Do supermarket openings reduce care home staffing?

Introduction

The adult social care sector in England faces long-standing and well-documented workforce challenges. Of perennial concern are adult social care providers’ struggles to attract and retain sufficient numbers of suitably skilled staff.

In an attempt to ‘address the recruitment and retention crisis in the sector’, the government plans to facilitate ‘Fair Pay Agreements’ between employers and employees, with the aim of improving pay and conditions, and attracting more workers.

In recent years, increases in the number of care staff have been driven by recruitment from abroad, rather than domestic recruitment. Following the recent decision to close the social care visa route to new applicants, efforts to attract more domestic workers via improved pay and conditions have taken on even greater significance. We need to understand, then, the factors which determine the ability of social care providers to attract and retain staff.

One frequently cited factor behind the sector’s struggles is that other sectors of the economy offer more attractive wages and/or working conditions than the financially constrained social care sector. A particularly common observation is that care workers would be paid more if they worked in supermarkets. The president of the Royal College of Emergency Medicine, for example, publicly called for additional funding for long-term care in order to ‘Pay people more than stacking shelves in Tesco gives them’. This summer, Sir Ed Davey MP argued in the House of Commons that we ‘still see people earning more in Amazon warehouses and supermarkets than in care homes, and that will mean our loved ones going without the care they need’. There are many other such examples.

Despite the prevalence of the narrative that social care providers are frequently losing staff to supermarkets, there is limited empirical evidence to support it. Previousresearch has found that relatively few workers move from adult social care into retail.

In a new IFS working paper, we build on this by examining the effects of supermarket openings on the local social care workforce. We also examine the role of new Amazon warehouses.

Our analysis suggests that supermarket and Amazon warehouse openings had no large impacts on staffing in local care homes (in terms of the number of staff, their turnover and vacancy rates, and various other measures). Our central estimate is that there was no statistically detectable effect. This runs contrary to the popular discussion and, with the introduction of Fair Pay Agreements on the horizon, provides new evidence on the workforce of this increasingly important sector.

Methodology

In our primary analysis, we focus on the opening of large supermarkets, such as those in out-of- town retail parks (rather than small, city-centre supermarkets). We use data collected on supermarkets to identify 178 openings of large supermarkets in England between 2016 and 2022. We also document 12 new Amazon warehouses over this period. While Amazon warehouse openings are much rarer, they also tend to be much larger employers than supermarkets.

We combine information on supermarket and Amazon warehouse openings with data on the adult social care workforce, collected by Skills for Care (covering around half of all care providers). We focus on staff working in care homes, which makes up around 40% of the total adult social care workforce, equivalent to 700,000 filled roles in 2023–24. For each care home, we define a local labour market based on where its staff live. We consider a care home to potentially be affected by an opening if a new supermarket or warehouse opens within its local labour market. Most care homes in our sample are too far away from any such opening, and therefore we define them as unaffected.

To estimate the effect of these openings, we must account for the fact that supermarket and Amazon warehouse openings are likely to depend on local factors, some of which are hard to measure, and may also be related to the workforce measures we study. For example, the decision of whether or not to open a new supermarket may be based on local workforce or economic conditions, which could also affect local care homes.

We therefore employ what is called a ‘difference-in-differences’ research design. This approach compares workforce outcomes (such as staffing numbers, turnover rates and vacancy rates) in affected and unaffected care homes before and after the opening. That is, we look at what happened to staffing numbers in care homes where a supermarket opened nearby, and compare this with what happened to staffing numbers in places where no supermarket opened. This approach allows us to identify the local impact of a new supermarket or warehouse by controlling for broader trends in the adult social care workforce and for persistent differences in the workforces of care homes in different areas. We also control for relevant population and economic characteristics of local areas. Finally, we examine whether care homes affected by the opening of a supermarket or warehouse had different workforce trends (e.g. in staff numbers) from other care homes prior to the opening. If there are no such trends, this is a reassuring validation that we are likely capturing the true effect of an opening.

Results

We begin by examining whether the opening of a new supermarket affects the number of staff employed by nearby care homes. Figure 1 shows our estimated effects at six-month intervals, from four years before the opening to four years after. Each green point represents the estimated effect at a given time point, and the vertical lines show 95% confidence intervals.

Figure 1. Effect of a supermarket opening on total care home staffing

The results show that supermarket openings have no meaningful effect on care home staffing. In the first year after opening, the average estimated impact is an increase of 0.1 staff per care home – a change that is not statistically different from zero.

For context, the average care home in our sample employed 33 staff in the six months before a supermarket opens. The 95% confidence interval of the effect over one year spans from a reduction of 0.3 staff to an increase of 0.5 staff, equivalent to –1.0% to 1.5% of the pre-opening average employment per care home. This means that we are able to rule out large changes in staff numbers. We also find no significant differences in staffing prior to a supermarket opening, which is a reassuring validation of our methodology.

We find similarly small effects over longer time horizons. For example, over four years, the estimated impact is an increase of 0.03 staff per affected care home, which is also not statistically different from zero effect. Across a range of robustness tests, we also find small to no impacts of openings. In a few cases, we find that new supermarkets may have significantly (in a statistical sense) reduced the number of staff in care homes. But in these cases, the magnitudes of the reduction are still very small, and therefore indicate that staff numbers in care homes were not meaningfully reduced following the opening of a supermarket nearby.

We repeat our analysis using the openings of Amazon warehouses (Figure 2). Here we also find little evidence of a large decrease in staffing in response to an opening – if anything, the effect is more positive than for supermarkets (i.e. implying a small increase in care home staffing levels after an opening), though it still remains statistically insignificant.

Figure 2. Effect of an Amazon warehouse opening on total care home staffing

While we find no impact on overall staffing, there could be other changes to the workforce that are still important. For example, it may be that some staff are leaving for these new job opportunities and care homes are hiring new staff to replace them, increasing staff turnover. However, we find no evidence that the rate of staff leaving or joining care homes changes after a supermarket or Amazon warehouse opens. This also suggests that openings do not reduce the pool of workers willing to move into working in care homes. Although vacancies are harder to interpret, we also find no evidence of a significant increase in vacancy rates.

What might explain our findings?

All in all, our analysis suggests that the opening of new supermarkets or Amazon warehouses has no meaningful impact on the local adult social care workforce. There are a number of potential explanations for why we do not find a large effect.

First, it may be that care homes are increasing their wages, or changing other aspects of their job offer, in response to the opening of new supermarkets or warehouses. However, when we examine this directly, we find no evidence that care homes raise their wages, or change their training provision, in response to new supermarket openings.

Second, contrary to the prevailing narrative, it may be that supermarket and Amazon warehouse jobs are not relevant or attractive outside options for care home staff, requiring different skills or appealing to a different section of the labour force (perhaps due to the out-of-town location of many large supermarkets).

Third, a new large supermarket may employ around 100–150 staff, which compares with 472 care home staff employed in the (average) local area in our sample.1 Amazon warehouses are even larger shocks, employing an average of 750 at each site. This suggests that these shocks are large relative to the local adult social care labour market. But they may be small relative to the total low-paid labour market in an area, and the impacts may therefore not be statistically detectable with our methodology.

Conclusion

Our analysis finds no meaningful impact of new supermarkets and Amazon warehouses on the local care home workforce. This likely reflects a combination of factors: these job opportunities may not be relevant to many care home workers, and these shocks may be relatively small in the context of the overall local labour market these workers are operating in.

While we cannot rule out that other aspects of these jobs – such as their level of pay – may affect the social care workforce, our findings suggest that the number of new retail positions created is not, in itself, a key driver of workforce challenges in adult social care. It is important to note that we study the impact of new supermarket jobs, not the impact of supermarkets in general, and so we cannot definitively rule out any relationship between the two sectors. Nonetheless, these results suggest that the popular narrative may be overstating the role of retail relative to other important outside options for care workers, including those in the NHS.

In forthcoming collective pay negotiations between employers and employees (which recent legislation seeks to encourage and facilitate), our analysis indicates that job opportunities within the supermarket sector ought not to be used as a central benchmark for that of care home staff. The drivers of recruitment and retention issues would appear to lie elsewhere. Other potential contributing factors – such as burnout, low professional status, limited opportunities for progression, workplace culture, and frustrations around pay more generally – might helpfully be explored in future research.  

_{We gratefully acknowledge financial support from the UK Economic and Social Research Council through the Centre for the Microeconomic Analysis of Public Policy (CPP) at IFS (ES/Z504634/1) and through Zaranko's new investigator grant (ES/Z503514/1); and the National Institute for Health and Care Research (NIHR) Policy Research Programme (NIHR206121 – NIHR Policy Research Unit in Health and Social Care Workforce). The views expressed here are those of the authors and not necessarily those of the NIHR, the Department of Health and Social Care, other funders or data providers. We also thank members of the PRU Public Contributor Involvement and Engagement group, in particular David Winskill, and the PRU Workforce Panel for their helpful comments.}