A Theory of Retirement

Abstract

We construct a life-cycle model in which retirement occurs at the end of life as a result of declining health. We show that improvements in life expectancy, coupled with a delay in the onset of disability, increases both the optimal consumption level and the proportion of life spent in leisure. The retirement age increases proportionally less than the increase in life expectancy.

Optimal choice of health and retirement in a life-cycle model

“…A positive link between longevity and the number of life-years spent in good health is empirically well established by now (for a discussion and references see Bloom et al 2007). According to the 'compression of morbidity' hypothesis the period spent in ill health towards the end of life is compressed (either in the absolute number of life-years or at least in proportion to total life-time).…”

“…These simulations show for instance, that workers who earn a higher base-line wage tend to (re-)invest earlier in their health and tend to retire later. Thus, in their model, and in contrast to Bloom et al (2007), the substitution effect in the consumption-leisure trade-off is dominant. In contrast, higher levels of initial health induce earlier retirement by a pure income effect, whereas, surprisingly perhaps, a greater deterioration of health induces later retirement (a negative income effect).…”

“…However, this is not reflected in theoretical work. While there are first efforts at gaining theoretical insights (Bloom et al 2007, Galama et al 2008) into the relationship between health and retirement, to date this is still underexplored, and even less is known so far on the relationship between pension funding (rather than retirement itself) and health.…”

“…The health improvements over the life course underlying the increase in longevity should also imply that individuals enjoy greater productivity (or lower disability from labour supply). Bloom et al (2007) employ this relationship within a life-cycle model with endogenous retirement to examine how increasing longevity bears on the optimal timing of retirement. They show that the income effect related to higher productivity dominates the substitution effect so that individuals respond to increases in longevity by demanding both more leisure and more consumption.…”

“…Hence, while individuals tend to increase the length of their working lives (in order to accumulate greater retirement wealth) they do so less than proportionately, implying that a greater share of the lifetime gained is spent in retirement. Bloom et al (2007) consider a set-up without public pension scheme. Individuals are able to spread their wealth over their lifecourse by transacting on a perfect annuities market.…”