Helen Spiropoulos and Rebecca Bachmann

Australian businesses fudge their numbers to show social and environmental ‘progress’

What’s the easiest way to improve a company’s social and environmental performance? The unfortunate answer, from our analysis of Australian public companies, is to change the way you measure it. In particular, by changing what you said last year to make this year’s performance look better.

Reporting on environmental, social and governance (ESG) performance is increasingly important to the fortunes of listed companies – under pressure from investors, regulators and other stakeholders.

In some cases, executive pay is tied to these metrics.

Our study – which examined reports from the top 500 companies listed on the Australian Securities Exchange (ASX) over the past 15 years – suggests progress can be little more than sleight of hand, facilitated by the lack of clarity around how ESG performance is measured and reported.

Helen Spiropoulos (University of Technology Sydney) and Rebecca Bachmann (Macquarie University)

Our final data set includes those companies that report on both their ESG performance and their executive remuneration practices. This numbered about 215 individual companies.

Of these, roughly one in six made adjustments to past reported ESG performance, particularly around social measures such as gender diversity or workplace safety. The average size of the adjustments was also significant, at 28% of the original value reported.

About 55% of companies tied a proportion of their chief executive’s bonus to ESG metrics. These companies were twice as likely to make one or more adjustments to past reported ESG numbers. In fact, 33.5% of all adjustments were of an ESG measure that was included in the chief executive’s bonus. The average size of these adjustments was also larger, at 36% of the original value.

This occurred across all industries but was most common in two areas: the financial sector and the materials sector, which covers mining and chemical, construction and forest products.

Linking executive pay to improvements

If retrospective changes were the result of previous mistakes or improvements in measurement systems, there should be no bias in the direction of changes to past performance (that is, it could go up or down). But we found a significant bias towards making past performance look worse, thereby making the current year’s look better.

Fuelling this seems to be the practice of tying executive bonuses to a simple improvement, rather than to a stipulated target. For example, executives might be rewarded for increasing the proportion of female or Indigenous employees or cutting injury rates, rather than hitting specific targets for these metrics.

We found about 17% of bonus pay is typically attributed to ESG targets. For the average chief executive in our study, this translates to around $200,000 in extra income. It’s not surprising that retrospective changes making current-year performance seem better were more likely to occur when greater weighting was given to these targets in the chief executive’s contract.

The case of Commonwealth Bank

The restatement of ESG measures is illustrated by the case of the Commonwealth Bank of Australia, which has been criticised for vague performance targets.

Over the past few years it has tied about 10-15% of the chief executive’s bonus to relatively intangible metrics. In its 2020 report, for example, “people measures” covering “culture, well-being, talent and capability” comprised 9% of the chief executive’s bonus.

These metrics have persistently changed retrospectively, as shown by employees’ “safety and well-being” results.

For example, the bank’s 2017 annual report showed 1.1 injuries per million hours worked (this is a standard measure, known as the “lost time injury frequency rate”).

Excerpt from 2017 Annual Report visualising the Lost Time Injury Frequency Rate for 2017 (CBA Annual Report 2017, page 47)

However, in its 2018 annual report, the bank revised the 2017 figure from 1.1 to 1.6, which meant the 2018 report showed injury rates as having decreased relative to the previous year.

Excerpt from 2018 Annual Report visualising the Lost Time Injury Frequency Rate for 2017 and 2018 (CBA Annual Report 2018, page 78)

The stated reason for the change was that the updated 2017 figure includes injury claims received after the year-end reporting date, as well as expanding the scope to include New Zealand employees when calculating this metric.

In its 2019 annual report, the 2018 figure was then revised up from 1.1 to 1.4. This meant the report showed no increase in injury rates for 2019.

Excerpt from 2019 Annual Report visualising the Lost Time Injury Frequency Rate for 2018 and 2019 (CBA Annual Report 2019, page 303)

In its 2020 report, the 2019 figure was revised upward to 1.6. If we compare the final adjusted figures for 2019 and 2018, there was actually an increase in rates between the two years.

Excerpt from 2020 Annual Report visualising the Lost Time Injury Frequency Rate for 2020 and 2019 (CBA Annual Report 2020, page 50)

This practice is persistent up to the 2023 reporting period, and it’s likely legitimate that some injury claims are indeed filed after the year-end reporting date. However, if this is the case, we might wonder why an incomplete number from a current report is being compared with a complete one from a previous report.

In response to queries from The Conversation, a spokesperson for Commonwealth Bank confirmed this is indeed how these figures are collated, but denied that there is pressure to compile the data in a way that gives the impression of constant improvement.

Of course, safety and well-being are important issues, but the discretionary nature of these metrics means retrospective changes can make it look like improvements have occurred. Therefore, we advise users of this information to exercise caution when comparing performance to adjusted numbers, particularly as little information is typically given to explain why the adjustment was made.

For example, Commonwealth Bank’s 2023 annual report features a change to the way full-time equivalents are calculated, but without explaining how or why the previous years’ figures were adjusted.

Manipulating vs. managing

As per the adage “what gets measured gets managed”, governments and investors have pushed sustainability reporting on the basis it will encourage companies to be socially and environmentally responsible.

This is partly due to a shift away from the view that companies exist purely to maximise shareholders’ wealth, and towards the idea they should be good corporate citizens with minimal negative impact on the environment and society. Some investors prefer to invest only in stocks that meet ESG performance criteria.

This creates incentives to exaggerate claims – made easier by the lack of uniformity in measuring and reporting such results.

Who sets the standard?

An international body to set global standards – the International Sustainability Standards Board – was established in 2021. It published its first set of ESG standards in June.

Australia is set to follow these standards, under a proposal by Treasury, with one catch: the regulations will focus only on climate-related disclosures, which will apply to Australian companies from 2024, overseen by the Australian Accounting Standards Board.

While these standards will create greater consistency in ESG disclosures, there will remain significant discretion in how ESG performance is measured – including the ability to change how ESG items are measured from year to year and to adjust previous years’ apparent performance.

Currently, ESG performance isn’t required to be audited and just 22% of the companies we looked at had their ESG performance audited. Last month, the International Auditing and Assurance Standards Board issued a draft sustainability assurance standard that will require auditors to provide assurance of reported ESG figures.

Treasury has also said Australia will follow this standard and expects auditors to provide reasonable assurance on all climate disclosures by 2027. Hopefully, these audits will also consider the legitimacy of revisions or changes in measurement. But regardless of this improvement in accountability regarding environmental performance, companies’ reporting on social metrics will still be unregulated.

The irony is striking. What was conceived as a mechanism to drive positive environmental and social change may instead act as an incentive to manipulate sustainability performance.

This article is authored by Helen Spiropoulos (Associate Professor, University of Technology Sydney) and Rebecca Bachmann (Lecturer in Accounting, Macquarie University). This article was originally published on The Conversation.


Inform your opinion with executive guidance, in-depth analysis and business commentary.