Why smoking should be regulated, but not by the government.

My previous post triggered a lot of comments accusing me of misunderstanding Coase’s Theorem.  It is certainly possible that I did, but I am going to argue that many of my commenters are misunderstanding that famously misunderstood Theorem.

When I was young I constantly breathed in second-hand smoke, and paid no attention to it.  Lots of people smoked.  My dad smoked all the time.  Even university professors smoked in class.  Then smoking began to decline and people started becoming much more sensitive to smoke.  I don’t know if it was generation X, Y, or Z, but at some point we became almost absurdly oversensitive to smoke.  I’m not a smoker but I find it almost comical that people don’t want to stay in a hotel room not because someone else in the room is smoking, but because someone in the past smoked in the room.  But enough reactionary whining from me.  As an economist it is my job to maximize utility.  If people say they are extremely bothered by even tiny amounts of second-hand smoke, it is my job to do something about it.  Even though I know deep down that all attempts to make society more comfortable from this point on are doomed to fail.  The more discomfort we remove, the more sensitive we become to the remaining annoyances.  It’s all about set points. 

Suppose the state office building in Boston has both smokers and non-smokers.  Over time the non-smokers become increasing numerous and increasingly assertive.  At what point should the owner of the building do something about the problem?  Coase understood that there were two ways of thinking about the issue.  If it was easy for people to negotiate side deals with office workers next door, the non-smokers would be able to bribe smokers into not smoking whenever the benefits from a smoke-free environment exceeded the disutility incurred by smokers having to go outside to light up.  All would be well.  But Coase also realized that transactions costs might easily make those side deals too costly to implement.  In that case he suggested that regulation might be appropriate.  Who should regulate the activity?  Whoever owned the space where the economic harm occurred.  If the annoying smoke occurred in a government office building, then the government should regulate the activity. 

Now consider the same example, but this time with a private office building.  Again, it might be best for the owner to simply let the employees work out side deals, and not have any regulation at all.  Or the best solution might be to regulate the activity.  But here is where it seems to me that some of my commenters went off course.  They seemed to assume that if regulation was called for, then it should be done by the government.  In fact, for two reasons government regulation would be far inferior to regulation by the owner of the office building:

1.  The first problem is that government regulations tend to be one-size-fits-all.  So they might ban smoking in all offices.  Or they might ban anti-smoking rules in all offices, claiming (accurately) that these rules discriminate against potential office workers coming from low SES backgrounds.  Either way, it will be a one-size-fits-all regulation.  But that regulation would not be optimal in all offices.  It is very likely that the best solution in some office buildings is to allow anyone to smoke, and in other office buildings to ban smoking.  It all depends on the relative proportion of non-smokers, whether the non-smokers are tough old guys like me who don’t mind smoke or delicate Gen X’s who can’t tolerate it, and how much smokers enjoy smoking.  (I might lose half my readers after this post.)

2.  Government regulators may have an incentive to maximize votes, not economic efficiency.  Suppose 80% of people are non-smokers, but the utility from smoking (in dollar terms) is ten times the disutility from second-hand smoke.  In that case it is socially optimal to allow smoking.  But since 80% of voters are non-smokers, the government might ban smoking in the workplace anyway.  

In contrast, the owner of the firm has an incentive to set in place the socially optimal smoking policy.  The easiest case to consider is where the owner is not even present, but adding an on-site owner doesn’t change anything.  If the owner is not present at the work site, he has an incentive to put in place a regulation that maximizes the aggregate utility of his workers (or customers, if we are considering a restaurant.)  This is because either choice of regulation can be imposed at zero direct cost, so the only thing that matters to the owner is the indirect cost of the regulation, which is the impact on his wage bill.  In generally, it is easier to find good workers willing to work at low wages if you have a pleasant work environment.  If you are a non-smoker you might assume that a pleasant work environment is one that is free of smoke.  But if you are a smoker then you might think that a pleasant work environment is one where you are free to smoke. 

The point many people miss is that all of these costs and benefits of smoking are internal to the firm.  A factory-owner might not care if its smoke bothers people downwind, as they generally don’t have to pay compensation.  That’s an externality to the firm.  But a company will care very much if second hand smoke annoys non-smoking workers, as they would then have to pay higher salaries to get people to work there.  Of course it is possible that people would sort into smoking and non-smoking firms, and that may be an efficient solution in some cases.  But even where it isn’t feasible the owner has an incentive to institute the smoking policy that makes for the happiest employees on average.  This will minimize his productivity-adjusted wage bill.

When deciding on optimal employee amenities one must consider costs and benefit.  For example, does beautifying the office with flowers and artwork pay for itself in higher employee morale?  A smoking regulation has zero direct cost, so you simply compare the negative consequences to one side with the positive consequences to the other side.

Part 2.  Beyond Villains and Victims

One thing I learned from Coase is that it is often misleading to think of social problems in terms of villains and victims.  Is the problem of bad smells from pork processing plants caused by the plants themselves?  Or by the suburban developments which arrive after the plant was already built?  Coase showed that this was a meaningless question.  The only relevant question is what set of property rights are most likely to produce an economically efficient outcome.

It seems to me that many people are so appalled by smoking that they assume smokers are the villains.  In contrast, I see the over-sensitive GenXers as being the villains.  Coase would say we are both wrong.  The only villains here are those who advocate inefficient government regulation when a much more efficient solution is to let the private market sort things out.  The government should (at most) regulate smoking in government buildings.  Note that I cannot be accused of being a reflexive opponent of regulation.  I support carbon taxes because the area of damage (the atmosphere) is commonly-owned, and the cost of privately negotiating an optimal emissions framework is prohibitively high.

It’s quite possible that everything I have written is wrong.  This is certainly not my area of expertise.  Or perhaps my argument is correct but has nothing to do with Coase’s Theorem, and was already widely known by the earlier Pigovians.   If a GMU blogger or someone like Steven Landsburg comes over here and sets me straight, I will gladly admit I am wrong.  But so far in the comment section I’m not seeing arguments that dissuade me from the views I expressed in the previous post.  I still say Richard Thaler is wrong, second-hand smoke does not call for government regulation.

Update.  I could have avoided wasting on hour writing this post if I had just read David Henderson first.  He links to an article that confirms the externality problem does not occur in privately-owned properties such as restaurants and offices.  He also answers my question about whether Thaler and Sunstein spend equal time trying to replace burdensome hard paternalism with softer libertarian paternalism.  The answer is yes.  That means I’m officially on board.  But first Thaler has to stop saying those silly things about second-hand smoke.


Tags:

 
 
 

62 Responses to “Why smoking should be regulated, but not by the government.”

  1. Gravatar of David Pinto David Pinto
    9. April 2010 at 18:56

    In the eighties, I worked for a small start up that rented space from a larger company. None of us smoked, so we instituted a no smoking policy, and future employees did not smoke in the office.

    We had to walk through the offices of the larger company to get to our area, and there was a combination of closed offices and open cubicles, but no apparent reason why anyone was assigned an office or cubicle. It turns out they solved the smoking problem by putting the smokers in offices, the non smokers in cubicles.

    It was also my understanding that smoking in offices was often stopped due to insurance issues. It was cheaper to get either building or health insurance if smoking was not allowed.

  2. Gravatar of Doc Merlin Doc Merlin
    9. April 2010 at 19:13

    I agree Scott, couldn’t have said it better.

  3. Gravatar of rob rob
    9. April 2010 at 20:14

    I don’t understand any of the theory of Coase’s Theorem, but in practice I have observed the unintuitive reactions of patrons to smoking bans in bars.

    Forget office buildings for a moment. Just think bars. Places where people go to sin. In my city, there was a huge uproar when a smoking ban in bars was proposed. None of the bar owners seemed to like the idea. And keep in mind that not a single bar ever, on the planet, as far as I’m aware, ever voluntary instituted a private smoking ban before the gubment stepped in.

    Then smoking gets banned in bars and the majority seems to favor it. Even I favor it, and I was against the ban before the ban. Perhaps people simply don’t know what they want until they get it.

  4. Gravatar of wcw wcw
    9. April 2010 at 20:43

    What Rob said.

    There is no hard paternalist law I have appreciated more in my entire life than the smoking bans. I would rather drop vehicle licensing and traffic laws than smoking bans.

  5. Gravatar of Winton Bates Winton Bates
    9. April 2010 at 21:06

    I agree with Doc Merlin!

  6. Gravatar of Doc Merlin Doc Merlin
    9. April 2010 at 21:08

    @rob:
    “Then smoking gets banned in bars and the majority seems to favor it. Even I favor it, and I was against the ban before the ban. Perhaps people simply don’t know what they want until they get it.”

    Well, yes, the smokers don’t go to the bars anymore, so it looks like all the people at bars are now in favor of it.

  7. Gravatar of Mikkko Mikkko
    9. April 2010 at 21:14

    There is one crucial difference between the owner mandating rules and government mandating rules that has to do with human social systems. I don’t know how to model it in economic terms, though.

    The point is that if it is the owner, who also happens to be the owner of the company that the smokers work for, then the workers will blame him. They may well be angry and this will reduce their willingness to work. If it’s the government, then no such drop should be expected as the owner cannot do anything about it, so it’s not his fault.

  8. Gravatar of scott sumner scott sumner
    10. April 2010 at 05:21

    David, All good points. BTW, I also oppose company health insurance subsidies, which do all sorts of harm. One of those side effects is that they cause firms to pressure employees into more healthy behavior.

    Thanks Doc Merlin.

    rob, I have to say that your argument is one of the more far-fetched that I have ever read. We all wanted to behave a certain way, there was nothing stopping us, but we just couldn’t do it until the government forced us to. It might be true, but is that a plausible argument for public policies in a world of great uncertainty? I don’t think so. Believe me, if millions of people were yearning for smoke-free bars then entrepreneurs would have supplied them.

    I think one thing people forget is that the big reduction in public smoking has caused people to become very sensitive to smoke. When I was young smoke didn’t bother me in the slightest, and I don’t recall many other people being annoyed either. Now people are much more sensitive, that makes it seem like the smoking bans did much more good than they really did.

    WCW, I understand your point, but remember that how you personally feel about the effects of a law are COMPLETELY IRRELEVANT TO WHETHER THE LAW SHOULD BE ENACTED. I recall conservatives I’ve met who support manned space exploration. When I ask why the government should waste money on these projects, when unmanned exploration is far cheaper, they say they are really interested in space. Yes, but one’s personal feelings about a public policy are irrelevant, all that matters is its impact on AGGREGATE UTILITY.

    Thanks Winton.

    Good point Doc Merlin.

    Mikkko, This is exactly why it’s best to have the owner make the decision. He knows that employee morale will affect productivity, thus has a strong incentive to come up with the policy that leads to the greatest employee satisfaction. The government has much less incentive to maximize welfare, their incentive is to maximize votes.

  9. Gravatar of DanC DanC
    10. April 2010 at 05:26

    When Ohio passed the ban on smoking in bars and other public space it was done to protect the workers at those establishments. They really ignored the issue of second hand smoke on other patrons.

    I suppose Coase could argue that workers could demand extra income to compensate for working in a smoke filled bar or attract workers who don’t care. But such arguments would lead to the end of OSHA.

    Moreover the bar owners are still at risk for lawsuits from employees claiming illness related to working in a smokey workplace. Given the casino nature of our legal system, it is hard for the owners of the pub to do anything but insure against future possible claims.

    The risk to employees was why a smoking and non-smoking section of the public establishment was rejected. That could protect patrons but not employees.

    Bar owners who seek to ban smoking in their bar would be at a competitive disadvantage. Heavy smokers are often heavy drinkers. While they could try to market themselves as a healthy drinking establishment, I’m not sure the market is that large.

    In that case, bar owners, who want to protect themselves from employee health claims, could prefer a government ban.

    It is a bit like the claim of Southern restaurant owners prior to civil rights laws. Some restaurant owners claimed that they would be willing to serve blacks but that they feared that whites, in great enough numbers, would switch to competitors who continued to ban black diners. Federal laws imposed a cost, whites who switched to private clubs, but it also gave a benefit, prevented restaurants from competing purely on the race of the diners.

    I don’t want to claim that smoking is a constitutional problem. Just that government regulation can, rarely, lead to a second best solution.

    It is just hard to imagine that society would permit a binding contract between employers and employees to enter a contract where the employee agrees to risk future health in return for employment. Such contracts are written or implied all the time but they are filled with risk for the employer. If a potential jury pool is 80% anti-smoking, do you want to take the risk as an employer. (Given the short run nature of most privately owned firms it is not surprising that the deeper pocketed chains tend to be more supportive of such bans.)

  10. Gravatar of Mark A. Sadowski Mark A. Sadowski
    10. April 2010 at 05:38

    Scott,
    I guess I’m trying to get inside Thaler’s skull and play devil’s advocate. The only way I think one can justify government regulation of smoking in public places is by legally “endowing” people with the right to clean air no matter where they go. Then in effect the employer, the pub/restaurant owner etc, effectively do not “own” the air in their buildings, the employees and customers do.

    Even with the concept of a legal right to clean air it seems to me that in self contained “semi-private” public environments with few people and little to no turnover in population (say like a small office with absolutely no public interaction) it might be feasible to subject it to a vote with everyone having a veto. However, in the vast majority of cases it would probably be most efficient to simply have an outright ban.

  11. Gravatar of Mark A. Sadowski Mark A. Sadowski
    10. April 2010 at 05:46

    By the way I grew up in a smoke filled house. Also I seem to remember a Korean War film where a doctor in MASH unit screamed at a medic to “get that man a cigarette” as though it were of medical necessity. Our culture has indeed changed quite a bit.

  12. Gravatar of StatsGuy StatsGuy
    10. April 2010 at 06:23

    Can we please differentiate between agreeing with a government ban on smoking in bars (which I don’t care for), and accurately representing the Coase Theorem?

    Your argument is that arguing for a government smoking ban is a “violation of the Coase Theorem”. I disagree.

    First, it’s a theorem – that means that to demonstrate a “violation” you need to show that if the assumptions are true the outcome fails. Second, Coase never claimed the assumptions are true all the time – indeed, he noted that most of the time they are not true to varying degrees.

    So really, the empirical argument is whether or not the assumptions hold. Yet they never do hold completely, so the pragmatic question is really whether they hold sufficiently that it’s more efficient to allow private negotiation/contracting than government regulation. (I think yes, private contracting is better here – but again, I don’t really care since I don’t frequent bars.) You attack this question above – but the previous post did not.

    Libertarians tend to like Coase’s theorem without really doing due diligence on the assumptions – and often make blanket statements of the type “Coase Theorem says private contracting is better” without adding the “when”. Moreover, they often miss some implications of the Coase Theorem – one of those implications is that if the assumptions really hold true throughout society, then IT DOESN’T MATTER whether you regulate or don’t regulate. The regulation would take the form of exactly what the private outcome would be.

    Above you discuss why that is not the case here when you talk about election incentives – but you are missing a key point. You are assuming that contracting and low transaction costs are not available in the political arena -that smokers can’t offer non-smokers side payments. That interest groups can’t marshall funds to pay for regulation. Public Choice theorists (aka, most regulation benefits the regulated industry) might disagree.

    In essence, your argument boils down to this: Transaction/contracting costs are RELATIVELY lower for the private firms than they are for the regulator in this case – even taking into account externalities and collective action costs etc.

    I happen to agree with you. In this case.

  13. Gravatar of StatsGuy StatsGuy
    10. April 2010 at 06:37

    One last point on Coase and rent seeking:

    One of the challenges to Coase’s Theorem in reality has to do with rent-seeking behavior. Let me ground it in two examples that are relevant to your post above, and offer you a little challenge:

    1) Some workers smoke. Other workers are highly allergic to one type of cigarette, but not another. The smokers don’t have a preference between types of cigarettes. Side payments are allowed, contracting is free, and transaction costs are zero.

    What happens?

    2) 80% of voters don’t smoke, and never visit bars. Nor would they visit bars even if smoking was disallowed, and let’s ignore “moral utility” or disutility. 20% smoke, and do visit bars. Side payments (contracting) are allowed in the political arena, and transaction costs are zero there.

    What happens now?

    What does this say about whether or not assignment of blame (aka the initial distribution of property rights) matters?

    [BTW, when I say "you attack this above", I mean attack in a good way - as in, engage.]

  14. Gravatar of StatsGuy StatsGuy
    10. April 2010 at 06:52

    One final amusing anecdote…

    I find that children are nascent economists. Mine especially.

    Yesterday, my daughter was playing with her younger brother. He’s not playing the game she wants him to play, so she gets upset. She looks around, and finds a toy that he’s not playing with but that she knows he likes. Even though she never plays with this toy normally, she picks it up, sits down next to him, and starts loudly playing with the toy. Her brother gets really upset, but she tells him that he left his toy in the living room (didn’t clean up) and so she’s allowed to use it (which is our house rule). She then offers to give him the toy back if he plays her game with her rules.

    As a parent, what should I do?

  15. Gravatar of Rob Rob
    10. April 2010 at 07:06

    Professor Sumner,

    I agree with your interpretation of the Coase theorem. However, it may be the case that firms may “solve” the externality problem by forcing workers to smoke outside, say in the company’s parking lot. Then the externality would cease to be internalized. Should, government action be called for? Depends who I read last, Samuelson or Tullock.

  16. Gravatar of rob rob
    10. April 2010 at 07:19

    You said: “rob, I have to say that your argument is one of the more far-fetched that I have ever read. We all wanted to behave a certain way, there was nothing stopping us, but we just couldn’t do it until the government forced us to. It might be true, but is that a plausible argument for public policies in a world of great uncertainty? I don’t think so.”

    No, I don’t believe it is a good argument for public policies. I’m not making that argument. I’m just making the observation that in this unique case the government happened to get it right, probably by sheer luck. I believe the government is more likely to get things wrong. But this specific example is a bad example of the government getting it wrong — because they didn’t. Just because the government is more likely to get it wrong doesn’t automatically mean they always get it wrong.

    Liberals often make the mistake of pointing to specific examples like this when the government gets it right and extrapolate it into a belief that the government is more likely to do good than harm. Perhaps that falls into TC’s theme about making the mistake of thinking too much in terms of narrative. The narrative version of stringing together different cases when the government gets things right can yield a persuasive argument that the government often gets things right.

    Your point about people being more sensitive about smoke these days can be turned on its head. Smoke doesn’t bother me, indeed I often hang out at illegal bars that allow smoking. But I have also learned to appreciate how much better it feels to wake up the next morning and to be able to smell other things besides smoke. If one is desensitized to smoke, they are often desensitized to other smells too. But if all you know is a world of smoke, you don’t realize this. Yes, it took a smoking ban for a lot of people to realize things could be better — because they had never experienced otherwise (in the case of bars).

  17. Gravatar of DanC DanC
    10. April 2010 at 07:28

    StatsGuy

    Not sure of your point. Coase argues that private contracts are usually preferred to government regulation unless transactions costs are too high for mutually beneficial private contracts to occur.

    That does not imply that government is low cost or easy. Just that government regulation may, sometimes, be an alternative. But high political transactions costs can also block mutually beneficial contracts. Indeed the political costs of executing a mutually beneficial contract are often higher then the private costs. If the social costs of the contract are too high, I think, Coase would just advise that the contract should not occur. A contract that might be mutually beneficial is not completed due to high transaction costs.

    Public choice theory is often about how actors use the political process to distort or bend regulations (regulatory capture) or that regulations often impose costs on third parties. Taken together we see that government regulations often impose hidden costs and unanticipated changes. These costs may encourage voters to block mutually beneficial private contracts or encourage mutually beneficial private contracts that impose costs on third parties.

    In the smoking ban example, given litigation danger to private contracts (high transaction costs), government regulation is a low cost alternative to force employers to compete based on other factors. A nudge toward a healthier workplace.

    However, in general, I think libertarian paternalism is just an excuse for unwanted intervention. I also think that many of the behavioral economists make large claims from small studies – lab experiments that fail in a real market. Not to mention that markets are far more dynamic and responsive then any nudges that government regulations can create.

  18. Gravatar of rob rob
    10. April 2010 at 07:40

    I don’t want it to sound like I think the smoking ban was a good idea because I personally think it was. I don’t personally care much one way or the other. But upon observing feedback from others, smokers and non-smokers, I’d bet if you polled bar patrons (those who were bar patrons before the ban, to be clear), you’d find that the majority do support the ban now, even though many of them did not before.

    Though perhaps it is a case of ownership bias.

  19. Gravatar of StatsGuy StatsGuy
    10. April 2010 at 08:14

    @DanC –

    I agree with the Coase Theorem – it’s a _theorem_. How do you ‘disagree’ with it? Coase was very careful to base his conclusions about real-world efficacy of regulation on empirical studies.

    “That does not imply that government is low cost or easy.”

    Who said it does? Certainly not me. But the Coase Theorem does mean (not merely imply) that if transaction costs throughout society are zero, then public regulation vs. private outcome yields the same outcome. Does it not? That’s why some people think political contracting is a good thing. Other people call it “bribery”.

    You note that transaction costs for regulation are usually higher than for public action/regulation (or the benefit of the contract). I agree. Usually.

    Re Public Choice – The typical Public Choice outcome is that regulation benefits the regulated, which prompts the question of whether the smoking ban benefits the bars (or the bars’ competitors). If not, then this would be exception to the typical Public Choice prediction. Hmm…. One would think this is related to the Coase Theorem in some way.

  20. Gravatar of DanC DanC
    10. April 2010 at 09:37

    It has been some years since I talked with Professor Coase, but with a desire for efficiency, then private contracting is preferred. If private contracting can be done without cost, then the legal “right” is unimportant.

    For Coase the government is a poor moderator of contracts. The government can impose taxes or regulations to encourage an outcome but they, in practice, can have no idea what the efficient outcome will be or should be.

    So public regulation and private contracts can create very different outcomes even with zero transaction costs.

    For example, the government can impose a tax on polluting firms. The firms will adjust to this tax for some optimal outcome, given the tax. But that does not mean that private parties left to a private contract wouldn’t have created a more efficient outcome absent the tax. Government regulation or taxes are a blunt tool that almost never create an efficient outcome. More often then not government regulators have no interest in an efficient outcome.

    With regard to Public Choice the regulated will often “capture” the regulating force. That does not mean that they will be better then they were pre-regulation. Capturing the regulators is not a costless exercise, I would just view it as an additional tax. However if the regulated were able to create monopoly power, for example, that tax is then passed to society.

    In the bars example, the owners are poor protectors of the health of the employees. The employer compensates the workers enough to get workers to work in smoke filled rooms. Smokers create a negative externality on the workers, for which the employer compensates the employees. Some employers would like to switch to a smoke free bar, to lower medical costs for employees. However competitive pressures prevent them from doing this. (They lose customers and their workforce isn’t cheaper.)

    Government comes in and imposes new regulations. Employees get a better workplace. Smokers are paying the new tax. Employers may have removed an unfunded liability from their balance sheets (future health expenses.) Some bars have seen a drop in revenue.

    Is this outcome efficient? Probably not. Left alone this would not have been the outcome. Did the government know the cost of an externalities that were present? No. They just imposed a tax on smokers because they taught it was more equitable to stick smokers with a tax on their negative externality.

    The prior state was consistent with Coase, the latter with Pigou

  21. Gravatar of mbk mbk
    10. April 2010 at 10:27

    DanC beat me to the broad idea but here we go from my perspective: Frankly this discussion is leaving me scratching my head. If you assume that government regulation can (sometimes or always) achieve an optimal outcome as efficiently as private markets would, then you must assume that government (and the rest of us) already knows ex ante which outcome is optimal.

    But conventionally, efficient market hypothesis anyone, don’t we hold that we have markets precisely for the discovery of optimal trade-offs in society and therefore, for the efficient coordination of current and future resource use? In other words, for optimizing utility in society?

    I’m not even going into the possibility that different societal arrangements could result in the same mythical total utility, or whether anyone can actually quantify utility, whether total utility should matter more than utility distributions (a small number destroyed for the sake of the happiness of a large number) or whether discontinuous variables such as life and liberty can even be quantified and therefore bargained with meaningfully etc.

  22. Gravatar of rob rob
    10. April 2010 at 11:19

    I’m going to stick with the theory that ownership bias explains why the majority of bar patrons were against the ban before the ban but for the ban after the ban. If so, a few questions:

    1. Why was permitted smoking the default position for bars? My guess is that it is only a coincidence of history: people smoked for hundreds of years before they knew it was bad their health.

    2. Imagine an alternate history in which everything else is held constant but smoking was invented only yesterday, and today a television commercial came out with Tiger Woods proudly smoking a cigarette with a caption beneath reading: “Where there’s smoke, there’s fire.” Would the default position of bars tomorrow be to permit or forbid this new thing called smoking?

    3. Before the smoking ban in bars, it was impossible to find a bar anywhere that forbid smoking. After the ban, we have places that abide by the law and smoker-speakeasies that ignore the law. So the ban has created more options for consumers, whereas the default position seemed more like a one-size-fits-all situation due to cultural inertia.

  23. Gravatar of Jason Jason
    10. April 2010 at 11:33

    Not to provide anecdotes as data, but while smoking does have its chemical pleasures and health problems, I think it is far more complicated than that. Some thoughts:

    1. Social signalling. In England and Ireland there were stories of people taking up smoking after the bans — so that they could hang out with the smokers outside. I’ve personally seen the exact same thing in Seattle.

    2. Territory. In all the bars I’ve been to across the country and in others that have smoking bans there are two kinds: bars that have a large portion of their clientele outside smoking and those that have almost zero. It’s almost like the smokers out front act as bouncers to keep “sensitive” people away.

    3. There was a sudden rash of smoking bans that after a wave of urban renewal — rich people don’t smoke and can displace the poorer residents/hipster kids from their previously dilapidated, but now trendy eateries.

    I think smoking bans may function more as “aggressive panhandling” laws than as any public health initiative (not that I don’t think smoking is bad for people). But it took awhile before smoking went from majority/dominant culture signalling in a 1930s urban-rural America to minority/subculture signalling in a post-1980s urban-suburban-rural America and the bans didn’t spring up until the suburban culture started to come back to reclaim urban areas.

    Of all of those cultures, suburban Americans are the most paternalistic. Gen-Xers probably seem ridiculously sensitive because they are signalling their suburban upbringing. I can assure you that the segment of that generation I hang out with smokes a lot.

  24. Gravatar of rob rob
    10. April 2010 at 11:49

    By ownership bias in this case I mean: Before the ban people valued the freedom they had to smoke in a bar or to enter a smoke-filled bar. After the ban, people value the freedom they have to enter a smoke-free bar. In either case, people value the freedom they have over the freedom they don’t. I don’t believe this is a novel idea.

  25. Gravatar of Doc Merlin Doc Merlin
    10. April 2010 at 12:08

    @rob:
    Never thought of it that way before.

  26. Gravatar of StatsGuy StatsGuy
    10. April 2010 at 12:43

    DanC:

    “So public regulation and private contracts can create very different outcomes even with zero transaction costs.”

    “Government regulation or taxes are a blunt tool that almost never create an efficient outcome.”

    I’m not sure why we’re arguing this – so let me offer the following. IF transaction costs for government were free, THEN government regulation would not be a blunt tool.

    In practice, they are not – but that has nothing to do with the theorem, that has everything to do with empirical reality.

    Coase covered this issue in his discussion of horizontal vs. vertical contracting (markets vs. firms). If transaction costs are free, choice of the organization of production does not matter. Think of government like a really big firm. Regulation is only blunt because transaction costs are not free…

    I think a lot of people have difficulty visualizing a world without transaction costs – it’s like a world without friction.

    Rob – in addition to ownership bias, also consider rational updating and risk aversion. Bar visitors know what the bar is like today. They don’t know what it’s like tomorrow (they may even have misinformation). Yet even without misinformation, uncertainty and risk aversion can cause them to prefer no ban. Once that uncertainty is resolved, the cost of risk goes away – and they may even gain new information that conflicts with prior expectations.

  27. Gravatar of DanC DanC
    10. April 2010 at 14:07

    Stats Guy

    You misunderstand Coase.

    For example
    “The ultimate thesis is that law and regulation are not as important or effective at helping people as lawyers and government planners believe (Coase 1960, p. V, 9). Coase and others like him wanted a change of approach, to put the burden of proof for positive effects on a government that was intervening in the market, by analysing the costs of action (Coase 1960, p. VIII, 23).”

    “Furthermore we have to take into account the costs involved in operating the various social arrangements (whether it be the working of a market or of a government department), as well as the costs involved in moving to a new system. In devising and choosing between social arrangements we should have regard for the total effect. This, above all, is the change in approach which I, am advocating.”

    “Analysis in terms of divergences between private and social products concentrates attention on particular deficiencies in the system and tends to nourish the belief that any measure which will remove the deficiency is necessarily desirable. It diverts attention from those other changes in the system which are inevitably associated with the corrective measure, changes which may well produce more harm than the original deficiency.”

    Coase is arguing against activist government action not saying that they are neutral in some ideal case – that never exist.

    “The government is, in a sense, a super-firm (but of a very special kind) since it is able to influence the use of factors of production by administrative decision. But the ordinary firm is subject to checks in its operations because of the competition of other firms, which might administer the same activities at lower cost and also because there is always the alternative of market transactions as against organization within the firm if the administrative costs become too great. The government is able, if it wishes, to avoid the market altogether, which a firm can never do. The firm has to make market agreements with the owners of the factors of production that it uses. just as the government can conscript or seize property, so it can decree that factors of production should only be used in such-and-such a way. Such authoritarian methods save a lot of trouble (for those doing the organizing). Furthermore, the government has at its disposal the police and the other law enforcement agencies to make sure that its regulations are carried out.

    It is clear that the government has powers which might enable it to get some things done at a lower cost than could a private organization (or at any rate one without special governmental powers). But the governmental administrative machine is not itself costless. It can, in fact, on occasion be extremely costly. Furthermore, there is no reason to suppose that the restrictive and zoning regulations, made by a fallible administration subject to political pressures and operating without any competitive check, will necessarily always be those which increase the efficiency with which the economic system operates. Furthermore, such general regulations which must apply to a wide variety of cases will be enforced in some cases in which they are clearly inappropriate. From these considerations it follows that direct governmental regulation will not necessarily give better results than leaving the problem to be solved by the market or the firm. But equally there is no reason why, on occasion, such governmental administrative regulation should not lead to an improvement in economic efficiency. This would seem particularly likely when, as is normally the case with the smoke nuisance, a large number of people are involved and in which therefore the costs of handling the problem through the market or the firm may be high.

    There is, of course, a further alternative, which is to do nothing about the problem at all. And given that the costs involved in solving the problem by regulations issued by the governmental administrative machine will often be heavy (particularly if the costs are interpreted to include all the consequences which follow from the government engaging in this kind of activity), it will no doubt be commonly the case that the gain which would come from regulating the actions which give rise to the harmful effects will be less than the costs involved in government regulation.

    Please read Coase

  28. Gravatar of StatsGuy StatsGuy
    10. April 2010 at 16:13

    DanC:

    “Please read Coase” I have, thank you, several times.

    I’ve basically argued that the Coase Theorem indicates that with zero transaction costs, legal regulation would achieve the same thing as private negotiation. However, empirically, transaction costs are high for government, so regulation is often inefficient. Since you demand citation, here’s Coase saying the same thing (again from his Nobel lecture)…

    “While we can imagine in the hypothetical world of zero transaction costs that the parties to an exchange would negotiate to change any provision of the law which prevents them from taking whatever steps are required to increase the value of production, in the real world of positive transaction costs such a procedure would be extremely costly, and would make unprofitable, even where it was allowed, a great deal of such contracting around the law. ”

    Rather than addressing this, you keep repeating over and over that Coase did not like regulation as a matter of empirical observation – something I absolutely agree with. I am totally at a loss what we’re arguing about.

    Since you already cited Coase discussing how government is a special case of an uber-firm, I don’t need to discuss that either.

  29. Gravatar of StatsGuy StatsGuy
    10. April 2010 at 16:19

    DanC:

    “Coase is arguing against activist government action not saying that they are neutral in some ideal case – that never exist.”

    OK, I guess we do disagree – see the Coase quote above. Coase explicitly acknowledges that activist government and private contracting are neutral in an ideal case, but that the ideal case never exists. It’s on the basis of this non-existence of the ideal case, and the empirical fact that the relative costs of transacting by changing the law are usually higher than for contracting in the market, that Coase argues against Pigou.

    But many people – you included, I suppose – seem to take the Coase Theorem to state (a priori) that government regulation is less efficient than private contracting. The Coase Theorem says nothing of the sort, a priori.

  30. Gravatar of TGGP TGGP
    10. April 2010 at 19:22

    Frank Knight made the point against Pigou a while before Coase. Private ownership resolves the dilemma.

  31. Gravatar of DanC DanC
    11. April 2010 at 01:50

    Stats Guy

    Ok , since you want to look at the Noble speech, let’s look at it

    ” I thought that I was exposing the weaknesses of Pigou’s analysis of the divergence between private and social products, an analysis generally accepted by economists, and that was all.”

    “Pigou’s conclusion and that of most economists using standard economic theory was, and perhaps still is, that some kind of government action (usually the imposition of taxes) was required to restrain those whose actions had harmful effects on others, often termed negative externalities. What I showed in that article, as I thought, was that in a regime of zero transaction costs, an assumption of standard economic theory, negotiations between the parties would lead to those arrangements being made which would maximise wealth and this irrespective of the initial assignment of rights. This is the infamous Coase Theorem, named and formulated by Stigler, although it is based on work of mine. Stigler argues that the Coase Theorem follows from the standard assumptions of economic theory. Its logic cannot be questioned, only its domain7. I do not disagree with Stigler. However, I tend to regard the Coase Theorem as a stepping stone on the way to an analysis of an economy with positive transaction costs. The significance to me of the Coase Theorem is that it undermines the Pigovian system. Since standard economic theory assumes transaction costs to be zero, the Coase Theorem demonstrates that the Pigovian solutions are unnecessary in these circumstances. Of course, it does not imply, when transaction costs are positive, that government actions (such as government operation, regulation or taxation, including subsidies) could not produce a better result than relying on negotiations between individuals in the market. Whether this would be so could be discovered not by studying imaginary governments but what real governments actually do. My conclusion; let us study the world of positive transaction costs.”

    So Coase was never as simple as you claim. Indeed your gross over simplification of Coase distorts what he wrote.

    Coase wanted to argue against Pigou so he used a simple model without transaction costs to show why he thought Pigou was wrong. And Coase only wrote about the initial assignment of property rights in this simplified model to argue against Pigou. He wasn’t, at that point, really pushing a complex notion of government regulation. It is a very simple model to make a simple insightful comment – Pigou is wrong.

    The zero transaction costs is an artificial construct to make a point.

    However I do agree with your later comment that

    “I’ve basically argued that the Coase Theorem indicates that with zero transaction costs, legal regulation would achieve the same thing as private negotiation. However, empirically, transaction costs are high for government, so regulation is often inefficient. ”

    And I do admit that I pushed Coase a bit far. However Coase does view government regulation as a blunt tool that has a hard time achieving efficient outcomes, not that it never can. For example, government may lower the transactions costs to aid a beneficial transaction amongst a large group of people. I admit that I think Coase is overly optimistic when he thinks that courts or governments will seek an efficient outcome when their goal is more often some notion of an equitable outcome. And Coase is aware of the problem but is offering an alternative path.

    Which returns us to the smoking in bars. The solution we have is consistent with Pigou. The government picks a side to blame for a negative externality, imposes a tax, and has the involved parties make the best of it.

    The Coase solution would have been private negotiation of the involved parties. Or the government would have sought out an efficient outcome. It didn’t. It picked a villain and imposed a tax.

    So the smoking ban as written does violate a Coase solution, the search for an efficient outcome, and returns us to a Pigou style solution.

  32. Gravatar of DanC DanC
    11. April 2010 at 02:08

    BTW

    I do disagree with your assertion that Coase argued that “legal regulation” with zero transaction costs leads to the same solution.

    The legal regulation in the simple Coase model was the initial assignment of property rights. The minute anything else is added to the mix, for example hiring a lawyer, the simple Coase model begins to breakdown. So beyond the simple assignment of property rights you don’t really have “legal regulation” in the model.

    For me the Coase Theorem is about finding the efficient outcome in contrast to deciding that x produces a negative externality on others and must be curtailed.

  33. Gravatar of StatsGuy StatsGuy
    11. April 2010 at 06:05

    DanC:

    So I agree with you 100% about the Coase Solution.

    And the Coase ‘model’ does break down if you hire a lawyer to change the assignment of property rights, _because_ hiring a lawyer is a non-zero transaction cost. The moment you need to hire a lawyer, you’ve broken the _assumptions_ in the Coase Theorem.

    But the Coase _Theorem_ doesn’t doesn’t break – not unless we’ve invented a new and better form of mathematics.

    For me, the Coase Theorem (which was framed by Stigler, if I recall?) is actually a straw man that Coase sets up. Coase was studying real world existence of Firms, and looking at all these contemporary models that described the beauty of pure fluid zero-transaction cost markets, and asking the question – if markets are SOOO perfect, why do firms exist? Competition and evolution should have wiped out the existence of firms, right? So there must be some reason that firms exist, and it must be that at some scale and under certain conditions, markets break down.

    In the context of this, the Coase Theorem emerged, and Coase basically says that all this mathematical modeling of ‘ideal markets’ is crud because if you use those same assumptions you can prove all sorts of crazy stuff. The Coase Theorem thus becomes this hugely important “null case” that serves as a backdrop for the entire new institutionalist field (vis a vis Oliver Williamson, etc.).

    Having said all of this, when Coase was asked whether any regulation is useful, his response was not “read my theorem”, it was something to the effect that he had read several hundred studies of specific regulations and couldn’t recall one where the regulation had improved the outcome. That’s the response of an empiricist.

  34. Gravatar of DanC DanC
    11. April 2010 at 07:04

    Coase, in the beginning, was a socialist. He wondered why government control of the economy couldn’t work like a firm.

    BTW Stigler was, in the beginning, in favor of government control to fix market failures.

    Coase disagreed with Pigou and used a simple model to illustrate the point. The lesson of the model is that private contracts would lead to the most efficient outcome without any government interference. (Price discovery, etc) You don’t always need the government to moderate or pick sides.

    What you are claiming as the Coase Theorem is such a gross simplification of Coase that it turns him into a distorted straw man.

    Coase was not against mathematical modeling. Just that the goal is economic efficiency and finding information superior to the information that a free market supplies is difficult. Still he encouraged research.

    Which brings me back to your comment that started this

    “Your argument is that arguing for a government smoking ban is a “violation of the Coase Theorem”. I disagree.

    The way the government instituted the smoking ban is a violation of Coase and is a throw back to Pigou.

    It is also why I dislike Libertarian Paternalism. As I said above

    However, in general, I think libertarian paternalism is just an excuse for unwanted intervention. I also think that many of the behavioral economists make large claims from small studies – lab experiments that fail in a real market. Not to mention that markets are far more dynamic and responsive then any nudges that government regulations can create.

    Government nudges are like robots in science fiction: they either continue to mindlessly perform functions long after they serve any real purpose or they take on a life of their own to destructive end.

    Libertarian Paternalism is often about equity and not efficiency in any case.

    In any case, I think nudges, more often then not, violate the path Coase suggested

  35. Gravatar of scott sumner scott sumner
    11. April 2010 at 07:20

    DanC; You said;

    “I suppose Coase could argue that workers could demand extra income to compensate for working in a smoke filled bar or attract workers who don’t care. But such arguments would lead to the end of OSHA.”

    That would be a dream come true. People who have studied OSHA find no impact on worker health and safety, and yet it imposes big costs on society.

    BTW, compensating wage differentials isn’t just something Coase might say, it is the standard model of labor economics, taught in every labor economics textbook.

    You said;

    “It is just hard to imagine that society would permit a binding contract between employers and employees to enter a contract where the employee agrees to risk future health in return for employment.”

    Coal miners?

    Mark Sadowski, If you take away the property rights to the air within a building, then you have the problem of the commons, and will get a much less efficient solution.

    Statsguy; You said:

    “In essence, your argument boils down to this: Transaction/contracting costs are RELATIVELY lower for the private firms than they are for the regulator in this case – even taking into account externalities and collective action costs etc.”

    No that’s not my argument. I think the transactions costs are lower for the government. They can ban smoking everywhere at the stroke of a pen. But I agree that the Coase Theorem is not the right way to discuss this problem. I never should have mentioned the Coase Theorem. The simplest answer is that there is no externality because the air inside restaurants and office is privately owned by a single entity. No externality implies no externality justification for taxes or regulations. So Thaler is simply wrong. There may be a justification for regulating smokers in private buildings, but it’s not a second hand smoke externality argument.

    Statsguy, I agree that if there were an externality, the problem of rent seeking would be very relevant. Of course there is not externality here, so it is not relevant to this problem.

    I’ll do the rest later.

  36. Gravatar of Jim Vernon Jim Vernon
    11. April 2010 at 08:12

    I still think you’re overlooking the symmetry of Coase’s logic.

    Yes, non-smokers could pay smokers not to smoke. Smokers could just as easily pay non-smokers to “allow” them to smoke. Which transaction takes place depends on property rights. Who owns the relevant rights to the clean air? Do smokers own the right to push smoke into it? Do non-smokers own the right to prevent smoke from being pushed into it?

    This is all the more difficult because we have a true common resource, which is non-excludable and apparently rival. (Who knew?) Solving Coase’s railroad dilemma seems a bit easier, since we can fence off the land around the tracks. It seems to me that we’re trying to fence off the air around the smokers.

    Best regards,
    Jim

  37. Gravatar of scott sumner scott sumner
    11. April 2010 at 08:23

    Statsguy#3, Congratulate your daughter for rent seeking.

    Rob, Smoking outside doesn’t cause any external harm either.

    rob. I still think they got it wrong. The cost to smokers greatly exceeds the trivial gains to non-smokers. I still think that if there had been a significant demand for smoke-free bars, someone would have provided it.

    mbk, I agree. Indeed the argument for smoking bans and OSHA is pretty much the same as the argument for government ownership of companies. It’s the argument that companies might make mistakes and that the government might avoid making those mistakes. Yes they might do better, but how likely is it?

    Jason. Those are good observations.

    DanC and Statsguy, we need to get Coase over here to settle the debate. Like that famous scene in the Woody Allen movie involving Marshall McLuhan.

    TGGP, Thanks for that link.

  38. Gravatar of scott sumner scott sumner
    11. April 2010 at 08:25

    Jim Vernon, The air in private office buildings is private property, so there is no externality problem. If someone can find an academic article addressing and refuting the David Henderson piece I linked to, I’d love to see it. Otherwise I’ll assume Henderson’s position is the standard view of economists.

  39. Gravatar of DanC DanC
    11. April 2010 at 09:56

    Papers against Libertarian Paternalism

    http://papers.nber.org/papers/w15139#fromrss
    http://online.wsj.com/article/SB117977357721809835.html
    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=473444

  40. Gravatar of Mike Sandifer Mike Sandifer
    11. April 2010 at 12:55

    Scott,

    I was not familiar with the Coase Theorem until today, which means I’m still not familiar with it, but I do have some questions and comments while I try to explore its implications.

    1.) On Rob’s first point above about favoring smoking bans once instituted, I think one problem with the blind economic approach is that it assumes much too much about the efficiency with which companies explore alternative business practices. The fact is, there are finite resources for considering alternatives and cognitive biases of managers, such as functional fixedness and groupthink, play roles. In fact, one can never explore most permutations in most cases. And even economists consider risk aversion.

    In your response to rob, you seem to discount the idea that the market can fail to recognize significant existing demand. Well, the market always fails to realize existing demand, until a new product, process, or service is introduced. Sometimes the thing demanded represents a revolution, but at others, just a patching together of different elements of long existing desired goods or services.

    Take Snuggies, for example. Someone decided to combine a blanket with a robe, and many millions have been sold. This is a simple idea, but how long did it take the market to recognize the demand? How long have blankets and robes existed, or even televisions to sit down and watch in comfort?

    So, can the government do a better job at anticipating demand than the market? Well, the anecdotes offered above about changes in support from against to for smoking bans, and before and after polls, seem to indicate that either people didn’t know what they wanted in the first place, and/or that demand is sometimes easily shiftable.

    In any case, it’s hard to argue that smoking indoors will make anyone healthier. And perhaps the above-mentioned shifts in demand indicate at least a perceived positive externality, which can sometimes produce positive externalties in self-fulfilling ways(even patrons who smoke preferring non-smoking establishments, as a shift in moral attitudes). And of course, since the government heavily subsidizes health care, there are at least some practical externalities for indoor smoking.

    2.) You mention above that smoking outdoors carries no negative externalities, but this is not necessarily true. Cigarette smoke contains carbon monoxide, which is a greenhouse gas, but more importantly has greater indirect effects that worsen global warming. If enough people smoke outside often enough, perhaps this could begin to have a measurable effect?

    So, the bottom line for me is, that given the required conditions for the Coase Theorem to hold true, it represents an illustrative limiting case at best, with little relevance to actual policy making, especially given normative concerns, which you mention.

  41. Gravatar of TGGP TGGP
    11. April 2010 at 18:07

    Speaking of Coase, Nick Szabo argues that his theorem has been greatly overextended, particularly by anarcho-capitalists. I’d like to get him arguing with David Friedman (his chief target), but neither seems too interested in that.

  42. Gravatar of Zepplin Zepplin
    12. April 2010 at 00:22

    Totally agree.

    Same for work safety mandates, health hazard regulations and fire hazard regulations which goes beyond reducing the chance of fire spreading to non-owned neighbors (I’m looking at you, stupid multiple exits rule)

    The market will efficiently decide by balancing costs against employees and customers choosing safer offices / coal mines / night clubs and also insurance and PR costs. One size fits all regulations are clearly inefficient.

  43. Gravatar of DanC DanC
    12. April 2010 at 03:01

    To Scott
    For clarification.

    I didn’t mean to imply that Coase was the only person who talks of compensating wages. I just was thinking of how Coase might approach this problem.

    With regard to the mine workers, my point is that our legal system is too often a casino. And I will bet you that the recent coal mining tragedy in West Virginia will lead to lawsuits over issues of negligence.

    And on the issue of tobacco how do lifetime smokers win a lawsuit that claims they were ignorant of the risks of smoking. I find it easy to imagine an employer being asked in a disposition if they knew that second hand smoke was dangerous for their employees and why didn’t they take steps to correct the situation. I think such potential litigation will occur and is not costless.

    The firms at greatest risk for these lawsuits are deep pocketed chains who may have a desire to impose smoking bans but fear losing customers to other firms that allow smoking but that are too small to be a target of lawsuits or view the risks of such lawsuits as small.

    For example this from the Detroit Free Press:

    “Michigan’s smoking ban might seem unfair to smokers or another blow to a wobbling restaurant industry. But others see an opportunity to do what they’ve wanted to do for years: toss out their ashtrays. More than 150 Michigan eateries have opted to go smoke-free since Dec. 10, when the Legislature voted for the ban. Some had wanted to do it for years, but feared losing customers to other restaurants. The looming May 1 ban snuffs out any chance of that. Light up in any restaurant or bar and most workplaces, and you can be fined $100 for the first offense, and up to $500 for repeated violations. All 23 National Coney Island sites in metro Detroit went smoke-free in January. Though customer response was overwhelmingly positive to a test ban last year, Tom Giftos, president and CEO of the chain his father started in 1965, said the economy had him worried. “Nobody wants to turn away business in any shape or form these days,” he said. The state ban “gave us a bit of justification.” ”

    And this from some study in Colorado
    1. A mixed effect is observed for restaurant owners. Based on data that I have researched and analyzed I suspect (but cannot and do not attribute the view to Ms. Castellanos) that negative impacts are felt most severely by small, independent neighborhood restaurants, whereas large franchise chain venues have experience less severe or positive impact.

    2. The most widespread negative economic impact is felt by small, independent neighborhood bars and taverns. For example, the Coalition for Equal Rights counts 50 bars and taverns that have gone out of business since the smoking ban took effect and declines on revenue for some bars and taverns are 20 percent or more. In addition, employee tips are significantly reduced.

    3. The economic impact also spreads over other venues such as bingo halls. Which raises the point that adverse economic impact is most severely felt in establishments where the activities are most directly associated with smoking, such as dinking alcoholic beverages, gaming, playing pool, etc.

    4. The negative economic impact of Colorado’s smoking ban is reduced for hospitality establishments that are able to provide outdoor smoking areas.

    5. The negative economic reported is measurably due to and causally associated with the statewide smoking ban and not other external influencing factors.’

    Still is there a negative externality. No. What you do have, I think, is a growing movement against smoking and an increased risk of successful litigation against employers who allow a smoke filled workplace. Deep pocketed chains are at greater risks and they don’t really fight that hard against smoking bans.

  44. Gravatar of scott sumner scott sumner
    12. April 2010 at 05:21

    Mike Sandifer, There are many incorrect assertions in your comment, let’s start with the last one:

    “So, the bottom line for me is, that given the required conditions for the Coase Theorem to hold true, it represents an illustrative limiting case at best, with little relevance to actual policy making, especially given normative concerns, which you mention.”

    There are no conditions required for the Coase Theoren to hold true. The Coase Theoren says nothing about whether government regulation is useful or not. It offers a way to think about problems, not a solution. It is true whether transactions costs are high or low. And even without the Coase Theorem there is no good reason to regulate workplace smoking, as there are no important externalities.

    I don’t know which “normative concerns” I mentioned. But the Coase Theorem is generally recognized as being an integral part of public policy formation. It underlies much of the field of law and economics.

    2. Your statement about outdoor smoking is completely false. The damage from cigarettes smoked outdoors is trivial. The optimal Pigovian tax on cigarettes from things like the greenhouse effect would be less than once cent per pack. Even indoor smoking is debatable. Many studies have failed to find any statistically significant effect from second-hand smoke. I’m pretty sure that there is a health effect, but it is probably rather small. The truth is that we just don’t know.

    3. Your statement about government health care costs being an external cost is misleading. There are some costs incurred by the government, but they are surprising modest (because smokers live much shorter lives.) So if looking at things from a public finance perspective, the optimal tax should be dramatically lower than the current cigarette tax.
    4. Yes, the market can occasionally make mistakes about what customers want. But government regulators are much less likely to reflect consumer interests, as they have very little inventive to get things right. If businesses don’t provide a good produce they lose customers. I still don’t think anti-smoking regulations helped, for all of the reasons I have already given. The fact that lots of customers now say they like the regulations probably just reflects status quo bias.

    In any case, even if you and rob are correct, now that non-smoking restaurants are in place, the regulation should obviously be repealed. If you are right, then restaurants will remain non-smoking so as not to antagonize their customers. That is clearly the logic of your argument.

    Thanks TGGP.

    DanC, I certainly don’t like our legal system, and smoking lawsuits are just one of many examples that I don’t think should be allowed. Of course the people that suffer from smoking lawsuits are mostly the smokers themselves, as the tobacco companies pass the costs on in higher prices.

    Suppose 80% of a city is non-smokers (I think this is about right. Suppose half of all non-smokers prefer restaurants with no second-hand smoke. In that case a restaurant that went smoke-free would HAVE 40% OF ALL CUSTOMERS TO THEMSELVES. So if there really was a strong demand for smoke-free restaurants then I don’t doubt that the market would meet that demand.

  45. Gravatar of DanC DanC
    12. April 2010 at 06:08

    I’m sure that the market can meet the need for smoke free restaurants.

    My point is that many restaurants might want to ban smoking to avoid lawsuit litigation danger from employees but fear that the loss in business could be greater then the gains from avoiding lawsuits.

    However if they are able to ban smoking at potential competitors then they don’t have to worry about the lost customers and still gain the reduction in liability.

    Since large deep pocketed firms are the ones most likely to be sued, they are the ones most likely to agree to such bans.

    For example, sometimes a large firm may welcome regulation, to some degree, because it serves as a barrier to entry from other firms. Large firms, or restaurant chains, may have more incentive to limit litigation dangers, in the smoking example, by transferring some of the cost to smaller competitors.

    Say the chain has $100 in litigation risk from employees – from allowing smoking by patrons. (Ignore for a minute then many in the restaurant business are smokers.) If they ban smoking and they lose more then $100 in business they are worse off.

    However if they can get the government to ban smoking at all restaurants and pubs they lose only $75 in business but still save $100 in litigation risks. The government has made the owner better off. And the smokers are in a huddle outside.

    Of course the small bar wasn’t worried about being a target of lawsuits. He placed his litigation risk at $1 and has seen his business drop by $75. He is much worse off.

    So the large chains have lowered their litigation risks and improved their bottom line, with the help of the government, and transferred some of their litigation costs unto their smaller competitors.

    The transfer of those litigation risks to smaller bars may lower their profits enough to force them out of business. Business which the chain restaurants pick-up.

    Society is left with fewer choices in bars, fewer options for smokers.

    On balance is this a good thing?

  46. Gravatar of DanC DanC
    12. April 2010 at 06:22

    BTW

    Assume that their are 200 bars, evenly divided between chain and independent.

    In my example, after the ban, the chains gain $250 while the independents lose $740.

    Society is made poorer (even ignoring the troubles for smokers).

    However if you really hate smokers the chain restaurants may help you ban smoking in all restaurants.

  47. Gravatar of Mike Sandifer Mike Sandifer
    12. April 2010 at 09:04

    Scott,

    First I apologize about the reference to a recognition of normative concerns from you. Upon second reading, I found no such statements.

    But, to continue my reply, what of the case in which an employer bans smoking in the workplace for moral reasons, despite the financial, and hence, macroeconomic cost? In the latter case, we’re surrendering GDP growth, wouldn’t you say? If so, can that be considered a negative externality? If so, would you then ban private bans on smoking in the workplace? Correct me if my logic is unsound.

    And with regard to smoking bans, many of them are passed via referendum, so can’t this be considered an expression of market opinion? Would you override the popular will of the voters in these cases?

    And perhaps owners of firms are initially afraid of banning smoking, given perceived competitive disadvantages if done so in isolation. Wasn’t the same true in the case of racial discrimination, which entailed macroeconomic costs, but not unlike those considered in the smoking example above?

    These are the sorts of issues that come to mind regarding Coase. In the real practical world, it doesn’t always matter, as human beings make decisions based on concerns other than economic.

    This is also my problem with libertarianism in general. It seems to assume optimal outcomes, despite some large, obvious examples in which they aren’t realized. More importantly, it’s hard to imagine libertarianism gaining too much ground, as there are always incentives by some to regulate others, especially with regard to wealth redistribution.

    I voted against a smoking ban in my state, but now that I can go to certain places without the smoke in my face and the residue on my clothes, I wouldn’t want to repeal it.

  48. Gravatar of Mike Sandifer Mike Sandifer
    12. April 2010 at 09:17

    I guess I should also link the micro with the macro explicitly. Even if one or a handful of firms banning smoking doesn’t affect overall GDP, what of the case of referenda banning it in all workplaces? Could this be construed to have a macroeconomic cost? Doesn’t the free market fail in this case?

  49. Gravatar of StatsGuy StatsGuy
    12. April 2010 at 12:17

    ssumner:

    I cannot help myself. CapsLock is engaging automatically…

    “There are no conditions required for the Coase Theoren to hold true.”

    FALSE. IT IS A THEOREM. BY DEFINITION, IT HAS CONDITIONS.

    “It offers a way to think about problems, not a solution.”

    TRUE.

    “The Coase Theoren says nothing about whether government regulation is useful or not.”

    TRUE.

    “It is true whether transactions costs are high or low.”

    FALSE. The theorem only applies when transaction costs are ZERO – that is, PERFECT COMPETITION.

    If you don’t believe me, here’s Stigler’s definition, which was the first recorded use of the term “Coase theorem”. (The 1966 Stigler… who despises regulation.)

    “The Coase theorem thus asserts that under perfect competition private and social costs will be equal. It is a more remarkable proposition to us older economists who have believed the opposite for a generation, than it will appear to the younger reader who was never wrong, here.”

    That is Stigler’s Coase Theorem. There is no empirical validation. It’s a THEOREM.

    http://en.wikipedia.org/wiki/Theorem

    “In mathematics, a theorem is a statement which has been proved on the basis of previously established statements, such as other theorems, and previously accepted statements, such as axioms.”

    As Coase later recognized – which I quoted in the Nobel speech above – this theorem has all sorts of crazy implications (like the Coase Conjecture re monpolies & commitment). Among them, with zero transaction costs (both inside and outside firms) and no barriers to contracting, markets and firms (of which govt. can be considered a unique kind) yield identical outcomes. [side note - there are a dozen "theories of the firm", and some consider the firm merely a "bundle of contracts"]

    I am peeling my thumb off of CapsLock.

  50. Gravatar of scott sumner scott sumner
    13. April 2010 at 05:37

    DanC, That is a good point about why big chains might want a blanket ban on smoking.

    Mike, You said;

    “But, to continue my reply, what of the case in which an employer bans smoking in the workplace for moral reasons, despite the financial, and hence, macroeconomic cost? In the latter case, we’re surrendering GDP growth, wouldn’t you say?”

    Measured GDP may fall but utility should rise. If satisfying their moral intuitions is worth sacrificing $X in profits, then perhaps they should do so. I suppose it depends how far one wants to take utilitarianism; I understand that there are cases where this moral criterion would be offensive (say racial discrimination) but I just don’t see smoking bans in the same light.

    You said;

    “And with regard to smoking bans, many of them are passed via referendum, so can’t this be considered an expression of market opinion? Would you override the popular will of the voters in these cases?”

    Two points. Markets are better than referenda because it should be one dollar one vote, not one man one vote. Smokers get much more pleasure from smoking than non-smokers get disutility from second-hand smoke. On the other hand I would not overturn the referendum, because if we are going to have democratic votes then I think we should abide by the results.

    You said;

    “These are the sorts of issues that come to mind regarding Coase. In the real practical world, it doesn’t always matter, as human beings make decisions based on concerns other than economic.”

    All concerns are economic, including feelings and preferences.

    You said;

    “This is also my problem with libertarianism in general. It seems to assume optimal outcomes, despite some large, obvious examples in which they aren’t realized. More importantly, it’s hard to imagine libertarianism gaining too much ground, as there are always incentives by some to regulate others, especially with regard to wealth redistribution.”

    Elsewhere I’ve argued that libertarianism will gain more ground in societies that are more idealistic. BTW, libertarians like me, Friedman and Hayek favor some redistribution.

    Statsguy, OK, but I was thinking of Coase’s version of his ideas, not Stigler’s. I accept your point about Stigler’s definition. But Coase’s ideas certainly apply equally well to a world with or without transaction costs. He said that the proper way to think of his idea was not to focus on the zero transaction cost case, but to think of it as an inquiry into the role of transactions cots. I was reacting to someone arguing that the Coase Theorem doesn’t apply to a world with transactions costs. I took that as implying Coase’s ideas were not relevant to a world of transactions costs, which is false. Coase showed that Pigou’s theory is wrong whether or not there are transactions costs. The entire externality problem must be restudied from a transactions cost perspective. If there are transactions costs then Pigou’s policy recommendations might be correct, but they might not be correct. Pigou had no way of knowing the answer, as he ignored transactions costs.

  51. Gravatar of StatsGuy StatsGuy
    13. April 2010 at 06:17

    - I agree with everything you said in the last comment. I am sorry for being so annoying about this too.

  52. Gravatar of Mike Sandifer Mike Sandifer
    13. April 2010 at 08:56

    Scott,

    I used to be very libertarian and still am relative to most Americans. I favor legalizing all drugs, getting government out of marriage recognition and incentives, ending programs like the Patriot Act and closing Gitmo, etc. But, I do believe the government sometimes knows better and can do better than markets, despite the inefficiencies and warped incentives, even sans externalities.

    Demand just doesn’t work the way many macro simplying assumptions imply. Government can seem to change demand and make better personal decisions for citizens and busines than they can for themselves. I consider racial discrimination. I also think about the country before Teddy Roosevelt and the beginning of the establishment of laws to protect consumers and workers. Yes, The Jungle was a piece of propaganda, but it wasn’t entirely unrealistic.

    I know you’re not an extremist in your views. I only meant to say that purer forms of libertarianism will never be widely accepted, as long as humans are greedy and irrational.

  53. Gravatar of scott sumner scott sumner
    14. April 2010 at 05:26

    Statstguy, No problem, You keep me on my toes.

    Mike. A few comments:

    1. I don’t believe in purer forms of libertarianism.

    2. The country was much poorer back in 1900, which explains why conditions were worse. If we had had a lot of regulation in 1900, you still would have been able to write books like The Jungle. India has always had lots of regulations, and look at health and safety conditions there. So regulation has little or nothing to do with improvements in regulation. Get a graph showing annual worker deaths on the job. Draw in the point where OSHA was founded. You won’t seem any break in the long downward trend.

    3. There is little doubt that the government can sometimes do better than markets, the question is whether regulation can be expected, ex ante, to do better. That is a much more doubtful proposition.

  54. Gravatar of Mike Sandifer Mike Sandifer
    14. April 2010 at 13:30

    Scott,

    At the end of my last post, I wrote about you:

    “I know you’re not an extremist in your views. I only meant to say that purer forms of libertarianism will never be widely accepted, as long as humans are greedy and irrational.”

    So, I did recognize that you’re not an extreme libertarian. Perhaps my statement was unclear.

    Frankly, I agree with you much more often than not, but often lack the sophistication to evaluate your macroeconomic opinions.

    With regard to India though, perhaps this example of hyper-regulation is inappropriate, but you are certainly in a better position to judge that than I am.

  55. Gravatar of rob rob
    14. April 2010 at 13:57

    You said: “In any case, even if you and rob are correct, now that non-smoking restaurants are in place, the regulation should obviously be repealed. If you are right, then restaurants will remain non-smoking so as not to antagonize their customers. That is clearly the logic of your argument.”

    Yes, that is exactly the logic of my argument and I agree entirely. I believe most restaurants and many bars would remain non-smoking now that that is the status quo.

  56. Gravatar of rob rob
    14. April 2010 at 14:02

    So now perhaps we agree, but a few days ago you claimed my argument was “one of the most far-fetched you’d ever read…”

  57. Gravatar of scott sumner scott sumner
    15. April 2010 at 04:55

    Mike, Yes, the India example may be inappropriate. But do you know of any big, low income countries that currently don’t have bad working conditions because of OSHA-type regulations? My argument is that poverty, not lack of regulation, causes bad working conditions.

    I appreciate your comments on my blog.

    rob#1, Great.
    rob#2, That was a different argument. I still think it is unlikely that there was a huge unmet demand for non-smoking restuarants in the old days. But I obviously agree with your view that we can now remove the regulations, it’s just that I still don’t think they were needed in the first place. In any case, it’s a nice compromise suggestion, in the spirit of libertarian paternalism.

  58. Gravatar of rob rob
    15. April 2010 at 21:58

    i don’t think my argument was different, tho perhaps less explicated.

    and for the record, i think i’m about to go to jail for not filing my income tax return for a long time, so not like i am a fan of the government here. yet, despite that i hate the tea baggers.they seem to be whining about all the trivial stuff, as opposed to complaining about the major stuff like why i am about to go to jail for not doing paperwork.

    i have no problem with paying taxes, but i am a criminal for not doing my government homework. something just seems wrong with that.

  59. Gravatar of ssumner ssumner
    16. April 2010 at 05:03

    rob, I completely agree. In Sweden they simply send you the income tax bill–that’s how it should be done here. Individual taxpayers should not have to compute their taxes. Even worse, when I called the IRS to have them explain a tax provision, they said they didn’t know—and yet they expect me to know!

  60. Gravatar of Demsetz’ critique of Coase is lame « Entitled to an Opinion Demsetz’ critique of Coase is lame « Entitled to an Opinion
    17. February 2011 at 23:25

    [...] Ronald Coase‘s paper on externalities and transaction costs. There is an appropriate nod to Frank Knight’s argument against Pigou, but the new critique of Coase is completely unconvincing. At the end he acknowledges the existence [...]

  61. Gravatar of TheMoneyIllusion » Ronald Coase, RIP TheMoneyIllusion » Ronald Coase, RIP
    3. September 2013 at 12:25

    [...] and got a lot of pushback.  Since I’m super busy now, I refer you to the comment section of this post, where all your questions will be [...]

  62. Gravatar of JAS JAS
    3. September 2013 at 16:23

    Thanks for responding even though busy. I read the post and comments. I know you write mainly to engage fellow economists, but well written blogs like yours also provide a great education for the lay reader.

Leave a Reply