What if you could print a perfect-bound volume for as little time as it takes to brew a cup of coffee? That is the premise behind the Espresso book machine, which turns digital PDF files into paperbacks in minutes. Jason Epstein and Dane Neller, chairman and CEO respectively of On Demand Books in New York, the company behind the Espresso book machine, believe their technology has the potential to transform book publishing. Epstein, who was editorial director of Random House for 40 years, recently wrote a widely read essay in The New York Review of Books about the impact of digital technology on publishing. In an interview with Stephen J. Kobrin, editor of Wharton School Publishing, and Knowledge at Wharton, Epstein and Neller discuss their views on where the publishing business is headed.

Epstein will speak at a conference on The Future of Publishing in New York on April 30. The conference is being organized by the Wharton Interactive Media Initiative, Wharton School Publishing, and Knowledge at Wharton.

An edited version of the discussion appears below:

Stephen J. Kobrin: Mr. Epstein, I was fascinated by your article in The New York Review of Books. My first question is what will a book be in the digital age? Will it simply be a traditional book digitalized? Will it take new forms? Will the term “book” still be relevant?

Jason Epstein: We are talking about two different things — content and format. As far as content is concerned, that has been extraordinarily stable over centuries. A book will remain a book in that sense, no matter what the technology is. That doesn’t mean there won’t be innovations. For instance, there may be short books that people can read on cell phones, the way they do in Tokyo, but basically content will remain what it always has been.

Authorship, too, will remain what it always has been. There is a lot of talk about community publishing, about groups getting together to write books. That’s impossible. Authors don’t encourage collaboration of that sort unless they are writing books that require collaboration, as historians sometimes do.

Eventually all the content in the world will be digitized and available everywhere in the world in that format. It will be downloaded in one way or another — either to be read online or to be printed in book form by Espresso or similar devices. What the proportion will be is impossible to say at this point. My guess is that the dominant format will be the traditional codex [the format used for modern books], which is convenient, durable and portable. At least in America, when you buy a physical book in the bookstore, you own it; you can do what you want with it. When you download a digital file onto your Kindle, you don’t own it for fear that it will get loose. It could become viral and then the copyright would be lost, so publishers have to protect that.

Dane Neller: The Financial Times recently published an article about e-books based on a study by PriceWaterhouseCoopers. In the U.S. in 2008 — just for consumer books — e-books were a little sliver of the overall market. This is estimated to grow by 2013 to about 3% worldwide, which leaves print at around 97% of the market.

Now this is the key statistic. The print market is broken down between offset and toner based — or digital and offset. The toner based market, which is what we are in, represents about 6% of the books market in the U.S. right now. It’s twice as big as the e-book market and estimated to grow to 15% in the next three to five years. What is often not understood is that the toner-based, digital print on-demand market is larger than e-books and is experiencing a growth curve that is higher than e-books.

Knowledge at Wharton: Book publishing is witnessing a shift from physical inventory to digital distribution. How do you see the implications for the different players in the value chain — authors, publishers, retailers and customers?

Epstein: When this all settles down in four or five years, much of the publishing infrastructure will have been eliminated. Digital files can go from the publisher to the end user in one step with no intervening process. To the extent that this becomes possible, obviously the price of a book to the consumer will be lower because there will be less cost involved. The reward to the writer will be greater for the same reason. The possibility of profit to publishers will also be greater. They won’t be investing in inventory and the infrastructure that goes with creating an inventory. In other words, the publisher’s risk and cost will be lower.

In the digital future, those three conditions will prevail. Books will be cheaper; more [money] will go to the author; and the publisher will get a better break. But those are the ultimate effects; there will be a lot of static between now and then.

Knowledge at Wharton: How do you find publishers responding to these changes? What is right and wrong about the way they are reacting?

Epstein: Let me tell you an anecdote. When I first joined the publishing business, I was 21 years old. It was in 1951. I worked at Doubleday, and it occurred to me that it might be a good idea to take some of the books that my colleagues in graduate school — where I had just come from — weren’t able to afford because they were only available in hardcover and put them out in paperback.

There was a new market because of the GI Bill; millions of people who had never gone to college before were now going to college to study. By some miraculous effect, the man I worked for said, “Okay, we’ll try that.” So we printed 12 of those books in that format. It was a permanent paperback format that would be sold in bookstores, not like paperbacks sold in newsstands. To make a long story short, it revolutionized the book trade.

Now, none of my colleagues — none of the people who had been in the book business all their lives — noticed that opportunity. In fact, when I suggested it, they said, “Oh, this will never work. What does he know about anything?” There is a lot of inertia in the publishing industry, and it hasn’t changed since 1951. Publishers are loath to see the next step.

I can talk about other innovations that ought to have come out of the book business but didn’t. For example, Amazon should have come out of the book business. The publishers should have seen that. In fact, I did see that when I created my Reader’s Catalog [during the 1980s], which in effect became Amazon.

The digitization process [of books] also should have begun with the publishers. They should have seized this opportunity. They should have taken the initiative and not left it up to Google. Eventually they will come around, whichever of them are left. This change is going to be revolutionary.

Look at China. University towns there have 2 million students. We just signed a deal with the China Publishing Group, the country’s largest publishing company. Their goal is to make available Espresso book machines throughout China and reprint Chinese books, English books, French books and others. Their focus now is mainly on scientific and professional volumes, but over time that will change. You are talking about 1.6 billion people, some of whom have never seen books before. Pakistan has 2 million people who go to Islamabad University. This is a phenomenon of enormous worldwide growth.

Knowledge at Wharton: Consider another aspect of books going digital. If all books exist essentially as single droplets in a vast digital cloud, how will each book be found? What implications does this have for the marketing of books?

Neller: There are multiple answers to your question. Let us take the university setting and consider library books. That is an enormous growth area. What will happen? One way these books will be found is through the library’s electronic catalog systems. If you are a student, you will have access to the entire database of the library as well as have the ability to choose new acquisitions. These could either be in e-book form or print form. So you will discover books through that mechanism.

Outside the library systems, books will be made accessible through companies like Google who will make them discoverable. Readers will find books through Google, Amazon or websites of special interest. These mechanisms of discovery are already in place.

Kobrin: You are talking basically about accessibility and availability. But what about awareness? Right now a lot of people become aware of books in a bookstore. They pick the book up on the shelf and leaf through it. When books are part of a vast digital cloud, how will people become aware of them? How will they know they exist?

Neller: Word-of-mouth has always been the best way to sell a book — and the web represents the word-of-mouth medium in spades. There’s never been anything like it. There will be countless opportunities for books to get viral publicity on YouTube and its successors. That is something we never had before. Amazon does a very good job of building awareness about books. Believe me, Google will as well.

Knowledge at Wharton: What role do you see for the Espresso book machine in this brave new world of digital publishing? Can you give us some background about how you came up with the idea for the machine?

Epstein: For my entire career I have been concerned about backlists. Backlists keep publishers in business. They are like annuities; these are books you sell year after year with no effort and at very little expense. Our civilization resides in our backlists.

As I wrote in my article in the New York Review, in the mid 1970s I noticed that backlists had begun to deteriorate rapidly — 60,000 or 70,000 titles a year or more and growing. There were two reasons for this. One was a change in the tax law that made it no longer possible to write off unsold inventory as a business expense. The other was the demographic shift from cities to the suburbs. The large independent bookstores that depended upon city markets were closing down and being replaced by smaller shopping mall stores that couldn’t carry backlists. They had to depend upon current books.

I tried to offset this problem by starting a project called the Reader’s Catalog back in the mid 1980s. This was a telephone-size directory of some 40,000 backlist titles that you could order through an 800 number. The Internet wasn’t yet commercialized. The Reader’s Catalog was a huge success, but we were losing money on shipping. It was very complicated, and I decided to give it up. Then Amazon picked it up and put it online — and it became Amazon.

That didn’t solve the backlist problem, though. When digitization came in, I started to see there was an opportunity to solve it on a grander scale. With digitization you can restore every book ever printed to print. You can have a vast — almost infinite — inventory at very little expense — and you can deliver a book anywhere in the world on demand.

I discussed all this in a series of lectures I gave at the New York Public Library in 1999. One of the things I said was what’s necessary to complete this technology is a device, like an ATM, that can receive a digital file on demand anywhere in the world and print it out as a library quality paperback book. That lecture was printed — and soon someone called me and said, “I know where there is such a machine.” I went to see it — and that’s how it all began.

Knowledge at Wharton: How has the rise of e-readers like the Kindle and the Sony e-reader redefined the book business? What impact will the iPad have?

Epstein: It is hard to predict how that will sort itself out. I think eventually one-purpose machines like the Kindle and Sony e-reader will be replaced by multi-purpose machines such as the iPad. People who like to read books on screens will read their books on those devices along with their newspapers. But as I said before, most people in most parts of the world will prefer physical books.

Kobrin: What will happen to territorial rights? If you have an Espresso machine, you can download and print a digital book from any place in the world. As the business is structured now, the rights are regulated by territory — English rights, Dutch rights, American rights, Canadian rights, etc.

Epstein: That will have to change. Territorial rights become superfluous when content is in a digital cloud and can be downloaded anywhere instantly. There will have to be another understanding. Territorial rights are an artifact of the old Gutenberg system and like so many of those artifacts, they will have to change. How is that going to happen, when will it happen, who is going to make it happen…who knows? It has to happen.

Kobrin: Have you run into that problem with Espresso machines in other countries downloading U.S. books?

Neller: We deal with it every day. Our Espressnet software product filters territorial rights. Many times we have rights in the U.S., sometimes we have rights in Australia. But we have to filter them. Even with public domain books, not every country has the same rules so we have to filter those.

Epstein: Eventually that aspect of traditional infrastructure will give way and there will be a more convenient system, just the way that copyright rules will have to become universal and simplified. We can’t have different copyright conventions everywhere in the world.

Knowledge at Wharton: As the publishing business is transformed, how will the structure of the industry evolve? Will there be a few large global monopolies or smaller companies?

Epstein: It won’t evolve; it will devolve. It will become the way it was before these conglomerates were put together. You have to understand that these conglomerates were not rationally assembled. That happened out of necessity because of certain kinks in the way the system worked.

Fifty years ago, publishing was essentially a cottage industry. Each company was run by, effectively, a handful of editors with minimal management. Many services were farmed out. That’s how Random House worked; that’s how Simon & Schuster worked; that’s how they all did. But when the marketplace changed from one based on large independent bookstores that could carry extensive backlists to one with stores in shopping malls, where the backlist didn’t figure, bestsellers became a necessity. Publishers had to provide bestsellers for these shopping mall stores to remain in business.

The rights to potential bestsellers were then put up for auction. Smaller publishers could no longer afford to compete so they joined hands with larger publishers. Random House picked up a number of imprints — Knopf, Pantheon and so on. Even that didn’t work because even those proto-conglomerates — like Random House and Simon & Schuster back in the 1990s — couldn’t compete. They couldn’t put up these $1 million- or $2 million-, $5 million-, or $10 million-risks, so they collapsed into the present conglomerates.

Conglomerates were not rationally assembled as the most efficient way of publishing books. They were assembled because someone had to be able to fund these auctions — these gambling situations — we got ourselves into. But these conglomerates don’t function. There’s much too much overhead and complexity. There’s too much unnecessary management. Lawyers get in the way of everything. That can’t last. Digitization I think will inevitably have to replace it.

Neller: Let me add one more nuance. This is not so much about the publishing house as it is about the supply of content. It is true that the consumption of content will be massively decentralized either through Espresso book machines or e-readers or whatever. The supply of content, though, is a different matter. There is growing concern about control over supply exercised by companies like Google and Amazon in the U.S. or the Gallica digital library in Europe [which claims to have more than 1 million digital titles]. These organizations are developing very large digital discovery mechanisms and digital warehouses. As such, though some parts of the publishing business such as consumption may return to a decentralized model, there also a trend towards centralized digital hosting and discovery.

Kobrin: With digitization and print on demand, do you believe we will move away from a focus on bestsellers and have a much larger number of books that sell far fewer copies?

Epstein: I think that’s true. Of course, there will still be bestsellers. There will still be Stephen King and John Grisham, but that is a separate activity from mainline bookselling. All John Grisham needs is a printing company and a truck. That’s all the publishing work that has to be done, because those books are presold. You just ship them to Costco and other distributors and there you have it. That’s a different arrangement than what it takes to publish a real book of substance. I don’t mean to say those aren’t real books. They are just different than the books we are talking about.

Culture and Marketplace Effects on Perceived Price Fairness: China and the USA