One man's rags are another man's riches: Identifying adaptive preferences using panel data

Tania Burchardt

Published December 2004

One of the motivations frequently cited by Sen and Nussbaum for moving away from a utility metric towards a capabilities framework is a concern about adaptive preferences or conditioned expectations. If utility is related to the satisfaction of aspirations or expectations, and if these are affected by the individual’s previous experience of deprivation or wealth, then utility cannot provide a basis for assessing well-being, equality or social justice which is independent of the initial distribution. This paper contributes to the identification of adaptive expectations by using ten years of panel data from the British Household Panel Survey to study the process of adaptation based on the individual’s own previous experience. Subjective assessments of financial well-being at time t, for individuals with a given income level, are compared according to the income trajectory of the individual over the previous one to nine years. Descriptive statistics are followed by multivariate analysis, introducing controls for changes in need (family size and composition, disability), and possible social reference groups (for example, ethnicity and employment status). Fixed effects regressions allow for individual variation in the scaling of satisfaction. The results show that year on year, individuals who have experienced a fall in income since the previous year are less satisfied than those who have a steady income, suggesting that subjective assessments may be made in comparison with previous experience. Surprisingly, individuals who have experienced an increase in income are also less satisfied. This suggests that income is a poor proxy for satisfaction but it does not provide firm evidence for the existence of adaptation over the short term. Over a longer period, those who have experienced falling incomes are less satisfied than those who have had constant income, while those who have experienced rising incomes are no more satisfied than those who have had constant incomes. This suggests that over a longer period, adaptation to changes in income is asymmetric: people adapt to rising incomes but less so falling incomes. The paper concludes that satisfaction with income is influenced by objective circumstances, and to changes in objective circumstances, in complex ways. In particular, the process of adaptation to rises in income masks long-term differences in outcomes for individuals and makes subjective assessments of well-being a flawed basis for judgements of inequality or social justice. An objective normative standard, such as is offered by the capabilities framework, avoids social evaluations being unduly influenced by individuals’ past experiences.