BMJ Open, along with a couple of other journals, published a statement a couple of days ago saying that they’d no longer accept papers based on research wholly or partially funded by the tobacco industry.  The gloss on the statement is damning:

The tobacco industry, far from advancing knowledge, has used research to deliberately produce ignorance and to advance its ultimate goal of selling its deadly products while shoring up its damaged legitimacy.  We now know, from extensive research drawing on the tobacco industry’s own internal documents, that for decades the industry sought to create both scientific and popular ignorance or “doubt.”  At first this doubt related to the fact that smoking caused lung cancer; later, it related to the harmful effects of secondhand smoke on non-smokers and the true effects of using so called light or reduced tar cigarettes on smokers’ health.  Journals unwittingly played a role in producing and sustaining this ignorance.

Some who work within public health and who buy the notion of “harm reduction” argue that the companies that now produce modified cigarette products and non-cigarette tobacco products, including electronic nicotine delivery devices (e-cigarettes), are different from the tobacco industry of old, or that the tobacco industry has changed. For “hardened” cigarette smokers who can’t or won’t quit cigarettes, the argument goes, new tobacco products could represent potential public health gains, and company sponsored research may be the first to identify those gains.

But one fact remains unassailably true: the same few multinational tobacco companies continue to dominate the market globally and, as smaller companies develop promising products, they are quickly acquired by the larger ones. However promising any other products might be, tobacco companies are still in the business of marketing cigarettes. As US federal court judge Gladys Kessler pointed out in her judgment in the case of US Department of Justice versus Philip Morris et al, the egregious behaviour of these companies is continuing and is likely to continue into the future.  And just this summer documents leaked from one company showed a concerted campaign to “ensure that PP [plain packaging of tobacco products, bearing health warnings but only minimal branding] is not adopted in the UK.”  The tobacco industry has not changed in any fundamental way, and the cigarette—the single most deadly consumer product ever made—remains widely available and aggressively marketed.

What should we make of the policy?

A bad argument against the ban – yeah, I know that that misses some linguistic subtlety, but it’s close enough – is that it’s a violation of free speech: it really is no such thing, for the simple reason that noone is trying to stop the tobacco industry making its case – a right to free speech doesn’t imply a right to a platform.  Of course, if every reputable publisher denies the industry a platform, than this might be a de facto rather than de jure curb on free speech – but that’s just the way it goes: just as noone gets to insist that a particular person gives them a platform, they don’t get to insist that they be provided with one at all.  (Also – though it doesn’t apply in this case – merely to splutter “B…b.. but free speech!” isn’t an argument anyway.)

Still, I guess I am uneasy about a ban.  There’s a range of reasons for this.  One of these has to do with a kind of fallacy about repeat offenders.  Let’s allow that the tobacco industry has a record of unreliable research – that’s not all that much of an allowance.  More: let’s allow that the industry has been systematically dishonest about what it does, and its funding habits have been an extension of that dishonesty.  Yet we wouldn’t be entitled to say based on that that every instance of research was unreliable: one swallow doesn’t make a spring, but the fact that it’s spring doesn’t mean that everything’s a swallow.  And, besides: the fact that research has been unreliable in the past isn’t an indication that future research will be; one would prefer to think, surely, that a given piece of research should be taken on its own merits, rather than on the merits of the person who did it, or the merits of the last piece of research.

So there’s at least a chance that there might be decent research carried out on tobacco industry’s dollar – and the slim possibility that there might be some hitherto unrecognised benefit that’s derived from tobacco.  There’s even a chance (vanishingly small, I admit, but not nothing) that we’ve got tobacco wrong all along.  Science sometimes works like that.  More worryingly, if we’re concerned that tobacco-funded research has an agenda, shouldn’t we also entertain at least the vague possibility that research funded by, say, a cancer charity also has an agenda?  And if we can adjust for the latter, why not the former?  Along these lines, one might say that journals have a duty to be disinterested, and that the policy violates that duty.

I think that these worries are non-negligible; but they don’t necessarily tip the balance for all that.  This is partly, but not wholly, because I don’t think that there is a duty to research anyway – and so to forego the opportunity to publish research isn’t anything more than just that: foregoing an opportunity.  So even if there is good research that won’t get published… well, meh.

And while it might be the case that we’ve got tobacco all wrong, or even that it has some minor mitigation, as it stands, the overwhelming balance of probabilities is that we haven’t.  And note that the policy is not that no papers will be published that question conventional wisdom about tobacco – those would normally have to pass a fairly high bar anyway (for reasons related to Hume’s reasoning concerning miracles: there’re certain things that we’re entitled to discount, even if not ignore, just because believing them would require such a grand suspension of disbelief), and it hasn’t been raised; it’s just that tobacco-industry sponsored stuff is ruled out.  That actually makes unexpected results more trustworthy on aggregate, because it’s more likely to be disinterested.  Industry-supported research is more likely to be favourable to the industry that supports it, not least because a sponsor might have the ability to sit on research that isn’t favourable.

The biggest worry has to do with the possibility of drowning out an unpopular bias with a popular one.  But it has to be said here that cancer charities do not have a track record of distorting research; even if there is the occasional dodgy paper – one ought never to say never – there’s no evidence of anything like the practices of big tobacco.  Of course, this takes us back to the point about the past not providing evidence of this particular piece of research.  But still – there’s room to be at least slightly wary of tobacco-sponsored research in a way that doesn’t apply to the other stuff.  Cancer charities don’t have a profit incentive, and not-smoking isn’t an addictive revenue-source, for one thing.  So even if you think that the BMJ statement is a bit too righteously angry, the general point stands.

Hence that particular line of attack on the ban, I think, fails.

What we do have reason to believe is that tobacco is harmful; and those with a fairly obvious interest in selling the stuff also have a corresponding interest in pushing research that talks down those risks.  Hence there’s a conflict of interest.  And since there’s no obligation to publish to begin with… well, it’s probably no great loss, all told.

Should we be wholly at ease with the policy?  Probably not.  But the idea that we should always be at ease with moral decisions is fallacious.  I think that the policy is, at the very least, defensible, and could well be right.