Reddit recently announced its plan to tokenize - i.e., converting to cryptocurrency - their so-called Karma points, which is a measure of reputation based on how much positive engagements (likes, shares and comments) an individual has received from other users on their posts. This development, which is also on the agenda of many other social media platforms, raises a crucial question: What is the effect of decentralizing a social media platform – in which users earn cryptocurrency based on their reputation – on online user behavior? To explore this mechanism, we recruit U.S. citizens via MTurk to run an online survey-experiment in which participants are exposed to several treatments to assess the extent to which a hypothetical monetary incentive affects the type of content they would post on their timeline. While in all treatments participants can earn a hypothetical reward for the content they choose to share, the treatments vary across two dimensions: a) whether they are penalized (i.e., losing points) or not for posting objectionable content such as misinformation and hate speech, and b) whether it is clear to participants exactly how rewards and penalties are linked to their behavior. Manipulating these two dimensions is important to understand whether participants pay more attention to the quality of the content they post depending on the extent to which they are aware of the true costs and benefits when posting objectionable content online. The findings provide first empirical evidence on the possible consequences of social media decentralization for politically relevant outcomes.