The problem of the value of digital service user data is methodologically difficult, but it is also used in debates about the direction of government policies. As a result, it is not difficult to encounter opinions based on unwarranted simplifications. The purpose of this report is to outline the main elements of the data value problem.
So far, attempts to estimate how much (zlotys, dollars) user data is worth (1) to individual Internet giants and (2) to the users themselves have failed.
This is due in part to the lack of adequate data (e.g., on the detailed income and cost structure of technology firms) and in part to the methodological difficulties of monetary valuation of goods in situations where the goods are not exchanged for money. Therefore, one should also be very careful about the information and arguments on which decisions on the direction of state policies are based.
The report separately presents the methodological problems of estimating data value from the perspective of service providers and from the perspective of users.
From the perspective of service providers, the value of data cannot be estimated by dividing the value of ad revenue by the number of users. Unfortunately, this type of error is repeated in non-scientific publications. Reliably estimating the impact of user data value on enterprise value or even enterprise profit itself is challenging, not least because of the difficulty in distinguishing the impact of data value from the impact of know-how, technology, etc.
From the users’ point of view, it would seem that the obvious way to observe the value they assign to their data is to observe the actual exchanges that users participate in (e.g., using social media). However, here too it is difficult to distinguish between the value of data and, for example, the value of user attention or actions. On the other hand, it may seem unattractive to some that services available at no charge do not allow for easy assignment of price.
Among other reasons, researchers use survey (declarative) or quasi-experimental methods that involve asking users about their willingness to pay (or accept payment) in hypothetical situations. Unfortunately, such research is sometimes applied incompetently, without consideration of the limitations of the method and the specific context in which it is to be applied. These methods are not adequate to determine how much PLN a Polish user is willing to pay per month for “access to Facebook without personalized ads”.
However, the dispute over “fairness of exchange” in this context is largely not an economic dispute, but a dispute over what we think is fair, morally justified. Focusing on trying to estimate monetary values distracts us from these more fundamental issues.
Focusing on the value of data also distracts from the, often more important, value of user attention or actions. Noting this second aspect allows for a better understanding of the morally and economically relevant similarities of digital and “non-digital” services. These similarities may lead us to conclude that, in fact, digital services are based on exchanges of a nature that we intuitively morally accept in “non-digital” services.
“Irrationality” of users has, according to some, a particular impact on the fairness of exchanges between users and digital service providers. It is not clear, however, that the impact of phenomena grouped under the label of “irrationality” is large enough that it should affect our moral assessment of the fairness of exchanges or the direction of state policies.
State interventions motivated by insufficient fairness of exchanges between users and digital service providers always carry costs. For example, the introduction of digital user fees enforced by the reduction in advertising funding for the service may exclude less affluent users. Moreover, some contemplated interventions may simultaneously in effect exclude less affluent users and provide “more privacy” only to those users who do not currently care enough about it to expend the small effort required to change the default situation.