Thursday, July 1, 2004

Valuing Bestselling Books

STANFORD GRADUATE SCHOOL OF BUSINESS—Each week millions of readers look at the New York Times bestseller list to see what everybody else in the country is reading. And as soon as a title hits the list, booksellers typically push the book to the front of the store and slash its price by as much as 40 percent.

So it seems reasonable to assume that once a book makes the list its sales will really take off—if not for the lower price then because readers might view bestseller status as a sign of quality or because they don't want to miss the action. According to Alan Sorensen, an assistant professor of strategic management who has studied the effect of bestseller lists on sales of hardcover fiction, the majority of book buyers seem to use the Times' list as a signal of what's worth reading. Relatively unknown writers get the biggest benefit, while for perennial best-selling authors such as Danielle Steel and John Grisham, being on the list makes virtually no difference in sales.

"There is an effect, but it's small—much smaller than most people would have expected," says Sorensen, who studied sales records for 2001 and 2002. "What's remarkable is there is this dominant downward [sales] trend," he says, describing the typical sales path over time in the fiction titles he studied. Most sales occur soon after a book hits the shelves and gradually peter out. "If anything, what appearing on the [bestseller] list does is not so much cause your sales to increase from one week to the next, but rather to decrease at a slower rate."

Of course, looking at bestsellers alone wouldn't prove that any week-to-week sales changes were caused by the list itself. To answer the causal question, Sorensen needed a comparison group: books that sold well but somehow missed the list. So he looked at data from Nielsen BookScan, a sales monitoring service that tracks retail sales of books across the nation. Unlike the New York Times, which samples sales from only some stores, Nielsen BookScan captures most actual sales.

Consequently, Sorensen found differences in the two lists. In fact, in the two years he studied, Sorensen found 109 different books that failed to make the Times list even though Nielsen reported they sold more copies than other titles on the Times' list. Thus, if Sorensen saw sales rise on the Nielsen BookScan data the week following that same title's appearance on the New York Times list but saw no similar increase for a different top-selling Nielsen book that wasn't on the Times list, he inferred that something about the appearance on the Times list caused the subsequent jump in sales.

Based on these comparisons, Sorensen estimates that previously best-selling authors got the least benefit from being on the New York Times list, while unknowns had the greatest jump in sales. On average, he estimates, appearing on the Times list might increase a book's first-year sales by 13 to 14 percent, but for first-time authors sales probably would increase by an impressive 57 percent. And for established authors like Danielle Steel or John Grisham whose every novel seems to become a bestseller, "the list has no discernible impact on sales," writes Sorensen.

This pattern, he says, suggests the bestseller list primarily tells consumers what may be worth reading. "It's free advertising for new authors who make it to the list," he says. With a well-known author, on the other hand, people don't need a bestseller list to help them decide whether to buy the book.

Careful to isolate the bestseller list from other likely causes of higher sales, Sorensen also looked at the famous "Oprah effect," the stunning way being chosen for Oprah Winfrey's on-air book club immediately catapults a title onto the bestseller list. Though reluctant to name numbers because there were few Oprah titles in the sample, Sorensen says the Oprah effect is "many times bigger" than the bestseller effect.

Looking at sales of individual titles is fun, but for economists, says Sorensen, a more interesting question is the effect of bestseller lists on overall product variety. If bestseller lists are indeed generating extra sales, are those sales stolen from other books? Or are those extra sales literally extra sales, bought by people who wouldn't otherwise have bought any book? "It's a really hard question, and I don't have the ideal data for answering it," concedes Sorensen. But he has indirect evidence suggesting that being on the list does indeed generate extra sales. If he's right, everyone but the perennially best-selling authors and their publishers may have a little bit less to grumble about.

Related Information

Bestseller Lists and Product Variety: The Case of Book Sales
Alan T. Sorensen, GSB Research Paper Series, 2004

Making the List: A Cultural History of the American Bestseller, 1900-1999, Michael Korda, Barnes & Noble Books, 2001

Book Business: Publishing Past, Present, and Future, J. Epstein, W.W. Norton & Co., 2001

Social Economics: Market Behavior in a Social Environment, G. Becker and K. Murphy, Harvard University Press, 2000

"Do Mergers Increase Product Variety? Evidence from Radio Broadcasting," S. Berry and J. Waldfogel, Quarterly Journal of Economics, 116(3), 1009-1025 (2001, August)