Definition
Disability-Free Life Expectancy (DFLE) is the average number of years a person is expected to live free of disability (or free of a specified health limitation, depending on the measure used). It is computed by applying the Sullivan method: age-specific disability prevalence rates from cross-sectional surveys are combined with age-specific mortality rates from a period life table; disability-free person-years in each age interval equal (1−prevalence)×person-years lived, and DFLE is the sum of disability-free person-years from a given age. DFLE decomposes total life expectancy into a disability-free component and a years-with-disability component, enabling direct assessment of whether mortality gains are accompanied by health gains or by longer survival with chronic impairment.
Key Ideas
- Sullivan method: The standard computational approach; uses period prevalence data (e.g., National Health Interview Survey (NHIS) activity-limitation questions) and period mortality tables. The result is a period measure — it does not track an actual cohort but summarizes current conditions. Sensitivity to the disability indicator used is substantial: activity-limitation measures produce different trends from Activities of Daily Living (ADL)-disability, mobility-functioning, or disease-prevalence measures.
- Compression vs. expansion of morbidity: The central empirical question is whether DFLE is growing faster than total life expectancy (LE) (compression — the Fries hypothesis) or slower (expansion — Gruenberg's "failure of success"). Evidence depends critically on the indicator: ADL-disability shows partial compression at age 65+ for elderly U.S. women; disease prevalence and mobility functioning show expansion (Crimmins and Beltrán-Sánchez 2010). See Compression of Morbidity.
- 40-year U.S. trend (Crimmins, Zhang, and Saito 2016): Using NHIS activity-limitation data 1970–2011: (1) at birth, DFLE and total LE grew in parallel — no compression at the population level; (2) at age 65, partial compression for women (DFLE grew faster), dynamic equilibrium for men; (3) the gender disability paradox holds — men have higher DFLE than women despite shorter total lifespans; (4) DFLE gains concentrated at age 65+, not at working ages, providing no justification for raising the retirement age.
- Distinction from HALE: Healthy life expectancy (HALE) uses Global Burden of Disease (GBD) disability weights (a continuous 0–1 scale based on disease burden) applied to all conditions simultaneously. DFLE uses a binary disability indicator from a single survey question or set of ADL questions. HALE is more comprehensive; DFLE is more sensitive to the specific disability indicator and reflects the policy outcome of interest for retirement and long-term care policy. See Healthy Life Expectancy.
- Working-age improvement: Crimmins, Reynolds, and Saito (1999) documented ≈24% improvement in work inability for ages 60–69 over 1982–1993 (NHIS), but no improvement for 50s — relevant to Social Security (SS) retirement age debates. Improvement was mostly compositional (rising education), with real biological contribution from declining cardiovascular disease (CVD) and arthritis prevalence.
Why It Matters
- Compression of morbidity debate: DFLE is the central empirical measure for testing whether population health improved or merely survival lengthened; the answer determines whether rising LE should drive up retirement ages.
- Retirement age policy: If working-age DFLE is not improving, raising the Social Security retirement age imposes disproportionate costs on workers with physical impairments who cannot work into their late 60s.
- Disability Insurance (DI) and disability program design: DFLE trends by education and race inform who reaches retirement with residual work capacity and who has been forced out of the labor force through disability.
Related
Sources