Tuesday 31 January 2017

Your opinions are assertions! My opinions are science!

This Guardian headline caught my eye on Twitter today:

Alcohol industry subverting science to prevent greater regulation, study finds

What had the drinks companies being doing, I wondered. Denying the link between drinking and liver cirrhosis? Disputing the number of drink-driving fatalities? Intrigued, I looked up the study in question, which was by some Australian academics, and found myself reading one of the weirdest things I have seen in a journal for some time.

One of its authors is Kypros Kypri who told the Guardian:

“While the tobacco industry can’t argue against the science anymore, the alcohol industry continues to, even though people don’t die from tobacco overdose but they do from drinking and they don’t become violent when they smoke but they do when they drink.”

Kypri is one of the many psychologists who are now on the 'public health' gravy train. Three of his four co-authors are also psychologists (I can't work out what the other one is) and they are united in a near-religious belief in the cause of lifestyle regulation. Theirs is way, the truth and the life. Anybody who disagrees with them is a liar, a deceiver and deserves to be cast out of decent society.

The topic under discussion is the regulation of alcohol advertising. For Kypri and friends, it is self-evident that alcohol marketing kills people, has no benefit and should be banned. Those who disagree are not expressing a contrary opinion based on a different reading of the facts or different values. They are engaged in 'framing' as part of a 'strategy to influence policies in ways likely to protect or generate profit.'

Their paper, published by PLos One and known hereafter as Martino et al., is loosely based on a 'systematic review' (which was nothing of the sort) by the thick-as-mince quackademic Anna Gilmore and her chums in 2014 which looked at tobacco industry lobbying. Gilmore's paper (Savell et al. 2014) shockingly revealed that tobacco tactics included such dirty tricks as trying to persuade politicians that 'the proposed policy will have negative unintended consequences, that there are legal barriers to regulation, and that the regulation is unnecessary'.

In other words, they make the kind of arguments that any industry - or group of individuals - would make if the government was pursuing a vendetta against them. It is lobbying, and shouldn't come as a surprise to people in 'public health' who spend half their lives complaining about industries lobbying (they spend the other half, er, lobbying).

Martino et al. are critical of Savell et al. for being too soft on Big Business. For them, any opinion that conflicts with 'public health' dogma is beneath contempt and, they say, should not even be dignified with the word 'argument'.

While Savell et al. use the term ‘arguments’ to refer to assertions or claims made by the tobacco industry in support of its position in favour of or against particular policies, we suggest a more suitable label. The word ‘argument’ is generally understood to refer to a connected series of propositions intended to establish a conclusion [33]. The validity of the conclusion depends on the veracity of the propositions and the soundness of the logic linking them. In our view, by using the term ‘argument’, Savell et al. elevate what are, almost without exception merely claims or assertions, to a status they do not deserve. Accordingly, we use the terms ‘claim’ and ‘assertion’ interchangeably in our analysis reflecting the class ‘argument’ in Savell et al.’s system.

After this Stalinist statement of intent, Martino et al. explain their assumptions. Sort of.

We started with the assumption to usis [sic] the reader meant to make of this? er, [sic!] cite evidence to support the claim. [sic!] should be made.n [sic!] of participants are likely to that corporations’ framing of alcohol problems, scientific evidence, and government policies, is part of a strategy to influence policies in ways likely to protect or generate profit.

It does not reflect well on PLoS One that this gibberish was published. The authors seem to have left in some words from the peer review in which the referee understandably asks what the reader is meant to make of this. What indeed? I think the authors were trying to say this:

'We started with the assumption that corporations’ framing of alcohol problems, scientific evidence, and government policies, is part of a strategy to influence policies in ways likely to protect or generate profit.' 

If so, their starting assumption is also their conclusion. Everything in between is padding. Viva academia.

Our intrepid seekers of truth claim to have identified 'five overarching frames to oppose increased alcohol marketing regulation'. These are the five claims:

1) Regulatory Redundancy, ie. the regulation is unnecessary
2) Insufficient Evidence, ie. there isn't good evidence that the policy will work
3) Negative Unintended Consequences
4) Legal, ie. the regulation breaches a law or the constitution
5) Corporate Social Responsibility, ie. not really an argument but refers to companies acting in a cuddly way by, for example, supporting charities

Martino et al. insist that these arguments are so horribly dishonest that they should not be called arguments. In other words, they are calling them lies. Are they right? Let's take each in turn:

The alcohol industry claims its marketing targets only adults, however, research shows that young people are also exposed to this marketing and are negatively affected by it [1, 2]. If the industry genuinely wishes to target only adults, the self-regulatory codes should include restrictions on sport sponsorship, outdoor media and product placement in films and music videos.

This is a misreading of what 'targeting' means. One would target children by advertising during children's TV programmes or using children's cartoon characters. Martino et al. portray any advertising that is seen by children as being targeted at them, whereas the target market for a beer commercial shown during a cricket match, for example, is adult males, who make up the bulk of the audience and, not coincidentally, the bulk of the beer market. Martino et al. are not actually complaining about alcohol adverts being targeted at children. They are complaining about children ever seeing alcohol adverts at all, or even just seeing alcohol ('product placement in films').

The alcohol industry repeats the mantras that ‘most people drink responsibly’ and that alcohol consumption can be ‘part of a healthy lifestyle’ [43], claiming then that the majority of the population should therefore not be ‘punished for the sins of the few’ through policies that reduce the promotion of alcohol [44].

The first two points are demonstrably true and the third point is a reasonable, mainstream opinion derived from those facts.

It was claimed within industry submissions that there was insufficient evidence to link marketing of alcohol products to increased alcohol consumption, and therefore, that marketing regulation would have no effect (eg #10). Some specifically stated that more research would be needed to prove this link.

Although this view conflicts with 'public health' dogma, it is consistent with the evidence. The 2014 Cochrane Review that looked at alcohol advertising concluded:

There is a lack of robust evidence for or against recommending the implementation of alcohol advertising restrictions.

Awkward. Remember that this is what the Guardian describes as industry 'subverting science'.

Martino et al. continue:

Submissions cited Australian government research purporting to show a decline in alcohol consumption in minors and pregnant women [37] and claiming “…there is no evidence to suggest that alcohol problems are on the rise which could justify further regulatory constraints on the alcohol industry” (Brewers Association of Australia and New Zealand (BAANZ)).

I don't know why the word 'purporting' is in there. They cite the government research which does indeed show a decline in alcohol consumption. Perhaps Martino et al. do not trust official statistics either? In any case, the claim is correct, as is the claim about the lack of evidence that alcohol problems were on the rise in Australia at that time.

Like the tobacco industry, the alcohol industry proposed that there is insufficient evidence to show that marketing influences consumption, asserting that it merely affects brand loyalty.

This is true, regardless of what single issue campaigners want to believe. It is true of all established products. If industries could grow the market by advertising, they would spend far more on advertising than they do.

Some asserted that alcohol production provides substantial economic benefit and employment within Australia, and that increased marketing regulation would adversely affect the economy.

Alcohol production clearly provides employment. It also provides economic benefit by any conventional economic analysis. Whether marketing restrictions would adversely affect the economy is somewhat contestable but given the benefits advertising brings in terms of lowering search costs and generating competition, this claim is highly likely to be true as well.

Martino et al. try to brush off the economic benefits by quoting a ludicrously high estimate of the cost of alcohol to society but, as have been explained umpteen times - most recently in a report from Menzies House - such estimates are not monetary costs, are not external costs and cannot be compared to the economic benefits associated with alcohol.

Biased public health advocates was a newly identified type of claim within the Insufficient Evidence frame. Submissions asserted that the Expert Committee on Alcohol, with whom ANPHA consulted to develop this report, was biased and anti-alcohol and that the research referenced in the report was not scientifically valid (eg #13). For example: “Lion believes that ANPHA should be careful to distinguish between research that is the best available, expert, peer-reviewed research and surveys that are produced by anti-alcohol activists…” (Lion).

One only needs to read Martino et al.'s study to find evidence of biased public health advocates. It is a bit rich to complain about ad hominem attacks from your opponents when your entire 'study' starts from the premise that your opponents are inherently dishonest.

Incidentally, they add:

Attacking the credibility of public health advocates in submissions to government appears to be increasing [58].

Their evidence that this 'appears to be increasing' (reference 58) is a study that was written 15 years ago.

A set of claims was also framed around the notion that increased regulation has negative unintended consequences. The key themes were: 1) manufacturers, who would, as consequence of regulation, have trouble maintaining or increasing market share (eg #14), or have difficulties introducing new brands (eg #15); 2) employment in associated industries (eg #16a, b); 3) loss of public revenue from alcohol tax and the alcohol industry’s direct contribution to the Australian economy (eg #17); and 4) loss of consumer sovereignty (eg #18).

It is a rare policy that doesn't have some unintended consequences, but this can never be admitted in 'public health' because it is a cult that sees itself as infallible and must portray every regulation as a win-win. In this instance, it is blindingly obvious that advertising bans make it more difficult for companies to fight for market share and launch new brands; only a moron who doesn't have any understanding of advertising whatsoever could fail to see that. Banning a major industry from advertising is also likely to lead to unemployment in the advertising industry. That, too, should be obvious.

The only dodgy part of the claims cited in this paragraph is the idea that the government will lose revenue from alcohol taxes if advertising is banned. This implies that the industry expects the market to shrink if advertising is banned, but if you look at the source (#17) you can see that the Brewers Association is not saying this at all, it is merely reminding people that it contributes to the economy. The source they cite doesn't even mention advertising. Martino et al. are playing fast and loose with the facts again.

Two claims were identified within the Legal frame. Some submissions asserted the need for a Regulatory Impact Statement before proposing new regulation, for example “…any proposals to further regulate alcohol advertising needs to clearly demonstrate that the social and economic cost it potentially introduces are outweighed by the benefits in an environment where rates of “at risk” consumption and harm are either stable or in decline. (#20, Winemaker’s Federation of Australia).

If it is a devious industry trick to ask for evidence that the benefits of a policy will outweigh its costs, I hope more politicians fall for devious industry tricks. Seems like good law-making to me.

Some submitters claimed they were ‘socially responsible companies’ by presenting involvement in efforts and programs to reduce harmful consumption. For instance, “Recent examples of our social responsibility initiatives include…a social marketing campaign, using the strapline ‘Don’t see a good night wasted’, aimed at 18–25 year olds socializing in and around licensed venues in Sydney. (#22; Diageo). Some emphasised their membership of DrinkWise (eg #23a; an industry funded ‘social aspects/public relations’ organization (SAPRO)), as evidence of their commitment to Corporate Social Responsibility [11]. For example, Lion stated that it “…is committed to…funding culture change initiatives, such as those developed by DrinkWise…”.

This isn't an argument at all and insofar as it is a claim, it is true. I suppose the companies are showing their credentials and trying to make a good impression, as most people would if they were dealing with a powerful government.

Having discussed a bunch of supposed lies that are either demonstrable facts or reasonable opinions, Martino et al. finally come to their conclusion...

Alcohol industry actors used multiple strategies to push their claims that increased marketing regulation in Australia is unnecessary, including claims that: there is ‘insufficient evidence for the effectiveness of increased regulation’; ‘there is insufficient evidence that alcohol marketing contributes to drinking’; ‘current regulation is satisfactory’; ‘there is no community concern’; and that ‘the alcohol industry markets its products in a way that minimise harmful consumption’. 

The five statements are, in order, (1) True, (2) True, (3) Matter of opinion, (4) Depends on circumstances, (5) Matter of opinion.

These assertions, at least regarding health, stand in contrast to the scientific literature regarding alcohol-related harm and continuing high levels of alcohol consumption in the community.

No they don't, but that doesn't matter to Martino et al. because the name of the game is censorship. And here it comes...

The science reveals the poverty of industry claims that industry actors put to public servants whose job it is to evaluate submissions... Continuing to engage with industry as stakeholders in public health policies increases their opportunities to present such claims.

Or, put bluntly: "Ban them from lobbying and listen only to us."

And that is your evidence, from a peer-reviewed journal, no less, that the alcohol industry is 'subverting science' (The Guardian). I don't know whether to laugh or cry.

No comments: