In 2015, Nicole Maestas, Kathleen J. Mullen, David Powell, Jeffrey Wenger, and Till von Wachtner fielded two surveys. One asked respondents about their working conditions; the second asked about their preferences for certain job characteristics. The nine working conditions were control over hours, paid time off, ability to telecommute, work pace, independence, physical demands, working by oneself or with a team, training opportunities, and social impact.

For their 2019 paper, “Understanding Job Transition and Retirement Expectations Using State Preferences for Job Characteristics,” Maestas et al. matched the 2015 results to information on respondents’ actual job changes collected three years after the initial surveys. They were curious about whether job preferences would manifest as transitions from jobs lacking preferred characteristics to new ones with those characteristics, or from working to retirement.

Small subsample sizes meant that the researchers weren’t able to determine which job characteristics affected transitions or retirement outcomes. They did see the following patterns overall, however:

• A strong tendency toward inertia: 60 percent of workers have the same job attributes three years after the 2015 survey. This may indicate that older workers have already selected jobs with their preferred characteristics.
• Generally, if respondents changed jobs, they switched to one with more of the attributes they valued.
• If the 2015 job lacked a desirable attribute, respondents were more likely to transition to unemployment/retirement by 2018, suggesting that people leave work if their job doesn’t align with their preferences.

One way to help close the OASDI trust fund gap is to encourage people to work longer. This research suggests that nonmonetary job characteristics might provide potential handles for those interested in incentivizing later retirement.