Heterogeneity in the association between retirement and cognitive function: a machine learning analysis across 19...

Area of Science:

• Gerontology
• Cognitive Science
• Public Health

Background:

• Rising state pension ages may impact cognitive aging by delaying retirement.
• The cognitive effects of retirement are not uniform and can differ across individuals and contexts.

Purpose of the Study:

• To investigate the heterogeneous association between retirement and cognitive function.
• To understand how factors like gender, socioeconomic status, and health influence retirement's cognitive impact.

Main Methods:

• Analysis of harmonized data from three longitudinal studies (HRS, ELSA, SHARE) across 19 counties.
• Included 12,811 individuals aged 50-80, assessing retirement status and cognitive function (word recall) over three waves (2014-2019).
• Employed instrumental variable causal forests estimation, using state pension age as an instrument for retirement.

Main Results:

• Retirees recalled 1.348 more words than workers on average.
• Significant heterogeneity in retirement's cognitive impact was observed.
• Greater cognitive benefits were linked to being female, higher socioeconomic status, better pre-retirement health, and engaging in physical activity.

Conclusions:

• The heterogeneous cognitive effects of retirement suggest a need for flexible pension systems.
• Policymakers should consider early retirement options to accommodate individual circumstances and optimize cognitive aging outcomes.