds:
In 1996, public school teachers earned approximately 93.9 cents on the dollar relative to comparably credentialed professionals in other fields. The regression-adjusted gap was 6.1 percent.
By 2024, public school teachers earned approximately 73.1 cents on the dollar relative to comparably credentialed professionals. The regression-adjusted gap had widened to a record 26.9 percent.
The Reason Foundation, using Census Bureau data, calculates that inflation-adjusted average teacher salary declined by 0.6 percent between 2002 and 2020 — a period during which inflation-adjusted per-pupil spending rose by approximately 25 percent. The money grew. Teacher salaries did not.
Where did the increase go? The Reason analysis answers: between 2002 and 2020, while inflation-adjusted salary and wage spending per pupil rose by approximately $573, inflation-adjusted benefit spending per pupil rose by approximately $1,745 — three times the salary increase. The bulk of the benefit growth was concentrated in pension obligations, the largest single beneficiaries of which are senior administrators, not classroom teachers.
The Texas Teacher Retirement System carried approximately $62.6 billion in unfunded pension liabilities as of 2024. The California State Teachers' Retirement System carried approximately $85.5 billion. The administrative class, in setting its own retirement benefits, promised itself compensation the system did not fund. The unfunded balance comes out of present and future classroom budgets.
Where ESSER Went
If the structural pattern were limited to operational budgets, the emergency $190 billion might have been used to correct it. It was not.
Sworn testimony before the U.S. House Committee on Oversight and Government Reform, February 2024:
Rep. McClain: "Do you know where the money was spent? Was it spent on tutoring and extended learning?" Witness: "No, most of it was spent on labor, so increasing the permanent staff, and on pay raises."
Independent confirmation, from Edunomics Lab at Georgetown University and from the U.S. Department of Education's own June 2024 ESSER spending report: approximately fifty-three percent of disbursed ESSER funds were spent on staff salaries and benefits. Nearly half of ARP ESSER funds, specifically, were directed to permanent labor additions. The largest one-time educational appropriation in American history was used, by majority share, to grow the institutional payroll.
The Government Accountability Office report on ESSER spending, GAO-24-106913, confirms that twenty states declined to provide expenditure detail beyond top-line district totals. Twenty states will not, as a matter of public record, disclose how the federal money was spent. The transparency failure is not accidental. It is structural.
IV. THE METHODOLOGY WAS WRONG ON PURPOSE
Misallocated money is one harm. Misapplied science is a separate and graver harm. The American education establishment, for approximately three decades, taught reading using a method that the cognitive-science literature had identified as ineffective by the year 2000. The institutional response was not correction. The institutional response was continued purchase.
The National Reading Panel, 2000
The National Reading Panel, convened by the National Institute of Child Health and Human Development at the direction of Congress, published its final report in 2000. The report's central finding, derived from a meta-analysis of the existing peer-reviewed reading-instruction literature, was unambiguous: systematic, explicit phonics instruction is one of the five empirically validated pillars of effective reading instruction. Children must be taught the letter-sound correspondence directly. They do not learn it incidentally.
That conclusion has been repeatedly affirmed in the subsequent quarter-century by independent cognitive-science research, by neuroimaging studies of the reading brain, and by international literacy comparison data.
The Method the System Bought Instead
The dominant American reading curriculum from approximately 1995 through 2022 was "balanced literacy," a rebranded continuation of the discredited "whole language" movement. Its central operational technique was the "three-cueing system," in which a child encountering an unfamiliar word was instructed to:
Look at the picture on the page for a contextual clue.
Look at the first letter of the word.
Guess what word would make sense in the sentence.
Decoding the word phonetically — applying the letter-sound correspondence the National Reading Panel had identified as essential — was not the primary instructional path. It was either omitted entirely or treated, in the words of one independent reviewer, "like salt on a meal: a little here and there, but not too much."
Education Week's January 2020 survey of kindergarten, first-grade, and second-grade teachers found that only twenty-two percent of respondents believed phonics should be taught explicitly and systematically. Sixty-eight percent reported subscribing to the balanced-literacy framework. The instructional preference of the working teacher had been shaped, over a generation, by the administrative purchase decisions of district leadership.
The Procurement Apparatus
The methodology was not adopted by accident. It was sold. The dominant balanced-literacy product line was published by Heinemann, an imprint of Houghton Mifflin Harcourt. The dominant curricular framework, Units of Study, was developed at the Teachers College Reading and Writing Project at Columbia University. Independent reporting from American Public Media's APM Reports estimated that approximately one in four American elementary schools used the Units of Study curriculum at the peak of its market penetration. Independent estimates place the total number of American elementary schools using balanced-literacy curricula at approximately sixty-seven thousand.
Public-records-request data, compiled by APM Reports from one hundred large school districts, documented multi-year contracts. New York City alone reported spending approximately twenty-one million dollars over ten years on Heinemann materials. Gwinnett County, Georgia, reported approximately fourteen and a half million over the same period. The aggregate national expenditure on the disproven methodology, conservatively estimated, runs into the hundreds of millions of public dollars.
The Reckoning
In 2022, following the broadcast of the APM Reports series "Sold a Story," the dominant publisher's sales declined by approximately seventy-five percent in a single year. The Teachers College Reading and Writing Project was rebranded in 2024. Its founding director stepped down.
No public refund has been issued. No accounting has been published of cumulative public expenditure on the disproven methodology. The decision-makers who ratified the purchases — district superintendents, curriculum directors, state education agency officials — remain employed, in most cases in the same institutions whose children are now reading at historic lows.
The institutional response, in short, was to retire the brand and continue paying the salaries.
V. THE HARDWARE COMPOUNDED THE HARM
Concurrent with the failure of methodology, a second institutional decision compounded the developmental injury. American school districts, between roughly 2012 and 2022, transitioned from paper-and-pencil instruction to one-to-one device deployment in the elementary classroom. The decision was made at the district administrative level, executed through procurement contracts, and accelerated by ESSER spending. The decision was not subjected to a developmental-science review. It was subjected to a procurement review.
The Cognitive Science the Decision-Makers Could Have Consulted
The neuroscience of handwriting versus typing was not a matter of recent or contested research at the time of the deployment decisions. The findings had been published, replicated, and were accessible to any administrator inclined to read them.
Karin James, Indiana University, published functional MRI studies in 2012 and again in 2017 demonstrating that children who learned letters by handwriting activated the adult reading-circuit neural pathway — left fusiform gyrus, inferior frontal gyrus, posterior parietal cortex. Children who learned the same letters by typing did not.
Virginia Berninger, University of Washington, published longitudinal research demonstrating that handwriting fluency predicted composition quality and length independent of IQ. Children who could not automate the motor act of letter formation burned working memory on the mechanics and had less cognitive resource available for the substance of writing.
Audrey van der Meer and colleagues at the Norwegian University of Science and Technology published electroencephalography studies in 2020 and 2023 documenting widespread brain connectivity during handwriting that was absent during typing on the same content. Their conclusion, published in Frontiers in Psychology: handwriting is critical for memory and learning at the developmental stages in which American school districts had elected to replace it with typing.
The Screen-Time Literature
The Journal of the American Medical Association's pediatrics journal (JAMA Pediatrics), the most influential peer-reviewed clinical pediatric publication, has documented in successive studies between 2019 and 2024:
A 2019 longitudinal cohort study associating higher screen-time exposure in children ages 24, 36, and 60 months with lower developmental-screening performance.
A 2021 cohort study associating high school-age screen time with worse cognitive, executive function, and behavior outcomes in children born extremely preterm.
A 2023 cohort study of infants in the Growing Up in Singapore Toward Healthy Outcomes population, using electroencephalography markers, documenting altered brain development associated with infant screen use, with downstream cognitive consequences at school age.
A 2023 cohort study associating screen time at age one with delayed communication and problem-solving at ages two and four.
A 2024 systematic review and meta-analysis of screen-use contexts in early childhood, finding associations with adverse cognitive and psychosocial outcomes.
The findings were published in the most-cited peer-reviewed pediatric journal in the world. The findings were available to any school administrator before any 1:1 device contract was signed. The findings were not consulted.
The Reverse Flynn Effect
The most settled finding in twentieth-century cognitive science was the Flynn Effect: average tested IQ in developed nations rose approximately three points per decade from the 1930s through the 1990s. The cause was environmental — better nutrition, lead removal, expanded schooling, smaller families.
That trend has reversed. Compulsory military-conscription IQ data from Finland, Norway, Denmark, Estonia, the United Kingdom, France, the Netherlands, and Australia all show a declining trend beginning in the mid-1990s. The decline rates, depending on the country, range from approximately 0.38 to 8.4 IQ points per decade. Estonia, with the cleanest longitudinal dataset, registers the steepest decline.
The downward inflection coincides chronologically with the introduction of personal screens in childhood. Causation in such a complex association is contested. Correlation is not. The first generation of children raised on personal screens is the first generation, in the documented history of cognitive testing, to test lower than its own parents on validated cognitive instruments. The American educational establishment, in the same period, expanded screen deployment in the elementary classroom.
The Comparative Reversal
In 2023, the Swedish Ministry of Education, on the public recommendation of the Karolinska Institute, reversed a decade of tablet-based early-elementary instruction. Sweden restored physical textbooks and reinstated systematic handwriting instruction in early grades. The Karolinska Institute issued a public statement opposing the displacement of paper-based learning by screens at developmentally sensitive ages.
No equivalent reversal has occurred in the United States. American district administrators continue to expand device deployment. The procurement contracts continue.
VI. THE EARLY-CHILDHOOD ARCHITECTURE
The same institutional pattern that produced the K–12 outcomes documented above operates at the pre-kindergarten level. North Carolina, as a representative state, has built an extensive publicly funded early-childhood apparatus over the past three decades. Examining the apparatus on its own published evidence is instructive.
The Smart Start Funding Stream
Smart Start was established by the North Carolina General Assembly in 1993 under the administration of Governor James B. Hunt. The program operates through the North Carolina Partnership for Children, Inc. ("NCPC"), a 501(c)(3) nonprofit that contracts with the North Carolina Department of Health and Human Services. NCPC, in turn, allocates state appropriations to seventy-five local 501(c)(3) partnerships serving all one hundred North Carolina counties.
The funding flow is therefore: General Assembly appropriation → DHHS → NCPC → local partnership. The funding source, by statute, is the annual General Fund appropriation. State funding for early-childhood initiatives is provided primarily through Smart Start and the companion NC Pre-K program, with NC Pre-K drawing additionally on state lottery revenue.
By the NCPC's own published budget history:
Smart Start public funding peaked in Fiscal Year 2000–01 at approximately $231 million.
Funding stabilized at approximately $151 million per year following the 2008 recession.
In FY 2024–25, the Smart Start state appropriation totaled approximately $157 million.
Cumulative state expenditure on Smart Start since program inception, computed from published appropriation history, exceeds $4 billion in nominal dollars. With the NC Pre-K appropriation included — approximately $197 million per year in recent budget cycles — total state spending on the publicly funded pre-K and early-childhood apparatus is approaching $5 billion cumulative, with current annual expenditure in the range of $354 million.
The Evidence Base — Honestly Stated
Peer-reviewed evaluation of the Smart Start and NC Pre-K programs has produced measurable findings. The Frank Porter Graham Child Development Institute, the principal program evaluator, and Duke University, in independent longitudinal work, have documented:
Positive short-term effects of NC Pre-K participation on children's language, literacy, and mathematics skills at the end of the pre-K year, with effect sizes (Cohen's d) ranging from approximately 0.22 to 1.16 across measures.
A measurable boost in school readiness at the fall entry into kindergarten, with effect size approximately 0.22 in language and literacy.
By the spring of the kindergarten year, approximately eleven percent of the pre-K boost remained measurable. Eighty-nine percent had dissipated.
Higher-quality kindergarten instruction was found to be necessary to sustain the pre-K effect into later grades. Where kindergarten instruction was of average quality, the pre-K effect substantially disappeared.
These are the findings of the program's own evaluators. The honest reading: the program produces a short-term measurable benefit. The benefit substantially dissipates within twelve months absent compounding instructional quality. The receiving K–12 system has not, by the NAEP record documented in Movement I, provided that compounding quality.
The Output Mismatch
North Carolina has invested approximately five billion dollars over thirty years in publicly funded early-childhood services. Concurrent with that investment:
North Carolina's 2024 NAEP fourth-grade reading and mathematics results are not significantly different from the pre-program baseline.
North Carolina's 2024 NAEP eighth-grade reading scores show the same historic decline documented at the national level.
North Carolina's racial and socioeconomic achievement gaps remain materially unchanged from the pre-program era.
This is not an argument that Smart Start and NC Pre-K produce zero benefit. The peer-reviewed record establishes that they produce a short-term, measurable benefit at the end of the pre-K year. This is an argument that thirty years and roughly five billion dollars of cumulative state investment have not produced — and on the published evidence cannot produce — a durable cognitive or academic advantage in a K–12 system structured as the present one is.
The Administrative Pattern Repeats
Independent investigation of NCPC's executive compensation, published by the Civitas Institute (now the John Locke Foundation), documented that the NCPC president's total compensation exceeded the salary of the Governor of North Carolina and the United States Secretary of Health and Human Services, the cabinet officer through whose department the federal portion of the funding ultimately flows. The Secretary of HHS oversees a federal budget measured in trillions. The NCPC president, at the time of the Civitas filing, oversaw a state-program budget in the range of $188 million.
This is not an argument that nonprofit administrative compensation is per se illegitimate. It is an argument that the structural pattern documented in Movement III — administrative payroll growing at rates uncorrelated with classroom outcomes — extends from the public school system into the publicly funded nonprofit early-childhood apparatus.
The Internal Contradiction
The same publicly funded local early-childhood organization, in May 2026, distributed a single promotional flyer featuring two adjacent program panels:
Panel one promoted hands-on manipulative play — building blocks, stacking rings, bead-stringing — for developmental purposes. The promotional copy emphasized fine-motor-skill development, hand-eye coordination, and the developmental science underlying tactile early learning.
Panel two promoted a free two-week laptop loan program for families. The promotional copy stated that the laptops could be used for "job searching" in addition to children's learning.
The same organization, on the same single sheet of paper, simultaneously promoted the developmental science of fine-motor manipulation and the distribution of personal-computing devices into the homes of families with children in the developmental window during which fine-motor manipulation is most consequential. The contradiction was not addressed in the promotional copy. The organization did not appear to recognize it as a contradiction.
The flyer is not an outlier. It is representative. It depicts, in a single artifact, the institutional incoherence at the center of the publicly funded early-childhood apparatus: the same actors who can cite the developmental literature on tactile learning are simultaneously deploying the technology that the same literature identifies as developmentally harmful at the same ages.
VII. THE LOCAL VERDICT
The national pattern terminates in local classrooms. The local data is therefore the verdict on the national pattern. Examining one representative North Carolina school district, drawing exclusively on data published by the National Center for Education Statistics, by Ballotpedia's aggregation of NCES filings, and by U.S. News & World Report's compilation of state proficiency assessments:
The District in Question — by the Numbers
Enrollment, 2023–24 school year: approximately 18,110 students across thirty-two schools.
Full-time classroom teachers: approximately 1,124.
District administrators: approximately 19.
School-building administrators: approximately 83.
Total administrative personnel: approximately 102 — roughly 3.2 administrators per school.
Annual revenue: approximately $223 million.
Per-pupil spending: approximately $12,322.
Free-or-reduced-lunch eligible: approximately sixty-six percent of student body.
The Output, by State Proficiency Assessment
The receiving classrooms produce, on the state's own proficiency assessment, the following outcomes:
Elementary-grade reading proficiency: approximately 36 percent at or above proficient. Sixty-four percent of elementary students are not reading at grade level.
Elementary-grade mathematics proficiency: approximately 37 percent at or above proficient.
Middle-grade reading proficiency: approximately 30 percent at or above proficient. Seventy percent of middle-school students are not reading at grade level.
Middle-grade mathematics proficiency: approximately 26 percent at or above proficient. Seventy-four percent of middle-school students are not performing mathematics at grade level.
High-school reading proficiency: approximately 44 percent at or above proficient.
High-school mathematics proficiency: approximately 31 percent at or above proficient. Sixty-nine percent of high-school students are not performing mathematics at grade level.
The Verdict, Stated Plainly
The district has $223 million in annual revenue. It employs approximately one hundred two administrators across thirty-two schools. Two-thirds of its elementary children cannot read at grade level. Three-quarters of its middle-school children cannot perform mathematics at grade level.
This is not a budget failure. This is not a teacher failure. This is not a child failure. This is an administrative-decision failure, distributed across a generation of curricular, procurement, staffing, and methodology choices made by the rotating occupants of approximately one hundred two administrative chairs and by the state and federal apparatus from which those chairs draw direction and funding.
The district is not unusual. The national pattern documented in Movements II through V predicts precisely this local result. The local result therefore confirms, rather than departs from, the national diagnosis.
VIII. THE GENERATIONAL COMPARISON
The most credible witness against the present system is the historical record of the system it replaced. American children educated under the pre-balanced-literacy, pre-screen, pre-administrative-bloat instructional regime tested measurably better than the children educated under the present regime. The system that produced the comparison cohort is the control group. The system that produced the present cohort is the experimental group. The experiment has failed.
NAEP Long-Term Trend
The NAEP Long-Term Trend assessment, by design, is held constant across decades to permit direct generational comparison. The 2023 administration produced the following findings, on the test administrator's own publication:
Thirteen-year-olds, 2023, scored an average four points lower in reading and nine points lower in mathematics than the same age cohort in 2020.
Compared to a decade earlier, thirteen-year-olds in 2023 had declined seven points in reading and fourteen points in mathematics.
Reading performance among thirteen-year-olds is at levels not seen, by the test administrator's published statement, since the 1970s.
Mathematics performance among thirteen-year-olds is at levels not seen since the 1990s.
Nine-year-olds, between 2020 and 2022, posted the largest reading decline since 1990 and the first measured mathematics decline since the assessment's inception.
The cohort tested in 2023 has not merely failed to advance over the past decade. It has retrogressed by a generation.
The SAT
The Scholastic Aptitude Test, administered by the College Board, was the principal cognitive selection instrument used for American college admissions throughout the post-war period. Its score trajectory is instructive.
Average SAT verbal scores declined approximately fifty points between 1963 and 1979. Average mathematics scores declined and partially recovered, then declined again from approximately 2005 forward. In 1995, the College Board, by its own published announcement, re-centered the SAT scoring scale to restore the appearance of a mean of approximately 500 per section. The re-centering was an administrative response to score decline. Subsequent redesigns in 2005 and 2016 further altered the instrument.
The instrument was modified. The underlying performance was not improved. In the present era, an increasing proportion of American colleges have abandoned the SAT requirement entirely. The test signal that constituted the comparison instrument for generations of American educational performance has been progressively masked.
The Reverse Flynn Effect, Restated
Restated for emphasis: the most-validated finding in twentieth-century cognitive science — the steady generational rise in tested cognitive ability — has reversed in every developed nation with longitudinal military-conscription IQ data. Children born in developed nations after approximately 1975 test, on average, lower than their parents on the same validated cognitive instruments. The reversal coincides chronologically with the introduction of personal screens in childhood and with the curricular and administrative shifts documented in Movements III, IV, and V.
This is the most consequential single finding in this document. The American educational establishment is, by every available measure, producing the first generation in the documented history of cognitive testing to test lower than its own parents. The establishment has done so while spending more money than has ever been spent on the project.
IX. THE DECISION-MAKERS ARE STILL IN THEIR JOBS
The administrative class that ratified the balanced-literacy curriculum purchases remains employed, in most cases by the same institutions whose children are now reading at historic lows. The administrative class that approved the 1:1 device deployment remains employed, in most cases by the same institutions whose children are now showing the developmental and cognitive deficits the peer-reviewed pediatric literature predicted. The administrative class that, in the ESSER period, directed an emergency $190 billion congressional appropriation primarily to permanent staff additions and benefit increases remains employed, in most cases at the institutions to which the funds were directed.
No officer of any major American school district, state education agency, or federal education office has been publicly held accountable for the curricular, procurement, or staffing decisions that produced the outcomes documented in this Armory. No public refund has been issued. No retrospective accounting has been published of cumulative public expenditure on the disproven balanced-literacy methodology. The publishers of the disproven curriculum have lost market share. The administrators who purchased it have not lost their offices.
The institutional structure rewards growth, insulates failure, and externalizes the cost of the failure onto the population least able to bear it: the children, who did not choose the curriculum, the device, the staffing pattern, or the methodology, and who will carry the cognitive consequences of those choices through the entirety of their adult lives.
The children in the thirty-second school of the district profiled in Movement VII, two-thirds of whom cannot read at grade level, cannot read this Armory. That is not a rhetorical figure. That is a literal statement of the casualty count, taken from the state's own proficiency assessment.
The educated casualty is not an unfortunate outcome of an underfunded system. The educated casualty is the predictable, statistically documented, peer-reviewed-literature-confirmed product of a fully funded administrative apparatus that, for three decades, prioritized its own employment over the cognitive development of the children in its custody.
That is the finding of this Armory. The sources are in the public record. The arithmetic is unforgiving.
In a world of sheep, be an outlaw.
— R. Bryant Starnes
SOURCES AND CITATIONS
All numerical and factual claims in this document are sourced to the following published records. Where multiple sources support a single claim, the most authoritative is cited.
Federal Government Data and Audit
National Center for Education Statistics, Digest of Education Statistics 2023, Tables 236.55, 236.60, 236.65, 236.75 (per-pupil expenditure history).
National Center for Education Statistics, Common Core of Data (CCD), National Public Education Financial Survey, 2010–11 through 2020–21.
National Center for Education Statistics, NAEP 2024 Reading and Mathematics Assessment Results, Grades 4 and 8 for the Nation, States, and Districts (January 2025).
National Center for Education Statistics, NAEP Long-Term Trend Assessment Results: Reading and Mathematics, 2023 (age 13) and 2022 (age 9).
National Assessment Governing Board, "The Nation's Report Card Shows Declines in Reading, Some Progress in 4th Grade Math," January 29, 2025.
U.S. Government Accountability Office, Report GAO-24-106913, "K-12 Education: School Districts Reported Spending Initial COVID Relief Funds on Meeting Students' Needs and Continuing School Operations" (2024).
U.S. House Committee on Oversight and Government Reform, Hearing Wrap Up, February 2024 (ESSER spending testimony).
U.S. Department of Education, ESSER Spending Report, June 2024.
Peer-Reviewed Research
James, K. H. & Engelhardt, L., "The effects of handwriting experience on functional brain development in pre-literate children" (Indiana University, 2012, with subsequent 2017 replication).
Berninger, V. W., et al., longitudinal research on handwriting fluency and composition (University of Washington, multiple publications 1990s–2010s).
van der Meer, A. L. H., & van der Weel, F. R., "Only Three Fingers Write, but the Whole Brain Works: A High-Density EEG Study Showing Advantages of Drawing Over Typing for Learning" (Frontiers in Psychology, 2017); further EEG work, 2020 and 2023.
Mueller, P. A., & Oppenheimer, D. M., "The Pen Is Mightier than the Keyboard: Advantages of Longhand Over Laptop Note Taking" (Psychological Science, 2014).
Bratsberg, B., & Rogeberg, O., "Flynn effect and its reversal are both environmentally caused" (Proceedings of the National Academy of Sciences, 115(26), 6674–6678, 2018).
Multiple JAMA Pediatrics cohort studies, 2019–2024, on screen-time exposure and developmental outcomes.
National Reading Panel, "Teaching Children to Read: An Evidence-Based Assessment of the Scientific Research Literature on Reading and Its Implications for Reading Instruction" (National Institute of Child Health and Human Development, 2000).
Carr, R. C., et al., "Investigating if high-quality kindergarten teachers sustain the pre-K boost to children's emergent literacy skill development in North Carolina" (Child Development, 2024).
Peisner-Feinberg, E. S., et al., NC Pre-K program evaluations (Frank Porter Graham Child Development Institute, multiple publications 2011–2019).
Independent Analysis and Reporting
Hanford, E., "Sold a Story" (American Public Media Reports, six-episode series, 2022); "At a Loss for Words" (APM Reports, 2019).
American Enterprise Institute, Mark J. Perry, animated chart analyses of public-school enrollment, staffing, and per-pupil cost, 1970–2018.
Friedman Foundation for Educational Choice, Benjamin Scafidi, "The School Staffing Surge: Decades of Employment Growth" (multiple editions).
Reason Foundation, "Why teacher salaries are stagnant" (2025); "Inflation-Adjusted K-12 Education Spending Per Student Has Increased By 280 Percent Since 1960."
Economic Policy Institute, Allegretto & Mishel, "The teacher pay penalty reached a record high in 2024."
Civitas Institute (now John Locke Foundation), "How Much Does Your Smart Start Administrator Make?" (executive compensation analysis from IRS Form 990 filings).
Edunomics Lab, Georgetown University, ESSER spending analyses.
Education Week, January 2020 reading instruction teacher survey.
State and Local Public Records
North Carolina General Assembly, annual budget appropriations for Smart Start and NC Pre-K, FY 1993–94 through FY 2024–25.
North Carolina Partnership for Children, Inc., Annual Report to the North Carolina General Assembly, FY 2022–23.
Smart Start Rowan, Inc., IRS Form 990 filings (EIN 56-1890324), available through ProPublica Nonprofit Explorer.
North Carolina Office of the State Auditor, Smart Start Rowan, Inc. financial statement audit report.
National Center for Education Statistics, Search for Public School Districts, district detail for Rowan-Salisbury Schools (district ID 3704050).
Ballotpedia, Rowan-Salisbury Schools, North Carolina (compilation of NCES Elementary/Secondary Information System data, 2023–24).
U.S. News & World Report, Rowan-Salisbury Schools district profile (state proficiency assessment compilation, 2021–22 through 2023–24).
The Outlaw Armory — Forensic Civic Accountability Publications — outlawlivin.com
Comments