EFFECTS OF RETIREMENT ON COGNITIVE DECLINE: Evidence that reluctant retirees suffer bigger losses

The large and negative effect of retirement on cognitive decline in older ages depends on how much individuals want to be retired and less on the time in retirement, according to research by Hendrik Schmitz (Paderborn University) and Matthias Westphal (TU Dortmund University). These effects arguably arise because individuals who prefer to be retired have a plan for the time thereafter and can thus maintain a certain skill-preserving level of cognitive stimulation. 


Because of the drastic ageing of Western societies, age-related diseases such as dementia are ever increasing. The roughly 50 million people with dementia today will have tripled by 2050 – and the economic costs grow even faster. This reinforces the need to understand the determinants of age-related cognitive decline, previously thought of as irreversible. 

Schmitz and Westphal study the impact of retirement – one of the major decisions along the life course – on age-related cognitive decline as measured by memory skills. Is retirement the rather start of a stimulating ‘good life’ or will it initiate new troubles such as the onset of an accelerated cognitive decay? 

Using data on individuals aged 55 and above from the United States and 15 European countries (including the UK), the authors find that retirement indeed induces a persistent decline in cognitive functioning of about 20& of the pre-retirement skill level. This effect is substantial given that the study also controls for the general effects of ageing that are unrelated to retirement. 

For precise policy implications, however, we need to understand the driving forces of the retirement-induced cognitive decline. Because the timing of retirement is largely a voluntary decision within institutional boundaries, it is important to understand whether the effects are different with respect to how eager the individuals want to be retired. 

After a long working career, does someone suffer a greater loss in skills if she cannot wait to be retired or is someone worse off who stalls the retirement decision? Relatedly, it is important to ask whether the main effect of 20% increases over time or stays constant after the working career. 

To come up with an answer, this study exploits a common property of retirement schemes across all the countries in the analysis – that the monetary incentives to retire increase with every year that one continues working. This makes it possible to infer the revealed preference to retire by evaluating the incentive at which they chose to retire. 

The authors then can estimate the effects along this revealed preference and find that the more reluctant one is to retire, the higher the cognitive loss she suffers after retirement. While there are only mild effects for those who retire as early as possible, there are substantial effects of more than 30% for relatively more reluctant retirees.

This effect cannot be explained by more accumulated wealth nor differences in the occupation nor other observables before individuals enter retirement. Rather, individuals who are looking forward to their retirement know what to do with their unprecedented leisure time and can prepare accordingly. 

Individuals, in contrast, who stall the retirement decision to the very last, need to make bigger adjustments after retirement to prevent a drastic loss in their cognitive stimulation. The duration of retirement causes a much weaker effect heterogeneity – ten years in retirement not even doubles the initial effect. 

Taken at face value, one policy implication is straight at hand: granting individuals even a higher freedom of choice of when to retire (and how much to work after retirement) is likely to induce a positive net effect on the skill decline of older people. Individuals with high preferences may even exhibit positive effects on their skills, while individuals with low preferences to retire may postpone their skill decay toward later ages and stay active longer.




Contact details: 

Matthias Westphal 

Assistant Professor 

TU Dortmund University 


+49 179 4764905