Friday, 19 July, 2024
HomeTalking PointsSA's alcohol bans were not 'not based on credible science'

SA's alcohol bans were not 'not based on credible science'

Research challenging the conventional wisdom over the South African government's alcohol bans has been met with “spittle-flecked fury” from the anti-alcohol establishment, writes Peter Bruce, former editor of Business Day and the Financial Mail.

Bruce writes in his regular column that the government’s ban on alcohol sales was fuelled by prejudice, not science:

However much the government’s absurd and ultimately useless coronavirus lockdown alcohol bans have cost, it has been briefly all worth it watching our sanctimonious anti-alcohol establishment explode with spit-flecked fury at the arrival of new research that very effectively challenges their conventional wisdom that alcohol bans kept hospitals going during the 19 weeks of bans the government has imposed so far.

Distell, one of South Africa’s biggest liquor producers, funded independent researchers Ian McGorian, a data analyst with Silverfox Consulting, and Mike Murray, associate professor in the School of Mathematics, Statistics and Computer Science at the University of KwaZulu-Natal, to pin down the actual facts about alcohol and the lockdowns. Why assume the alcohol bans are what is working when there are simultaneous restrictions on movement (the curfews), gatherings and travel?

Their report, peer-reviewed by one of the country’s leading statisticians, Professor Graham Barr of the University of Cape Town (UCT), is explosive. It is obviously necessary to make clear that the exercise was funded by Distell and that its critics have rounded on that fact. But I don’t think it matters. The facts they produced are powerful and the responses so far have been pathetic.

The liquor industry had little choice but to produce this report. The government has repeatedly failed to engage with it and the raw patient data of hospital admissions since the pandemic hit South Africa have been kept a secret. But the costs to society have been staggering – about R36bn in lost retail sales to date, according to reports, and about R30bn in lost tax revenues.

McGorian and Murray try to pin down a pattern using five hospitals in the Western Cape. Some caution is required because, as they say, “the assertion that alcohol bans … are causal in the reduction of trauma is now firmly entrenched and repeated regularly by senior members of government” and in the mess of informal information coming out of hospitals. They establish a series of scenarios with the information they are able to get and conclude that no matter how you mix up the variables the curfew or movement restriction always dominates.

This is not surprising. In the UK (-57%), the US (-54%), Italy (-56%) and Ireland (-62%), restrictions on movement and gathering also cut trauma admissions dramatically. In South Africa a working estimate seems to be a 60% drop in trauma admissions. So we are bang in the middle. Except the others experienced those falls without shutting down an entire industry, ruining lives and losing vital revenues. So what has happened here? Essentially nothing.

There’s no way the alcohol bans in South Africa have been based on credible science. They’re based on prejudice. The industry hopes its data will help open some sort of dialogue with the government, but I think they’re mad. No-one in the anti-alcohol establishment will talk to them. They have too much to lose. I think they should sue the state for lost revenue.

It seems clear on reading this report that there is no clear evidence whatsoever that banning alcohol works at all in reducing hospital trauma admissions. The researchers look at the 2019/2020 SA Police Service report, which says alcohol is “confirmed” in only 5.4% of sexual offences, 5.3% of assaults and 6.7% of murders. Only 5.5% of fatal road accidents can be attributed to alcohol.

The Global Burden of Disease, a collaborative study involving more than 3,000 researchers from 145 countries, has in its 2019 report South Africa improving, not getting worse, on a range of diseases normally associated with heavy drinking: cirrhosis of the liver, liver cancer, pancreatitis and alcoholic cardiomyopathy.

I contacted Barr to ask him what he thought of the research he had been given to review. “I would rate it as good,” he replied, “especially given the data constraints.”

I quoted Professor Charles Parry, an anti-alcohol campaigner and director of the Alcohol, Tobacco & Other Drug Research Unit at the SA Medical Research Council from the weekend’s Sunday Times. As the raw data on trauma admissions from the five Western Cape hospitals had not been released, Parry asked “where did they get it from … the logical conclusion is that they have imputed the data from screen grabs of graphs that have been released periodically”.

Given that Parry does have access to the raw data, I asked Barr what he thought of that. “Somewhat disingenuous,” he replied. “The authors were desperate for the raw data, but the health department wouldn’t release … So the data that was used had to be obtained from a screenshot. This is not ideal, but not a deal-breaker; it clearly lends a certain amount of extra uncertainty to the process, but because the data can be pretty closely estimated from the screenshots, an acceptable amount.”

Barr, now emeritus professor at UCT, taught statistical science there for 40 years until 2018, produced 80 peer-reviewed articles and himself peer-reviewed hundreds of other papers. His speciality was the application of statistics to economics, or econometrics. I’ll go with him on this.


This article was originally published on BusinessLIVE and is republished with kind permission.


Full Business Day opinion (Restricted access)


See also MedicalBrief archives:

Critics say SA alcohol ban study ‘lacks methodological detail’


MedicalBrief — our free weekly e-newsletter

We'd appreciate as much information as possible, however only an email address is required.