Breathalysers and drunk tanks

This morning I saw a story about breathalysers having been used in Norwich city centre to test people entering nightclubs.  I wasn’t excited or outraged by this, just confused.  I couldn’t see how it could be helpful.  However, it’s been bouncing around in my head, and after some helpful Twitter responses, I seem to have come round to thinking that it might be worthwhile after all.

Most of my posts on this blog are somewhat reactive and negative – talking about something that’s annoyed me, whether it’s Brighton’s focus on happiness in their Public Health strategy, the government’s decision to move substance misuse services into local authority Public Health teams, or the tactics of both the alcohol industry and public health campaigners.  The list could go on.  So it’s nice to actually write something vaguely positive for a change – and to feel that I’ve listened and had a think and changed my mind.

In the Norwich example, one of the points that made me initially sceptical was that as far as I could tell (and I watched the video through twice to try to make sure), there was no limit at which people would be turned away.  It seems more likely that bouncers would focus on the usual techniques of judging whether someone is drunk by their behaviour.

This raises all sorts of questions about the adequacy of breath tests.  First, there’s the claim that they are not an accurate indication of pharmacological intoxication.  Then, there’s the issue of whether a test that measures one’s blood alcohol concentration (BAC) gives us an accurate idea of what we might call ‘drunkenness’.  Although some previous research (thanks Matt Hennessey) has found a pretty decent match between BAC readings and subjective judgements of drunkenness, it’s clear from the approach taken here that, as entry will still be based on observed behaviour, there’s something of a PR exercise going on here.

As I’ve been reminded several times, we shouldn’t overstate the rationality or calculated nature of policy.  Sometimes, policymakers are as keen to be seen to be doing something as to actually do something effective.

However, let’s assume that it’s possible there’s something more to this initiative.  Even if it only runs for one weekend, it has still gained some press coverage and might highlight the issue of venues serving intoxicated customers, and the possibility of them being turned away at the front door.  As James Nicholls points out, clubs would have an interest in turning people away not just to please the police but if it changed people’s expectations about how much they could drink before going out while still getting into a club: the clubs will have an interest in initiatives that reduce pre-loading and push people to go out earlier and/or wanting more alcohol to get to their desired level of drunkenness.*

That is, the aim is that people will change their behaviour, seeing that they’re likely to get turned away if they arrive drunk, but I fail to see how this is likely unless there’s a fundamental change in the admission/serving process – which seems unlikely when bouncers are to rely on the same tried and tested observation techniques.

Moreover, drunkenness is seen as desirable by many in the night-time economy.**  That study that showed how observations of how drunk people seemed generally matched BAC figures also noted that for most people knowing their BAC would be irrelevant to their drinking levels over the rest of the night.  Where it would change people’s behaviour, it would be mostly for them to drink more (24.7% of those surveyed), rather than less (just 3.5%).

However, it’s possible that this might reduce drinking levels amongst those going out, as with observational assessments there’s always the feeling that you can beat the bouncers, pretending to be more sober than you are.  People are likely to think it’s more difficult to fool a bouncer who’s liable to breathalyse them.

So by this logic the point of breathalysing people isn’t to actually change their immediate desires for drunkenness; it’s to push them to arrive at the club more sober.

At the same time, this doesn’t undermine the initiative.  We know that interventions that deal with immediate desires can in the end reshape culture as different drinking practices just become normal.  Think of licensing restrictions originating in the First World War that seem to have had a long influence on this country and what forms of drinking were considered normal.  And the current context for policymaking is arguably one in which we’re more likely to accept that people are irrational, and so changing the ‘choice architecture’ that affects their decisions is sensible.

At first sight it would be easier if there were a clear line of BAC that would be unacceptable, as this would reduce the opportunity for just taking a chance with drinking beforehand, assuming that your level won’t be too high, or the bouncer won’t check you, or you’ll be able to explain away a high reading.  However, there are issues with this in practice.  Just to give a couple of examples, the same BAC won’t imply the same effects for all in terms of intoxication, affecting coordination, reasoning and so forth – those things that make it undesirable for a club to let in people who are very drunk.  Second, drunkenness is learned, social behaviour, so even the same level of strictly defined ‘intoxication’ might produce more or less problematic behaviour in different people.

So, having begun the day thinking this was a pointless PR exercise, I’m wondering if there might be some mileage in it after all.

Thinking in this way about an idea that initially seemed unhelpful prompted me to look again at drunk tanks.  (This was also of course so I could end the post on a typically negative note.)

‘Drunk tanks’ as I refer to them here are distinct from triage style units that take the heat off emergency departments because, according to the ACPO proposal, the individual drinker pays for their confinement.

Such a scheme can have two possible aims (which may be connected).  First, to recoup the costs associated with drunkenness.  In this sense, as I discussed in my previous post, the responsibility for drunkenness lies almost entirely with the individual.  This shouldn’t be taken for granted, particularly if we are in this new policymaking world of understanding individuals as inherently irrational.

In this sense, the drunk tanks would work perfectly well if run efficiently.  However, ideas such as a late-night levy, or (in order to capture the off-trade’s role in this issue) simply funding these initiatives through taxation on alcoholic drinks would seem perfectly reasonable.  Such approaches would acknowledge the role of availability of wider structures in shaping culture and desires.

One concern of mine is the involvement of private companies, implying the addition of a profit incentive.  Even if one accepts that an individual should pay for the cost of their care, it’s another step to suggest that they should generate profit.  Of course the argument would run that the private company could provide the care and the profit at a cheaper rate than the existing state institutions, but I’d be baffled as to why that couldn’t already be done within the current system***, or why a third sector provider wouldn’t be able to provide the same or a cheaper service than the private sector.

However, there’s a second possible aim of the drunk tank, and that’s to re-shape behaviour.  It’s here that I think other approaches are likely to be more effective.

The drunk tank as deterrent relies on people reacting to the spectre of having to pay for their care by being more careful about their behaviour.  However, such attempts to re-shape people’s decision-making are not always effective.  Women in particular are frequently reminded of the importance of safekeeping strategies.  You might think of articles like this or campaigns like this, this or this.  However, research (not just mine) suggests that often what is the more likely consequence is ‘safety talk’.  One example in my research was a young woman who took a fixed amount of cash out with her, and no debit or credit card, so she couldn’t drink more than she felt was good for her.  She also took out an extra tenner, hidden in a ‘secret pocket’ so that if she lost her bag, or drank the full amount she had with her, she’d still have the money for a cab home.  Only, in practice, she knew that money was there and her drunken self would end up spending it on alcohol and deciding to walk home.  Regularly.

This highlights Harry Sumnall’s issue with the idea of the participants in the research cited above who said they’d drink more if they were given a BAC reading: “at time of interview subjects had drunk 10-16 units. I dont trust their answers”.  It’s a good point: we are different when we’ve had 10 units or so.

And that’s why I’m not sure about drunk tanks changing behaviour.  They might, as a spectre, but we’re also very capable of explaining away our drinking and underestimating our risk when we’re thinking of things we like.  Moreover, it’s unlikely to seem such a frightening prospect when we’re getting out that ‘secret tenner’.  At this point, we maybe do need some ‘nudging’ rather than an appeal to our rational, better natures.

If we’re serious about changing behaviour, I’d be interested in more structural solutions – and that could include initiatives like the Late Night Levy, even – or perhaps especially – if the costs are passed on to the consumer.  Certainly there are problems with how this would apply if it only affects the premises that are open late at night, ignoring the role of the off-trade; that’s one of the reasons MUP is attractive.  At the moment, though, I remain fascinated by the Gothenburg model of alcohol retail, largely removing the profit motive from the trade.

*This doesn’t necessarily mean they’d be keen on a strict definition of who they should be letting in and serving and agree with the police.  In a 2008 study, only around 60% of staff surveyed knew it was illegal to serve drunk people alcohol, and even of those who did know, over 8% of those still felt it was part of their job.

**There’s loads of stuff on this, but as well as my work, these two articles are particularly useful.

***For example, if it’s expensive because the police/ambulance staff are overqualified for the role they’re performing, a solution like the Cardiff triage centre solves precisely this problem by employing people and resources to fit the precise issue.