Ill health and retirement in Britain: A panel data-based analysis

Abstract

We examine the role of ill-health in retirement decisions in Britain, using the first eight waves of the British Household Panel Survey . As self-reported health status is likely to be endogenous to the retirement decision, we instrument self-reported health by a constructed 'health stock' measure using a set of health indicator variables and personal characteristics, as suggested by . Using both linear and non-linear fixed effects estimators, we show that adverse individual health shocks are an important predictor of individual retirement behaviour. We compare the impact of our constructed health measure on economic activity with that arising from the use of other health variables in the data set. We also examine the impact of the 1995 reform of disability benefits on the retirement decision. Key WordsIll health Retirement Disability insurance JEL classification H55 I12 J26 AcknowledgementsWe are grateful to the Nuffield Foundation for funding the project: 'A study of health and the labour market behaviour of older workers', and to Sharon Witherspoon in particular. Executive SummaryThis paper examines the role of ill-health in retirement decisions in Britain, using the first eight waves of the British Household Panel Survey . To tackle the problem that self-reported health status is likely to be endogenous to the retirement decision, a two-stage method, as suggested by , is adopted. The first stage is to construct a 'health stock' measure that is cleansed of the effects of reporting behaviour reflecting labour market participation. This measure is then introduced into a reduced form model of labour market (in)activity. At this latter stage, modelling how labour market transitions are related to time variation in health and other characteristics, helps to eliminate any unchanging person-specific association between characteristics and the decision to work.The 'health-stock' measure is constructed by regressing self-reported health on a set of more objective health indicator variables and a set of other personal characteristics.Few personal characteristics, other than those related to health, are found to be significant in explaining self-assessed health. Many of the health indicators are significant in the model, and as expected having health problems is associated with reporting poor health.The panel structure of the data set is exploited to estimate a non-linear 'fixedeffects' model that allows an exploration of how time variation in various characteristics relates to transitions out of (and in to) work. Deterioration in an individual's health is found to be strongly positively associated with movement out of work. Sensitivity analyses suggest that there may be some asymmetry between the respective effects of health deteriorations and improvements on transitions out of, and in to, work.The final section briefly considers a reform to the public disability insurance programme in 1995, which both tightened formal eligibility conditions and reduced the economic incentives to retire via the disability insu...

“I want to break free”. The role of working conditions on retirement expectations and decisions

“…In the literature, there is a concern about the use of self-reported measures, in particular for health, to explain retirement decisions (Bound 1991, Dwyer and Mitchell 1999, Disney et al 2006. The general concern is that the responses to subjective judgements may not be independent of the labour market outcome they are used to explain (Bound 1991) and this can lead to an over estimation of their effect on the outcome of interest.…”

I feel good! Gender differences and reporting heterogeneity in self-assessed health

“…Moreover, these errors seem to be smaller for those in the labour force while they are larger if the assessment of health leaves room for subjective interpretation. Disney et al [7] focus on retirement behaviour in Britain and the influence of ill health. They use self-assessed health as the predictor of health status and instrument this variable using an or- A review about the relationship between labour force status of older workers and health is presented by Lindeboom [8].…”

Long-Term and Spillover Effects of Health Shocks on Employment and Income

“…The effectiveness of these measures depends on the extent to which ill-health constrains the earnings capacity and employment opportunities of the 50+ population. Ill-health is frequently reported to be a leading cause of labor market withdrawal in middle-age (Bound et al 1999; Currie and Madrian 1999; Dwyer and Mitchell 1999; French 2005; Disney, Emmerson, and Wakefield 2006). However, the existing evidence base to gauge whether improved population health, or at least reduced disability, alongside improved financial incentives, could increase labor market participation is weak or incomplete in three important respects.…”

“…Studies of the 50+ population that rely on self-reported indicators of health generally find a strong negative impact of ill-health, or disability, on earnings and income that operates through employment and work hours rather than wages (McClellan 1998; Bound et al 1999; Smith 1999; Riphahn 1999; Charles 2003; Wu 2003; Disney, Emmerson, and Wakefield 2006; Lindeboom and Kerkhofs 2009). Our contribution to this evidence base is both through the use of a robust estimator identified from acute hospital admissions providing variation in health that is more likely to be exogenous than self-reported measures and by tracing the employment and income effects over an extended follow-up period using tax records.…”