3 What the Existing Research Literature Tells Us
3.1 In the United Kingdom (UK), the Living Wage has been promoted by the Living Wage Foundation and others, and adopted by local authorities, universities and numerous other public and private bodies, including the Scottish Government for its own employees. The Living Wage in various forms has also been promoted in other countries, including the United States (US) and New Zealand. The increased interest in adopting a Living Wage has been accompanied by a need to understand its costs, benefits, and implications, including its impact on public budgets, labour demand, Living Wage recipients and society in general.
3.2 In investigating such effects, we need from the outset to be clear about what we mean by a Living Wage. In the UK today it is principally understood to mean the voluntary adoption by public and private employers of a wage floor based on what is needed for people to meet minimum living costs. However, in the US the term can also refer to a standard that has some form of statutory backing and/or is not purely voluntary. In the US, for example, some cities have adopted Living Wage Ordinances that impose a higher compulsory minimum wage than required by state laws, mainly on employers receiving public funding through contracts or grants. At the same time, a number of states as well as some cities set minimum wages for all employees substantially above the federally required minimum, and arguments about living standards can influence this too. As a consequence, the US provides by far the best test-bed and solid long-term evidence of the effect of compulsory wage floors at different levels. The overall effects of voluntary standards remain much harder to research.
3.3 Individual studies of the effects (on companies and the labour market) of adopting wage floors provide varied and, to some extent, contradictory findings. This is partly because of the range of contexts in which these wage floors have been implemented, and may also be influenced by bias related to the controversial nature of the subject, which is influenced by political and economic beliefs. This has led to accusations of data mishandling and of misinterpretation of evidence between authors. Nevertheless, studies that have reviewed the evidence as a whole, based on a range of research, have come to reasonably clear-cut conclusions.
3.4 This literature review aims to assess what existing evidence can tell us about the introduction and extension of the Living Wage in Scotland, based mainly on evidence from the UK and the US. The fact that studies have found varying impacts in different contexts and that the experience of the Living Wage in the UK, especially outside London, is relatively new and limited means that there is no fully reliable way to predict the effects of a widespread implementation of the Living Wage in parts of the United Kingdom.
3.5 However, the evidence on wage floors found elsewhere, especially in the US, does give some strong indications of their impacts on:
1) labour demand and employment levels
2) wage costs, and ways in which employers absorb any increases
3) employees' wellbeing, income and poverty rates
3.6 A fourth important impact is on the cost and level of public services, where public bodies and their contractors raise wages at the bottom. These effects are harder to project from research, but experiences of bodies who have introduced the Living Wage in the UK are nonetheless relevant.
The effect on labour demand and employment levels
3.7 The debate on whether raising the minimum wage is detrimental for the economy is not new. In the US, academic and political discussions on the minimum wage go back a century (Kaufman, 2010; Neumark and Wascher, 2007). Classical liberal and neo-liberal theories argue that imposing wage floors leads to a reduction in employment rates due to the increased cost of labour. In the United Kingdom, this has produced a cautious approach to the setting of a national minimum wage (NMW). Throughout the 15 years that the Low Pay Commission has supervised the uprating of the UK Minimum Wage, it has found no evidence of significant negative employment effects (e.g. Draca et al, 2011). However, this may be because the NMW has never been at a level high enough to affect labour demand, and that there is bound to be some higher level at which it would do so. In the United States, the setting of minimum wages at up to 50% above the federal level of $7.25 ($10.74 in San Francisco) offers much greater scope for researchers to test whether a relatively high wage floor is associated with less employment.
Have higher wage floors reduced employment levels in the US?
3.8 Two decades of intensive research comparing employment levels under different minimum wages in the US has failed to confirm the hypothesis that higher minimum wages reduce employment. On the one hand, Card and Kreueger's (1994) influential study of the effect of minimum wage increases in fast food restaurants in New Jersey found no discernible employment effect. Subsequent research in the US (Fairris, 2007; Fairris and Fernandez, 2008; Reich, Hall, and Jacobs, 2005) has continued to show no consistent evidence of changes in labour demand as a result of increasing wages. Recently, Laura D´Andrea Tyson, who served as chairwoman of the Council of Economic Advisers under President Clinton, reflected that despite the increase in the minimum wage implemented during Clinton´s administration, "a higher minimum wage did not impede robust employment growth; it did contribute to healthy income gains for low-wage workers" (D´Andrea, 2013).
3.9 On the other hand, some studies show significant negative employment effects in certain cases. For example, Neumark and Adams (2003) found reductions in employment associated with programmes that set wage floors for recipients of public business assistance. Also, Tolley, Bernstein, and Lesage (1999) estimated that a proposed Living Wage Ordinance in Chicago would cost over 1,300 jobs - based on a survey of public sector contractors to whom it would have applied.
3.10 One approach to summing up the effects of various studies has been to carry out "meta-studies" of employment effects identified by a large range of individual studies (Allegretto et al, 2013; Doucouliagos and Stanley 2009; Giuliano, 2009; Lester and Jacobs, 2010). In general, these studies have found there is no consistent relationship between wage floors and employment levels (i.e. there are considerable differences in individual studies' findings of whether employment effects are positive or negative), but the key finding is that the vast majority of individual studies show the effects to be very small.
3.11 Some authors have argued that traditional approaches have failed to account for ways in which employers may not behave in manners predicted by competitive models and find channels of adjustment (Brenner, 2004; Schmitt, 2013; Thompson and Chapman, 2006) or contextual differences in employment growth (Dube, Leicester, and Reich 2010).
Would the Living Wage reduce employment levels in the United Kingdom if implemented generally?
3.12 Studies in the United Kingdom have pointed out that the association between higher wages and employment often contradicts the theoretical expectations of classical economic thinking (Bryan, Salvatori and Taylor, 2012; Dolton, Rosazza and Wadsworth, 2012; Stewart, 2004). Metcalf (2008) found that the introduction of the national minimum wage (NMW) is not associated with employment trends during the last two decades and exposed twelve possible reasons for small or non-existent employment effects of the NMW in the UK.
3.13 A number of studies have sought to project what would happen to employment levels in the UK if the current Living Wage were generalised, either through some form of compulsion or through very widespread take-up and possibly a situation where larger companies feel obliged to implement it for reputational purposes. At present, none of the evidence on the Living Wage as implemented, including in London over the period of a decade is able to address that question, because it has not been sufficiently generalised.
3.14 Lawton and Pennycook (2013) sought to project "the likely economic and social impact of more extensive Living Wage coverage". One aspect of this was to model the impact on labour demand of a situation where the present Living Wage levels became compulsory. It estimated that "labour demand" would decline by 80,000 net, and by 160,000 for young people (some of whose jobs could go to older more experienced workers). This theoretical calculation is based on previously known relationships between pay and employment levels. It is important to emphasise that this is not a prediction of how many jobs would actually be lost; rather, it is based on observed historical relationships between pay and employment levels, which may not apply to the unknown context of increasing significantly the wages paid in all lowest-paid jobs.
3.15 Lawton and Pennycook's (2013) analysis is also helpful in setting out which job sectors might be most at risk, which is largely based on the distribution of jobs where pay is well below the Living Wage. A move to the Living Wage would increase the wage bill most for bars and restaurants (6.2%), general retailers (4.9%) and food and drug retailers (4.7%). In most other sectors the effect on the overall wage bill would be very low or negligible, because only a small minority of jobs would be affected and/or because those affected pay not much below the Living Wage at present. It is also notable that workers under 30 - and particularly those under 20 and young workers with low skills - have a much higher than average chance of being on low pay, and are therefore more likely to be beneficiaries of a Living Wage, but also face some risk that they will be replaced by more experienced or better qualified workers. Finally a regional breakdown shows that the risk of being below the Living Wage is somewhat lower in Scotland than on average in the UK - at 18%, lower than in any other UK region except London and the South East.
3.16 Reed (2013) has attempted to make a comprehensive estimate of effects of expanding the Living Wage to all employees in the United Kingdom, taking account of macroeconomic knock-on effects. Arguing that the net fiscal benefits derived from implementing the Living Wage universally would be of £1.5 billion, which, through a multiplier effect, would increase GDP by £3.15 billion, he estimates employment effects of between 45,000 jobs lost and 58,000 jobs gained. This can only be seen as a very broad estimate of what would actually happen, but shows the degree to which the uncertainty but generally low level of projected employment effects can make it hard even to know whether there would be a small net reduction or small net increase in employment.
3.17 A report for the Welsh Assembly Government (Marsh et al, 2010) estimated the consequences if Neumark and Adams' (2003) estimate of an employment elasticity of -0.12 were to apply to introducing a Living Wage to 9,000 low-paid workers in public administration in Wales. This estimate means that a 10% wage increase is associated with a 1% cut in employment, and on this basis it was estimated that job losses would be of the order of 300. However, it is by no means clear that an estimate applying to contracted and publicly-aided employers in the US is valid for predicting the employment outcomes for public administration in Wales (or, indeed, Scotland).
3.18 Local reports, such as Holden and Raikes (2012) for Manchester, have to some extent drawn on the evidence from London and the United States and concluded that there is unlikely to be a serious impact on jobs. However, these studies have largely considered the Living Wage in terms of a voluntary initiative, in which firms have to make their own judgements about whether and how to implement it, and so have not considered in detail possible economic impacts of a mandatory increase in the wage floor.
The effect on wage costs, and ways in which employers absorb any increases
3.19 There is considerable literature in both the US and among Living Wage employers in the UK detailing how employers have adapted to higher wage floors. This evidence helps explain why it has been hard to discern significant reductions in employment levels. In general, other changes create a context where the economic principle that "other things being equal, higher minimum wages means fewer jobs" does not apply, for the simple reason that other things are not equal.
3.20 These other potential changes, each of which have been documented in US studies (Schmitt, 2013), are:
- Employers reducing labour costs in other ways, such as by reducing hours, non-wage benefits, or training; or by cutting the wages of higher-wage workers
- Employers suffering reduced profits as they absorb the costs of the Living Wage, cutting returns to their owners
- Employers passing the costs of a higher pay bill on to consumers in the form of higher prices
- Productivity increasing as a result of a higher minimum wage, for a variety of reasons e.g. reduced employee turnover (resulting in lower recruitment and training costs) and increased employee effort, or as a result of employer action (such as increasing training or shifting the composition toward higher skilled workers).
3.21 There is little consensus in the economics literature about which of these effects dominates in practice. Most of the real world changes that have occurred because of minimum wage rates have been quite small, which makes their effects difficult to measure against a background of constantly changing macroeconomic conditions and long-term trends in the regional and industrial makeup of the economy (Northern Ireland Council for Voluntary Action, 2014).
3.22 A study of the effects of the implementation of the National Minimum Wage found that there were no effects on productivity or on employment levels, but some reduction in profits (Draca et al, 2011, pages 130, 149). However, the OECD has concluded that an increase in the ratio of the minimum wage to the median wage by ten percentage points can boost productivity by almost two percentage points (OECD, 2007). And Schmitt's (2013) review of US evidence concludes that:
"…the cost shock of the minimum wage is small relative to most firms' overall costs and only modest relative to the wages paid to low-wage workers … probably the most important channel of adjustment is through reductions in labour turnover, which yield significant cost savings to employers." (p23)
3.23 Evidence in the UK also highlights staff turnover as the single most widely-cited change associated with introducing the Living Wage. Wills and Linneker (2012) found that in a sample of London employers taking up the Living Wage, turnover had reduced by an average of 25%, causing a 0.3% reduction in overall wage costs. London Economics (2009) came to a similar conclusion, and also noted a range of other effects: reduced absenteeism and sickness, increased worker effort and productivity and reputational benefits. While some issues had arisen about wage structures and differentials, the study found that there were no reported lasting disadvantages from a business perspective. Wayland (2011) refers to a study conducted by Barclays Bank in 2004, where absenteeism was reported to reduce from 30% to 4% since the introduction of the Living Wage.
3.24 It is important to note that these London studies are based on small samples and must be regarded as indicative. Moreover, results may depend on how and to whom questions are asked. The London Chamber of Commerce & Industry (2008), asked 275 business representatives about the London Living Wage and 42% of respondents said they would consider job cuts, 49% could cut back on investment, and 26% reduce their training budgets. However, this reports the views of employers not volunteering for a voluntary standard, and does not report on actual behaviour.
3.25 Overall, the US and UK evidence suggests that firms' experiences of costs and benefits of the Living Wage, and the ways in which costs are absorbed, are highly diverse. Although increased productivity is likely to outweigh the higher wage cost for many firms, this does not mean that all firms will experience increased productivity, or that it will fully cover the cost for all those that do experience it.
The effect on employees' wellbeing, income and poverty rates
3.26 Studies of minimum wages and of Living Wages find, not surprisingly, that they increase the earnings of the employees affected. The amount obviously varies according to the size of change in the wage floor. Importantly, the impact on net incomes also varies according to the rate at which increased taxes and reduced in-work benefits result from higher individual earnings, as well as the extent to which low individual earners are located in households with low overall incomes. This makes it impossible to generalise effects on poverty levels from one context to another.
3.27 A number of US studies suggest that, despite the fact that low pay is not the same as poverty, raising minimum pay can have significant effects in reducing household poverty and improving outcomes such as health. For example, the Living Wage Ordinance in San Francisco has been associated with less sickness absence, improved subjective health, reduced premature death rates (Bhatia and Katz, 2001) and a 15% reduction in poverty rates (Howes, 2002). Similarly, Neumark, Thompson, and Koyle (2012) found that implementing business assistance Living Wage laws is associated with a reduction in the poverty rate by 2.4% in twenty-six American cities.
3.28 In the United Kingdom context, gains from Living Wages are bound to be shared between the recipient and the Exchequer, particularly in the case of families on tax credits (who typically lose 73% of additional earnings in reduced tax credits and increased taxation). Partly as a result of this, some UK authors have urged caution about the size of the direct impact of the Living Wage on poverty. Holden and Raikes (2012) argue that it would be a mistake to rely on the Living Wage on its own to address in-work poverty, and that it is essential to combine it with other policies designed to improve career progression and to reduce the costs faced by people on low incomes, for example in paying for housing, transport and childcare (which could potentially be funded from the gains to the Exchequer of a Living Wage). In addition, changes to benefit withdrawal and tax settings could ensure that lower income households keep a greater proportion of the increases in earned income.
3.29 The ways in which employers absorb any increases in wage costs are also likely to influence the impact of increased hourly wages on employees. For example, as noted earlier, some employers respond to higher wage bills by reducing hours, non-wage benefits, or training (Schmitt, 2013).
3.30 Nevertheless, other authors have pointed to tangible ways in which those who receive the Living Wage in the UK experience a higher living standard than those who do not. Flint, Cummins, and Wills (2013) found that those who received the London Living Wage scored 3.9 units higher on a 70-point well-being scale than those who did not, controlling for other socio-economic and demographic characteristics. The introduction of the London Living Wage has also been associated with higher commitment and better working atmosphere (Sokol et al, 2006) and with higher satisfaction at the workplace (Wills, Kakpo, and Begum, 2009).
The amount that a Living Wage raises the direct pay-bill of public bodies
3.31 It has already become evident in the United Kingdom that the direct effects of public sector bodies adopting the Living Wage for their own staff is very small. This depends on the extent to which, before the policy is implemented, public employees are earning below the Living Wage. In general, only a relatively small minority have such low earnings and often they are not too far below the Living Wage.
3.32 A Freedom of Information request in 2011 by the Poverty Alliance produced a list of the numbers below the Living Wage in each Scottish local authority at that time. On average this represented 7% of local authority direct employment, with a range from zero to 15% (Poverty Alliance, 2011). As an illustration, a local authority paying say 7% of its employees an average of £1 below the Living Wage would have to increase its overall wage bill by well under 1% in order to lift their wages to that level. In practice, local authorities deciding to pay the Living Wage to all staff have experienced only very minor increases in total wage costs. For example, in Glasgow it is an estimated £1.2 million (Marsh et al, 2011, p 58) and in Birmingham £1.3 million (Birmingham City Council 2012); these local authorities have total budgets of over £2 billion and over £3 billion respectively. Similarly, Elmore´s (2003) study in twenty American cities found that the actual cost was less than 0.1% of the total city budget (see Table 2 in Part 2, section B below).
3.33 An important reason for this situation is that many of the lowest paying jobs funded by the public sector, such as cleaning and care staff, are not carried out by directly employed personnel. This makes the issue of the Living Wage for staff in contracted services important both in terms of the benefits of the Living Wage to staff in public services and the potential cost to the public purse.
The scope for introducing the Living Wage among contractors to the public sector
3.34 Setting wage floors for directly-employed staff only can potentially create a perverse incentive to contract out to cheaper suppliers who do not pay the Living Wage. At present, the Living Wage Foundation recommends that Living Wage employers should set achievable milestones towards ensuring that their suppliers also pay their staff a Living Wage.
3.35 While the US has widespread experience of using city ordinances to enforce a Living Wage among all public contractors (and in some cases all bodies receiving any public grants), the situation is different for public bodies in the UK. If a public sector body in Scotland (or the rest of the UK) made the payment of staff at a higher rate than the National Minimum Wage a mandatory requirement as part of a competitive procurement process, this would run the risk of breaching European Law and European Procurement rules. It is, however, possible to include Living Wage considerations in procurement exercises without making payment of the Living Wage a mandatory requirement, and a number of public bodies in the UK have done so (Ready for Business, 2015). For example, one contracting authority asked bidders to include proposals to enhance workforce retention/cohesion/performance", and reference was made to the authority having an interest in the promotion of equality and measures towards ensuring good working conditions for those performing the contract. The successful bidder included a commitment to paying the Living Wage within their response to this question. In Lewisham Council, as each contract comes up for renewal or extension, the corporate procurement team ascertains whether staff involved in providing the services are paid above or below the Living Wage. If staff are paid below the Living Wage then the tender documents include two pricing schedules for the tenderer to complete: one which included paying all staff the Living Wage and one without. Evaluation of the bids is carried out and an award recommendation produced, identifying the proposed service provider and also whether the Living Wage should be implemented (the actual cost can be seen by the difference in the two price schedules) (Public Health England, 2014).
3.36 The approach taken in the Procurement Reform (Scotland) Act 2014 will see Scottish Ministers publish statutory guidance for public bodies on how workforce related matters, which can include such matters as a company's approach to recruitment, remuneration, including Living Wage and terms of engagement, should be considered when selecting bidders and awarding public contracts, where such matters will affect the quality of service that the bidder may provide.
3.37 The Act, once provisions are commenced, will see those public bodies required to prepare procurement strategies set out (among a range of other priorities) what their general policy is on payment of a Living Wage to persons working on their contracts.
3.38 A sector where workforce terms and conditions are key to the delivery of quality public services, and where it is particularly challenging to encourage suppliers to pay a Living Wage as part of a procurement exercise, is social care. This is a sector with large numbers of low-paid staff, where there has been a severe squeeze on public budgets, and where the relationships between local authority funding and delivery is complicated by factors such as personal budgets and purchase of residential provision to out-of-area suppliers who may have charges applying to the residents of several local authorities. For example, Birmingham City Council estimated potential increased costs of circa £11m per annum if the Living Wage was to be applied to its social care contracts (not including personal budgets) (Birmingham City Council, undated). In Lewisham, legal advice was that to include the Living Wage in its residential and nursing care homes contract, Lewisham Council would have been required to purchase all available rooms. The council only block-books a proportion of rooms or spot purchases individual units and the legal advice was that by paying the Living Wage in these circumstances it would be subsidising self-funders and placements by other councils, which is beyond the council's powers (Public Health England, 2014).
3.39 In the US (which, unlike the UK, is able to use city ordinances to enforce a Living Wage among public contractors), evidence on additional contracting costs suggests at most very modest effects. Brenner (2004, p 22) reported on the basis of a range of local studies that "although cities have had a wide range of experiences with Living Wage laws, the preponderance of evidence indicates that Living Wage ordinances are unlikely to cause large increases in city contract costs". The bidding process, these studies find, can be influenced in a variety of ways. In some cases, it may narrow the field of bidders. In some, it replaces competition on the basis of undercutting on price by paying low wages with competition based more on quality. Importantly, many bidders welcome the "level playing field" provided by a wage floor, and in some cases this actually attracts more bidders. Overall, Brenner's review found that Living Wage ordinances have not produced an observable increase in public sector costs, but are compatible with a competitive bidding process in which profit margins are trimmed. In addition, Adams and Neumark (2005) found that in order to mitigate the costs of complying with Living Wage laws in the US, some employers have reallocated their higher-wage labour to the contract work and their lower wage labour to the non-contract work - producing better quality for the taxpayer if not always an overall improvement for employees.
3.40 Where the additional contracting costs have been quantified they have often been found to be low. Studies by Weisbrot and Sforza-Roderick (1996) and Niedt et al (1999) of Baltimore's Living Wage indicate that even though the cost of contracting increased due to increased labour costs, inflation was higher during that period. The former study found that nominal costs of contracting increased by 0.2%, but that after adjusting for inflation, they declined by 2.4%. Similarly, the latter concluded that city contracting costs increased by 1.2%, but that real costs actually decreased once inflation was taken into consideration.
The overall fiscal consequences and where they fall
3.41 Some UK analysts suggest that any increased public costs consequent on the Living Wage will be largely or wholly offset by other effects on the public finances, derived from increased income tax and national insurance contributions and reduced benefits and tax credit payments and a multiplier effect of additional disposable income spent by low-income groups. Reed (2013) estimates that expanding the Living Wage to 572,000 workers in the United Kingdom would result in a net increase of around £1.5 billion in the public finances. Wills and Linneker (2012) calculate that by paying the London Living Wage to those who are currently being paid lower rates (around 580,000 workers in London), the Exchequer would save around £823 million per year. This is because the overall gains in reduced benefits and tax credits and higher tax revenues from public and private employers switching to the Living Wage would be greater than the additional wage bill for just the public employers who adopt it.
3.42 Evidence in the US also suggests that additional public costs will be offset by fiscal benefits, but evidence on the overall net effect varies. Studies like Pollin, Brenner, and Wicks-Lim (2004) estimated net fiscal savings of around $3.4 million per annum, resulting from additional $12.6 million in wages and implementation of the law and savings of $16 million due to reduced Medicaid expenditures and increases in tax revenues. One important difference between the US and the UK in these calculations is that the American cities may require their contractors to pay all their staff, not just those working on public contracts, a higher wage. The fiscal impact of implementing the Living Wage universally or only through public contracting would be associated with a range of issues, including not only the costs of paying higher wages and administrative costs, but the tax and social benefit rules in place. In Pollin, Brenner, and Wicks-Lim's (2004) study, higher wages were associated with lower public cost in medical assistance to low income families - but as with tax credits in the UK, these savings will not all accrue directly to the local authority paying the higher contract costs. For instance, Tolley, Bernstein, and Lesage (1999) calculated that introducing the Living Wage in Chicago would cost the city around $20 million, and that of the $5 million recouped in fiscal gains, less than 20% would remain in Chicago.
Email: Alison Stout