MAY 10 MAY 10
3:00 PM - 4:30 PM

Speakers | Amal Harrati, PhD and Robert Willis, PhD: 1) The Role of Cognitive Decline on Early Retirement and Disability 2) Does Job Complexity Offer Protection against the Mental Retirement Effect?

PHS Seminar Series


Li Ka Shing Center, Berg Hall B
291 Campus Drive
Palo Alto, CA 94305


The Role of Cognitive Decline on Early Retirement and Disability

Amal Harrati, PhD

While there are many reasons to believe that age-related cognitive decline in mid-life may be a driver of retirement decisions, little attention has been paid to the effects of cognitive capacities on labor market participation. A recent paper by Rohwedder and Willis (2010) showed a causal relationship between early retirement and subsequent cognitive decline; whether there exists a causal relationship in the other direction remains an open question. This paper examines the role of age-related cognitive decline in mid-life on the decisions to withdraw early from the labor force. First, the author characterizes the trajectories of cognitive decline by individual-level and group-level characteristics including occupation and socio-demographics. Second, the author employs an instrumental variable approach using a genetic risk score for cognitive decline to identify the causal impact of mid-life cognitive decline on early retirement decisions.

Does Job Complexity Offer Protection Against the Mental Retirement Effect?

Robert Willis, PhD with Dawn C. Carr, Melissa Castora-Binkley, Ben Lennox Kail & Laura Carstensen

In prior work, Rohwedder & Willis (2010), using cross-national data and instrumental variable methods, found that retirement has a negative causal effect on cognitive performance, measured by a test of immediate and delayed word recall. They hypothesized that this “mental retirement effect” (MRE) arises because working life tends to be more cognitively stimulating than retired life. This paper investigates how the MRE varies by the cognitive complexity of the skills, tasks and context of the longest occupation of respondents in US Health and Retirement study (one of the surveys used by Rohwedder-Willis) using a measure of complexity constructed from the O*NET data base of occupational requirements from the Bureau of Labor Statistics.  We employ a potential outcomes framework to estimate the effect on the trajectory of cognition of several forms of retirement relative to remaining employed, controlling for selection into occupation and retirement. We find a large, highly significant negative MRE effect for individuals in low complexity jobs, but only a small and insignificant effect for those in medium and high complexity jobs.  This is consistent with recent theories and evidence in cognitive psychology and neuroscience (Park & Reuter-Lorenz, 2009) which suggest that long term engagement in complex activities results in the creation of durable neural circuity that helps maintain cognitive function.