Every day, we interact with two kinds of goods. The first kind is acquired and shared instantly, is weightless, impervious to damage, easy to customize, and impossible to lose. Even a child can carry thousands of it at a time. The second kind requires travel to acquire or share, is difficult to alter, cumbersome, easily lost, and can be damaged in a myriad of ways. Only a few of its kind can be crammed into a single bag. Despite the many advantages of the first kind –– digital goods — companies find again and again that people value and are willing to pay considerably more for the latter –– their physical counterparts. Our research aims to explain this puzzling behavior.

Modern life has been transformed by the widespread digitization of many consumer goods, from books, to magazines, newspapers, music, movies, airplane tickets, and calculators. Digital photographs, first commercialized in 1990, are now taken more often than print photographs. Yet, despite the many advantageous features of digital goods, physical goods appear to retain greater allure. Print books are still the dominant format, Blu-ray and DVD sales continue to grow, as does demand for physical prints of digital photographs people already own.

We began exploring this paradox on a bright summer day in Boston, by offering tourists on the Freedom Trail a souvenir photograph with an actor dressed as Paul Revere. We did not charge for the photograph, but asked each tourist to make a donation to the historical group maintaining Old North Church in exchange for it. Unbeknownst to participants, they were randomly assigned to one of two groups. One group received an instant digital photograph immediately via email, the other group received a still-developing Polaroid picture. While tourists in the first group chose to give an average of $2.29, remarkably, tourists in the second group gave $3.39­ — 48% more for the Polaroid that they could touch and hold.

We find in our research that these tourists are not unique. Despite the many advantages conferred by digital goods, comparable versions of physical goods are valued more. When a physical good such as a paper book, a printed photograph, or a DVD is digitized, it loses some of its value to buyers. Our experiments suggest that the key driver of this value loss is not the resale value of the good, or how much it costs to make, or how long it can be used, or whether it’s unique or popular. We find that the key difference is that digital goods do not facilitate the same feeling of ownership that physical goods do.

The very feature that imbues digital goods with their unique abilities—their immateriality—is also what impairs our ability to develop a sense of ownership for them. Because we cannot touch, and hold, and control digital goods in the way that we interact with physical goods, we feel an impaired sense of ownership for digital goods. They never quite feel like they are ours, and when we feel that we own a thing, we psychologically inflate its value. As a result, digital goods don’t enjoy this premium we extend to things that we own.

In additional experiments we ran with online samples and business students, we found that the tendency to value digital goods less than physical goods extends to a variety of products, from popular films to novels to course textbooks. We observed the same effect whether we measured value by allowing research participants to pay whatever they wanted for goods, having them report the most they were willing to pay, or by measuring their likelihood of buying a digital or physical copy of a good when both were sold at the same price. The greater value ascribed to physical than digital goods persisted when we accounted for people’s estimates of production costs and retail prices. It even held for goods with no resale value. Plausible alternative explanations, such as physical goods lasting longer or being more enjoyable to use than digital goods, also failed to explain this difference.

Only a difference in the extent to which people feel a sense of ownership for digital and physical objects explained their preference for the physical format. Indeed, the value gap disappeared for goods participants rented and expected to give back. Students were not willing to pay more for a physical than digital textbook they would return at the end of a course.

This difference in ownership also allows us to identify when people will value digital goods no less than physical goods. Because ownership entails a link between a person and an object, we found the gap in their value increased when that link was easy to form and disappeared when that link was hard to establish. Participants valued a physical copy of The Empire Strikes Back more than a digital copy, for instance, only if they considered the Star Wars series to be films with which they strongly identified. Participants who weren’t Star Wars fans valued physical and digital copies similarly.

We think that the difference in ownership that people feel for digital and physical goods may be due to a weaker sense of control over digital objects. People feel ownership for a good to the extent that they can touch, hold, and control it. Indeed, we found that people with little need to feel in control over their environment reported no preference for physical relative to digital goods, whereas people with strong need to feel in control reported a substantial preference for physical relative to digital goods. Participants in one experiment reported how much they would be willing to pay for a new copy of a book by their favorite author, how much they anticipated feeling ownership for that book once they acquired it, and their need for control – an index of the extent to which a person needs to be able to exert control over his or her environment. We found that participants with a higher need for control were most likely to believe they would feel greater psychological ownership for digital than physical copies of the book, which in turn lead them to place a greater value on physical copies than digital copies of the book, even when their estimates of its retail price and production cost were included in the statistical model. Perceived ownership and valuation of physical and digital books did not differ as much for participants with a lower need for control.

Our results both identify the source of value lost through digitization, and how goods can be designed to recapture this lost value. Digital goods can be made to visually resemble their physical counterparts — using skeuomorphism — like Apple’s display of digital books on a virtual wooden shelf in its iBooks application.

Ownership may be achieved by increasing users’ feeling of control through touch interfaces, and customization opportunities that involve users in the production or design of the product. In this vein, our findings foreshadow a looming friction in the adoption of autonomous devices such as self-driving cars. Just as modern automotive purists reject automatic transmissions, which are technically superior but allow them less control than manual transmissions, people may devalue autonomous devices that require little or none of their input. This lack of physical input is likely to dampen their feeling of being in control of autonomous devices. As a result, they should be less likely to establish a feeling of ownership for autonomous devices, and those devices will not benefit from the value premium extended to goods for which people feel psychological ownership.

Finally, our findings provide a window into the different perspectives that people apply to the theft of physical and digital goods. Because perceived ownership is impaired for digital goods, people may not feel that their piracy causes the same harm to their owners as does the comparable theft of physical goods.

The digitization of content frees us – and the environment – from the burden of accumulating material objects for our information and entertainment. Despite the considerable gains in consumer welfare that digital goods thus provide, our preferences are not solely determined by how useful or practical we find these innovations. Our research demonstrates how new technologies can collide with the architecture of cognition, and the cognitive taxes that the mind can impose.