Day Twenty Three

Days Without a Cigarette: 22.9375
Days Without Nicotine: 0
Dollars Saved: $61.45
Time Saved: 32 hours, 30 minutes

Edit: There are some bad and misleading arguments in this post. After I posted it, a reader followed up and set me straight on a lot of the stuff I said here. I’m keeping the post here because I (a) don’t want it to look like I’m hiding from my mistakes and (b) I want to preserve the conversation where I was talked out of this opinion. But if you read this post, please read the comments section as well, where much of what I say here is refuted.

Yesterday I wrote a post where I calculated how much money I might have saved if I’d quit smoking in an area with higher tobacco taxes. But after chatting with a few of you and reflecting on the post overnight, I want to clarify the point of it. I only did that because I wanted to make the numbers look a bit better for people who might be considering giving up cigarettes. My numbers don’t look very impressive, so I wanted to emphasize that their results would probably be better than mine in terms of money saved.

What I did not intend (and what at least a few of you seemed to take away from it) was to present an argument for greater taxation of cigarettes. While I’ll concede that there are many good arguments for high taxes on tobacco, there are also plenty of good arguments against it, and I feel the latter outweigh the former.

Now, before I dive into any of those arguments, I should concede that my brain is still in “team smoker” mode. As a smoker, I constantly found myself in debates where I was pitted against the anti-smoking argument. My brain has been running the “pro-smoker” confirmation bias filter for thirty years and it’s hard to step away from that. I don’t have a dog in the race anymore, but all the opinions I’ve bothered to formulate on these topics over the years bias towards the smoker’s position. And while I try to account for my biases in every argument, I also recognize that one can only accomplish that task to a certain degree.

So let me start with the arguments in favor of higher tobacco taxes. As near as I can tell, one of those arguments is not that it reduces smoking. Nations and states with higher cigarette taxes do not have lower rates of smokers, as the map below demonstrates. While there is some indication that young people are less likely to start smoking in states with higher tobacco taxes, there’s nothing definitive. And since the overall rate of smokers doesn’t seem to go down faster in those states with the highest taxes, it’s hard to argue that’s doing much more than delaying the onset of smoking. Now, don’t get me wrong, that’s beneficial – but it’s worlds apart from actually reducing the overall number of smokers.

I should note that there is a correlation between high cigarette taxes and low rates of smoking, but not a causal one. For reasons that should be obvious, areas with lower rates of smokers are less likely to elect people who campaign on higher tobacco taxes or to vote directly to increase those taxes. But you can see from the map that even this correlation isn’t particularly strong.

Of course, the other argument in favor of upping those taxes is that smokers create a new burden on the public coffers and thus they should have to disproportionately pay for it. And that makes sense, but it’s not something that we do with other groups of people that engage in unhealthy behavior. We don’t tax people for being overweight or for failing to exercise sufficiently. That doesn’t mean the argument is invalid, but it’s obviously not such a good argument that it’s spilled into other potential uses.

Now, the argument against raising tobacco taxes should be obvious. It’s a regressive tax that disproportionately effects the poor and the mentally ill. If I told you “Hey, I’ve got a great idea for a tax that will really hammer the poor and mentally ill and rich people will scarcely notice it”, I certainly hope that you’d be primed to reject it before I went any farther. In fact, even if there was a strong correlation between lowering the number of smokers and raising the taxes, this fact alone should make us hesitant. Not that we should abandon that effort, of course, but we should demand a damn high reduction before we consider it an overall good.

Look, it’s obvious why politicians love tobacco taxes. You can raise them indefinitely and virtually nobody complains. Even the smokers who are paying them by and large just nod along and say “yeah, I kind of deserve this.” But what we’re effectively doing is driving poor people further into poverty by adding cost to a substance that they’re already addicted to.

Don’t get me wrong, I’m happy to change my opinion on this. I formulated this argument when I was ‘Team Smoker’, and while it still strikes me as a pretty good argument, I get that my perspective steered my thinking. Feel free to disagree below, and keep in mind that my opinion on these matters has never been more malleable.

Published by Noah Lugeons

Noah Lugeons co-hosts a bunch of podcasts: The Scathing Atheist, God Awful Movies, The Skepticrat, and Citation Needed

30 thoughts on “Day Twenty Three

  1. I had a vague recollection that I had read an article by the Australian Department of Health that claimed that the tax was a major factor in reducing the smoking rates in Australia from ~25% to ~12% in 20 years. I can’t find that article so maybe it doesn’t exist.

    I did find this: https://www.nber.org/papers/w18326

    It claims that a tax of 100% would reduce smoking by 5%. The tax in Australia is far higher than 100% so maybe that’s the key? I don’t know.

    As to the fact that it’s a tax on the poor – no question about it, and it’s a significant problem. On the other hand, if it reduces smoking rates among the lower SES it might be a net positive, but I don’t know that it does.

    As I mentioned previously, I was never a smoker and pretty much no-one around me smokes, so my desire to see smoking rates drop is based on a desire for harm minimisation.

    Like

    1. Yeah, that seems in line with what I’ve seen. Huge tax increases may lead to very small decreases in tobacco use, but even that isn’t definitive. The only thing that seems clear from the data is that if it does have an effect, it’s not very big.

      Like

      1. You know what’s funny? We both read the same article, but our different perspectives led us to see very different things!

        What I took from it was that a high but not unimaginable tax (hundreds of % in the case of Australia, but let’s say 100%) can achieve great reductions, because 5% is literally millions of Americans.

        Bias comes in all kinds of forms. Mine appears to be the Australian experience…

        Like

  2. Noah

    I’m surprised you have come to the conclusion after a cursory examination of a few data points. One of the few things in economics which actually works is that, assuming consumers have a choice, raising the price of something causes demand to fall. In other words if you want to discourage a behaviour you tax it. Like I said, this is one of the very few things which is actually correct in economics so economists are generally very proud of the fact that it is true. There is massive and overwhelming evidence this is the case as an examination of the literature will show, and a lot of that literature is specifically associated with the impact of taxation on tobacco consumption.

    In general the conclusion is that you need really high taxes to discourage smoking and, to the best of my knowledge these do not exist in the US. (True fact: there are entire countries that are not in the US!)

    Another factor is that nicotine is highly addictive and most smokers are hooked as children. Making a pack of cigarettes cost $25 is a sure way to discourage that. Of course we have the slimy bastards at Juul who have figured a way around that little problem and are getting an entire generation addicted to the stuff but that’s another matter.

    The argument that tobacco taxes disproportionately impact the poor is incomplete because they also suffer disproportionately from tobacco related health issues and because they also are disproportionately smokers (and, in the US, generally lack access to medical services to treat those issues). You use nicotine patches, correct? Does it not follow that increasing the cost of tobacco (which is really, really, bad for health) and decreasing the cost of alternatives such as nicotine patches (which is bad, but much less bad) would disproportionately help the poor by getting them to quit?

    To assume that taxation does not impact consumption also implies that, for example, carbon taxes should not be imposed – or is tobacco “special” in some way?

    Like

    1. Everything I’ve read from a credible source suggests that the benefit of sin taxes is either very low or non-existent. As I’ve said, I’m happy to re-examine that conclusion; but all you’ve provided here are a couple of condescending statements that tangentially impact my argument. As a rule of thumb, if you find yourself summarizing another person’s argument in a way that they would all but definitely disagree with (“taxation does not impact consumption”), you’re not doing it right.

      Couple of other issues: I think it’s pretty obviously an addictive substance is ‘special’ compared to a non-addictive one when it comes to purchasing habits. In the 3rd paragraph of your comment, you disprove your own point at the end and then dismiss it as ‘another matter’. But I think your most egregious error comes in paragraph 4 where you claim that increasing cigarette taxes would disproportionately help poor and mentally ill people quit. This is only correct if I’ve already accepted your conclusion, and thus cannot be offered up as an argument in favor of your conclusion. Also, you site no sources and offer no links to credible refutations of my argument.

      You’ve done a poor job stating your position. That doesn’t mean it’s wrong, but it’s wholly unconvincing.

      Like

      1. Noah

        I realize that my comment was excessively snarky but it was written from the perspective of someone who listens to your podcasts. Evidently that is unacceptable here and I apologize. It also reads a bit incoherent, however, that was likely due to the fact I had (and still have) not slept much in several days and that your comments against tobacco taxes are contrary to the views of the overwhelming majority of health policy experts globally.

        I do not understand the first and last sentence of your first paragraph of your reply:

        “Everything I’ve read from a credible source suggests that the benefit of sin taxes is either very low or non-existent. … As a rule of thumb, if you find yourself summarizing another person’s argument in a way that they would all but definitely disagree with (“taxation does not impact consumption”), you’re not doing it right.”
        Unless I am completely missing the point it seems to me that you are arguing that sin taxes don’t work and that yet you agree with the statement that (removing the double negative) taxation impacts consumption. Because you explicitly state the benefit of sin taxes is very low, I assume the second statement reflects your reaction to my snarkyness rather than your position.

        Let me try again.

        You only provided a snapshot of a single data set at a single point in time. You state
        “For reasons that should be obvious, areas with lower rates of smokers are less likely to elect people who campaign on higher tobacco taxes or to vote directly to increase those taxes.”
        This is a potentially valid conclusion but unsupported. Although I am not intimately familiar with US demographics, alternative plausible explanations would be that lower smoking rates are correlated with high income and higher education by state as they are in the general population, or that states with higher income and education have been more committed to public education about tobacco, etc..
        Taxation is a necessary evil: it not only funds government but it is an important aspect of policy. Most, but not all, tax policy is crafted after internal studies and consultations which determine the best path to meet policy objectives. If you have ever done corporate taxes you know there is a web of credits, etc., which have been put in place and frequently modified in order to encourage or discourage certain types of investments and activities. Consumer taxes have to be much simpler in order to be manageable but in any case such studies are done by economists and related experts in the field. These policy advisors suggest the most effective path to achieving the objectives.

        I’ve worked with policy advisors and economists (though not on health related issues) and they are usually remarkably competent. In fact, although I have never worked for government, I have met many more idiots in high office in the private sector than in government. The global consensus of experts is that high taxes are a suitable way to meet the policy objective of reducing tobacco use. There evidence is clear that taxation reduces smoking (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3228562/ , https://www.who.int/bulletin/volumes/94/4/15-164707/en/ , https://www.nber.org/papers/w18326, etc.), and I have never seen a study which concludes differently. You say your reading suggests otherwise so if you have references supporting that view I’d like to see them. In fact I suspect global policy advisors would like to see them as well.

        Simply stating that

        “Look, it’s obvious why politicians love tobacco taxes. You can raise them indefinitely and virtually nobody complains”

        is akin to a conspiracy theory where diabolical politicians exploit hapless smokers. It may be the case at some times and places but it is far more likely that the politicians have been convinced that reducing tobacco consumption is a valid policy goal, and their experts have told them that a high tax on tobacco is a good way to accomplish that goal. Again, it seems that all the evidence is on their side.

        In your original post you say

        “We don’t tax people for being overweight or for failing to exercise sufficiently.”

        These are not analogous to taxing tobacco: nobody is proposing to tax smokers for being smokers. Moreover, people are overweight because they consume too many calories. While the popular view is that the obese are gluttons, over consuming metabolic requirements by a few hundred calories a day (i.e. a latte or a couple of bananas or a doughnut, etc) leads to morbid obesity within a few years. Unfortunately, we have not developed a means to tax excess calories.

        Getting back to sin taxes, because you didn’t cite anything in particular I don’t know what sin taxes you are referring to. Perhaps you were referring to attempts to tax soda. These in fact result in reduced soda consumption but, because they were implemented as a result of a flawed hypothesis that soda consumption is the cause of obesity, the impact on obesity has been negligible. Unsurprisingly, since the cause of obesity is excess calories, people simply replaced their excess calories from soda with excess calories from other foods and beverages.

        One thing that is quite clear from the literature is that modest increases in taxes have a negligible impact on adult cessation. This is besides the point unless one assumes the objective is exclusively to get adults to completely stop smoking. In fact, a primary objective is to reduce the number of tobacco addicts: since the majority of people get hooked when they are young, and the youth tend to have limited resources, making tobacco expensive fulfills that objective quite well. If I were to guess that is a major reason overall smoking rates have declined over the past few decades: few people start smoking in their 30s. A secondary objective is to reduce consumption by smokers since health effects are proportional to amount smoked and years smoked. A tertiary objective is to get adults to stop smoking, which it does, though not as much as would be hoped. This likely reflects the highly addictive nature of nicotine and the limited utility of smoking cessation aids.

        Treating the issue as binary as your charts do (i.e. looking at percentage of smokers at a specific time) does not capture the impact on fewer new smokers or the health benefits smoking fewer cigarettes.

        If we take it as a given that tobacco consumption should be reduced there are two practical ways to reduce smoking: through public education and taxes. Nonetheless, although it has been well established since the 1950s that cigarettes are a major health issue, smoking rates only really began to decline as governments implemented high taxes on tobacco ( see https://ourworldindata.org/grapher/sales-of-cigarettes-per-adult-per-day?country=JPN+DEU+USA+FRA+CHE+ESP+HRV+AUT+CAN ). Countries which implemented high tobacco taxes typically also engaged in public education, banned advertising, etc., so it is hard to say what specific measure helped but it is safe to say high taxes didn’t hurt.

        All consumption taxes, unless explicitly targeting luxury goods, have a disproportionate impact on the poor. This is true whether discussing tobacco, alcohol, gasoline, carbon taxes, or fees on plastic bags. It happens that smoking is far more prevalent among lower income strata than higher income strata https://www.cdc.gov/tobacco/disparities/low-ses/index.htm. The chart stops at 2x the poverty rate but in my experience very few rich people smoke. In fact, I know a lot of them and I don’t recall meeting a smoker from a wealthy demographic over the past 20 years. This wasn’t the case 40 years ago when a large proportion of adults smoked, including the well off.

        It seems that public education, bans on advertising, etc., have less of an impact on the poor than on the rich. Because many more poor people smoke it is clear that anti-smoking policies literally have to target the poor. In fact, because very few well-off people smoke, any attempt to target them is a waste of money.

        Given that the health burden of smoking is largely born by the poor, what would you propose policy makers do? With respect to taxation should governments lower taxes to lessen the burden of addiction, or increase taxes to reduce the number of poor smokers picking up the habit and reduce amount smoked, or leave things as they are?

        If one truly cares about the poor, one doesn’t want them to die young so the logical choice would be to work towards reducing the number of smokers and amount smoked. Taxation does that.

        Best of luck in your journey to quit smoking and keep up the good work.

        Like

      2. Now that’s what I’m talking about. Great job.

        So would you agree with the following statement: “Raising taxes on tobacco helps reduce the future number of smokers at the expense of current smokers”?

        Like

      3. Noah

        This reply seems out of sequence so I’ll quote your reply here:

        “So would you agree with the following statement: “Raising taxes on tobacco helps reduce the future number of smokers at the expense of current smokers”?”

        I agree partially. More correct would be

        “Raising taxes on tobacco helps reduce the future number of smokers, reduces the amount smoked, and results in some smokers quitting (though not that much) at the expense of current smokers”

        Based on what I have read it is not clear to me whether quit rates cited in research are for a particular smoker for a particular period (i.e. 5% quit in a year, resulting in 20% fewer smokers are 4 years or these are total (i.e. 5% quit after being followed for 20 years). If it is 5% per year it is obviously far more significant than only 5%.

        I don’t even have anecdotes to frame this: with a single exception every smoker I’ve known has either quit or died of smoking related illnesses – including my father and brother.

        Like

      4. I have no idea why WordPress makes me approve every one of your comments individually. With everyone else, once I approve a comment they can just comment on the blog. Yours always show up awaiting approval. No idea why.

        Just want you to know I’m not purposely censoring your replies.

        Like

      5. No worries. Maybe they know I’m a dick and figured they save you the trouble. It takes me to a WordPress site after I sign submit the comment so you’d think they’d know me. After all, I’ve have Worpress sites for a few years and never had an issue.

        Like

  3. Noah – Just curious – I submitted a comment to you but I’m having issues with WordPress. Did you reject it or not get it?

    Like

    1. Yeah; not sure why, but it kicked it to the spam folder. Regardless, all comments have to be approved (it’s the only way to keep the comments from being an endless series of bots), so pretty much everybody has to wait for their first comment to show up regardless.

      Like

      1. No worries. Maybe they know I’m a dick and figured they save you the trouble. It takes me to a WordPress site after I sign submit the comment so you’d think they’d know me. After all, I’ve have Worpress sites for a few years and never had an issue.

        I should add: when showing how X may or may not cause Y, a simple point in time is not sufficient. You are better showing trends. So, for example, if a state had a smoking prevalence of 29% with a tax of 20%, and then they raised the tax to 50% and the smoking prevalence changed to 20%, that would strongly suggest an association. If the rate remained unchanged that would argue against causation. It should be clear that a simple point in time provdes much less information of rates of change, which are inherently time-related.

        This chart for South Africa provides one such example https://www.tobaccofreekids.org/global-resource/tobacco_tax_success_story_south_africa I found another, much more ancient one for Canada when we bizarrely decided to cut tobacco taxes and it showed, as would be expected, that consumption rose. https://www.cmaj.ca/content/cmaj/156/2/187.full.pdf

        Like

  4. No worries. Maybe they know I’m a dick and figured they save you the trouble. It takes me to a WordPress site after I sign submit the comment so you’d think they’d know me. After all, I’ve have Worpress sites for a few years and never had an issue.

    I should add: when showing how X may or may not cause Y, a simple point in time is not sufficient. You are better showing trends. So, for example, if a state had a smoking prevalence of 29% with a tax of 20%, and then they raised the tax to 50% and the smoking prevalence changed to 20%, that would strongly suggest an association. If the rate remained unchanged that would argue against causation. It should be clear that a simple point in time provdes much less information of rates of change, which are inherently time-related.

    This chart for South Africa provides one such example https://www.tobaccofreekids.org/global-resource/tobacco_tax_success_story_south_africa I found another, much more ancient one for Canada when we bizarrely decided to cut tobacco taxes and it showed, as would be expected, that consumption rose. https://www.cmaj.ca/content/cmaj/156/2/187.full.pdf

    Like

    1. You’re overstating the case against these statistics. Let me be clear here – if the data showed no correlation between higher tobacco taxes and rates of smoking in US states at any one time, that would be sufficient to prove my point. What I’m conceding here is that the data don’t show that. But you’re assertion here is incorrect.

      You would be correct going the other way, of course. Showing a single chart with correlation (for example, the South Africa chart) would not be sufficient, as this could be explained by confounding factors.

      Like

      1. I think you’re making a logical mistake here, Noah. It is possible for something to be true in general but not in some cases. Confounding factors work in both directions. This is not, of course, a comment on the specific graphs presented here,

        Like

      2. “Does is still sound like a logical mistake?”

        When you put it this way, no, but that’s a disingenuous way of phrasing it. Here’s an analogy for you:
        Arsenic levels in the water varies significantly between the states, yet there is no difference in the rate of arsenic poisoning between the states.

        Here are two possible explanations:
        1. Arsenic is not toxic
        2. The level of arsenic in the water is below toxic levels in all states. Had the levels been higher we might have seen differences

        Your logic precludes #2 being correct.

        The same applies to your rejection of Brian’s arguments. You may still be correct in your conclusion because the evidence is weak, but the logic of “single negative sample” is not appropriate in this case (and it’s worth noting that there are many cases where it would be completely appropriate).

        Like

      3. It is impossible to discuss such things on this time horizon but let me try again.

        Causation is like a rope connecting something: you pull on the rope and the thing at the other end moves because there is a causal relationship. Of course, if there is a black box between you and the other end of the rope you can’t be sure that there is a causal relationship but if there is a reasonable theoretical framework for what you are seeing then, assuming causation is a reasonable first order approximation to what is going on.

        Note that the position of the thing at the end of the rope at a particular time does not provide any information regarding causation. Causation has, by its very nature, a time dimension.

        In the South African and Canadian statistics, taxes were initially high and smoking rates were a particular value. This tells us nothing about the impact of taxes on smoking rates. However in both instances (and I would wager in every example one could find) decreasing taxes led to increased smoking rates and then increasing taxes resulted lower smoking rates. This is as would be expected by basic economics: raising the price of something typically results in lower consumption and lowing the price results in higher consumption. There could have been (and likely were) other issues at play but these experiments strongly support the hypothesis that there is an inverse causal relationship between tobacco tax rates and smoking rates. They do not support the hypothesis there is no relationship.

        In contrast, your snapshot in time (such as a map of tax rates and smoking rates at a particular time) provides no information regarding causality because there is no time dimension. If we had a graph of, for example, tax rates and smoking rates over time by state, or a population weighted graph of overall tax rates and smoking rates over time and that showed that there was no impact then one might be able to conclude otherwise.

        I did manage to find a US study https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6147505/ which looks at taxes and cessation and concludes:

        “It appears that increased taxes also have an association with the percentage of smokers that tried to quit the past 12 months. Our analyses suggest that an additional $0.25 in cigarette tax was associated with an estimated 0.67% increase in the percentage of active smokers that reported trying to quit smoking (Table 1).”

        Unfortunately, it also notes

        “Paradoxically, when stratifying by income, cigarette taxes had the least impact in smoking prevalence for those with the lowest income, with a $0.25 tax increase associated with a minimal change (0.09% reduction) in smoking prevalence among participants with income <$25,000”

        Which supports your assertion that taxes essentially target the poor.

        Like

      4. You’re simply incorrect on this. If I’m trying to establish whether there is a geographical correlation in the name Fred, simply knowing the percentage of people named Fred in each state is sufficient information to draw a negative conclusion. Unless all the taxes were changed really recently, looking at fifty data points at a single moment and knowing both the tax rate and the rate of smokers would be plenty of information to reject the assertion that tax rate reduced smoking.

        Like

      5. In order to know whether increasing taxes results in decreased smoking you have to know the tax rate and smoking rate at a particular time, then change the tax rate and look at the smoking rate at some time after.

        So, it State A has a smoking rate of 25% and a tax rate of 20%, then increases the tax rate to 30% and smoking goes to 22% that is suggestive of a causal relationship. If the smoking rate stays at 25% that is suggestive of a non-causal relationship. If State B starts at a smoking rate of 19%, raises taxes from 10% to 15% and ends up with a smoking rate of 18%, that is suggestive of a causal relationship. However, looking at State B’s smoking rate of 18% and tax rate of 15%, and State A’s smoking rate of 22% and tax rate of 30% at that specific time tells you nothing whatsoever about whether raising taxes has an impact on smoking rates.

        Changes are inherently time related.

        Like

  5. I don’t know if this will be tied to this reply,

    ” If I’m trying to establish whether there is a geographical correlation in the name Fred, simply knowing the percentage of people named Fred in each state is sufficient information to draw a negative conclusion. ”

    But I’ll try. We are trying to establish whether raising tobacco taxes is a good policy move, assuming the objective is to reduce smoking. Because of that, what we need to know is the change of smoking relative to the change in tax, which is necessarily time dependent. If a low tax/low rate state raises its tax and the rate declines further (as per essentially all relevant studies) then that is meaningful. Simply knowing the current or even past smoking and tax rates do not inform as to whether changing the rates had an effect.

    Like

    1. No, but it does tell us whether or not the two numbers are correlated. There CANNOT be a causal effect without a correlation. Again, assuming that tax rates weren’t identical and then changed shortly before our snapshot, this would be sufficient data to reject the argument.

      Like

      1. You’re oversimplifying the situation. What the numbers show is that there is no correlation under specific circumstances (eg tax rates below a certain threshold, and potentially other factors). It’s definitely enough to make any serious observer think very hard about what’s going on, but it’s not enough to reject the hypothesis.

        Like

      2. That’s fair. But I think it speaks more to a carelessly worded assertion on my part. I don’t think anybody would argue that a tax rate that amounted to thousands of dollars per pack wouldn’t reduce consumption; so some threshold should be assumed. That being said, I didn’t state what it was.

        That being said, if one were to show that there was no correlation between tobacco taxes and consumption, it would show that (within the bounds of exiting differences in taxes within the sampled area), raising tobacco taxes is not an effective way to reduce consumption. Even if we grant additional factors, unless we can identify and eliminate them, that would not affect my assertion.

        Like

      3. I think it’s a little more than that, because Brian and I presented a plausible case, with evidence, that taxes at realistic – though high – levels seem to be effective, at least in some circumstances.

        Just to clarify (and probably end this thread): I don’t think it’s an open and shut case either way, but I do think the evidence for the impact of taxes can’t be dismissed based on the information you provided.

        Like

      4. Yeah; to be clear, I conceded that point based on the additional evidence you provided. At this point we’re just arguing about the A-B =/= C kind of stuff. I’m only defending the point (as modified by your previous comment) that if no correlation is found between tax rates and consumption, then (within the range represented by the data) there is no causal link between taxation and consumption. This is what you referred to as a logical error earlier.

        For clarification, it was presented in response to Brian’s claim (or suggestion) that one cannot determine a causal relationship without looking at multiple times. So unless you’re defending that proposition (a causal link can not be determined without sampling data over a range of time), we’re arguing past each other at this point.

        As to the larger argument, I think Brian has done a good job presenting enough evidence to tentatively conclude that increasing taxes decreases smoking; though to be honest, the effect seems to be pretty small even with massive increases, so I don’t know that the argument over whether it’s a good way to go about it is settled.

        Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: