Do experimental auction estimates pass the scope test?

Abstract

Scope insensitivity is a popular anomaly in many valuation studies. Although scope insensitivity is a problem that may be present in any valuation method, most previous literature has focused on evaluating scope sensitivity within the context of contingent valuation applications. Nevertheless, it is necessary to understand the demand-revealing properties of experimental auctions since they are increasingly used to value products, such as quasi-public goods. In this paper, we test explicitly whether estimates coming from experimental auctions may pass a scope test. We conduct experimental auctions on products with a subset of attributes (part) and a comprehensive set of attributes (whole) related to animal welfare using two multi-product auction approaches: sequential and simultaneous. Results show that estimates pass the scope test when multi-product auctions are conducted simultaneously but not when they are conducted sequentially for all valued products. Implications of these findings are discussed.