The Effects of Minimum Wage Policy on the Long‐Term Care Sector in England

1 Introduction

The introduction of the National Living Wage (NLW) in April 2016 was a major labour market intervention, with the target of increasing wages for minimum wage workers aged 25 and over to 60 per cent of median earnings by 2020. Between April 2015 and April 2016, the year‐on‐year wage growth for minimum wage workers aged 25 and over was boosted by more than 10 per cent, through the aggregated October 2015 National Minimum Wage (NMW) rate increase (from £6.50 to £6.70) and the new April 2016 NLW rate of £7.20. This represented more than three times the 2015/16 growth in median earnings (of 3.1 per cent), and was one of the highest increases in NMW since its introduction in April 1999 (Low Pay Commission 2016).

This wage increase affected employers’ payroll costs; and stakeholders in several labour intensive, low‐pay sectors — including long‐term care (LTC) — expressed concerns about future sustainability (Low Pay Commission 2016). Compared to other sectors, the LTC sector faced an additional constraint to adjusting to the payroll cost increase: reduced public expenditure. A large proportion of services users are publicly funded and austerity measures during the 2010s meant that there were substantial cuts to adult social care budgets by local authorities (Fernandez et al. 2013; Luchinskaya et al. 2017). Therefore, care providers lacked the flexibility of passing cost increases on to charged fees, and changes to minimum wage legislation were expected to have a noticeable impact on the sector. The main concerns were that the increase in minimum wage will have a negative effect on employment, wage distribution, wage differentials, promotion prospects, employment conditions, quality of care, market sustainability, and/or noncompliance with minimum wage regulation (Allan and Vadean 2020; Gardiner 2015a,b; Gardiner, Hussein 2015; Low Pay Commission 2016).

The aim of this study was to assess the effects of minimum wage policy on the LTC sector in England. Using the Adult Social Care Workforce Data Set (ASC‐WDS),¹1 Previous known as the National Minimum Data Set for Social Care (NMDS‐SC).
a large and rich employer‐employee linked dataset, we were able to identify effects on type of care homes (i.e. with and without nursing) as well as on domiciliary care providers. Within both residential and domiciliary care, we also differentiated between private (i.e. for‐profit) and voluntary (i.e. not‐for‐profit) establishments. We believe the distinction to be important, as the voluntary sector has traditionally paid relatively higher hourly wages and offered better employment conditions (Skills for Care 2016a,b). As employment effects are usually elusive (see next section for a brief literature review), we also explored the effects of minimum wage policy on an alternative potential adjustment mechanism: the employment of workers on contracts without guaranteed working hours (i.e. zero‐hour contracts (ZHC)). Care providers could use ZHCs to increase the efficiency of labour input, by managing more flexibly the fluctuations in demand for services (including loss of contracts). Nonetheless, the extensive use of this contract type (in particular by domiciliary care providers) has received substantial criticism due to the low job and financial security it offers employees (CIPD 2013; Skills for Care 2019b).

Our results show that the minimum wage increase had a strong and positive effect on the increase in the average wage at establishment level between April 2015 and October 2016. The minimum wage also generated quite a substantial compression at low wage levels. In care homes, for example, the share of care workers aged 25 and over at (or below) minimum wage increased from about 35 per cent in April 2015 to about 49 per cent in October 2016. With everything else being equal, the share of staff paid initially under the new minimum and the wage gap (i.e. the average increase in wages needed to bring staff pay to the mandatory minimum) had a short‐term negative effect on total employment and a longer‐term negative effect on weekly working hours for care homes, in particular for care homes without nursing and those owned by private (i.e. for‐profit) providers. We also found some evidence that the minimum wage increase had a short‐term positive effect on the change in the share of care workers on ZHCs in both residential and domiciliary care.

To our knowledge, a previous version of this study made available as a discussion paper in July 2017, was the first to assess the impact of minimum wage policy on employment contracts without guaranteed hours; see (Vadean and Allan 2017).

In the next sections, we discuss literature relevant to the topic, give a brief overview of the minimum wage policy in the UK and the English LTC labour market, we present the dataset and the sample analysed, the empirical strategy, and then discuss the results and conclude.

2 The effects of minimum wage policy on wages and employment

Most studies on the impact of minimum wage have focused on employment effects. Starting with the ‘new minimum wage research’ at the beginning of the 1990s (Card 1992; Card and Krueger 1994, 1995), a large strand of literature developed showing mainly that although minimum wage policy had significant positive effects on wages, it had no or very small negative effects on employment. Employment effects were ‘elusive’ despite studies focusing on groups with high shares of employees on minimum wage (e.g. teenagers or low educated workers) or low‐wage sectors (e.g. hospitality, retail, and LTC), for which one would expect an upward bias due to selection. A number of literature reviews and metastudies give a good overview of this research strand (Belman and Wolfson 2014; Chletsos and Giotis 2015; Doucouliagos and Stanley 2009; Linde Leonard et al. 2014; Metcalf 2008; Schmitt 2015).

Economic theory offers a few explanations for the ‘elusive employment effect’, depending on the theoretical approach. The standard competitive model predicts the main adjustment — to a wage level set above the competitive wage — to be through declining employment, but allows also for adjustment through increasing prices to consumers, reduction in non‐pay benefits, reduction in training, and/or changes in the skill mix. In the institutional model, firms are assumed to operate below maximum efficiency because it is costly to identify, implement, and maintain practices that continuously maximize efficiency. In this context, a minimum‐wage increase would give employers an incentive to improve efficiency (Hirsch et al. 2015; Kaufman 2010). Higher wages may also increase workers productivity by inducing them to work harder to ensure that they keep their job or by increasing job satisfaction (Akerlof 1982; Hirsch et al. 2015; Shapiro and Stiglitz 1984). Moreover, by increasing the spending power of low wage workers, a minimumwage increase might act as an economic stimulus, increasing demand for the firms’ outputs (Aaronson et al. 2012; Hirsch et al. 2015). Alternatively, in the dynamic monopsony model, market power comes from labour market frictions (i.e. employers face real costs associated with hiring new workers and workers incur costs to find new jobs). Therefore, a minimum wage set below the market competitive wage could raise both wages and employment (Hirsch et al. 2015; Manning 2016; Schmitt 2015).

The first study that looked at the employment effects of minimum wage in the English LTC sector was by Machin et al. (2003). They surveyed care homes in England before and after the introduction of the NMW in April 1999, and obtained data from about 2,000 care homes on workforce, care home characteristics, and managers’ views about minimum wage; their balanced panel included 641 care homes. They found that the NMW introduction had a compression effect at lower levels of the wage distribution, generating an about 30 per cent spike at minimum wage. Moreover, they found a positive and significant effect on average wages and a modest negative effect on employment and hours, but no effects on prices or productivity.

Given the low effects on employment, a few studies tried to identify mechanisms that care homes might have used to adjust to the NMW introduction. Machin and Wilson (2004) found that minimum wage had no effect on home closures either. Georgiadis (2013) showed that the wage costs generated by the April 1999 NMW introduction were at least partly offset by lower monitoring expenses, measured as the ratio of supervisors to supervised employees, while Draca et al. (2011) using data from the Financial Analysis Made Easy (FAME) database, found evidence that the NMW introduction reduced care homes’ gross profit margin.

The introduction of the NLW in April 2016 generated a new interest in the analysis of effects of minimum wage on the LTC care sector in England. Gardiner (2016) using data from the ASC‐WDS found no evidence that LTC providers reduced employment hours to offset the increased payroll cost after the NLW introduction, but an increased compression of the wage distribution at lower wage levels. A further study commissioned by the Low Pay Commission (Giupponi et al. 2016) aimed to replicate the earlier work of Machin et al. (2003) and Machin and Wilson (2004). Their results from primary data analysis of a survey of care homes in England show that the share of care assistants (i.e. care workers) paid below £7.20 decreased from about 60 per cent to about 4 per cent with the NLW introduction, with significant increases in average wages. However, despite the large effects on wages, they found no effects on employment, prices, or productivity (Low Pay Commission 2016).

An earlier version of this study was the first to assess the impact of minimum wage policy on employment arrangements without guaranteed hours (i.e. ZHC) and found a positive effect; see (Vadean and Allan 2017). A later study using both the ASC‐WDS, the Quarterly Labour Force Survey (QLFS), and survey data on alternative work arrangements similarly found a positive effect of minimum wage increases on ZHCs in the LTC and other UK low wage sectors (Datta et al. 2019). Moreover, a further study using the ASC‐WDS found that although the introduction of the NLW in April 2016 had no negative effect on employment in care homes in England, it had a negative effect on changes in Care Quality Commission (CQC) quality ratings (Giupponi and Machin 2018).

This article extends this literature by analysing an aggregated effect of two increases in minimum wage that took place only six months apart (i.e. October 2015 and April 2016), and thus, improving the likelihood of identifying any potential effects on employment and other outcomes. Moreover, we differentiate between effects on either private (i.e. for‐profit) or voluntary (i.e. not‐for‐profit) providers, and within residential care between care homes with nursing and without nursing. The distinction between these groups of care establishments is important, due to the differences in pay levels and structure, skill mix, and employment conditions.

3 The National Minimum Wage in the UK

The NMW was introduced in April 1999 with two rates: £3.00 for workers aged 18 to 21, and £3.50 for workers aged 22 or older. Since October 2000, the rates have been adjusted yearly on advice from the Low Pay Commission, new rates have been introduced (i.e. one for workers aged 16 to 17 in October 2004, and one for apprentices in October 2010), and the age bands have been slightly changed. In October 2015 (i.e. the last adjustment before the introduction of the NLW, there were four NMW hourly rates: £3.30 for apprentices, £3.87 for workers aged under 18, £5.30 for workers aged 18 to 20, and £6.70 for workers aged 21 and over.

The NLW was introduced in April 2016 for workers aged 25 and over at a rate of £7.20, representing 56 per cent of median hourly earnings, and since April 2017 both the NLW and NMW rates are adjusted yearly every April. The NLW is different from the other NMW rates in that it has the objective of reaching 60 per cent of median earnings by 2020 and two‐thirds of median earnings by 2024. Conversely, the NMW rates are adjusted based on affordability, that is, rates are set to be ‘as high as possible without damaging the employment prospects of each group’ (Department of Business, Energy and Industrial Strategy 2020).

4 The long‐term care workforce

Skills for Care estimated that the number of people working in the LTC sector in England in 2015 was about 1.43 million, filling about 1.55 million jobs. The majority of these, or about 80 per cent, were jobs with independent sector employers (i.e. about 60 per cent in the private sector and a further 20 per cent in the voluntary sector). In terms of service groups, most jobs were in residential care (i.e. care homes with or without nursing; 43 per cent) and domiciliary care (e.g. home care; 42 per cent) (Skills for Care 2016a).

Due to rising demand for care services, the LTC workforce has grown steadily. Even though public spending on LTC services in England dropped by about 17 per cent, between 2009 and 2015, the number of LTC jobs has increased by about 3 per cent per year (The Health Foundation 2017; Skills for Care 2016a).

Despite the strong increase in demand and employment, and probably reflecting reduced public funding, pay and conditions in the care sector in England rank rather poorly (Gardiner and Hussein 2015). The average hourly wage for care workers — the frontline, direct care staff representing over half of the jobs in the LTC sector — was £7.46 in 2015/16; this pay rate being at about the 10th percentile and less than half the mean UK hourly earnings. Care workers were paid slightly better in domiciliary care (£7.58; without taking into account travel time between clients) than in residential care (£7.20), and in statutory local authority jobs (£9.67) than in the independent sector (£7.35) (Low Pay Commission 2016; Skills for Care 2016a, b). The distribution of wages in the LTC sector are also quite significantly compressed at low wage levels compared to other sectors, with the ratio of minimum wage to median pay (i.e. the ‘bite’) at about 78 per cent in 2013, compared to only 52 per cent overall (Gardiner and Hussein 2015).

Although the majority of the LTC workforce are employed on permanent contracts (90 per cent), a quarter (24 per cent) were employed in 2015/16 on ZHCs (i.e. employment contracts that do not guarantee a minimum level of working hours), with the highest proportion in domiciliary care (49 per cent) and among care workers (58 per cent) (Skills for Care 2016b). Opportunities for progression are also reported to be rather limited, with a flat hierarchy in which the ratio of senior care workers to care workers in domiciliary services declined from 7 to 4 per cent between 2008 and 2012 (Gardiner and Hussein 2015).

Given such pay and conditions, it is rather surprising that employment in the sector rose. The evidence seems to show that the majority of workers entering the sector have low education levels and limited access to higher paid jobs and/or are looking for part‐time or flexible working time jobs that can be fitted around other (caring) responsibilities. The majority of social care workers are female (80 per cent), with a mean age of about 43, having a low level of formal qualifications, and a growing number are migrants (Gardiner and Hussein 2015). However, despite the increase in employment, care providers are reporting high turnover (27 per cent) and vacancy rates (7 per cent), providing an important challenge to the provision of good quality services (Allan and Vadean 2020; Skills for Care 2016b).

5 Data and descriptive statistics

We used data from the ASC‐WDS, the leading source of workforce information for the LTC sector in England collected and managed by Skills for Care. It holds information on over 20,000 establishments and 700,000 workers, therefore, covering about 50 per cent of the LTC market. The dataset is updated regularly by employers: public employers update data on a mandatory basis in September each year, while independent employers submit data on a voluntary basis and are incentivized by access to workforce development grants. All data in the ASC‐WDS have been updated or confirmed to be up to date within the last two years and about 90 per cent of employers have updated their data in the past 12 months. Although the dataset does not cover all independent sector establishments, it does have a large enough sample to provide a solid basis for reliable workforce estimates at both national and local level. All ASC‐WDS data have been validated at source and has undergone rigorous data quality checks (Skills for Care 2019a,b).

We focused the analysis on the period April 2015 to October 2016, capturing both the October 2015 NMW adjustment and the introduction of the NLW in April 2016. We allowed a six‐month period after the NLW introduction, to capture changes for providers who updated their ASC‐WDS entries with a delay. We excluded provider records that were not updated for more than six months, that is, for the April 2015 cut‐off, records that were not updated since the last NMW adjustment (i.e. October 2014), and for the October 2016 cut‐off, records that were not updated since April 2016 (i.e. NLW introduction).

There are two ASC‐WDS analysis files, one at establishment level and one at worker level, both being fully anonymized. These files include rich information about establishments (e.g. location, type of care provided, client type, count of employees, and job roles) and employees (e.g. demographic characteristics, job role, contracted hours, pay, qualifications, and training). We mainly utilized establishment level data for the analysis, but data in the employee‐level file was used to generate establishment level variables capturing the workforce mean age, gender structure, mean hourly wages, total weekly hours, and the share of workers on ZHCs.

Some employers had incomplete data on their employees. When computing the mean hourly wage, we set this to ‘missing’ where the employer had hourly wage information for less than 75 per cent of its care workers. Similarly, we set to missing the total weekly hours and the share of workers on ZHCs where there was no complete information for all care workers.

LTC establishments are heterogeneous, offering a variety of services (e.g. residential care, domiciliary care, day care, meals on wheels, supported living, etc.). We focused our analysis on establishments offering one of three types of services which together employed about 80 per cent of the total LTC workforce: care homes with nursing (19 per cent), care homes without nursing (18 per cent), and domiciliary care (42 per cent) (Skills for Care 2016a). After excluding establishments with incomplete data, our sample included 4,461 residential care and 1,620 domiciliary care establishments with observations for both April 2015 and October 2016.

The main variable of interest is the hourly wage for care workers aged 25 and over. We focused the wage variable on care workers, as they are the main group of staff in LTC (over 50 per cent). They were also being paid the lowest wages and, therefore, likely to be most affected by the minimum wage policy. Figure 1 presents kernel density distributions of the hourly wage of care workers aged 25 and over by service type, for April 2015 and October 2016. These illustrate that the increase in NMW generated quite a substantial compression of the wage distribution around the new NLW level in residential care, with rather limited spillover to higher wage rates. While in April 2015 around 35 per cent of residential care workers had an hourly wage at (or below) the adult NMW rate of £6.50, in October 2016 about 49 per cent of them were paid at or below the adult NLW rate of £7.20.

In the case of domiciliary care workers aged 25 and over, the wage distribution compression has also been substantial: in April 2015 around 15 per cent of them had an hourly wage at (or below) the adult NMW rate, while in October 2016 about 30 per cent had hourly wages at (or below) the adult NLW. It is also interesting to note that while the April 2015 wage distribution for domiciliary care employees had the highest peak at £7.00 per hour (i.e. £0.50 above the NMW rate at that time), the October 2016 distribution peaked at NLW level. This could indicate that domiciliary care providers paid wages above the minimum to compensate for travel time, but since the introduction of the NLW, due to increased operation costs, the wage premium declined. Nonetheless, Figure 1 reveals some spillover to higher hourly wages of domiciliary care employees, mainly around £7.50 and £8.00.

Table 1 presents differences between April 2015 and October 2016 for average values of the outcomes of interest by service type and sector. We can observe that in April 2015 the lowest average hourly wages for care workers were paid by for‐profit care homes with nursing (£6.80) and without nursing (£6.99). Hourly wages were higher in voluntary (i.e. not‐for‐profit) establishments (£7.40 to £7.52, depending on care type) and in domiciliary care (£7.42 to £7.52). Average care worker hourly wages have increased significantly between April 2015 and October 2016, and increased asymmetrically. The highest increase was in establishments with the lowest initial hourly wages: private care homes with nursing (increase of 48p or about 7.0 per cent) and private care homes without nursing (increase of 41p or 5.9 per cent). The minimum wage increase led, therefore, not only to a general compression in the wage distribution (Figure 1), but to a narrowing in wage differentials by service type and sector as well.

Care establishments are also quite heterogeneous in terms of the other outcomes of interest considered. Establishments with the highest staff levels are in domiciliary care, with an average of 66 to 71 staff (depending on sector and period), while the lowest are in care homes without nursing (27 to 28 staff). With the exception of employment in private care homes without nursing, where employment slightly increased (by about 1.2 staff, p‐value 0.056), the changes in employment between April 2015 and October 2016 were not statistically significant.

Employment can be substantially affected by the demand for services. We, therefore, assessed the minimum wage effects on staff per service user ratios as well. The highest levels were in care homes without nursing (1.7 to 2.1 staff per service user), compared to 1.5 to 2.0 in care homes with nursing and 1.0 to 1.5 in domiciliary care. The changes over the analysed period were not statistically significant for any employer group.

Many staff work only part time or only occasionally, in particular staff on ZHC. We also assessed, therefore, the impact of minimum wage on total weekly hours. Hours were captured in ASC‐WDS in two variables: ‘contracted hours’ and ‘additional hours (during the previous week)’, with hours for employees on ZHCs reported mainly as ‘additional hours’. About 80 per cent of employees on ZHCs had both zero ‘contracted’ and ‘additional’ hours reported. This might be due to a large variation in weekly hours worked over time for workers with no guaranteed working hours; see also Datta et al. (2019) and Giupponi (2018). For all employer groups, the average value of total weekly hours (i.e. ‘contracted’ and ‘additional’ hours) did not change significantly over time. The only exception was private care homes with nursing, for which weekly hours slightly increased (about 7 per cent, p‐value 0.066).

Despite public perception, the share of care workers on ZHCs is not high in all LTC settings. The share was rather low in care homes, with an average of only between 1.7 and 3.2 per cent in voluntary and 4.7 to 5.8 per cent in private establishments. However, ZHCs were more established in domiciliary care, with an average share of 57 to 59 per cent in private and 30 to 31 per cent in voluntary establishments. For three of the six employer groups, the share of ZHCs decreased in the period analysed, most probably due to the substantial criticism the use of ZHCs received in the media and government attempts to limit employers abuse of such contracts (e.g. ban of exclusivity clauses) (Department for Business, Innovation and Skills 2015). None of the changes were, however, statistically significant.

Characteristics of the LTC establishments included in the sample are presented in Table 2. The share of establishments with at least one care worker paid in April 2015 less than the April 2016 NLW was high: about 95 per cent of care homes with nursing, 82 per cent of care homes without nursing, and 69 per cent of domiciliary care establishments. Care homes with nursing also had the highest share of care workers aged 25 and over paid less than the April 2016 NLW (79 per cent) and the highest average wage gap (i.e. the relative increase in wages needed to bring workers being paid less than the future April 2016 NLW up to the that mandated minimum; about 7.6 per cent), compared to 63 and 5.8 per cent, respectively, for care homes without nursing and 48 and 3.1 per cent, respectively, for domiciliary care providers.

In terms of other characteristics, the mean worker age for all establishments at April 2015 was 42, with about 80 to 86 per cent of staff being female, and the majority of establishments being located in an urban area (i.e. 85 to 90 per cent). There are some characteristics that vary considerably between the three service types. For example, care workers make only about 50 per cent of employees in care homes with nursing, but 61 per cent of employees in care homes without nursing, and 75 per cent of staff in domiciliary care. Moreover, the majority of care homes with nursing were medium or large (i.e. had over 50 employees; 57 per cent), while the majority of care homes without nursing and domiciliary care establishments were small (i.e. 10 to 49 workers; 77 and 49 per cent, respectively). In terms of service user type, the majority of care homes with nursing specialized in looking after older people (i.e. aged 65 and over; 55 per cent), while most care homes without nursing were supporting young adults (i.e. aged 18 and 64; about 45 per cent) and the majority of domiciliary care providers looked after service users of all ages. In terms of sectoral split, only about 10 per cent of care homes with nursing were voluntary (i.e. not‐for‐profit), compared to 24 per cent of care homes without nursing and 16 per cent of domiciliary care establishments.

6 Empirical strategy

Following Machin et al. (2003), we used two measures for the importance of minimum wage policy for individual establishments: (a) the share of care workers employed by establishment being paid less than the future minimum wage rate; and b) the ‘wage gap’, that is, the relative increase in labour cost for establishment needed to bring care workers being paid less than the future minimum up to the minimum, or:

(1)

where denotes the hourly wage earned by care worker in establishment , and stands for the future minimum wage rate. Both the share of care workers paid less than the future minimum wage rate and the wage gap equal zero for establishments with no care workers paid less than the future minimum wage.

In the empirical analysis, we aimed to establish first if, everything else being equal, establishments that were more likely to be affected by the change in minimum wage (i.e. had a higher share of care workers paid less than the future minimum wage and higher wage gap) had the highest increase in wages after the increase in the wage floor.

The wage equation has the form:

(2)

where is the change in the log of average hourly wage for establishment between (i.e. April 2015) and (i.e. October 2016), stands for one of the two measurements defined above for capturing the importance of minimum wage policy for establishment in the period , are workforce, establishment, and local area characteristics in the period that could affect wages, and is the error term.

Secondly, we looked to establish the effect of minimum wage changes on other outcomes of interest (i.e. employment, share of employees per service user, weekly hours, and the share of care workers on ZHCs). The reduced form equations used to assess the effects of minimum wage increases on these outcomes, denoted by , were similar to the wage equation (2):

(3)

As in previous studies, equations (2) and (3) are estimated using Ordinary Least Squares (OLS); see (Card 1992; Card, Krueger 1994; Datta et al. 2019; Giupponi 2018; Giupponi and Machin 2018; Machin et al. 2003). The set of covariates included to account for the heterogeneity in care establishments are: the mean age of employees, the share of female employees, the share of care workers in total staff (i.e. skill mix), the establishment size (by number of employees), the main service user type (i.e. young adults, older people, and mixed), sector (i.e. private [for‐profit] or voluntary [not‐for‐profit]), group/chain ownership, the unemployment rate at local authority level, the log of tariff paid by the local authority for a week of residential care, the log of tariff paid by the local authority for an hour of domiciliary care the urban/rural status at postcode district level, the log of the geometric mean of house prices at postcode district level as well as dummies for the nine regions of England.

There are 152 upper tier local authorities in England, responsible for commissioning publicly funded LTC (i.e. about 50 per cent of the total LTC provision) and shaping their local LTC market. The average weekly pay for residential care and hourly pay for domiciliary care at local authority level should capture the local public LTC policy and funding. The local area house prices were included to control for the geographic variation in wealth, which may affect the demand for privately funded LTC and the possibility for care establishments to adjust their income by increasing the fees for their self‐funded clients. To capture other potentially unobserved local area characteristics, we also estimated models in which we replaced the above local area controls with dummies for the 326 local authority districts.

7 Results

The identifying assumption for and is that in the absence of the minimum wage increase there would be no relationship between the initial wage level () and the wage growth () or the change in outcomes (). In other words, the trends between establishments affected and those not affected by the minimum wage policy should be parallel in the period before the increase in minimum wage. We explored this by looking at the trends between April 2015 and October 2016 in wages, employment, weekly hours, and the share of care workers on ZHCs; see Figure 2. There are two sets of graphs, for each type of service (i.e. residential and domiciliary). Establishments were also divided in three groups: those with a zero wage gap (i.e. 15 per cent of care homes and 29 per cent of domiciliary care establishments, respectively), the bottom half of establishments with a positive wage gap (i.e. 43 per cent of care homes and 36 per cent of domiciliary care establishments), and the top half of establishments with a positive wage gap (i.e. 43 per cent of care homes and 35 per cent of domiciliary care establishments). The sample includes a balanced panel of establishments (i.e. with observations for all data points included: April 2015, October 2015, April 2016, and October 2016).

As mentioned earlier, the ASC‐WDS is a dataset continuously updated by care providers and updates are sometimes made with a few months delay. From the four data points included, we consider, therefore, that April 2015 and October 2015 would capture the period before the two increases in minimum wage (October 2015 and April 2016), while April 2016 and October 2016 would capture time points (during and) after the minimum wage increase. We believe that it is not suitable to include data points before April 2015, as these would capture adjustments following the previous (i.e. October 2014) minimum wage increase.

Average wages did not change much between April and October 2015 (which shows that only few care providers updated their records without delay after the October 2015 NMW increase), but they increase afterwards and continue to increase even after April 2016 (the introduction of the NLW), as more establishments updated their worker records. The graphs illustrate that, as expected, the average wage increased relatively more for the establishments in the top half of the ‘wage gap’ (i.e. establishment paying lower wages), while it changed rather little for establishments with a zero ‘wage gap’. Nonetheless, for all groups of establishments the total employment level and total weekly hours were quite stable over the period observed.

There does not seem to be evidence of non‐parallel trends in the share of staff on ZHCs either. The share was relatively stable between April and October 2015 for residential care establishments and slightly decreasing for domiciliary care establishments. However, after April 2015 the share of staff on ZHCs decreased for residential care establishments with a zero ‘wage gap’, while it remained stable for care homes in the bottom half of the wage gap and even slightly increased for care homes in the top half of the ‘wage gap’. For domiciliary care establishments in all three groups, the downward trend in the share of staff on ZHCs continued over the whole observed period.

Ordinary Last Squares estimation results for residential care establishments are presented in Table 3, while those for domiciliary care are presented in Table 4. For brevity, we present only coefficients for the two variables capturing the importance of minimum wage policy for individual establishments, that is, the share of workers paid less than the April 2016 NLW and the wage gap as well as marginal (i.e. slope) effects of interactions of these variables with sector and service type. Robust standard errors are computed to account for heteroscedasticity. We also estimated difference‐in‐difference models with establishments with no care workers paid below the future minimum wage at as control group (i.e. = 0 if and = 1 if ). The results were quite similar to those using the share of workers paid less than the April 2016 NLW and were not reported. The full set of estimation results is available from the authors upon request.

Both the initial share of care workers paid less than the April 2016 NLW and the wage gap had a strong positive effect on wage growth for establishments in all sectors and all care types (columns (1) and (2) in Table 3 and 4), showing that establishments with the highest likelihood to be affected by the increase in the minimum wage floor were indeed those who experienced the largest wage growth. For example, a residential care establishment with 67 per cent of its care workers being paid in April 2015 less than the April 2016 NLW (i.e. an average residential care establishment) experienced an average wage growth of about 3.3 per cent higher than one that had 10 per cent of its care workers paid less than the future minimum wage.²2 The calculation is 0.058 × (0.67 − 0.1) = 0.033.
This is sizable given that the average wage growth for care workers in residential care was about 5.8 per cent between April 2015 and October 2016, and is comparable with the estimated effect of minimum wage increases on wage growth in care homes found in previous studies (Datta et al. 2019; Machin et al. 2003).

Although the share of care workers paid less than the April 2016 NLW and the wage gap was on average highest in care homes with nursing, followed by care homes without nursing, and domiciliary care establishments (see Table 1), we found that the size of the minimum wage effect by service type was in reverse to this order. This shows that domiciliary care providers increased wages relatively more than they had to, potentially to keep competitive in the labour market for care workers. As already mentioned, care workers in domiciliary care traditionally have somewhat higher wages, most probably to compensate (at least partly) for travel time between service users’ homes. The size of the minimum wage effect was higher for private than for voluntary care homes, showing that the initial wage differential (i.e. lower average wage in private care homes) became smaller with the introduction of the NLW.

Similar to many previous studies on the effects of minimum wage policy — for an overview see Manning (2016) — we find no significant effects on either employment or the share of staff per service users. This is not really surprising. Due to cuts in publicly funded social care, providers were likely to have used staffing rather efficiently to begin with. Decreasing staffing further may have affected the service users’ safety.

However, similar to Machin et al. (2003), we did find a significant negative effect of minimum wage policy on total weekly hours in care homes. The average elasticity for a minimum wage increase from £6.50 to £7.20 was −0.67 when calculated using coefficients from estimations with the share of care workers paid less than the future minimum wage and −1.02 when using the coefficients from the estimation controlling for the wage gap.³3 The average elasticity for a minimum wage increase from £6.50 to £7.20 was calculated as .
We also found that the effect on hours differed significantly by sector and service type, and was significant only for private care homes and care homes without nursing.

The minimum wage effects on hours were quite large when compared to findings in other minimum wage studies, maybe due to the substantial impact of the minimum wage increase on wages and the wage distribution. The effect on hours was also rather surprising in the context of the LTC industry, as staffing levels are regularly monitored by the CQC to ensure that providers employ sufficient staff with the right skills for their service users’ care and treatment needs. One possible explanation is that the number of weekly hours employed might not be as easy to monitor compared to the number of staff employed.

With respect to ZHCs, we found a significant positive effect only for care homes, and only in the specification with the share of staff paid less than the April 2016 NLW. Moreover, when differentiating by sector and service type we found slightly higher positive effects on ZHCs for care homes in the voluntary sector and care homes with nursing. However, these were significant only at 10 per cent level. As the overall trend in employment on ZHCs in residential care was negative over the analysed period, these results show that the increase in minimum wage slowed the downward trend in the share of employment on ZHCs.

We also estimated specifications in which we replaced the local area controls with dummies for the 326 local authority districts in England.⁴4 The estimation results are available from the authors upon request.
While in this specification the effects on wage growth and weekly hours were still strong and significant, the effects on ZHCs were not significantly different from zero. Our results on ZHCs differ from those of arecent study using the same dataset (i.e. ASC‐WDS) that found a positive effect on ZHCs for domiciliary care establishments (Datta et al. 2019). The difference in results between the two studies might be due to the slightly different period analysed and the somewhat different sample size.

To assess the start and duration of the effects on wage growth, employment, hours, and ZHC, we also estimate models in which we consider the ‘after’ time point () at 6, 12, and 24 months from April 2015, in addition to 18 months (i.e. October 2016) in the main analysis. The estimated effects are presented in the Appendix: Table A1 for residential care and Table A2 for domiciliary care. The results show that the wage growth for both residential and domiciliary care establishments was small at the time of the NMW increase (October 2015); at least partly due to the fact that many providers had not updated their establishment and worker records at that time. The effect on wage growth increased in April 2016 (introduction of NLW), increased further in October 2016 (after most providers updated their records), and was rather stable afterwards.

In care homes, there seems to have been a short‐term negative effect on employment (in October 2015 and April 2016; both in terms of total employment and share of employees per service users) and a long‐term negative effect on weekly hours (i.e. over the whole 24 months). The early onset (October 2015) of these two effects raises the question on whether they are due to the LTC providers’ anticipation of the minimum wage increase or other reasons not related to the minimum wage policy. However, as there was no other economic shock or market intervention during that time, we believe these to be rather minimum wage effects. As mentioned earlier, we believe the reason for the short‐term effect on total employment versus long‐term effect on hours to be due to the likely difficulty of monitoring working hours by the CQC. The absence of both the employment and hour effects in Giupponi and Machin (2018) is probably because at the start time considered in their analysis (March 2016), the effect on total employment had already become weaker and turned insignificant, and the effect on hours had already occurred and did not change much later on.

The positive effect on ZHC found for care homes did occur only after the increase in the minimum wage floor (October 2016), but seems to have been rather short‐term, as in April 2017 it was not significantly different from zero. We note in Table 4 that there has been a short‐term positive effect on ZHCs in domiciliary care as well. It took place in October 2015 and, therefore, in anticipation of the minimum wage increase. However, by October 2016, it was no longer significant.

8 Conclusions

The introduction of the NLW in April 2016 at £7.20 was a major policy intervention in the labour market, with the aim of increasing the minimum wage for workers aged 25 and over to 60 per cent of median pay. Combined with the increase in NMW from £6.50 to £6.70 only six months earlier (October 2015), it generated a significant increase in the wage floor and has prompted concerns about potential negative effects on employment and sustainability, in particular in labour intensive and low‐pay sectors such as LTC.

The results of our analysis confirm the findings of previous studies which show that despite the sizable effects on wages, the minimum wage policy had only a short‐term effect on total employment. Nevertheless, we were able to find evidence that care homes which were more affected by the minimum wage policy — in particular private (i.e. for‐profit) care homes and care homes without nursing— reduced weekly working hours. This may not be surprising, as the number of staff employed is regularly monitored by the CQC and may be more difficult to reduce compared to hours. According to Regulation 18 of the Health and Social Care Act 2008 (Regulated Activities) Regulations 2014, providers must employ ‘sufficient numbers of suitably qualified, competent, skilled and experienced staff’ to make sure that they can meet people's care and treatment needs.

We found no effects on employment and hours for domiciliary care establishments. However, we found some evidence that for both residential and domiciliary care establishments the minimum wage increase had a positive effect on the share of staff employed on contracts without guaranteed working hours (i.e. ZHC), probably to keep labour input more flexible to meet fluctuations in demand. This was, however, only a short‐term effect.

The ‘before and after’ analysis employed in this study has its limitations and the results should be interpreted with care. As shown by our analysis as well as the differences in findings to other recent studies using the same dataset (i.e. ASC‐WDS), the results can be affected by the time points chosen ‘before’ and ‘after’ the policy intervention.

As NLW rates will apply to younger workers in the coming years (i.e. the age limit for the NLW will be reduced from April 2021 to 23 years and over and from April 2024 to 21 years and over) and the NLW rate is expected to increase further to two thirds of median wages by 2024 (Department of Business, Energy and Industrial Strategy 2020), the effects of minimum wage policy on the LTC sector will remain important for future analysis.