Closing Time?

I’ve been doing a series on drunk driving and alcohol-related harm (which can also include impaired work performance, domestic trouble, violence, crime, risky sex, fetal disorders, brain damage and many other health problems). Can – and should – we do more to control alcohol?

Last time, I wrote about programs designed to curb alcohol abuse through persuasion. But the evidence that they work is, well, less than persuasive. So is it time to start cutting people off?

With the help of a fine recent literature review by Peter Anderson, Dan Chisholm, and Daniela C. Fuhr in the medical journal The Lancet, we can look at the totality of the evidence on alcohol control, including policies to limit its availability through regulations.

Here’s one we already have: the minimum drinking age (MDA) of 21. We have a really good idea what the effects of MDAs are because we have neat natural experiments to study; states started reducing age limits (starting in 1969) and then turned around and began raising them (starting in 1976). What happened when they did?

As reported by Ann T. McCartt and Bevan B. Kirley, a sample of 33 studies found that forbidding alcohol for 18-21 year olds reduced alcohol-related crashes in that age group by between 10 to 16 percent. This effect has been consistently demonstrated, though perhaps it’s smaller than one might have hoped for.

Raising the MDA further is probably out of the question. But there are ways we could enforce it more strictly, like extending initiatives that allow parents to be sued if underage drinkers leave their home and get into fatal crashes. Worthwhile, or worringly draconian?

In the other direction, there is a movement out there to lower the MDA to 18, on the fairly logical grounds that if you’re old enough to vote or register for Selective Service you should be old enough to drink. Some college presidents think that campus drinking could actually be brought under better control if all students were legal.

And it should be noted that in Europe most MDAs are between 16 and 18, and their DUI problems are less severe than ours (though as I’ve noted, their other DUI-related policies, like blood alcohol content limits, are generally stricter).

Still, whatever the merits of these arguments, the evidence unsurprisingly shows that lowering the MDA would result in more road deaths.

(It’s interesting how much wiser we grow as we age. As a philosophically immature 18 year old, I considered the MDA one of our most iniquitous public policies, but just three years later I had a profound theoretical breakthrough and suddenly realized it is resoundingly just. Today, I think a cutoff at age 41 would be ideal.)

I went to college in Boston, where liquor stores were closed on Sundays. I can attest to the fact that this cut down on our Sunday alcohol consumption. (I guess we must have cleaned out the stash on Saturday nights.)

Studies have backed this observation up. I have not seen numbers on whether limiting the hours alcohol can be served would decrease drunk driving per se, but Sergio Duailibi and colleagues found that when the Brazilian city of Diadema enforced a cutoff in alcohol sales at 11pm (most bars had been open 24 hours), there was a 29 percent drop in the murder rate.

As for a reasonable step in this direction that might not totally ruin our good times, what about allowing bars to stay open as late as they do now but requiring them to stop selling alcohol an hour before closing time? This would give people a chance to sober up before they hit the road. A similar policy seems to work reasonably well for sporting events.

What’s the evidence on regulating the number of alcohol outlets? Fewer bars and liquor stores might theoretically translate into lower alcohol consumption and DUI. However, the limited data are inconclusive. Indeed, it is possible that having fewer alcohol outlets might actually lead to more driving, and thus more accidents, as people head to those places that are open.

Then again, a persuasive body of research has indeed linked a high density of alcohol outlets with other forms of alcohol-related harm. Nobody wants to return to Prohibition, but some tightening of the granting of liquor licenses might at least be worth some consideration.

When was the last time a bartender refused you service or took away your keys on the grounds you’d had one too many? Katharine Ker and Paul Chinnock have found very little evidence that programs designed to get bars and restaurants to limit alcohol abuse have proven effective.

In large part this is because serving places don’t follow through on the initiatives, which is understandable enough given the economic incentives they face. And it is consistent with evidence I showed you earlier, which demonstrated that educational material produced by the alcohol industry sometimes actually promotes positive views of drinking, and that promises by the alcohol industry to restrict advertising usually don’t seem to bite.

So in sum it seems as if certain types of restrictions on alcohol availability might bear fruit. But are these the best way to go? Economists might not think so. For the most part, they believe that price signals (e.g. through taxation), not regulations, are a more promising tool for shaping behavior.

Why? Here’s one example. At heart, many programs designed to mitigate alcohol-related harm involve raising the “generalized costs” of drinking (in terms of time, money, hassle, etc.). But taxing people in the currency of time or discomfort (e.g. by making them drive farther to buy beer) would raise no revenue and indeed would ultimately cost the government money (fewer liquor license fees and less sales tax revenue). On the other hand, increasing alcohol taxes would actually raise money for the public coffers even as it limits destructive behavior.

Another issue is that these regulatory policies might be considered an unnecessarily blunt instrument. Cutting off bar service an hour early would penalize all drinkers, not just those who have shown a predilection for mixing alcohol and dangerous behavior.

Coming up I’ll look at the impacts of raising alcohol prices, and of policies that target drunk driving more directly.


I don't think it's appropriate to compare the MDA, which is a total prohibition for those under it, with temporarily restricting the availability of alcohol to those of legal status. Those who are prohibited from drinking will go to great lengths to get liquor and then consume it quickly and irresponsibly (binging) to avoid having it confiscated. If a bar stops serving alcohol earlier (an interesting idea), I can see how it will have an incremental benefit by keeping legal consumption from becoming overserving.

As the parent of a high school sophomore, I have begun to see the effects of binge drinking in our affluent suburban community. Fortunately it hasn't affected my family yet (at least not that I know of). That being said, I support a reduction in the MDA to 19 from 21. Such a move would at least cut 2 years from the use of alcohol as an illegal substance, while still allowing alcohol to be banned at the high school level.


Phil Scott

One thing in my city that definitely pushes drunk driving is the suburbs and the lack of a neighborhood bars. You simply MUST drive to a bar, which the NIMBY people have forced into particular neighborhoods often 20 minutes away. Sadly more and more people are choosing the risk driving drunk than pay the $80 it would take to get a cab ride both ways. Not to mention cabs are typically not in the suburbs, so getting a cab company to come out to pick you up is an ordeal in and of itself. So many people simply drive to the bars with good intentions, but when left with the choice of "risking it", or leaving their car at a bar and cabbing it home their alcohol impaired minds will pick risking it almost every time.


What is the incentive for a bar to remain open for an hour after serving the last drink?

I think this would effectively just change the closing time by an hour, I know if I was a bartender or bar owner and I had made all the money I was going to make for the evening I'd be looking to show people the door.

David Benson

A government sponsored study by a colleague in the early 70's, the Alcohol Safety Action Program, showed a number of interesting results. Surveys were made of drivers throughout the countrry at all times of day and night. Drivers were stopped at random and asked to provide breath and urine samples in return for freedome from prosecution and a free taxi ride home if their breath results showed excess alcohol. Surprisingly, most stopped did indeed volunteer.

One result I remember from this study was that your probability of encountering an impaired driver (number of impaired drivers per minute of driving time) was about the same day or night. The number of drivers on the road at 2 AM was very low; but the probablility that they were impaired was very high. This result suggests that a change in closing time might not help the drunk driving problem very much.


Personally, I couldn't picture myself staying in a bar for a full hour after they stop serving alcohol. At a baseball game, when they stop serving around the 7th inning, you still have the game to watch. But what are people expected to do in a bar between 3-4a.m. if they can't drink?


Would cutting people off an hour before closing really limit consumption? I'd probably finish my drink and then go home as soon as it was done. I love bars, but a large part of it is that I love drinking in bars. The two go together.

And further, I doubt an hour gives you much time to sober up. In fact, say this is New York and bars close at 4. If I'm still in the bar at 2:55 (which I'm often not - often I'll be quite drunk enough by 2 or 2:30 and just go home anyway). I'll order another drink and drink it while they are no longer serving (similar to buying another beer or two right before alcohol sales shut off at a baseball game). Say it takes me a half hour to drink the drink (I'm nursing it or whatever). Then I wait another half hour...probably just enough time for the effects of my last drink to set in.

I think #2 has a good point. Better public transit and more evenly distributed bars would help a lot. I know when I leave New York and go to my home town it becomes a much bigger hassle to go out drinking. I still do it, but it stinks to be stuck as the DD for the night, and I can see why a lot of people would resist it.



"what about allowing bars to stay open as late as they do now but requiring them to stop selling alcohol an hour before closing time? This would give people a chance to sober up before they hit the road. A similar policy seems to work reasonably well for sporting events."

This is not a very good analogy - there is a good reason people stay at a baseball game after the 7th inning, that would not apply to a bar after they stop serving alcohol.


If the issue is drink driving, then tackle that through lower limits, more check points etc.

Other policies such as changing opening hours and days are just infringing on people who are doing nothing wrong. If someone is prepared to get a taxi home / has a dd, why should they be restricted for other peoples lack of responsibility.

Try enforcing the laws you have first without bringing in more. Random checkpoints where every driver is breathalyzed are very effective, they have really cut down on drink driving in my home country in Ireland.

Mike B

Instead of restricting people's freedom to responsibly consume alcohol perhaps there is a technological solution that could be implemented to detect driver intoxication based on continuous monitoring of driving behavior (as opposed to BAC at time of engine start). We all know the sort of weavy driving pattern of a drunk driver so if the car's on-board computer could detect a drunk driver by this an other metrics then various steps could be taken such as limiting the top speed to 25mph until some action is taken that would be hard for an intoxicated driver to accomplish.

BTW, do you have any metrics for how many drunk drivers take their own lives vs the lives of innocent people killed by drunk drivers? You treat all driving deaths as a negative cost on society, but an alcoholic that removes him or herself from the gene pool may actually impart positive benefits on society.

Mark Wolfinger

How about zero tolerance? Not for using alcohol. Use it as much as you want. But no allowance for irresponsible use.

My solution: Get caught/convicted driving when intoxicated one time and lose your driver's license for life in all 50 states.

Caught driving while under that sentence of no driving - serious jail time. If drunk, life in prison, no parole.

Larry Scorsby

Your ideas have been tried and talked about before. Next


I remember going to College in a dry county. As underage students, we beat a path to the next county to buy our cheap beer and wine. When we turned 21, the proprietor of the closest liquor store gave us a split of wine for our birthdays. What I am saying is that the law didn't work. When the liquor establishments were closed on Sunday, we stocked up on Saturday. Then I moved to a different city, and I am positive it is better to have people drinking in a neighborhood pub and walking home, no matter what the day of the week or hour of the day.

I should also mention that I am VERY against churches who pay no taxes having any say about how citizens who do pay taxes behave. They shouldn't force their beliefs on other people. That is constitutional, however in the South, the constitution is often ignored.


Let me get this straight--18-21 year old's shouldn't drink because of the harm it may cause, but they can enlist in the military and go places where people can shoot at them and vice versa?


I believe that in most areas bars are allowed to be open as long as they want, but cannot sell alcohol between certain hours. They close their doors because their money is made on alcohol, not other services.
Early cutoff at a sporting event is different. There is the rest of the game to watch. Also, football games I've been to take another hour to creep your car out of the lot. Assuming even drunk people have reasonable success driving at 3mph, the early cutoff at sporting events could account for as much as 2 or more hours before captain-drunko hits the open road.


Force bars to stop serving an hour before they close? That's silly. Unlike sporting events, there's not much more to do at a bar once they stop serving alcohol (unless it's a nightclub). Otherwise, bars would already stay open an hour or more later than the time limit for booze.

On the contrary, I think extending alcohol-serving hours, at least where I live, where it's 2am, to say 4am such as in NYC, makes more sense. Instead of crowds of drunks all filtering out into the streets (and cars) at the same time, people would peter out between 2 and 4. This seems like it would result in fewer drunken fights, and while it may not cut down on DUIs, there would at least be fewer drunks on the road at the same time (and maybe a few more extremely drunk drivers out after 4).

I don't have any numbers to back up my suppositions, but I do think the 2am limit is lame.

For DUIs specifically, why not just tighten up the restrictions and the enforcement to European levels?



I believe that the concept of a bar staying open an hour later then the last call is interesting....on the one hand there are people who perhaps would sober up to a condition that is acceptable for driving. However, I believe that it would lead to more people believing that they are okay to drive when they actually are not. (I have worked in a bar)

I believe that the lack of local bars and the uneasiness people feel when walking home or using public transportation late in the night play a large role in the prevalence of drunk driving.

As for lowering the MDA, I think it would be good for kids to first experience alcohol in a controlled environment where they can be taught responsibility via their parents. In all honesty, if a high school student wants to get alcohol they will find a way. Perhaps the thought of not having to wait as long to drink legally may persuade more kids to wait.



I'd like more information on how regular, slightly excessive alcohol intake affects middle aged bodies. I read that too much drinking can cause cancer, but the information is so vague. I know a lot of people my age who are not drunks but drink more than is considered safe. These people would never get in a car after drinking but I don't really think they see any harm to their health in drinking the way they do.

Andy W

Criminological literature on changes in drinking hours has shown what I would consider reasonable evidence that changing drinking hours later has an adverse effect on assaults (and the same vice-versa that lowering closing time tends to decrease assaults)
[Do relaxed trading hours for bars and clubs mean more relaxed drinking? A review of international research on the impacts of changes to permitted hours of drinking
by Tim Stockwell and Tanya Chikritzhs]

I think the studies need to look at who goes to the bar after changes in hours though as well. Later hours may attract more deviant clientele (or just more patrons in general). I'm not sure if the research I've seen is more suggesting of bars shaping behavior (like drink specials encouraging larger consumption), or that just certain bars tend to attract bad cats.

I agree though that neighborhood studies on the number of bars and adverse outcomes are pretty weak. I have not come across any good quasi experiments examining changes in bars and crime (at least not any that had anywhere near a large change in the number of bars in a neighborhood in a short amount of time).



How can you advocate increasing alcohol taxes in one sentence and then, in the very next sentence, call regulation "an unnecessarily blunt instrument" because it affects all drinkers? This smacks of arguing from the desired conclusion rather than from rational examination.


I think that if bars have last call an hour early but remain open for the next hour everybody will just leave the bar when they stop serving. Of course a few people will stay to try and seal the deal but most people will leave once they finish that drink