Measuring Individual Risk Attitudes when Preferences are Imprecise

Abstract

The Service Systems research group at WMG works in collaboration with large organisations such as GlaxoSmithKline, Rolls-Royce, BAE Systems, IBM, Ministry of Defence as well as with SMEs researching into value constellations, new business models and value-creating service systems of people, product, service and technology.The group conducts research that is capable of solving real problems in practice (ie. how and what do do), while also understanding theoretical abstractions from research (ie. why) so that the knowledge results in high-level publications necessary for its transfer across sector and industry. This approach ensures that the knowledge we create is relevant, impactful and grounded in research.In particular, we pursue the knowledge of service systems for value co-creation that is replicable, scalable and transferable so that we can address some of the most difficult challenges faced by businesses, markets and society. Research StreamsThe WMG Service Systems research group conducts research that is capable of solving real problems in practice, and also to create theoretical abstractions from or research that is relevant and applicable across sector and industry, so that the impact of our research is substantial.The group currently conducts research under six broad themes: (Australia) as well as at the 42 nd Australian Conference of Economists in Perth (Australia) for many insightful comments. Finally, the authors very much appreciate the efforts of referees and an editor, which have greatly improved the final version of this paper. 4 Measuring Individual Risk Attitudes when Preferences are Imprecise By Graham Loomes and Ganna Pogrebna AbstractThere is widespread interest in measuring risk attitudes and incorporating such measures into broader econometric analyses. We consider three elicitation procedures currently in use. We find considerable variability within -and even more, between -the results they produce. We suggest that this reflects the way that different instruments interact with imprecise underlying preferences. The short run implication is that such procedures need to be used with caution and are likely to be highly context-specific. The longer run implication is that adding 'white noise' to deterministic models is inadequate: we need to develop models that allow for imprecision and procedural variation.

Do Preference Reversals Disappear When We Allow for Probabilistic Choice?

“…17 One disadvantage is that responses appear to be susceptible to various systematic framing effects connected to the ranges and frequencies selected by the researcher and the order in which rows are presented. Parducci (1965) provides an early discussion of range-frequency effects, Levy-Garboua et al (2012) and Loomes and Pogrebna (2014a) report such effects in tables intended to elicit risk attitudes, and Sanchez Martinez et al (2015) find evidence of such effects in the context of health state evaluation. Of course, if decisions are the result of some process of sampling and accumulation, it would not be surprising to find that "cues" provided by MPLs are liable to have various systematic effects.…”

Elicitation of preferences under ambiguity

“…In evaluating and comparing them there is a fundamental problem: the experimenter does not know the 'true' attitude to risk of the subjects, nor their 'true' preference functional. All we can conclude from Loomes and Pogrebna (2014) and Zhou and Hey (unpublished) is that context matters. Further work needs to be done to discover how and why.…”

“…These include Holt-Laury Price Lists (Holt and Laury 2002), Pairwise Choice questions (Hey and Orme 1994), the Becker-DeGroot-Marschak (BDM) Mechanism (Becker et al 1964), the Bomb Risk Elicitation Task (Crosetto and Filippin 2013), and the Allocation Method, pioneered originally by Loomes (1991), revived by Andreoni and Miller (2002) in a social choice context, and later by Choi et al (2007) in a risky choice context. Some of these are contrasted and compared in Loomes and Pogrebna (2014) and in Zhou and Hey (unpublished). We describe them briefly here.…”

Estimating Risky Behavior with Multiple-Item Risk Measures: An Empirical Examination

“…However, an unsettling result derived from the literature indicates that the degree of risk aversion varies for the same person depending on the elicitation technique used (e.g. Isaac and James, 2000;Berg et al, 2005;Anderson and Mellor, 2008;Loomes and Pogrebna, 2014). Evidence suggests that the existence of noise in survey items or experimental measures could distort findings (Harless and Camerer, 1994) and affect the ability of predicting economic outcomes (Cutler and Glaeser, 2005;Beauchamp et al, 2012).…”

“…Deck et al, 2009;Hey et al, 2009;Loennqvist et al, 2015;He et al, 2016;Loomes and Pogrebna, 2014). They find degrees of variability and disparity that are difficult to explain within the terms of any expected utility model.…”