This is a tale of two committees, and the tangled web of interactions between them and the current UK government. All the actors in it claim to serve the same mistress: Science. This is not the lower-case ‘science’ of the 0.9 impact factor, or the technician looking at the bench trying to remember which DNA went into which tube. This is full blown upper-case Science. The sort of thing that ministers talk about when they appear on Newsnight to declaim, in the face of some new threat, that “what is imperative is that we follow the Science”. (“There,” they think, “that’ll shut Paxman up. Let’s see him take on Science!’

Except of course, as I have commented before, what these guys call Science often bears only a passing resemblance to what I would recognise by the word.

The question under discussion is how to deal with an increasing incidence of bovine tuberculosis. This is a big problem, especially for an agricultural industry under the cosh. It is driving farmers to the wall in areas where it is common (like the South West of Britain) and the cause is clear. Badgers. Yes, friendly old Brock is not so much Wind in the Willows as La Boheme, wheezing away consumptively in his lair before popping out to find a cow to cough all over¹.

Given the fact (and I am afraid it is so, animal lovers) that badgers can and do give cows TB, it seems that control of the badger population is a reasonable way to get at this. And control is a euphemism for killing. Enter Committee number one, stage left. This is the Independent Science Group² or ISG (note the S-word, right there in the middle), convened almost a decade ago to oversee a trial of three different culling strategies: what happened to TB in cattle if you did nothing; what happened when you went out and killed all the badgers you could lay your hands on; or what happened if you only got medieval on their little furry black and white asses in response to a case of TB in a herd.

The upshot of this was that killing badgers displaced the TB problem to adjoining areas where there was no cull, by interfering with the natural territory structure and causing surviving badgers to range over larger areas, infecting more cattle and indeed other badgers.

This is an awkward finding. Yes, badgers do transmit TB. No, culling them doesn’t make things better, and could in fact make things worse. There is a light in the tunnel though – it is easier to control TB in cattle. After all, we know where they are and it is easy to catch and test them, and indeed cull them if necessary. The report suggested that cattle alone are responsible for the majority of disease (badgers only account for up to 40%) and measures targeted at them could be effective.

This created great wailing and gnashing of teeth among the farming community, entertainingly summarised in this blog. However, how can you argue with such a meticulously designed trial? Overseen by a panel of professors dripping with academic kudos?

Well, bring on David King, the Chief Scientific Adviser to the UK government, a man with a distinguished record in chemistry, and charged with the presentation of and analysis of scientific argument to ministers. And after a day and a half talking over the ISG report with some other scientists, he came to almost the exact opposite conclusion: culling of badgers is apparently essential.

This is very strange, to say the least. Two groups of distinguished scientists produce such divergent opinions. How on earth is this possible? How can we reconcile the two?

The same question occurred to the relevant select committee of MPs, who questioned King closely on this topic when they met a few days later. Confusingly, he claimed that there was little or no disagreement between the two reports. Having looked at both of them, this seems a pretty odd definition of ‘no disagreement’ to me.

However, there are some subtle differences in the style of the reports. The ISG’s is a sledgehammer of 199 pages, minus appendices, which is meticulously, not to say in places pedantically and somewhat turgidly, written and argued. Not unlike your average scientific paper in other words. In contrast, King’s report is only 18 pages long, and speaks more in generalities. For example, it suggests a cull in “areas of the country where there is a high and persistent incidence of TB in cattle”, without defining exactly how high or persistent this might be. Elsewhere “strong action” is called for, with no regard for what counts as ‘strong’ – impressive looking, or demonstrably effective? King’s contribution has been heavily criticised, both by members of the the ISG and others. In fact, to me it seems more vague than anything else. The sort of thing which you might expect from a government office, with plenty of handles to allow it to be spun as required.

More importantly, King’s report contains a statement which borders on statistical illiteracy. Annex one deals with confidence intervals. These are statistical values for things like reduction in disease, estimated from the data, in which you can be 95% sure the true value lies. The ISG reports many of these, and they are frequently quite large, reflecting the noise in the data. Through confidence intervals, the reader is always kept aware of how (in)secure the data under discussion are. It is hence rather odd that King criticises the ISG for providing figures containing decimal points because “[G]iven the often very wide confidence intervals, I consider that this level of precision may give the impression of more certainty than is the case.”

But the confidence intervals have been presented, and so the reader is always exactly aware of how much certainty should be attached to any particular figure. I really have no idea what he is on about here. I ran King’s comments past a few undergraduates, not telling them the context, and they immediately leapt on them as nonsensical. When I told them the source, they were surprised and a little worried. Me too. Perhaps even more strange, when asked about cost effectiveness of a cull, King implied that this was not a scientific issue – which would be news to all the statisticians who crunch scientific data in order to answer questions such as this.

I find it hard, despite statements to the contrary, to see this as anything other than a means of providing wriggle room to ministers. Now they can order a cull, or not, and go on Newsnight claiming they have the Science to back them up. Of course, the fact that Science also could be used to point in the exact opposite direction will be forgotten.

It looks to me like we need a Chief scientific Adviser, rather than a Chief Scientific Adviser.

Related information

1. Please note that I am not actually suggesting that this is the mode of transmission. It is a joke.

2. I should declare an interest: I work in an academic department alongside Christl Donnelly, one of the members of the ISG. I can, however, honestly declare that I have not discussed these issues with her, and that what follows are entirely my own views and I am responsible for them.