Bateman, Ian J. Cooper, Philip Georgiou, Stavros Poe, Gregory L
Year of Publication:
CSERGE Working Paper EDM 01-01
In this paper we argue that the burgeoning empirical debate over scope sensitivity within contingent valuation studies is fundamentally incomplete in that the effect of study design upon observed scope has largely been ignored. In particular we highlight the frequently overlooked fact that in many common study designs the choice set initially offered (or 'visible') to respondents is changed in a stepwise disclosure manner as they progress through a valuation exercise. Conversely, other designs give advance disclosure regarding the full extent of the final visible choice set prior to any choices or values being elicited. Although this issue has been raised within this journal by several commentators (e.g. Kahneman and Knetsch, 1992; Smith, 1992) it has not been formally tested. We present laboratory and field tests of the impact upon contingent values of varying the visible choice set through stepwise and advanced disclosure modes using exclusive choices throughout (i.e., providing a situation in which procedural variance is not expected). These dimensions of design are interacted with changes in the order in which nested goods are presented (bottom-up versus top-down). We find that when a stepwise disclosure procedure is adopted the observed scope sensitivity is substantially and significantly affected by the order in which goods are presented but such procedural invariance is not observed within advanced disclosure designs. Conjectures regarding the origin and implication of such findings are presented.