[Paper Review] Age and Cognitive Skills: Use It or Lose It

Eric A. Hanushek, Lavinia Kinne+2 more|arXiv (Cornell University)|Oct 1, 2024

Retirement, Disability, and EmploymentSocial Sciences3 citations

TL;DR

This study uses longitudinal data from Germany's PIAAC-L survey to examine how cognitive skills change with age, correcting for measurement error. It finds that skills generally increase into the forties and decline only among older adults with low skill usage, while high-usage white-collar and highly educated workers maintain or improve skills into later life, challenging the 'use it or lose it' narrative for most populations.

ABSTRACT

Cross-sectional age-skill profiles suggest that workers' cognitive skills start declining by their thirties if not earlier. If accurate, such age-driven skill losses pose a major threat to the human capital of societies with rapidly aging populations. We estimate actual age-skill profiles from individual changes in skills at different ages. We use the unique German longitudinal component of the Programme of the International Assessment of Adult Competencies (PIAAC-L) that retested a large representative sample of adults after 3.5 years. Two main results emerge. First, correcting for measurement error, average skills increase into the forties before decreasing slightly in literacy and more strongly in numeracy. Second, skills decline at older ages only for those with below-average skill usage. White-collar and higher-educated workers with above-average usage show increasing skills even beyond their forties. Women have larger skill losses at older age, particularly in numeracy.

Motivation & Objective

• To reassess cross-sectional age-skill profiles that suggest early cognitive decline by analyzing actual individual skill changes over time.
• To correct for measurement error in skill assessments to produce more accurate age-skill trajectories.
• To investigate whether skill usage moderates age-related cognitive changes, particularly in different occupational and educational groups.
• To examine gender differences in cognitive skill decline, especially in numeracy, across older age groups.

Proposed method

• Utilized the German longitudinal component of the Programme for the International Assessment of Adult Competencies (PIAAC-L), which retested a representative sample of adults after 3.5 years.
• Applied longitudinal fixed-effects models to estimate individual skill changes over time, controlling for time-invariant individual characteristics.
• Corrected for measurement error in cognitive skill assessments using statistical techniques to improve reliability of age-skill profiles.
• Stratified analysis by skill usage levels, occupation (white-collar vs. non-white-collar), education, and gender to assess differential aging patterns.
• Used a multi-wave panel design to isolate within-individual changes, reducing bias from cohort and selection effects.
• Employed instrumental variable approaches and reliability corrections to enhance precision in estimating true skill change over time.

Experimental results

Research questions

• RQ1 Do cognitive skills decline in the thirties as suggested by cross-sectional data, or is this an artifact of measurement error?
• RQ2 How does the frequency of skill usage moderate age-related changes in literacy and numeracy skills?
• RQ3 Do white-collar and highly educated workers experience sustained cognitive skill development beyond their forties, even in older age?
• RQ4 Why do women show larger skill losses in numeracy at older ages compared to men, and is this related to usage patterns?
• RQ5 To what extent do measurement errors in skill assessments distort the observed relationship between age and cognitive performance?

Key findings

• After correcting for measurement error, average cognitive skills in literacy increase into the forties and only show a slight decline thereafter.
• In numeracy, skills decline more strongly after the forties, but only among individuals with below-average skill usage.
• Workers in white-collar jobs and those with higher education who report above-average skill usage show increasing cognitive skills even beyond age 50.
• Women experience significantly larger declines in numeracy skills at older ages, particularly when skill usage is low.
• The overall pattern contradicts the 'use it or lose it' hypothesis for high-usage, well-educated, and white-collar workers, who maintain or improve skills with age.
• The study demonstrates that measurement error in cross-sectional data can lead to misleading conclusions about early cognitive decline.