Economists are often accused of being a dour lot, whose grubby focus on molding behavior with? carrots and sticks ignores what is noble in the human spirit: higher cognition, altruism and innate goodness. Does the fight against alcohol abuse, particularly drunk driving, show that man can be reasoned with, or does economics – aka the “dismal science” — offer a better guide to human nature?
A huge number of studies from around the world have looked at the effectiveness of alcohol control measures. Peter Anderson, Dan Chisholm, and Daniela C. Fuhr have done a nice summary very recently in the medical journal The Lancet. Over the next couple of pieces, I’ll fill you in on what they report. This time, I’ll look at whether it is possible to persuade people to drink responsibly.
Strategy one: can we reach potential DUI candidates when they’re young? Unfortunately, the evidence on teaching sobriety in the classroom is not too encouraging. A large body of research has shown that the vast majority of programs have largely been ineffective. L. Jones and colleagues found that only six of 52 high-quality programs have been able to show results.
Moreover, those results tend to fade over time. For example, one of the success stories – the School Health and Harm Prevention Program – managed to reduce dangerous drinking an impressive 25.7 percent in the short haul, but only 4.2 percent 32 months later (see this from N. McBride and colleagues). Because of this, the program is not particularly cost-effective, with a cost of over $2500 for each case of hazardous drinking averted at 32 months.
If teachers can’t get the job done, can “Smart Mom?” The good news on alcohol-related parenting programs – which are designed to foster parent/child communication or otherwise improve parenting skills through discussions, videos, coaching, internet programs, etc. – is that the evidence on them is brighter. The bad news is that it’s not much brighter. Only six of 14 studies on parenting programs reviewed by J. Petrie, F. Bunn and G. Byrne come up with statistically significant evidence that they have any effect on future drinking.
As M. Stead, R. Gordon, K. Angus, and L. McDermott report, evidence on the effectiveness of social marketing campaigns (which borrow tactics from the private sector like market research and messages designed for the target audience) is also mixed; only about half of the programs they analyzed showed any effect.
Public information campaigns (e.g. advertising about the dangers of alcohol) can focus our minds on the problem but probably don’t have much effect on actual drinking, or so the limited evidence thus far indicates.
C. Wilkinson and R. Room have found that warning labels on alcohol may perhaps make us feel a little guilt, but have little if any effect on actual alcohol consumption (though they do seem to work for cigarettes).
If more persuasion to not drink isn’t too effective, what about less persuasion to drink? It seems self-evident that less alcohol advertising, sports sponsorships, etc. would lead to more sobriety, but here again the evidence is not overwhelming. The bulk of the literature,? as reviewed by C. Gallet, shows a surprisingly weak link between alcohol advertising and consumption, though some studies, particularly those that track subjects over time, have shown that less advertising does work, particularly for the young (see this from P. Anderson, A. de Bruijn, K. Angus, R. Gordon, and Gerard Hastings).
However, we are unlikely to see less advertising, particularly if we wait for the alcohol industry to take the lead; self-regulation initiatives have not led to results in the past. Ironically, responsible drinking education programs produced by the alcohol industry have been shown to sometimes have the exact opposite of the (presumably) intended effect: they actually promote positive views about alcohol and its makers.
Workplace policies like interventions have been poorly studied; though G. Webb, A. Shakeshaft, R. Sanson-Fisher and A. Havard find in their review that such techniques have “potential,” there is as yet little reliable evidence that they work.
As E.F.S. Kaner and colleagues report, one form of persuasion has proven to be effective in controlled trials: health provider intervention. In this method, doctors, nurses or psychologists screen patients and identify those who have unhealthy levels of alcohol consumption. Then the medical professionals provide information on alcohol dependence and aid in formulating a plan to cut back. The problem with this method is that it is time-consuming and costly, and does not reach, or work for, all drinkers.
It is much harder to prove that something doesn’t exist than to prove that it does exist, and it is quite possible that there are some persuasion programs out there that might be a silver bullet. Perhaps the fact that half of the programs in some of these studies had an effect is good news, not bad: we can now build on these strategies.
On the other hand, it should be noted that even in cases where statistically significant results can be shown, the actual magnitude of the effects can still be disappointingly small, as in the School Health and Harm Prevention Program cited above.
In all, we’d definitely like to see more conclusive evidence that these methods work, and thus far we don’t quite have it. My two cents is that part of the fun of drinking is knowing that you’re doing something rebellious and vaguely anti-social, and thus societal exhortations to not do it in a way only add to the perverse thrill.
Time to get the dusty old stick and carrot out of the closet? More on this next time.
and 
. You can also watch it on 
Financial incentives make me drink less. I added up what I spent in March on booze and booze-related activities and Whoa. Big money!
I really want to pay off some debt so I’m cutting into my booze allowance to do so. Hopefully I can shed a few pounds along the way.
Once I’ve paid off the debt I’l undoubtedly go on a bender, so maybe these results aren’t actually long-term.
Have good studies been done on the effectiveness of a punitive approach? If we cannot increase the fraction of drunk drivers apprehended, can increasing the penalties for the ones who *are* caught change behavior? I ask because Finland’s rather draconian drunk-driving laws have been mentioned here and elsewhere, but not in context of whether it works better for them than for us.
Very interesting I’ll read all the links. Was about to blog on this topic myself but will just tweet your post. Two things: Parents – I wonder if it is not so much the relationship but what the parents do with their own drinking – i.e. setting the benchmark! Wrong message – It’s not if drunks kill/maime themselves it’s the innocents whose lives and families they destroy. The problem though is how to get that through.
Is alcoholism really a sickness or is it simply a “bad” lifestyle choice? Can one inherit it? Is it a self-destructive bent in an otherwise reasonable and very intelligent person? Where do you go from “just say no”.??
Shirley
All nonsense.
In vast areas of the world, water and sex are the big killers, and if the folks there had free beer, there would be many fewer deaths from cholera and all the other serious water-borne diseases.
The prohibitionists here are not really humanitarians–they are knee-jerk fundamentalists who take great delight in forcing their silly morals on all others.
The closest thing to a solution to drunk-driving is to provide better public transportation. When we were young and drank heavily in Munich, we never even had a car for five years. We could either walk or travel efficiently by bus, streetcar, U-bahn or S-bahn for a cold beer or some wine. Kids had the right to drink beer and wine starting at age 14.
Trouble is, a drunk pedestrian’s getting hit by a sober Baptist is counted by Mad Mothers (and the OP here) as an “alcohol-related” accident, and, as has been pointed out right here on Freakonomics, a drunk is safer driving home than walking!
It’ll be a cold dry day in hell that I cooperate with a society that continues taking my money to build ever more roads that keep me from walking to my favorite bar. City planners here could learn a lot from Jamie Lerner of Curitiba, PR Brazil.
There is probably some percentage of drunk drivers that could potentially be affected by further training (e.g. school-based education and public awareness campaigns). Such offenders would be non-repeat offenders, either “learning their lesson” in the classroom or as a result of a drunk driving event.
However, I think it likely that the majority of drunk driving offenses are not randomly distributed throughout the population, but are concentrated in a narrow segment of the population. Such habitual offenders are unlikely to be affected by additional training, but would be more likely to respond to individual interventions and legal enforcement.
If this is correct, then an evaluation of anti-drunk-driving programs should find little (if any) success in broad educational or awareness programs – due to the low incidence of drunk driving in the general population – and more success in programs that target known offenders.
i recently underwent a residential treatment for alcoholism, for which the success rates (nationally in the uk) was something around 1%. (I have tried to verify this – the closest I could find was “4-26%”, in a somewhat related study)
To my mind, there is an underlying issue: if people really want to drink they will.
Good luck trying to solve that through pamphlets…
I have to agree with Malthus on this. It seems that, typically, people will believe themselves to be above the norm – i.e. that while perhaps someone else might be unable to handle themselves while drunk, they will have no problems. What this means is that informing people of the negative consequences, unless they can basically be conditioned to think in a specific way (I’m thinking along the lines of Burgess’ “A Clockwork Orange” here), will likely not have a substantial effect. If someone firmly believes that they will not cause any problems, then why would they be moved by hearing about the consequences that would result from such problems?
On that note, then, I agree that the most effective way to deal with drunk-driving is to eliminate the driving, rather than the drunk. Whether it be public transit, some kind of subsidized taxi program (or perhaps a taxi fee paid upon entering the bar, which may also lead to limiting drinking), etc, there need be some options available that aren’t overly expensive or time consuming (which I would guess is likely why people drive, as the typical bus trip is much longer than the car trip, and taxis are quite expensive).
On a total aside, I just finished the book, and I’m glad I found this blog – it seems quite an interesting read.