OP-ED

The medical schemes probe: How racial discrimination denialists get it wrong

By Pierre De Vos 20 January 2021

The industry is now being looked to, to assist in saving the people of the country from the ravages of Covid-19 by helping to finance the acquisition and distribution of vaccines.(Photo: 2oceansvibe.com/Wikipedia)

Medical schemes may be hard-pressed to convince a court that the way they are dealing with fraud, waste and abuse by medical practitioners does not constitute unfair racial discrimination.

This week the Government Employees Medical Scheme (GEMS) lost an urgent legal bid to interdict the release of an interim report (authored by a panel headed by Advocate Tembeka Ngcukaitobi) on allegations that medical schemes are racially profiling black, coloured and Indian medical practitioners and that they are therefore unfairly racially discriminating against such practitioners. Predictably, many racial discrimination denialists have questioned the legitimacy of these findings, often displaying a profound ignorance of South African discrimination law in the process.

“Discrimination” – much like “freedom of expression” – is a commonly used, but much misunderstood, term. Everybody thinks they know what discrimination means and how the law regulates it, while only a few ever take the trouble to read the relevant legal texts. No wonder then that so many people have no clue what the prohibition of unfair discrimination in section 9(3) of the Constitution (and a similar prohibition in the Promotion of Equality and Prevention of Unfair Discrimination Act (PEPUDA) actually entails. No wonder, also, that many people confidently (but wrongly) claim that redress policies are always discriminatory and amount to “reverse discrimination”. 

The same kind of ignorance has marred reaction by racial discrimination denialists to the Ngcukaitobi panel’s interim report in which it identifies clear patterns of racial profiling of medical professionals by medical schemes. Because they are working with a misguided and outdated view of what constitutes discrimination (a view rejected by the Constitution and the Constitutional Court), they do not understand that our law does not require proof of discriminatory intent to make a finding that a person or institution unfairly discriminated against anyone in conflict with the law.

Ngcukaitobi’s panel did not find “evidence of explicit racial bias in the algorithms” used to identify potential fraud, waste and abuse by healthcare practitioners, or in the methods that the administrators and schemes use to identify such practices. However, the panel found that “there is a substantial difference in fraud, waste and abuse outcomes between black and non-black practitioners over the period January 2012 to June 2019”. It also found that over “this period, across all disciplines… black practitioners were 1.4 times more likely to be classified as having committed fraud, waste and abuse than those identified as not black”. 

These findings suggest that the medical schemes may be in serious legal trouble. Black doctors (or an organisation acting on their behalf) may well approach the Equality Court for a declaration that these schemes are unfairly discriminating against black doctors on the basis of their race. For reasons I explain below, medical schemes may be hard-pressed to convince a court that the way they are dealing with fraud, waste and abuse by medical practitioners does not constitute unfair racial discrimination.

In terms of PEPUDA, medical schemes (like all other private institutions or individuals) are prohibited from unfairly discriminating against anyone on any ground, including race. The Act defines discrimination as “any act or omission, including a policy, law, rule, practice, condition or situation which directly or indirectly (a) imposes burdens, obligations or disadvantage on; or (b) withholds benefits, opportunities or advantages from, any person on one or more of the prohibited grounds.”

In terms of this definition, the focus is on the impact of the “policy, law, rule, practice, condition or situation” on the group complaining of unfair discrimination, not on the intention of the person or institution accused of discrimination. As long as it can be shown that the “policy law, rule, practice, condition or situation” had a disproportionate negative impact on the complaining group based on one or more of the 16 grounds listed in the Constitution (for example, race, sex, gender, sexual orientation), or on a ground analogous to the listed grounds (for example, HIV states and non-citizenship), the court will accept that discrimination had occurred. (Whether the discrimination is unfair – a second requirement that must be met – is dealt with below.)

As those who discriminate against others often defend themselves by claiming they had not done so intentionally, and as it can be very difficult to prove that somebody had the requisite intention to discriminate, PEPUDA’s focus on the impact of the “policy, law, rule, practice, condition or situation” (instead of on the intention of those who cause the discrimination) makes it more difficult for those who discriminate to escape legal liability. It also ensures that where racism and racist assumptions are so deeply entrenched that those who discriminate do not realise that they are discriminating by making racist assumptions about a group of people, the law will still be able to hold the discriminators accountable by focusing on the discriminatory impact of the wrongdoers.

Thus, the finding in the interim report that medical schemes were 1.4 times more likely to classify black medical professionals as having committed fraud, waste and abuse than white practitioners, in effect amounts to a finding that the medical schemes discriminated on the basis of race because the way it handled these cases disproportionately impacted on black medical professionals. 

This is not the end of the matter though. Discrimination is only unlawful if it is found to be unfair (although it is presumed to be unfair once it has been shown that racial discrimination took place). The onus is on those accused of unfair racial discrimination to prove that the discrimination is fair and not unfair. In this case, medical schemes would have to prove that it was entirely justified that black practitioners were 1.4 times more likely to be classified as having committed fraud, waste and abuse than white practitioners. 

When considering whether the discrimination is fair or unfair, the Equality Court will have to consider all relevant factors (many of these factors are listed in section 14 of PEPUDA). Most notably for the present case, it will ask whether the discrimination has a legitimate purpose and whether the importance of this purpose outweighs the negative effect the discrimination has on the complaining group.

When the purpose of the discrimination is extremely important and sensible, and the impact of the discrimination is minimised, a court is likely to find that the discrimination was fair. For example, a rule prohibiting young children from owning firearms would discriminate against children on the basis of their age, but as the purpose of protecting both the relevant children and the larger public is a rather pressing one, it is unlikely that a court will find that such discrimination is unfair.

In this case, the medical schemes will have to show that the aim of the discrimination was to root out fraud, waste and abuse by healthcare practitioners. To do so, a medical scheme would have to show, at the very least, that black practitioners fingered for fraud, waste and abuse were indeed guilty. But it would also have to show that white practitioners in similar situations were treated in the same manner as black professionals. In other words, I suspect the only way medical schemes would be able to escape a finding that they are guilty of unfair racial discrimination, would be to show that in fact, they treated black practitioners and white practitioners in very close to an identical manner when investigating them. 

Practically, it would be rather difficult for any medical scheme to prove that it investigated black and white practitioners in close to an identical way and held them accountable to the same degree. But given the fact that racism is built into the very structure of our (or any society), and given the fact that medical scheme fraud investigators are unlikely to be less racist or racially biased than the population at large, I would not be surprised if, based on the facts, it turns out that medical schemes are unable to prove that they investigated black and white practitioners in close to an identical way and held them accountable to the same degree.

The findings in this interim report should make us think about what other types of systemic discrimination may be uncovered by similar investigations in other contexts. For example, an investigation of the salaries of employees with similar experience, qualifications and responsibilities in a specific company may reveal that women and black people systematically earn less than their white colleagues. Or an investigation of lending practices by banks may reveal that they disproportionately deny bond applications by black or female applicants. 

But such investigations are costly and – in the absence of political will – are easily thwarted by powerful business interests, which means they are likely to occur less often than they should. DM

Gallery

Comments - share your knowledge and experience

Please note you must be a Maverick Insider to comment. Sign up here or if you are already an Insider.

Everybody has an opinion but not everyone has the knowledge and the experience to contribute meaningfully to a discussion. That’s what we want from our members. Help us learn with your expertise and insights on articles that we publish. We encourage different, respectful viewpoints to further our understanding of the world. View our comments policy here.

All Comments 16

  • ” To do so, a medical scheme would have to show, at the very least, that black practitioners fingered for fraud, waste and abuse were indeed guilty. But it would also have to show that white practitioners in similar situations were treated in the same manner as black professionals. ”
    But isn’t that exactly what they say and the panel conceded they could not find otherwise.
    There could be many reasons why black practitioners are proportionally more likely to commit fraud than white ones ( bear in mind that is not only black ones who were implicated), without being a race baced reason. Fur example there are proportionally more black ones who are not top earners or established and therefore more susceptible to the opportunities to engage in fraud.

    • Agreed in principle. The rate of occurrence of an event in population A may be different from that in population B, and that will reflect in sample statistics. Kahneman illustrates that people have a poor appreciation of the effects of base rates, and I wonder whether this needs more ventilation. To cite a different situation, Herman Mashaba is called a xenophobe. Now we already know that some foreigners are illegals. Secondly, with pressures to favour locals in employment, would it be any wonder if desperate foreigners turned to crime more often? How can we assume the base rates are the same? Political correctness meets statistics.

  • “Everybody thinks they know what discrimination means and how the law regulates it, while only a few ever take the trouble to read the relevant legal texts.”

    The legal domain is not the only discipline contemplating issues such as discrimination. Ethics, for instance consider aspects from various perspectives, also other angles than what is written in law, and hopefully these discussions might serve to ammend laws in time. Like another Prof once said: “It might be legal, but is it right?”

  • This report is misleading. Surely we can assume that the black practitioners “fingered” did actually commit fraud. If they did then there is no discrimination as the numbers speak for themselves and
    Ngcukaitobi is being mischevious and playing to the gallery. If they did not commit fraud but were “fingered” by the schemes then they have a case against the schemes.

  • I would love for Pierre to explain in layman’s terms what his counter to this argument is:

    The medical aid schemes use algorithms to flag potential cases, as well as whistleblowers. These people are not targeted by anything other than behaviours which trigger a warning from the algorithm, or people calling in to warn that someone is possibly committing a crime. The medical aid schemes then act, and in their sample size audit, it was found that predominantly “black” (I use “” because this includes other races, such as Indian people) are more likely to have transgressed and committed acts of fraud.

    So the panel assessed the algorithms and found no indication of bias. They questioned the medical aids on how they gather information from whistleblowers and information comes from a third party (i.e. another company that is independent from the medical aids) who gets information from whistleblowers, which is a possible area that needs to be looked at but there is currently no evidence of bias.

    So how in all hell does the panel come to the conclusion that there is racial bias against “black” doctors when there is no evidence pointing to the fact? I’m a “black” person in this country and even I can see that their argument is ludicrous!

  • Is the mere act of questioning such a report immediately enough to make someone a denialist? I find that a little warped but fully in line with the current near Kafkaesque form of Identity Politics, where the act of denying something is proof of what is being denied.

  • I find it bizarre that in 2021, medical aid schemes can still identify medical practitioners not just by “race”, but by all the categories of apartheid. I am not arguing for no redress. I just find it supremely ironic that the dream of a non-racial society is still getting bogged down by the metrics of redress. How can we remedy racism except by continuing to measure it?

  • These are orthogonal issues, and while I see the point, the argument is – from a practical standpoint – rubbish (IMHO).

    Let’s assume out of everyone 100 people, 20% will *attempt* to engage in fraud. Some will get away with it, some will be caught, but most will not persist if sufficient controls are in place. These controls can be sophisticated IT systems with automated auditing and security etc. And these controls are more likely to be employed at practices in more affluent areas – places that have access to the internet etc. So if you create an algorithm that says “look more closely at practices who don’t have these security controls, or practices who didn’t supply these audit reports etc” (and this is a perfectly fine thing to do), then just by demographics alone you would expect them to more find practices in “poorer” areas who are guilty of fraud. And this implies more non-white practices.

    However, nowhere in the algorithm profile was race ever defined. It was brought in as a correlation *after the fact*. This is where the problem is – just because there is some correlation, you now have to show that you investigated implications of this correlation equally, which is complete rubbish because this was never part of your heuristic in the first place.

    Furthermore, the media isn’t doing this report any justice at all. There is little to no context for the statistics involved. What if I find that a non-white person in South Africa is more likely to be sentenced for murder than a white person? Are courts now going to have to prove that each case was handled equally, and if not, non-whites would need to have their sentences reduced?

  • fascinating that the media automatically latches onto a sensationalist view of what are most likely simple facts. I assume that all instances of fraud, no matter the race of the possible miscreant, are investigated. If, post these investigations, it turns out that the statistics of those found guilty happen to show a greater propensity to commit fraud on the part of black practitioners, that would then simply be a fact. No racial bias at all if that is the case – simple facts!

  • Just wrong. The filter is colour blind and the process thereafter is unfair across all demographics. Those are 2 key findings. How is that systemic discrimination?

  • I thought the key factor in this whole affair is that there is a demographic basis to the differences in the ways that Doctors operate, generally, based on the communities they serve. By which I mean that there are far, far more “black” Doctors who have to claim directly from the Medical schemes for work they have carried out, because their patients tend not to be able pay for their treatment at the time they receive it & then claim it back from their schemes themselves. This leaves the Doctors claiming their fees back from the schemes themselves. So, the demographics actually lead to the “discrimination” rather than the schemes’ methods of tracking down fraud.
    Needless to say, this is not racial discrimination at all – the outcomes may not be equal & the law may treat that as discrimination but saying that there is no “discrimination” (as understood by the majority – who have NOT studied law) is not racist denialism either!

  • MAGASHULE ON THE MOVE

    Ace going nowhere, new ANC step-aside guidelines reveal

    By Ferial Haffajee