Abstract

Background and aim: Social care is a highly labour-intensive industry and therefore issues around wage rates, turnover, and contracts are likely to impact on quality. Currently the UK care home sector has relatively high levels of staff turnover and vacancy rates (Skills for Care, 2015) and there is a potential future workforce shortage (ILC, 2015). US evidence (e.g. Bostick et al., 2006; Castle et al., 2008; Castle 2009) has generally found that poor workforce characteristics (e.g. higher turnover, lower levels of staffing) have a significant negative effect on quality indicators, but there is very little evidence for the UK.

The aim of this paper is to estimate the impact of workforce composition on the quality of English care homes. We hypothesise that better work conditions (e.g. higher pay, lower staff turnover, fewer temporary workers) will lead to higher quality ratings in care homes.

Data and methods: The analysis uses the national health and social care regulator (CQC) database of registered care homes and their CQC quality rating at 1st March 2016, matched with care homes data available from the market specialists Laing & Buisson, and the Skills for Care National Minimum Dataset for Social Care (NMDS-SC) from 2014/15. The NMDS-SC is a rich source of information on care provider staff characteristics and staffing levels in England. Data was available on over 7,500 independent sector (i.e. for profit/voluntary) care homes in the year from May 2014 to April 2015.

We estimate probit models of a binary measure of CQC quality rating and examine how this is impacted by local area workforce characteristics (average care worker wage, staff turnover rate, percent of temporary workers, all at postcode-district level). The regressions include controls for care home- (e.g. type of care home, size, sector) and local area-level (e.g. level of needs, older population) characteristics, and the level of competition that a care home faces.

Two assumptions are made: first, that the impact of staff pay and conditions will take time to impact on care home quality; and second that local social care market conditions will impact on individual care homes. Whilst the time-lagged and geographical nature of the workforce data should mitigate potential reverse causation issues (i.e. quality impacting on staffing decisions), we use spatial instrumental variables to control for the potential endogenous relationship between quality and both staff conditions and competition.

Results and policy implications: The findings suggest that local workforce characteristics do have a significant impact on quality. The results are based on a cross-section and only confirm a correlation between quality and workforce characteristics. Causation would have to be examined using a longitudinal analysis. This is an interesting finding for policy given that care providers are currently facing both income pressures as a result of the continued climate of public sector spending reductions and cost pressures due to the introduction of the new national living wage.

Slides