TL;DR. Most of the claims about famous leaders’ reading habits that circulate online do not rest on a primary record. Of the 14 claims in the old version of this article, only 4 could be verified in the leader’s own writing or in a dated interview transcript, 1 is partly contradicted and 9 could not be verified and were removed. The verified ones are these: on 4 December 2017 Bill Gates described reading as his favorite way to indulge his curiosity; Mark Zuckerberg’s 2015 “A Year of Books” page stated a goal of a new book every two weeks; in an oral history interview dated 26 June 2008, Elon Musk said he had read books about Edison and had taught himself to program as a kid. At the level of society the picture is less bright: according to US National Endowment for the Arts (NEA) data, 48.5% of adults read at least one book in the past 12 months in 2022, down from 54.6% in 2012. Adults over 50 who read books for 3.5 hours or more a week had a lower 12-year mortality risk, but this finding is correlational. CEO move: plan reading not as a hobby but as an investment tied to a quarterly decision agenda. Student move: start every book with a single question and test what you read in real work within two weeks.
This article is part of the Skills series. We covered in detail why deep reading has become a differentiator in the AI era, along with meta-analytic data on print exposure, in deep reading and competitive advantage; for methods that make what you read stick, see the evidence ranking of 12 study techniques, and for deciding what not to read, see the article on strategic ignorance.
Why was this article rewritten and what changed?
The first version of this article was published in 2020. It listed quotes and figures attributed to Warren Buffett, Bill Gates, Elon Musk, Mark Zuckerberg and Oprah Winfrey: 5 hours of reading a day, at least one book a week, close to 10 hours of science fiction a day as a teenager, an encyclopedia finished at age 9. None of these claims came with a source.
CEOtudent’s sourcing policy requires every statement attributed to a real person to be verified against text that person published, a dated interview transcript or an official record. In this refresh we did the following:
- We searched for every attribution in a primary record. Those we found stay in, with date and source; those we could not find are marked “not verified” and removed from the text. The audit table below shows every decision.
- We removed ad code, broken images, clickable external links and automatically inserted internal links.
- We added a section explaining why famous examples alone do not count as evidence.
- We added published data on adult reading rates, the measured effects of reading and the paper versus screen comparison.
- We added a framework that fits a reading habit into a working adult’s calendar, plus a 4-week protocol.
The topic and the question of the article did not change: Why do these people place so much importance on reading, and what does reading actually give a person?
Which claims about leaders’ reading habits can actually be verified?
The table below lists every claim from the old article and the primary record we looked for. “Verified” is used only for statements we found in the person’s own writing or in a dated transcript. News sites, listicles and summary blogs were not counted as sources; this is why some very widely repeated figures also appear as not verified in this table.
| # | Claim in the old article | Primary record sought | Result |
|---|---|---|---|
| 1 | The quote “The person you will be in 5 years will be shaped by the books you read and the people you’re with” | Not attributed to anyone in the old article; no source found | Not verified, removed |
| 2 | Warren Buffett: “The more you learn, the more you earn” | No text published by Buffett himself could be opened | Not verified, removed |
| 3 | Warren Buffett reads 5 hours a day | Same; no primary record could be opened | Not verified, removed |
| 4 | Bill Gates finishes at least one book a week (commonly given as “50 books a year”) | Gates’s own statement was searched for on Gates Notes and not found; the figure appears only on news and list sites | Not verified, removed |
| 5 | For Gates, reading is a way to indulge his curiosity | Gates Notes, 4 December 2017 | Verified |
| 6 | Gates avoids fiction and rarely reads novels | The same 2017 list recommends a novel (John Green, Turtles All the Way Down) | Partly contradicted |
| 7 | Zuckerberg will finish a book every two weeks | 10 January 2015 archive of the “A Year of Books” page and the Marquette University Zuckerberg Files record | Verified |
| 8 | Zuckerberg reads about beliefs, cultures, history and technology | The same page’s description | Verified |
| 9 | Zuckerberg’s long quote saying books let you explore a topic in depth | The archived copy of the original post sits behind a login wall; the text appears only in secondary publications | Not verified, not used as a quote |
| 10 | As a teenager Musk read close to 10 hours of science fiction a day | No dated transcript that could be opened was found | Not verified, removed |
| 11 | Musk finished the Encyclopaedia Britannica at age 9 | No dated transcript that could be opened was found | Not verified, removed |
| 12 | Asked how he learned about rockets, Musk said “I read a lot of books” | No dated transcript that could be opened was found | Not verified, removed |
| 13 | Musk grew by reading and teaching himself | The Henry Ford oral history transcript, 26 June 2008 | Verified (not in the old article, added) |
| 14 | Views on reading attributed to Oprah Winfrey | No primary record could be opened | Not verified, removed |
CEOtudent analysis: the old article’s claims were checked against Gates Notes (2017), the “A Year of Books” page (Internet Archive, 10 January 2015), Marquette University Zuckerberg Files record 365 and The Henry Ford Musk oral history transcript (2008). Of the 14 rows, 4 were verified, 1 is partly contradicted and 9 were not verified.
Bill Gates: What does he write in his own words?
Gates’s year-end book list on his blog Gates Notes, dated 4 December 2017, opens with this sentence: “Reading is my favorite way to indulge my curiosity.” Right after, he writes that although his work lets him meet interesting people and visit fascinating places, books are still the best way to explore new topics. The same list puts a history book on the rise of ISIS, a book examining the role of federal policies in racial segregation in US cities and a novel side by side. This weakens the old article’s portrait of a man who “avoids fiction”.
We could not find the very widely repeated figure “Gates reads 50 books a year” in Gates’s own writing. Those who cite the figure are news and list sites; because we do not use those sites as sources and could not reach Gates’s own statement, we removed the figure from the article. We are not saying the figure is wrong; we simply could not verify it.
Mark Zuckerberg: What did “A Year of Books” say?
Marquette University’s Zuckerberg Files collection holds a Facebook post dated 2 January 2015, with Mark Zuckerberg recorded as its author. According to the record’s abstract, the post kicks off the “Year of Books” challenge and announces the first book: Moisés Naím’s The End of Power. In the 10 January 2015 archived copy of the challenge’s Facebook page, the description reads: “We will read a new book every two weeks and discuss it here. Our books will emphasize learning about new cultures, beliefs, histories and technologies.” The same archived copy shows 237,497 likes for the page.
A book every two weeks works out to a target of about 26 books a year (52 weeks divided by 2; CEOtudent calculation). The long Zuckerberg quote in the old article, by contrast, appears only in secondary publications. Because the archived copy of the original post sits behind a login screen, we do not present this quote as his words.
Elon Musk: What record exists?
Three popular claims about Musk (10 hours of science fiction a day, an encyclopedia at age 9, learning rockets from books) circulate in secondary publications. We could not find them in any dated transcript we were able to open and read, so we removed them. We also found the much-quoted line “I was raised by books” only in secondary publications and did not use it for the same reason.
The primary record we could open is the transcript of an oral history interview that The Henry Ford museum recorded at SpaceX on 26 June 2008. Asked whether he had studied Edison’s life, Musk says: “Yeah, I read books about him, absolutely.” Asked whether he is a programmer, he explains that he taught himself to program computers as a kid. In the same interview he says reasoning has to be rebuilt “from first principles” in the sense used in physics, that is, from the ground up. This record does not say “how much” Musk read. What it shows is that he used reading as part of a habit of teaching himself and rebuilding a subject from its foundations.
Do famous readers prove that reading brings success?
No. The inference “successful people read a lot, so reading brings success” contains three separate errors.
Survivorship bias. The leaders whose reading habits get discussed are discussed because they are already successful. The millions of people who read a lot and did not reach the same success never make these lists. If a sample consists only of winners, a trait the winners share may not be the cause of success; it may simply be common in that group.
Reverse causality and common causes. Reading rises with education, income and free time. In Pew Research Center’s 2021 survey, 39% of adults with a high school education or less said they had not read a book in the past year, compared with 11% of those with a bachelor’s or advanced degree (a gap of about 3.5 times; CEOtudent calculation). The share is 31% among those with household income under $30,000 and 15% among those at $75,000 or more. A wealthy, educated leader reading a lot does not show that reading made them wealthy; both can grow out of the same conditions.
Selective recall. Popular articles pick the most striking figure and repeat it. The fact that most of the claims in the old version of this article could not be found in a primary record is a concrete example.
This does not mean reading is worthless. The value of reading should be inferred from measured effects, not from famous examples; what leaders’ records show is how they connect reading to a learning system.
How much do adults actually read?
The table below shows published figures from two US sources. The NEA’s Survey of Public Participation in the Arts (SPPA) was conducted in 2012, 2017 and 2022; the most recent data we could access is from the 2022 round, and that round’s detailed technical report was published in August 2025. The survey excludes books read for work or school.
| Indicator | 2012 | 2017 | 2022 | Source |
|---|---|---|---|---|
| Adults who read at least one book in the past 12 months | 54.6% | 52.7% | 48.5% | NEA SPPA, Figure 8 |
| Adults who read novels or short stories | 45.2% | 41.8% | 37.6% | NEA SPPA, Figure 9 |
| Adults who read books and/or listened to audiobooks | – | 55% | 51.9% | NEA SPPA, Figure 10 and text |
| Indicator (Pew, January-February 2021, 1,502 adults) | Share |
|---|---|
| Adults who have not read a book in the past year (print, e-book or audio) | 23% |
| Among those with a high school education or less | 39% |
| Among those with a bachelor’s or advanced degree | 11% |
| Among those with household income under $30,000 | 31% |
| Among those with household income of $75,000 or more | 15% |
Sources: National Endowment for the Arts, Arts Participation Patterns in 2022: Highlights from the Survey of Public Participation in the Arts (October 2023); Pew Research Center, Who doesn’t read books in America? (21 September 2021). Figures are as published.
Three conclusions follow from the NEA data (calculations are CEOtudent’s). First, between 2012 and 2022 the share reading at least one book fell 6.1 points and the share reading novels and short stories fell 7.6 points; the relative decline in novel reading is about 16.8%. The NEA rounds this decline to 17 percent in its report and describes the 2022 novel-reading rate as the lowest share on record for the survey. Second, in 2022 more than half of adults (51.5%) did not report reading a single non-required book in the past 12 months. Third, adding audiobooks softens the picture slightly but does not reverse it: the share who read books and/or listened to audiobooks was 55% in 2017 and 51.9% in 2022.
The direction matters: in the US, regular book reading is becoming a minority behavior. Those who keep up a habit that is growing rarer build up knowledge and attention that set them apart more.
What does research say reading is good for?
Reading research answers two different questions: What is reading associated with, and under what conditions do we understand better? The table below brings together four findings specific to this article. We did not re-tabulate the meta-analytic data measuring the effect of print exposure by education level here; that data is in the deep reading article.
| Study | Design | Published finding | How to read it |
|---|---|---|---|
| Bavishi, Slade and Levy (2016), Social Science & Medicine | US Health and Retirement Study; 3,635 participants over 50; up to 12 years of follow-up | After all control variables, 20% lower mortality risk among book readers (HR 0.80); HR 0.77 in the group reading 3.5 hours or more a week; 23-month survival advantage in the unadjusted model | Observational and correlational; does not prove causation |
| Delgado, Vargas, Ackerman and Salmerón (2018), Educational Research Review | Meta-analysis; 54 studies from 2000-2017 (38 between-participants, 16 within-participants); 171,055 participants | Advantage of reading on paper for comprehension (Hedges g = -0.21); advantage larger under time-constrained reading and for informational texts; none for purely narrative texts | Small but consistent effect; grows with hurried reading |
| Clinton (2019), Journal of Research in Reading | Meta-analysis; 33 randomized studies, 2,799 participants (reading performance) | Negative effect of screen reading on performance g = -0.25; g = -0.32 for informational texts, g = -0.04 for narrative texts; paper readers estimate their own performance more accurately (g = 0.20) | Screens are fine for narrative; the gap shows up with dense informational text and self-assessment |
| OECD (2021), 21st-Century Readers (PISA 2018) | 15-year-old students, OECD countries | Compared with those who never or rarely read books, digital book readers read for enjoyment about 3 hours more a week, print book readers about 4 hours more and those who balance both formats about 5 hours or more; knowledge of strategies for assessing source credibility is the strategy most strongly associated with reading performance | Student data; should not be transferred directly to adults |
Sources: Bavishi et al. (2016), PMC5105607 author manuscript; Delgado et al. (2018), abstract only; Clinton (2019), abstract only; OECD (2021), executive summary. Figures are as published; the “how to read it” column is CEOtudent’s interpretation.
Three practical conclusions come out of this table.
Time works as a threshold, not as a promise. Bavishi and colleagues’ highest reading group read books for 3.5 hours or more a week; that comes to about 30 minutes a day (CEOtudent calculation). The authors report that the advantage among book readers was clearly larger than among newspaper and magazine readers and that cognitive scores mediated the relationship. Still, the study is observational: people who choose to read may differ in other ways too. This section is for general information and is not medical advice.
Choose the format to fit the text. Both meta-analyses point the same way: paper has the edge for informational and dense texts, and there is no difference for narrative texts. In Clinton’s findings, the gap in effect between informational and narrative texts is about 0.28 standard deviations (CEOtudent calculation). For a novel, a screen or an audiobook is a perfectly good choice; for a contract, a technical report or a dense strategy book, slower, focused and, where possible, print reading makes more sense.
Reading is not enough; you have to evaluate. In the OECD data, the strategy most strongly associated with reading performance is knowing how to assess a source’s credibility. In an era when AI tools produce summaries in seconds, this is the least copyable part of reading: telling what is true from what is contested.
What common pattern emerges from leaders’ records?
The verified records answer the question “how”, not “how much”. Gates starts reading from curiosity, that is, from a question. Zuckerberg ties reading to a cadence (two weeks), a theme (cultures, beliefs, history, technology) and a public commitment. Musk describes reading as part of teaching himself and rebuilding a subject from the ground up. Combined with the research findings, these three elements turn into a workable system for a working adult.
| Component | CEO move (direction and decision) | Student move (learning) | Basis |
|---|---|---|---|
| 1. Question | Tie the quarter’s reading to 1-2 real decisions or questions | Start every book with a single written question | Gates 2017: reading starts from curiosity |
| 2. Theme and cadence | Pick one theme and a fixed rhythm (e.g. a book every two weeks or 30 minutes a day); block it in the calendar | Target time, not page count | Zuckerberg 2015 page; Bavishi 2016 threshold (correlational) |
| 3. Format | Read dense informational texts with focus and, where possible, in print | Audiobooks or screens are fine for narrative and exploratory reading | Delgado 2018; Clinton 2019 |
| 4. Processing | Draw a one-page decision note from every book | Summarize from memory after closing the book; check the sources of its claims | OECD 2021 credibility strategy |
| 5. Application | Test one idea in real work within two weeks and note the result | Rebuild the idea from first principles in your own words | Musk 2008 oral history |
| 6. Pruning | Set a rule for dropping a book that is not serving you | Keep a not-to-read list | Strategic ignorance approach |
CEOtudent editorial framework: built on the findings of Gates Notes (2017), “A Year of Books” (2015), The Henry Ford Musk transcript (2008), Bavishi et al. (2016), Delgado et al. (2018), Clinton (2019) and OECD (2021). The framework is not a research result but a proposal for applying these findings.
The logic of the framework comes from the CEO and student duality. The CEO side decides what reading serves: which question, which theme, which rhythm, when to stop. The student side makes sure what is read sticks: retrieval, source checking, rebuilding in your own words. A reader with only the CEO side “consumes” many books but learns little. A reader with only the student side learns well but may read the wrong things. If you want to turn reading into a learning curriculum, the 90-day structure in building a personal curriculum with AI can be used alongside this framework.
Practical tool: a 4-week reading protocol
The protocol below spreads the framework above across four weeks. The 30 minutes a day is an illustrative assumption; adjust it to your own calendar. Under this assumption you read 3.5 hours a week, 14 hours in total over four weeks (CEOtudent calculation).
| Week | Goal | What to do | Measure |
|---|---|---|---|
| 1 | Question and selection | Write down the 1-2 most important decisions of the next 3 months; form a reading question for each; pick 2 books that serve these questions and deliberately cut 1 from the list | Is there a written question? Is the book tied to the question? |
| 2 | Cadence | Put a fixed 30-minute daily block in the calendar (illustrative assumption); read the dense informational book in a focused setting; turn off notifications | On how many days was the block kept? (target: at least 5 of 7 days) |
| 3 | Processing | After each session, close the book and write a 3-sentence summary; check the source of the most important claim the book makes | Number of summary notes; number of claims checked |
| 4 | Application and review | Test one idea from the book in real work; write a one-page decision note; decide whether to continue to the next book or drop it | Number of ideas tested; is there a decision note? |
Weekly 5-question checklist
- Is the book I read this week still serving the question I wrote down? If not, am I dropping it?
- Did I keep at least 5 of the 7 reading blocks I planned?
- Did I read dense informational text in a focused setting and narrative text in a flexible format?
- Did I check a claim I read all the way back to its source?
- Did I test an idea I read in real work this week, or set a date to test it?
The goal of the protocol is not to increase the number of books but to tie reading to a decision and learning loop.
What this does not mean
- Imitating leaders does not bring success. The verified records show the form of the habit; they do not guarantee its outcome. The warning in the survivorship bias section applies.
- Reading does not guarantee a longer life. Bavishi and colleagues’ finding is observational and concerns US adults over 50. This article does not contain medical advice.
- Screens are not bad and paper is not good. In the meta-analyses the difference is small and depends on text type; no meaningful difference was found for narrative texts.
- A claim that is not verified is not necessarily false. “Not verified” means we could not reach a primary record. We are not declaring these claims false; we are only not using them as evidence.
- US data is not universal. The NEA and Pew figures are for the US; reading rates may differ in other countries.
Frequently asked questions
Does Bill Gates really read 50 books a year?
We could not find this figure on Gates’s own blog; it appears only on news and list sites. What is verified in Gates’s own writing is that he describes reading as his favorite way to indulge his curiosity and that he publishes book lists on Gates Notes.
What was Mark Zuckerberg’s “A Year of Books” challenge?
It was a reading challenge launched on 2 January 2015. According to the description on its Facebook page, a new book would be read every two weeks, with an emphasis on cultures, beliefs, histories and technologies. The first book was Moisés Naím’s The End of Power.
What can be verified about Elon Musk’s reading habits?
In the 2008 oral history transcript we could open, Musk says he read books about Edison and taught himself to program as a kid. We could not verify claims such as reading 10 hours a day or finishing an encyclopedia at age 9 in a primary record.
How much should you read a day?
There is no universal threshold. In Bavishi and colleagues’ study, the highest group read 3.5 hours or more a week, or about 30 minutes a day; but this relationship is correlational. In practice, a sustainable amount of time that is tied to your calendar and serves a question is worth more than an ambitious goal you do not keep.
Paper, screen or audiobook?
It depends on the text. Two meta-analyses show that paper has the edge for informational and dense texts and that the gap disappears for narrative texts. For novels and exploratory reading a screen or an audiobook is fine; for dense texts you will make decisions from, focused reading is the safer choice.
Can AI summaries replace reading books?
A summary conveys what a text says, but in the OECD data the skill most strongly associated with reading performance is assessing a source’s credibility. Following a book’s argument and checking its claims exercises that skill; reading a summary does not. A summary can be a useful filter for choosing which book to read in depth.
Sources
- National Endowment for the Arts, Office of Research & Analysis (2023). Arts Participation Patterns in 2022: Highlights from the Survey of Public Participation in the Arts. October 2023.
- National Endowment for the Arts and 2M Research (2025). Arts Participation in 2022: A Technical Summary Report (By All Means, the Arts: A Full Report of the 2022 Survey of Public Participation in the Arts). August 2025.
- Gelles-Watnick, R. and Perrin, A. (2021). Who doesn’t read books in America? Pew Research Center, 21 September 2021.
- Bavishi, A., Slade, M. D. and Levy, B. R. (2016). A chapter a day: Association of book reading with longevity. Social Science & Medicine, 164, 44-48. PMC5105607 author manuscript.
- Delgado, P., Vargas, C., Ackerman, R. and Salmerón, L. (2018). Don’t throw away your printed books: A meta-analysis on the effects of reading media on reading comprehension. Educational Research Review, 25, 23-38. (Abstract only.)
- Clinton, V. (2019). Reading from paper compared to screens: A systematic review and meta-analysis. Journal of Research in Reading, 42(2), 288-325. (Abstract only.)
- OECD (2021). 21st-Century Readers: Developing Literacy Skills in a Digital World. PISA, OECD Publishing, Paris. (Executive summary only, Internet Archive copy.)
- The Henry Ford (2009). Collecting Innovation Today: Transcript of a Video Oral History Interview with Elon Musk, 26 June 2008, SpaceX, Hawthorne, California. (Internet Archive copy.)
- Gates, B. (2017). 5 amazing books I read this year. Gates Notes, 4 December 2017. (Internet Archive copy.)
- Zuckerberg, M. (2015). Zuckerberg Facebook post about A Year Of Books: The End of Power, 2 January 2015. Zuckerberg Files, Zuckerberg Transcripts 365, Marquette University e-Publications. (Record page and abstract.)
- A Year of Books, Facebook page, “About” section and posts dated 2 January 2015. Internet Archive copy, 10 January 2015.
Note: The “Leader reading claims audit” table is a CEOtudent analysis and the “Reading system for a working adult” table is a CEOtudent editorial framework; the 30 minutes a day in the 4-week protocol is an illustrative assumption. The percentage point differences, ratios and the 26 books a year and 30 minutes a day calculations are CEOtudent’s. Unverified claims removed from the old version: the unsourced “in 5 years” quote, the quote and the 5 hours of reading a day attributed to Warren Buffett, Bill Gates’s one book a week (50 books a year) figure, Elon Musk’s 10 hours of science fiction a day, Britannica at age 9 and the line about learning rockets from books, Mark Zuckerberg’s long “media diet” quote and the views attributed to Oprah Winfrey. The much-repeated line “Musk was raised by books” was also not used because it comes only from magazine sources.
This content was compiled with the support of AI following in-depth research, then written and prepared for publication by the CEOtudent editorial team.
This post is also available in:
){kind=link}
&description=&image=https://ceotudent.com/wp-content/uploads/2021/07/ctu83s-1024x682.jpg){kind=link}