The Gini in the executive pay bottle

💡

This post was originally published on the Sense of Fairness blog (original post here)

What’s the best way to assess the fairness – or unfairness – of executive pay? It’s a lively debate around the world. It has a particular bite in countries where overall inequalities are especially pronounced.

Usually that fairness is assessed, or at least made manifest, through disclosure of pay ratios. The US and the UK now both require disclosures by public companies of the ratio between the CEO’s total pay and that of the median worker* (the US requires the comparison to be across the global employee base, with some limited exclusions permitted, the UK only considers the UK employees; neither considers the circumstances of the workforce who are not employees). Campaigners and legislators have focused on this ratio: see, for example, San Francisco’s Overpaid Executive Gross Receipts Tax, which imposes additional taxation on companies that pay their CEOs more than 100 times their median employee; and the Max 12:1 campaign. But there are broader measures that could be considered.

For various reasons I have read a few South African annual reports of late, for the first time for some years. And I was very struck by some of the disclosures made by miner Impala Platinum (usually called Implats for short). There is an explicit section of its latest Annual Report called ‘Our approach to fair pay’ which also references the company’s Fair Pay Policy Statement. This policy statement includes Guiding Principles, among them Principle 3 on Equity, ‘Reward given to different employees is fair, consistent, and justifiable’:

Principle 3a: All jobs are appropriately graded to reflect required technical knowledge, skills and experience
Principle 3b: Reward policies are designed to enable necessary variation depending on local contextual factors, such as in hardship locations, fragile states or absence/scarcity of necessary skills
Principle 3c: There is a commitment to policies being applied systematically

These fine words also appear to be reflected in practice. The annual report notes recent pay deals both for its refinery workers and its miners, and asserts that “The Implats guaranteed minimum wage for permanent full-time employees remains significantly higher than a ‘living wage’, and in addition our employees are eligible for progressive variable pay arrangements which are generally above the median for the industry.” The pay ratio that it discloses is between the CEO and the lowest paid employee, and is 58:1. While this is far above the 12:1 some campaigners are seeking, many US and UK companies report pay ratios far in excess of this – and those are ratios to the pay of the median employee, not to the lowest paid.

But still more unusual are the company’s disclosures of income inequality metrics most often applied by economists, the Gini coefficient and Palma ratio (both explained at the foot of the image):

All of this discussion of fairness in executive pay is a response to the expectations of South Africa’s corporate governance code, the latest update of which – referred to as King IV – was published in 2016. As some of the introductory wording to the new Code states: “An important introduction in King IV is that the remuneration of executive management should be fair and responsible in the context of overall employee remuneration. It should be disclosed how this has been addressed. This acknowledges the need to address the gap between the remuneration of executives and those at the lower end of the pay scale.”

And fairness is in fact included in King IV’s Glossary of Terms:

“Fairness refers to the equitable and reasonable treatment of the sources of value creation, including relationship capital as portrayed by the legitimate and reasonable needs, interests and expectations of material stakeholders of the organisation.”

Putting this into practice, Principle 14 reads “The governing body should ensure that the organisation remunerates fairly, responsibly and transparently so as to promote the achievement of strategic objectives and positive outcomes in the short, medium and long term.” Under it, recommended practice 27 states: “The governing body should approve policy that articulates and gives effect to its direction on fair, reasonable and transparent remuneration.” And one expected disclosure in remuneration reports is: “An explanation of how the policy addresses fair and responsible remuneration for executive management in the context of overall employee remuneration.”

This focus on fairness matters particularly in South Africa, which because of the unfairnesses of its past is one of the most unequal societies in the world. The nation’s Gini coefficient and Palma ratio are still worse according to the OECD than those set out in the Implats annual report, though the calculations are from 2017 rather than the latest years: 0.62 and 6.9 respectively. In both cases, the nation sits at the most unequal spot among those countries for which data is disclosed. It is recognised that business must play a significant role in addressing this unfairness.

Not all South African companies agree though that disclosure of these ratios for their employees is the best way to express their attempts to press for fairness. The most blunt that I found was this from the annual report of retailer Woolworths: “While the Gini coefficient or index is widely considered to be the most scientific and accurate measure of income disparity and many commentators use it as a proxy for fair and responsible remuneration as envisaged by King IV, the committee agreed to focus on strategic initiatives to drive and address fair and responsible pay.” It emphasises that it is not tied to South Africa’s minimum wage, nor even to the concepts of a living wage, but rather is aiming for a ‘just wage’, “informed by many data points, including minimum wage rates, market rates, CPI, and our EVP [Employee Value Proposition – the firm’s overall approach to staff retention and reward] strategy”.

Woolworths also produces a striking waterfall chart showing the impact of its own approach:

I have railed previously about the problems that come from a focus on just one metric, and I’ve suggested that the fact considerations of inequality are often reduced simply to a discussion of the Gini coefficient is narrowing and unhelpful. So I don’t rush to approve of Implats’ disclosure of Gini. But it is very clear that disclosing the Gini coefficient, as well as median pay ratios and pay ratios to the lowest paid employee, and also the Palma ratio, offer a much richer understanding of the spread of pay across the whole organisation than disclosing any one of those numbers alone. The US and UK requirements to disclose pay ratios are positive, but it is becoming harder to say that they go far enough; what they do indicate is that many companies have many more issues of unfairness than both Implats and Woolworths seem to have, notwithstanding the difficult and unfair national context that both of those companies face.

It’s particularly welcome that both companies looked at in a little detail for this post disclose their number of employees and that all of them receive at least the living wage. As I’ve said in a previous blog, I think all companies everywhere should be disclosing these datapoints as the start of baseline information on social factors that can be aggregated across portfolios. The South African remuneration approach pushes companies to think much more about fairness, and companies are rising in a variety of positive ways to this challenge. Other countries have something to learn from King IV’s call for fairness, and companies elsewhere in the world from how South African businesses are responding to that call.

‘median worker’ is clearly a shorthand for the worker who is at the middle of the pay distribution (often calculated by shorthand methods). No worker is average.