Judge Posner, respectfully I think your discussion of this subject would be substantially improved if you removed the tangential remarks about prostitution, specifically:

"There is no dearth of persons willing to sell the right to see them performing or otherwise appearing in the nude, so there is a well-functioning market without need to coerce anyone to so appear."

If your assertion was correct, we would not be seeing frequent reports and articles about the huge size of the international sex trade and other similar problematic elements.

Judge Posner, I very respectfully disagree with the view that you take of information privacy.

First, by stating:

All that privacy means in the information context (for I am not talking about the Supreme Court�s usage of �right of privacy� to describe the right to an abortion and other sex-related rights�the Court appropriated the word �privacy� to describe these rights presumably because of its positive connotations) is that people want to control what is known about them by other people

you are ignoring the fact that modern technology allows for "information" to be used to evaluate movement and location. This monitoring (dare I use the word "surveillance") goes directly to issues of autonomy privacy - a clearly constitutionally protected right. Tracking and monitoring cases (e.g. US v. Karo, US v. Knotts, and Dow Chem. v. US) have already shown the line between use of electronic information used in tracking - as "low-tech" as these may seem today. Cases involving physical tracking using Internet and electronic transactions will inevitably arise at some point. The physical rights of privacy are blurring with the informational rights of privacy. Control of this information may be as important as physical freedom.

Secondly, you state that people often want to "conceal facts about themselves that are embarrassing or discreditable." While this is true, people also want to conceal facts which may be flattering or neutral as well. The point being that a person ought to have some control over how they are presented or represented to society to some degree of certainty. Allowing a person control over personally identifiable information is a natural result of this desire. While information cast into society and made public should not be given such protection, the Court has regularly recognized a private person's ability to do so - possibly, most notably, in NY Times v. Sullivan.

Often this involves concealing information that would cause potential transacting partners to refuse to transact with them or to demand better terms as a condition of doing so. Such concealment is a species of fraud.

More often, this likely does not and protection into the seclusion of one's personal life advocated for in Warren and Brandeis' seminal work, the Right of Privacy.

While your view of privacy may hold true in some instances, basing policy decisions upon such a view in a era of such technology growth would subject the general public to unprecedented scrutiny by the government as well as other citizens, and would greatly weaken the privacy rights which have been established and relied upon.

Best Regards, CS.

"The second motive for privacy, however�the desire to conceal discreditable facts�is more questionable from a social standpoint."

Questionable, yes, but not wholly illegitimate. Consider evidentiary rules, where the bulk of all possible information available about a witness is excluded for reasons of confusion, time wasting, or prejudice. Society frequently has a substantial interest in preventing certain people from knowing relevant information about the credibility of other's since the possibility of any of those three concerns is so high. Valid arguments can be for disclosing rape victim's sexual histories or tortfeasors insurance status, yet valid arguments can be made for excluding both, largely that the knowledge of either is very likely to short-circuit rational, empirical thought on behalf of the factfinder.

That is, I believe, a missing point of your analysis: that parties frequently (and validly) conceal information to prevent confusion or prejudice on the other party's behalf. Consider an African-American not revealing their race to a mortgage refinance corporation because they've read reports that that African-Americans pay higher rates, even when all other variables such as socioeconomic status are considered. This information is certainly "material" to the refi company since it changes their statistical profile, and yet few would say the mortgagee is unjustified in avoiding racism. Or consider a lawyer concealing HIV-positive status from his firm partners for fear of a "Philadelphia" scenario.

If, as the evidentiary rules suggest, we as a society believe that concealing material information from persons to avoid confusion, inefficiency, and prejudice is valid, then why should it not be valid in the information privacy context?

I think there is a useful distinction to make between privacy (the right to not walk around naked and to not be observed at home, the 4th amentment and similar concepts) and anonymity. I am very skeptical that the concepts are the same or overlap.

I think that privacy should be protected, but that anonymity is a historical anomaly that does not deserve protection. My heuristic for distinguishing between privacy and anonymity is to imagine a pre modern village. In the middle of the village is a marketplace where old ladies sat, sold vegetables and gossiped betimes.

You were born in the village, lived there and probably never went more than a few miles away. The old ladies in the marketplace knew who you were, where you lived and the identiies of all of your ancestors and kin. They also knew whether you liked brocolli, whether you paid your bills on time and if you were charitable to the poor. You were not anonymous, but they had no way of knowing what you did in private.

In the 19th century, cities began to grow so large that most people began to live in places where they were anonymous. This was a historical anomaly.

Now, anonymity has its uses, especially if you want to comit a fraud or a crime or overthrough the government. And many people began to enjoy anonymity, especially among radical political groups who dreamed of the "Revolution," that final redemtion from history.

In the 21st century, the electronic data processing revolution has begun to dispell the fog of the 20th century city that engendered anonymity and create a new global version of that ancient village square. It is not of itself evil, it is merely a return to the common condition of mankind. Revolutionaries and their sympathisers are deprived of a usefull tool and the world is not yet redeemed, but for most of us life will go on as it always had.

You give two broad motives for wanting privacy (protecting yourself from embarrassment and carrying out fraud), but I think you miss several others that are at least as important. One of the more obvious ones involves not giving potential criminals extra information about myself: If you know I have lots of available cash, I may be a good target for various scams, kidnapping of relatives, etc; if you know I'm out of town next week, you may want to steal my things while I'm gone.

--John Kelsey

"people will not speak freely if they think they are being overheard by strangers, and there is value in frank communications, including being able to try out ideas without immediate exposure to criticism."

Isn't there a concern that people will not speak openly/be frank with their doctors (something that clearly has tremendous social value if the privacy of medical records is not protected?

check out http://blawgandecon.blogspot.com

First of all, I find the reference to Kennedy an unnecessary cheap shot. Is he really the only Presidential candidate ever to conceal unfavorable information about himself? I can think of more recent examples.

I think there are good reasons to want to maintain privacy. I have little confidence that information that I provided to an insurance company, for example, would be used solely for the purpose of evaluating my application. Lots of people will see it, and I personally don't trust them to maintain its confidentiality. And the insurance company will be eager to take commercial advantage by selling the information.

Most important, I see no need to justify my reasons. There is a difference between keeping quiet and lying.

Judge Posner, with all due respect, "freedom" is not merely a "strongly positive connotation ... which obscures analysis."

On a more analytical level, however, I take major exception with your suggestion that "the desire to conceal discreditable facts" is troublesome even as an economic factor. I think that even looking at things economically, people tend to overweight these discreditable facts once they're exposed.

For example, take the Bill Clinton thing. I think there's a strong argument that Clinton should have been able to "conceal [the] discreditable fact" of his sexual relations with Monica Lewinsky, because the cost of revealing those facts and the subsequent overweighting of those facts by Congress, the press, etc., caused serious harm to the national interest. If Clinton's right to conceal that fact had been honored, millions of dollars of government funds would not have been wasted.

You sort of acknowledge this point when you suggest that the contrary would "drown society in trivial and distracting information," but that isn't completely satisfying. Failing to require blanket disclosure of private facts (to avoid drowning) but at the same time refusing to forbid the collection of those same facts by people with the resources to do so would likely have terrible distributive effects. Small, low-resource actors would be victimized by these "minor frauds," while large actors with the resources to collect and control this information would not. That's unjust. (Shameless plug: I discussed some of these issues in a student book note 6 years ago in the Harvard Journal of Law and Technology... although I'm now not too fond of the writing style I chose at the time...)

Finally, you fail to consider that many of the facts being concealed under the rubric of "privacy" are historical facts which may be of limited relevance in predicting present behavior. Credit reports are one good example. It is of only limited rationality to use the tainted credit report of someone who lost their job or had major medical problems five years ago to determine their reliability today, when they're healthy and securely employed. Yet, again, persons with access to those reports are likely to overweight that damaging historical information.

Also, what about simple forgiveness? Again, this goes to historical damaging facts. Should people be functionally removed from their maximum productivity just because of distant errors? Permitting people to conceal past errors by, for example, expunging criminal records, permits them to be fully employed. Do we really want a society where 20 year old criminal records follow everyone around and there's a bunch of unemployable people? Or do we want to institutionalize forgiveness by (except for the most henious crimes) permitting one's history to be concealed after a certain amount of time?

Even as to present damaging facts: do you think it would have been better, or worse, for America and the world if the "damaging fact" of Franklin Delano Rossevelt's polio had been more widely known?

Whatever one's opinion of the extent to which information about an individual should be property of that individual, I think that this situation provides a good example of the adage that "there's no situation so bad that it can't be made worse by a solution."

Take HIPPA as an example. The medical center where my physician works (and where my wife works) used to employ a computerized phone system to contact patients to remind them of upcoming appointments or of the necessity fro them to schedule an appointment. In the past year or two the system has been modified so that it reminds patients of their appointment with "the clinic" rather than a particular physician or department, since doing so might divulge to other people who overhear the message that this individual might have a particular category of medical concern. Of course it also eliminates the patients' ability to know which of the several hundred physicians distributed over several buildings they're supposed to see.

My wife recently told me that they're also remodeling one of the labs to eliminate the necessity of calling patients by name when they can be served, since that also would divulge information about the patients. "But wait a minute," I observed. "Other patients sitting in the waiting area of this lab and any other waiting area in the clinic can already see the other patients waiting there, thus conveying the exact same information. Shouldn't they enter masks (or blindfolds) as soon as patients arrive to prevent this obvious violation?" She, being the sort of intelligent person whom I would choose to marry, agreed, noting the additional problem of cars openly parked in the lot, displaying their license plates for any private detective, reporter, or pharmaceutical salesperson to record for any purpose whatsoever.

We both laughed briefly, only to come to our senses when we realized, based on several decades of experience with government bureaucrats, that there's absolutely no proposal so absurd that some earnest cog, hoping to become a bigger cog, won't advocate, and that politicians in search of a few more votes won't demand.

Thanks for your comments Judge Posner, I always enjoy your refreshing take on matters. I agree with you totally about 'privacy' being a word of positive conotations making it difficult to analyze like freedom. In response to the individual who said that there are important reasons to support freedom I agree, after all pointing out that a word has positive conotations does not imply that these are always undeserved. I take it that the point Posner was trying to make, and the one I agree with, is that words like privacy or freedom really mean something like 'privacy when it is good' or 'freedom when it is good'. As Posner points out we don't call wiretaps on mobsters violations of privacy nor do we call laws preventing nudity or murder a restriction of freedom. In short the point is that these words have a moral component in their meaning so one can't analyze freedom to see if it is good or not since the word just means those freedoms which are good.

However, I think you miss one very important concern with privacy. In particular you don't address the dangers of differential privacy which I think is the true danger. Unfortunatly, people of privleged racial backgrounds or better socioeconomic classes often recieve more privacy than those in less favored situations. For instance racial profiling to do police stops is one example of such differential privacy and in chicago they are mounting microphones to listen for gun shots but only in the high crime (and hence low rent) neighborhoods. Thus attempts to combat crime and the increased ability of the well off to pay for forms of privacy (owning ones own house versus living in an apartment with thin walls or with a roomate who can let the police in) threaten to create a society with differential availibility of privacy.

So why is this differential availibility of privacy a problem? The problem is that we as a society condem and even criminalize many private actions which occur quite commonly. By preferentially revealing these activities in only the poor and minorities one produces a skewed view of both the disadvantaged group and the behavior (both are smeared by association). For instance I suspect that many of societies negative attitudes about both marijuanna (and over drugs to different extents) and blacks arises because of the preferential revelation of drug use amoung the poor and black despite the statistically comparable use by the well off and white. By virtue of this preferential revalation stereotypes about both drug users and poor blacks are inappropriate strengthened.

If this situation does not convince you consider the (fading) moral prohibitions against sodomy (in this usage meaning sex acts of a non-procreative type). Now this sort of behvaior has been both condemed and made illegal in many states but nevertheless its practice has been very common. However, if we now imagine a situation where the poor or otherwise disadvantaged lost a great deal of privacy but the rich did not we would of course except to discover many instances of sodomy amoung these groups for the simple reason that it is common amoung all groups. Yet this is a ready made smear on these groups and no one in the advantaged group with privacy is going to speak up to defend them lest they too be identified as sodomites. Thus by preferentially revealing the private behavior only of minority/disadvantaged groups you play directly into stereotypes about these groups being morally inferior and perhaps even result in the arrest of these individuals for a socially common activity.

In short people have a disgusting habit of publicly agreeing (or not disagreeing) with moral and legal prohibitions they violate frequently in private (I certainly no people who smoke weed who wouldn't stand up when the evils of MJ are being discussed). If we don't carefully make sure that privacy is degraded equitably we risk both enforcing stereotypes about the prohibited behavior (it isn't the sort of thing good upstanding citizens do) and the disadvantaged groups (people taking the fact that they engage in the prohibited activity as proof of deliquency). So while I think many of the concerns about 'privacy' in the abstract are overplayed I do worry greatly about making sure privacy isn't something reserved only to the rich and privleged.

Also I think it is important in this discussion to realize there are really two sorts of concerns which people have. Sure for some sorts of information people want it to be truly private, meaning the have near complete control over its dissemination. However, for many other types of information people only care about it being obscure, by which I mean it is not easily accesible, i.e., requires a non-trivial expenditure of effort or money to discover. In fact I submit that it is a loss of obscurity which is most troubling to people not privacy.

Several examples spring immediatly to mind. For instance while an individual might not have a big problem with their divorce records being availible to interested parties at the courthouse, and might even support their information being included in studies on divorce outcomes but would be very disturbed to find out that the same information was availible by googling their name. It is one thing for faceless strangers to be able to review this sort of information and another for every potential date or employer to have easy access.

Another personally relevant example. I don't really care if my medical information is viewed by outsourced employees in india (heck I don't really care if it is posted on public bulletin boards in india) but I would be very disturbed if my medical history was availible on google for anyone who was interested.

So while desire for obscurity does hide potentially relevant facts from parties who might have some concerns (employers and medical records and dates and divorce records) it can still be availible for situations where it is truly important. While there might be some slight advantages to making this data easily accesible I think these are strongly outweighed by our societal preferences to keep different parts of our life distinct. That is people have a strong conviction that what sort of thing you do in bed at night, or what sort of medical treatments you might be recieving (so long as it doesn't interfere with job performance) shouldn't affect your position at your job. As someone else pointed out this desire is often rationally based on the idea that employers or aquantances might draw incorrect conclusions from emotionally biasing information (the same worries as with differential privacy plus the problem that scandalous information travels faster than exculpatory information, for instance the information that someone was addicted to opiates is likely to spread and be damaging but the information that it was a result of intense pain medication and the person has been a model citizen since are unlikely to spread. Plus the concern that character can be assasinated by one isolated event which is easy to relay but is only exaulted through a long pattern of avoiding temptation that is hard to relay). However, rational or not this strong preference exists.

Still one might seek to object that the advantages of seperating work and 'private' life can best be balanced against the advantages of better information by the free market. That is people who don't want their private lives to be held against them by their employers will only work for employers who don't make use of this information. Thus the free market will put the approriate price on the use of this sort of personal information.

This response is lacking for several reasons. First of all individuals often simply don't have enough information to make informed deciscions. Any employee will likely not know if his employer is looking at information from his credit report or other sources when making promotion or hiring deciscions. Preciscely because of the strong preference people have that their private life shouldn't affect their work it is in the employers strong interest to hide or cover up their use of personal information in these situations. In more personal situations we have the problem that people are often quite irrational about learning information about other people, even knowing that their friend or date might be very put out many people simply cannot resist looking at easily availible dirt about the person. The same can even happen in buisnesses where the religious convictions of the owner influence what sort of information they collect from employees (all blockbuster employees are drug tested because of the conservative christian owner with probably only negative economic consequences).

Furthermore the situation with obscurity is a classic prisoners dilema like paying taxes and thus warrants government intervention to prevent defection. While one might support the general idea that work life and private life ought to be distinct if offered a pay raise to reveal one non-embarasing facet of personal information the temptation to defect is quite high. That is even though you might be benefited by a general inclination of people to avoid companies that make use of invasive information if you are offered a raise to reveal your HIV status and know you are negative the incentive to defect is high. So while people in total are better off when companies observe a strict line between the personal and professional in most situations they have an individual incentive to abandon some obscurity/privacy. Thus like any tragedy of the commons the perfectly rational action of the free market can lead to a less advantageous result.

In short I think we ought to give real serious consideration to the idea of keeping certain information obscure but not private. The fairest way to do this is likely to require some kind of physical presence and buerocratic form filling to recieve many types of publicly availible information. I don't think we can count on the free market to come up with the right answer in this case since the danger is that information will generally become easily accesible and encourage buisnesses to adopt invasive hiring and promotion practices. Thus each individual has little benefit in expending money/effort to keep their individual information obscure (and in fact doing so when others are not may give people the impression you are hiding something) and may recieve benefits of convience and even monetary reward for allowing this information to be easily accesible but suffers when everyone does this.

Posner: "I believe that many people would be uncomfortable to learn that their medical history had been disclosed to people living in distant countries, people with whom the possessor of the medical history will never transact."

I think it's the opposite, as other commenters have also pointed out. Who cares if some guy in India who has no idea who I am looks at my medical records? We don't want our friends and neighbors to know that we saw a psychiatrist once.

When I was working at a medical company, I saw people's health records... how incredibly boring.

When I began practicing law in the early 1970s, I quickly learned that privacy was ephemeral. Even then, both criminal investigators for the government, private investigators, and well-funded or obstreperous lawyers in civil litigation could ascertain almost everything about an individual that the individual might not want anyone else to learn. Privacy was protected to some degree by the cost of obtaining such information, which required face-to-face visits, phone calls, subpoenas, and, since the 1960s, time consuming formalities for criminal investigators who wanted to monitor oral communications. Digitizing information on financial transactions and medical data, however, vastly increases the risk that such nformation will become available to the general public or to those who gain access to it for improper reasons -- as the recent reports on fraudulent access to databases maintained by Chocepoint and LEXIS-NEXIS demonstrates. In the case of medical information, our society has made a value judgment that the benefits of frank patient communication to health-care providers are so great that such communications should be encouraged even by making them privileged in prosecutions and other litigation with attendant costs of blocking truth-seeking in those proceedings. It seems to me that the threat to candid communications and even to consultations with health-care providers from the risks of broader and unauthorized exposure of digitized data to the public is the relevant factor to analyze.

Very interesting discussion! Paul Gowder has made a point pretty similar to the one I was about to post.

He notes that the kind of things we tend to believe are protected by privacy are the kind of things which people tend to overvalue or act irrationally about. For example, a man or woman who has HIV may wish to conceal this fact because of irrational fears and/or prejudice that disclosure may elicit. Now, certainly there are some situations (medical field) where that fact is important to disclose, yet in the vast majority of cases the relevance of the fact is far outweighed by the irrational and prejudicial way in which the information is treated.

If at oral argument the good Judge Posner was asked if he had ever had anal warts, would the Judge even answer such a question? I would hope he would not. Yet the judge seems to take a position which would make such information available--because, hey, we don't want to commit fraud--to the general public, one's employers, fellow employees, etc. Which begs the question, is the Judge who refuses to answer such a question committing fraud in the same way keeping those medical records private commits "fraud"?

The two competing interests, as I see it, are the relevance of the information and the the interest in reducing irrational discrimination or embarrassment. As the relevance of the information grows, the case for keeping records private diminishes. The second interest is in protecting individuals from irrational or prejudicial treatment. Why someone's STDs or their irritable bowel syndrome status should be available to employers, business partners, or even the general public is a concept which escapes me. The relevance of the information like this is very low, yet the likelyhood the information will unduly embarrass, damage, or stigmatize the individual is high. Privacy makes a whole lot of sense in most cases, Judge Posner.

Though in your conclusion you seem to concede some of this with your "materiality" qualfication, I think it is important to remember that because people are often irrational or prejudicial, what is "material" to them is in fact immaterial to the business transaction at hand from an economic standpoint.

"It seems to me that the threat to candid communications and even to consultations with health-care providers from the risks of broader and unauthorized exposure of digitized data to the public is the relevant factor to analyze."

Excellent point!!!

While I agree with Palooka that many things we care about in respect to privacy are those things people are inclined to act irrationally about, in some sense, I'm not sure if this is sufficent to defend government sponsored privacy rights or properly deliminate those rights. In particular I am worried that we are confusing two distinct and very different notions of irrational. There is the first sense of irrational we use in everyday language refering to feelings which we judge to be reasonable. This is the sense in which one can have an irrational fear of flight. We judge a fear of flight to be irrational because we don't feel it is reasonable to have a primitive fear of being in flight and the chance of death is quite small. However, we don't call a fear of being seen naked in public in a foreign city with no nudity laws irrational even though there may be no more risk of a bad outcome than when flying.

Alternatively we have the technical notion of irrational that we use in economics. That is an action is irrational if it does not serve to maximize the actors ends no matter how silly those ends might be. Thus in the economic sense of rationality it is perfectly rational to pay an extra 50 dollars to take the train rather than to fly since your 'irrational' fear of flying will cause you 50 dollars worth of anguish if you took the plane. Since the government is often rightly in the position of choosing not to inflict unnecessery worry on the public even if that worry might be irrational (in the first sense) it seems only this second sense of being irrational is relevant to government action.

Having made this distinction the quesiton now becomes whether the knowledge of private information really makes people behave in economicly irrational ways or just irrational in the first sense. That is it is not sufficent merely to note that people are scared of those with HIV even when there is no chance they won't get it. Even if we don't think this fear is reasonable it still is a real fear and these people may be truly happier knowing certain of their friends are HIV free.

In short observing that people act 'unreasonably' in response to certain knowledge is not sufficent to establish that their possesion of this knowledge doesn't increase total utility. Conversely to establish that this information made people act irrationally in the economic sense is just to say that this information tends to make whoever knows it act so as to achieve less of their ends. Yet to establish this we would already have to have established that knowing this information didn't provide more benefit than harm which is just begging the question.

Basically, I'm worried that this argument relies on a subtle confusion of the two notions of irrationality. However, there does seem to be something there so maybe I am just missing it.

Furthermore, I have to object to the analogy between lack of privacy for personal information and being asked embarassing personal questions. While these subjects might also be embarassing for others to know the answers to this is entierly distinct from the reason we are bothered by people asking the quesitons in professional situations.

The reason why asking a judge if he has anal warts is upsetting is because the question is rude and implies a disrespect of the person involved not because we feel strongly about the privacy of the information. After all such questions are unpleasent even when they are not answered so it can't be the revelation of the information which is bothering us. Moreover, often we are totally unperturbed if that same information is publicly accesible. For instance Judge Posner would likely be quite offended if an attorney asked him in court if he had anal warts even if he wrote publicly on his blog that he did or did not. Or more personally I have no problem admitting here or in other public forums that I don't have HIV but I would be quite disturbed if one of my students asked me this in section. In short one needs to be careful about distingushing between the impacts of the information being public and accesible in some manner and the social implications of directly asking about that information. (As an aside I think this is what the protestor who asked Scalia if he sodomized his wife was missing)

"Furthermore, I have to object to the analogy between lack of privacy for personal information and being asked embarassing personal questions."

I agree that such a question represents a great deal of disrepect. But WHY does it do so? Because we have expecations that certain subjects are "off limits" and certain information is for our own use only. It's rude and disrespectful because it's NOBODY'S BUSINESS!

"Or more personally I have no problem admitting here or in other public forums that I don't have HIV but I would be quite disturbed if one of my students asked me this in section."

I think you're confusing something here. When one chooses to unveil one's privacy on an issue, then of course they would not be offended! A voluntary disclosure is not analogous to mandatory disclosure to the public. Sure, the kind of forum makes some difference. People don't like to be humiliated or embarrassed in person, but neither do they want their neighbors or employers gossiping about their conditions behind their backs (or worse, acting on that information in a discriminatory manner).

It is true that many personal or embarrassing medical details may be relevant to an employment or contractual decision. A more tailored solution would be to provide confirmation for relevant medical facts rather than make medical information widely availabe. I by no means think privacy is absolute, but I find the level of transparency implicit in Posner's analysis troubling and deeply flawed. He doesn't seem to analyze what positive effects--for example, limiting irrational or prudish discrimination--privacy produces.

LogicNazi,

One more thing: I agree that pyschic disutility can be included in economic analysis. The problem is, in place of increased "pyschic utility" on the part of the discriminator we get many costs, which in my estimation greatly exceed the gains. The increased utility somebody gets from discriminating against X group is unlikely to exceed the loss suffered because of discrimination (decreased income, opportunity lost for group X, disutility for group X, AND efficiency suffers, reducing income of the discriminator as well).

I think there are several aspects of privacy beyond those contemplated so far. To name one case, I'm amazed that the implications of attorney-client privilege have not come up.

To name other examples, people commonly desire privacy in the details of their contemplation and discussion of important decisions. Should I move? Should I have a child? Should I take that job? This makes sense from two possible lines of quasi-economic analysis: it's a bargaining strategy and it preserves the value and quality of news about oneself. If we announced our every idle whim, our friends and acquaintances would soon be overwhelmed with tiny, frequently incorrect details about us, which they would no doubt learn to ignore as a nuisance. If we reserve discussion of our intentions until we have made up our minds, we show due respect for the attention and interest of our friends and colleagues. In this case, privacy is used to establish a sort of quality-control over information.

In a similar manner, some of our ideas are not so much embarrassing as they are unfinished. Privacy in both these cases affords us the opportunity to produce a finished product.

Then there is the use of privacy to establish a sort of fund of the personality. Many of us cherish ideas, dreams, preferences, intimacies, opinions, favorite entertainments, etc. that we discuss with others once we feel we have met a kindred spirit. There's nothing embarrassing or discreditable about this information, but sharing and discussing such things can form part of building a friendship.

LogicNazi: I'm inclined to reject your second sense of irrationality for any number of reasons discussed all over the place that basically boil down to "not everything someone wants is good." Your "economic irrationality" defines the concept of irrationality out of existence, since people will do that which brings them pleasure and everything that's pleasurable is rational under that definition. That way is nihilism.

Plus psychic disutility is very difficult to measure and value against all the other harms that Palooka rightly pointed out.

However, I'll assume it arguendo, because I have something more interesting to say:

Here's the thing: the virtue of privacy is that it can avoid this psychic disutility. I'm going to behave exactly like an economist here, even though (as noted above), I reject the principles I'm about to apply. Lets see if I can't defend privacy even accepting them arguendo.

Assume (god, I love playing economics) that a person is irrationally (in your first sense) afraid of HIV-positive persons to the point where they lose utility valued at $5,000 if someone working for them is HIV-positive.

Assume that the disutility of someone denied a job because they're HIV-positive can be valued at $4,500. (Which, I suppose, requires us to assume that the employer is REALLY bigoted, or the employee has a bunch of people beating down his/her door, but what the hey, it's economics.)

Now assume that the disutility of the employer who doesn't actually know that any employee is HIV-positive, but instead is simply deprived of information on which to base his fear, is $2,000. This makes some sense: it should be scarier to know a phobia is present than it is to merely be ignorant of whether it's present or not. People who are afraid of snakes can still walk in tall grass.

In a no-privacy regime, when that employee applies for a job with that employer, $4,500 of utility is lost (assuming away antidiscrimination laws). The employee won't be willing to pay the employer to hire him because he values the job less than the employer values the coddling of the phobia, so he suffers that loss.

On the other hand, in a privacy regime, only $2,000 of utility is lost, because the

Obviously, since I just made up the numbers (in the finest economic tradition as demonstrated by Posner's first post on this very blog), this doesn't prove that privacy is in FACT consistent with utility maximization. However, it DOES prove that secrecy of overweighted discreditable facts CAN BE consistent with utility-maximization even accepting your "economic definition of irrationality."

Expressing this thought in real terms: the emotional cost of silly fears can be avoided with privacy because the people who would otherwise experience the silly fears don't know that there's anything there to fear.

Woo-hoo! Depriving economic actors of information increases their net gain! (I can't wait to see Judge Posner's response to this point, should be interesting.)

Sorry, I forgot to finish a sentence in there. (Now you see how I write...) In the privacy regime, only 2k of utility is lost because the... employee doesn't lose the job and the employer doesn't fear his exposure to a HIV-positive person. The only loss is the general suspicion of the employer that someone in his employ MIGHT be HIV-positive, which we valued at 2k.

Posner argues that we have an obligation, in some transactions, to surrender privacy, to reveal facts we would prefer to keep secret.

But this seems to be a one-way street. Take the case of an HIV-positive job applicant. If the applicant is somehow ethically required to reveal his condition, why isn't the employer required to reveal its position in regard to hiring HIV-positive workers? Isn't it just as deceptive to keep this policy secret?

Might that cause the company embarassment? Well, the policy is either defensible or it isn't. If it's not then maybe the company should be embarassed inot changing it.

Why isn't it a requirement that the health insurance company publicize its underwriting rules? That way some applicants would be spared the trouble of applyng for insurance they are not going to get, or that will cost too much. Others will be able to select an insurance company that has the best deal for them. Does this work to the disadvantage of insurance companies? Yes, but that doesn't seem to be a concern in the case of individuals. Why should the companies be any different?

judge,

your concern is concealment of "facts that would be material to the willingness of other persons to transact with the concealer on terms favorable to him."

i agree, but i am wary for the following reason: it may be the case that persons, if given full information regarding prospective transacting persons, would improperly process the information. embarrassing facts from someone's past may be enough to not hire that person for a job or not date that person. but if you view the individual as a series of individuals over a life span, as i do (like Nagel), then it seems clear that past poor qualities or judgment may very well be a poor proxy for current qualification.

because i think people would put to much emphasis on information from one's past, privacy can help to control the amount of irrelevant information an information-seeker receives.

there are responses, of course. privacy also protects current, relevant information. further, if everyone lacked privacy, then perhaps people would give less weight to embarrassing facts from the past because it would be understood that the past, in many instances, is a poor predictor of conduct.

thanks.

Bernard hit the nail on the head.

Sad to say, but I think real human beings should have at least the same rights to privacy as ficticious human beings (corporations).

Judge Posner is worried that real humans might hide relevant information that could affect economic transactions. Ficticious human beings routinely hide relevant information up to the point of endangering the health of buyers (at which there is at least some redress but not much given the disparity of information as well as financial and political power).

Lets bring the privacy rights of real people into line with the rights of artifical ones. That is a good start in my opinion.

Some very interesting thoughts raised by the HIV+ job applicant.

If that information is made public, couldn't that have, in the long run, some very positive effects?

In the beginning, the individual is not hired by an employer because of irrational fears (or, perhaps, a rational and ever increasing desire to keep health care costs down in the workplace - see below). There is disutility here, but it also creates value for rational actors.

The next employer who evaluates this candidate in a presumably rational way will recognize the good deal that they are getting - the market value of the applicant's skills has been depressed because of the irrational fears, but not the actual value of the skills, so the applicant may get a job at less than their skills' actual worth, and the employer gets a better value than a rational market valuation would have given him. Thus, he has a market advantage over the irrational actor, his competitor.

The competitor can either face the facts and recognize that he has acted irrationally, and thus discontinue his fear-driven hiring policies, or can maintain them to his own detriment.

I would expect that, eventually, enough employers would see their irrationality (when faced with the competitive disadvantage that it puts them in), and change their ways, thus restoring the HIV+ job applicant's skills to a more rational market valuation.

However, in a scheme where such information is kept private, irrational actors are never confronted with their irrationality, and thus they maintain it.

But, this of course falls apart if health care costs, lost productivity from illness, or other factors do, in fact, make an applicant with HIV (or any other illness, or, arguably unhealthy habit e.g. smoking) less valuable to an employer. In such a case, it is a value judgment on society's part whether we want that person to suffer the detriment of the valuation, or if that information should be kept private, and thus let the loss fall on potential employers.

�But once records are digitized rather than existing solely in hard copies in the office of the patient�s primary physician, the risk that unauthorized persons will gain access to them is increased.�



This assumption that digitization automatically creates increased risk instead of decreased risk is of serious doubt.



The UK is currently building out a digitized medical record system that will keep track of everyone who views a medical record. Keeping an access log like this can reduce the risk someone uninvolved in the care of a patient will view without punishment the records of celebrities, politicians or even neighbors.



Similarly, the Department of Defense has had great success in the creation of its secure computer network, largely in part because the network uses computers that are only attached to the secure network and no part of the network is connected to any outside system or the internet. In offices with computers on the secure network you will see two computers one for secure work and one insecure work.



It seems like paper records can be viewed almost at the will of anyone working in a medical facility without record of copies being made or who has accessed a file and when. Digitized records reputation has been tarnished by a few high profile lapses in security that is not indicative of the real promise of the technology.

There is something to the nudity taboo, as this might perhaps be the psychological basis for privacy instincts. You ignore for the most part the psychological effect of 'unveiling' certain aspects of a persons life. The immediate example would be rape victims who for obvious reasons would not want people to know they have been raped. Though there might be elements of shame involved, many recognize that by definition the victims did not have the agency to prevent it. The victims would perhaps want to avoid special treatment by everyone around them, which would be a constant reminder of the rape. How is that different from an alcoholic who wishes that his problem drinking not be known by his co-workers, employer, etc. ? Or say a diabetic? I had a diabetic friend when I was younger who really hated when other parents treated him differently when he came over, preparing for him special foods and such, even though he was in a far better position to recognize his own blood sugar level anyway.

I am not sure I can interact in a society where everything about me is easily accessable. Criminal or not, you'd worry about being identified as a fraud, especially in academic/professional circles. I'd like to have my doubts over my abilities and knowledge as private, because if they were out in the open I wouldn't have the confidence to speak out. e.g. "How can I talk about ethics, if everyone knows I was caught stealing candy in elementary school?" Perhaps Augustine's way is best, admit you were a sinner, and somehow package it that you are all the more better for recognizing it.

Interesting post, but I think Judge Posner conflates a number of concepts that are analytically distinct. I wish I had more time to refine my thoughts, but I'll give it a quick shot.

First, "privacy" in a legal sense has nothing to do with the "right" to not be seen nude in public. That is simply a social convention that people follow by wearing clothes and that society facilitates by providing "private" areas where people can change clothes, etc. This social convention is just a custom -- a taboo, perhaps -- far removed from the "privacy" debate about medical records or, say, abortion. It's more akin to the social convention that one wears a suit and tie to court. Indecent exposure laws create a "right" not to see others nude in public, but I think that just demonstrates the depth of the social convention; it doesn't create a "privacy" right.

"Privacy" in the 4th or 5th Amendment sense is a limitation on state power: the state cannot unreasonably spy on its citizens. The purpose is to create a zone of liberty where "big brother" cannot penetrate. There is no doubt that the 4th and 5th Amendments contemplate such a zone of "privacy," though the outer edges of that zone (e.g., abortion rights) are a matter of great debate.

"Privacy" in the sense of keeping medical records (or other personal records) confidential is a positive right created by legislation that bars private actors from passing on certain information. It is more akin to trade secrets or state secrets than to 4th or 5th Amendment rights. The idea is that personal dignity is diminished if strangers know certain very personal details. Such rules expand the zone of privacy beyond the body to the realm of facts and knowledge.

These various forms of "privacy" have different legal roots and grow out of different social conventions. We should take heed of the underlying reasons for the privacy concerns when deciding where to draw the line. I would err on the side of greater privacy when the concern is infringement by the state, and I would be less inclined to recognize a "privacy" right that prevents private actors from sharing information. However, the prevelance of identity fraud and other "information" crimes these days, not to mention email "spam" and telemarketing, might indicate a need for greater "information" privacy.

As to concealing bad qualities from a dating partner, that should remain well beyond the realm of law. Of course, as Judge Posner notes, concealment is doomed to fail in the long run.

David, you do realize "privacy rights" in the context of the Court's jurprudence on abortion or sodomy or sexual contraceptives is nothing more than a rhetorical guise--the Court has chosen the word because of its positive connotations. That's why, I think, Posner explicitly said he was NOT addressing those issues. Privacy as embodied in our Constitution has nothing to do with creating a "zone" or "penumbra" of behavior which is unacceptable for government regulation. It has to do with reasonable adminstration of justice, probable cause, and the like. If the Framers wished to recognize a right to something as inexplicable as "privacy" and "private behavior" then I am certain they would have done so. I suspect Posner didn't want this turn into a debate about those kind of "privacy rights." So I'll leave it at that.

Many of the above posts have made useful critiques of whether privacy really promotes efficiency. Several of those critiques raise the issue of privacy hiding either past non-predictive behavior or predjudicial behavior.

Recent psychological research suggests that contrary to the purely efficient market hypotheses, the human brain has evolved to put greater weight on past patterns then pure rationality would suggest. Furthermore, the human mind has also evolved so that prejudices are a barrier to rational behavior, even in decision making. Evolution favored animals that chase away or kill those outside their group!

Finally, even if the human mind were efficient, we might use privacy rights as a form of social insurance. Thus it might be efficient for an employer to avoid hiring people with a past history of cancer as screening for cancer efficient way of weeding out potential laggards (e.g. if 1/3 are likely to relapse).

However, as a society we don't want to deprive cancer patients of the opportunties that people with perfect health histories have, just because some of them will become too sick to perform on the job. We thus disperse the risk from the minority of individuals with cancer through employers and insurance companies. Just like I purchase insurance at a net negative present value, society may decide to protect someone's history of cancer treatment so that those individuals costs are someone relieved.

Chris Willis,

But consider the alternative to your not-so-lovely scenario.

HIV status is secret. HIV-positive workers are paid the market wage rather than being forced to accept below-market pay.

The irrational employer stays irrational, but this has no effect, in this case at least, so the company loses no competitive advantage. So the many other people who work for the irrational employer do not suffer by having the company go out of business, or otherwise perform poorly.

What was that about positive effects?

There is, "no presumption of privacy." There is an "absolute presumption of privacy." Either-or, the devil and the deep blue sea. How can two such mutually exclusive ideas co-exist in the world and in the affirs of man? Quite simply, life is a box full of dilemmas, contradictions, and paradoxes. And as informal logic has shown, there are two methods to deal with such dilemmas. One is to break one of the horns of the dilemma and deal with the other. The second is to escape through the horns.

Let's try and escape through the horns. But how?
By recognizing that "privacy" and it's issues are "situational" in nature and not categorical imperatives; nor are they a part of the "either-or" scenario outlined above. By accepting such a dodge it allows the use of either horn depending on the situation. Take for example, the case of a band of sabotuers and cut-throats operating at will in the country because we have adhered to the principle of "absolute presumption of privacy." Remember 9-11? I don't think we want that to happen again. In this case it would have been better to apply the principle of, "no presumption of privacy" and acted accordingly.
Just as in the case of a typical family going about it's normal affairs. We certainly don't want to apply the principle of "no presumption of privacy." In this case I think we can all agree that the principle of "absolute presumption of privacy" should prevail.

The question now becomes, who is to make the decision as to which principle to apply in each given situation? Customs and conventions that have been developed over a long period of time have established a method and all we need do is to stand on them. In terms of the questions of privacy, judges and courts have made these decisions in times past and a whole body of law and procedures has been developed in response.

Even in this age of international terrorism and special organizations to deal with it, judges and special tribunals have already been set up to deal with the issues arising out of the privacy paradox. Let's just hope we don't recreate the episodes of the Court of Star Chamber.

Hmmm... I know some federal judges who are making a strong effort to get their home addresses out of the public domain. Goodness, I wonder what illegitimate motives they could have to do do such a thing. Embarassment? Discredit? One can only speculate.

Some responses,

First in response to Paul Gowder I was just explicating two different notions that people use the word rationality to mean. The notion I called economic rationality is meant to capture the idea that someone who is rational is one who best optimally pursues his ends whatever those ends might be. The second notion, informal rationality if you will, is that one is not only (more or less) economicly rational but also has 'reasonable' ends. Surely you are not denying that economic rationality is a meaningful concept or that the word is sometimes used in this fasion (consider someone talking about a white supremicist rationally acting to eliminate blacks, the end is clearly unreasonable but we still might use the word rational to describe them if they pursued this end in a very efficent manner).

I simply don't understand what you mean when you say you reject the notion of economic rationality. Do you mean that people never use the word in this manner or something else? How can it be that using the word rationality to refer to economic rationality leads to nihlism? Is there some important moral rule that tells us the letters 'r', 'a' and so forth put together must refer to informal rationality? I certainly never meant to imply that rationality was a sufficent condition to be moral (it may not even be necessery) so I fail to see how this could lead to nihlism. Just as one might decide that being honest doesn't preclude you from being a murderer yet not think murder was okay so too can you define economic rationality to include white supremicists and others with unreasonable goals without condoning their behavior.

Perhaps though I wasn't clear enough in my original post. I do think people are better off with certain information protected by government regulation keeping it either private or obscure. I merely meant to say that I don't think we can use the notion of whether people react rationally to the information to either pick out those situations that ought to be protected by privacy laws or to defend those laws. In particular I am alleging that a rule that said we should protect people's privacy when others would be likely to react irrationally to the information would not be of much use.

Now we have two choices. If the rule is supposed to say that we should protect information when it makes people act economically irrationally this is much too strong a burden to meet as you correctly point out. It seems nigh impossible to show that some information made some act irrationally in this sense and didn't just have some bizarre strong preference. Furthermore (and perhaps this is what you were alluding to with your nihlism comment) it would fail to protect many situations we think deserve protection.

So what about the other case where we test information to see if it will make people act irrationally in the informal sense? This unfortunatly seems to be almost no guide at all. The notion of a reasonable end seems so vague as to be useless. For instance consider an employer who has mounted cameras in the restroom to make sure employees aren't using drugs in the restroom or discover who was defacing the bathroom. Should he have a right of privacy to hide the existance of these cameras because not wanting to be seen nude is an unreasonable end? After all being seen nude has a lower risk of harm than touching someone with HIV and we seem to want to call the desire not to be around HIV+ individuals irrational. If you don't believe employers deserve privacy rights just rephrase the situation to involve individuals (someone who put cameras in his bathroom to make sure his kid didn't fall down). Does someone who runs a daycare in a house built on low level radioactive tiles (some tiles used to be made with uranium paint) have a privacy right to keep that information secret as people react irrationally about radioactivity?

Conversly consider someone who wants to keep their terminal illness private. It seems that any reasonable privacy protections would respect this right but it doesn't seem like there are an irrational reactions to this information. Even better consider the privacy of communications between husband and wife. I think we rightly protect such communications in the law in not making spouses testify against each other about such communications. However, here the concern is preciscely that the DA will use the information perfectly rationally to prosecute the individuals.

In summary I don't think the notion of irrational response can be used to deliminate privacy or make a case for it. Either the notion is much to weak as in economic rationality or even more vague than the idea we are trying to analyze. I think we have a better sense of what sort of information ought to be private that we do about what ends are reasonable and hence rational. Is it rational to not want to be seen naked, not to be around radioactive substances, stop people from using drugs, to not want to walk home at night in a neighborhood that just had a murder even though statistically it is still safer than driving? In short this notion of rationality really just reflects our social value judgements not some objectively perceptable difference between the ends under consideration. Yet if we are going to do this why not just directly make these judgements about what is private information and what isn't which seems more direct and clear than this roundabout fashion.

In particular I think the compelling reason why rationality cannot be a guide is the following. We certainly do not think that a cook in a restaurant has a right to privacy about putting pork in the food or whether he allows a dog in the kitchen while cooking even though people react strongly to these things because of arbitrary religious prohibitions. However, if these arbitrary religious prohibitions are considered rational for the purposes of privacy it seems we must also deny the cook a right of privacy about whether she is menstrating as some religions find this to be unclean.

So yes I agree that certain information may in fact make everyone happier if it is kept quite like your HIV example. However, there isn't any notion of rationality which can provide us a guide to determining which information deserves protection. We have to actually look at all the effects and psychological reactioins and see on which side the balance falls including questions of whether lack of privacy would deter honest communication (reason for privacy of doctor patient relations). I do think that our intuitive reactions that some things should be private and others not are helpful but I don't really think our intuitions about what is rational and what isn't are very helpful. Some things which are completely irrational like not wanting to be seen naked are still worthwhile since they don't hurt anyone and correcting the irrationality would just cause suffering.

Finally to Palooka I agree that some of the reason one might be upset about your particular question is a sense that the information should be private but I don't think this is the full reason such a question would be disturbing. I mean imagine two situations one in which an employer accidently posts the medical information of all the employees and another where you are asked by another employee whether you have anal warts. I at least would be much more disturbed and offended by the second than the first. So while I agree finding the question invasive is a guide to whether we view the information as private I think it is incorrect to assume that stripping privacy protections is just as bad as being publicly asked. Often with this sort of information we find it embarassing and the actual act of being questioned in public can be much more unpleasant than someone just knowing.

Logicnazi:
I simply don't understand what you mean when you say you reject the notion of economic rationality. Do you mean that people never use the word in this manner or something else? How can it be that using the word rationality to refer to economic rationality leads to nihlism? Is there some important moral rule that tells us the letters 'r', 'a' and so forth put together must refer to informal rationality? I certainly never meant to imply that rationality was a sufficent condition to be moral (it may not even be necessery) so I fail to see how this could lead to nihlism. Just as one might decide that being honest doesn't preclude you from being a murderer yet not think murder was okay so too can you define economic rationality to include white supremicists and others with unreasonable goals without condoning their behavior.

I think I was unclear. I reject the notion not because it's incoherent, but because we're not just talking about theoretical "economic rationality," we're talking about how policy should be made.

Let me rephrase. I reject the notion that policy can be made on the basis of "economic rationality." Because if we make policy on the basis of assuming that our goal is to, as it were, remove "transaction costs" and treat all desires as "utility" deserving of respect in accorance with the value the desirer places on it, we wipe out centuries of ethics in social policy. That's nihilism.

Nonetheless, your criticism of the application of rationality in your more recent comment are well-taken: in particular, it's hard to answer your analogy between forbidding a chef to keep private the presence of a dog in the kitchen (concern about which, assuming the dog doesn't get into the food, is irrational under my definition, albeit rational under yours) and forbidding a chef to keep private their mensturation (concern about which would also be irrational under my definition and rational under yours). Because, unfortunately for my own defense of that principle, I agree that chefs shouldn't be allowed to keep dogs in the kitchen privately. (Perhaps, however, this is merely a measure of my own human irrationality.)

As much as it pains me to do so, then, I'm going to partially concede my previous rationality position to you. There are some human irrationalities which ought to be coddled (i.e. the religious refusal to consume pork) and others which should not (the hysterical refusal to associate with HIV+ people) and I don't have a principled distinction between them.

However, I maintain in opposition to Posner's point that there are some private facts that may be "discreditable," such as HIV status (or FDR's ailments, or JFK's ailments, or Clinton's fellatio, or 5-year old negative credit items under wholly different circumstances, or 20-year old criminal convictions) that should be kept private. I further maintain that one way to describe the reason that those facts should be kept private is because the persons who learn them will likely them to punish the person about whom the facts pertain in a far greater degree than is warranted by the actual personal risks to the punisher that the facts in question indicate. Perhaps we can use a different word to describe that, since irrationality isn't working. "Excessive risk-aversion" might be a good phrase.

The problem with excessive risk-aversion as applied to choosing to enter transactions with other humans, is that if such risk-aversion is truly excessive, it causes a massive culumative harm on the person at issue.

Consider, for example, the hypothetical case of one lower-middle class person rejected by 100 wealthy corporate employers because of his 20 year old conviction for stealing a pizza. This is not improbable: assuming two candidates are fairly close, almost any employer will likely (common-sense here) reject the one with the old conviction.

Once again, I'll play economist and make up some numbers (this is fun!). Each employer presumably takes a slightly less enriched pool of candidates than if they hadn't "irrationally" or "excessively risk-aversely" rejected the pizza-thief. Lets say their loss is valued at $100.00 each. Assume further that the risk-averse manager at each corporation incorrectly views the probability of a theft from such person as .05 (it's really .001), and the potential loss from such a theft as $1,000.00 -- leading to the incorrectly expected loss from hiring such a person at $200.00. Each employer then saves a net $100.00 from not hiring the pizza-thief, and so each employer will make an "economically rational" decision (since their misestimation of the risk is in accordance with their risk-averseness, which is psychic disutility) to not hire the person.

HOWEVER, the loss to a person who is thus rendered unemployable is greater than the loss to one employer, so lets say that loss is $2,000.00. Plus there's some marginal additional harm from each rejection (wasted effort, misery, etc.).

Thus society as a whole is put in this weird nonlinear position that to me seems to defy the basic economic assumptions we're working with. If any one employer had simply hired the guy, he would have taken an expected $200 hit instead of an expected $100 hit (although he really would have profited under the assumptions above since his expected loss was based on an overly adverse estimate of the risks), but the pizza thief would have gained $2000+, so society would have had a net gain if he'd been hired. (Do we have a game theory expert in the house??)

And don't just pull a Coase on me and say that the employee could have paid the employer (or posted a bond or something) to compensate him for those risks, since getting a job is worth more to the employee than not hiring him is to the employer. That's why I used the example of an unemployed, convict. He can't afford to pay the employer, and he certainly can't get a loan with no job and a criminal record. One of the big, huge, MASSIVE flaws with the Coase theorem is that it assumes unlimited wealth is available to everyone.

Far better, it would seem to me, to just expunge the criminal record and thus deny the employers even the opportunity to cause that net social economic loss by virtue of their misestimation of the risks inherent in hiring someone with a 20 year old conviction for stealing pizza.

Do I have a principled distinction between the pizza stealing/menstration and the dog in the kitchen? Alas...

Addendum: the other point that I was thinking about (and forgot to write in my speed-typing-between-work-tasks) in the previous comment's hypothetical is that the employers also collectively lose in that situation.

Each employer, after all, lost $100. Now, admittedly, each employer did so in order to avoid a $200 loss to themselves, so under an individual reckoning they gained $100 net. However, only one employer need have taken that $200 loss from hiring him. Thus, among the hundred employers:

NOT HIRING:
100 employers lose $100
Total loss = $10,000.

HIRING:
1 employer loses $200
99 employers lose $0*
Total loss = $200

Thus even if we fail to take into account the loss of the employee, the employers have a collective action problem: each individually makes an "economically rational" decision to not hire, but that decision leads to a net social loss.

I submit to you that this is the case for all forms of discrimination that cause an employer to reduce their applicant pool: even if you give them credit for "economic rationality," there's a net social loss that needs to remedied by law, and one of the laws that can do it is allowing people to keep "discreditable" (a.k.a. "discriminatable") facts private.

------
* The reason the other 99 employers don't suffer the same "less enriched applicant pool" that was the source of their loss in the non-hiring scenario is that I'm assuming that the employee gets an interview before he discloses the conviction, or otherwise uses some of the employer's limited screening resources. Thus, if the employer could consider 20 people with him in the applicant pool, their rejection of him reduces their applicant pool to 19. On the other hand, if one hires him, he's not applying to all the other employers, who can then fill that slot in the applicant pool they can afford with another person, giving them the full opportunity to find the best candidate and avoiding the $100 of applicant pool loss.

To my view, privacy rights between equally situated individuals are ideally irrelevant. Disclosure only conveys a bargaining advantage to other persons if the information disclosed is something criminal, embarassing, or socially stigmatized. People can avoid the harmful effects by not doing anything criminal unless they are willing to accept the consequences (civil disobedience 101), by growing up about bodily functions, and by working to counter unfair social stereotypes.

If you are not a criminal, have rejected puritanical American repressions, and are committed to outing bigots and profilers rather than working for them, you can't be hurt by disclosure. Privacy between individuals can be limited to things like PIN numbers and passwords that protect real/personal property rights.

I think the main justification for privacy rights derives from the potential for Orwellian government use of personal information. The government wants information about you because they want to individualize their mechanisms of control. They want to maximize their tax revenue, predict your voting patterns, and preempt any revolutionary impulse you might have.

There hasn't been a lot of talk here yet about the PATRIOT act, Arab profiling, and airport security. Yet the worst impositions on all of our privacy rights over the last 5 years have been right there. Everyone say hello to CARNIVORE in your next email.

The government thinks most Americans are willing to give up their 4th and 5th Amendment rights to some degree because the people are scared of Arabs. You all may actually feel like you have nothing to hide and so it is OK for the NSA to read your email along with the evil bearded terrorists you imagine are out there. The people who want to push back against government incursions into personal privacy risk being characterized as an apologist for the terrorists.

History shows that this is exactly the climate for the erosion of individual rights. A populace that is afraid, increasing economic disparity, an "enemy" that is reportedly among us... and a slow silent slide into facism.

For these reasons I think conflating privacy between individuals and privacy vs. the government is a bad idea. (Or if you prefer, privacy b/t market actors and privacy in unequal power relationships) It is certainly possible to want more of the latter and less of the former. Lumping them together under the global term "privacy" limits the discussion.

However, I maintain in opposition to Posner's point that there are some private facts that may be "discreditable," such as HIV status (or FDR's ailments, or JFK's ailments, or Clinton's fellatio, or 5-year old negative credit items under wholly different circumstances, or 20-year old criminal convictions) that should be kept private.

=======

Though we have been in much agreement on this topic, I do not think JFK's or FDR's ailments should have been kept private. The people have a right to know if their president is healthy because the president's health is vital to the job he is conducting. Where a medical condition is immaterial to his performing his duties, his records should be kept private in the same way as the general public. From an economic standpoint, the harm to the people, because of the importance of the job of the presidency, is much greater than the harm done to a given presdidential candidate's ambitions. We have an asymetry here--one man's ambitions versus the well being of the country. This asymetry is not present in the common case under discussion in this thread.

It's wrong to conflate these two situations with the subject of medical privacy broadly for several additional reasons. First, the "job" of being POTUS is an exceptional one. It is mentally and physically exhausting for even a man in excellent health. The American people have a right to know if a candidate suffers from a severe physical or mental impairment because it is absolutely material. Now, this doesn't mean that it was irrational to elect JFK or FDR, it just means that their physical conditions are one of many factors which the public should weigh. Given two similar candidates, one should choose the healthier one. Second, it is widely recognized today that as a public servant one surrenders a great deal of their "privacy." Moreover, in the case of JFK or FDR you are suggesting that they have MORE, not the same, privacy of the average citizen. Would it be conceivable that FDR conceal his impairment to an employer in private practice? Would it be conceivable that JFK could have done the same? Of course not.

Lastly, let me deal with your drivel about Clinton. He had the same privacy that any other citizen would have if they were being sued for sexual harassment. It is absolutely relevant to a sexual harassment suit to investigate his pattern of sexual activity with subordinates. His privacy was not breached anymore than YOUR privacy would be breached if you were being sued for sexual harassment, and had embarrassing details of sexual relations with subordinates come to light. The appropriate question, however, is if the office of the presdidency should be burdened with any kind of civil suits, even if credible.

"Finally to Palooka I agree that some of the reason one might be upset about your particular question is a sense that the information should be private but I don't think this is the full reason such a question would be disturbing. I mean imagine two situations one in which an employer accidently posts the medical information of all the employees and another where you are asked by another employee whether you have anal warts."

Yes, there are differences, but the underlying principle of privacy is present throughout. An additional grievance to such a question is that it implies that there is a chance the individual has anal warts. If I asked you in front of your students if you had HIV, I think part of the harm done is that some students may think you do, even if you answer that you do not. It is especially true if you refuse to answer (the principled response to a rude question).

But I believe that the main reason these subjects are considered rude and inappropriate is because this kind of information is considered "private." It is for one's use only, unless they personally choose to disclose it. If the question of whether Judge Posner has anal warts or not is immaterial, rude, and completely inappropriate during oral argument, then I think it is immaterial, rude, and inappropriate to disclose this information to the public in any forum (even if some fora are more damaging than others).

Palooka:

That's as may be, but I really only meant the FDR thing at least (I'm not sure about JFK) as s good example of the risks from overweighting (being excessively risk averse). Notwithstanding his polio, I think almost everyone, no matter their political perspective, agrees that he was one of a handful of the U.S.'s greatest presidents. If he hadn't been elected because the people overweighted his polio, well, some of the more extreme possibilities if a lesser president had been elected are rather dramatic. How's your German?

Perhaps it was a bad example, since I suppose we have to trust the people to give due weight to those facts about presidential candidates, as unlikely as I think that is.

(I'm not going to address the whole Clinton thing and my "drivel" for the same reason I didn't address your abortion point earlier: in your correct words: "I suspect Posner didn't want this turn into a debate about those kind of 'privacy rights.'" However, suffice it to say that I'm certainly not agreeing by silence...)

Bernard:

I think there is a positive benefit any time people don't feel that they have to keep something secret about themselves that they don't want to conceal, and I don't think that people should have to hide those things from irrational actors to maintain the market value of their skills.

Since reading Judge Posner's excellent essay on Huxley's Brave New World and Orwell's 1984 (not complete yet, I don't think), I am curious to know, if Judge Posner has any sociological indicators to watch out for, least our society starts (we have already started in my opinion)to mass produce psychologies of infantilism (the ultimate societal damage according to Jeremy Bentham's Penopticon)from a complete loss of "secrecy and solitude" -- at what level of nihilism do we attempt a true analysis of privacy lose causation. This is another situation were libertarianism truly clashes with paleo-conservative values of guarding against de-humanization at all levels. I think it is a difficult argument to make that personal information privacy should be viewed solely or largely through an economic lense in the Information Age without granting a property right to the producer of the personal information. Got to go...I off to a "feelie" -- otherwise known as America's Pub in Kansas City, Missouri -- true "Ecstasy" awaits me at the door.

Chris,

I agree, but you're describing an ideal world. The fact is that sometimes concealment is necessary, because there are irrational actors in the world. And of course some people simply prefer to keep things to themselves.

First to Palooka I think I misunderstood your original point. I thought you were saying that the harm of breaching privacy is tantamount to the harm we would suffer if someone publicly asked us these questions. I agree that the fact we find such questions disturbing is rooted in our attitude that the information asked for is private. I was just pointing out that much of our motivation for keeping information private is not fear of real consequences but just embarasment and that embarasment is far more intense when you are personally present than when you merely knew something happened in the abstract. This is why many people would prefer to handle embarassing encounters via email or letter and view dealing with it in person as a chore.

So yes I agree that this suggests we have an emotional attachment to keeping certain information private. However, I was just saying that when weighting the costs and benefits of keeping information secret we should be careful to think of it as the information being passively made availible and not equate it with the embarasment of being personally involved in the disclosure.

As for the FDR and JFK deals I have to second what Paul has said and agree that the revelation of this information might hurt the choice of president instead of help it. That is I think it is silly to suppose the people have some prima facia right to any information that might be relevant to someone's role as a president. In fact I think it is somewhat silly to think that there is some fundamental moral obligation to democracy or an open society, rather we favor democracy and open information because it makes for a better country and has good consequences. As such the question of whether presidental medical information should be public depends on whether this will make the population make better or worse choices for president. In particular given the limited resource of public attention I think it is quite possible that public information about medical conditions or sexual misadventures blots out more important matters of policy consideration.

In other words I agree totally with the point that who ends up as POTUS has a huge impact on society. Thus if in fact knowledge of private information made the public make better choices this trumps the presidents interest in privacy. Conversly though if people make worse choices when this information is present (and it isn't hard to come up with examples where more information makes people worse at deciscion making) then it should be kept private for the same reason.

Finally, to Paul I simply disagree that making public policy deciscions based on the actual preferences exhibited by individuals irrespective of the nature of those preferences leads to nihlism. Just the opposite it leads to utilitarinism. This sort of attitude simply says the government should do whatever creates the most happiness regardless of why it creates that happiness.

In fact far from leading to nihlism it warrants exactly the sort of distinctions society finds reasonable and caused problems for you with the dog and menstration case. Why is it that we feel restaurants shouldn't have a right of privacy about keeping a dog in their kitchen but the cook should have one about whether she is menstrating? I think the answer is simple, because the discomfort and emotional unpleasentness of having to reveal the presence of your dog is much less than that of having to reveal your menstraul status. That is we draw a distinction between these situations not because there is some truly objective difference between the presence of dogs and menstration but because the strength of people's reactions to them differs.

The same principle applies to our government laws on discrimination. Why is it that religous belief is a protected class but things like your favorite band, amount of income or opinion on the war of 1812? I would allege that it isn't anything objectively special about religion, the idea that some guy rose from the dead 2000 years ago isn't intrinsicly different than any other ridiculous historical belief or other choosen affiliation rather it is the fact that people have very strong feelings about religion and religous discrimination that they don't have about these other beliefs and affiliations. This is also the same reason we make a big deal out of discrimination based on race, gender and to some extent sexuality but not about other inherent characteristics like having a grating voice or an asymetrical face (it is illegal to refuse to hire a recptionist because they are of a race which wouldn't be pleasing to your customers but not because they have a voice that would be irritating to your customers even though both are inherent qualities).

So even though there is no objective reason why people should define themselves in terms of their race and religion rather than in terms of groups with a longer big toe than second toe or those who believe the Rosenburgs were guilty of spying the fact that people do have these preferences authorizes the government to treat these categories quite differntly. In fact not only do we decide our laws on this basis I can't think of any other way it would be acceptable for us to decide our laws. To deny this principle would be to say that we should make deciscions that we KNOW will cause more overall suffering (in the long run) because the grounds for this suffering is not reasonable.

If you don't accept the idea that the government should consider completely arbitrary but strong emotional reactions when making policy how could you possibly defend the privleged position the government grants to religion?

Logicnazi: I think we're in the midst of one of the Big Unsolveable Political Theory Problems.

Cause, see, it's like dis: You, quite reasonably, posit that policy should be made in a democracy on the basis of what the people want. Hence (I think) you take the position that (a) a lot of things should be disclosed, and that (b) we have to decide on a case-by-case basis which rights deserve protection based on the strong feelings engendered in the people involved. The logical conclusion of (b), of course, is (c) Either the voting booth or the market, or some combination thereof, should decide.

I, by contrast, demand the right to draw certain lines and say that X, Y, and Z matters (which include nondiscrimination on the basis of certain things, privacy in certain matters, etc.), are beyond challenge. I posit that if we simply leave all decisions to the collective decisionmaking power of the polis and the market, we have an absence of no-compromise values -- which I call "nihilism."

However, in order to achieve my line-drawing goal, I have to find some value, some principled reason, to point to to justify my refusal to just hand those values over to the tender mercies of the market and/or the polis.

You seem to, als, be shooting holes in many (but not all!) of my principled reasons.

Fortunately, I've still got my Psychedelic Economics to rest on. I don't think you've answered the last little economic hypo I put forward, where, even if we accept the "economic rationality" as a basis for valuing decisions, we've still got collective action problems. If many people act in an "economically rational" fashion and expend resources discriminating against someone because of their "discreditable" information, the collective loss is possibly higher than if one individual had bitten the bullet and taken a slightly larger individual loss and failed to discriminate. The beauty of my Psychedelic Economic example is that it applies whether or not we permit people to assign whatever silly value they want to the private information. It's simply a matter of game theory.

So I'm going to cling to that with my teeth. (In fact, I may sit down and see how far I can take that particular form of collective action problem, and perhaps publish.)

Supposing, however, that you shoot holes in that too. Then I'm left with this: the impossibility of perfect information. Lets re-frame the discussion some. ("Come here, my pretties...heh heh heh") Why is it that the "economically rational" actors choose to ascribe a large disutility to HIV+ status? I suggest that it's because they have imperfect information about the risks of HIV+ status.

By giving them information as to the HIV+ status of an employee, without being able to convincingly give them information as to the risks of that HIV+ status, we merely replace one absence of information -- as to the virus in that employee's bloodstream -- with another absence of information -- as to the effects of the employer of that virus.

Is that really meaningful? What good does it to provide an economic actor with incomplete and misleading information?

But wait! There's more! Another little trick of my psychedelic game is that the players... don't know each other! So they can't follow the usual way to break collective action problems, which is to make enforceable contracts with each other. I suppose entire industries could get together, in the particular hypo, and create risk pools of employees (sorta like insurance company high-risk pools), but, well, are they gonna?

The only other way to bust a collective action problem generally is by... force. And therein is the justification for drawing a line around certain privacy rights by just that.

Bernard:

I'm not sure we're even on opposite sides of the coin here.

I believe some people do prefer to keep things secret because there are irrational actors. If opening that information up to everyone will lead, in the long run, to fewer irrational actors, but some discomfort and/or embarrassment in the present, I think that the benefits would weigh in favor of disclosure.

"I have to find some value, some principled reason, to point to to justify my refusal to just hand those values over to the tender mercies of the market and/or the polis."

Just stick it to morality. You are allowed to have moral values and ethical beliefs that guide your life. All of our culture heroes do, which is why we worship/respect them. Your distrust of simply handing all of life over to the mechanical preference summing machine of the marketplace does not and can not require mechanical preference summing to justify itself.

If you must, attack the market for its numerous pathologies (offer-asking problems, sensitivity to initial distribution, transaction costs), but really, these are just ways to enter the morally bankrupt land and argue within the terms allowed by the cynics. Those who findamentally reject "skeptical liberalism", "social darwinism", "L&E" or whatever it is called this decade do so on a base, human, spiritual level and would do well to admit that.

Those who insist on re-labeling "values" as "preferences" simply reveal their cynicism, which is very unattractive. This is why the People reacted so badly to Bork. I posit that real human beings believe in things without proof based on learned cultural values and this is the right and proper way. To call these expressions of cultural truth "preferences" demeans their derivation (in the minds of those who hold them) from a higher, natural or even mystical source.

It is one of the characteristics of higher education that the more one learns the harder it is to ground anything on faith or instinct or inspiration. Universities and scholarly discourses adopt an institutionalized cynicism.

For example, I recently learned that one of the most famous collections of "existential" literature (used to teach undergrads) includes only the first half of Dostoevsky's Notes From Underground. In that book, he manages to both present the problem of cynicism and suggest a humanistic redemption. Modern scholarship has deemed the latter to be worthless when in fact it was the whole point to me.

But back to law, as between the market or the polis, the choice depends on whether one prefers one-dollar-one-vote or one-person-one-vote. And that choice really depends on the extent to which you believe that it is possible to acculmulate lots of dollars in a principled, praiseworthy manner. As my contracts professor said, "People don't get rich by exchanging goods of equal value in the marketplace, yet we praise the rich."

I would point out that the desire to reduce "values" to "preferences" is an excellent way to avoid the above question. To the extent that we allow moral content to determine policy, then there is a strong argument for real equality. If "fairness has no content", then those with many dollars can justify their unequal status on the grounds that preferences are best expressed in monetary terms.

That's how I read the subtextual poly-sci debate here most weeks.

"Why is it that the "economically rational" actors choose to ascribe a large disutility to HIV+ status?"

Even given perfect information about the risks, it could still be possible that there is a large disutility to hiring HIV+ employees. (In terms of higher health insurance costs, since HIV treatment is expensive)

However, most people would agree that employers should not be able to discriminate on this ground.
(Applying their moral/cultural equality norms)

This situation REQUIRES a (as you say) "no-compromise" value in order to achieve the result that most people would prefer. The market, if it functions at all, will discriminate against the unhealthy if left unchecked.

Adhering to a privacy right would work, however, there is perhaps a better way... simply adopting a universal equal access right to health care! The costs of treating HIV+ patients would be spread and there would be no incentive for a particular employer to discriminate in order to affect premiums.

Depending on the privacy right is a less desirable way to approach this because it simply says to employers "no, you can't cost-optimize in this particular area."

Similarily, use of privacy norms to protect bargaining power in the market... which could be better protected by equality norms regarding fair prices. (Or simply by a more equitable initial distribution of bargaining power)

The Supreme Court uses "privacy" as a stand in term for liberty interests in abortion, contraception, sodomy. Perhaps as Posner suggests because privacy is more universally recognized as a "good".

Corey: you're completely right, of course. I'm merely trying to beat them at their own ballgame.

Doesn't the concept of "reasonable expectation" have a great deal of importance in understanding both emotional responses and the costs/benefits of privacy rules? For instance, I have no expectation of privacy when I post this theory on a blog, but would have a greater expectation were I to email this directly to Judge Posner and ask him not to identify me by name, and a still greater (in fact, legally enforceable) expectation if I were to send it to my lawyer. This expectation will determine in large part the opinions I am willing to express and how I express them. Indeed, context has such a powerful effect on the content of speech (gossip vs. personal opinion vs. official or professional statements) that I believe that most people make the appropriate adjustments totally unconsciously. (Except, famously, for a certain politicians/college presidents who misjudge their audience occasionally and make remarks that later blow up in their face -- i.e. about Strom Thurmond's legacy or the academic capabilites of women scientists).

This idea seems quite broad to me. For instance, I may request an HIV test from my doctor if it is anonymous (cannot be reported to private insurers) rather than just confidential (can be, tho only if I agree to let them review my charts, as I must in order to begin a policy). And I may take great care in taking a law school exam but little care in writing a seminar response paper, because I know that future employers will have access to the results of one but not the other.

Put formally, the model that Judge Posner outlines looks only at whether the social utility that results from a set of transactions would be increased or decreased if certain information were disclosed. But I believe the model could usefully be expanded to consider whether the privacy rules will affec whether that information would exist in the first place -- for the reasons noted above, I think there are plenty of situations where it would not exist without an expecatation of privacy.

It also seems to me that two conclusions flow from this amendment to the model. First, in terms of understanding peoples' emotional responses, the level of discomfort about disclosure of private information can be seen as a type of betrayal of expectations. I don't want my nude picture displayed because I had an expectation that my pose on a nude beach would not be reproduced. If I had known that it would be, I might have exercised more this winter in order to look more attractive. (I concede that there is also a more basic discomfort with nudity operating here -- but I do think that the idea of being "violated" is in some ways driven by the dislike of being "surprised".)

But more important, the cost-benefit analysis of privacy rules also changes if we account for expectations. To take an easy example, knowing that I am "off the record" may permit me to have discussions with people that are socially useful but that would not occur without that expectation. The lawyer-client privilege is one example, as is the informal chat between academic colleagues (the privacy rule in the latter case is of course a social convention rather than legal obligation). But more broadly, consider the rule that the police have less authority to look at what you do in your house than what you do in the street. The obvious cost of this rule is that it leads to fewer drug arrests, since criminals can assume that their home is a safe base for, say, turning cocaine into crack. But the benefit is that people feel comfortable in their homes viewing non-obscene pornography or walking around naked or drinking heavily, all of which are perfectly legal and utility-enhancing activities that they might not want other people (including police) viewing. More generally, and this does get into the idea of privacy expressed in Griswold and the due process cases, creating a sense of comfort in certain specified realms is important to allowing people to engage in lots of behavior that they would be shamed into stopping if they had no expectation of privacy. Thus, if we want to allow that class of semi-shameful but utility-enhancing behavior, then we need to set clear privacy rules and enforce them. And we need to think about how various degrees of privacy would affect the amount of such behavior. Of course, the best outcome of all would be to stop people from fee