Markets Aren’t Perfectly Competitive: The Game Theory of Why a Smoking Ban Might Make Sense
Every Tuesday is a Game Theory article at Mind Your Decisions.
This article will arm you with a powerful critique. In no time you’ll be second-guessing how fiscal conservatives talk about economics.
Here’s one of the most common criticisms of a new law:
“My opponent supports law X. While I agree with the spirit of law X, I do not think the government should get involved. If change is really better for us, the free-market economy will make it happen.”
It sounds like good logic. It appeals to people versed in introductory economics. The subtext is that perfect competition always creates better outcomes than regulation. Sorry to burst their bubble, but this is simply not true.
Perfect competition doesn’t always exist. Imperfect competition is commonly seen and allows for strategic behavior (negotiating job offers, pro sports, health care, elections, the used car market…I can go on and on). In these cases, competition might lead to a sub-optimal outcome. For instance see my article about on how price matching guarantees can help businesses avoid price wars.
Don’t blindly think markets will solve problems; always consider the possibility of strategic interaction.
I’ll illustrate the phenomenon through two examples: restaurant leftovers, and the Illinois smoking ban.
Dining out, taking home
Restaurant meals are so large that I often have leftovers. The first inefficiency is the box size. Restaurants usually use the same boxes for an entire entrée as for the leftover portion. This means I use a box that’s too large. If I ate half my meal, I’m using a box that’s twice as large as needed.
But the story gets worse. Because the boxes are big, they usually won’t fit conveniently in my refrigerator. I have to transfer the food to another container. So here’s the recap: I get a box twice as large as needed, take my food home, transfer the food to my own container, and proceed to throw away the leftover box. The box only serves as a way to transport food from the restaurant to my home. This has happened three times in the last month.
I’m sure the restaurant is not happy about this situation. The leftover box costs them money and they would rather not unnecessarily create waste. If customers ate all of their food, or just brought their own containers, it would be easier for restaurants to manage.
Now, there are many solutions to this inefficiency. The simplest is that I anticipate leftovers and bring my own container. I wouldn’t waste food or material, and the restaurant wouldn’t have to pay for boxes.
But I’m lazy and forgetful. This past week I was in a rush and didn’t bring my own container. It’s because my failure carried no individual penalty—I just asked for a leftover box. I would only have to live with the guilt of being wasteful, which is pretty easy to get over.
Is there a way to fix the system?
I’ve seen some restaurants go as far as charge customers for packing leftovers. Needless to say, this is not well received. Customers find it appalling especially because other places don’t have this policy.
The system is clearly broken—the market for restaurant leftovers is failing. There is simply not a natural market for take-out boxes. Restaurants who try to create a market individually look unpopular. Customers have little reason to take action. It’s a big coordination game. And that’s where government action could help.
Imagine there was a law along the following lines: restaurants must charge customers a 25 cents tax for take-out boxes. It would be an environmental law to offset the cost of waste.
If this law existed, a lot more customers like me would remember to bring their own containers. And just as important, restaurants could charge without feeling their competitors could undercut them. It wouldn’t be a popular bill, and I’m not sure it would ever pass. But it’s a fun thought experiment.
Don’t blow smoke in my face
Illinois went smoke-free in 2008. That means no smoking in indoor workplaces or public places, including bars, clubs and casinos. Smokers are annoyed and some are coping by switching to chewing tobacco. Non-smokers are generally pleased at the refreshingly clean atmosphere of nightlife.
Do I think it is a good law? My gut reaction is yes, but the truth is that I don’t know. The cost/benefit analysis is complicated. It could save the state money by improving public health, but it could also cost the state money if businesses like casinos lose revenue. I have not read enough, and the economics are not simple.
People who think they understand the economics really anger me. Many opponents of the law believe the government is wrong to limit choice. They think business should decide and let the free market take its course. Here’s an example of that logic (original pdf):
Aimone said that the public could settle the issue by deciding whether or not to frequent a place that allowed smoking.
“It’s a personal choice not to come in,” he said. “If enough people came in for the bar to pay the bills, the place would stay open. It’s basically economics.”
Oh, this pseudo-economic argument is appealing. Heck, I’ve even uttered it many times when I have a shallower economic understanding.
The claim is analogous to this one: if smoke-free bars were a valued service, wouldn’t competition create a market?
As the restaurant example tells you: no, this is not necessarily true.
Bars are susceptible to local competition. If one bar alone tried to ban smoking, it would be excluding customers and probably losing revenue, like the restaurant that tries to charge for containers.
Non-smoking customers face a coordination problem too in trying to frequent non-smoking bars. If you were in a group of ten people, and just one person really wanted to smoke, but the rest of you did not care as much, you would probably end up going to the smoking bar. It’s not a simple cost-benefit analysis—there is strategic behavior and coordination involved.
My friend was witness to these incentives. Several years ago, his downtown city had exactly one non-smoking bar. Even though it was a clean and great place, it was never as packed as the smoking places. That’s not to say its smoke free policy was the only problem, but it wasn’t helping the bar’s traffic. The bar couldn’t sustain business and shut down.
The smoking ban is a way to coordinate. I have often heard that some bars and restaurants did want to go non-smoking long before the law. It’s a personal concern: bartenders and managers had to breathe in smoke for their long shifts. But they could never coordinate with enough bars to change to non-smoking without the risk of losing customers. Using this logic, you can see that it’s even possible that all bars could desire non-smoking but they could not cooperate. It’s possible that government is needed to limit options and improve the overall social outcome.
In short
Don’t let politicians justify inaction with bad economics. There may be other perfectly good reasons not to get the government involved—bureaucracy, not wanting higher taxes, etc.
But if you see something broken, do not assume that competition will fix it. As I argued in two examples, even if everyone wants a new outcome, it might not happen without a law. Don’t let misinterpreted economics get in the way of better results.





16 Responses to “Markets Aren’t Perfectly Competitive: The Game Theory of Why a Smoking Ban Might Make Sense”
1. “The subtext is that perfect competition always creates better outcomes than regulation. Sorry to burst their bubble, but this is simply not true.”
Both statements “perfect competition always creates better outcomes than regulation” and “Perfect competition doesn’t always exist” are true. Neither statement “is simply not true.”
2. “I’m sure the restaurant is not happy about this situation.”
Probably not. If the restaurant were truly unhappy, they would not offer the boxes at all (they exist!). If the restaurant were somewhat unhappy, they could offer different size boxes or adjust portions, optimizing on obvious constraints. Both of these could be competitive advantages: obviously a consumer like you would prefer a restaurant with the “correct size,” and many people would prefer a restaurant at which they possibly pay less for less wasted food. Given that these restaurants have neither policy, and relying on the risky assumption of a rational manager,
one can conclude that the restaurants are operating at an optimal level.
3. “It’s not a simple cost-benefit analysis”
The value of hanging out with friends at a bar is more positively valued than the negative sum of the smell of smoke or health deterioration. While it’s true that personal values are dependent (and hence include “strategy”), that does not mean the decision is not a simple cost-benefit analysis.
By Glenn on Mar 4, 2008
I have to agree with a lot of Glenn’s points, and also toss into another issue I see, specifically your use of “optimal” seems to be a personal judgment.
For the restaurant, its probably most optimal from cost-basis to buy a single-size container in bulk vs buying different containers. Sure, as you point out Presh, there’s a lot of waste, and you might not like the huge containers showing up in your fridge, but…for the restaurant its cheaper to have a single take-home box.
Also, allowing customers to use their own containers opens them up to a potential lawsuit (ie, unclean personal containers could lead to food-cleanliness lawsuits).
The smoking issue is more complicated, because it could be a matter of public health, in which case…I’d argue that the optimal solution is one that promotes health.
Likewise, with emissions from factories, its probably cheaper to not care about where the waste goes…but as a public health concern the government needs to make sure the “optimal” solution of clean air / water / land is reached, which the free-market may not facilitate.
By RohoMech on Mar 4, 2008
I think you are right. Social surplus of cleaner work environments probably outweighs any inconvenience caused to smokers… and the beauty is restaurants don’t lose their smoking customers in the long run. This is a very complicated problem though. I’m sure some economic advisor has modeled it and come to the conclusion that it’s worth a shot.
This is such a great post Presh. Rock on.
By Monica O'Brien on Mar 4, 2008
@Glenn: Good counterpoints. You caught me on some less precise language. Here’s what I meant to say.
Point 1 should read: “The subtext is that perfect competition exists for the topic at hand and will create a better outcome than regulation. Sorry to burst their bubble, but this is simply not true.” My statements in the article are misleading as you point out.
Point 2 is about externalities and strategic behavior. I don’t care about waste, and neither does the restaurant. I agree each of us is individually making a rational choice, but it’s not optimal socially. I create unnecessary waste by failing to bring my container. A real-life example on this topic is how cities sometimes ban using styrofoam containers–neither the consumer or the restaurant would otherwise necessarily care to change.
Point 3 addresses the passage saying free choice would lead to either non-smoking bars or smoking bars closing down. I agree all choices ultimately can be seen as a cost-benefit analysis. Sorry if I was unclear about that.
By Presh Talwalkar on Mar 4, 2008
@Rohomech: “Also, allowing customers to use their own containers opens them up to a potential lawsuit (ie, unclean personal containers could lead to food-cleanliness lawsuits).”
Ack, I forgot about that. Good point. It seems like everything in America is about minimizing lawsuits. More on that in another article
By Presh Talwalkar on Mar 4, 2008
@Monica O’Brien: I wonder how I’d feel about this if I were a non-smoker…Thanks for the positive feedback. I needed to hear something nice after getting deflated from Glenn and Rohomech’s smart criticisms
By Presh Talwalkar on Mar 4, 2008
Presh;
Isn’t the choice between joining a smoker’s bar/non smoker’s bar a game with two stable equilibrium, all smoker and no smoker, and an unstable equilibrium at some mixed level of smoke/smoke free?
From a political point of view, there would seem to be very little impetus to create stability for the mixed equilibrium.
By michael webster on Mar 4, 2008
@michael webster: Nicely said. I think you’re right and I enjoy your phrasing. I could not have written it better myself
Using your interpretation, my article argues that the market might end up in an all-smoking equilibrium even if it’s not a socially optimal one.
This brings up all sorts of memories about developmental economics. I’m going to have to brush up on that and write something…
By Presh Talwalkar on Mar 4, 2008
I will have to respectfully disagree with the equilibrium comment.
In theory it makes absolute sense. But practically, smoke and pollution are negative externalities. There is no motivation for a government intervention that leads to an all-smoking equilibrium. That would create more loss of economic surplus. But a government would be motivated to do the opposite, because then there is a gain - the benefits outweigh the costs, basically. Or at least that’s what we hope here in Illinois!
Governments are in place to correct for free markets that don’t come to a socially optimal equilibrium (I think Presh already said this in his post). The government imposes taxes or creates laws to shift the demand and supply curves to reach that socially optimal equilibrium.
By Monica O'Brien on Mar 5, 2008
Monica; just how would decided that “the benefits outweigh the costs” if you didn’t allow people to make the choice from which you could deduce their preferences?
By michael webster on Mar 5, 2008
Readers of this post may want to refer to Chapter 9 in Dixit and Nalebuff’s “Thinking Strategically” for a nice introduction into games with two equilibrium which are stable and one unstable equilibrium. It is a very good non technical introduction.
By michael webster on Mar 5, 2008
Presh, you have a great discussion going here.
Michael, as I said before, I do like what you’ve written, and I think in theory it makes sense. It is a very interesting possibility to consider.
My answer to your question is the cost-benefit analysis doesn’t depend on people’s preferences alone, if we are looking at “is the law best for society.” Maybe you are looking at this differently, which I understand.
The way I’m defining the problem, I also see health risks to be a factor. Even if a worker prefers to enhale second-hand smoke during their shift, the smoke also affects their health negatively. This could raise the cost of health insurance as a whole in Illinois.
Maybe people only smoke when they are at bars (the casual smoker, I am one) and these people would decide not to smoke at all anymore, which could benefit their health positively.
Then maybe cigarette companies would sell less cigarettes and it would be bad for the economy. I have no idea.
I think what Presh said is right - the cost-benefit analysis for society would be some crazy modeling exercise. I find it difficult to believe that in reality, with all the information we have about second-hand smoke on health, that we could get to an all-smoking equilibrium. But maybe it is possible and I am projecting my own preferences as a non-regular smoker onto the problem.
By Monica O'Brien on Mar 5, 2008
Going back to the notion of government influencing choices at a fundamental level, Ronald Reagan had it right. He said “Governments exists to protect us from one another, where government exceeds its mandate is when it tries to protect you from yourself”.
By Mahesh on Mar 5, 2008
Monica, Michael, and Mahesh–thanks for excellent discussion. There are so many ideas going on here, so I’ll try to summarize:
1. Equilibria exist in theory for all bars smoking, none smoking, or some mix. Where we end up depends on initial conditions. The mixed one might is not stable–we’ll likely observe all or no smoking. (michael webster’s point)
2. Politicians and economists suspect that all-smoking is worse than no-smoking. A law is needed to create a no-smoking equilibrium and increase social welfare (Monica O’Brien)
3. Governments can improve imperfect markets. When should the government step in? The government is better suited to protect us from externalities but not from free choice when we only harm ourselves (Mahesh’s point).
By Presh Talwalkar on Mar 5, 2008
Life and its joys can’t be reduced to numbers. These bans area about decreasing options and social engineering. Not to forget, many places with bans (Ireland, Scotland, Italy) report rises in smoking, especially in the home around children.
The dangers of smoking are further open to debate. Austria and Japan have higher rates of smoking, and longer life expectancies than he US or Britain.
By chris on May 9, 2008
chris: This is certainly one for the law of unintended consequences:
“Not to forget, many places with bans (Ireland, Scotland, Italy) report rises in smoking, especially in the home around children.”
I agree it is hard to separate causal effects for health studies. I think better measures are direct individual impacts. I feel much worse after going to a smoky bar and my athletic performance decreases. Should that warrant a ban? I don’t know, so I leave that up to the experts who I hope do their due diligence.
By Presh Talwalkar on May 9, 2008