Care talk, DEI talk, and the quits rate
Corpus-wide care talk (rest, balance, mental health, sustainable pace) and DEI talk by year, overlaid on the JOLTS quits rate — how freely workers were leaving jobs. All three standardized onto one scale.
Source: Care talk climbs from 2019 and peaks in 2021–22, in near-lockstep with DEI talk and the quits peak. The pandemic years are partly confounded with lockdown wellness talk; the clean, unconfounded part is the deflation — as quits fell 2.8 → 2.1 (2022–2024), care talk fell with it. Quits rate is total-nonfarm, not the information sector, and the overlay rests on twelve annual points.
The same pair, one company at a time
Each firm's own care and DEI talk, standardized within the company. Pick a company.
Source: The paired rise-and-recede shows in all 20 firms (median within-company correlation +0.65, 95% CI [0.57, 0.84]), from GitLab (r = +0.95) down to Shopify (+0.13); the panel has no negative case.
Which themes spiked in 2020
Each language axis's 2020 deviation from its own historical average.
Source: Only the worker-concession axes spike. Performance, meritocracy, and control language barely move — the surge was targeted, not a general values-forward lift.
Care benefits, split by who absorbs the cost
Advertised mental-health benefits (EAP, therapy stipends, wellness apps — individual-locus in 19 of 20 codings) against family and caregiving benefits (paid leave, childcare — structural in 105 of 125), by year.
Source: Mental-health benefits climb to the end of the record; family/caregiving benefits crest around 2019 and recede. Locus codebook hand-validated at Krippendorff's α = 0.90. Counts are modest, and all of it is self-presentation — what firms advertise, not what they provide.
Whole-life benefits as a share of everything advertised
Benefits that support life outside work — parental leave, childcare, elder and dependent care, fertility, lactation, identity-positive coverage, bereavement — as a share of all advertised benefit mentions, by year.
Source: The share peaks in 2017–18 at roughly one mention in ten, before the pandemic-era talk spike, and shrinks after. Parental leave is the sturdiest thread; fertility totals fifteen benefit mentions ever (an anecdote, not a trend); lactation, bereavement, and identity-positive coverage barely register.
What this can’t say
- The aggregate individualization index this study was designed around is too sparse to test: only 8 of 15 companies have enough enumerated benefits on both sides of 2022, at 2–5 items each.
- Benefits data throughout is self-presentation, not provision.
- The strong claim is about language only: a worker-concession bundle (care + DEI talk) that inflated with the labor market and deflated as leverage returned, while performance language never moved. The benefit shapes are texture — vivid where clean, silent where the counts run out.