George Mason Talk and Paper on Economics of Reputational Information
By Eric Goldman
Last week I presented a version of my Economics of Reputational Information talk at the Third Annual Conference in the Law and Economics of Innovation series, an event co-sponsored by George Mason Law School and Microsoft. My talk slides. Apologies for the slide formatting, but I honored the organizers’ request to use a standardized template. If you’ve seen my slides from previous presentations of this topic, these slides will look familiar, but I did change them around a bit for this audience.
In conjunction with the talk, I also posted a short draft essay giving an overview of the research project generally. This is my first time I’ve organized my thoughts about this project into an essay, so those of you interested in a high-level overview of the project might find this interesting. Then again, it’s a rough draft, incomplete and not very decisive.
In response to the talk, I got some excellent questions from the audience, including:
Q: How do we know service providers will honor their promises in the future? A: life is uncertain, so there will always be future risk. However, my goal is to help design systems that gather the relevant information available at the decision-making time that allows the best decision about future performance.
Q: Why don’t platforms have more incentives to police against fraud? A: 47 USC 230 provides some of that incentive by insulating websites for their fraud-policing efforts. With this regulatory freedom, we have seen many websites voluntarily adopt increased anti-fraud efforts. For example, consider how much more eBay does to mediate the buyer-seller transaction today than it did 10 years ago (see, e.g., this article). Indeed, we’ve repeatedly seen reputable websites increasingly undertake more voluntary efforts to protect their users over time.
Q: Reputational information is strewn all over the web. Why isn’t there one-stop shopping for reputational information? A: One-stop shopping would be convenient. But competition among reputational systems is also useful.
Q: Why shouldn’t reputational systems turn over information about anonymous posters? A: First, reputable websites already have incentives to manage content from anonymous posters to avoid garnering a reputation as a cesspool of content. Hence, JuicyCampus went out of business because it lacked credible content. Second, with respect to turning over information, websites already do in many cases, but only if they keep IP address information. However, we may not need a law requiring websites to retain IP address info. As proceedings in the AutoAdmit lawsuit showed, plaintiffs sometimes can overcome the information deficiency to unmask anonymous posters through other means.