Same Firm, Different Staff: Dementia Group Homes Employ More Experienced Carers Than Day Services

For-profit ownership has long been linked with thinner staffing in long-term care, but most studies compare providers within a single service type. A new study in BMC Health Services Research takes a different angle: it asks whether the experience composition of care staff differs between two service lines operated by the very same firm. Using a census of 24,567 adult day services and 14,236 dementia group homes drawn from Japan’s Long-Term Care Service Information Disclosure System, the researcher modelled the shares of full-time care staff with at least three and at least ten years of occupational experience using fractional logit regression with prefecture fixed effects and firm-clustered standard errors.

The between-firm comparison first confirmed a familiar pattern. In adult day services, for-profit establishments had lower shares of staff with at least three years’ experience than social welfare establishments, with an odds ratio of 0.718 (95% CI 0.675–0.765), and an even lower share of staff with at least ten years, at 0.617 (95% CI 0.572–0.665). The ten-year gap was smaller in dementia group homes, suggesting that the experience penalty associated with for-profit operation is not uniform across service lines.

The core of the study, however, lies within firms. Among the 2,346 firms operating both an adult day service and a dementia group home, firm fixed-effects models compared the two lines directly, holding firm-level characteristics constant. The group home’s share of staff with at least ten years’ experience exceeded the day service’s by 2.81 percentage points (95% CI 1.46–4.16) when establishments were weighted equally, and by 2.61 points (95% CI 1.19–4.04) when firms were weighted equally. For the three-year experience threshold, the differences were 2.72 and 1.60 points respectively, although the median firm-level difference was zero, indicating that the average gap is driven by a subset of firms rather than a universal pattern.

The results were not identical across ownership types. Among social welfare corporations, the group-home advantage in experienced staff held under both weighting schemes and remained after excluding the largest chains. Among for-profit firms, the picture was more fragile: the three-year difference depended on the largest chains, shrinking to 0.08 points when firms were weighted equally, and the ten-year difference of 2.62 points (95% CI 0.17–5.08) was not significant after Holm correction for multiple comparisons (p = 0.073).

Two further contrasts in working conditions emerged. The full-time share of care staff was 9.87 percentage points higher in group homes than in day services within the same firms, and turnover was 1.90 points higher. The higher turnover weighs against, without excluding, a simple retention explanation for the group-home experience advantage: if group homes simply kept staff longer, one might expect lower, not higher, turnover there.

The author is careful to frame the finding as descriptive rather than causal. Heavier regulation of dementia group homes is one plausible explanation for their more experienced workforces, but the two service lines also differ in the nature of the work, the users they serve, capacity and working conditions. Staff movement between establishments is unobserved, so the data cannot show whether experienced workers migrate toward group homes or are recruited differently. The study also relies on establishment-level disclosure data, which capture shares of experienced full-time staff but not the underlying employment dynamics.

The methodological point may be the study’s most transferable contribution. Service-by-service evaluation, the standard approach in ownership and staffing research, cannot compare a firm’s own service lines, yet public establishment-level data of the kind Japan publishes make such within-firm comparisons feasible. By conditioning on the firm, these designs remove the confounding that arises when multi-service operators differ systematically from single-service providers in resources, management or location.

For long-term care policy, the modest but consistent experience gap within firms suggests that workforce composition is shaped partly at the service-line level, not only by firm ownership. Whether similar within-firm differences appear in other countries with multi-service long-term care providers, and whether they translate into differences in care quality, remain questions for future research.

Subject of Research: Within-firm differences in care-staff experience between adult day services and dementia group homes in Japan

Article Title: One firm, two workforces: within-firm differences in care-staff experience between adult day services and dementia group homes in Japan

Article References: Furuya, A. (2026). One firm, two workforces: within-firm differences in care-staff experience between adult day services and dementia group homes in Japan. BMC Health Services Research. https://doi.org/10.1186/s12913-026-15936-7

Image Credits: AI Generated

DOI: 10.1186/s12913-026-15936-7

Keywords: long-term care, dementia group homes, adult day services, for-profit ownership, care staff experience, Japan, staffing, turnover, fractional logit, firm fixed effects, care workforce, firm