Despite its abandonment in theoretical work, a literature search shows that variation in consumer surplus (VCS) is the overwhelming choice in applied work -- not compensating variation (CV) or equivalent variation (EV). How can this be explained? Besides the obvious ease of computation, there are three good reasons for the persistence of VCS. (1) The Willig bounds on VCS usually give close upper and lower bounds on CV and EV, respectively, and are thus <italic>conservative</italic> in the estimation of EV. (2) Without integrability, all three measures are inaccurate. Common quasi-linear utility assumptions for VCS, however, imply integrability. (3) Even with integrability, the expected values of highly nonlinear CV and EV measures cannot be determined by substituting prices or quantities into the estimated equations; simulations are required. Thus, VCS is not only simpler, it may also be more accurate.
