Introduction

How can we effectively aggregate disparate pieces of information that are spread among many different individuals? In other words, how can we best access the ‘wisdom of the crowd’? Prediction markets, which are speculative markets created for the purpose of aggregating information and making predictions, are a means of addressing this issue. Their theoretical underpinning derives from the efficient-­markets­hypothesis­and­stems­from­the­view­that­relevant­ information­ concerning the likelihood of future events is dispersed among the opinions and intuitions of many people. While the mechanisms underlying prediction markets vary, they all offer a means of aggregating this information. Many of these markets are open to the public, while others are open to particular groups. Prediction markets can be used to provide forecasts of the probability and the mean and median outcomes of future events, as well as correlations between these events. They also have many potentially valuable applications for public policy (Paton et al., 2009, 2010). The markets have already been used to forecast uncertain outcomes ranging from influenza­outbreaks­ (Wang­et al., 2009) to the spread of infectious diseases (Polgreen et al., 2007), the demand for hospital services (Rajakovich and Vladimirov, 2009)­the­box­office­success­of­movies,­climate­change,­vote­shares,­election­outcomes­(Rhode­and­Strumpf,­2004;­Wolfers­and­Zitzewitz,­2004;­Snowberg­et al., 2005) and the probability of meeting project deadlines at Google (Leigh and Wolfers, 2007). Prediction markets may also be used as a mechanism to help market participants hedge their exposure to risk (Athanasoulis et al., 1999). Even so, some researchers have questioned how far prediction markets are able­to­outperform­other­means­of­forecasting­(Erikson­and­Wlezien,­2008).­It­is­ also suggested that they may be open to manipulation (Wolfers and Leigh, 2002), though this might actually aid prediction market accuracy (Hanson and Oprea,­2009).­Again,­they­may­not­provide­efficient­forecasts­of­low­probability­ events (Smith et al.,­ 2006;­Wolfers­ and­Zitzewitz,­2004),­ and­may­be­open­ to­ systematic biases, such as optimism bias (Cowgill et al.,­2009)­and­the­favourite-longshot bias (Vaughan Williams and Paton, 1997). The effective use of prediction markets has the potential, however, not only to help forecast events at a national and international level, but also to assist companies­in,­for­example,­providing­improved­estimates­of­the­potential­market­size­

for a new product idea or the launch date of new products and services. Examples of companies that have used internal prediction markets for a range of business forecasts­include­Hewlett-­Packard­(Chen­and­Plott,­2002),­Google­(Cowgill­et al., 2009) and General Electric (Spears et al., 2009). The success and potential of these markets in predicting public events and corporate outcomes has therefore generated substantial interest among social scientists, policymakers and the business community. The insights gained also have many potentially valuable applications for policy more generally, not least when accurate forecasts are required in relation to­quantifiable­targets.­Moreover,­the­information­provided­by­prediction­markets­ will have value in the advance warning managers may be given of weak performance­in­identifiable­areas.­This­can­help­improve­resource­allocation. Important research questions include the impact of prediction market design choices on performance (Spann and Skiera, 2003) and the impact of the nature of­rewards­on­ the­ level­of­accuracy­of­prediction­markets­(Servan-­Schreiber­et al., 2004). Indeed, the design of the incentive programme may be critical to optimising performance, insofar as people may invest more thought and energy into expressing their opinion when there is a meaningful incentive to do so. Overall, the balance of opinion provided by previous research suggests that well-­designed­ prediction­ markets­ can­ offer­ substantial­ promise­ as­ a­ tool­ of­ information aggregation and forecasting, whether alone or as a supplement to other mechanisms like surveys, group deliberations and expert opinion. Moreover, they can be applied at a macroeconomic and microeconomic level to yield information that is valuable for government and commercial policymakers, and that can be used for a number of social purposes. ­ This­volume­of­original­readings­marks­a­significant­addition­ to­ the­base­of­ knowledge about this fascinating subject area. What is provided is a collection of readings that draws on the expertise of many of the leading contributors in the field.­The­chapters­are­not­only­novel­and­original,­but­also­set­the­subject­within­ the existing framework of literature. As such, this book should serve as a valuable asset for those who are coming fresh to the subject, as well as for those who are more familiar with the subject matter. The contributors hail from a host of prestigious­institutions­located­as­far­afield­as­Australia,­Denmark,­Germany,­the­ Netherlands, Israel, Taiwan, the United Kingdom and the United States. In many cases, the contributions would, in my opinion, have gone on to be published­in­top-­ranked­journals,­but­the­authors­lent­their­support­instead­to­the­idea­ of­a­single­volume­that­would­help­promote­this­field­of­research­to­a­wider­audience. In all cases, the authors have provided contributions that are valuable and important,­and­which­contribute­something­significant­to­help­meet­the­burgeoning­ growth of interest in the theory and applications of prediction markets. It has been a pleasure to edit this book, and my deepest gratitude goes to all involved.