Measuring healthy working life expectancy: implications for estimates and socioeconomic inequalities

Populations in high-income countries are aging at an unprecedented pace, raising concerns about the sustainability of pension and social security systems. Healthy working life expectancy (HWLE) is a key indicator linking population health to economic and social sustainability, but its use is complicated by heterogeneous health and work measures affecting estimates and comparability. This study provides the first comprehensive assessment of how measurements shape estimates of healthy working life expectancy, including differences by gender and education and variation in HWLE across time. Based on the U.S. Health and Retirement Study (1998–2006, 2008–2022), we estimated multistate life table models and compared healthy working life expectancy at age 50 across health measures – ADL limitations, IADL limitations, self-rated health, functional limitations, working limitations, and chronic conditions – and work measures – employment status, labor force participation, and labor income. Healthy working life expectancy is highly sensitive to measurement choice: it is highest when measured by ADL/IADL limitations (13 years for men, 11 years for women, when based on labor force participation), followed by when measured by self-rated health and working limitations (11 and 10) and functional limitations (6.5 and 4.4), and it is lowest when measured by chronic conditions (2). Estimates are ~1 year lower when based on labor income and are also ~1 year lower when based on employment status. Men’s values are 0.5–2 years higher than women’s values, educational inequalities are as high as nine years, and temporal trends are heterogeneous. This study strengthens the conceptual and methodological basis of healthy working life expectancy research by guiding the selection and interpretation of measures. It highlights the need for harmonized approaches to enable meaningful comparisons, inform policies to extend working lives, and identify inequalities.