Sources
This is the working file behind the arguments on this site. For each figure you get the exact group that was measured, the organization that produced it, when it was collected, and a link. Where a number has a weakness, it is written next to the number rather than left for you to find.
Not all research is the same weight, so the figures below are grouped by how they were produced rather than by topic. A federal survey that randomly selects households and publishes its margin of error is stronger evidence than a company survey of people who volunteered to answer. Both can appear in an argument. They should not be quoted as if they were equal.
Survey sentiment, meaning what people say about their degree, is kept separate from records-based measures like underemployment, wages and debt. Both matter. They are not the same kind of evidence.
That is also why this page is not ordered to match the problem page. Reading it in story order would tell you where a figure was used; reading it in this order tells you how far each one can be pushed, which is the more useful thing to know when you are checking somebody’s argument. Entries that appear on the problem page carry a small tag showing where, so you can still find any specific claim.
Random or address-based recruitment, published methodology, stated margin of error. The strongest evidence on this page.
Act 3 · Part 03
of young adults said they lacked “meaning or purpose” in their lives in the previous month. On what was harming their mental health: 56% named financial worries, 51% achievement pressure, 50% not knowing what to do with their life, 45% a sense that things are falling apart and 44% a sense of not mattering to others. 36% screened positive for anxiety and 29% for depression, against 18% and 15% of teens.
Act 3 · Part 03
“Far too many young adults told us that they feel on edge, lonely, directionless, and that they worry about financial security. Many are ‘achieving to achieve’ and find little meaning in either school or work.”
Act 3 · Part 03
the United States’ world ranking for the life evaluations of people under 30, against 10th for those aged 60 and over. The U.S. placed 23rd overall, down from 15th the year before. In the authors’ words: “the young are now, in 2021–2023, the least happy age group. This is a big change from 2006–2010, when the young were happier than those in the midlife groups.”
Act 3 · Part 03
“Across every dimension of well-being that we looked at; happiness, health, meaning, character, relationships, financial stability, each one was strictly increasing with age. Those who are 18 to 25 felt they were worse off across all these dimensions. It was pretty striking, pretty disturbing.”
Act 3 · Part 03
countries in which the long-standing U-shape of wellbeing across age has disappeared. Ill-being no longer peaks in midlife; it now falls steadily with age, because the young have got so much worse since roughly 2013.
Act 2 · Part 03
the rise in the price of college tuition and fees between 1985 and 2025, against roughly 3× for consumer prices generally. CPI for college tuition and fees went from 119.9 to 955.3; all-items CPI went from 107.6 to 321.9. Both series share the 1982–84 = 100 base.
Act 2 · Part 02
of U.S. teens say anxiety and depression are a major problem among people their age at their school. It tops the list, ahead of bullying (22%), drug use (19%), physical fights (17%) and drinking (14%).
Act 2 · Part 02
“When parents ask me where all of this pressure is coming from, I ask them: Where is it not?”
Act 1 · Part 01
of U.S. adults age 25 and over held a bachelor’s degree or higher in 2023, against 25.6% in 2000, 23.0% in 1995, 11.0% in 1970, 7.7% in 1960 and 4.6% in 1940.
of recent college graduates are underemployed, working a job that does not typically require a degree.
unemployment among recent college graduates, against about 4.2% for adults overall.
recent-graduate unemployment, against 4.2% for all workers and 7.2% for same-age workers without a degree.
of Americans say the U.S. higher education system is going in the wrong direction, up from 56% in 2020.
say a four-year degree is worth it even if someone has to take out loans. 47% say worth it only without loans; 29% say not worth it.
Act 2 · Part 02
of teens say they face a great deal or a fair amount of pressure to get good grades.
Act 2 · Part 03
of U.S. adults rate a college education as “very important,” down from 75% in 2010.
of U.S. workers stay in their current role primarily because leaving would be too difficult or costly.
of adults under 30 currently have or are being treated for depression, up from 13.0% in late 2017.
the fall since 2023 in young Americans saying it is a good time to find a job locally: 70% in 2023, 55% in 2024, 43% in 2025.
of adults 18 to 29 report doing okay or living comfortably financially, against 83% of adults 60 and over.
Act 3 · Part 02
cumulative rise in consumer prices between July 2019 and July 2026.
Act 3 · Part 02
of American renter households spend more than 30% of their income on rent and utilities; 22.7 million households, or 49% of all renters.
Built from employment records, transcripts or completed transactions rather than from opinion. Different strengths: large and concrete, but each one measures a narrow, specific thing.
the Fed’s baseline total cost of a four-year degree, rising to $272,000 if it takes five years and $364,000 at six. One extra year to finish “reduces the return by about a quarter”; two extra years, “by more than 40 percent.”
the estimated rate of return for a 25th-percentile college graduate. In the authors’ words: “for at least a quarter of college graduates, college does not appear to pay off.”
Act 1 · Part 01
the approximate year demand for high-skill cognitive work stopped rising and reversed, after decades of growth, pushing graduates down the occupational ladder into jobs previously done by less-educated workers.
Act 3 · Part 01
of graduates whose highest degree is a bachelor’s are underemployed one year after finishing. 45% are still underemployed a decade out.
the shortfall in employment among workers ages 22 to 25 in the most AI-exposed occupations, relative to where it would stand had it kept pace with employment of similarly aged workers in less-exposed occupations. Experienced workers in the same occupations show no comparable gap.
of undergraduates who had declared a major changed it at least once within three years.
is the average published cost of one year at a public four-year college, in-state, all in. Out-of-state public is $49,080 and private nonprofit is $62,570. Over four years, roughly $120,000 to $250,000.
Act 3 · Part 02
is the median age of a first-time U.S. homebuyer, an all-time high. The first-time buyer share fell to a record-low 21%.
of 15-year-olds have no clear career plans.
Respondents volunteer rather than being randomly selected, so these cannot carry a margin of error and can skew. Useful for sentiment, weaker for prevalence.
Act 1 · Part 03
of the skills a job requires are expected to change by 2030, with job disruption equal to 22% of jobs; 170 million roles created, 92 million displaced.
Act 1 · Part 03
“Trends such as generative AI and rapid technological shifts are upending industries and labour markets, creating both unprecedented opportunities and profound risks.”
Act 2 · Part 03
of parents say they would prefer their child attend college right after high school: 40% a four-year college, 19% a two-year or community college.
of parents who do not want a four-year college for their child name cost as the main reason, unprompted. 17% cite return on investment.
of student borrowers who have not completed a degree say they have delayed a major life milestone because of their loans.
of the rise in U.S. unemployment since mid-2023 came from recent graduates, who are about 5% of the labor force.
Act 3 · Part 02
of Gen Z degree-holders say their own degree was a waste of time or money, against 41% of millennials and 20% of baby boomers.
Produced by businesses with a commercial interest in the subject. Methodology is often unpublished. Treated here as indicative, not as evidence.
Act 1 · Part 02
applications per job posting in 2025, up from 116 in 2022; a 111% rise in three years.
applications handled per recruiter per year in 2025, against 146 in 2022, while recruiters per organization fell from 10.4 to 4.6.
Act 1 · Part 02
of business leaders say they have raised the qualifications and experience required for the same roles; 59% expect that to continue.
Act 1 · Part 02
“While companies are hiring more, they’re also receiving a flood of applications, making it harder for employers to select the right candidate and for employees to stand out.”
of graduates feel unprepared to apply for entry-level jobs in their field, while 89% of educators believe their students are ready.
Act 3 · Part 02
the gap between what students expect to earn ($101,500) and what recent graduates actually average ($68,400).
Act 1 · Part 03
of young adults say they are not in the career they intended to pursue.
Act 2 · Part 01
of college juniors and seniors say they chose their major with a specific job or career path in mind. In the same survey, 88% say their coursework is adequately preparing them for entry-level roles, while more than half of hiring managers surveyed a year earlier said recent graduates are unprepared.
Act 2 · Part 01
“Students feel ready because they’re doing well in class, but you can’t prepare for a marathon just by reading about running.”
of registered voters say a four-year degree is worth the cost; 63% say it is not. In 2017 the same question split 49% worth against 47% not.
college graduates wish they had pursued a different path (community college, an apprenticeship) or skipped college altogether. Separately, 46% say they could have got their current job without a degree.
What we took down
These were on this site or in the drafts behind it. Each was checked against the originating source and removed. They are listed here because a page that only shows what survived is not a source page.
“70% say college did not prepare them for the job market.”
Cengage Group
The real 70% in that report is graduates saying basic AI training should be added to courses. It is not a verdict on preparation, and the same report found 73% of graduates said their education was worth the cost.
“Up to 75% of students change majors at least once.”
Widely circulated, often miscredited to NCES
Traces back to an uncited assertion in a 2016 advisory paper whose own analysis never measured switching rates. NCES, the body it is usually attributed to, publishes 30%.
“85% of the rise in unemployment is recent graduates.”
Oxford Economics
The 85% describes all new labor-market entrants. Oxford’s figure for recent graduates specifically is 12%.
“43% of first-time buyers cite student loans as the top obstacle to a down payment.”
Attributed to the National Association of Realtors
This appears in no NAR publication. NAR’s own figure for the obstacle question is 36%, and high rent outranks student loans. The 43% appears to be a garbled version of a different statistic about how many Younger Millennials carry student debt.
“7 in 10 teens say anxiety and depression are a major problem.”
Pew Research Center
Real, but fielded in fall 2018 and stated without its qualifier. Pew asked whether it was a major problem among people their age in the community where they live, so it measures what teens see around them rather than what they report about themselves. Pew’s current wave asks a differently worded question and finds 30%.
“Four in ten teenagers are unclear about their career, double the rate of a decade ago.”
OECD
The 39% is real but it is an OECD-country average for 15-year-olds, not a United States figure. The decade comparison does not exist; the underlying question was not asked that far back.
What cuts the other way
The argument on this site is not that college is worthless. It is that college should not be the unexamined default. Here is what a fair skeptic would raise, and should.
19.3% of U.S. job postings required a bachelor’s in November 2025, up from 16.6% two years earlier; and that holds after controlling for which jobs exist, so it is not simply a shift in the mix. It sits against the skills-first narrative, and it is reported here for that reason. Indeed’s own reading is that the degree may now be serving as a shortlisting filter in a crowded market rather than as a measure of skill.
The New York Fed puts the lifetime return on a degree near 12.5%. And on the comparison its own economists prefer, recent graduates at 5.7% unemployment still do better than their same-age peers without degrees at roughly 7.2%.
Federal Reserve Bank of New York, 2026
Our daughter left college after a semester, spent seven months going nowhere, and only turned it around after we told her she had to move out. It is the case I know best and it proves nothing on its own. The figures above came first. We read them when she was a child and parented accordingly.
For the first time in a decade, the share expressing a great deal or quite a lot of confidence rose from 36% to 42%. Still well below 57% in 2015, but the direction matters and it is reported here for that reason.
Gallup and Lumina Foundation, 2025
Around 59% of degree holders say their education helped prepare them for a well-paying job. Regret about a specific major sits near 20%, far below the headline “waste of money” numbers.
Pew Research Center, 2024; USA Today, 2024
The Gen Z subsamples inside them are small and mostly unpublished. The distance between low major-specific regret near 20% and high “waste” sentiment at 36% to 51% suggests the anger is about cost and debt at least as much as about the education.
Methodological caveat
Roughly a third of graduates of any age have worked in jobs not requiring a degree across the past three decades, in good economies and bad. Many move out of it. The NY Fed series has run between 38% and 48% since 2008.
Federal Reserve Bank of New York
Where a statistic circulates through secondary outlets, it is attributed here to the organization that produced it rather than to the news story that amplified it. Several of the claims in the section above were removed precisely because that trail led somewhere different from where the popular version said it did.
Several figures update on a fixed schedule. The New York Fed labor series publishes in February, May, August and November. The Federal Reserve household survey publishes each May. The National Association of Realtors profile publishes each November. When those land, the numbers here should be rechecked rather than assumed.
Pathbreaker Parenting source file. Figures verified against originating sources in August 2026.
Found an error? seth@pathbreakerparenting.com