◆ The Lost Art
JesseW22.com ↗

An investigation · JesseW22.com · October 2026

The Lost Art of
Critical Thinking

When did American students’ reasoning skills start slipping, what caused it, and who made the decisions? This piece follows the evidence from 1957 to 2026.

2012Long-term-trend peak for 13-year-olds
2015First broad national score drop
−15Points in age-13 math, 2012→2025
146Numbered sources
The record

What the documents and data show.

Analysis

Interpretation and judgment, labeled as such.

Figures above come from NAEP long-term-trend data (Chapter 09). About 42 minutes to read · 9,722 words · Works offline; videos load only on request.

Part IThe question
Chapter 01

Introduction

What this piece covers, and how to read it.

Many parents and grandparents ask the same question: are American kids being taught to think less well than they used to be? If so, when did it start, and who made the decisions? This piece follows the evidence from 1957 to 2026. It takes positions where the evidence supports them and stays careful where it doesn’t.

The record shows the slide started years before the pandemic. Measured skills peaked around 2012–2013 and were already falling by 2015 . COVID made it much worse, and it did something else that matters: it put the classroom on the kitchen table, where millions of parents saw what their kids were being taught. So COVID gets a big chapter, but it isn’t the starting point.

The piece also tests the most common explanations people offer: the 2009–2017 period, teachers’ unions, education schools, gender policy, critical race theory, the claim that tests or “objectivity” are white supremacy, and several others. Each is checked against primary documents: federal filings, test data, court rulings, the actual training materials, and peer-reviewed studies. Some claims hold up, some hold up only in part, and a couple do not.

How to read this

Two kinds of boxes run through the piece, kept deliberately separate:

The record

What documents, data, and studies show. Every claim carries a numbered source chip. Tap or click one to see the source and open it.

Analysis

Interpretation and judgment. You can disagree with these without disputing any fact in the record boxes.

Specific organizations, documents, policies and people, from any side of politics, are named when the record ties them to a decision. Motives such as hatred or jealousy are not assigned without evidence, and no such evidence turned up. Most importantly: students are never the cause in this story. They’re the ones who paid for other people’s decisions.

Use the contents to jump around, the search box to find a topic, and the Save place button to come back later. The interactive timeline lays out the whole story from 1957 to 2026.

Chapter 02

What we mean by critical thinking

A working definition, and why knowledge is part of it.

John Dewey described reflective thinking in 1910 as holding a belief up to the grounds that support it and the conclusions it leads to, rather than accepting it because it is familiar . Edward Glaser’s 1941 definition added three parts: a willingness to think problems through, knowledge of how to reason and weigh evidence, and the skill to apply both .

Cognitive scientists add something crucial. You cannot think critically about a subject you know nothing about. Daniel Willingham’s summary of the research is blunt: critical thinking is not a free-floating skill that transfers from one topic to another; it depends on knowing the subject . A large meta-analysis finds it can be taught, especially through explicit instruction, real discussion, and authentic problems .

Analysis

That gives a practical test for any policy in this essay. Does it help children (1) learn real content, (2) learn how to check claims against evidence, and (3) practice disagreeing in good faith? A policy that weakens any of those weakens critical thinking, whatever its intentions.

So “critical thinking declined” here means the measurable foundations declined: reading comprehension, mathematical reasoning, the ability to evaluate a source, and the willingness of classrooms to argue over hard questions. No national test measures “critical thinking” directly. These are the closest honest proxies.

VIDEO · TED-Ed · Samantha Agoos

5 tips to improve your critical thinking

A short, practical five-step process for working through a decision.

Loads from the video site only when you click. The page works offline without it.

Open on the video site ↗

Timeline

1957–2026 at a glance

Filter by type, jump by decade, and open any event for its sources.

44 events

  1. 1957Sputnik spurs science-education pushpolicy

    Federal money flows to NSF curriculum projects such as BSCS Biology (1958).

    Read the chapter →

  2. 1971NAEP long-term trend beginsscores

    9-year-olds average 208 in reading; 13-year-olds 255.

    Read the chapter →

  3. 1983A Nation at Riskpolicy

    Warns of “a rising tide of mediocrity”; the standards movement begins.

    Read the chapter →

  4. 1989NCTM math standardspolicy

    Reasoning and problem solving become official goals.

    Read the chapter →

  5. 1999Okun’s “white supremacy culture” listculture

    Late-1990s list, later used in trainings; includes “objectivity” and “worship of the written word.”

    Read the chapter →

  6. 2000National Reading Panelscores

    Systematic phonics confirmed as essential for beginning readers.

    Read the chapter →

  7. 2002No Child Left Behind signedpolicy

    Annual testing in grades 3–8. Math rises; 44% of districts later cut other subjects.

    Read the chapter →

  8. 2005Dispositions disputesculture

    FIRE challenges Washington State and Brooklyn College over dispositions evaluations.

    Read the chapter →

  9. 2006NCATE drops “social justice”policy

    Accreditor removes the term from its dispositions glossary (June).

    Read the chapter →

  10. 2009Race to the Toppolicy

    About $4 billion to 12 grantees; rewards new standards and tests.

    Read the chapter →

  11. 2010Common Core releasedpolicy

    Final standards published in June; most states adopt them.

    Read the chapter →

  12. 2011California FAIR Education Actculture

    First state law requiring LGBT-inclusive history instruction.

    Read the chapter →

  13. 2012Long-term-trend peakscores

    13-year-olds hit their high point: math 285, reading 263.

    Read the chapter →

  14. 2012Smartphones spread among teenstech

    Teen ownership 23% (2011) → 37% (2012); 73% access by 2015.

    Read the chapter →

  15. 2012Texas GOP platform opposes “critical thinking skills”culture

    Party later calls it an “oversight.”

    Read the chapter →

  16. 2013Main NAEP peaks; “learning recession” beginsscores

    Grade-8 reading 267.6 and math 284.6. Mississippi passes its literacy law.

    Read the chapter →

  17. 2013NCTQ: “an industry of mediocrity”policy

    Fewer than 10% of rated prep programs earn three stars or more. NCATE and TEAC merge into CAEP.

    Read the chapter →

  18. 2014Federal school-discipline guidancepolicy

    Issued January 8; rescinded December 2018.

    Read the chapter →

  19. 2015First broad NAEP dropscores

    Math falls at grades 4 and 8; reading at grade 8.

    Read the chapter →

  20. 2015ESSA keeps annual testingpolicy

    Replaces NCLB; states get more control over accountability.

    Read the chapter →

  21. 2015Obergefell v. Hodgesculture

    Supreme Court recognizes same-sex marriage nationwide.

    Read the chapter →

  22. 2016Title IX transgender-student letterpolicy

    Issued May 13; withdrawn February 22, 2017.

    Read the chapter →

  23. 2016Kendi: tests a “racist weapon”culture

    Essay calls standardized tests “the most effective racist weapon ever devised…”

    Read the chapter →

  24. 2017Reading slide acceleratesscores

    2017–2019 yearly reading losses rival the pandemic’s.

    Read the chapter →

  25. 20191619 Project published; historians objectculture

    Five historians’ letter in December; Times revises a key line in March 2020.

    Read the chapter →

  26. 2019NYC DOE “white supremacy culture” slidesculture

    Administrator training lists objectivity as a trait.

    Read the chapter →

  27. 2020Schools close; tests waivedcovid

    Spring testing waived nationwide; remote starts in fall track politics more than case rates.

    Read the chapter →

  28. 2020Smithsonian “whiteness” chartculture

    Lists “objective, rational linear thinking”; removed with an apology in July.

    Read the chapter →

  29. 2020Equitable Math Stride 1culture

    Calls insisting on right answers a sign of “objectivity” in white supremacy culture.

    Read the chapter →

  30. 2020UC goes test-blindpolicy

    Regents suspend SAT/ACT requirement (May 21).

    Read the chapter →

  31. 2021CDC guidance and AFT languagecovid

    FOIA emails show union-proposed language adopted in February guidance.

    Read the chapter →

  32. 2021Moms for Liberty foundedcovid

    January, in Florida; a record 92 school-board recall efforts follow that year.

    Read the chapter →

  33. 2021Oregon SB 744policy

    Suspends graduation proficiency requirement; later extended through 2027–28.

    Read the chapter →

  34. 2021NSBA letter, Garland memo, apologycovid

    Sept. 29 letter; Oct. 4 memo; Oct. 22 apology.

    Read the chapter →

  35. 2021Virginia governor’s racecovid

    McAuliffe’s debate line; Youngkin wins on parents’ rights.

    Read the chapter →

  36. 2022Steepest drops in decadesscores

    9-year-olds down 5 in reading, 7 in math vs 2020.

    Read the chapter →

  37. 2022MIT reinstates SAT/ACTpolicy

    Dartmouth, Yale, Harvard and others follow in 2024.

    Read the chapter →

  38. 2022Florida HB 1557culture

    Restricts instruction on sexual orientation and gender identity in early grades.

    Read the chapter →

  39. 2023Reading reversalspolicy

    Teachers College restructures Calkins’s project; NYC Reads; states ban cueing.

    Read the chapter →

  40. 2023Exit exams down to eight statesscores

    From about half the states in 2014.

    Read the chapter →

  41. 2024Cut scores lowered in some statespolicy

    Wisconsin “recalibrates”; Oklahoma later restores higher bars.

    Read the chapter →

  42. 2024Adult skills fall; NAEP reading down againscores

    PIAAC (released Dec. 2024): adults at the lowest literacy level rise to 28%. NAEP 2024 (released Jan. 2025): reading down again.

    Read the chapter →

  43. 2025Title IX rule vacated; NCES guttedpolicy

    Court vacates 2024 rule (Jan. 9); ~$900M in research contracts cancelled.

    Read the chapter →

  44. 2026LTT 2025 results releasedscores

    9-year-olds rebound; 13-year-olds stuck below 2012.

    Read the chapter →

Part IIBaseline, 1950–2001
Chapter 03

1950s–1970s: Sputnik, inquiry, and the basics

A compact baseline of what classrooms were trying to teach.

After the Soviet Union launched Sputnik in 1957, Washington poured money into science and math education through the National Science Foundation . Programs such as BSCS Biology (1958) asked students to investigate, not just memorize . Jerome Bruner’s The Process of Education (1960) argued that any subject could be taught in some honest form at any age, if teachers taught its underlying structure . Bloom’s 1956 taxonomy gave teachers a vocabulary for moving from recall toward analysis and evaluation .

The historian Larry Cuban’s research is a useful reality check: in most classrooms, most of the time, teaching stayed teacher-centered: lecture, recitation, seatwork . The reforms changed ambitions more than daily routine. Debate and forensics, a real training ground for reasoning, had been a school tradition since the 1920s .

The record

The NAEP long-term trend test began in 1971 (reading) and 1973 (math). In 1971, 9-year-olds averaged 208 in reading and 13-year-olds averaged 255 .

This era shouldn’t be romanticized. Many children, especially Black children in segregated schools, were denied the same education. The point of the baseline is simpler: schools of this era said, often and explicitly, that reasoning from evidence was a goal.

Chapter 04

1980s–2001: alarms, standards, and steady gains

From A Nation at Risk to the eve of No Child Left Behind.

A Nation at Risk (1983) warned of “a rising tide of mediocrity” and launched the standards movement . The math teachers’ council (NCTM) made problem solving and reasoning official goals in 1989 . By 2000, two major federal syntheses summed up what research knew: expertise rests on organized knowledge , and early reading requires explicit, systematic phonics along with vocabulary, fluency and comprehension .

The record

On the long-term trend test, math scores rose steadily from the 1970s through 1999 at ages 9 and 13. Reading stayed fairly flat: 9-year-olds averaged 212 in 1999, and 13-year-olds 259 .

In short, the country entered the 2000s with rising math, flat reading, and a clear research consensus on how to teach children to read. Whether schools acted on that consensus is a big part of what comes next.

Part III2002–2015: the turn
Chapter 05

No Child Left Behind: gains, and a narrower school day

2002–2008: the first upstream change, and what it did and didn’t do.

No Child Left Behind (signed January 2002) required annual reading and math tests in grades 3–8 and penalties for schools that missed targets. Two things happened at once.

The record
  • Math went up. Dee and Jacob’s national study found NCLB raised grade-4 math by about 0.22 standard deviations by 2007. It found no comparable gain in reading .
  • The school day narrowed. By 2007, 44% of districts reported cutting time from other subjects, such as social studies, science, art and recess, to make room for tested reading and math .
  • Long-term trend scores rose from 1999 to 2008 and again to 2012 .
Analysis

The common belief that something changed around 2005 has a real basis here. NCLB squeezed history, civics and science, the subjects where students most often argue from evidence. But the measured scores kept rising through 2012. NCLB planted a risk. It didn’t produce a measurable decline in this period. The responsible parties for the narrowing are Congress and the Education Department, who built the incentives, and the states and districts that chose to cut other subjects rather than integrate them.

Chapter 06

Reading: the quiet failure

Settled science in 2000, ignored in many classrooms for two decades.

In 2000 the National Reading Panel said beginning readers need systematic phonics, sounding words out . Many schools and education schools did something else. The best-known alternative, “three-cueing,” taught children to guess unfamiliar words from pictures, context and the first letter. It was promoted for decades by the Reading and Writing Project at Teachers College, Columbia University, led by Lucy Calkins, and sold to districts nationwide .

The record
  • NCTQ’s 2013 review called teacher preparation “an industry of mediocrity”; fewer than 10% of the programs it rated earned three stars or more .
  • Ten years later, NCTQ reviewed course materials from 702 programs: only 28% adequately covered all five core reading components, and 22% covered none .
  • After APM Reports’ Sold a Story podcast, states began banning cueing . Teachers College restructured the project in September 2023 . New York City switched to research-aligned curricula through NYC Reads .
  • Mississippi passed a phonics-centered literacy law in 2013. Its grade-4 public-school reading average rose from 208.5 (2013) to 219.3 (2019) while the nation stalled .
Analysis

This is the most damning, best-documented failure in the whole story, and it isn’t a left–right issue. A child who can’t decode fluently can’t read the documents, data and arguments that critical thinking requires. Responsibility sits with the curriculum authors and publishers who sold cueing, the preparation programs that kept teaching it, and the district leaders who bought it after the research said otherwise. Mississippi proves it was fixable.

Chapter 07

Race to the Top and Common Core

2009–2015: federal incentives, rushed rollouts, and small negative effects.

Race to the Top (2009) was a federal grant contest, funded by the stimulus, that awarded about $4 billion to 12 states. It rewarded states for adopting “college- and career-ready” standards and new tests, among other things . The Common Core standards were released in June 2010 , and most states adopted them, many partly to compete for the money. The standards’ text actually emphasizes evidence-based argument and close reading .

The record

The best national study, by the federally funded C-SAIL center, compared states by how much their standards changed. It found small negative effects on NAEP. Grade-4 reading fell by roughly 0.06–0.11 standard deviations, and grade-8 math showed a negative seven-year effect of about 0.10. The losses were larger for students with disabilities and English learners. The authors caution that the results mostly reflect the transition period .

Federal law limits Washington’s role in curriculum , so the standards were technically state choices. But the incentives were federal. That is one place where a specific presidential-era policy fairly enters the ledger: Race to the Top, under the Obama administration and Education Secretary Arne Duncan, pushed states to adopt new standards and tests quickly.

Analysis

The rollout was the problem more than the paper. New standards, new tests, new textbooks, and new teacher evaluations all landed on teachers within a few years, starting in 2010–2014, right when scores started to turn. The evidence says the net effect was a small negative, not a catastrophe. It’s a real contributor, not the main culprit.

Chapter 08

Smartphones arrive

2012: the device that changed childhood, and the debate over what it did.

The record
  • Teen smartphone ownership jumped from 23% in 2011 to 37% in 2012 . By 2015, 73% of teens had access to one .
  • In 2025, 53% of public-school leaders said phones hurt academic performance, and 73% said they hurt attention .
  • In PISA 2022, about 3 in 10 U.S. students reported being distracted by digital devices in most or all math lessons .
  • The Education Recovery Scorecard team names the rise of social media, along with the retreat from test-based accountability, as a “likely candidate” for the pre-pandemic slide. It also finds phone bans have small positive effects .

Jonathan Haidt argues that the move to a phone-based childhood between 2010 and 2015 rewired adolescence . The psychologist Candice Odgers counters in Nature that the causal evidence is much weaker than the book claims .

Analysis

The timing fits the decline better than almost any other explanation, and it fits internationally too. The decline in reading for fun is striking. Among 13-year-olds, 35% read for fun almost every day in 1984; by 2025 it was 14% . But timing isn’t proof, and phones are not a school policy. The schools’ share of responsibility is narrower: handing every child a device (88% of schools have one-to-one programs ) without guardrails, and waiting years to restrict phones in class.

Chapter 09

The scores turn

The data: a peak around 2012–2013, a drop by 2015, and a slide before COVID.

Here is the core evidence, straight from the federal test. The long-term trend assessment has asked the same kinds of questions since the 1970s, so it is the best ruler we have.

CHART

13-year-olds peaked in 2012 and were falling before COVID

NAEP long-term trend average scale scores, age 13

Age 13 mathAge 13 reading
13-year-olds peaked in 2012 and were falling before COVIDNAEP long-term trend average scale scores, age 1325026027028029019711980199020002012202020252012 peak → slideCOVIDAge 13 math, 1973: 266.0Age 13 math, 1978: 264.1Age 13 math, 1982: 268.6Age 13 math, 1986: 269.0Age 13 math, 1990: 270.4Age 13 math, 1992: 273.1Age 13 math, 1994: 274.3Age 13 math, 1996: 274.3Age 13 math, 1999: 275.8Age 13 math, 2004: 281.0Age 13 math, 2004 (updated test format): 278.8Age 13 math, 2008: 281.5Age 13 math, 2012: 285.0Age 13 math, 2020: 279.7Age 13 math, 2023: 270.7Age 13 math, 2025: 270.3Age 13 reading, 1971: 255.2Age 13 reading, 1975: 255.9Age 13 reading, 1980: 258.5Age 13 reading, 1984: 257.1Age 13 reading, 1988: 257.5Age 13 reading, 1990: 256.8Age 13 reading, 1992: 259.8Age 13 reading, 1994: 257.9Age 13 reading, 1996: 257.9Age 13 reading, 1999: 259.4Age 13 reading, 2004: 258.7Age 13 reading, 2004 (updated test format): 256.7Age 13 reading, 2008: 259.7Age 13 reading, 2012: 262.8Age 13 reading, 2020: 260.1Age 13 reading, 2023: 255.7Age 13 reading, 2025: 256.1

← Swipe left for earlier years

Source: NAEP long-term trend average scores, 1971–2025, from the official results appendix . Gap in each line at 2004 marks the switch to an updated test format (both 2004 values shown). Hover or focus a point for its value. Scales are separate for math and reading; compare trends, not levels across subjects.

CHART

9-year-olds: long rise, 2012 plateau, COVID drop, partial rebound

NAEP long-term trend average scale scores, age 9

Age 9 mathAge 9 reading
9-year-olds: long rise, 2012 plateau, COVID drop, partial reboundNAEP long-term trend average scale scores, age 920021022023024025019711980199020002012202020252012 peak → slideCOVIDAge 9 math, 1973: 219.0Age 9 math, 1978: 218.6Age 9 math, 1982: 219.0Age 9 math, 1986: 221.7Age 9 math, 1990: 229.6Age 9 math, 1992: 229.6Age 9 math, 1994: 231.1Age 9 math, 1996: 231.0Age 9 math, 1999: 232.0Age 9 math, 2004: 241.2Age 9 math, 2004 (updated test format): 239.4Age 9 math, 2008: 242.9Age 9 math, 2012: 244.0Age 9 math, 2020: 241.4Age 9 math, 2022: 233.9Age 9 math, 2025: 237.7Age 9 reading, 1971: 207.6Age 9 reading, 1975: 210.0Age 9 reading, 1980: 215.0Age 9 reading, 1984: 210.9Age 9 reading, 1988: 211.8Age 9 reading, 1990: 209.2Age 9 reading, 1992: 210.5Age 9 reading, 1994: 211.0Age 9 reading, 1996: 212.5Age 9 reading, 1999: 211.7Age 9 reading, 2004: 218.7Age 9 reading, 2004 (updated test format): 215.7Age 9 reading, 2008: 220.1Age 9 reading, 2012: 220.8Age 9 reading, 2020: 219.7Age 9 reading, 2022: 214.6Age 9 reading, 2025: 218.4

← Swipe left for earlier years

Source: NAEP long-term trend average scores, 1971–2025, from the official results appendix . Gap in each line at 2004 marks the switch to an updated test format (both 2004 values shown). Hover or focus a point for its value. Scales are separate for math and reading; compare trends, not levels across subjects.

The record
  • Long-term trend, age 13: math peaked at 285.0 in 2012 and fell to 279.7 in 2020, before the pandemic. It reached 270.3 in 2025. Reading peaked at 262.8 in 2012, slipped to 260.1 in 2020, and was 256.1 in 2025 . The governing board says 13-year-olds have declined since 2012, “a high point” .
  • Main NAEP: grade-8 reading peaked at 267.6 in 2013 and grade-8 math at 284.6 in 2013 . In 2015, math dropped at grades 4 and 8, and reading dropped at grade 8 .
  • Learning recession: the Education Recovery Scorecard dates a “learning recession” to 2013, and finds the yearly reading losses of 2017–2019 were as large as during the pandemic .
CHART

Main NAEP, grade 8: peak in 2013, slide before 2020, drop after

National average scale scores, 2003–2024

Grade 8 mathGrade 8 reading
Main NAEP, grade 8: peak in 2013, slide before 2020, drop afterNational average scale scores, 2003–2024255260265270275280285290200320072011201320152019202220242013 peak → 2019COVIDGrade 8 math, 2003: 277.6Grade 8 math, 2005: 278.8Grade 8 math, 2007: 281.3Grade 8 math, 2009: 282.9Grade 8 math, 2011: 283.9Grade 8 math, 2013: 284.6Grade 8 math, 2015: 282.2Grade 8 math, 2017: 282.8Grade 8 math, 2019: 282.0Grade 8 math, 2022: 274.3Grade 8 math, 2024: 273.8Grade 8 reading, 2003: 263.3Grade 8 reading, 2005: 262.2Grade 8 reading, 2007: 262.8Grade 8 reading, 2009: 264.0Grade 8 reading, 2011: 265.2Grade 8 reading, 2013: 267.6Grade 8 reading, 2015: 265.4Grade 8 reading, 2017: 266.6Grade 8 reading, 2019: 263.1Grade 8 reading, 2022: 260.5Grade 8 reading, 2024: 258.0

← Swipe left for earlier years

Source: main NAEP national averages via the NAEP Data Service ; 2015 decline . Grade 4 shows the same shape with a later reading peak (222.5 in 2015).

Chapter 10

When did it start? The verdict

Testing the common 2005–2015 window against every candidate date.

A common view holds that the decline started between 2005 and 2015, probably closer to 2005. Here is each candidate date, tested against the data.

CandidateWhat happenedDoes the data show a decline then?
2002–2008NCLB narrows the school day No. Scores rose, especially math
2005–2006NCATE “social justice” dispositions fight No measurable effect found
2012LTT peak; smartphones pass a third of teens Peak, not decline
2013Main NAEP peak; Scorecard’s “learning recession” begins Turning point
2015First broad NAEP drop Yes, first measured decline
2017–2019Reading losses as large as during the pandemic Yes, accelerating
2020–2022COVID closuresYes, the sharpest drop in decades
Best-supported answer

The measured decline began around 2012–2013 and was visible nationally by 2015. Its roots reach back into the 2000s: the curriculum narrowing after 2002, and reading instruction that ignored the 2000 research consensus. The 2005–2015 window is right. The 2005 end of it marks when risks were planted; the test data says the harvest started around 2013.

Analysis

One caveat matters. No national test measures “reasoning instruction” directly, so there is no proof that teaching of critical thinking dropped in 2005. What the record does show is that the earliest documented change in practice is the post-2002 narrowing, and the earliest documented change in outcomes is 2012–2015. Anyone who claims a precise earlier date is claiming more than the evidence shows.

Part IVThe usual suspects: what the evidence says
Chapter 11

Was 2009–2017 the key period?

The turn happened inside that window, but so did forces that had nothing to do with one administration.

A common claim is that the damage was done mainly between 2009 and 2017.

Verdict: Partly
The record
  • The score peak (2012–2013) and the first broad drop (2015) fall inside 2009–2017 .
  • Several federal policies from that period are documented: Race to the Top and its standards incentives (2009–2010) , the January 2014 school-discipline guidance , and the May 2016 transgender-student guidance .
  • But the slide accelerated in 2017–2019 , under a different administration, and the 2016 guidance was rescinded in February 2017 .
  • Smartphones, the biggest non-school change, arrived on their own schedule .

A factual correction, offered gently: same-sex marriage became the law nationwide through the Supreme Court’s decision in Obergefell v. Hodges (June 2015), not through presidential action . President Obama’s own stated position changed in May 2012 . Neither event has any documented connection to reading or math instruction.

Analysis

The evidence points to a window of roughly 2012–2019, followed by the COVID collapse. That window crosses two administrations and both parties. Presidents belong in the ledger only where a specific, documented federal policy applies, and three do: Race to the Top’s incentives, the 2014 discipline guidance, and the 2016 Title IX letter. Each is weighed in its own chapter.

Chapter 12

Did teachers’ unions get much richer?

The unions’ own federal LM-2 filings say: richer, no; more politically active, yes.

A common claim is that teachers’ unions grew much richer and more powerful during 2009–2017.

Verdict on “much richer”: Doesn’t hold up · on political spending: Holds up

The NEA and AFT national headquarters each file an annual LM-2 financial report with the U.S. Labor Department. The cleanest measures are dues (NEA) and per-capita receipts from affiliates (AFT), because “total receipts” gets inflated in some years by investment sales. The NEA’s 2020 and 2021 totals topped $588 million for exactly that reason .

CHART

Dues income: flat in dollars, shrinking after inflation

Millions of dollars, national headquarters LM-2 filings

NEA dues (nominal)NEA dues in 2010 dollarsAFT per-capita (nominal)
Dues income: flat in dollars, shrinking after inflationMillions of dollars, national headquarters LM-2 filings10015020025030035040020072010201420172020202420262009–2017NEA dues (nominal), 2007: $314.6MNEA dues (nominal), 2010: $362.5MNEA dues (nominal), 2014: $363.0MNEA dues (nominal), 2016: $366.9MNEA dues (nominal), 2017: $370.1MNEA dues (nominal), 2020: $375.0MNEA dues (nominal), 2021: $377.4MNEA dues (nominal), 2024: $381.4MNEA dues (nominal), 2025: $390.8MNEA dues in 2010 dollars, 2007: $330.8MNEA dues in 2010 dollars, 2010: $362.5MNEA dues in 2010 dollars, 2014: $334.3MNEA dues in 2010 dollars, 2016: $333.3MNEA dues in 2010 dollars, 2017: $329.3MNEA dues in 2010 dollars, 2020: $315.9MNEA dues in 2010 dollars, 2021: $303.7MNEA dues in 2010 dollars, 2024: $265.1MAFT per-capita (nominal), 2012: $175.1MAFT per-capita (nominal), 2014: $169.4MAFT per-capita (nominal), 2016: $192.4MAFT per-capita (nominal), 2017: $177.4MAFT per-capita (nominal), 2020: $186.0MAFT per-capita (nominal), 2021: $196.7MAFT per-capita (nominal), 2026: $235.1M

← Swipe left for earlier years

Sources: NEA and AFT LM-2 filings . Inflation adjustment uses BLS CPI-U annual averages ; 2025 is not adjusted. Fiscal years: NEA ends Aug. 31, AFT June 30.

CHART

Political activities and lobbying: up sharply from 2016

Millions of dollars, LM-2 Schedule 16 totals

NEAAFT
Political activities and lobbying: up sharply from 2016Millions of dollars, LM-2 Schedule 16 totals01020304050607020072010201420172021202420262009–2017NEA, 2007: $32.2MNEA, 2010: $29.7MNEA, 2014: $31.4MNEA, 2016: $43.0MNEA, 2017: $53.3MNEA, 2020: $50.7MNEA, 2021: $66.0MNEA, 2024: $39.2MNEA, 2025: $51.7MAFT, 2012: $21.5MAFT, 2014: $24.9MAFT, 2016: $28.6MAFT, 2017: $40.0MAFT, 2020: $31.3MAFT, 2021: $48.8MAFT, 2026: $37.0M

← Swipe left for earlier years

Same filings as above. Spending peaks in fiscal years that include a presidential election (NEA FY2017 and FY2021 include November 2016 and 2020).

The record
  • NEA dues: $314.6M (FY2007) → $362.5M (FY2010) → $370.1M (FY2017) → $381.4M (FY2024) → $390.8M (FY2025) .
  • From 2010 to 2017, NEA dues grew about 2% while consumer prices rose about 12%. Through 2024, dues were up about 5% against inflation of about 44% . In real terms, NEA’s dues income shrank.
  • Members and fee payers reported by NEA fell from about 3.29 million (2010) to 3.07 million (2017) and 2.84 million (2024) .
  • AFT per-capita receipts: $175.1M (FY2012) → $177.4M (FY2017) → $196.7M (FY2021) → $235.1M (FY2026) . That is growth of about a third in nominal dollars over 14 years, roughly the pace of inflation or less .
  • Political activities and lobbying rose clearly. NEA went from $29.7M (FY2010) and $31.4M (FY2014) to $53.3M (FY2017) and $66.0M (FY2021). AFT went from $21.5M (FY2012) to $40.0M (FY2017) and $48.8M (FY2021) . Peaks track presidential-election years.
  • The 2018 Janus decision ended mandatory agency fees for public employees . NEA’s dues line still rose afterward.
Show all the LM-2 figures behind the charts
FilingDues / per-capitaPolitical & lobbying
NEA FY2007$314.6M$32.2M
NEA FY2010$362.5M$29.7M
NEA FY2014$363.0M$31.4M
NEA FY2016$366.9M$43.0M
NEA FY2017$370.1M$53.3M
NEA FY2020$375.0M$50.7M
NEA FY2021$377.4M$66.0M
NEA FY2024$381.4M$39.2M
NEA FY2025$390.8M$51.7M
AFT FY2012$175.1M$21.5M
AFT FY2014$169.4M$24.9M
AFT FY2016$192.4M$28.6M
AFT FY2017$177.4M$40.0M
AFT FY2020$186.0M$31.3M
AFT FY2021$196.7M$48.8M
AFT FY2026$235.1M$37.0M

NEA fiscal years end August 31; AFT’s end June 30. Figures are nominal dollars from the national-headquarters filings only. State and local affiliates file separately and are not included. Sources: .

Analysis

The “much richer” story doesn’t match the unions’ own filings. What did grow was political spending, and the moment union power clearly shaped what happened to children came in 2020–2021, over school reopening (Chapter 21). There, the evidence is strong: union strength predicted remote starts better than local COVID rates did . For 2009–2017 itself, there is no evidence linking union finances to a drop in reasoning instruction. If you want to hold unions accountable, hold them accountable for the closures, where the record is real.

Chapter 13

Did education schools train ideology instead of thinking?

The 2005–06 dispositions fight was real. The stronger evidence is about what ed schools failed to teach.

A common claim is that education schools trained teachers in ideology instead of critical thinking.

Verdict: Partly
The record
  • NCATE, then the main accreditor of teacher-preparation programs, listed “social justice” among example values that could guide a teacher candidate’s “dispositions.” The National Association of Scholars and FIRE argued this invited political screening of future teachers .
  • In 2005, FIRE intervened in two disputes. At Washington State University, a student’s political views figured in his dispositions evaluation . At Brooklyn College, a professor was threatened with an investigation after criticizing the college’s dispositions program .
  • In June 2006, NCATE’s president, Arthur Wise, denied that NCATE ever had a “social justice” requirement but announced that the term would be dropped, saying it “is susceptible to a variety of definitions” .
  • NCATE merged with TEAC to form CAEP in 2013 . A “teaching for social justice” literature remains influential in teacher education, for example Marilyn Cochran-Smith’s Walking the Road (2004) .
  • Was reasoning instruction displaced? The clearest displacement is in reading science, not ideology: most programs failed to teach it .
  • Effect on learning: a careful study of Missouri programs found that differences between preparation programs explain only a small share of differences in teacher effectiveness . No study located for this piece links social-justice coursework to lower student achievement, or to higher.
Analysis

Screening future teachers for political beliefs is wrong, whatever the beliefs. Education schools that did it deserved the pushback they got. But on measurable harm, the bigger failure of education schools is the decades they spent not teaching future teachers how children learn to read. That failure is documented, it touched millions of kids, and it fits the timeline.

Chapter 14

The gender-policy timeline

The policy changes are real. A link to declining reasoning skills isn’t in the evidence.

A common claim is that gender policy in schools helped drive the decline.

Verdict as a cause of skill decline: Doesn’t hold up · as a source of classroom conflict: Holds up
The record
  • 2011: California’s FAIR Education Act (SB 48) became the first state law requiring LGBT-inclusive history instruction .
  • May 13, 2016: the Education and Justice Departments issued a Dear Colleague letter saying Title IX covers gender identity, including access to restrooms and locker rooms . It was withdrawn on February 22, 2017 .
  • 2019: New Jersey and Illinois passed inclusive-curriculum laws; Illinois’s took effect in July 2020 .
  • 2022: Florida’s HB 1557 restricted classroom instruction on sexual orientation and gender identity in early grades, and other states followed with restrictive laws .
  • 2024–2025: a new Title IX rule took effect in 2024 and was vacated nationwide by a federal court on January 9, 2025 .
  • In 2023, 41% of teachers said debates over race, gender and sexuality had made their jobs harder .
Analysis

The timing doesn’t support a causal story. Most of these laws came after scores had already turned, and the 2016 letter lasted nine months. These policies are legitimate subjects for parents to debate. What is documented is their effect on climate: teachers on every side of the issue report avoiding discussion, and that is bad for teaching kids to argue well. Each of these policies deserves to be judged on its own merits, not as the cause of falling reading and math scores.

Chapter 15

CRT, K-12 antiracism, and the 1619 dispute

Academic CRT and classroom antiracism materials are different things. Both deserve precision.

A common claim is that critical race theory changed what schools teach, including history rewritten to fit the writer’s opinion.

Verdict: Partly

Two different things under one label

Academic critical race theory began in law schools in the 1970s and 1980s, and Ladson-Billings and Tate brought it into education research in 1995 . It is a graduate-level body of theory, and few K-12 classrooms teach it directly. K-12 antiracist materials are something else: trainings, workbooks and posters, often built on Tema Okun’s list of “white supremacy culture” characteristics (late 1990s; workbook 2001) . These spread through school trainings, especially between 2016 and 2021. Chapter 17 covers them in detail.

The 1619 Project dispute

The record
  • The New York Times Magazine published the 1619 Project in August 2019. Nikole Hannah-Jones won the 2020 Pulitzer Prize for Commentary for her lead essay .
  • In December 2019, five historians (Gordon Wood, James McPherson, Sean Wilentz, James Oakes and Victoria Bynum) wrote to the Times objecting to several claims, including that protecting slavery was a primary reason the colonists declared independence. The editor, Jake Silverstein, defended the project .
  • Historian Leslie M. Harris, who supports the project overall, wrote that she had warned the Times against that claim before publication .
  • In March 2020, the Times changed the passage to say protecting slavery was a motivation for “some of the colonists” .
  • The project reached schools through free lesson plans from the Pulitzer Center .
Analysis

The concern that history was changed to fit the writer’s opinion is fair in one specific way. A flagship claim was published over a fact-checker’s warning and only softened after prominent historians objected, while lesson plans were already in circulation. The right lesson for students isn’t “never read the 1619 Project.” It’s “read it next to the historians’ letter and the Times’s correction, and decide which claims the evidence supports.” That is critical thinking. Classrooms that taught it as settled fact failed students. Honest evidence also cuts the other way: a careful San Francisco study found an ethnic-studies course raised attendance and graduation , so race-conscious material isn’t automatically harmful.

Chapter 16

“Tests are white supremacy,” and did states drop tests?

Every state still tests. What changed was the stakes and the bar.

A common claim is that some states no longer test at all, because tests were called a form of white supremacy.

Verdict: Partly

A factual point first: no state has stopped annual testing. Federal law (ESSA, 2015) still requires statewide reading and math tests every year in grades 3–8 and once in high school . Testing was waived nationwide only in spring 2020 . What did happen is documented below, with what people said kept separate from what policy did.

What was said

The record
  • In 2016, Ibram X. Kendi wrote that standardized tests had become “the most effective racist weapon ever devised to objectively degrade Black minds and legally exclude their bodies” .
  • In 2021, the NEA’s magazine published “The Racist Beginnings of Standardized Testing,” tying testing’s origins to eugenics-era figures and urging less reliance on high-stakes tests .

What policy did

The record
  • Exit exams: about half the states required one in 2014; eight did for the class of 2023 .
  • Oregon: SB 744 (2021), signed by Gov. Kate Brown, suspended the requirement that students prove proficiency in reading, writing and math to graduate. The State Board later extended the suspension through 2027–28 .
  • Lower bars: several states lowered or “recalibrated” proficiency cut scores after the pandemic, including Wisconsin (2024), New York and Oklahoma . Oklahoma later restored higher standards .
  • College admissions: the University of California suspended SAT/ACT requirements in May 2020 and went test-blind for California applicants . MIT (2022), Dartmouth (2024) and others reinstated tests, saying scores helped them find talented students from disadvantaged backgrounds .
Analysis

“Some states don’t have tests anymore” isn’t accurate. “Many states quietly lowered the stakes and the bar” is. A test is just a measurement, and a fair test is one of the best friends a disadvantaged student has: it can spot talent that a biased grade or a well-connected recommendation would miss. MIT and Dartmouth said exactly that. Calling the thermometer racist doesn’t cure the fever. Removing the measurement mostly hid the decline from parents.

Chapter 17

“Objectivity is white supremacy”: who actually said it

No mainstream source found says “critical thinking is white supremacy” word for word. Some came close, and they are named here.

A common claim is that critical thinking itself is now labeled “white supremacy.”

Verdict: Partly (the ideas circulated; the exact phrase did not)

A search for anyone saying “critical thinking is white supremacy” in those words turned up no mainstream source. Here is what the record does show, in order of how close it comes.

The record
  1. Tema Okun’s list of “white supremacy culture” characteristics includes “objectivity,” “either/or thinking,” and “worship of the written word” . Okun’s current site warns against using the list as a checklist or a weapon .
  2. New York City’s Department of Education, 2019: training slides shown to administrators listed objectivity and “worship of the written word” as white-supremacy-culture traits. Officials said the slide was not part of the citywide bias training . The same list circulated to other school audiences, for example through a Berkeley schools foundation in 2020 .
  3. Smithsonian museum, 2020: the National Museum of African American History and Culture posted a chart of “Aspects and Assumptions of Whiteness” that included “objective, rational linear thinking,” “hard work is the key to success,” and “the nuclear family” . After criticism, the museum removed it and said: “We erred in including the chart. We have removed it, and we apologize.”
  4. Equitable Math, 2020: the Stride 1 workbook lists “There is a greater focus on getting the ‘right’ answer than understanding concepts and reasoning” as a sign of white supremacy culture, and says “Upholding the idea that there are always right and wrong answers perpetuate objectivity as well as fear of open conflict.” In a sidebar, it also concedes that “most math problems have correct answers” . Its authors say math is not racist and standards shouldn’t drop . Oregon’s Education Department promoted the related course in a February 2021 newsletter .
The other side of the ledger: a party platform against “critical thinking skills”

The 2012 Texas Republican Party platform opposed teaching “Higher Order Thinking Skills (HOTS) … critical thinking skills and similar programs.” A party spokesman later called it an “oversight” . Hostility to critical thinking as a school goal has shown up on more than one side.

On motives: some suggest people say these things out of hatred of white people, or jealousy. There is no evidence for either, and this piece doesn’t assign a motive it can’t document. The authors’ stated motive is to make classrooms more inclusive . You can reject the idea without guessing at what’s in anyone’s heart.

Analysis

The claim that objectivity, rational thinking, or the existence of right answers is a feature of “white supremacy culture” is wrong, and it is harmful. Objectivity, the discipline of checking your beliefs against evidence other people can also inspect, belongs to every human being. Assigning it to one race insults everyone else, and it hands students a reason to dismiss evidence they don’t like. The Smithsonian was right to apologize. There is no evidence that these materials caused the national score decline; they mostly spread after it started. Their harm is to the culture of reasoning, and that harm is real.

VIDEO · EdTrust-West · primary source

Dismantling Racism in Mathematics Instruction (webinar)

The proponents explain Stride 1 in their own words. Watch it alongside the quotes above and judge for yourself. Inclusion is not endorsement.

Loads from the video site only when you click. The page works offline without it.

Open on the video site ↗

Chapter 18

Other pre-COVID causes, scored

NCLB, Common Core, phones, grade inflation, discipline policy, reading, and the retreat from accountability.

Other common explanations point to testing pressure, Common Core, phones, grade inflation, and discipline policy.

The record
CandidateEvidenceStrength
Reading instruction (cueing, weak teacher prep)NCTQ audits; the Mississippi turnaround; states banning cueing Strong
Retreat from test-based accountability (after ~2015)Scorecard names it a likely candidate ; exit exams dropped Moderate
Smartphones and social media (2012+)Timing fits; causal evidence contested Moderate
Common Core rollout (2010–2015)Small negative effects Moderate–small
Grade inflationACT: average math GPA rose while test scores didn’t Moderate (hides decline)
NCLB teaching to the testNarrowed the curriculum; math still rose Mixed
2014 federal discipline guidanceIssued in 2014, rescinded in 2018 ; studies of suspension policy are mixed Weak / unsettled
Analysis

Ranked by weight of evidence, the pre-COVID culprits are: first, how reading was taught; second, phones colliding with schools that didn’t protect class time; third, the shift away from measuring and acting on results, including inflated grades and lowered bars. The culture-war items are real, but they’re smaller and later in the measurable story.

Part VClassrooms under pressure
Chapter 19

The chill: when discussion becomes risky

Pressure from every direction taught teachers to avoid hard questions.

The record
  • In a 2024 RAND survey, 65% of teachers said they had limited instruction on political and social topics, including many in states with no legal restrictions .
  • Pew found that 41% of teachers said debates over race, gender and sexuality had made their jobs harder .
  • PEN America counted 10,046 instances of books removed or restricted in school districts in 2023–24, using its own broad definition .
Analysis

The pressure came from both directions. Some schools promoted one-sided trainings that treated disagreement as harm; some legislatures and activists pushed rules vague enough that teachers avoided whole topics. Either way, students lose the thing critical thinking needs most: a room where a hard question can be argued from evidence. A teacher afraid of the next email is not going to run a real debate.

Chapter 20

The online-reasoning gap

Students who grew up online are poor at judging what they see there.

The record
  • In a national study of 3,446 high-school students, about two-thirds couldn’t tell sponsored content from news, 96% didn’t consider why ties between a climate website and the fossil-fuel industry mattered, and 52% believed a grainy video was “strong evidence” of voter fraud .
  • Short lessons in “lateral reading,” leaving a site to check who is behind it, improved students’ source evaluation in a field study .
  • In a randomized experiment, high-school students with unrestricted AI help did 48% better on practice problems and 17% worse on the exam once the AI was taken away. A tool designed as a tutor avoided most of that harm .
Analysis

Being fluent with a phone is not the same as being able to judge what’s on it. AI tools can make kids worse thinkers if they do the thinking for them. The evidence favors AI as a tutor that asks questions, not as an answer machine.

VIDEO · Crash Course · with Stanford researchers

Check yourself with lateral reading

How professional fact-checkers judge a website: leave it and see what others say about it.

Loads from the video site only when you click. The page works offline without it.

Open on the video site ↗

Part VICOVID: amplifier and awakening
Chapter 21

The closures: who decided

COVID didn’t start the decline. Decisions about how long to close made it far worse.

The record
  • A study of thousands of districts found that partisanship and union strength predicted fall-2020 remote starts much better than local COVID case rates did . A study of large districts reached a similar conclusion .
  • Emails released through a FOIA request show that AFT staff proposed language that the CDC adopted in its February 2021 reopening guidance . AFT president Randi Weingarten confirmed to a House subcommittee that the CDC accepted the union’s suggestion on accommodations . The White House press secretary called such consultation “best practice” .
  • Across 12 states, math pass rates fell about 14 points in spring 2021, and the losses were about 10 points smaller in districts that were fully in person . A study of 2.1 million students found remote instruction was a primary driver of widening achievement gaps .
  • U.S. students were more likely than the OECD average to attend schools closed for more than three months (64% vs. 51%) .
Analysis

Here the responsibility is direct and well documented. District leaders and school boards made the call; in many places, union leadership shaped it; and the CDC let a union help write guidance it should have written on the evidence alone. Many of these people were acting under real fear and uncertainty. But the cost landed on children, and the poorest children paid the most.

Chapter 22

Learning loss and missing students

The damage, measured.

The record
  • Long-term trend scores for 9-year-olds fell 5 points in reading and 7 in math from 2020 to 2022, the sharpest drops in decades .
  • For 13-year-olds, the governing board’s chair said scores are down “almost 10 points in math and 4 points in reading” since the pandemic .
  • Chronic absence (missing 10% or more of the school year) hit about 31% in 2021–22 and 28% in 2022–23 . The Scorecard team puts it at 23% in 2024–25, against about 15% before the pandemic .

The good news is in the same data: 9-year-olds rebounded 4 points in both subjects by 2025, and more than 100 districts improved fast . Recovery is possible when schools focus on it.

Chapter 23

The kitchen-table classroom and the parent awakening

Remote school let parents see the lessons. Many didn’t like what they saw, and they organized.

For a year, many parents could hear the lesson from the next room. Some were alarmed by how little was being taught; others by what was being taught. Whatever you think of what came next, it was one of the largest waves of parent involvement in school governance in modern memory.

The record
  • School boards: Ballotpedia tracked a record 92 recall efforts against 237 school-board members in 2021 .
  • Moms for Liberty was founded in Florida in January 2021 by former school-board members Tina Descovich and Tiffany Justice, and became a national political force .
  • The NSBA letter: on September 29, 2021, the National School Boards Association asked the federal government for help with threats against school officials, saying some acts could be “the equivalent to a form of domestic terrorism.” On October 4, Attorney General Merrick Garland directed the FBI to coordinate on threats against school personnel. On October 22, NSBA apologized, saying “there was no justification for some of the language” . Garland’s memo addressed threats and violence, not parents speaking at meetings .
  • Virginia, 2021: in a September 28 debate, Terry McAuliffe said, “I don’t think parents should be telling schools what they should teach” . Glenn Youngkin, who campaigned on parents’ rights, won on November 2 .
Analysis

Parents asking to see the curriculum is healthy, and the NSBA’s “domestic terrorism” language was a serious mistake; the group itself admitted it. One caution applies in the spirit of this whole piece: the parents’-rights movement is at its best when it demands more knowledge, real reading instruction, and open debate, and at its worst when it removes books wholesale . Same rule for everyone: more evidence, more argument, fewer forbidden questions.

Part VII2024–2026: where we are
Chapter 24

The latest scores, adults, and the data itself

The latest national results, adult skills, and the cuts to the statistics agency that measures them.

The record
  • NAEP 2024: grade-4 reading averaged 215.0 and grade-8 reading 258.0, both below 2019 and well below their 2013–2015 peaks . Grade-12 reading and math hit lows for their trend lines . Eighth-grade civics declined for the first time in 2022 .
  • Long-term trend 2025: 9-year-olds recovered 4 points; 13-year-olds did not. Only 23% of 13-year-olds were taking algebra, down from 34% in 2012. Just 14% read for fun almost every day .
  • Adults: between 2017 and 2023, the share of U.S. adults at the lowest literacy level (Level 1 or below) rose from 19% to 28%, and in numeracy from 29% to 34% .
  • The measuring stick: in early 2025, the Education Department cancelled roughly $900 million in research contracts . The statistics center (NCES) lost most of its staff, and the age-17 long-term trend test was cancelled .
Analysis

The children who were 9 when scores peaked in 2012 are now in their twenties, and the adult data is starting to show it. Cutting the agency that measures all this, at the moment it matters most, makes accountability harder no matter who is in charge. You can’t fix what you stop measuring.

Part VIIIAccountability and what to do
Chapter 25

Who is responsible

A tiered conclusion, pre-COVID causes first. Students are the harmed party, not the cause.

This ledger names specific organizations, documents and decisions, and not political parties. Strength refers to the evidence linking each decision to measurable harm to students’ learning or reasoning.

Tier 1 · Strong evidence of direct harm

  • The reading-instruction establishment, 2000–2023: the authors and publishers of cueing-based programs (most prominently Teachers College’s Reading and Writing Project), the preparation programs that didn’t teach reading science, and the districts that bought these programs after 2000 .
  • Extended school closures, 2020–2021: district leaders and boards who kept buildings closed longest; union leaderships that pressed for closure (including AFT’s role in CDC guidance); and the CDC for letting outside lobbying into its guidance .

Tier 2 · Substantial contribution

  • Lowering the bar and hiding the decline: state officials who dropped exit exams, suspended graduation proficiency (Oregon SB 744) or lowered cut scores, and grading practices that inflated GPAs .
  • Phones in class without guardrails: platforms built for compulsive use, and schools that waited years to protect class time .
  • Federal incentive design: NCLB’s pressure that narrowed the curriculum (2002–2008), and Race to the Top’s push for a fast standards-and-testing overhaul (2009–2015), which had small negative effects .

Tier 3 · Documented damage to the culture of reasoning; national effect on scores unmeasured

  • Materials that racialize objectivity: trainings that used Okun’s list as doctrine (e.g., NYC DOE 2019), the Smithsonian chart (2020, since retracted), Equitable Math’s framing, and agencies that promoted it (Oregon ODE 2021) .
  • Ideological screening in teacher prep: the “social justice” dispositions era, dropped by NCATE in 2006 .
  • Classroom chill from all sides: one-sided trainings, vague restrictive laws, mass book removals, and a 2012 party platform that opposed teaching critical thinking .
  • Curriculum taught as settled when it was disputed: classroom use of the 1619 Project’s contested claims without the historians’ critique .

Not supported by the evidence as causes

  • A surge in union wealth during 2009–2017: NEA dues fell in real terms .
  • Same-sex marriage or gender policy as causes of falling reading and math skills; the timing and mechanism don’t fit .
  • Students. Never students.
Analysis

In one sentence: American schools stopped doing the boring, proven things (teaching reading systematically, building knowledge, measuring honestly, protecting class time) at the same moment phones arrived, and then closed for too long when COVID hit. The ideological fights are real and worth having, but they are mostly symptoms of that drift and a distraction from fixing it.

Chapter 26

What to stay away from

A practical checklist for parents, school boards and voters.

These are red flags worth watching for in any school, whatever its politics:

  • Reading programs that teach children to guess words from pictures or context instead of sounding them out .
  • Lowering graduation requirements or proficiency cut scores instead of raising achievement .
  • Grades that keep rising while test scores fall .
  • Any training that calls objectivity, written evidence, or “right answers” a racial trait .
  • Any rule, from any side, so vague that teachers stop discussing hard questions .
  • Political screening of teachers or teacher candidates .
  • Phones in class, and AI that gives answers instead of asking questions .
  • Disputed history taught as settled fact, whoever wrote it .
  • Decisions about kids made by adults’ interest groups without the evidence on kids’ learning .
Chapter 27

What works

The evidence-backed fixes are not mysterious.

The record
  • Systematic phonics and knowledge-rich reading: Mississippi’s 2013 law and its results .
  • Explicit teaching, then guided practice: full guidance works better than discovery for beginners .
  • Teach reasoning inside subjects: document-based history improved both reasoning and reading ; explicit critical-thinking instruction plus dialogue works .
  • Lateral reading for online sources .
  • Honest measurement and attendance: fast-recovering districts focused on both .
  • Phone limits in class show small positive effects .
Chapter 28

Closing: the lost art isn’t lost

The last word.

Analysis

Critical thinking isn’t a left-wing value or a right-wing value. It’s the habit of asking “how do you know?” and being willing to hear the answer. Over the last 25 years, adults with very different politics made choices that weakened that habit in kids. Some did it through neglect of proven methods, some through fear, some through ideology, some through convenience. Most meant well. The children paid anyway.

The fix starts with the same habit. Ask your school what reading program it uses and why. Ask to see the evidence behind a training. Ask how grades compare with test scores. Ask, politely and persistently, “how do you know?” That’s the lost art, and it’s teachable.

Sources

Every source, numbered

146 sources, in order of first citation.

About the sources and methods

Sources were chosen in this order of preference: primary documents (laws, court rulings, federal filings, the training materials themselves), official data (NAEP, NCES, PIAAC, PISA), peer-reviewed or quasi-experimental studies, and then reputable reporting or advocacy-group documents, which are labeled as such. Each source notes what kind of evidence it is. Links open the source document; a few government, journal and news sites block automated tools but open normally in a browser.

The record boxes contain only sourced facts. Analysis boxes contain interpretation. Verdict badges on common claims mean: Holds up (the evidence supports the claim), Partly (some of it is supported, some is not or is overstated), and Doesn’t hold up (the evidence contradicts it or does not support it). Union finance figures are nominal dollars from national-headquarters LM-2 filings unless marked as inflation-adjusted.

  1. 001
    Federal assessmentNAEP · Long-Term Trend 2025 results ↗

    Ages 9 and 13; 2025 vs 2022/2023/2020 and the 1970s.

    Cited in: Ch. 01 Ch. 10 Ch. 22 Ch. 24 Timeline

  2. 002
    Federal assessmentNCES · 2015 reading and mathematics results ↗

    First broad main-NAEP drop: math at grades 4 and 8, reading at grade 8.

    Cited in: Ch. 01 Ch. 09 Ch. 10 Ch. 11 Timeline

  3. 003
    Research team releaseEducation Recovery Scorecard · May 2026 ↗

    “Learning recession” from 2013; 2017–19 reading losses; absenteeism.

    Cited in: Ch. 01 Ch. 08 Ch. 09 Ch. 10 Ch. 11 Ch. 18 Ch. 22 Ch. 25 Ch. 27 Timeline

  4. 004
    Primary textJohn Dewey · How We Think (1910) ↗

    Reflective thought, suspended judgment. Public-domain full text.

    Cited in: Ch. 02

  5. 005
    Author excerptEdward Glaser · definition of critical thinking (1941) ↗

    Glaser’s three-part definition: disposition, knowledge of methods, skill in applying them.

    Cited in: Ch. 02

  6. 006
    Cognitive-science essayDaniel Willingham · “Critical Thinking: Why Is It So Hard to Teach?” (2007) ↗

    Critical thinking depends heavily on subject knowledge; it is not a free-floating skill.

    Cited in: Ch. 02

  7. 007
    Meta-analysisAbrami et al. · strategies for teaching critical thinking (2015) ↗

    341 effect sizes; explicit instruction, dialogue, and authentic problems help.

    Cited in: Ch. 02 Ch. 27

  8. 008
    Educational videoTED-Ed · 5 tips to improve your critical thinking ↗

    Samantha Agoos; five-step process.

    Cited in: Ch. 02

  9. 009
    Institutional historyNSF · institutional history ↗

    Post-Sputnik federal investment in science education.

    Cited in: Ch. 03 Timeline

  10. 010
    Institutional historyBSCS Science Learning · history ↗

    Inquiry-based biology curriculum launched 1958.

    Cited in: Ch. 03 Timeline

  11. 011
    Book recordJerome Bruner · The Process of Education (1960) ↗

    Structure of disciplines; the spiral curriculum.

    Cited in: Ch. 03

  12. 012
    University teaching guidanceBloom’s taxonomy · Cornell teaching resource ↗

    The 1956 taxonomy and its 2001 revision.

    Cited in: Ch. 03

  13. 013
    Historical researchLarry Cuban · How Teachers Taught, 1890–1990 ↗

    Classroom practice changed slowly; teacher-centered instruction dominated most decades.

    Cited in: Ch. 03

  14. 014
    Institutional historyNational Speech & Debate Association · year-by-year history ↗

    Debate and forensics history since 1925.

    Cited in: Ch. 03

  15. 015
    Federal data fileNAEP LTT 2025 · technical and results appendix (xlsx) ↗

    Average-score tables 1971–2025 used for the chart in this essay.

    Cited in: Ch. 03 Ch. 04 Ch. 05 Ch. 09 Ch. 10 Ch. 11 Ch. 22 Timeline

  16. 016
    Federal report recordA Nation at Risk (1983) ↗

    Commission report warning of “a rising tide of mediocrity.”

    Cited in: Ch. 04 Timeline

  17. 017
    Contemporary recordNCTM standards implementation (1989) ↗

    Problem solving and reasoning as goals of school mathematics.

    Cited in: Ch. 04 Timeline

  18. 018
    Research synthesisNational Research Council · How People Learn (2000) ↗

    Expertise rests on organized knowledge.

    Cited in: Ch. 04

  19. 019
    Federal research synthesisNational Reading Panel · findings (2000) ↗

    Phonemic awareness, phonics, fluency, vocabulary, comprehension.

    Cited in: Ch. 04 Ch. 06 Ch. 25 Ch. 26 Ch. 27 Timeline

  20. 020
    Quasi-experimental studyDee & Jacob · the impact of No Child Left Behind (NBER w15531) ↗

    Math gains (about 0.22 SD at grade 4 by 2007); no reading gains.

    Cited in: Ch. 05 Ch. 10 Ch. 18 Timeline

  21. 021
    District surveyCenter on Education Policy · instructional time (2007/2008) ↗

    44% of districts cut time from other subjects after NCLB.

    Cited in: Ch. 05 Ch. 10 Ch. 18 Ch. 25 Timeline

  22. 022
    ReportingNew York Times · Teachers College closes Calkins’s project (Sept. 2023) ↗

    Background on the cueing controversy.

    Cited in: Ch. 06

  23. 023
    Advocacy-organization auditNCTQ · Teacher Prep Review (2013) ↗

    “An industry of mediocrity”: fewer than 10% of rated programs earned three stars or more.

    Cited in: Ch. 06 Ch. 13 Timeline

  24. 024
    Advocacy-organization auditNCTQ · Reading Foundations (2023) ↗

    702 programs; 28% cover all five components; 22% cover none.

    Cited in: Ch. 06 Ch. 13 Ch. 18 Ch. 25

  25. 025
    Investigative reportingAPM Reports · reading laws after Sold a Story (2025) ↗

    States restricting three-cueing.

    Cited in: Ch. 06 Ch. 18 Ch. 25 Ch. 26 Timeline

  26. 026
    Institutional statementTeachers College · Reading and Writing Project restructured (2023) ↗

    TC dissolves the project as constituted.

    Cited in: Ch. 06 Ch. 25 Timeline

  27. 027
    District policyNYC Public Schools · NYC Reads ↗

    Citywide shift to research-aligned reading curricula.

    Cited in: Ch. 06 Timeline

  28. 028
    Federal assessment dataNAEP · Mississippi grade 4 reading profile ↗

    Mississippi’s rise after its 2013 literacy law.

    Cited in: Ch. 06 Ch. 18 Ch. 27 Timeline

  29. 029
    Government oversightGAO · Race to the Top (GAO-11-658) ↗

    About $4 billion to 12 grantees; standards and assessments criteria.

    Cited in: Ch. 07 Ch. 11 Ch. 25 Timeline

  30. 030
    Standards developer historyCommon Core · development process ↗

    Final standards released June 2010.

    Cited in: Ch. 07 Timeline

  31. 031
    Standards textCommon Core · English language arts standards ↗

    Argument from evidence and close reading are explicit goals.

    Cited in: Ch. 07

  32. 032
    Quasi-experimental studyC-SAIL · Song, Yang & Garet, effects of college- and career-ready standards (2019) ↗

    Small negative effects on NAEP, e.g. grade-4 reading and grade-8 math.

    Cited in: Ch. 07 Ch. 18 Ch. 25

  33. 033
    Federal explanationEducation Department · what is not part of ED’s role ↗

    Curriculum is set by states and districts, not Washington.

    Cited in: Ch. 07

  34. 034
    National surveyPew · Teens and Technology 2013 ↗

    Teen smartphone ownership 23% (2011) → 37% (2012).

    Cited in: Ch. 08 Ch. 10 Ch. 11 Ch. 18 Timeline

  35. 035
    National surveyPew · Teens, Social Media & Technology 2015 ↗

    73% of teens had access to a smartphone.

    Cited in: Ch. 08 Timeline

  36. 036
    School Pulse PanelNCES · phones and devices (Feb. 2025) ↗

    Leaders report harm to academics, attention, mental health.

    Cited in: Ch. 08 Ch. 25 Ch. 26

  37. 037
    International assessmentOECD · PISA 2022 U.S. country note ↗

    Math down, reading and science steady; device distraction.

    Cited in: Ch. 08 Ch. 21

  38. 038
    Book / author siteJonathan Haidt · The Anxious Generation (2024) ↗

    Argues the 2010–2015 shift to phone-based childhood harmed youth.

    Cited in: Ch. 08 Ch. 18

  39. 039
    Critical reviewCandice Odgers · review of The Anxious Generation, Nature (2024) ↗

    Argues the causal evidence is weaker than the book claims.

    Cited in: Ch. 08 Ch. 18

  40. 040
    Federal assessment releaseNAGB · June 2026 long-term trend release ↗

    Declines since 2012 for 13-year-olds; reading for fun; algebra enrollment.

    Cited in: Ch. 08 Ch. 09 Ch. 22 Ch. 24 Timeline

  41. 041
    Federal data serviceNAEP Data Service · national average scores ↗

    Official API used for the main-NAEP values in this essay (example query: grade-8 reading, 2013).

    Cited in: Ch. 09 Ch. 24 Timeline

  42. 042
    ReportingInside Higher Ed · “A Spirited Disposition Debate” (June 6, 2006) ↗

    Wise: NCATE will drop “social justice” from its glossary.

    Cited in: Ch. 10 Ch. 13 Ch. 25 Timeline

  43. 043
    Federal guidance (rescinded)ED/DOJ · 2014 school discipline guidance ↗

    Issued Jan. 8, 2014; rescinded Dec. 21, 2018.

    Cited in: Ch. 11 Ch. 18 Timeline

  44. 044
    Federal guidance (rescinded)ED/DOJ · Dear Colleague Letter on Transgender Students (May 13, 2016) ↗

    Interpreted Title IX to cover gender identity in schools.

    Cited in: Ch. 11 Ch. 14 Ch. 25 Timeline

  45. 045
    Federal guidanceED/DOJ · withdrawal of the 2016 letter (Feb. 22, 2017) ↗

    Rescinded the 2016 guidance.

    Cited in: Ch. 11 Ch. 14 Timeline

  46. 046
    Supreme Court decisionObergefell v. Hodges (2015) ↗

    Same-sex marriage recognized nationwide by the Supreme Court.

    Cited in: Ch. 11 Ch. 25 Timeline

  47. 047
    ReportingABC News · Obama announces support for same-sex marriage (May 9, 2012) ↗

    The president’s position change in a television interview.

    Cited in: Ch. 11

  48. 048
    Federal financial disclosureNEA · LM-2, FY ending Aug. 2020 ↗

    Dues $375.0M; political & lobbying $50.7M. Total receipts inflated by investment sales.

    Cited in: Ch. 12

  49. 049
    Federal financial disclosureNEA · LM-2, FY ending Aug. 2021 ↗

    Dues $377.4M; political & lobbying $66.0M.

    Cited in: Ch. 12

  50. 050
    Federal financial disclosureNEA · LM-2, fiscal year ending Aug. 2007 ↗

    Dues $314.6M; political activities & lobbying $32.2M.

    Cited in: Ch. 12

  51. 051
    Federal financial disclosureNEA · LM-2, FY ending Aug. 2010 ↗

    Dues $362.5M; political & lobbying $29.7M.

    Cited in: Ch. 12 Ch. 25

  52. 052
    Federal financial disclosureNEA · LM-2, FY ending Aug. 2014 ↗

    Dues $363.0M; political & lobbying $31.4M.

    Cited in: Ch. 12

  53. 053
    Federal financial disclosureNEA · LM-2, FY ending Aug. 2016 ↗

    Dues $366.9M; political & lobbying $43.0M.

    Cited in: Ch. 12

  54. 054
    Federal financial disclosureNEA · LM-2, FY ending Aug. 2017 ↗

    Dues $370.1M; political & lobbying $53.3M.

    Cited in: Ch. 12 Ch. 25

  55. 055
    Federal financial disclosureNEA · LM-2, FY ending Aug. 2024 ↗

    Dues $381.4M; political & lobbying $39.2M.

    Cited in: Ch. 12

  56. 056
    Federal financial disclosureNEA · LM-2, FY ending Aug. 2025 ↗

    Dues $390.8M; political & lobbying $51.7M.

    Cited in: Ch. 12

  57. 057
    Federal financial disclosureAFT · LM-2, FY ending June 2012 ↗

    Per-capita receipts $175.1M; political & lobbying $21.5M.

    Cited in: Ch. 12

  58. 058
    Federal financial disclosureAFT · LM-2, FY ending June 2014 ↗

    Per-capita $169.4M; political & lobbying $24.9M.

    Cited in: Ch. 12

  59. 059
    Federal financial disclosureAFT · LM-2, FY ending June 2016 (amended) ↗

    Per-capita $192.4M; political & lobbying $28.6M.

    Cited in: Ch. 12

  60. 060
    Federal financial disclosureAFT · LM-2, FY ending June 2017 ↗

    Per-capita $177.4M; political & lobbying $40.0M.

    Cited in: Ch. 12

  61. 061
    Federal financial disclosureAFT · LM-2, FY ending June 2020 ↗

    Per-capita $186.0M; political & lobbying $31.3M.

    Cited in: Ch. 12

  62. 062
    Federal financial disclosureAFT · LM-2, FY ending June 2021 ↗

    Per-capita $196.7M; political & lobbying $48.8M.

    Cited in: Ch. 12

  63. 063
    Federal financial disclosureAFT · LM-2, FY ending June 2026 ↗

    Per-capita $235.1M; political & lobbying $37.0M; receipts $299.7M.

    Cited in: Ch. 12

  64. 064
    Federal statisticsBLS · Consumer Price Index ↗

    CPI-U annual averages used for inflation comparisons.

    Cited in: Ch. 12 Ch. 25

  65. 065
    Supreme Court decisionJanus v. AFSCME (2018) ↗

    Ended mandatory agency fees for public employees.

    Cited in: Ch. 12

  66. 066
    Peer-reviewed researchHartney & Finger · Politics, Markets, and Pandemics (2022) ↗

    Politics and union strength, not case rates, predicted remote starts in fall 2020.

    Cited in: Ch. 12 Ch. 21 Ch. 25 Ch. 26 Timeline

  67. 067
    Observational studyMarianno et al. · power in a pandemic (2022) ↗

    Union strength and reopening decisions in large districts.

    Cited in: Ch. 12 Ch. 21

  68. 068
    Advocacy-organization releaseFIRE · NCATE drops “social justice” language (June 2006) ↗

    Quotes NCATE’s prior wording on dispositions.

    Cited in: Ch. 13

  69. 069
    Advocacy-organization releaseNational Association of Scholars · NCATE drops “social justice” (June 5, 2006) ↗

    NAS objected that the term invited ideological screening.

    Cited in: Ch. 13

  70. 070
    Case fileFIRE · Washington State University dispositions case (2005) ↗

    A student’s political views figured in his dispositions evaluation.

    Cited in: Ch. 13 Ch. 25 Ch. 26 Timeline

  71. 071
    Case fileFIRE · letter to Brooklyn College (Aug. 2005) ↗

    Dispute over a dispositions program and a professor’s criticism of it.

    Cited in: Ch. 13 Timeline

  72. 072
    Accreditation standardsCAEP · standards ↗

    Accreditor formed from NCATE and TEAC.

    Cited in: Ch. 13 Timeline

  73. 073
    Book / publisher recordMarilyn Cochran-Smith · Walking the Road: Race, Diversity, and Social Justice in Teacher Education (2004) ↗

    A central text of the teaching-for-social-justice literature.

    Cited in: Ch. 13

  74. 074
    Peer-reviewed researchKoedel et al. · are there real differences across teacher prep programs? (2015) ↗

    Program differences explain only a small share of teacher-effectiveness variance.

    Cited in: Ch. 13

  75. 075
    Research compilationMAP · LGBTQ curricular laws, citations ↗

    State-by-state inclusive and restrictive curriculum laws, incl. California SB 48 (2011) and New Jersey (2019).

    Cited in: Ch. 14 Timeline

  76. 076
    State lawIllinois Public Act 101-0227 (HB 246, 2019) ↗

    Requires instruction on LGBT contributions in history, effective July 2020.

    Cited in: Ch. 14

  77. 077
    State lawFlorida HB 1557 (2022) · Parental Rights in Education ↗

    Restricted classroom instruction on sexual orientation and gender identity in early grades.

    Cited in: Ch. 14 Timeline

  78. 078
    Court decisionTennessee v. Cardona (E.D. Ky., Jan. 9, 2025) ↗

    Vacated the 2024 Title IX rule nationwide.

    Cited in: Ch. 14 Timeline

  79. 079
    National surveyPew · race and LGBTQ issues in K-12 schools (2024) ↗

    41% of teachers say these debates made their job harder.

    Cited in: Ch. 14 Ch. 19

  80. 080
    Theoretical articleLadson-Billings & Tate · Toward a Critical Race Theory of Education (1995) ↗

    Brought legal CRT into education scholarship.

    Cited in: Ch. 15

  81. 081
    Primary text, preserved copyJones & Okun · Dismantling Racism workbook (2001) ↗

    The 1999/2001 list as it circulated.

    Cited in: Ch. 15 Ch. 17 Timeline

  82. 082
    Advocacy frameworkTema Okun · White Supremacy Culture characteristics ↗

    Lists “objectivity,” “worship of the written word,” “either/or thinking”; warns against weaponizing the list.

    Cited in: Ch. 15 Ch. 17 Ch. 26 Timeline

  83. 083
    Prize recordPulitzer Prizes · Nikole Hannah-Jones, Commentary (2020) ↗

    Prize for the project’s introductory essay.

    Cited in: Ch. 15

  84. 084
    Publisher responseNew York Times · “We Respond to the Historians…” (Dec. 20, 2019) ↗

    Prints the five historians’ letter and the editor’s reply.

    Cited in: Ch. 15 Ch. 25 Timeline

  85. 085
    Commentary by historianLeslie M. Harris · “I Helped Fact-Check the 1619 Project. The Times Ignored Me.” (Politico, 2020) ↗

    Supports the project overall; says she warned against the Revolution claim.

    Cited in: Ch. 15

  86. 086
    Publisher correctionNew York Times · “An Update to The 1619 Project” (Mar. 11, 2020) ↗

    Changed the Revolution passage to “some of the colonists.”

    Cited in: Ch. 15 Ch. 25 Ch. 26 Timeline

  87. 087
    Policy journalEducation Next · “The 1619 Project Enters American Classrooms” (2020) ↗

    How the project reached schools through Pulitzer Center lesson plans; critical review.

    Cited in: Ch. 15

  88. 088
    Regression-discontinuity studyBonilla, Dee & Penner · ethnic studies in San Francisco (2021) ↗

    Local gains in attendance and graduation.

    Cited in: Ch. 15

  89. 089
    Federal policy explanationU.S. Education Department · ESSA ↗

    ESSA (2015) kept annual statewide testing.

    Cited in: Ch. 16 Timeline

  90. 090
    Federal letterEducation Department · 2020 assessment waivers ↗

    Spring 2020 testing waived nationwide.

    Cited in: Ch. 16 Timeline

  91. 091
    Essay by authorIbram X. Kendi · “Why the Academic Achievement Gap is a Racist Idea” (AAIHS, 2016) ↗

    Calls standardized tests “the most effective racist weapon ever devised…”

    Cited in: Ch. 16 Timeline

  92. 092
    Union publicationNEA Today · “The Racist Beginnings of Standardized Testing” (2021) ↗

    Traces testing to eugenics-era figures; argues for less high-stakes testing.

    Cited in: Ch. 16

  93. 093
    ReportingEducation Week · states have soured on the exit exam (2023) ↗

    Eight states in 2023, about half in 2014.

    Cited in: Ch. 16 Ch. 18 Ch. 25 Timeline

  94. 094
    Advocacy-organization countFairTest · states requiring graduation tests fall to eight ↗

    Count of states with exit exams, 2023.

    Cited in: Ch. 16

  95. 095
    State policyOregon ODE · Essential Skills requirement suspended ↗

    Suspension extended by the State Board.

    Cited in: Ch. 16 Ch. 25 Ch. 26 Timeline

  96. 096
    ReportingThe 74 · states lowering proficiency cut scores (2024) ↗

    Wisconsin, New York, Oklahoma, Alaska and others.

    Cited in: Ch. 16 Ch. 25 Ch. 26 Timeline

  97. 097
    State agencyWisconsin DPI · new cut scores (2024) ↗

    DPI calls it recalibration.

    Cited in: Ch. 16 Timeline

  98. 098
    State agencyOklahoma OEQA · 2025 assessment results ↗

    Oklahoma restored higher cut scores.

    Cited in: Ch. 16 Timeline

  99. 099
    University system releaseUniversity of California · Regents change testing requirement (May 21, 2020) ↗

    SAT/ACT suspended; test-blind for California applicants.

    Cited in: Ch. 16 Timeline

  100. 100
    University statementMIT Admissions · reinstating the SAT/ACT requirement (Mar. 2022) ↗

    Scores help identify prepared students, including disadvantaged ones.

    Cited in: Ch. 16 Timeline

  101. 101
    University statementDartmouth · reactivating the SAT/ACT requirement (Feb. 2024) ↗

    Faculty research found scores helped identify talent.

    Cited in: Ch. 16 Timeline

  102. 102
    Advocacy frameworkTema Okun · One Right Way / objectivity ↗

    Okun’s current text on objectivity.

    Cited in: Ch. 17

  103. 103
    ReportingNew York Post · NYC DOE “white-supremacy culture” training (May 2019) ↗

    Slides listed objectivity and “worship of the written word.” Officials said the slide was not part of the citywide bias training.

    Cited in: Ch. 17 Ch. 25 Timeline

  104. 104
    Primary documentWhite Supremacy Culture summary · Berkeley Public Schools Fund (2020) ↗

    Example of the list circulating to school audiences in 2020.

    Cited in: Ch. 17

  105. 105
    Government legal opinionArkansas Attorney General Opinion 2021-042 ↗

    Preserves the chart’s wording (footnote 10).

    Cited in: Ch. 17 Timeline

  106. 106
    Archived primary documentNMAAHC “Whiteness” page · Internet Archive (June 2020) ↗

    Snapshot of the museum’s Talking About Race page.

    Cited in: Ch. 17

  107. 107
    ReportingMiami Herald · museum removes “whiteness” chart (July 2020) ↗

    “We erred in including the chart… and we apologize.”

    Cited in: Ch. 17 Ch. 25 Timeline

  108. 108
    Institutional statement, preservedNMAAHC · July 2020 removal statement ↗

    The museum’s posted explanation.

    Cited in: Ch. 17

  109. 109
    Training workbookEquitable Math · Stride 1 (2020) ↗

    Printed p. 66: right-and-wrong answers “perpetuate objectivity.”

    Cited in: Ch. 17 Ch. 25 Ch. 26 Timeline

  110. 110
    Authors’ explanationEquitable Math · authors’ FAQ ↗

    Authors deny that math is racist or that standards should drop.

    Cited in: Ch. 17

  111. 111
    State agency newsletterOregon ODE · Math Educator Update (Feb. 2021) ↗

    Promoted the Pathway to Math Equity micro-course.

    Cited in: Ch. 17 Ch. 25

  112. 112
    Political platformTexas Republican Party · 2012 platform ↗

    Opposed teaching “Higher Order Thinking Skills… critical thinking skills.”

    Cited in: Ch. 17 Ch. 25 Timeline

  113. 113
  114. 114
  115. 115
    Proponents’ videoEdTrust-West · Stride 1 webinar ↗

    Proponents explain the workbook.

    Cited in: Ch. 17

  116. 116
    Assessment-provider researchACT · grade inflation in math (2023) ↗

    Average math GPA rose while scores did not.

    Cited in: Ch. 18 Ch. 25 Ch. 26

  117. 117
    Peer-reviewed researchLacoe & Steinberg · do suspensions affect student outcomes? (2019) ↗

    Philadelphia evidence on suspension policy changes.

    Cited in: Ch. 18

  118. 118
    National teacher surveyRAND · restrictions and spillovers (2024) ↗

    65% of teachers limited political and social discussion.

    Cited in: Ch. 19 Ch. 25 Ch. 26

  119. 119
    Advocacy-organization countPEN America · 2023–24 school book bans ↗

    10,046 instances in 2023–24 by PEN’s definition.

    Cited in: Ch. 19 Ch. 23 Ch. 25

  120. 120
    Study summaryDigital Inquiry Group · students’ civic online reasoning ↗

    3,446 high-school students; widespread failures to evaluate sources.

    Cited in: Ch. 20

  121. 121
    Field study recordWineburg et al. · lateral reading field study (2022) ↗

    Short lessons improved students’ source evaluation.

    Cited in: Ch. 20 Ch. 27

  122. 122
    Randomized field experimentBastani et al. · generative AI can harm learning (PNAS, 2025) ↗

    Unrestricted AI help raised practice scores and lowered exam scores.

    Cited in: Ch. 20 Ch. 26

  123. 123
    Educational videoCrash Course · Check Yourself with Lateral Reading ↗

    Made with the Stanford History Education Group.

    Cited in: Ch. 20

  124. 124
    Primary documentsAFT–CDC emails, Feb. 2021 (FOIA release) ↗

    AFT proposed language CDC adopted in its reopening guidance. Site may block automated access.

    Cited in: Ch. 21 Ch. 25 Timeline

  125. 125
    Congressional recordHouse Select Subcommittee · Weingarten testimony (2023) ↗

    Republican-majority committee summary; Weingarten confirms the accommodation suggestion was accepted.

    Cited in: Ch. 21 Timeline

  126. 126
  127. 127
    Peer-reviewed researchHalloran, Jack, Okun & Oster · schooling mode and test scores (2023) ↗

    Larger losses where schooling was remote.

    Cited in: Ch. 21 Ch. 25

  128. 128
    Large observational studyGoldhaber et al. · remote and hybrid instruction (NBER, 2022) ↗

    Remote instruction a primary driver of widening gaps.

    Cited in: Ch. 21

  129. 129
    Federal dataEducation Department · chronic absenteeism ↗

    Chronic absence rose sharply after 2020.

    Cited in: Ch. 22

  130. 130
    Election dataBallotpedia · school board recalls ↗

    2021: record 92 recall efforts against 237 board members.

    Cited in: Ch. 23 Timeline

  131. 131
    Organization statementMoms for Liberty · about ↗

    Founded January 2021 in Florida.

    Cited in: Ch. 23 Timeline

  132. 132
    ReportingAP · Moms for Liberty rises as a power player (2023) ↗

    Growth, influence and controversies.

    Cited in: Ch. 23

  133. 133
    Fact-checkFactCheck.org · the NSBA letter and the Garland memo (2022) ↗

    Dates and wording of the NSBA letter, the memo, and NSBA’s apology.

    Cited in: Ch. 23 Timeline

  134. 134
    Federal memoAttorney General Garland · memorandum on threats against school officials (Oct. 4, 2021) ↗

    Directed FBI-led coordination on threats and violence.

    Cited in: Ch. 23 Timeline

  135. 135
    Video recordRealClearPolitics · McAuliffe debate remark (Sept. 28, 2021) ↗

    “I don’t think parents should be telling schools what they should teach.”

    Cited in: Ch. 23 Timeline

  136. 136
    Election resultsVirginia Department of Elections · Nov. 2, 2021 governor results ↗

    Youngkin won the governorship.

    Cited in: Ch. 23 Timeline

  137. 137
    Federal assessment releaseNAGB · 2024 grades 4 and 8 results ↗

    Reading down again; math mixed.

    Cited in: Ch. 24 Timeline

  138. 138
    Federal assessmentNAEP · grade 12 reading, 2024 ↗

    Lowest grade-12 reading average in the trend.

    Cited in: Ch. 24

  139. 139
    Federal assessmentNAEP · grade 12 mathematics, 2024 ↗

    Lowest grade-12 math average since 2005 framework.

    Cited in: Ch. 24

  140. 140
    Federal assessment summaryNAGB · 2022 U.S. history and civics ↗

    First-ever civics decline at grade 8.

    Cited in: Ch. 24

  141. 141
    Adult skills assessmentNCES · PIAAC 2023 U.S. results ↗

    Share of adults at Level 1 or below rose in literacy and numeracy.

    Cited in: Ch. 24 Timeline

  142. 142
    ReportingInside Higher Ed · IES contracts cancelled (Feb. 2025) ↗

    About $900 million in contracts.

    Cited in: Ch. 24 Timeline

  143. 143
    Professional associationAmerican Statistical Association · NCES in 2025 ↗

    Staff cuts and cancelled studies.

    Cited in: Ch. 24 Timeline

  144. 144
    Research reviewKirschner, Sweller & Clark (2006) ↗

    Minimal guidance is less effective for novices.

    Cited in: Ch. 27

  145. 145
    Instructional guidanceBarak Rosenshine · Principles of Instruction (2012) ↗

    Explicit, research-based teaching practices.

    Cited in: Ch. 27

  146. 146
    Quasi-experimental studyReisman · Reading Like a Historian (2012) ↗

    Document-based history improved reasoning and reading.

    Cited in: Ch. 27

JesseW22.com · Facts are sourced; analysis is labeled. Corrections welcome.