Background: England's population with long-term care (LTC) needs is increasing rapidly due to population aging and the rise in life expectancy of people with physical, learning disabilities, and mental health problems. To meet the growing demand for care services, the number of LTC jobs increased by about 20 per cent over the last decade and is estimated to increase by another 32 per cent by 2035 (Skills for Care, 2020). The increase in demand for LTC workforce has been accompanied by lagged labour supply and high staff turnover, leading to concerns about the sector's capacity to meet the increasing demand. One of the factors assumed to be related to low recruitment and retention in LTC is low pay, often at minimum wage.
Objectives: The aim of this study is to bring much needed evidence on explained and unexplained wage differentials between LTC (direct care) job roles and employers as well as on the responsiveness of LTC labour supply to wages.
Methods: We use worker level data for 2016 to 2019 from the Adult Social Care Workforce Data Set (ASC-WDS). Wage elasticities of labour supply are derived from elasticities of job separation, using a dynamic monopsony model. The identification strategy is based on controlling for confounding factors at individual, employer, and local market level, and using panel data methods to account for unobserved worker and employer hererogeneity.
Results: After accounting for worker, job, and employer characteristics, we find substantial wage differentials between sectors, with wages being about 20 percent lower in the private and 15 per cent in the voluntary sector compared to the public sector. Wage differentials between sectors are higher for senior care workers, compared to care workers, mainly due to the lack of pay progression in the private and voluntary sectors. When looking at wage elasticities, we found that LTC labour supply was substantially more responsive to wages than found by previous studies. Using longitudinal data and being able to control for time-invariant unobserved heterogeneity, we could show that unobserved heterogeneity bias was important, reducing estimated wage elasticities by two to three times, depending on care setting. We show that wage elasticities are higher at lower wage levels, for workers over age 25, and in the private (for-profit) sector.
Conclusions: Overall, our findings show that wages could be used to increase labour supply in LTC in England and should be considered in any policy aimed at improving the sustainability of the sector.