On ShareAction’s evidence in favour of the Sainsbury’s Living Wage resolution

5 Jul

5 July 2022

ShareAction has cited several papers in support of its case that financial benefits, not costs, would arise from supporting their Living Wage resolution at Sainsbury's. Here’s why I don’t think they show that at all.

Last week I published an article outlining why I felt that shareholders should vote against ShareAction’s Living Wage resolution at Sainsbury’s. This followed Schroders’ public declaration that they would vote against the resolution. I analysed the proposal against three tests that London Business School developed in a project we undertook with The Investor Forum and a number of its members, entitled “What does stakeholder capitalism mean for investors?”.

I should say that I don’t have any personal stake in the Sainsbury’s vs ShareAction stand-off. Nor do I have any kind of connection to Schroders. My interest in the case was twofold.

First, I now devote my life to largely unpaid roles promoting the evidence-based practice of responsible business, by connecting academic research, public policy, and corporate action. I care about this stuff being done right.

Second, having developed a decision-making framework for shareholders to assess ESG issues, as part of a collaboration between London Business School and The Investor Forum, I was keen to try it out on a live case. I also said in my original article that “…all views in this article are my own and cannot be attributed to London Business School or The Investor Forum or to any of the companies or investors mentioned, and of course reasonable people can come to different conclusions on these matters.” I had intended to leave it at that – a little case study. But I now seem to have been drawn into the debate more broadly.

Based on the private messages I’ve received, a number of board members and investors found the framework helpful, as well as the illustration of its application. Nonetheless, I was surprised and, let’s be honest, flattered (or should I say “humbled”?) to see that ShareAction has published a new article by Simon Rawson, Director of Corporate Engagement, that is largely devoted to rebutting my arguments. This suggests that, at the very least, I made some points that some people think some other people (however misguided) may view as sensible.

ShareAction make a lot of points in their article, some of which I agree with. I plan to address all of these in a second blog at some point. But I want to focus here on the point that I think is most important, which relates to the evidence for whether there is a financial business case for shareholders to vote for the Sainsbury’s proposal.

The evidence that ShareAction cites, unfortunately, nowhere near supports the strength of the claims they make. I’ll explain why.

Doing well by doing good?

ShareAction make broadly two points in support of the resolution. First, it’s morally the right thing to do. Second, it is to the financial advantage of shareholders both through the positive impacts on Sainsbury’s itself from becoming a living wage employer and because it contributes to addressing the “systemic risk” of inequality.

This second strand, the financial business case, is particularly important. If there is a financial business case, then asset managers have an immediate reason to support the resolution. If there is not a financial business case, but only a moral case, then it’s more tricky. This is because asset managers have a fiduciary duty to act in their clients’ interests.

What are those interests? It’s generally agreed that most clients would, all else equal, prefer higher rather than lower financial returns. Therefore, actions by the asset manager that increase financial returns would generally be viewed as consistent with their fiduciary duty in the normal course of business. There would be no need for the asset manager, acting as a fiduciary for their client, to refer back to their wishes.

But of course, clients don’t just care about financial returns. They have non-financial preferences too. But the issue for the asset manager is that to a pursue a non-financial goal that has a cost in terms of financial returns requires them to have a very clear and explicit mandate from the client that this is how they want their money to be used. Because if there is not a financial business case for, and indeed a potential net financial cost of, Sainsbury’s becoming an accredited Living Wage Employer, then in effect clients’ money is being used to fulfil that social objective. To use their client’s money in this way without their explicit consent is a misappropriation of their property.

Most asset managers simply don’t have such consent from their clients to incur costs in pursuit of social goals (a few might, in relation to very specific funds in the social sphere). Mandates just don’t have the requisite detail (another point we made in our work with The Investor Forum). And they can’t just assume that their clients would be happy with this outcome. Some of their clients may not care about inequality at all and think it’s none of their business. Some, like me, might care about it a lot but think that using shareholder powers to impose Living Wage accreditation on individual companies is not the most effective way to go about it. Some may be all for the approach.

So by itself, the moral case is not sufficient to create a mandate for asset managers to act in line with their fiduciary duty. ShareAction knows this, which is why they are so keen to assert the financial business case for the resolution. Indeed, this question of fiduciary duty motivates a whole industry of analysis that aims to show that ESG activities of various types pay for themselves by enhancing long-term shareholder value.

Some indeed do, but some just cost money.

Bring out the big names

In my blog I wrote that:

“It is possible that all of these companies and management teams are missing an easy win-win of increasing wages to increase profits. But the onus must be on those making this case to prove it. The shareholders supporting the resolutions have notably not done so.”

I also wrote that:

“There is indeed some evidence for the negative impact of inequality on growth. However, the evidence for this is stronger in low income than high income countries, and overall the evidence is extremely mixed. So the case for it being a “systemic risk” seems contestable.”

ShareAction responded to this concern by asserting that “there is a strong company-level and system-level case for Living Wages.” They went on to cite:

“Research by the MIT Sloan School of Management, by Cambridge University, and by Cardiff Business School, which we’ve cited throughout our campaign, suggests the opposite. Business leaders themselves are recognising the case. A recent paper by the CEO-led Business Commission to Tackle Inequality, highlights how inequality undermines social cohesion, erodes trust in institutions, and fuels unrest.”

These are some big names! Surely they can’t all be wrong?

The thing is, I’ve clicked through to read the underlying papers (it’s the sort of thing I spend my time doing these days) and they really don’t make the case that ShareAction want them to. And some of them even argue against it.

Here’s why.

Evaluating the evidence

My colleague Alex Edmans, Professor of Finance at London Business School and a leading authority on the evidence-base for sustainable business, has written an excellent paper on how to evaluate research. The audience was time-poor journalists, to help them filter valid claims from those that really can’t be supported based on the evidence provided.

As well as checking whether the research actually exists (sometimes it doesn’t!) he recommends considering the following five questions:

Is the research based on actual data? Or is it based on stories and opinion?
Is it published in a top peer-reviewed journal? So that it has been through a rigorous testing and review process.
What are the credentials of the authors? Do they have relevant expertise and are they motivated to support a particular argument?
Are there alternative explanations? Could the results easily be explained by reverse causation (B causes A rather than the other way round) or by an omitted variable that causes A and B to be correlated?
Is it balanced? Is it designed to educate or to make a case?

Bearing these questions in mind, let’s look at ShareAction’s evidence.

MIT Sloan School of Management

The first thing to note about this article is that it was published in Harvard Business Review. This is a publication where academics and practitioners come together to provide insights for the real world. As such it is very helpful, but it is not a publication where data-driven results are shared and subject to the scrutiny of academic peer review. So we need to look through to the underlying work of the authors on which the article is based.

The authors say that the financial business cases for better pay in retail can come from cost reductions, higher revenues, and labour productivity gains. This certainly can be true. But is it? The authors present no data to support their case, but only highlight two case studies, which are stories rather than data, and a fictional company that they’ve made up. This is not really evidence in any normal sense of the word. But we wouldn’t expect detailed evidence in Harvard Business Review, so it’s important to look at the background work of the authors on which the article will be based.

Looking at the background of the authors, two work at the Good Jobs Institute, a non-profit whose mission is to help companies thrive by providing good jobs. Good for them. Making work more humane and fulfilling is hugely important for society. But it also means they are likely to be prone to motivated reasoning when it comes to research.

The connection to MIT Sloan School of Management comes from Zeynep Ton, who is Professor of Practice at the School and a cofounder of the Good Jobs Institute. (So really this should be referred to as a Good Jobs Institute study rather than an MIT Sloan School of Management study.) She also appears to do consulting in the area of improving work quality. Most of her articles are thought pieces in Harvard Business Review and the like but I have also looked at her academic papers. One of them seems relevant to the question at hand: If higher pay is profitable, why is it so rare? Modeling competing strategies is mass-market services. This paper has not yet been published so has likely not yet been subject to peer review. But in it she does look to build a strategy-based model for why higher pay strategies in sectors like retail are not so common.

In brief, the authors find that there are two strategy choices that create profitability peaks, one based on a strategy with low pay and low task richness and one based on a strategy with high pay and high task richness. However, the high pay and high task richness strategy is more difficult to achieve because of three challenges, which I summarise as:

The need for multiple strategic components simultaneously to be maintained, requiring significant customisation for each company context in which it’s delivered.
A dependence on long-term and synergistic investments, with slow-moving feedback loops, which are hard to learn, leading to misleading performance feedback.
Efforts to adjust labour supply to highly variable demand compromise the strategy.

In other words, this paper suggests that a high pay high productivity strategy is possible, but is difficult. And is certainly not one-size-fits all (unlike the ShareAction proposal). Companies may be inspired to take such a strategy on and helping such companies is the work of the Good Jobs Institute.

Ton has written a book called The Good Jobs Strategy. I haven’t read this, but it is clear from its description that the book is a set of inspiring case studies saying it is possible to find a way towards a high pay high productivity model rather than data on how frequently this can be achieved. Indeed it is possible. But while such a strategy can be successful, Ton’s own research demonstrates why it is not inevitable that it will be. Indeed her research suggests that the risks of failure are quite considerable.

Cambridge University

The Cambridge paper was published by the Cambridge Institute for Sustainable Leadership with the support of Business Fights Poverty. It is not an academic study providing new data per se, although it has elements of academic review within it. Two of the authors are from Business Fights Poverty, which appears to be an organisation bringing businesses together with the aim of using the power of business to help improve the lives, livelihoods and learning opportunities of the most vulnerable people and communities. This is indeed a very worthwhile aim, which I commend. However, it does mean that they have a stake in the conclusions of the report.

The lead academic author appears to be Dr Anna Barford a geography lecturer who is also the Prince of Wales Global Sustainability Fellow at the Cambridge Institute for Sustainability Leadership.

The paper is called “The Case for Living Wages”, which rather indicates the stance it will take. I have to say I found it a very good report, and I would recommend it to any CEO who is looking at implementing a living wage strategy globally. It contains some great content on how to overcome the challenges of implementing a living wage strategy, particularly across the world where the mere calculation of living wages is itself a formidable challenge. It’s a great “how to” guide for companies going down this commendable path.

However, what we’re concerned with here is the business case for living wages. Here the report is less persuasive, based more on exhortation than evidence. The authors themselves admit that: “…it remains hard to isolate how living wages directly affect financial performance”. The list of business benefits is what you would expect to be: higher employee wellbeing, lower turnover, increased productivity, feedback of greater spending into the economy, reputational benefits and so on. All of these can indeed be benefits of higher wages, but cannot apply without limit and need to be balanced against the cost. There must be trade-offs between different stakeholder interests: employees, shareholders, customers, suppliers, those not yet employed and so on. Otherwise business would be easy! These trade-offs apply at the company and system level, and the fact that a benefit can exist does not mean that it does, nor that it outweighs the cost.

When it comes to the evidence cited in support of the financial performance claims, the report seems largely to rely on three academic studies. One of these is published in an academic journal and is a paper from 1999 looking at the impact of introducing a living wage in Baltimore in the US. Two are research reports from the last decade, one from Queen Mary University and one from Cardiff University. These last two have not been published in journals and so have probably not been subject to peer review.

I cover the Cardiff study separately, as ShareAction directly cite this. I will therefore briefly comment on the other two papers.

The Baltimore paper studied the impact of the state introducing a living wage requirement for contractors providing services paid for by the taxpayer. It focusses on the implications for the state. This is much more akin to an economy-wide minimum wage measure and so does not really address the business case for voluntary adoption by companies. In summary, the study found that the costs were manageable for the state and had some benefits for workers. This is consistent with a lot of work on statutory minimum wages, which on an economy-wide level often end up being more manageable than economists predict. A brief review suggests that there is other research on similar state-mandated living wage requirements for public procurement, which reaches similar conclusions.

The paper on the London Living Wage is interesting and more relevant as it relates to optional rather than mandated introduction. It struck me as being a carefully constructed study using mixed methods combining detailed company case studies with employee surveys. The authors admit that a shortage of matched non-living wage employers participating in the study is a weakness and sample sizes are small, which prevents statistical significance tests in the corporate side of the study. However, the study finds overall that for most employers and suppliers the living wage commitment was manageable, with increased wage costs offset by a combination of: fixed price contracting; use of service audits potentially with financial penalties; reduction in headcount and / or hours; alterations in service specifications and supplies and in some cases reduced margin. The study did not include a measure of net benefit to employers and so provides no conclusions on whether they were positive or negative.

These are encouraging findings for those pushing the case of the Living Wage at Sainsbury’s, even though the Baltimore findings, in particular, are not so relevant. However, they far from prove the business case in general or in the specific. They do, however, give encouragement to those, like me, who favour moving faster rather than slower on statutoryeconomy-wide minimum wages, as the Low Pay Commission has now been mandated to do.

Cardiff Business School

The Cardiff study is a survey study, which asks Living Wage adopters about their experiences of adopting the Living Wage. It’s an interesting read but the research design poses two immediate problems for those wanting to use it as evidence for the financial benefits of the Living Wage. It is perhaps for this reason that it is presented as a research report in conjunction with the Living Wage Foundation and Citizen UK, rather than being published in an academic journal. The research design would likely have been immediately rejected on peer review.

First, there is selection bias. Those companies who have voluntarily chosen to be Living Wage Accredited are very likely those for whom the benefits are greatest and costs lowest. Therefore, the impact on these companies can in no way be extrapolated to those that have yet to adopt. Particularly they cannot be extrapolated into a sector where no major employer has yet become Living Wage accredited.

Second, there is the well-known phenomenon of people becoming invested in decisions they make. It can cause cognitive dissonance to admit that we’re wrong, and confirmation bias leads us to cherry-pick evidence that supports our case.

So overall, the study is a bit like asking Leave voters whether Brexit has been a good idea.

Nonetheless, with these potential biases in mind, such a survey can still provide useful insight and raise important issues, even if it will never provide conclusive proof of the impact on a firm of adopting Living Wage accreditation. And the report is full of fascinating insights, which I can’t cover here. I’ll focus on just on benefits and costs.

Focussing first on benefits, most adopters felt they captured reputational benefits from living wage adoption, which was the key benefit outlined. Around half of respondents felt that there were advantages along various dimensions relating to employees: employee relations; commitment, motivation, retention, and recruitment. Around one-third felt that these benefits were of major or moderate significance.

The main disadvantage, cited by around two-thirds of respondents, was increased labour costs (with just over 40% saying the disadvantage was of major or moderate significance). Around one-third separately cited the increased cost of contracted services.

The authors claim that the benefits outweigh the disadvantages. But given that the cost disadvantage was viewed as significant by more respondents than the tangible employee-related benefits it is not clear how they come to this conclusion and there is no methodology to justify this.

Moreover, they found that the challenges were more pronounced in the main low-paying sectors of the economy: retail, hospitality, social care and cleaning. As the authors state, for companies in these sectors, like Sainsbury’s: “where the bit of the Living Wage is relatively powerful, as is likely in low-wage industries, then more challenges ensue.”

Overall, the study is interesting and for many businesses holds out hope that moving to pay Living Wages is a manageable strategy. But it says nothing about how the costs and benefits compare for a retailer that has not adopted Living Wage accreditation to date.

Business Commission to Tackle Inequality

The final study referenced is Tackling inequality: the need and opportunity for business action by the Business Commission to Tackle Inequality (BCTI), including a number of prominent CEOs including Alan Jope, CEO of Unilever, which some years ago adopted a global living wages policy.

This is very much a piece of advocacy (and a powerful one at that) rather than a research report. It is a call for leadership from the corporate sector to tackle the “systemic issue” of inequality. The report does make some claims for the specific business benefits of paying living wages, but these claims are not based on academic evidence (consultancy reports are cited), and the report focusses much more on the systemic risks of global inequality.

In this context, the relevance of this analysis to Sainsbury’s is questionable. In my initial blog I highlighted that the evidence in favour of a negative link between economic growth and inequality is mixed and seemed strongest in low-income countries. In the UK, in part because of significant increases in National Living Wage, inequality after tax and benefits has been broadly stable following an increase through the 1980s. Indeed a study by researchers at the Bank for International Settlements that does make a case for a negative feedback loop between inequality and recessions also notes that the UK’s level of inequality has plateaued in the last 20 years. Of course Sainsbury’s has international supply chains, although the Living Wage proposal only applies in the UK.

Overall, the BCTI report is three parts advocacy one part research. That’s not a criticism – it’s a call to leadership, and one with which I have sympathy across many dimensions. But is says nothing about the business case for a UK retailer such as Sainsbury’s to become a Living Wage employer.

Where does this leave us?

The research cited by ShareAction is far from conclusive, and overall has little to say on whether there would be business benefits from Sainsbury’s becoming a Living Wage employer that would outweigh the costs. The research suggests that there is a pathway for some companies to move to a higher pay model in a way that is manageable in the context of business performance and even positive. But it also suggests that there will be significant risks with this approach, particularly in low margin industries.

What the research clearly does not do is to provide a sufficiently compelling financial case for shareholders to mandate the Sainsbury’s directors to take a course of action that they do not believe to be in the interests of the company or its stakeholders overall. The research cited also does nothing to show that there is a systemic inequality risk in the UK that warrants investor action of this type.

Of course, other compelling research that I’m not aware of may be available to prove these points. But ShareAction haven’t provided it.

So the financial business case is not made by the research that ShareAction cites. That’s ok, we still have the moral, intrinsic business case. It remains perfectly legitimate for shareholders to vote for the resolution on these grounds provided they understand that there may be, perhaps probably will be, non-recovered financial costs involved (even though ShareAction seems to deny the reality of this possibility). However, for an asset manager intermediary to vote in this way, they need a crystal-clear mandate from their clients that this is a cause against which they are willing to put their money at risk.

ShareAction’s case is really a moral one. And that’s fine. But presenting it as a financial case for Sainsbury’s is unconvincing. And this isn’t just “academic”. Investors have a fiduciary duty to act in their clients’ interests, and cannot, consistent with that duty, take a certain moral position on the vote just because ShareAction says so. This is the money of ordinary people saving for their futures for which those asset managers are fiduciaries.

As importantly, voting against this resolution is not an abrogation of moral responsibility, but can simply represent a difference in the proposed means. Attempts to impugn the morals of those on the other side of the argument isn’t constructive. The National Living Wage, supported by the Low Pay Commission is in my view proving a relatively effective mechanism for ratcheting up minimum in-work wages in the economy, in a way that takes all stakeholder interests in the economy into account. Of course, there are enormous strains on society from the cost-of-living crisis as ShareAction have graphically reported. But these are not going to be addressed in an effective or timely way by a campaign of Living Wage adoption. As Torsten Bell, Director of the Resolution Foundation has pointed out, we have a well-developed system for getting money quickly to those that most need it: the benefits system. Shareholders can’t solve every problem.

I personally care about inequality and would rather live in a more equal society where there was greater social mobility. At every election in my adult lifetime I have voted for the party I thought most likely to bring that about regardless of the personal implications for me and my family’s financial position. But while I can agree with cajoling and encouraging a company to consider whether they can find a pathway to being a Living Wage employer, forcing them to do so against the recommendation of their board is not something I can support, as I believe it will likely not be in the interests of the company and its stakeholders overall.

But this doesn’t mean I don’t care about the less well off in society. ShareAction and those that support their resolution don’t have exclusive rights to that.