Multiattribute Utility Theory, Intertemporal Utility and Correlation Aversion

Convenient assumptions about qualitative properties of the intertemporal utility function have generated counter-intuitive implications for the relationship between atemporal risk aversion and the intertemporal elasticity of substitution. If the intertemporal utility function is additively separable then the latter two concepts are the inverse of each other. We review a simple theoretical specification with a long lineage in the literature on multi-attribute utility, and demonstrate the critical role of a concept known as intertemporal correlation aversion. This concept is the intertemporal analogue of a more general concept applied to two attributes of utility, but where the attributes just happen to be the time-dating of the good. In the context of intertemporal utility functions, the concept provides an intuitive explanation of possible differences between (the inverse of) atemporal risk aversion and the intertemporal elasticity of substitution. We use this theoretical structure to guide the design of a series of experiments that allow us to identify and estimate intertemporal correlation aversion. Our results show that subjects are correlation averse over lotteries with intertemporal income profiles, and that the convenient additive specification of the intertemporal utility function is not an appropriate representation of preferences over time. † Department of Finance, Copenhagen Business School, Copenhagen, Denmark (Andersen); Department of Risk Management & Insurance and Center for the Economic Analysis of Risk, Robinson College of Business, Georgia State University, USA (Harrison); Department of Economics, Copenhagen Business School, Copenhagen, Denmark, and Durham Business School, Durham University, UK (Lau); and Center for the Economic Analysis of Risk, Robinson College of Business, Georgia State University, USA (Rutström). E-mail contacts: [email protected], [email protected], [email protected] and [email protected]. Harrison and Rutström are also affiliated with the School of Economics, University of Cape Town, and Lau is also affiliated with IZA – Institute for the Study of Labor. We thank the U.S. National Science Foundation for research support under grants NSF/HSD 0527675 and NSF/SES 0616746, the Danish Social Science Research Council for research support under project 275-08-0289, and the Carlsberg Foundation under grant 2008_01_0410. We also thank two referees, the Editor, Antoine Bommier, Anke Leroux, Harris Schlesinger, and seminar participants for many useful comments. Additional appendices are available in CEAR Working Paper 2011-03 available at http://cear.gus.edu/papers.