The cleanest direct measure of late-career retraining outcomes comes from an American Institute for Economic Research survey of US adults age 47 and older who had attempted a career change. In that survey, 82% of older workers who wanted to change careers were able to make the move, and among those movers, 90% reported that the change was successful — leaving roughly 10% who did not, the action-side regret proxy. (The 82% is a transition/placement rate, not a satisfaction rate.) The Center for Retirement Research’s summary of that survey documents a striking stress reduction: respondents reporting work-related stress dropped from 65% before the change to 36% after, and 72% agreed with the statement “I feel like a new person.” Recent AARP Public Policy Institute data (n=2,249, fielded mid-2023) shows that 34% of adults age 50+ participated in job-related skills training in the past two years, and 60% report willingness to retrain if asked — a population scale that makes midlife retraining a common rather than exotic decision.
On the inaction side, an AARP survey of 3,580 US adults age 50+ fielded in late 2024 found that 24% are planning a job change in 2025, a 10-percentage-point jump from 14% in 2024. The figure is used here as a latent-dissatisfaction proxy: workers still in their current role who actively want out are signalling near-term regret about the stay-put choice. The Center for Retirement Research’s documentation of the stress and identity gap between changers and stayers supports the directional reading. Both sides are constructed proxies, and the entry is published with proxy_only: true — no nationally representative survey directly asks “do you regret retraining at 40+” or “do you regret not retraining.”
The inaction-dominates pattern (regret_delta -0.14) is consistent with Gilovich and Medvec’s general finding that long-term regret tends to skew toward inactions, and here the gap sits in the range of the typical 10-15 percentage point inaction premium — but that comparison is fragile, because midlife retraining carries real action-side costs that the proxies understate: tuition, foregone earnings during retraining, age-discrimination headwinds in hiring (74% of AARP respondents cited age as a barrier), and uncertainty about whether new credentials will pay off in a remaining career horizon of 15-20 years. The expected-value math is highly sensitive to retraining type. Short certifications and bootcamps in tech and healthcare show fast placement and wage recovery; a second bachelor’s or master’s at 45+ may take a decade to pay back. Above age 60, the picture shifts again — shorter horizons and stronger age-discrimination compress the payoff window further, narrowing the action-side advantage that drives the modest inaction-dominates pattern at 40-50.