Friday, October 12, 2012
No Coercion, Maybe?
Imagine that you haven't been following the news for a month (humor me). You walk into a Best Buy browsing HDTVs. You have already decided on a limit on how much you would pay for a new HDTV, say, $1000. While browsing, you notice that all of the HDTVs are now $800 or less. Giddy with joy, you buy your dream HDTV.
As you walk out of the store, you see a headline plastered on the front page of the Miami Herald that states: "New Mandate to Purchase HDTVs Lowers Prices." You look up the article and discover that, a month earlier, Congress passed a law requiring all persons to purchase an HDTV, and that the law caused TV manufacturers to lower prices. Were you coerced by Congress into buying the HDTV? What if you knew about the mandate before buying the TV?
In a new paper I co-wrote with Ralph Boleslavsky, an economist at Miami, we argue that there is no coercion in both cases.Instead of TVs, our paper, entitled "Does the Individual Mandate Coerce?," focuses on the individual mandate provisions of the Affordable Care Act, which require certain individuals to buy health insurance or pay a certain amount in taxes.
As probably everyone who reads this blog knows, last term the Supreme Court upheld the individual mandate, but a majority of justices concluded that Congress cannot use its Commerce Clause power to coerce individuals to buy a product. Much of the commentary about the case has assumed that the mandate coerces (it is called a "mandate," after all). However, we use a simple game-theoretic model to show that the mandate may not coerce if the market for health care and health care insurance has certain features. The basic intuition is reflected in the hypothetical above: the individual mandate may cause insurers to sell health insurance at a price that everyone subject to the mandate would be willing to pay. Thus, no coercion.
Admittedly, someone may not want to purchase affordable health insurance for ideological reasons. We are skeptical, however, that anyone would be that ideologically rigid. Moreover, the Act excludes from the mandate those who do not want to purchase insurance for religious reasons, excluding the very individuals most likely to forgo purchasing health insurance, no matter the price, because of their beliefs.
In the paper we concede that the model is stylized. But we also point out that certain features of the model -- relative risk homogeneity among the persons subject to the mandate, and economies of scale that insurers can use to purchase health care from providers -- have some empirical support. Now, I cannot force you to accept this conclusion, but it is certainly food for thought, and suggests that maybe supporters conceded the coercion of the individual mandate too easily.
Posted by Sergio Campos on October 12, 2012 at 11:13 AM | Permalink
TrackBack URL for this entry:
Listed below are links to weblogs that reference No Coercion, Maybe?:
I have not read the article, with apologies, but a core example cited are young people who choose not to buy insurance because they feel they don't need it and instead use the money for other reasons. The 20 something who rather use the money used for insurance to go to a club each weekend is a hard nut to crack. It is perfectly fine in my opinion to tax their paychecks, especially since (unlike Roberts) I readily think they are not "inactive" here but actively involved in the stream of commerce in respect to the insurance market and should not free ride in various respects. But, everyone is not likely willing to pay here no more than everyone is willing to do any number of things, even if in their best interests, that regulation "coerces" them to do to some degree.
Posted by: Joe | Oct 12, 2012 12:15:11 PM
Wait a minute. Mere game theoretic analysis can't solve the question of coercion, and neither can the hypo you give. In the hypo, it's not coercion because you didn't buy the HDTV because of the threat ("buy a tv or get punished"), you bought the TV because of the price. Lets change the hypo. Now, you know that the law requires you to buy a HDTV, but you don't know economics, so you grudgingly go in, prepared to pay through the nose. Arriving at the store, you're overjoyed to find that the TV you were going to buy anyway -- because there's a gun pointed at your head -- is now cheap enough that you WANT to buy it. Still, this sounds more coercion-like.
Or at least we need a full-scale philosophical argument to suggest otherwise... is there one in the paper?
Posted by: Paul Gowder | Oct 12, 2012 12:32:50 PM
Thanks for the comments. @Gowder Why is it coercion if you have knowledge of the mandate, but would have done the action voluntarily anyway? @Joe I agree, and a person who would not have purchased the insurance "but for" the mandate is coerced. We only show that, under some conditions, the mandate may lower the price of insurance enough so that someone would buy the insurance even in the absence of the mandate. It is worth noting, moreover, that the mandate does not apply to individuals who would be willing to pay, but can't afford it, which is similar (although not exactly like) your opportunity cost example.
Posted by: Sergio Campos | Oct 12, 2012 2:03:47 PM
I have no particular desire to kill anyone. Nonetheless, it seems to me that the government coerces me with the laws against murder. Why? Because quintessentially coercion is the removal of options from someone's choice set by means of threats. It doesn't matter that I wouldn't want to make those choices, nonetheless, the government takes them away. And the threats might well factor as a partial cause in my decision making process. Suppose someone cuts me off on the highway, and it makes me really really mad. I might have the following thought process: "why, that &$@@&@#% -- I should run him off the road! But no, that would be wrong, I'd never be able to live with myself afterward, and, moreover, I' go to prison." That kind of thinking -- which I'll suggest to you is common -- makes it rough to sort out what caused my refraining from the murder; we might say my moral qualms and the law both caused my restraint, even though each would be sufficient on its own. The same goes, perhaps, for the individual mandate.
I'm loosely tracking Nozick's (really influential) conceptipn of coercion, incidentally -- nicely summarized in the Stanford Encyclopedia of Philosophy article on the subject.
Here's the danger of the alternative view -- of saying that laws don't coerce when they just demand you do something you were gonna do anyway. It seems to strip ALL regulatory laws of their coercive power except against the Holmesian "bad man" or against citizens who disagree. Why? Because laws have a persuasive power as well. Some laws may persuade the citizen that the legislature's judgment was right. Others may persuade the citizen to follow them just because she believes she has a moral duty to obey the law. But either way, there's a sense in which she'd "do it anyway" as a result of the law's non-threat incentives. Just like the health care case.
But from this it would follow that many if not most regulatory laws fail to coerce many if not most citizens. And that seems to be biting a pretty big bullet...
Posted by: Paul Gowder | Oct 12, 2012 3:22:34 PM
@Gowder We are defining noncoercion as doing something voluntarily in the absence of the legal requirement. Your examples are interesting, but the key question is whether, when you strip away the requirement, you would have done it anyway (including stripping away the law's persuasive power). Admittedly, counterfactuals are hard, but modeling gives you some idea of what the counterfactual might be. In fact, this definition of noncoercion is used constantly in the law. The "reliance" requirement in fraud, for example, essentially asks if you would have done something in the absence of the fraud.
Posted by: Sergio Campos | Oct 12, 2012 3:40:06 PM
Sergio, I am trying to dispute that "key question" (and that definition of noncoercion) with those "interesting" counterexamples...
Also, does that definition of noncoercion work for what you want to do with it? After all, people who buy insurance because the mandate lowers prices aren't doing it "in the absence of the legal requirement" -- they're doing it because of the legal requirement, in the sense that but for the legal requirement the prices wouldn't be lower, so they wouldn't do it. I think your argument needs noncoercion to be more individualized, something like "legal requirement L doesn't coerce agent A if A would do what L commands even if L did not require A to do it."
Posted by: Paul Gowder | Oct 12, 2012 4:08:36 PM
I don't understand your last point. I'll admit that the mandate can cause consumers to purchase health insurance, but causation is not synonymous with coercion. Let me put it this way. According to the Nozick definition of coercion you reference, coercion only obtains if "(6) Part of Q's reason for not doing A is to lessen the likelihood that P will bring about the consequence announced in (3) (Nozick 1969, 441–445)." See http://plato.stanford.edu/entries/coercion/#NozNewAppCoe. That is the definition we are using. I call your examples "interesting" because they raise counterfactual uncertainty about what caused someone to do something. But counterfactual uncertainty does not prove there is coercion, only that there is uncertainty about whether there is coercion.
Posted by: Sergio Campos | Oct 12, 2012 4:18:43 PM
We should distinguish motivation from causation -- examples like the murder case suggest that one can be motivated by the commands of a law without it being the case that they wouldn't have done it without the law. And one thing I'm suggesting is that Nozick's (6) is about motivation rather than causation. Put differently, the fact that they will be punished for not doing so can be one of the reasons people buy health insurance even if it happens to be the case that they would buy health insurance were they immune from punishment.
But I agree that is is uncertain factually whether these mixed motives obtain -- maybe people are just indifferent to the mandate penalty.
Posted by: Paul Gowder | Oct 12, 2012 6:50:00 PM
Thanks for the reply.
"We only show that, under some conditions, the mandate may lower the price of insurance enough so that someone would buy the insurance even in the absence of the mandate."
That's understandable and lowing costs is an aim but reference is made to "everybody" and some subset appears to not be covered by your analysis. That is, people who otherwise would have to pay or be taxed under the law in question.
"can't afford it, which is similar (although not exactly like) your opportunity cost example"
The opportunity cost I cited does not really seem that "similar" to the criteria set forth in the legislation, which is based on income. Twenty-somethings often spend freely, not thinking about the future, and this can be totally discretionary income that could readily be applied to insurance costs.
Posted by: Joe | Oct 12, 2012 7:52:45 PM
I had the same reaction as Paul. I don't want to get into the weeds of a philosophical debate, but I will just note that the way that you present the hypothetical in the main post is really downplaying the contrary intuition. That is, you FIRST come with a hypothetical where one does not even know there is a mandate (and where you emphasize the low price), and then insert a TINY sentence where you note there is a mandate, and even then you downplay its coercive impact by not mentioning any kind of penalty.
So suppose I say that a government agent holds a gun to your head and tells you to buy a HDTV or he'll shoot. You go to the store (with the gun to your head) and find that the HDTVs are so cheap you would have bought one anyway. Now I ask the audience whether you have been coerced into buying a HDTV. I dare say many people, as a matter of ordinary language, would say yes. You can come up with philosophical counterarguments--all centered on the fact that you would have bought the TV anyway--but Paul's point is not that you can't make the argument. It is that you don't make the argument and instead seem to frame the hypo to dodge it.
Posted by: TJ | Oct 12, 2012 9:27:07 PM
I don't know how important to your argument is the claim that the Act "excludes from the mandate those who do not want to purchase insurance for religious reasons." But I wanted to pop in to point out that the exemption does not apply to all those who have religious reasons for failing to purchase insurance, as the DC Circuit litigation in Seven-Sky v. Holder demonstrates: http://aca-litigation.wikispaces.com/file/view/District+Court+opinion.pdf
Posted by: William Baude | Oct 12, 2012 11:02:55 PM
Thank you all for the wonderful comments. I really appreciate you all engaging with the argument. We simply make the argument that, by lowering the price of health insurance, the mandate may not be a but-for motivation for purchasing health insurance. I understand Nozick's argument for coercion (and, really, any common sense notion of coercion) to require the credible threat to be a but-for reason for doing an action (thus, both causation and motivation).
I am not framing the hypo to dodge the argument. I am framing the hypo to start with the counterfactual - the threat of the mandate is not present, so we know that you would buy the HDTV even in the absence of the threat. Both Paul and TJ's arguments concern either examples of counterfactual uncertainty, or examples where the mandate "penalty" is itself harmful even though it is not the but-for motivation for buying the HDTV. If someone holds a gun to my head and asks me to take a big bag of money, the gun does not coerce me to take the money (I would gladly take it in the absence of the gun), but it is upsetting regardless. Critics of the mandate are not objecting to the upsetting nature of the mandate. They are arguing that Congress cannot coerce you to buy a product you otherwise wouldn't purchase voluntarily. By the way, I say "pay a certain amount in taxes" in the hypo to avoid the tax/penalty debate, not the minimize the penalty. It can be a billion dollars, and the argument would be the same.
Will, the comprehensiveness of the religious exemption is not essential. The exemption would obviously not exclude someone who simply hates Obama. But we wanted to show that the Act is actually pretty meticulous about excluding people who would be coerced, thus showing that the coercion of the mandate is very small or, as our model suggests, possibly nonexistent.
Posted by: Sergio Campos | Oct 13, 2012 5:31:52 AM
To the extent that you explicitly define "coercion" as "but for motivation," then the analysis itself follows logically. But that only raises the question of whether your definition is the same as that of mandate opponents -- otherwise you are simply talking past each other. And I would be quite skeptical that "[c]ritics of the mandate are not objecting to the upsetting nature of the mandate." It is true that they are mainly arguing that "Congress cannot [mandate] you to buy a product you otherwise wouldn't purchase voluntarily" (because that is easier to argue on the facts); but it is far from clear to me that they wouldn't object to Congress putting a gun to someone's head to make them buy insurance that they would purchase voluntarily, or that they wouldn't consider that "coercion," if those were the facts being presented. Do you actually know any mandate opponents who actually say that putting a gun to someone's head to make them buy a product they would otherwise purchase anyway is not coercive?
Posted by: TJ | Oct 13, 2012 6:52:18 AM
TJ, we do define coercion as but-for motivation. I also understand the Constitutional argument as using this definition of coercion. In fact, as we note in the paper, that is usually how the law defines coercion, as I suggested by my fraud example. To the extent critics have a different definition of coercion, then 1) how is it relevant to their Constitutional argument, and 2) frankly, how is it coercion? I also object to having someone put a gun to my head, but I would object even if he wasn't coercing me to do something. I don't like people pointing guns at me. But coercion is forcing someone to do something against their will, which, to me, requires but for motivation.
Posted by: Sergio Campos | Oct 13, 2012 7:36:02 AM
TJ's hypo occurred to me as well. I'll add a twist to it: you're already on your way into the store to buy a TV when someone *shows* you a gun and says buy one or else he'll use it. That still seems like coercion, because even though you intended to buy a TV, you could have decided not to; now you have no reasonable choice. (You have an unreasonable choice: take your chances with the thug outside.) It's the difference between being ordered to do something and doing that same thing voluntarily; even though the end result is the same, one is coerced (i.e., not voluntary) and the other is not. I suppose you could argue that even seeing the gun in the hypo is separately upsetting, and that I am conflating the objectionableness of that with coercion, but I'm unclear how you could ever have coercion to do X without a threat of something objectionable happening if you don't do X; the unpleasantness of the threatened action is precisely what makes it coercion.
Posted by: Bruce Boyden | Oct 13, 2012 6:11:40 PM
Again, thank you all for the comments. I really appreciate them. Bruce, we do not dispute that a threat is a necessary condition for coercion. But, in our view, as supported by the law, it is not a sufficient condition for coercion. Paul and TJ contend that it is. We argue that you also need for the threat to be a but-for reason for acting. I take condition 6 of Nozick's argument to also require the threat to be a but-for reason for acting: "Part of Q's reason for not doing A is to lessen the likelihood that P will bring about the consequence announced in (3) (Nozick 1969, 441–445)." http://plato.stanford.edu/entries/coercion/#NozNewAppCoe. As the entry further states, "Nozick's analysis emphasizes the alteration in the coercee's choice of actions which results from the way the coercer's proposal affects her reasons for acting." So do we.
Posted by: Sergio Campos | Oct 13, 2012 6:49:38 PM
One clarification. I (and I read Paul) am less contending that "it is," than "it could be." I am not saying that you cannot define coercion as but-for motivation and a threat as a necessary-but-insufficient element of coercion. I am saying that this is not the only definition, likely not the definition that the opponents you are trying to refute--either mandate opponents or the conservative judges who ruled for them on that issue--have in mind, and one that affects whether people will be persuaded by your claim of demonstrating "non-coercion". My problem is not that your definition is wrong, it is that you seem to find an opposite definition impossible (i.e. "frankly, how is it coercion?") and therefore kind of run right by the issue with unacknowledged embedded assumptions in framing the hypothetical and making the rhetorical claim. Of course, you might not want to address it anyway--a single article can't address every embedded dispute--but you should at least acknowledge this is a debatable issue when making an ambitious claim about "non-coercion."
Posted by: TJ | Oct 13, 2012 7:44:19 PM
TJ, I think you do a good job of describing the disagreement. The paper is only 16 pages long, so, needless to say, we do not address this disagreement in any great detail. We are, however, clear about our definition of coercion. We also point out that this is how the law defines coercion (see, e.g., fraud), so it is weird to me that any judge would use our definition of coercion in one context but not another. Finally, I just don't get the alternative conception of coercion you describe. Putting a gun to my head may be many things, but if I was going to do the action anyway, how does the gun coerce me to do it?
Posted by: Sergio Campos | Oct 13, 2012 8:30:21 PM
Sergio, the fact that you "just don't get the alternative conception of coercion" is precisely the problem, and I can't explain more than I have. One note, however, is that you are also wrong that "but for" motivation is "how the law defines coercion"--if by that you mean it is the ONLY way that the law defines coercion. There are other definitions. If we look at, say, fraud or duress as a method of voiding a contract, the Restatement Second of Contracts (s 167 cmt. a) says: "It is not even necessary that he would not have acted as he did had he not relied on the assertion."
Posted by: TJ | Oct 13, 2012 10:04:55 PM
TJ, you have provided an intuition, and argued that others have that intuition. I don't have that intuition, questioned whether others have it, and have defined what I mean by coercion. So, like you, I also "can't explain more than I have." As to the legal point, duress as a contract defense "has been defined as a 'threat of harm made to compel a person to do something against his or her will or judgment; esp., a wrongful threat made by one person to compel a manifestation of seeming assent by another person to a transaction without real volition.' - Black's Law Dictionary (8th ed. 2004)." http://en.wikipedia.org/wiki/Duress. As I read it, the duress defense in contract law focuses on attempted coercion, not on actual coercion, which is why proof of actual coercion (e.g., but for motivation) is unnecessary. Moreover, and I think this is reflected in the law you cite, a per se rule against attempted coercion makes sense in certain situations given the difficulties of proving actual coercion. (see, e.g., the fraud on the market presumption). So I don't see your example as a counterexample to my claim.
By the way, I am really enjoying this exchange, and I thank you all for taking the time to respond to my argument. It is greatly appreciated.
Posted by: Sergio Campos | Oct 13, 2012 10:52:54 PM
"but if I was going to do the action anyway" Sergio, how do you know this? It's being presumed in the hypo in the post (and also mine), but it's not something that people actually know with certainty in the real world -- what they would have done in some sort of counterfactual universe. Absent compulsion, I would be free to change my mind at the last minute and decide that I don't really want to buy a TV today, and preserving that option is something that is usually important to me, even if I don't exercise it. With compulsion, I no longer have the option. My freedom is poorer by whatever the option price was.
I'm also not persuaded that the way the law defines proof of compulsion, e.g. for some affirmative defenses, should govern how we think of actual compulsion. (I'll take your word that the law requires some sort of showing of an impact on behavior in many cases.) There's compulsion, and then there's proving it sufficiently clearly to overcome some prior obligation. I don't think at the macro level of a constitutional power case we really need to be concerned about the micro level of whether person A could prove compulsion, and person B could prove compulsion, etc.
But perhaps we're just running into some sort of incommensurability problem. It seems frankly bizarre to me to think that in the hypo of the person who's going to buy the TV anyway (and let's say we have omniscient knowledge that that is in fact the case) that they are not compelled by the person who orders them at gunpoint to purchase the television. It seems a very odd and nonstandard use of the word "compel."
Posted by: Bruce Boyden | Oct 13, 2012 11:38:29 PM
Sergio, you can't say in your article that the law defines coercion as but for causation and and cite fraud as your primary analogy, and then when someone gives you an authority that says fraud does not require but for causation, come around and say that fraud is not really about actual coercion but about attempted coercion. That makes your fraud analogy an unfalsifiable argument.
Posted by: TJ | Oct 14, 2012 2:49:21 AM
TJ, I was only saying that, in some contexts, courts do not require a showing of but for motivation to prove fraud because it is too hard to prove. They use per se rules - the fact of the threat is enough. I interpret your example of duress as a contract defense and the fraud on the market presumption as examples of courts using per se rules, not using a different conception of coercion. Bruce, in a way, proves my point. It is difficult in a real world case to prove what actually motivates a person to do something. But economic models provide guidance on the counterfactual. Courts use such models to determine the counterfactual all the time, as I have written in a previous article. See http://ssrn.com/abstract=1999691. The hypo was set up to clearly establish the counterfactual, so that, when we include the threat, it is clear that the threat was not the but for reason for buying the HDTV. Finally, how is your freedom poorer if an option is removed that you valued at zero?
At this point, I have outlined a definition of coercion that is generally accepted in the law and in philosophy, as Paul pointed out with the Nozick definition. But, as with the Nozick conception of coercion, "[o]thers have disagreed with the success condition, though Nozick's is apparently a majority view." http://plato.stanford.edu/entries/coercion/#NozNewAppCoe I also have a success condition for my definition of coercion, and you guys don't. You all have a different intuition, and I don't have it. I am not sure that much more can be said at this point. Again, thank you so much for your comments on this.
Posted by: Sergio Campos | Oct 14, 2012 7:23:43 AM
"Finally, how is your freedom poorer if an option is removed that you valued at zero?" But I don't value it at zero -- as I said, ordinarily it's important to me to have the option to change my mind. And I doubt anyone values that sort of option at zero for most transactions. Certainly not people who actually pay money for options; and many of those people believe it was money well spent, *even if they don't exercise the option*.
Posted by: Bruce Boyden | Oct 14, 2012 10:14:06 AM
Quick footnote based on skimming the last few comments:
- I wasn't suggesting that the mere presence of a threat is sufficient for coercion -- rather, a threat that actually constrains one's choice set may be.
- I'm also trying to open the issue of partial, multiple, and overlapping motivation. That is, regardess of whether you accept the previous idea, T can be "part of Q's reason for not doing A" without being the but-for cause of Q not doing A, when Q, in deciding not to A, considers multiple sufficient reasons, one of which is T.
Posted by: Paul Gowder | Oct 14, 2012 11:40:44 AM