Many seminal experiments in behavioral economics have been performed using small groups of undergraduates at elite universities as subjects. Drawing robust scientific conclusions from these experiments is difficult. Student test subjects are, in general, whiter, richer, younger, and more American than the world populace taken as a whole. In addition, the campus laboratories that conduct such experiments are expensive to run, limiting the number of students that can be tested. Given the new possibilities opened up by the advent of the Internet, there should be easier ways to gather behavioral data using large numbers of participants from all over the world. This grant supports a project by Harvard's Erez Liberman Aiden to develop a user-friendly platform for creating, performing, and tracking large-scale economic experiments online. Called the "Laboratory for Online Research in Economics" (LORE), the platform will invite online visitors to participate in economic experiments, beginning in with classic "matrix games" such as the Prisoners' Dilemma or the Public Goods game and eventually expanding to include auction and market simulations with complex matching protocols and population structures. Harvard has funded the creation of a preliminary version of the LORE platform. Funds from this grant would pay for hardware, the hiring of a programmer, and provision of player incentives.

Funds from this grant support the work of Stanford economists Jonathan Levin and Liran Einay, who have obtained unprecedented access to a massive dataset on consumer behavior data collected by the internet retailing giant and auction site eBay. The eBay data is a goldmine of information containing records of hundreds of millions of transactions over ten years, including the histories of every seller, details on every item ever listed on the site, and records of every click made by site users. Grant monies will support Levin and Einay's work analyzing this data, which will initially focus on three distinct issues: how buyer and seller behavior have changed over time particularly with regard to auctions; how to model seller learning; and the impact of changes in online sales taxes on buyer and seller behavior. The depth and richness of the dataset they will be analyzing promises to shed new light on our understanding of what happens when people go shopping.