« Broken Windows | Main | A Jujitsu Defense? »

October 18, 2005


Feed You can follow this conversation by subscribing to the comment feed for this post.


Prof. Epstein,

Great first posting on the blog. I hope you will keep it up. My favorite blog is Becker-Posner blog, and I'm sure you have at least as much to add as those gentlemen.

I'm not very familiar with this topic. I was wondering if you could address the health insurance question of how current employees are grouped, how the addition of these potentially "high risk" hires affect their insurance rates in the absence of genetic screening, and whether the prohibition on testing will lead companies to drop health insurance completely or make the employees pay higher premiums, co-payments and so forth. Thanks. And keep posting!

Daniel Schwei

I think that this post provides an interesting approach to a soon-to-be incredibly controversial issue. With that said, I think that there are critical flaws in Professor Epstein's argument.

First, Professor Epstein presents a solid case for the employee's potential benefits to genetic testing. Testing may easily allow an individual to prolong her lifespan by enabling her to take preventive measures. However, this hardly justifies allowing employers to probe into the matter. Simply because the good experienced by the employee (genetic testing leads to a longer life) outweighs the bad (employers now get to ask about this information) does not justify allowing the bad.

Secondly, Professor Epstein argues that there is no difference in principle between the kind of information normally asked in interviews and genetic information. His examples are blatantly self-serving. Professor Epstein assumes that employers will only look at relevant information for the job. A basketball player's physical health is certainly germane to his employment. I would even be willing to admit that the presence of a heart condition is relevant to a job in which temporary and spontaneous incapacitation could lead to catastrophic results. However, for the vast majority of jobs, I think it is too far of a stretch to argue that these types of factors are relevant. A manual laborer's risk of developing Alzheimer's is too far removed from his job function to justify this type of inquiry. To allow employers to weigh these factors in assessing potential employees would be so offensive to our public conception of justice and so disruptive to our social order that permitting employers to utilize genetic testing would not be worth whatever minute gains the market would achieve in efficiency. Provided all other characteristics equal, do we really want to permit job offers to be determined by tiny differences in risk for developing medical conditions decades later? Provided the same desire to get pregnant, should we hire the less fertile woman because she is less likely to get pregnant? By this logic, we should hire a man because he has no risk of getting pregnant, and surely this offends us.

Professor Epstein may respond that I should have more faith in the market. It would only make sense for employers to spend the time and money on genetically testing their potential employees if the efficiency gains were large, or if the factors they were searching for were actually quite relevant to job performance. It would not be efficient for employers to test every employee. My response is that while this may be true in our day and age, the one fact that cannot be disputed is that technology is advancing at an incredibly rapid pace. In fact, the day might not be so very far off when genetic analysis becomes so cheap that it is efficient for employers to test everyone. My argument is that the gains in efficiency this may provide are not worth the privacy violations and morally repugnant discrimination that would follow from it.


Questioning whether employers should or should not be allowed to use genetic testing to screen potential employees is another step down a disastrous path for businesses and health policy. The important question is not whether genetic testing allows increased efficiency in the market. It is how we have gotten to the point where businesses are so willing to pervert their hiring practices that they will use a criterium highly irrelevant to job performance to screen employees. The example of a professional basketball player that Professor Epstein uses, while interesting, is hardly a typical one, or an illuminating one when it comes to general issues of genetic testing and employers. The real issue is if employers' responsibility for their employees health plans is having such a distortionary effect on hiring practices, why we still believe health care should be connected to employment. If your main concern is really efficiency in the job market you should be arguing that health care should be independent of employment, not that genetic testing should be allowed by employers.



[And with apologies to Prof. Epstein, for whom I should be doing some work right now. Welcome to the blogsphere!]

In response to your first point, it shouldn't matter whether we let employers ask these questions so long as the employees are free to answer them. Consider two potential employees, one with perfect health and one with a disease detectable through genetic testing, such that the first employee is worth $100 to a particular employer and the second is worth $60. Let us further assume that neither the employees nor the employer know of either employee's genetic disposition, and that the distribution of their dispositions in the labour market as a whole is even (that is, 50% with and 50% without the genetic disposition). In this situation, we can expect that each employee will be paid $80 in a competive labour market.

But now let us assume that the first employee--with the perfect health--somehow learns of his fortunate health. The catalysing event doesn't much matter, so long as there might be one (perhaps he is a basketball player wondering if it is safe to keep playing). If we allow him to report this fact to his employer, he will be able to--rightly--demand the full $100 that he is worth to the employer. In turn, the employer will lower the pay of the second employee--regardless of his health--under the implicit assumption that if he doesn't affirmatively demonstrate his health he must be hiding his poor health.

Under such a model, we would expect that the employees would provide this information to the employers, even without their demanding it. Indeed, for many this is what higher education is all about: signalling to employers that one has a higher earning potential and therefore should be paid more. The simple rule is that confidentiality rules don't work, if the parties they are meant to protect can benefit through disclosure. Note, as Prof. Epstein does, that an employee might value his privacy over his salary, in which case it is perfectly rational for healthy employees to take the lower salary--it's their choice.

It seems, then, that the concern shouldn't be whether employers can ask these questions--or for that matter whether employees can provide such information with or without asking--but whether they should be allowed to make employment or salary decisions based upon them. (Note that, under this model, the second fellow wouldn't lose his job at the employer's bidding. He's worth $60 to the employer, and so long as he'll take that much, he's got a job. It is rather the case that he would only lose his job if he decided to remove himself from it, perhaps because he found a job less sensitive to his genetic disposition (and that therefore would pay him more)).

Note, too, that there might be a paternalistic argument to be made that both employees should be paid the same amount, because they have equal moral worth or some such--that is, it's not the second guy's fault, so society shouldn't let people pay him less. But then we need to call it what it is: a wealth transfer from those well off to those less fortunate. But why do this through such indirect means? It so happens that, due to an accident of fate (our inability to do genetic testing), we have necessarily done this wealth transfer. But we shouldn't maintain an ineffecient market simply because it's what we've always had to use before. Better it seems, if we do decide that society should concern itself with this, to tax the first employee $20, or find some way to incentivise a direct transfer.

This transitions us to your second point. Why should a construction company care if I'm going to develop Alzheimer's in 50 years? If a company decides to make employment decisions based upon such information, it is making bad decisions, and, one would hope, would take a hit in the pocketbook as a result. Note that, if such testing were implemented in general, it is likely that the employer would bear the cost of testing, which provides a natural incentive not to require it arbitrarily (and, as a very positive externality, would be a free trip to the doctor for the job seeker). But, perhaps this won't in happen in fact (esp. where there is a labour surplus and minimum wages). It seems the better response is not to hamstring the market by denying it access to this information, but to hold it accountable for the responsible use of it: treat genetic disposition discrimination as a form of prohibited discrimination. Of course, this carries with it the myriad faults with the anti-discrimination laws; but why multiply faults when you can condense them?

Perhaps, however, a more important point comes out of your second concern. If it is such an affront to societal norms that an employer would ask such questions, making it legal to do so won't change the status quo in the vast majority of occupations. Only in those occupations where such information is of particular use would we expect it to be used.



The second and third comments are a bunch of crap. It's not your g-d damn business if private employers want to use genetic testing in their hiring practices. They're the ones who have payroll, overhead and other expenses that have to be paid, not you. If you don't like it, go choose some other place to work. And who the hell is "we" find this offensive anyway. I think it's offensive to force government to interfere with private decisions or to run other peoples' lives. It's a free country, not a Marxist one. Government jobs - we can let the voters or elected represetatives decide. Private jobs - we let the employer decide.


I think the argument made by prof. Epstein has a lot of merit from a standard law-and-economics point of view. Gus has done enough to flesh it out a little in response to some points of criticism, so I won’t go into that. There is of course ample room for debate on this level alone, but I would like to ask prof. Epstein a question that takes the issue to, I think, a more fundamental level.

What impact, if any, does the advent of vastly more sophisticated forms of genetic testing and mapping have on your arguments in favour of a (broadly) free society (as for example developed in your books)? You do not (any longer) believe in natural rights arguments in favour of the type of (simple) rules that govern a free society. The utilitarian argument you make in favour of, for example, the rule of individual autonomy and self-ownership is, I agree, more convincing. As a utilitarian argument, however, it depends on social conditions that may be subject to change. We are approaching the point where it is imaginable that a person’s genetic make-up, and many of the factors that will impact his ability to lead a full and productive live, are known in advance. It is easy to argue that, from an economic perspective, this can only serve to improve efficiency, allow efficient sorting and so forth. But the results in terms of what such a society would look like, may be very different than what the intellectual heroes who defended the rule of individual autonomy two and three centuries ago, imagined. I think you have (elsewhere) made a strong utilitarian argument that in a world where no-one knows what each person’s initial endowments are and what the future has in store for them, while at the same time knowing that comprehensive redistribution schemes come with enormous costs, a strong individual ownership rule works out best for everyone ex ante. This state of the world results in a lot of “tough luck” being spread out very thinly and randomly over the market. Full genetic disclosure, on the other hand, opens the prospect that genetic bad luck will become fully internalized, and hence extremely focused on the individuals afflicted. Historically, this was the case only for the (relatively small) subset of handicaps that where visible to all to see. But it is exactly in those cases that “veil of ignorance” type arguments (to which you are methodologically sympathetic) in favour of redistribution are strongest. Will it, with much larger parts of a person’s medical future visible for all to see and internalized through market mechanisms, not become much harder to say that, ex ante, everyone is better off (or at least no-one worse off) by allowing the free exercise of individual autonomy in these matters. And if the number of people who lose out under such a system would indeed turn out to be substantial, wouldn’t your approach at least warrant some form of compensatory transfers to the losers, which would then have to be organized on a much larger scale than anything currently envisaged. Viewed in this light, a “no genetic testing” rule could be the less intrusive alternative (although I think it would be untenable in practice in view of the self-sorting mechanisms Gus touched upon).

I am deliberately taking one side of the argument here. It’s just that I am quite sympathetic towards prof. Epstein’s arguments in general (even when I was in his class and called upon) and have lately been struggling with this issue.

Jeroen Delvoie
University of Brussels Law School


in response to gus, the only reasons why an employee could be "worth more" to an employer in your example are either because the health condition would affect their employment, or because the employer is responsible for insurance. Instances of (1) - where the genetic defect will affect employment - seem unusual, though we can imagine employers won't necessarily want to hire people who will die in a few weeks due to training costs or whatever. But it's really the insurance that's the rub. Get rid of the employment-insurance nexus, and why does an employer (training/admin costs aside) care whether his employee has some exotic genetic condition? I guess I'm confused by the logic.


incidentally, the timestamp on this blog is off by a few hours


Timestamp should be fixed now. Thanks.

Scott Scheule

It's posts like these that make me wonder whether I should call myself an Epsteinian or an Epsteinite. Epsteiner, perhaps.

Regardless, perfectly delightful, Professor; please keep up with blogging.


I agree with the general theme of Ryan’s response, though I would not characterize the second and third responses as “crap”. I will take the approach of John Stuart Mill and recognize the inherent value of differing opinions as they compete with each other, regardless of how incorrect, wrong, or offensive they may seem.

That said, however, I must severally note the very presumptuous nature of Daniel Schwei’s entire post. As Ryan similarly asked, what is this “we” you keep using? It is as if “we” have any say in the private, voluntary transactions between two private entities: (potential) employer and (potential) employee. You posit yourself (or society) as the moral arbiter of private action; such a position is obviously wrong, given the fact that you don’t merely express disapproval, but call for the state to intervene in order to prevent a line of inquiry that you find distasteful.

First, you first say that a patient’s personal health benefits from genetic screening don’t justify “allowing” employers to “probe”. “Probe,” however, is a misleading word. Employers no more probe potential employees than university recruiters probe potential students. Granted, employers may require more or different information, but in neither case is such gathering of information conducted by force, which is what “probe” seems to imply. You also use “allow” for the first of several times, along with “permit” later, both of which are problematic. It is not up to us to “allow” employers certain lines of inquiry, and they need no “permission” to engage in them.

Second, you write, “Professor Epstein assumes that employers will only look at relevant information for the job.” On the contrary, Epstein just does not presume in his post to decide for the employer what is “relevant” the way you do. Since you think you have the keen eye to discern the relevant information from the irrelevant, it would have been useful to share with us what your systematic criteria for judgment are. You do not do this, however. You only provide examples, which is insufficient in order to demonstrate your qualifications as the judge of what is “relevant” in what should be considered private affairs. In addition, your designation of certain information as irrelevant posits yourself as one more qualified (not to mention authorized) than the employers themselves to make business decisions for all employers—it rarely gets more presumptuous than that.

Third, you fallaciously write: “A manual laborer's risk of developing Alzheimer's is too far removed from his job function to justify this type of inquiry.” I doubt an employer who is hiring manual labor is going to worry about one disease of the brain that affects primarily those who are too old for such labor. That you use such a weak example here is to commit the straw man informal logical fallacy, for it is unlikely that even a sizable fraction of employers would base personnel decisions on a single disease unrelated to their field of work.

A much better example would be to perhaps require Alzheimer’s screening for Supreme Court Justices. Their jobs are highly dependent on their brains being in tip-top shape, which is all the more relevant given the fact that these justices can get quite old when serving lifetime appointments. Argue why this is not a good idea, and then when you cannot, then argue why public employers should be able to use such screening, but why private employers should be prohibited by law from doing so.

Fourth, you write, “To allow…would be so offensive to our public conception of justice…so disruptive to our social order…that permitting employers to utilize genetic testing would not be worth whatever minute gains the market would achieve in efficiency.” Not only do you continue with the problematic “allow” and “permit” language, but also you pull from who-knows-where “public conception of justice”. What does that phrase even mean? It seems to be little more than a feel-good PR device meant more to elicit emotional agreement than to convey real information. It is also fallacious, for the public’s conception of many things can be way off, and it seems that it is by virtue of the conception as a public one that we must assent to it; that is little more than the ad populum informal logical fallacy.

You also use, “disruptive to our social order”, but never really articulate what it is about our social order that will be disrupted. Since we do not know, we cannot assume that it is something good that we wish to keep, or that if it is good, that a greater benefit will not outweigh the loss of that good, not to mention that one person’s presumably harmonious social order can be another’s stifling status quo that needs to be shaken up a bit. It is another ambiguous phrase that does not advance your argument. And I will just file “minute gains [in] efficiency” under unqualified usurpation of employer autonomy, whereby you presume to judge efficiency better than the employers themselves, which is highly unlikely.

Fifth, you write, “…do we really want to permit job offers to be determined by tiny differences in risk…,” offering more of the same, that employers need permission to ask certain questions and that the differences in risk would be small. Again, businesses do not need permission, and you are neither qualified nor authorized to supplant employers risk assessments with your own.

In reference to your pregnancy example, such considerations are not out of bounds, especially for some demanding jobs. A CEO, for example, will more than likely be a man for precisely this reason—along with pregnancy comes leaves of absence. The nature of some jobs requires that people who are less likely to take such leaves fill those jobs. In addition, medical examinations of women’s fertility are virtually useless, since technology exists to compensate for weaker fertility. Pregnancy is much more a result of personal choice than biology more so than you imply.

Finally, you write, “…the gains in efficiency this may provide are not worth the privacy violations and morally repugnant discrimination….” Nothing in Epstein’s post indicates any privacy violations, because the potential employee is never forced to give such information. The potential employee has the right to refuse to give it, and in turn, the potential employer has the right to find someone else who will give it, and it does not matter how “morally repugnant” to you something is. The phrase itself is too ambiguous. Most would agree that cheating on a partner is morally repugnant, yet the law properly has nothing to say on the matter.

Employers should be able to make their hiring decisions based on whatever criteria they choose. It is not up to us to “allow” employers to conduct their business as they see fit, nor do they need permission. To argue that they do is to claim that potential employees have a greater claim on employment from the employer than the employer does in making the best decisions that will yield the greatest benefit. Unless coercion is involved somewhere, either party to the transaction has equal right to give or seek whatever information each requires to maximize the benefit to each.

Ken Arromdee

"One argument is that people will shy away from finding out their genetic position if they are subject to the test. Don’t believe all this. By assumption, there is good reason to believe that the information that is acquired from genetic sources is of value not only to the employer but also to the employee. ... Hard to believe that workers would take that kind of risk with their own lives."

The problem with this idea is that being able to earn a living is itself a big part of someone's life. If taking a test lets someone take prophylactic measures, that has value, yes. But if taking the test decreases their chance of getting a job, they lose value. It is possible that the loss in value from not being able to work outweighs the gain in value from being able to take prophylactic measures.

A (hopeful) UC applicant

"We are approaching the point where it is imaginable that a person’s genetic make-up, and many of the factors that will impact his ability to lead a full and productive live, are known in advance."
One could argue that IQ does this already, though it is nowhere near exact.
Jeroen, your argument applies to employers determining IQ.
"everyone is better off (or at least no-one worse off)." Those with superior genes are being taxed. No person is given the perfect salary, as no employer has perfect information. Again, Epstein's position is far more agreeable and obvious when you put the argument in terms of IQ and not genetic information. Why should the losers be compensated for their inferiority? Do I possess a partial claim on your wages because you are smarter, and will thus earn more than I?
Just as screening for intelligence makes certain otherwise impossible ventures possible, so too I suspect that genetic screening has a similar effect, though not as obvious.

It makes possible greater employment by lowering health care costs, so long as no law requires that employers provide health insurance.

BTW: How has Epstein changed over the years? I read his Simple Rules, but you said he has since changed tack. I read in one conservative rag that current libertarian streams of thought were insufficiently convincing, and that a new one has emerged in a journal called Critical Review. I read a Salon article covering the journal, and it sounds quite interesting.

Possible Member of Class '09

Prof. Epstein mentioned briefly the Americans with Disabilities Act but did nothing to address the violation of the Act that might occur by allowing companies to ask genetic information. Employers are not allowed to discriminate based on a disability. However, if a prospective employee is tested and found to have Multiple Sclerosis, he or she might be denied employment or be discriminated against through salary. This might even occur if the employee refuses to provide the results because the assumption would be that they were bad. Right now, employers cannot ask a potential hire if he or she has a disability. Allowing genetic information to be used in hiring would blur the lines the ADA has drawn, and this concerns me.

Money management

Using a Global Positioning System receiver, they log into a computer the latitude and longitude of each fish, plus their names, the date, the water temperature and the depth at which the fish was caught.


the idea of employers doing genetic testing as a condition of employment just makes my skin crawl.

The comments to this entry are closed.