Does Privatizing Retail Alcohol Sales Increase or Decrease Consumption?

If you’ve ever bought alcohol in Pennsylvania, you know what a weird, controlled system it has. As 1 of 19 “alcoholic beverage control” states in the U.S., Pennsylvania has some of the more strict, if not bizarre, laws regulating the retail sales of booze. Wine and spirits are sold only by state-owned stores, which don’t even have names; they’re designated by call numbers instead. Prices are kept uniform in all locations. Many of the stores operate at a loss. If you’re under 21, you’re not even allowed inside, and until recently, you couldn’t buy alcohol on a Sunday. Oh, and check out its “Kafka-esque” system of grocery store wine vending machines.

This may all be about to change. Pennsylvania is debating whether to privatize its liquor sales as a way to raise money and bridge its budget gap. Estimates show the state could raise anywhere from $1 to $6 billion by selling off its stores to private buyers, though that would be a one-time cash infusion and the state would give up its stream of roughly $90 million in annual profits; though it would still collect sales tax revenues.

Whether that’s good fiscal policy is debatable. But so is the projected effect that privatization would have on alcohol consumption. A 2009 study by the Commonwealth Foundation, a Harrisburg, Pa. think tank, which examined the effects of deregulation in Iowa and West Virginia, found that per capita alcohol consumption decreased by as much as 5.9 percent. One explanation cited by the study’s authors, John Pulito and Antony Davies of Duquesne University, was that with more convenient hours and less restrictive purchasing guidelines, consumers bought less alcohol per trip.

But a new study refutes those findings, and shows that privatization efforts increase alcohol consumption by an average of 48 percent. The report, conducted by a task force of public health experts, analyzed 21 other privatization studies in seven states, two Canadian provinces, and two European countries. Stephen Herzenberg, an MIT-trained economist and executive director of the Keystone Research Center, reviewed the report and wrote a policy brief. The highlights of his findings:

After privatization, consumption of privatized beverages increased “substantially” (the median increase was 48.2%). Meanwhile, consumption of non-privatized beverages increased very little (the median change was a fall of 2%). Thus, overall consumption of alcohol increased a lot.

Having documented large increases in consumption with privatization, the Task Force inferred based on “extensive evidence” that “when privatization results in substantial increases in alcohol consumption, there are at the same time substantial increases in excessive consumption.” The extensive evidence referred to consists of a body of literature documenting across many societies that most people drink a small or moderate amount, while a few people drink a large amount. Since most adults drink less than once per month, and the top 10% of drinkers account for the bulk of consumption, any sizable increase in consumption means that excessive consumption must have increased. This conclusion was borne out in Finland, where, after the sale of medium-strength beer was privatized, there were increases in alcohol consumption at all levels, including among the heaviest drinkers.

From its review of the most careful studies, the Task Force finds consensus in the scientific and public health communities that privatization will increase alcohol consumption and is likely, therefore, to increase problems associated with excessive consumption.

The Commonwealth Foundation study also compared control states to license states, and found that rates of underage drinking and underage binge drinking were roughly the same. There was also no difference in alcohol-related traffic deaths in license states compared to control states, according to the study.

Washington state is also considering privatizing its state-run liquor business. In Virginia, lawmakers floated the idea earlier last year before backing away from it.

So, what do you think?

[poll id=”9″]





I'm in Pennsylvania, and I'd love to see alcohol sales privatized. I'm inclined to think binge drinking would go down, especially among those who drink beer. At present you have to buy beer by the case, at a beer distributor. There are a few places that circumvent the law and sell six-packs by identifying themselves as restaurants, but those are relatively sparse, difficult to locate, and have limited selection. So most people who buy beer have to buy massive amounts of beer, which I assume leads to people getting drunk because it's already conveniently in the house in large quantities.

It might also reduce drinking and driving, since fewer people would go to delaware or new jersey to buy alcohol.


Beer sales are already privatized. I'm sure the beer distributors and their lobby will NOT let the new, privatized stores sell beer in any quantity, period. The state had to give some decent concessions to the other sellers of beer, wine, etc to let the beer distributors be open on Sundays. I'm guessing that the only thing this would do would privatize the state stores and that's it. I very much doubt that the private stores would be allowed to sell anything besides what they sell now.

Michael Fisk

I think asking the question "will alcohol consumption increase or decrease as a result of liquor store privatization?" is asking the wrong question, and that perhaps the more valid one from the perspective of crafting public policy with regards to the issue is that of "what effect will liquor store privatization have on rates of binge drinking?". From that perspective, it is possible to harmonize some of the conclusions of both the Herzenberg report and the Commonwealth Foundation report - namely, people may both drink more often but also drink less each time they do drink due to the increased convenience of purchasing. From this perspective it could be argued that, in reducing the likelihood of binge drinking, the proper public health policy decision would be to privatize liquor stores in the state even if overall alcohol consumption increases.


As a former Pennsylvanian, my mind is flooded with ridiculus stories of trying to by alcohol. Eventually, we discovered the best place to get alcohol in Pennsylvania is New Jersey. It's not just liquor, you can't get beer anywhere but "Beer Store." You are also limited to how much you can by at once. So to prepare for a large party, you go to Beer Store by two six packs, bring them to your car, then go back in and buy two more, and repeat.
Also, many restaurants find it so difficult to get a liquor license that they give wine away for free (with a meal).
And the "Wine Machines" in the grocery stores make you pass a breathalyzer test. On more than one occaision the machine found everyone in the store including the manager to be over the limit, but the machine has to be recalibrated by the State Regulators who wouldn't be in 'til Monday.


The last commenter is a little confused -- the beer stores sell only by the CASE as an earlier commenter noted. The 6-packs are sold at restaurants and eating places that have a license to sell beer.

The first time we tried to buy beer for a get-together after moving to PA, my husband and his visiting brother's little errand took forever. They'd, unknowing, gone to the liquor store (having lived in places where beer, wine, and liquor were all sold out of the same store). No, no, they were directed to the nearest beer distributor -- only to learn that they had to buy in no less than a case. So they went to buy a selection of 6-packs at a nearby deli/hotdog shop, only to decide that was too expensive, so back they went to get two "mixed" cases -- handily packaged at the brewery as combo packs.

Not being beer drinkers ourselves, we remembered the leftovers months and months later and ended up emptying them down the drain.

In cities, there are lots of bars, restaurants, etc. where you could buy a 6-pack. They buying out of state remedy only works for people who live near a border, as well. If you live out in the country and drink, I imagine you stock up.



Why would people be more or less inclined to drink because of the location of purchase point? I understand it would be more convenient to purchase at a more local location, but just because it is closer to your home doesn't mean you will drink more.
Ohio did away with state stores years ago. This didn't cause anyone I know of to purchase more alcohol. Maybe I just don't run in the right crowds.
And if the PA state stores are operating at a loss, doesn't that mean the state is subsidizing alcohol consumption? Maybe that's how they keep people from leaving the state.


"Since most adults drink less than once per month..."

Where on earth did they get that idea? I drink less than most of my friends & neighbors (excepting the few ex-alcoholics), and I still have a glass of wine with dinner more often than not.

Eric M. Jones

No wonder Pennsylvanians have to drink!

Marcus Kalka

1. "Prices are kept uniform in all locations. Many of the stores operate at a loss."

--Sounds like inefficient price controls and price manipulation -- bad for business...and tax revenues as well.

2. "Oh, and check out its “Kafka-esque” system of grocery store wine vending machines."
From the Wired hyperlink: "Each machine is connected to a state employee in Harrisburg, via video-camera. A customer chooses their wine, swipes their ID, puffs into a breathalyzer and faces the camera. The state employee checks that the ID matches the person and, if they’re not already intoxicated, the person is allowed to buy the wine (the machine vends only wine right now)."

--Sounds much more Orwellian rather than Kafkaesque...each machine equipped with its own telescreen. "Victory Gin, anyone?"


I live near a grocery store that has a wine kiosk. I've never used it but I have seen people so confused by it they just give up and leave. The grocery store has actually petitioned the state to remove them from all of their stores because of the confusion they cause. I'm sure the store would love to sell wine if it could do it the same way they sell beer. That section of the store is always busy and their selection is great.

I'd be all for privatizing the state stores if that meant opening up the market to grocery stores to sell all alcohol.


If most adults drink less than once per month, and total consumption is increased by 48.2%, then the result would be that most adults would drink less than 3 times every 2 months. WOW, let's bring back the paddy wagons.

Last year I went camping in PA where I was unfamiliar with the laws (I live in NY). I was forced to buy a case when I only needed a 6 pack. since gluts incentivize overuse, I drank the entire case in 3 days.

PA is beautiful and I plan to go back, but I'm bringing my own beer.


Did the study showing a 48% increase actually measure consumption, or did it measure sales?

It seems reasonable that in-state sales would increase because those who were buying from out of state -- to save money, or for better selection or hours etc -- could now get some of those benefits locally.

Unless you can postulate a mechanism, it's implausible that merely changing the method of distribution would result in a 48% increase in real consumption. I can understand a minor increase just from pure convenience, but 48%? Show me how.


I don't see a mechanism mentioned for this 48% increase. All it gives is numbers, some based on what sound like guesses. Also, the paragraph right after it says no increase in underage drinking and no increase in alcohol related traffic deaths. So what does that mean for this report? "Yeah, there's more drinking going on, but at least two of the results you'd expect aren't happening."

The latest I heard on Washington State privatization was that they were planning to find a private company to handle distribution. That state still gets a cut and can cancel after 3 years. This does not sound like privatization to me, it sounds like government contracting.

Wes Y

"most adults drink less than once per month"

Are you kidding me? I realize I probably drink more than most, but once a month? I might believe once a week, but even that seems low.

Joshua Northey

That seems somewhat low I agree, though I would believe 35-40%. I think it would surprise you how many teetotalers and near-teetotalers there are out there. My wife and I usually have some alcohol in the house, but I would bet the number of consumption incidents a month might look something like: 2,0,0,1,4,5,0,0,0,3,4,6?

Zero is definitely the mode.

Joshua Northey

"After privatization, consumption of privatized beverages increased “substantially” (the median increase was 48.2%). Meanwhile, consumption of non-privatized beverages increased very little (the median change was a fall of 2%). Thus, overall consumption of alcohol increased a lot."

I feel like this statement is very bad science without a lot more explication. Are the two types of beverages from the same areas (i.e. Jurisdiction A privatized beer but not wine and beer sales grew.)? Are there controls from neighboring jurisdictions with no laws/no changes? How long were the study periods, was the effect temporary? Were there difference in consumption increases between places where all beverages were deregulated vs only some?

I feel like the statement as written just tells me that when beverage A has a transaction cost reduction it outperforms similar beverage B which doesn't have a transaction cost reduction? Is that really the story here? If so it is not very interesting or helpful!



Apropos of nothing, I find it interesting that conservatives initially supported the ABC system as a way to control the sale of alcohol, but now support privatization in the name of smaller government, at least here in Virginia.

I do agree with their sentiment, however: why should the state be in the business of retail and wholesale alcohol sales?


One weird comparison. In Ireland there has been a long-standing prohibition on the sale of alcohol on Good Friday (the Friday before Easter), a remnant of the strong role Catholicism once played here. I noticed that my friends tended to buy massive amounts of alcohol on the Thursday in preparation for that Friday! I'd love to see the statistics, but I suspect that people might actually consume MORE alcohol on Good Friday because of this prohibition, through a mix of stockpiling on Thursday and a sense of anti-government or anti-Church rebelliousness.


Privatizing might change the rate of consumption, or it might not. It might change DUI rates, or it might not. But that's all fodder to the hand-wringing stat slingers. I really don't care what the effect of privatizing would be. I just know that the State has no damned business sticking its hands in commerce that should be managed by private enterprise.