Abstract
We rely more and more on reputation systems when making decisions: the score of an eBay seller, the rating of a hotel on TripAdvisor, or the reviews of a book on Amazon. As Craig Newman, the founder of Craigslist, states in the foreword of the anthology, trust and reputation will become more important than money and nominal power (p. ix). After this strong claim, reputation is examined in a wide variety of contexts and from different perspectives in the next six sections. The collection goes beyond the well-known examples such as eBay, Amazon and Yelp, and addresses also academic impact factors or less known reputation systems such as the Whuffie. Part I, Understanding Reputation, comprises of three chapters that define the field. Chrysanthos Dellarocas, for example, gives a short and clear introduction in central questions of reputation system design, pointing also to the negative sides. Part II, Regulation Societies, focuses mainly on the regulation of reputation systems, taking examples from biology (John Whitfield), the secondary and tertiary invisible hands at the marketplace (Eric Goldman), or the landlord–tenant relation (Lior Jacob Stahilevitz). Together, the three chapters show not only how reputation can solve problems, but also point to mistakes in designing or regulating reputation systems. Part III is titled Amplifying Signals, but it focuses mainly on the role of reputation systems in stimulating good behaviour – in online communities – but also philanthropy. Part IV, Supporting Science, discusses the benefits of usage-based metrics above simple citation scores in science. In part V, Improving Policy, the attention goes to societal problems such as citizen engagement. Part VI, The Reputation Society, closes the volume with three utopian, more pessimistic chapters. Whereas Michael Zimmer and Anthony Hoffman address the issue of privacy and the dangers of oversharing in social media as a reputational challenge, Jamais Cascio, and also Madeline Ashby and Cory Doctorow provide intriguing, but somewhat scary, scenarios about the future reputation society.
As can be seen in this overview, the book touches reputation in many domains. It is also written by an interesting mix of authors. Many are academics, but there are also founders or managers of major players in this field (e.g. Mendeley, GlobalGiving.org). This leads to an interesting mix of chapters. The two chapters on philanthropy, for example, complement each other nicely. Whereas Alex Steffen gives a passionate plea for attention philanthropy (p. 89) to help innovative philanthropical projects to find donors, some of his compelling sounding ideas remain somewhat vague. How does providing a ‘series of tasks both fun to do and useful to the attention philanthropy system’ (p. 93) look like exactly? Marc Maxson and Mari Kuraishi provide in their chapter detailed information about what GlobalGiving.org did to increase to increase donations, such as ‘visitor postcards’, informal reports from travellers and volunteers that might be more trustworthy than the formal reports of the organization. The same holds for the chapters on science. Michael Nielsen argues against centralized metrics such as the H-index by focusing on the downsides of such metrics, such as providing strange incentive systems. However, he links the topic only weakly to online reputation. Victor Henning, Jason Hoyt and Jan Reichelt make concrete suggestions for new usage-based reputation metrics that they are also building into Mendeley, their research and reference management system. This chapter provides a thorough discussion of the advantages and limitations of usage-based metrics.
Paolo Massa points out the two extremes that can be created by different trust metrics: ‘tyranny of the majority or echo chambers’ (p. 151). He argues that a shift from one global reputation to several local reputations is needed to make a reputation attack-resistant. To stress his point, an analysis of Epinions’ web of trust is reported that clearly shows that roughly 20% are controversial users, that is, users who are trusted by some other users and distrusted by others (p. 156). Although the chapter does not provide an answer to the question of whether reputation systems will be able to mediate between these two extremes, it raises interesting questions for the design of reputation systems.
As said, the chapters cover various perspectives – economics, (evolutionary) biology, information systems and even law. My own background is in social psychology, so I enjoyed the input from these divergent perspectives. However, I missed the social sciences a little bit. The chapters by Cliff Lampe and Marc Maxson include some social psychological references, but reputation plays a major role as a strategic solution to social dilemmas, so more focus on sociological or game theoretical approaches would have been nice. On the other hand, the anthology nicely complements more social scientific literature such as the Russell Sage Series on Trust. The interdisciplinary mix as well as the combination of academics and practitioners makes the collection attractive for a large readership.
The book does not only cover many perspectives, but also different forms of reputation. Most chapters are on the reputation of a person or organization, but sometimes the quality and trustworthiness of a specific activity (Marc Maxson and Mari Kuraishi) or an argument (Luca Iandoli, Josh Introne and Mark Klein) is judged. I missed a concluding chapter that summarizes the findings on all these different forms of reputation and from these different perspectives, especially because overarching themes come back repeatedly, such as multiple domain-specific reputations and the corresponding demands for multiple identity systems or the issue of meta-reputation. Nevertheless, I enjoyed reading the book. It does not go so much into depth, but it provides a lot of food for thought. Especially the Orwellian scenarios at the end stimulate thinking about the dangers of the reputation society. After reading those, I hope that Newman’s claim doesn’t become true before the designers of reputation systems read this book and incorporated the advice gained from reading it.