Adult social care labour supply: pay increase impact assessment

2. The Scottish ASC labour market

2.1 Aggregate trends in employment and pay

To understand the context for the new policy, Figures 1 and 2 plot the trends in employment and wage levels for the residential (Figure 1) and non-residential (Figure 2) adult social care (ASC) subsectors in Scotland. The left-hand panel in each figure plots annual employee headcounts while the right-hand panel plots the first five deciles of real hourly pay in the subsector, i.e., taking inflation into account.^[2],[3] Both figures indicate that real hourly pay (excluding overtime) increased between 2016 and 2022. The 10^th percentile of hourly pay increased by approximately 16.4% in residential care and 15.5% in non-residential care, while median pay increased by 14.0% and 11.6%, respectively. These increases reflect longer term trends, as opposed to the temporary spike in hourly pay above the 30^th percentile in 2021, which is likely an artefact of a larger share of lower-wage workers leaving the sector that year, thereby mechanically shifting upwards the wage distribution.

On the other hand, employment growth has been unsteady, with trends across subsectors diverging. Between 2016 and 2022, employment in residential care decreased by approximately 4.9% while employment in non-residential care services increased by about 5.0%. While the sharp fall in employment in both groups between 2020 and 2021 may be pandemic-related, the pattern of decreasing residential care and increasing non-residential care employment appear to be part of a longer-term trend. In the years leading up to 2020, employment in residential care had decreased by approximately 1.1 percent between 2016 and 2019, while employment in non-residential care services increased by about 2.4 percent. The net effect of these changes is a roughly 1% increase in total (i.e., residential and non-residential care) employment between 2016 and 2022.

2.2 Aggregate trends in retention

Changes in employment levels are driven by underlying changes in recruitment and retention. Year-on-year employment growth figures can thus mask significant levels of churn. That is, employment levels can appear stable even with high rates of employee turnover. While turnover data is not available for the Scottish social care workforce, the annual SSSC Workforce Data Report publishes Stability Indices (SI), which measure the percentage of staff retained in the same post from the previous year. For example, a SI of 75 in 2022 implies that 75% of staff employed in 2021 are still in the same post in 2022. Figure 3 plots the stability indices for the entire Scottish social services sector and for the Care homes and Care at home subsectors. It shows a general decline in SI between 2016 and 2022. Within this period, all three indices increased sharply in 2020 (likely due to initial job protection policies during the early waves of the pandemic) but subsequently dropped to below pre-pandemic levels in 2021.

While not directly comparable, for a fixed level of employment, a high SI implies fewer leavers and hence lower staff turnover. The Scottish social care workforce, therefore, appears to show a similar longer-term trend as the English ASC workforce, where the turnover rate in England increased from 27.1% in 2016/17 to 29.1% in 2022/23 (Skills for Care, 2023).

To sum-up, between 2016 and 2022, median pay in the Scottish social care sector increased between 11 and 31% (for non-residential and residential care, respectively), while employment growth in adult social care services has been weak and retention appears to be worsening. However, aggregate trends in pay and employment are, in general, affected by various factors such as wider labour market (i.e., non-social care) shifts, changes in demand for different types of care services (e.g., residential vs. non-residential care), and possible systematic differences across subsectors and localities. The preceding discussion therefore cannot be used to infer the relationship between employment and pay, which is the key quantity of interest. Instead, such an analysis would require more granular data (i.e., at employer and employee levels) and an approach to account for simultaneous non-wage factors that affect employment.

In the next section, we outline how we use employee-level data on pay and job/employer characteristics to address the confounding factors described above, thereby isolating the relationship between wages and employment.