The dangers of smoking cannabis are hotly debated
Different ways of seeing stats
Scientific advice may suggest how dangerous things are - like smoking cannabis, taking ecstasy and horse riding - but risk is not all about numbers, says Michael Blastland in his regular column.
Sir David King, a former chief scientific adviser (not to be confused with Professor David Nutt of cannabis, ecstasy and horse riding fame - we'll come to him in a minute), said the government should press ahead, urgently, with a new generation of nuclear reactors because of the threat of climate change.
Professor Nutt was sacked from his job as drugs adviser
Two conclusions in just one policy, but sufficient to skewer many people. Some will agree with his assessment of the threat to the climate, while recoiling at his remedy. Others will be comfortable with nuclear power but not share the rationale. Sir David also famously argued that genetically modified crops were safer than conventional ones.
If you find yourself resisting any of the policy implications here, you are guilty of rejecting scientific advice. For though opinions vary, Sir David was, after all, the chief scientific adviser.
The point is a simple one, that people's attitude to the validity of scientific advice often shifts according to how congenial they find it. It's a point worth bearing in mind in the case this week of Prof Nutt versus the Home Office on the classification of cannabis and other drugs.
For though there is a statistical basis to Prof Nutt's claim that ecstasy is no more dangerous than horse riding (though even these numbers can be interpreted differently and are contested), and he seems correct that the health risk of drinking and smoking is greater, risk is not now and never has been a question only, or even primarily, of numbers.
The numbers begin in sources such as a catalogue of deaths, this one for England and Wales.
Deaths by drug poisonings can also be found separately and more easily at the
Office for National Statistics
The tables tell us, among myriad morbid details, that through a variety of causes 23 people drowned in the bath in 2008, 108 died from "inhalation of gastric contents" and five starved to death.
Of course, death is not the only measure of the harm drugs can do. Prof Nutt's team considered eight others - including the likelihood of psychotic illness - to arrive at their suggested classification.
To see why even this exhaustive effort will struggle to be definitive, here is a true story.
Sixty years ago, a young woman named Mary Douglas studying tribal custom in Africa, came across the following belief: if a woman had been unfaithful, she risked miscarriage in pregnancy. This was viewed not as bad luck or magic, but as a law of nature, as today we might regard gravity.
At the time she thought the tribe primitive, exercising a crude means of asserting power and ignorant of science. Later, Douglas, a celebrated anthropologist, came to the view that their behaviour was not primitive at all but normal in all societies. That is, what we call "risky" is often an expression of disapproval - though we might not be aware that this is what it is - as much as a measure of real danger.
Her work suggests that even if Prof Nutt is right about the numbers - that there are many nasty things society smiles on and this is inconsistent, hypocritical you might say - this is only the beginning of the baggage carried by the concept of risk.
Risk is also about power and taboo, about cultural habits and our sense of control, about class, prejudice and values. Among these cultural complications is the fact that many people find "unnatural" risks somehow worse than natural ones, even though swine flu and arsenic are natural enough. Risks also depend on the political implications we attach to them. Otherwise, why is a suitcase left at an airport frightening?
Risks we hope to control are thought less nasty than those imposed on us, so we might be paranoid about pesticides in our food but sometimes unconcerned as we spray the garden where the children play. Spectacular but highly unlikely events worry us more than routine and more likely hazards that are just around the corner.
For some, of course, risk is an allure. Tell them it might be dangerous and they can't wait to try. That is one reason they climb big mountains, drive cars too fast, smoke underage. There is evidence that the more we try to protect some people from the risks they take, the more danger they seek.
This is known as risk homeostasis and acts like a personal thermostat, so that if someone gives you an airbag, you feel more at liberty to put your foot down. (Funnily enough, one psychological trait among the young is a tendency to overestimate the likelihood of bad things happening. They think the chance of HIV-Aids is far greater than it really is. But does that stop unprotected sex?)
There can also be confusion between the probability of a harm occurring and the severity of that harm. So if a chief scientist says nuclear power isn't risky, he might mean it's highly unlikely to go wrong. Someone else might think, "but if it does..."
And no measure of risk is complete without a calculation of the potential benefits.
For all these reasons, when in answer to either of the scientific Davids the government demurs, does it have a point that this is politics, not science? It has produced a classification list which is ostensibly about degrees of harm, but might also be a list of "things politicians think voters don't like". Or should we dismiss cultural considerations as anti-science?
When the classification system is designed not only as a ranking of harm, but as the basis of a policy that tries to satisfy public preferences, maybe it tries to do too much, stretched between two conflicting objectives, one about numbers (approximately), the other about culture (in all its contradictory vagueness). Maybe that is why the sides fail to see eye to eye. As with all measurements, it pays to be clear what we are measuring.
Below is a selection of your comments.
The thing is, with most high risk activities that don't endanger other people (think base jumping, for example), the person who is at risk is considered to be at liberty to decide whether they want to take that risk and for that purpose they should be fully informed. Unless they are considered by a Dr to be rationally incapable of making that decision then they're allowed to do it. No matter what the risk.
So why should drugs be any different?
John Stewart, Worcester, England
This article serves to reinforce the belief that we somehow have some kind of knowledge which is apart from, but as useful as, science - namely "culture (in all its contradictory vagueness)". This is utterly and shockingly false. If that "cultural knowledge" exists (which it does) then if it is to be useful it must be measured (which it can be). This process yields scientific results, and it is this analysis which we use to know things about the henceforth mysterious "cultural knowledge". An example of *exactly* this is featured in Professor Nutt's lecture - where he details the survey they undertook of people's attitude to drugs and drug classifications. It is unsurprising that the leading expert in this field undertook the required research to demystify this area of knowledge - it is, however, surprising that the author of this article didn't think to examine his work before writing about how it needs to be done.
The aforementioned survey shows exactly the "cultural knowledge" this article refers to - the public perception of harm and the public perception of the use of the classification system as a deterrent - boiled down to useful, scientific results. This knowledge was part and parcel of that lecture - there was no distinction made between science and "other knowledge" - as that distinction does not exist. The fact is that the "cultural knowledge" which the author seems to think should be given equal weight with science is already a part of the science. There is no distinction. "Should we dismiss cultural considerations as anti-science" - no, we should quantify them and integrate them as an important part of the bigger picture - as the professionals already have been doing.
This article is fundamentally incorrect. Risk is a definable quantity that can indeed be expressed purely in numbers, and is done so by people like me on a daily basis. The HSE provides information on how to quantify risk. I agree with the point that this risk must then be interpreted against our desire to accept it or otherwise based on qualitative factors. Risk is a combination of how often an undesirable event occurs and what the consequences of its occurrence are. So the frequency of a nuclear meltdown is very low, but the consequences can be large. This may or may not result in a low risk. Equally, less significant events may occur more often. It may be that this is more of a concern.
The "scientists" (I hate that expression by the way) may state numerical fact, but it is up to the politicians to interpret it. The problem in this case appears to be that the politicians start with an agenda and fire the experts when the numbers they generate don't prove the desired outcome.
Chris C, Manchester
Here in Australia we are required to wear bicycle helmets when cycling (although the law is not enforced much). The rationale is that this will lessen head injuries - this is backed up by accepted facts. However people in cars suffer more head injuries than cyclists and would benefit even more by wearing helmets (yes, inside the car - as rally drivers do). But of course the government would not dare make that compulsory! As you point out, politics influences risk perception and action.
Terence, Sydney, Australia
There is also the rate of luck burn to consider. One trip on the Space Shuttle carries about the same risk as commuting to work for 10 years on US highways. You find people who say "I would never risk the shuttle" happily taking an equal risk over a longer period of time with no thought the risk is statistically the same. Happy luck burning; I can guarantee you will run out in the end.
Sandy Almond, Knoxville, Tennessee USA
How many children die each year in road accidents on their way to and from school? I suspect it's significantly more than are murdered by strangers they meet on the internet, say, yet no one is suggesting we shouldn't send our children to school! Our perception of risk is heavily influenced by fear of the unknown, moral judgements, and prejudice. That's not new. What's new is the amount we obsess about risk avoidance, and the often irrational steps we'll take in the interests of making the world a "safer" place.
Rolf Howarth, Warwickshire
This article makes a good point. Certainly we seem to live in a society which has managed to steer its way to accepting some things without too much discomfort yet turns to hysteria at the thought of other things. prime example is alcohol versus cannabis. I'm fairly sure there is more cost to the taxpayer as a result of abuse of alcohol (in all the ways it can be abused) than use of cannabis. Embracing alcohol or at least being passive about it incurs health costs, the societal cost of unpleasant behaviour, cost to our youth etc. Yet we all seem capable of accepting it at a general level. The same is not true for cannabis yet the impact upon society I'm sure is much less destructive. It's a bit of a strange position to have arrived at.
What you're describing here is superstition, ie fake risk. That's just what your "cultural risk" is - superstitious mumbo-jumbo. No, there's no such thing as "unnatural". All REAL things are natural. Any explanation which ISN'T natural, we call "supernatural" - magical unicorns, Santa, the Tooth Fairy and so on. I shouldn't need to say it, but these aren't real. The reason all sides aren't seeing "eye-to-eye", then, is that politicians are away with the fairies.