Capturing Flexible Heterogeneous Utility Curves: A Bayesian Spline Approach

Abstract

Empirical evidence suggests that decision makers often weight successive additional units of a valued attribute or monetary endowment unequally, so that their utility functions are intrinsically nonlinear or irregularly shaped. Although the analyst may impose various functional specifications exogenously, this approach is ad hoc, tedious, and reliant on various metrics to decide which specification is "best." In this paper, we develop a method that yields individual-level, flexibly shaped utility functions for use in choice models. This flexibility at the individual level is accomplished through splines of the truncated power basis type in a general additive regression framework for latent utility. Because the number and location of spline knots are unknown, we use the birth-death process of Denison et al. (1998) and Green's (1995) reversible jump method. We further show how exogenous constraints suggested by theory, such as monotonicity of price response, can be accommodated. Our formulation is particularly suited to estimating reaction to pricing, where individual-level monotonicity is justified theoretically and empirically, but linearity is typically not. The method is illustrated in a conjoint application in which all covariates are splined simultaneously and in three panel data sets, each of which has a single price spline. Empirical results indicate that piecewise linear splines with a modest number of knots fit these data well, substantially better than heterogeneous linear and log-linear a priori specifications. In terms of price response specifically, we find that although aggregate market-level curves can be nearly linear or log-linear, individuals often deviate widely from either. Using splines, hold-out prediction improvement over the standard heterogeneous probit model ranges from 6% to 14% in the scanner applications and exceeds 20% in the conjoint study. Moreover, "optimal" profiles in conjoint and aggregate price response curves in the scanner applications can differ markedly under the standard and the spline-based models.choice models, utility theory, heterogeneity, splines, Bayesian methods, Markov chain Monte Carlo

Extracting multistage screening rules from online dating activity data

“…Utility functions for each activity (browsing and writing) are decomposed into three portions: an intercept, a two-knotted piecewise linear spline for each continuous (or "continuizable") attribute (e.g., age), and dummy variables for intrinsically categorical attributes (e.g., ethnic group). Specifically, in standard notation (9,11), utility (V B ij ) for user i browsing (B superscript; an analogous formulation holds for writing W) potential mate j is…”

Bayesian analysis of color preferences: An application for product and product line design

“…In other words, the surface fitted by natural cubic spline are smooth in the entire feature space. In contrast to our method's focus on a smooth function, the linear spline basis in Kim, Menzefricke, and Feinberg's conjoint analysis of bathroom scales data yielded a nonsmooth utility function over the features . The color space is incredibly large and is composed of three different visual vectors acting simultaneously.…”

Key Features of Electric Vehicle Diffusion and Its Impact on the Korean Power Market

“…A multivariate adaptive regression splines (MARS) technique was used to estimate representative the daily power demand patterns for 24 h. MARS is a non-parametric regression technique that can automatically model data nonlinearities [33][34][35]. The representative daily power demand patterns for the four seasons were obtained through the MARS technique that can be represented as the following functions for each hour: …”