National Academies Press

• 1. Introduction

  ›

  The National Academies Press publishes nearly 500 titles a year, covering sciences, engineering, medicine, and transportation. Its mission sounds like a riddle: disseminate the work of the National Academies as widely as possible, and remain financially self-sustaining through sales at the same time. For most publishers, giving away your product and selling it are simply incompatible goals. But for the NAP, that tension became the engine of two decades of bold experimentation. What does it actually mean to give knowledge away and still stay afloat? And what happens when millions of people start reaching for science the moment it becomes free?

• 2. First on the Web

  ›

  In 1993, the organization was still called the National Academy Press. That year it became the first self-sustaining publisher to make its material available on the Web, free of charge, in an open access model. By 1997, a thousand reports were online as sequential page images, beginning with page i, then ii, then iii. Navigation was minimal. The interface was spare. But the experiment was already yielding a counterintuitive result: by 1998, internal evidence suggested that free online access was increasing print sales, not cannibalizing them. That finding would shape every strategic decision the NAP made for the next quarter century.

• 3. Building the Openbook

  ›

  From 1998 onward, the NAP developed a tool called the Openbook, an online navigational envelope that gave each page a stable URL and allowed readers to search within and across chapters. Over time, page images gave way to HTML chunks, letting users browse freely rather than flip through static pictures. This page-by-page navigation arrived long before Amazon introduced its Look Inside feature or Google launched Book Search. The NAP kept layering on capabilities: Related Titles in 2001, Find More Like This Chapter in 2002, Chapter Skim in 2003, Search Builder and Reference Finder in 2004, and Active Skim along with an enhanced Search Builder in 2005. To help search engines find the material, the first page of every chapter was built to include the first and last ten pages of OCR-extracted text, making chapters indexable from outside the site.

• 4. The Cannibalization Experiment

  ›

  In 2003, the NAP published the results of a pricing experiment designed to measure the so-called cannibalization effect: would giving a book away as a PDF kill sales of the print version? The experiment was funded by a Mellon grant and conducted with the University of Maryland Business School. Researchers intercepted buyers just before they completed an online order and offered a randomly generated discount on the PDF: 50%, 10%, 100%, or 70% off list price. If the buyer declined, the NAP offered one additional step down in price. The result: 42% of buyers accepted the free PDF when it was offered, while 58% went ahead and paid for a printed copy. A related finding sharpened the picture further. When the NAP offered free PDF access to low-selling titles, roughly the bottom half of its catalog by sales, print sales fell by only 33% over eighteen months. At the same time, dissemination increased by a factor of one hundred. Through the middle of 2006, as reported at the AAUP annual meeting, the NAP remained financially self-sustaining even as it expanded its online offerings. Barbara Kline Pope, executive director of the NAP, and Michael Jon Jensen, who served as director of publishing technologies from 1998 through 2008, both wrote and presented extensively on the evolving strategy of staying viable in a world of abundant free content.

• 5. Making PDFs Free

  ›

  On the 2nd of June 2011, the NAP announced that every report from the National Academies of Sciences, Engineering, and Medicine would be available as a free PDF download. A 2011 analysis framed the decision not as a leap of faith but as the outcome of more than a decade of careful testing. The press had already tried free access in developing countries, studied how digital availability shifted print demand, and run the Mellon-funded pricing experiment. Internal research acknowledged that free PDFs would cut into some revenue. But projections suggested those losses could be absorbed by shifting marketing online, outsourcing printing and distribution, and discontinuing products that were not profitable. The rationale was straightforward: expand dissemination dramatically while keeping the finances intact. By early 2025, those PDFs had been downloaded 25 million times. By April 2026-15,485 books were freely available online or for download.

• 6. Who Actually Reads These Reports

  ›

  A 2022 study published in Proceedings of the National Academy of Sciences took a close look at 1.6 million downloads of National Academies reports, using machine-learning techniques to classify how readers were actually using them. The findings split almost evenly. Roughly half of all usage was academic: research and teaching. The other half came from people using the reports to improve professional practice, inform personal decisions, or simply learn. That second half is the more striking finding. Scientific consensus documents, often written by committees for specialist audiences, were being read by people making decisions in their own lives and careers. The data offered concrete evidence that open access to high-quality scientific information reaches well beyond traditional expert audiences.