An Empirical Investigation of Quasi-hyperbolic Discounting

There is a growing consensus based on extensive experimental evidence that people are generally influenced by myopia when making decisions of an intertemporal nature. The view that people are myopic has now become sufficiently main-stream that it now features in contemporary debate regarding public policy reform (e.g. Pensions Commission, 2005, pp. 68-69, and DWP, 2006a, p. 42). This is of practical importance because the assumption that preferences are time inconsistent suggests a potential role for paternalistic policy intervention, which is absent from the classical rational agent model of behaviour. Despite its growing influence, however, very few studies have attempted to investigate the empirical evidence for myopia beyond controlled laboratory experiments, which leaves open the question of how important myopia is for decisions taken in the field. 1 Here I address this issue by considering how far behavioural myopia is supported by econometric estimates for a structural model of household savings and labour supply calculated on nationally representative survey data for the United Kingdom. I find no firm evidence in support of the proposition that myopia is of practical importance. Economic analyses of intertemporal decision making are usually conducted within the life-cycle framework (LCF). This analytical approach focuses attention on the nature of intertemporal preferences over utility, which have typically been subsumed into a single discount factor, δ, following the seminal study by Samuelson (1937). 3 Unfortunately, it is difficult to reconcile this understanding of time preference with most of the associated empirical estimates that have been reported in the literature. This is because most of the available estimates for discount factors Ñ when viewed through the lens of the LCF Ñ are based on highly unrealistic assumptions (often implicit) regarding preferences, circumstances, and/or beliefs.