Insights from the IRAS data

Using data from two sectors to illustrate gaps in employee well-being

The indicators assessed are as follows:

• Absenteeism
• Employee turnover
• Income disparity ratio
• Lost-time injury frequency rate
• Number of employees trained
• Rand spend on employee training
• Percentage of employees who are permanent employees  
• Lost days due to strike action

The challenge is that companies are only now coming to grips with how to accurately report much of this information. This can throw into question the quality of some of the data discussed, in spite of the fact that it has all been collected out of public reports that have presumably been signed off by qualified assurance experts. As such, it’s not always clear whether the counterintuitive findings that may occasionally surface are a function of misreporting, misunderstanding the basis for reporting, or genuinely important conclusions that may have eluded our analysis in the past.
 

THE EIGHT ISSUES IN MORE DETAIL:
 

1. Absenteeism

Absenteeism has long been viewed as an indicator of an organisation in – or out of – control. Both job type and industry can have an impact on absenteeism, so industry comparisons become important.

What constitutes ‘absenteeism’ can often be misunderstood. In many cases, companies will report sick leave, not understanding that people who are ‘absent without official leave’ and late arrivals are significant contributors to total lost production time. Systems are therefore structured around ‘official leave’, consisting sick leave, maternity leave, parental responsibility leave and study leave, but this results in a gross under-reporting of actual absenteeism.

In one case, a company reporting only sick leave learned that the true extent of absenteeism was greater than 6%, not the 2% they’d been reporting. Given that their total turnover was roughly R40 million per year, their total cost of uncalculated absenteeism was roughly R1.6 million, across an employee population of just over 200 people. With the average employee earning R600 per week, absenteeism was resulting in an additional payroll cost of 25.6% of total cost to company.

To address this, the company considered a quarterly bonus for employees. If they could reduce total absenteeism to below 2%, every employee would receive R300, or 50% of a week’s wages (a total bonus pool of R240 000 per year). Did it make sense to pay employees extra to do what they should have already been doing? In this case the economics were irrefutable. The 4% regained production time translated into R1.6 million of revenue ‘purchased’ at a cost of R240 000.

The problem is fewer than 12% of all JSE-listed companies provide absenteeism data in their public reports. Either that means companies aren’t paying enough attention to this key statistic, or the data is not something companies want to advertise.

Comparing Metals & Mining with Retail

The Metals & Mining sector had the highest reported absenteeism rate at 7.2%. The Retail sector reported a much lower absenteeism rate of less than 1%. However, only seven (13.3%) of the 45 companies in the Metals & Mining sector provided comparable absenteeism data, while only 8.7% of Retail companies disclosed their rate, yet both sectors stand to lose significantly from lost productivity. Worse still is that within the supplied data, significant issues arise.

Some significant players in the mining industry reported absenteeism rates of as high as 17.1%. Either these rates have been miscalculated (a possibility as companies scurry to improve their reporting standards) or they signal potential productivity issues for the company – particularly if the absent employees possess critical skills, as they often do in mining.

In the Retail sector, only one company appears to have adequate controls in place to both manage absenteeism and report reasonable data. While 19 companies provide no absenteeism data, other companies submit data that is highly improbable. No company, regardless of how well they might manage absenteeism, can record only 1 685 lost days due to absenteeism across an employee population of 13 375 (roughly 3 million available person days).

There is a need for South African companies to be much more transparent about the cost of absenteeism within their workforces, noting the manner in which researchers can assign negative ‘proactive employer ratings’ for the absence of data, or the presence of data that calls into question the company’s ability to encourage employees to show up for work and be duly (and correctly) counted.

2. Employee turnover

Do companies pay enough attention to their employee turnover (ET) rates¹? Do they adequately interrogate causes of excessive ET, or make efforts to differentiate between ‘controllable’ and ‘uncontrollable’ causes, or, most importantly, attempt to find solutions to rectify those causes?

Employee turnover constitutes a key area of business risk, particularly with those companies requiring skills, or where continuity of service to customers is paramount to business success. More importantly though, ET can point to other deeper concerns that might be emerging for employees.

In 2008, a small garment manufacturing company in Cape Town identified a significant increase in ET and determined that one of the leading causes was garnishee orders, or emoluments attachment orders (EAOs). Employees had been given access to credit as a direct function of the National Credit Act (NCA), but were not financially literate enough to understand the significant risks and challenges of accepting retailers’ offers of ‘buy now, pay later’. Worse still, many had borrowed money from less than scrupulous lenders who were charging exorbitant interest rates.

Many workers simply applied their own solution to the problem by resigning from their place of employment as soon as the net sum of wages minus all garnishee orders decreased to below a level that was deemed a reasonable living wage. At a certain point, it no longer paid to go to work.

In some cases, companies opt to simply ignore garnishee orders, even when the EAOs were handed down as judgements in court proceedings. In other cases, such as the clothing manufacturer, the management team made an effort to provide debt awareness training to all employees, and debt counselling to those identified as ‘at risk’, ultimately observing a significant decrease in their ET rate. This decreased the direct and indirect costs of ET through recruitment, training, and lower productivity rates during the transition from old to new employees. Thus, the benefit of being a proactive employer with respect to ET is equally accrued to both employer and employee (for those willing to take the advice of the supplied debt counsellors).

Comparing Metals & Mining with Retail

In terms of the employee turnover data, the Retail sector (with 23 companies in the sector) scores relatively well with a reporting rate of 47.82%, while Metals & Mining (with 45 companies in the sector) scores rather poorly, with a rate of only 35.56%.

Given the higher proportion of unskilled labour working in the Retail sector, coupled with the sheer volume of employees (one company with more than 110 000 employees had a FY2013 ET rate of 17.3%), the challenge isn’t so much around cost of training as it is about the cost of constantly recruiting new employees to replace those who have left. For some well-known brands within the sector, ET rates are reported to be as high as 30.80% and 28.91%. Managing this metric provides some important insights into a company’s willingness to engage. The broad range of outcomes suggests that peer to peer, there appears to be considerable variability in attention being directed towards this metric.

By contrast, in the Metals & Mining sector, the highest ET number hovers around 17%, with an industry average of only 8.09%, well below the median of the Retail sector. This could be a function of the significant cost of recruitment and training in the sector, such that employees can be considered ‘less expendable’.

3. Income disparity ratio

In this age of continuous access to information, where anyone can access almost anything at any time, there are few societal concerns greater than income disparity. Economists like Thomas Picketty and Joseph Stiglitz have illuminated the growing discord resulting from too few people possessing more than the too many. In the United States of America, the Occupy Wall Street campaign was precipitated by a need to question how ‘the 99%’ can possess less collective wealth than those who make up the 1% of the population who own the rest.

For the past few years, annual general meetings (AGMs) around the world have been inundated by disgruntled stakeholders who argue that those at the top should not earn a disproportionate amount of the total wages paid by the company. In Switzerland, the government even attempted, but failed, to stop the practice of over-compensation by setting limits to what executives can earn. Yet with all of the public discontent, there are still too many company directors who can comfortably turn a blind eye to the anger welling up within their workforce. Perhaps this is a case of ‘in the land of the blind, the one-eyed man is king’, where only the bravest of MDs are willing to squint when forced to observe how income disparity is affecting the overall sustainability of their companies.

Comparing Metals & Mining with Retail

In 2014, the Metals & Mining (six companies) and Retail (five companies) sectors dominated the list of least equitable companies when comparing average executive director compensation (excluding gains on share options) to average employee compensation. In some of the least equitable companies, executive directors earned 188.34 times what their employees earned. This means that those at the top need only report to work for two days to earn what it would take the average worker to earn in an entire year of working.

As you might expect, the vast majority of JSE-listed companies provide data about how much everyone – executive directors and employees alike – earn. However, there remain several companies who manage to skirt around Generally Accepted Accounting Principles (GAAP) and International Financial Reporting Standards (IFRS), and avoid disclosing compensation data. Roughly 9.5% of the 311 IRAS-reviewed companies did not disclose executive director compensation, while 5.95% failed to report employee compensation data. This further calls into question what those companies felt the need to hide.

In fairness though, high income disparity should not necessarily force one to conclude that a company isn’t equitable or fair. One of South Africa’s largest retailers, who have over 111 000 employees, has an income disparity ratio of 188.34, while a smaller retailer with only 4 000 employees has an income disparity ratio of only 15.64, roughly 12 times less than that of the larger retailer. A medium-sized retailer with an employee population of 23 538 reports an income disparity ratio of 75.86. This suggests that of the three retailers, the largest one is perhaps doing the best job of giving more people access to employment, a fundamental component of what constitutes being a responsible employer. It’s simple math, the more people you hire at the lowest possible wage level – replete with those deemed ‘unskilled’ – the greater the possibility that those at the top of the income pyramid will be earning an inordinate amount of money in comparison.

4. Lost-time Injury Frequency Rate (LTIFR)

Nothing says, ‘We appreciate our employees’ more than an active commitment to keeping them alive.

Investing in safe machinery, adequate personal protective equipment, training, and adequate oversight and control procedures to ensure that every worker makes it home at the end of the day is tantamount to the ultimate demonstration of appreciation.

We can’t assume that any industry, whether it’s mining or construction, is inherently dangerous, and that employees must be willing to put their lives on the line in exchange for a pay cheque. Rather, the health and safety of each and every employee ought to be the first consideration whenever any company chooses where to operate, how to operate, and under what sociopolitical constraints. However, the data provided by some companies over the past few years tends to suggest that safety is far too frequently overlooked as a material concern.

Here is a particularly interesting example. In 2013, when IRAS launched its sustainability data analysis, many readers were shocked to learn that three of the country’s largest banking and financial services companies, FirstRand, Absa and Standard Bank, were more dangerous than some of the country’s largest mining companies. Even Sasol, which most people assume is a dangerous company based on highly publicised tragic events that occur less frequently than anyone estimates, reported a LTIFR roughly half that of the banks. Perhaps more astonishing was the fact that Nedbank even reported a fatality, or that Vodacom reported the tenth highest fatality injury frequency rate (FIFR), or the number of fatalities per 200 000 person hours worked (PHW).

The great challenge with LTIFR data is trying to determine whether the numbers reflect a genuine misunderstanding of how to calculate the data or whether something is genuinely amiss. Ironically, the data might also suggest that those industries that objectively appear to be more dangerous give far more attention to this aspect of employee protection than industries where risk of injury on the job should theoretically have a lower probability of occurring.

Comparing Metals & Mining with Retail

In the mining industry, safety continues to create cause for concern, but not necessarily because mining companies are the most dangerous to work for. Only two mining companies report an LTIFR that places them within the top 10 most dangerous companies in South Africa. Three of the top 10 are actually from the Food & Beverages sector. That said, of the 45 JSE-listed Metals & Mining companies reviewed in 2014, only 31 companies reported an LTIFR.

The problem? Mining actually is dangerous, regardless of what controls are in place, and therefore all mining companies should publicly disclose what they are doing to manage safety as a critical risk. The absence of such information must therefore call into question one of the two possibilities:

• The company does not have the policies, procedures, systems and/or controls in place to effectively manage safety risks
• The data is so bad that the company is unwilling to publicly disclose the information.

In either case, the problem is immensely troublesome, and investors – if not Government, unions and other key stakeholders – should be deeply concerned and should call the companies to account.

In the Retail sector, the problem is even more worrisome, as only three companies disclosed an LTIFR. All three reported an LTIFR of 0.00, which is highly improbable (if not impossible). With more than 300 000 people reportedly working for the 23 JSE-listed retailers – including people transporting goods from one location to another on South Africa’s notoriously dangerous roads, use soapy mops to clean already slick floors, or work on lighting and ventilation often perched at significant heights – we might conclude that the lack of injury data could mean that retailers aren’t committed to workplace safety.

5. Number of employees trained

In almost every ranking of ‘Best Companies to Work For’ or ‘Most Responsible Companies’, investment in employees is deemed a fundamental ‘must have’. More than wanting to believe in the company, or align to its mission, vision and values, employees want to know that their employer values their capabilities and will give them an opportunity to grow. New recruits, particularly those starting out on their first post-schooling ‘real job’, tend to assign more inherent value in training and development than in the size of their pay packet. They want to know that their future potential within the company will be identified and nurtured through company-sponsored training, and companies tend to understand that recruiting talent is relatively easy compared to retaining the most desirable of employees. As such, global leaders in employer surveys are quick to market their views on the importance of training.

In South Africa, where skills are scarce – and as a result valued at a premium – the ability to retain competence and promote from within is all the more critical to the ongoing sustainability of any company. As a result, it’s fair to assume that companies would ensure that they publicly disclose comparable data regarding the extent to which employees are trained. However, this doesn’t appear to be the case.

Comparing Metals & Mining with Retail

Of the 45 Metals & Mining companies, only 19 provide data on employees trained. Six reported obviously incorrect data, as their data suggests that they train more people than they employ.

In the Retail sector, the data is equally disappointing, with only 12 of the 22 companies providing training data. Of those, the data for three is obviously incorrect (or suggesting even greater problems like extremely unsustainable rates of employee turnover). Several large retailers reported training that exceeded the number of people they actually employ, with one company reporting training figures that represented eight times their reported employment numbers.

The problem here, based on IRAS’s experience as an assurance provider – or auditor of non-financial data – is that far too many companies over-report training data based on the application of an overly simplistic definition of ‘training’.

When we refer to ‘training’, or ‘training and development’, the issue is not whether or not every employee is inducted into the company with training on the company’s HR policies and procedures, but whether or not employees are given access to new skills. The count should not include the number of people participating in mandatory safety inductions, but rather those persons who have participated in specific training initiatives. This would include the number of people participating in anything that would create new ‘portable skills’.

In some cases, the opposite is true. Some companies are overly harsh in their definition of what training is, or is not, reporting that fewer than 5% of all employees have been afforded access to skills development. However, it is often the case that this data is a much fairer reflection of the extent to which the companies invest in comprehensive talent pooling and personalised skills development programming to develop and retain critical skills.

6. Rand spend on employee training

Several years ago, a large JSE-listed mining company battled with consistently poor safety performance. Not only were employees regularly getting injured, but lives were lost at such an alarming rate that the company had to do something. They had to place the blame on an appropriate set of shoulders. Unfortunately for the Group Health & Safety Manager, those shoulders were his, which should have meant that his access card would soon be recalled. Somehow though, his career was miraculously saved by a request to attend a company-sponsored PhD programme in the United States. Rather than pay him severance, the company chose to pay relocation costs for him and his family, to pay his salary for no fewer than four years, and to pay the substantial tuition fees demanded by an American university.

What was the benefit to the company? They were able to report all associated costs as ‘skills development spend’.

This is surely not what the government intended as skills development for South Africa as set out in the Mining Charter, the Skills Development Act and the Broad-Based Black Economic Empowerment (B-BBEE) Codes of Good Practice set out by the Department of Trade and Industry (dti). Rather, this is an abuse of otherwise extremely useful legislation. The above example is relatively rare, but it nevertheless represents one such case of abuse.

Nonetheless, the rand value of training spend can be an extremely useful benchmark for determining which companies are ‘more responsible’ than others, as well as which ones have sufficiently adequate financial controls in place to publicly report how much money they spend on training.

Comparing Metals & Mining with Retail

Of the 45 Metals & Mining companies, 30 companies provide training spend data. Unfortunately though, the data reported for many of the 30 companies is either incomplete or incorrect, such that the data for only 17 companies can be used to calculate a ‘rands per person trained’ value. The suggested value for 7 of the 17 exceeds R60 000 per person trained. Only 10 companies reported within a reasonable range of R1 700 to R15 000 per person trained.

Once again, the data for the Retail sector is less than impressive. Of the 23 companies in the sector, only 12 provide training spend data. Of this, only 8 provided enough data to calculate a ‘rands per person trained’ value, with two companies reporting obviously incorrect information.

7. Percentage of employees who are permanently employed

In recent years, much has been made of the fight for and against the use of labour brokers. On the one hand, companies argue that local legislation makes the hiring of permanent employees an unbearable risk, with employees controlling a disproportionate advantage in the balance of power needed to hire and fire people who ought, or ought not, to be given access to employment. At the same time, our workforce is regarded as one of the least efficient in the world, severely hindering our rankings in the World Economic Forum’s Global Competitiveness Index, where South Africa’s labour market efficiency is ranked 113th out of 144 countries in the latest report².

On the other hand, unions and other stakeholders argue that the corporate sector is shirking its responsibility to society by opting to recruit people on a temporary, rather than permanent, basis. Restricting the length of contract, or in many cases, limiting the number of hours someone can work in a week (which classifies them as ‘part-time’ rather than ‘full-time’ employees), permits companies to limit employee access to much-needed benefits. Labour brokers – or companies that act as an intermediary to recruit temporary workers on behalf of a company fearing the risks associated with hiring people on a permanent, full-time basis – have therefore become a blessing for larger corporates, and the bane of everyone else’s existence, including the government.

Comparing Metals & Mining with Retail

As per the data reviewed for IRAS’s 2014 research report, 29 of the 45 Metals & Mining companies (64.4%), and 17 of the 23 Retail companies (73.9%), recognised the significance of disclosing data about the percentage of employees who are deemed ‘permanent’, while six Metals & Mining companies and one Retail company didn’t even report how many people they employed (let alone ‘% permanent’).

Once again, the data suggests that while most companies understand the importance of providing data, the quality of what they place in the public domain remains questionable. For example, it’s highly improbable that one mining company is able to operate with only 8.87% permanent staff, unless contract employees are somehow considered ‘employees’ rather than reported separately. At the same time, it’s highly improbable that four retailers can report a labour force based exclusively on 100% permanent staff.

The issue is one of managing the ‘social licence to operate’: particularly when unions have made it clear to all concerned that they will not tolerate any practice that even suggests a two-tiered employment system. Whether it’s the string of endless strikes delaying Eskom’s Medupi and Kusile power station projects, or the five-month-long platinum strike that devastated platinum belt communities in 2014, unions have put on record an unwillingness to allow companies to underestimate the need for workers’ rights (be those rights real or perceived). Companies need to be much clearer about what constitutes a ‘permanent employee’, and be clear when presenting labour statistics to the public. If there’s a genuine need for temporary labour, inclusive of labour-broking services, then they must be clear in their description of where benefits accrued to permanent employees do not substantially differ from those of temporary workers.

8. Lost days due to strike action

Perhaps the most startling of all observations made in 2014 was the fact that only 17 of 311 companies, across all sectors, provided data pertaining to the number of days lost due to strike action. This in a country declared in an August 2014 article published by Engineering Weekly as ‘one of the world’s most violent, strike-prone countries³.’

Citing research by John Brand of Bowman Gilfillan law firm, the article declared that South Africa lost more than 17 million person hours due to strike action of which more th–an 16 million were in the mining sector – as a result of 99 strikes, 45 of which were unprotected. Between 2007 and 2011, there was an average of 65 strikes per year.

Using the international standard of measuring the number of working days lost due to industrial action per 1 000 employees (or two million person hours worked), Brand reported that Denmark, France and Belgium registered the highest work stoppage rates, with a respective 159.4, 132 and 78.9 yearly average days lost between 2005 and 2009. The global yearly average for the same period was 30.6 average days. However, for every 1 000 working South Africans, from the period 2006 to 2011, about 507 working days were lost a year due to strikes. This is 16.5 times more than the global average, and three times the rate of France.

Perhaps the cynical will assert that the reason for such a deplorable rate of reporting the number of days lost due to strike action is that nobody wants shareholders to back away for fear of investing in a lost cause. However, companies less affected by strike action ought to be much bolder in their reporting in this area. For example, Royal Bafokeng Platinum (RBPlats) reported in their most recent annual report (to be reviewed as part of IRAS’s 2015 research report) that while their largest platinum mining peers all suffered enormous losses during the 2014 platinum strike, RBPlats didn’t lose a single day, suggesting that genuine empowerment reflects a critical component of the company’s ongoing sustainability strategy.

Ultimately, all JSE-listed companies ought to recognise that the attraction of foreign investment is predicated on their ability to demonstrate publicly that adequate controls are in place to manage critical risks, one of which is being painted with a common reputational brush. Where South Africa is viewed as non-competitive from a labour market perspective, the only way to combat ‘guilt by association is to ensure that companies make available accurate and reliable data.

Concluding thoughts

In the matter of ‘wellness’, the above discourse is but a sample of how we can use comparable quantitative data to assert whether companies are – or are not – acting in a manner that can be deemed ‘responsible’ or ‘proactive’ from no less than a human capital perspective.

Without touching on country-specific indicators like whether or not companies give special consideration to historically disadvantaged South Africans (or HDSA candidates or suppliers), or on environmental or societal issues such as corporate social investment, the above demonstrates how researchers who evaluate ‘responsibility’ ought to look for information that they can use to compare companies with one another. In the same manner, companies ought to view their annual reports (and supporting online content) as an opportunity to demonstrate that they are not only ‘a going concern’ as from the last day of the previous reporting period, but that they have given due consideration to what it takes for companies to be regarded as both ‘responsible’ and ‘sustainable’.

References

1 Employee turnover (ET) is defined as the number of employees who left the company during the year relative to the number at the end of the year.
2 World Economic Forum (2014)
3 Odendaal (2014)