This article offers a model integrating heterogeneously moral preferences to overcome the seemingly irrational tendency of individuals to freely share data and knowledge. We build on recent literature showing that moral preferences are favored by evolution theoretically, and have a strong explanatory power empirically, to model individual sharing behavior. Our analysis highlights the limit of financial incentives and the importance of promoting a sharing culture by enhancing awareness. Shedding light on how people respond not only to financial but also moral motives, we contribute to the ongoing policy debate on the design of effective open science policies.
| INTRODUCTIONRegularly, when confronted with a new concept or when we want to verify a piece of information, we head to consult the corresponding Wikipedia page. As teachers, we consult available online classes to get inspiration for the design of our courses. As programmers, we seek precious debugging advice on Stack Overflow. And as researchers, we benefit from the freely available research and voluntarily shared data. All these informational public goods bring great value in terms of social welfare. However, the availability of this information largely depends on the willingness of contributors who most often voluntarily share without any financial compensation. In this context, it is particularly relevant to better understand the incentives that motivate individuals to freely share knowledge and data with the broadest audience. This article aims to address this question.The act of sharing knowledge and data has the properties of a social dilemma (Lichbach, 1996;Olson, 1965). Contributing to publicly available knowledge maximizes social welfare, but the time and effort required and the lack of compensation drive individuals out of it. With the common assumption of self-centered homo oeconomicus agents, individuals restrain from contributing and the level of sharing is suboptimal. To address this issue, economists have developed various incentive mechanisms, such as taxes, intellectual property rights (IPRs), and public subsidies (Arrow, 1962;Samuelson, 1954;Stiglitz, 1989). However, a significant share of the publicly available knowledge lies outside of these designed incentive mechanisms. Although often acknowledged by economists working on knowledge (Arrow, 1962;Dasgupta & David, 2002;Stephan, 1996), few models integrate non-self-centered motives. The classical economic approach thus fails to explain the contribution to crowdsourced initiatives, such as open-source software and public repositories.Voluntary contribution to public goods is not limited to knowledge and data sharing. In fact, whether by financing charity actions that provide no direct personal benefit (Andreoni, 1988;Carpenter & Matthews, 2017), or by performing