I have always wondered if Matt has ever given any thought to the State of New Hampshire selling hard liquor at highway rest areas in both direction on I-95 between Maine and Massachusetts. I assume Matt must have traveled this section of road many times on the way to Maine.
I've often been surprised by the chart that others have pasted a few times in the comments here, suggesting that a third of Americans basically don't drink, a third average less than 1 a day, and the remaining third is divided between one a day, several a day, and ten a day. I'm fairly surprised by the high end, but even more surprised at how big the low end is.
I've rationalized this by thinking that as a childless urbanite, my social environments select heavily for people who drink (i.e., I've assumed that non-drinking is largely a phenomenon of suburban parents).
But the income charts in this post suggest that there might be a separate dimension of skew in my social experiences as well, that I hadn't considered.
Teetotalers are much more common than people realize. A lot of people basically give up alcohol and their friends never notice.
I used to be a regular drinker. Over the last 3 years I don't think I've had a drink at home, never order one when I eat out for dinner. I have at most 10-15 drinks a year but if you ask my friends most would assume I still drink at normal levels.
I think this would be a wonderful way to start a crime spree of epic proportions. The idea that you could raise taxes high enough to discourage consumption without recreating moonshiners, bootleggers, hijackings, robberies etc is interesting. There is also the not insignificant problem of people switching to other potentially more harmful recreational drugs. Driving while high is a bit more difficult to police than alcohol. The additional costs of trying to partially recreate prohibition I do not see in your calculation.
Many of the taxes mentioned in the article don't seem high enough to create a lot of moonshiners. Many states that only just legalized cannabis have taxes in the 20-30% range and still seem to have a majority of sales in the legal market now, even though there was already an established black market. An alcohol tax in the 10% range, seems unlikely to create a bigger black market than already existed for cannabis (particularly given that taste is a bigger selling point for alcohol than for cannabis).
I am flabbergasted. Matt who also argues (correctly but it still makes me want to cry into my glass of wine) that even modest Gun Control bills - that would directly save lives! - is a bad idea b/c it costs Dems votes in swing states but cannot actually pass, this same Matt now wants to raise alcohol taxes because it will have slight indirect benefit to the death rate!?
Booze is ALOT more popular than guns! This truly baffles me. Of all the things to go after why on earth would any politician pick this one?
2. His case is that gun control is really a euphemism for private gun ownership abolition. Living in urban circles surrounded by gun control fans this hits me as correct. Most people who support gun control really mean bans guns. This comes with a heavy culture war component since it greatly mobilizes the opposing coalition but doesn't generate nearly as much energy in pro gun control camps..
By contrast raising the alcohol tax isn't a stalking horse for abolition and doesn't bring any of the culture war components.
I think the issue is that gun control is a type of bill that is talked about A LOT A LOT A LOT by certain politicians, while alcohol taxes basically never are. If he thinks both should be talked about somewhat, then he's pushing in the right direction on both.
I believe Matt has observed in other posts that increasing penalties for crimes rarely works anywhere near as well as increasing probability of being caught. People who are committing crimes are rarely thinking rationally, but if they learn from experience that they're likely to get caught, they'll respond to that subconsciously.
I think specifically increasing penalties on intoxicated people is likely to be the least effective type of increasing penalties, because intoxicated people (particularly on alcohol!) are even *less* likely to be thinking about this. (Perhaps increasing penalties for crimes committed while on cannabis would work better, because of the anxiety that cannabis induces, as opposed to the obliviousness of alcohol.)
I have long thought that it would be great to implement broad based consumption taxes a la a VAT just as the baby boom is drawing upon it's tax deferred savings and going into a phase of significant consumption relative to income in retirement. I guess this would be a baby step. As a consumer of not expensive but not cheap wine and pretty expensive scotch the ad valorem aspect would bite into my wine and spirit budget. Not sure if I would seek out lower cost substitutes or just consume less.
These sorts of takes always remind me of the push to ban large sodas: An idea that has merits on a certain level of policy analysis but turns out to be an utter fiasco when it comes to politics in the real world.
I live in TN and they have changed the laws recently with allowing wine to be sold on Sunday (crazy, I know). But as MY keeps talking about the intersection of alcohol consumption and crime, it would be interesting to see if these spikes in violent crimes lately have had anything to do with a loosening of alcohol laws generally speaking, as we become a more secular society.
Thinking back to my Econ 101 course, I remember Sin Taxes like this being touted as highly efficient due to their price inflexibility of demand which would render them highly efficient. Perhaps that's just outdated, but it seems the precise opposite of this thesis.
I think there is little widespread recognition of just how much some people drink. If you have a beer after work every day and then like 5 drinks on friday and saturday you are definitely drinking too much per health recommendations, that's drinking too much per health recs. The type of person who does that probably is thinking "I drink too much but raising the price of a beer by .25 cents would make no difference.
BUT apparently about 75% of the alcohol is consumed by the 10% of people who consume more than the person above. Someone who drinks 10+ drinks a day would surely have their consumption curtailed by a price hike of 25 cents per beer --we're talking about like 1,000/year. And surely it's in this area where you also see a disproportionate share of liquor related deaths. If I remember correctly even for drunk driving the lions share of deaths are from people who are at ~.15 BAC or higher which is very, very drunk.
Back when that "10% of drinkers drink 10 drinks a day" study came out, I looked into it. The survey that they gave people actually had the top 10% of drinkers saying that they had a more believable five drinks a day. However, the survey also was such that if you multiplied it out, it only accounted for about 50% of total alcohol purchased in the United States.
So the authors multiplied everyone's responses by two.
That... seems like a kind of dodgy methodology to me.
I haven't looked further into it to see if anyone has gathered additional evidence/addressed this objection/etc. But I think it's worth looking into before you take it as gospel.
Good critique, but I don't think it undermines the principle of the point even if it is just millions of people having 5 drinks a day.
Also, I would say the DOT stats above that show 70% of the drunk driving fatalities are from drivers at .15 or over indicates there is a HUGE amount of extremely heavy drinking. You need to have 7-8 drinks at a sitting to get to that level... the fact that this cause 6700 deaths in 2010 (sorry the old paper is what I found) indicates tons of people are drinking a lot.
If you are an occasional/moderate drinker I don't even think you'd think to drive after 8 drinks, I'm pretty sure I probably wouldn't be able to find my keys at that point.
I mean, if we said that 0.01% of drives where the person was blind drunk resulted in a fatality, then 6,700 deaths would be 67,000,000 drives in one year. You clearly do not need to have 24M of people getting that drunk each and every day of the year in order to result in 67M drives per year.
Another way to slice it is that on average you get one death per 100 million miles driven-- if .15 drunk driving is 100x more dangerous than average that means 1 death per million miles... 67,000,000 very drunk miles driven. If an average trip is like 4 miles, that would mean 17,000,000 trips, I suppose that could be as few as a million people or less who are making those trips.
In my experience, most people who drink too much are aware in at least some rough sense of how much their habit is costing them financially, and many make periodic attempts (successful or not) to reduce the amount they drink. I find it very plausible that they'll make these attempts more often and more vigorously when alcohol is more expensive.
I'm sure they could work in some exemptions for distilleries that sell below X amount of beer a year, of course too many exemptions will just swallow the rule itself.
At least we should increase it on big companies like Budweiser, they aren't even American owned anymore!
Also, in all seriousness, Matt often writes about the political palatability of various ideas, how they are sold to the public, and how that interacts with actually achieving real world policy victory.
I'll just say that "I'm going to double the price of your beer" is a massive political loser that would be immediately weaponized by the opposition.
I think "buck a beer" was a winning campaign slogan in Canada recently, by Rob Ford's equally populist brother complaining about nanny state liberals taxing your alcohol.
a major factor in firearm deaths is alcohol...
Wonderful work, thank you.
I have always wondered if Matt has ever given any thought to the State of New Hampshire selling hard liquor at highway rest areas in both direction on I-95 between Maine and Massachusetts. I assume Matt must have traveled this section of road many times on the way to Maine.
I've often been surprised by the chart that others have pasted a few times in the comments here, suggesting that a third of Americans basically don't drink, a third average less than 1 a day, and the remaining third is divided between one a day, several a day, and ten a day. I'm fairly surprised by the high end, but even more surprised at how big the low end is.
I've rationalized this by thinking that as a childless urbanite, my social environments select heavily for people who drink (i.e., I've assumed that non-drinking is largely a phenomenon of suburban parents).
But the income charts in this post suggest that there might be a separate dimension of skew in my social experiences as well, that I hadn't considered.
Teetotalers are much more common than people realize. A lot of people basically give up alcohol and their friends never notice.
I used to be a regular drinker. Over the last 3 years I don't think I've had a drink at home, never order one when I eat out for dinner. I have at most 10-15 drinks a year but if you ask my friends most would assume I still drink at normal levels.
Maybe we just don't talk about it very much? I'm a teetotaler now, but when I did drink before, it was still rare and very little.
I think this would be a wonderful way to start a crime spree of epic proportions. The idea that you could raise taxes high enough to discourage consumption without recreating moonshiners, bootleggers, hijackings, robberies etc is interesting. There is also the not insignificant problem of people switching to other potentially more harmful recreational drugs. Driving while high is a bit more difficult to police than alcohol. The additional costs of trying to partially recreate prohibition I do not see in your calculation.
Many of the taxes mentioned in the article don't seem high enough to create a lot of moonshiners. Many states that only just legalized cannabis have taxes in the 20-30% range and still seem to have a majority of sales in the legal market now, even though there was already an established black market. An alcohol tax in the 10% range, seems unlikely to create a bigger black market than already existed for cannabis (particularly given that taste is a bigger selling point for alcohol than for cannabis).
I am flabbergasted. Matt who also argues (correctly but it still makes me want to cry into my glass of wine) that even modest Gun Control bills - that would directly save lives! - is a bad idea b/c it costs Dems votes in swing states but cannot actually pass, this same Matt now wants to raise alcohol taxes because it will have slight indirect benefit to the death rate!?
Booze is ALOT more popular than guns! This truly baffles me. Of all the things to go after why on earth would any politician pick this one?
1. Alcohol kills more people than guns.
2. His case is that gun control is really a euphemism for private gun ownership abolition. Living in urban circles surrounded by gun control fans this hits me as correct. Most people who support gun control really mean bans guns. This comes with a heavy culture war component since it greatly mobilizes the opposing coalition but doesn't generate nearly as much energy in pro gun control camps..
By contrast raising the alcohol tax isn't a stalking horse for abolition and doesn't bring any of the culture war components.
I think the issue is that gun control is a type of bill that is talked about A LOT A LOT A LOT by certain politicians, while alcohol taxes basically never are. If he thinks both should be talked about somewhat, then he's pushing in the right direction on both.
Why not increase seriousness of crimes committee while drunk? Assault. Aggravated assault if drunk.
I believe Matt has observed in other posts that increasing penalties for crimes rarely works anywhere near as well as increasing probability of being caught. People who are committing crimes are rarely thinking rationally, but if they learn from experience that they're likely to get caught, they'll respond to that subconsciously.
I think specifically increasing penalties on intoxicated people is likely to be the least effective type of increasing penalties, because intoxicated people (particularly on alcohol!) are even *less* likely to be thinking about this. (Perhaps increasing penalties for crimes committed while on cannabis would work better, because of the anxiety that cannabis induces, as opposed to the obliviousness of alcohol.)
I have long thought that it would be great to implement broad based consumption taxes a la a VAT just as the baby boom is drawing upon it's tax deferred savings and going into a phase of significant consumption relative to income in retirement. I guess this would be a baby step. As a consumer of not expensive but not cheap wine and pretty expensive scotch the ad valorem aspect would bite into my wine and spirit budget. Not sure if I would seek out lower cost substitutes or just consume less.
These sorts of takes always remind me of the push to ban large sodas: An idea that has merits on a certain level of policy analysis but turns out to be an utter fiasco when it comes to politics in the real world.
I live in TN and they have changed the laws recently with allowing wine to be sold on Sunday (crazy, I know). But as MY keeps talking about the intersection of alcohol consumption and crime, it would be interesting to see if these spikes in violent crimes lately have had anything to do with a loosening of alcohol laws generally speaking, as we become a more secular society.
As a god fearing Methodist, I support this position.
Also this is a reason why inflation is bad no one in the 70s was like let’s lower alcohol taxes and yet we did.
Thinking back to my Econ 101 course, I remember Sin Taxes like this being touted as highly efficient due to their price inflexibility of demand which would render them highly efficient. Perhaps that's just outdated, but it seems the precise opposite of this thesis.
I think there is little widespread recognition of just how much some people drink. If you have a beer after work every day and then like 5 drinks on friday and saturday you are definitely drinking too much per health recommendations, that's drinking too much per health recs. The type of person who does that probably is thinking "I drink too much but raising the price of a beer by .25 cents would make no difference.
BUT apparently about 75% of the alcohol is consumed by the 10% of people who consume more than the person above. Someone who drinks 10+ drinks a day would surely have their consumption curtailed by a price hike of 25 cents per beer --we're talking about like 1,000/year. And surely it's in this area where you also see a disproportionate share of liquor related deaths. If I remember correctly even for drunk driving the lions share of deaths are from people who are at ~.15 BAC or higher which is very, very drunk.
https://www.washingtonpost.com/news/wonk/wp/2014/09/25/think-you-drink-a-lot-this-chart-will-tell-you/
https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/811654
Back when that "10% of drinkers drink 10 drinks a day" study came out, I looked into it. The survey that they gave people actually had the top 10% of drinkers saying that they had a more believable five drinks a day. However, the survey also was such that if you multiplied it out, it only accounted for about 50% of total alcohol purchased in the United States.
So the authors multiplied everyone's responses by two.
That... seems like a kind of dodgy methodology to me.
I haven't looked further into it to see if anyone has gathered additional evidence/addressed this objection/etc. But I think it's worth looking into before you take it as gospel.
Good critique, but I don't think it undermines the principle of the point even if it is just millions of people having 5 drinks a day.
Also, I would say the DOT stats above that show 70% of the drunk driving fatalities are from drivers at .15 or over indicates there is a HUGE amount of extremely heavy drinking. You need to have 7-8 drinks at a sitting to get to that level... the fact that this cause 6700 deaths in 2010 (sorry the old paper is what I found) indicates tons of people are drinking a lot.
If you are an occasional/moderate drinker I don't even think you'd think to drive after 8 drinks, I'm pretty sure I probably wouldn't be able to find my keys at that point.
I mean, if we said that 0.01% of drives where the person was blind drunk resulted in a fatality, then 6,700 deaths would be 67,000,000 drives in one year. You clearly do not need to have 24M of people getting that drunk each and every day of the year in order to result in 67M drives per year.
Another way to slice it is that on average you get one death per 100 million miles driven-- if .15 drunk driving is 100x more dangerous than average that means 1 death per million miles... 67,000,000 very drunk miles driven. If an average trip is like 4 miles, that would mean 17,000,000 trips, I suppose that could be as few as a million people or less who are making those trips.
Yeah and we tax anyway might as well tax bad things.
In my experience, most people who drink too much are aware in at least some rough sense of how much their habit is costing them financially, and many make periodic attempts (successful or not) to reduce the amount they drink. I find it very plausible that they'll make these attempts more often and more vigorously when alcohol is more expensive.
"But alcohol consumption rises with income, and as these CDC simulations show, higher taxes on alcohol would hit more affluent people harder."
Does more higher income = higher alcohol consumption explain upper-crust people twittering?
elm
surely it explains old people and facebook
I own a craft distillery and this is the by far worst Yglesias take I've ever seen. I'm just being objective.
As someone who doesn’t drink I think this is a great take. I am Totally objective lol.
Well pulling people away from the cheap low quality stuff and over toward the pricier artisanal stuff might be just the nudge your business needs!
I'm sure they could work in some exemptions for distilleries that sell below X amount of beer a year, of course too many exemptions will just swallow the rule itself.
At least we should increase it on big companies like Budweiser, they aren't even American owned anymore!
Also, in all seriousness, Matt often writes about the political palatability of various ideas, how they are sold to the public, and how that interacts with actually achieving real world policy victory.
I'll just say that "I'm going to double the price of your beer" is a massive political loser that would be immediately weaponized by the opposition.
You said this better than I did, wish I'd read your comment first :)
I think "buck a beer" was a winning campaign slogan in Canada recently, by Rob Ford's equally populist brother complaining about nanny state liberals taxing your alcohol.
It's been tried. Jacking up consumption taxes is generally pretty popular, or at least not SO unpopular as to cause political problems.
Popular with the Clinton and Bush administrations, and also tons of state legislatures.
Now, hiking the gas tax at the pump, THAT is unpopular.
elm
voters, unlike drunks, are not politically committed to the god-given right to drive under the influence