Why College Students Sell Back Their Textbooks

We analyzed 30,941 decisions to sell a used textbook. The most important factor was price. Higher prices increased the probability of selling. Age, credits completed, in-major status and being male all reduced the probability of selling. Ethnicity and financial aid status were insignificant. ********** A large number of undergraduate students in the U.S. sell their textbooks back to campus bookstores every semester. These used college textbooks in turn constitute a large proportion of the textbooks sold in subsequent semesters. These used book sales have profound effects on the behavior of textbook publishers and students. Publishers are powerfully motivated to price textbooks higher because they cannot recoup their development costs by selling a new book to every student who will be assigned to read the textbook. Publishers are also powerfully motivated to bring out new editions in order to "kill off" the used book market that depends on the earlier edition. Students are encouraged to pay higher prices for textbooks because they know that some textbooks can have significant resale value. Other students do not buy the textbooks at all because of the high prices. Previous Research Economists have theorized how buybacks and used goods affect markets. Fudenberg and Tirole (1997) modeled textbook sales as an example of "monopoly pricing of overlapping generations of a durable good." Talaga and Tucci (2001) surveyed college students to discover that in-stock status at the campus store, campus buyback policies, and price were very important in the students' choice of buying at the campus store versus buying textbooks online. Shulman and Coughlan (2005) presented a model that showed that a renewing population of consumers (new students) can lead to a profitable secondary market for a durable good in which it is advantageous for the manufacturer (publisher) to abandon the market entirely after the initial sale period (first semester or year of a new edition). Most of this work is theoretical but the issue so salient that some social scientists have waded into the fray by proposing that textbooks be free. Faculty members even act on this impulse by posting on the Internet "open source" textbooks they have written (DeLespinasse 2008). We believe that our results are first census of textbook sell/don't sell decisions that are correlated with individual sellers' characteristics. Data We collected data from the campus store at the end of the fall and spring semesters in academic years 2005-2006 and 20062007. The campus store runs a special counter service to buy back textbooks from students during the final exam period of each semester. The database thus begins with a list of students and the textbooks they sold back to the campus store. To this list we add every student in every course who did not sell these textbooks back to the campus store. It is important to note that we therefore automatically underestimate the total number of textbooks sold back because we did not capture sales in other venues, such as direct sales to another individual student or through an Internet site, etc. However, anecdotal evidence indicates that a vast majority of textbook sales in this time period were back to the campus store because of its convenient location, quick settlement, and high prices, especially for books that could be used again on campus. Students must show their college identification card to use this service in order to prevent non-students from using the counter to sell stolen textbooks (a significant problem on some campuses). Thus the buyback list contains sellers' student id numbers. We used these identification numbers to correlate the students' decisions to sell with information from their campus academic record, which includes grades, age, financial aid status, ethnicity, gender and other information. The college Institutional Review Board approved the database merge and analysis in August 2006. …