TL;DR
- The story comes in two shapes: Yale, class of 1953, where the 3% with written goals were later worth more than the other 97% combined; and Harvard, MBA class of 1979, where the same 3% earned ten times the rest put together.
- Yale University Library keeps a standing answer, whose first line says no goals study of that class occurred. The class secretary, his classmates and a set of administrators were asked, and the records of several offices searched.
- A magazine reporter chased the citation in 1996 and found a loop. Tony Robbins pointed at Brian Tracy, Tracy pointed at Zig Ziglar, and Ziglar’s office pointed back at Robbins.
- Four bibliographic indexes turn up nothing that matches. The queries and the hit counts are printed below so you can run them yourself.
- Goal setting does have a real literature, and it is far less dramatic. Across 384 effect sizes from randomised trials, setting a goal shifted behaviour by d = 0.34, which is small.
Somewhere in a hotel conference room in the mid-1990s, a man in a suit is telling a few hundred salespeople about a piece of research. The story was a fixture for speakers such as Zig Ziglar, the American motivational speaker who sold goal setting to corporate America. Three per cent of a graduating Ivy League class had written their goals down. Twenty years on, those three were worth more than the other ninety-seven put together.
It is a wonderful thing to hear at nine in the morning with a lanyard round your neck. It is also why a business journalist named Lawrence Tabak spent part of 1996 on the telephone, trying to find the study. His account ran in the December 1996 issue of Fast Company under the title If Your Goal Is Success, Don’t Consult These Gurus, and it restates the anecdote the way consultants tell it before delivering the verdict: “It’s also completely untrue” (Tabak, 1996). Thirty years on it is still circulating, still without a citation, and now with a second university attached.
Two versions of the same numbers
There are two stories, not one, and they share a suspicious amount of structure: three per cent, a famous university, a small group beating everyone else combined.
| Yale version | Harvard version | |
|---|---|---|
| Class | Yale College, 1953 | Harvard MBA, 1979 |
| Split reported | 3% yes, 97% no | 3% written, 13% unwritten, 84% none |
| Follow-up | Twenty years later | Ten years later, so 1989 |
| Result | The 3% held more wealth than the other 97% combined | The 3% earned ten times the other 97% put together |
| Credited to | Ziglar, Robbins, Tracy | Mark McCormack’s book about Harvard Business School |
When one story exists in two mutually exclusive versions with the same headline number, one of them is wrong at minimum. Both are.
Following the citation backwards
Tabak started with a 1995 book called No Excuse! by the consultant Jay Rifenbary, which cites the Yale story. Rifenbary could not document it. His researchers, Mike and Marjie Markowski, offered instead a measure of how widespread it had become: “We’ve attended hundreds of motivational seminars and heard it many times.”
Next came Tony Robbins, the American self-help author, whose bestseller Unlimited Power refers on page 200 of its thirty-fourth printing to “A study of the 1953 graduates of Yale University”. Tabak rang Robbins Research International in La Jolla, California, and learned that the files for that book had “met a disastrous end”. The company sent him to the personal effectiveness consultant Brian Tracy. Tracy sent him to Ziglar.
Ziglar had indeed used it, on his knees, in a video called Goals, Setting and Achieving Them on Schedule: “Those 3%,” he says there, “accomplished more than the 97% combined who had not set those goals.” Asked for the source, Ziglar’s Dallas office said it would be hard to establish, because “Mr. Ziglar is always reading. He reads more than two hours a day!” Then the office made a suggestion: try Tony Robbins.
That is a closed loop. Robbins to Tracy to Ziglar to Robbins, with nobody at any point holding a paper.
Sixteen years later the British journalist Oliver Burkeman, then a feature writer for The Guardian, repeated the exercise for his book The Antidote and got the identical circle (Burkeman, 2012). A citation that survives two independent attempts to trace it, sixteen years apart, and closes on itself both times, is a rumour with footnote formatting.
What Yale says when you ask
Tabak’s last call was to the university. Beverly Waters, described in his article as a research associate at Yale, said a run of articles citing the study in magazines as varied as Dental Economics and Success had pushed her into an exhaustive search of the Yale alumni archives. She found nothing. Her summary: “We are quite confident that the ‘study’ did not take place. We suspect it is a myth.”
Burkeman rang the same person around 2012, by then describing her as a senior Yale archivist, and heard a more tired version. “I did a systematic check, years ago, when this first arose, and there was nothing,” she told him. The secretary of the class of 1953 had run his own check, and nobody he spoke to had ever filled out such a questionnaire. Then she said the thing that explains the whole phenomenon: “It’s just too good not to be true, I guess.”
That class secretary was Silas Spengler, a retired lawyer who had held the post since graduation. He told Tabak he had never written down any personal goals, and that neither he nor his classmates had taken part in research on them. He sent extracts from the 1953 yearbook, which contain no personal goals at all, only guesses about future employment. Forrest Mars Jr., later chairman and chief executive of Mars, listed his employment possibilities as “no”.
Yale University Library keeps a standing answer, because the question arrives so often. Its first line: “no ‘goals study’ of the Class of 1953 actually occurred” (Yale University Library, 2020). The rest describes the work behind that sentence. The class secretary did not know of the study, nor did the classmates he asked. Several Yale administrators were consulted and the records of various offices examined. There was no relevant record, and nobody recalled such a study of that class “or of any other class”.
The same page traces one outbreak to a September 1991 article in Success magazine by Ken Thuerback, then president of Alpine Log Homes. His source was an audio tape by Ziglar. Ziglar’s source, asked in turn, was that he had heard the numbers mentioned over the years.
The book everyone points at
The Harvard version has a named origin: What They Don’t Teach You at Harvard Business School by Mark McCormack, an American sports agent and author. Pages repeating the Harvard numbers almost always name that book, several as the sole source anyone has. Two things about the attribution deserve a closer look.
The first is arithmetic on the calendar. The story describes a class surveyed in 1979 and interviewed again ten years later, which puts the follow-up in 1989. McCormack’s book was first published in 1984, according to the catalogue record and edition list held at Open Library, which shows printings in 1984, 1986, 1989 and 1994. A book that came out in 1984 cannot report a follow-up conducted in 1989. Either the passage sits in a later edition, or the dates in the retelling drifted, or both.
The second is the missing page number. Of every page examined for this article that credits McCormack, not one gives a page, a chapter or a line quoted from the book. They give the title and move on.
Looking for it where studies live
Testimony from a university archive is strong, and it is still testimony. The next step is mechanical: search where a real study would leave a trace, and print the searches so anyone can repeat them.
Four indexes were used. OpenAlex, which catalogues scholarly works of all kinds, was queried through api.openalex.org/works?filter=title_and_abstract.search: followed by the query. Europe PMC and PubMed cover the behavioural and biomedical literature. Crossref holds the registration records behind almost every academic DOI, and its search is looser, so instead of a hit count the first thousand records were scanned for the literal phrases.
| Index | Query as typed | Hits |
|---|---|---|
| OpenAlex | “written goals” AND “class of 1953” | 0 |
| OpenAlex | “goals study” AND “class of 1953” | 0 |
| OpenAlex | “written goals” AND “Harvard” AND “MBA” | 0 |
| Europe PMC | “written goals” AND “class of 1953” | 0 |
| Europe PMC | “written goals” AND “Harvard” AND “MBA” | 1, a 2008 book of conference abstracts |
| PubMed | “written goals” AND “class of 1953” | 0 |
| PubMed | “goal setting”[All Fields] AND “Yale”[All Fields] AND “1953”[All Fields] | 0 |
| Crossref | query.bibliographic=Yale class of 1953 written goals study, first 1000 records | 1 record contains “class of 1953”, a caption from an alumni reunion banquet |
Nothing. Not a paper, not a conference abstract, not a technical report. A search tool that returns zero is only useful if you know it can return something else, so an invented identifier, 10.9999/nonexistent.2026.0001, went through the same Crossref lookup used for every real paper cited here. It returned a 404, as it should. The zeros above are zeros, not a broken pipe.
What the numbers would have to mean
One more check needs no archive. Take the claim at face value and see what it demands of the world.
Imagine a hundred graduates: three wrote goals down, ninety-seven did not. In the Yale version those three hold more wealth than the other ninety-seven combined, which means each is worth more than thirty-two times the average of everyone else. Put the ninety-seven on a hundred thousand dollars of net worth each and every one of the three needs more than three and a quarter million.
The Harvard version multiplies that by ten: the three earn ten times what ninety-seven people earn together, so each sits at roughly three hundred and twenty times the average income of the rest. Put the ninety-seven on a hundred thousand dollars a year and each of the three is on thirty-two million a year, in 1989.
Ratios like that do occur in real populations, and they come from a founder whose company floats. A class with two near-billionaires in it will do this to any statistic you like, including handwriting, breakfast cereal or shoe size. The anecdote asks you to read a wealth distribution as a treatment effect.
Still on Google
Twenty search queries were pooled, from “Yale 1953 goal study” to “harvard mba goal study snopes”, and 124 pages opened and read; another 40 would not open and were dropped from the count. Fifty-three present the study as a finding with no hint of doubt, among them a business-coaching site (“The results were striking: only 3% of the students had written goals”) and several Goodreads quote pages lifted straight from Brian Tracy and Tony Robbins. Ten more give the numbers in full, then bury a line of doubt, one of them literally as a footnote: “The Harvard Study mentioned in this post is in question.” Forty-six say it never happened, and 27 of the 124 trace it back to the 1996 Fast Company investigation. Fifteen more pages mention the story only in passing without taking a side.
So the story still wins, roughly five to four, thirty years after it was shown to be empty. The split has a shape: the pages stating it as fact are coaching sites, sales-training articles and book-quote pages, written to be read aloud from a stage, while the corrections are mostly personal blogs and one university library.
The unglamorous thing that does exist
Goal setting has been studied properly for sixty years. Edwin Locke, of the R. H. Smith School of Business at the University of Maryland, and Gary Latham, of the Rotman School of Management at the University of Toronto, built goal-setting theory out of laboratory and field experiments.
Their 1981 review with Karyll Shaw and Lise Saari went through the studies published between 1969 and 1980 and reported that “specific, challenging goals led more often to higher performance than easy goals, ‘do your best’ goals or no goals”, with 90% of studies showing positive or partially positive results (Locke, Shaw, Saari & Latham, 1981). Two decades on they summarised the programme as a thirty-five-year odyssey and set out the mechanisms: goals direct attention, mobilise effort, extend persistence and push people to invent strategies (Locke & Latham, 2002). A later summary adds goal choice, learning goals, framing and group goals (Locke & Latham, 2006).
Notice what this literature is about. Difficulty and specificity, measured over tasks and weeks. A hard, clear target beats a vague one. It says nothing whatsoever about lifetime net worth, and it never promised to.
For the size of the effect, the cleanest number comes from a review by Tracy Epton and Christopher Armitage of the Manchester Centre for Health Psychology and Sinéad Currie of the University of Stirling. They collected randomised trials that isolated goal setting from everything bundled around it, extracted 384 effect sizes from 16,523 people, and found “a small positive unique effect of goal setting across a range of behaviors, d = .34”, with a confidence interval of 0.28 to 0.41 (Epton, Currie & Armitage, 2017). Difficulty, a public setting and a shared target each nudged that figure upward.
What does d = 0.34 feel like? Line up two rooms of thirty people, one with goals and one without. The average person in the goal room lands around the 63rd percentile of the other room. You would have to count carefully to see it: real, worth having, and a very long way from three per cent of a class owning the world.
One nearby finding beats the goal itself. Watching your own progress, rather than naming a target, moved goal attainment by d+ = 0.40 across 138 randomised studies and 19,951 participants (Harkin et al., 2016). Two conditions raised it further: making the outcome public, and recording it on paper rather than in your head.
The shift: treat “a study at Harvard found” as a claim about a document, and ask which document, before arguing about the finding.
First move: when a statistic has no author and no year, follow the attribution one hop. Two hops usually closes the circle.
Second move: do the arithmetic the story implies. Three per cent out-earning ninety-seven per cent combined means a ratio above thirty to one per head, which describes a wealth distribution rather than a cause.
Third move: point the pen at tracking rather than declaring. Recorded progress carries the larger measured effect.
The same failure mode produced the rule that 80 per cent of results come from 20 per cent of effort, named after an economist who never proposed it. Two more claims about who succeeds get the same treatment here: whether emotional intelligence predicts success, and where the twenty-one days figure for habits came from. The full list of claims with grades and dates lives in the registry of checked claims. The goal-setting bestseller that built a whole chapter on the same statistic is checked in our notes on «The Power of Focus». The other famous goal number, the 42 per cent, comes from a real study that never printed it: are you 42% more likely to achieve a written goal.
The boring bottom line
Yale looked for the study of its class of 1953 and could not find it. The class secretary had never been surveyed, the administrators had no record, and the archive search came back empty twice, years apart. Four bibliographic indexes return nothing that matches. The Harvard version credits a book first published five years before the follow-up it describes, and nobody who cites that book gives a page.
Two reporters, in 1996 and again in 2012, asked the three men who told the story where they got it, and were sent round in a circle. One of them, asked to respond to Yale’s finding, said: “Heard this story originally from Zig Ziglar. If it’s not true it should be.”
Write your goals down anyway, on the strength of a small, measured, unglamorous effect, and keep an eye on your progress while you are at it, because that part scores slightly higher.
Sources
- Tabak, L (1996). If Your Goal Is Success, Don't Consult These Gurus. Fast Company. Issue 06, December 1996/January 1997; traces the Yale class of 1953 story through Rifenbary, Robbins, Tracy and Ziglar, and quotes Yale research associate Beverly Waters: we are quite confident that the study did not take place fastcompany.com
- Burkeman, O (2012). Why Setting Goals Could Wreck Your Life. Fast Company. Excerpt from The Antidote (Faber and Faber, 2012); repeats the source-chasing sixteen years after Tabak and gets the same circle, and quotes Waters on her systematic check of the archive fastcompany.com
- Yale University Library (2020). Where can I find information on Yale's 1953 goal study?. Ask Yale Library. Answered by Laura Galas, last updated 28 September 2020; states that no goals study of the Class of 1953 occurred and describes the checks with the class secretary, classmates and university administrators ask.library.yale.edu
- McCormack, M. H (1984). What They Don't Teach You at Harvard Business School. Open Library catalogue record. Work record OL820252W; editions listed for 1984, 1986, 1989 and 1994, which places first publication five years before the 1989 follow-up the Harvard version of the story describes openlibrary.org
- Locke, E. A., Shaw, K. N., Saari, L. M., & Latham, G. P (1981). Goal setting and task performance: 1969-1980. Psychological Bulletin. 90(1), 125-152; review reporting that specific, challenging goals led more often to higher performance than easy goals, do your best goals or no goals, with 90% of studies positive or partially positive doi:10.1037/0033-2909.90.1.125
- Locke, E. A., & Latham, G. P (2002). Building a practically useful theory of goal setting and task motivation: A 35-year odyssey. American Psychologist. 57(9), 705-717; summary of the goal-setting programme and its mechanisms (attention, effort, persistence, strategy development) doi:10.1037/0003-066X.57.9.705
- Locke, E. A., & Latham, G. P (2006). New Directions in Goal-Setting Theory. Current Directions in Psychological Science. 15(5), 265-268; later summary covering goal choice, learning goals, goal framing and group goal setting doi:10.1111/j.1467-8721.2006.00449.x
- Epton, T., Currie, S., & Armitage, C. J (2017). Unique effects of setting goals on behavior change: Systematic review and meta-analysis. Journal of Consulting and Clinical Psychology. 85(12), 1182-1198; 384 effect sizes, N = 16,523 from randomised controlled trials, random effects d = .34, CI [.28, .41]; larger when the goal was difficult, set publicly or a group goal doi:10.1037/ccp0000260
- Harkin, B., Webb, T. L., Chang, B. P. I., Prestwich, A., Conner, M., Kellar, I., Benn, Y., & Sheeran, P (2016). Does monitoring goal progress promote goal attainment? A meta-analysis of the experimental evidence. Psychological Bulletin. 142(2), 198-229; 138 studies, N = 19,951; progress monitoring promoted goal attainment at d+ = 0.40, CI [0.32, 0.48], with larger effects when outcomes were made public and physically recorded doi:10.1037/bul0000025