A Gillette-Torvik Conversation™: End employment discrimination against the unemployed?

TORVIK: Mr. Gillette, what is your take on this aspect of Obama's new jobs bill proposal:

Ending Discrimination Against the Unemployed:   Recent reports have highlighted companies that are increasingly expressing preferences for applicants who already have a job. Specifically, some companies are posting job listings that include language such as “unemployed candidates will not be considered” or “must be currently employed” or “must be employed within the last 6 months.”  The exclusion of unemployed applicants is a troubling and arbitrary screen that is bad for the economy, bad for the unemployed, and ultimately bad for firms trying to find the best candidates. This is particularly true at a time when so many Americans have found themselves out of work through no fault of their own. New Jersey has passed legislation to address this practice, and members of Congress have also introduced legislation. The President’s plan calls for legislation that would make it unlawful to refuse to hire applicants solely because they are unemployed or to include in a job posting a provision that unemployed persons will not be considered.'

I understand the sentiment, but it seems like a bit of a hornet's nest to me. Currently firms work hard not to ask questions about "suspect classifications" to prevent any claim that someone wasn't hired on such a basis. But an employer has to ask about employment history!

GILLETTE:  Are you saying that the proposed claim is problematic because one can eliminate potential liability on a failure-to-hire racial discrimination claim by not asking about race (setting aside that some things, like zip codes, correlate strongly to race) but that an employer cannot similarly eliminate potential liability on a failure-to-hire- the-unemployed claim because they need to know job history?  If so, I don’t really see that as a problem with the bill.

In most cases, an application includes at least one thing that identifies the person as a member of a protected class. A person’s name tells you their sex 99% of the time. Just as knowing job history is relevant to assessing a candidate, a potential employer has to know the name of their potential employee. I have never heard anyone suggest that needing to ask applicants their name somehow exposes an employer to liability. Job history seems more akin to asking a name then, say, asking applicant their race.

Assuming the word "solely" means what I think it means, it should be pretty easy for an employer to come up with a reason besides employment status to not hire someone. If people who don’t have work aren’t being hired simply because they aren’t currently employed, I don’t have a problem with legislation to address the problem. It will be somewhat odd to have a cause of action, besides whistle-blowing, where the discrimination isn’t based on some immutable trait.

TORVIK:  You are of course correct that firms can usually learn about some suspect classifications based on information that they have to ask. But it is also true that one way firms insulate themselves from bogus claims is trying not to ask questions that would elicit such information. I remember being instructed at Dorsey not to ask about marital status during interviews of potential associates -- even though it of course would often come out anyhow. 

So I take your point that this particular aspect -- the inability for firms to shield themselves from the knowledge of the suspect classification -- isn't a deal killer for the bill because we already have other suspect classifications, like race, where the same thing is true. And I don't particularly have a problem with the idea that firms shouldn't refuse to hire people solely because they're unemployed. But I think the bill has some issues in practice. As you mention, it seems like an impossible claim to prove. How would someone even go about it?  I guess you'd have to show first that they hired somebody who already had a job. Then the employer is going to say something like, "we hired that guy because he had a better work history."  Surely that's still an acceptable basis on which to discriminate among applicants. And then the plaintiff would have to show that even though he had been unemployed he had a superior work history to the other guy/gal.

In other words, I think this bill is just pointless -- it won't do anything because it can't do anything. It's just for show.

I guess the one thing it could do is to prevent firms from making it a formal requirement that applicants be employed. Maybe that's the only point. But the side effect is going to be frivolous lawsuits, which I don't think are going to do much for the economy -- since the only way to prevent such lawsuit entirely is not to try to hire anybody...

GILLETTE:  Since fear of lawsuits under Title VII, the ADEA, the ADA, and various state employment laws doesn’t seem to inhibit employers from hiring, I doubt this new bill (in the unlikely event that it passes) will either. That said, I am sure that opponents of the bill will rely on similar reasoning.

As for how claims would be proved, I assume Court would simply adopt the McDonell Douglas test to the new law:

1. Unemployed plaintiff is applies, and is qualified for, the position;
2. Unemployed plaintiff is rejected for the position;
3. Employer hires an already employed applicant for the position (or continues to seek applicants for the position);
4. If 1-3 are met, then there is a presumption of discrimination and the employer must offer a legitimate non-discriminatory reason for hiring someone else (or continuing to look for applicants). If the employer does so, then the presumption of discrimination dissipates;
5. If the employer offers a legitimate non-discriminatory reason, the plaintiff must present facts to show an inference of discrimination. The plaintiff may do so either by showing that the defendant’s explanation is only a pretext for discrimination. I am not sure your superior qualifications argument would necessarily work in meeting this third stage as qualifications are often subjective and an employer is allowed to be wrong in their subjective assessments of performance. Courts are usually reluctant to second guess HR decisions absent a very compelling reason to do so. 

As for whether the bill is pointless because employers will find ways to circumvent the law, I do not agree with that. Presumably in 1964, a lot of people said the same thing about Title VII—that it can’t prevent discrimination, can’t change minds, and will only stop employers from explicitly refusing to hire on the basis of sex, race, religion, and national origin but won’t prevent them from hiring the way they want to hire. In 2011, however, I don’t think you would find many people who would say that Title VII didn’t work. Employment opportunities for women and minorities are much better than they were 50 years ago.

As for whether the side effect is frivolous lawsuits, I assume that there will be more lawsuits. I don’t have any reason to believe that this law will cause a higher percentage of frivolous lawsuits than any other law (statutory or common). As you probably know, your old firm got a Title VII lawsuit dismissed last week as being frivolous. While I think that the lawyer who drafted the complaint ought to be reported to the Board, I don’t view that as grounds for repealing Title VII.

TORVIK:  I think that frivolous lawsuits are definitely a side effect of Title VII and other anti-discrimination laws. Those laws also inhibit hiring somewhat by making it more risky and expensive, in the long run, to do so. The question is -- are these side effects worth it? In the case of Title VII most people would say yes, me included, for moral reasons. But it doesn't do any good to pretend that there are no economic costs to this regulatory regime.

So the question for this new proposed anti-discrimination law is -- are the side effects worth it? My preliminary feeling is that they are not. I don't see what good it is going to accomplish -- particularly as part of a jobs bill. As far as I know, this is the only kind of anti-discrimination law that would apply exclusively to the act of hiring. (Obviously you can't fire or refuse to promote someone for being unemployed.) So all the potential costs of this regulation will fall on the decision whether or not to hire. At the margin (i.e., in a borderline case) some companies are going to decide to make do with the employees they have rather than to subject themselves to potential liability for refusing to hire an unemployed person. If discouraging discrimination against the unemployed were a great moral cause on par with discouraging discrimination against blacks and women, etc., this cost might be worth it on that basis. But I for one do not get the moral case. In fact, I outright disagree with it, since I think that it is rational for an employer to prefer to hire an employed person over an unemployed person, all else being equal.

Do I think this law will be a disaster or have a major impact on the economy? No. I just think it does (a) no good and (b) some harm. No good plus some harm equals bad law.

GILLETTE:  I wasn’t aware I was pretending that there were no economic costs to the bill. I take it as a given that, as you put it, “regulatory regimes” have costs. I don’t have any evidence to suggest these costs are particularly high on this bill. For the most part, failure-to-hire discrimination cases are rare. This is because the applicant typically doesn’t have enough information about the hiring decision to believe that something is amiss about the hire. A discriminatory termination or failure-to-promote is more common because the employee—because they are already employed and have access to more information about the process—realizes that something strange played a part in the promotion decision or termination decision. While I recognize it is anecdotal evidence, in eight years of practice I have handled hundreds of employment cases and talked to a much larger figure of potential cases (I believe the rate for taking a case to talking to potential cases is about 3 percent). In that time, I have handled one failure to hire case and talked to one potential failure to hire case. The only reason those two occasions existed was because the employer was the State of Minnesota and the plaintiff, through use of public records laws, was able to see the qualifications of the people selected for the position.

Presumably the good it would accomplish is prohibiting employers from using unemployment as a reason for not hiring someone. 

Why, if all things are equal, is it rational to prefer an employed person to an unemployed person? In the current economy the unemployed person may be unemployed for reasons that have nothing to do with anything the unemployed person could control. It does not seem rational to base decisions on things that are out of the applicants control.

TORVIK:  For the record, I accused you of pretending there were no costs because you said that Title VII, etc., "doesn't seem to inhibit employers from hiring."

I think it is rational to favor the employed person because it is reasonable to suppose that the employed person is more employable than the unemployed person. You're right that it's not always the unemployed person's fault that he or she's unemployed, and there are lots of reasons that someone could be unemployed. So sometimes it will turn out to be a bad decision to favor the employed person, and it may be a dumb idea to refuse to interview anyone who's unemployed. But sometimes it is the unemployed person's fault that they're unemployed, and it's usually very hard to figure out whether this is so (in part because former employers will never say anything bad about ex-employees for fear of being sued for defamation). Thus, as a general tie-breaking rule, it seems at least rational to me to favor the employed over the unemployed. That it is a rational tie-breaker does not mean that there are no exceptions, and that it always works -- or even that it's necessarily a good idea -- just that it is a rational way to make a decision. As an analogy, I consider it rational for an employer to favor college graduates over college dropouts. Are there plenty of good reasons that people drop out of college, or at least reasons that have no bearing on their ability to do a particular job? Of course. But I still think it is a rational tie-breaking rule.

Compare this to a general preference for whites over blacks, or men over women. These are irrational prejudices. Preference for the employed (and the college graduate) are rational prejudices.

I would agree that the prejudice for employed over unemployed isn't as compelling as the prejudice for graduate over dropout. So if there were a truly compelling reason to use the force of federal law to prohibit discrimination against the unemployed, I'd be willing to support it. But I see no compelling reason. Most anti-discrimination laws have a moral foundation. There isn't one here. The justification seems to be economic. My argument, however, is that this bill would be a small net economic loss. So there is no justification for the bill.

I have no reason to doubt your anecdotal evidence about the rarity of failure-to-hire cases, which makes perfect sense. But two points: (1) this bill is different because it is exclusively a discrimination-in-hiring bill; and (2) if correct, your anecdotal evidence tends to show this bill is pointless, because it will never be enforced.

As for there being no evidence that the costs of this new law would be "particularly high," I think that misallocates the burden of proof. I think proponents of this bill have to prove that its benefits will outweigh its costs, not the other way around. Particularly under the circumstances of this being part of a jobs bill during an unemployment crisis, I think there's a heavy burden on Obama to show that this bill will actually reduce employment rather than, you know, increase it.

GILLETTE:  I am not sure it is rational to prefer college graduates over college drop outs. That seems like it depends on the position. However, I recognize the point that employers need to have tie-breakers for deciding which applicants to pick. I just don’t think that being currently employed is a good tie-breaker.

Your second point, that the bill will not be enforced and is therefore pointless also cuts against your view that there is an economic cost to the bill. If there isn’t litigation, what is the cost? If the benefit is that employers start offering the unemployed interviews, and the cost is negligible (because there isn’t much litigation) then that seems like a net gain because, I assume, sometimes the unemployed will get hired based on their interview where before they couldn’t get an interview.

I don’t think we are going to agree with where the burden of proof ought to lie. If the bill’s opposition is claiming that there is a cost, then they should be able to point to something that shows what the cost will be and how that cost will actually inhibit an employer from hiring. Just saying that all regulations have costs, doesn’t do much for me. The question isn’t whether there is a cost but rather whether the cost is somehow harmful. I don’t see how this creates a harmful cost. 

As for whether there is a need (or if you prefer a moral need), according to the Bureau of Labor statistics, there are 14 million unemployed persons in this country and nearly half, 6.76 million Americans, are considered long-term unemployed. That is a lot of people. As Salon.com notes, there are 4.8 workers for every job slot. Before the recession there were 1.5. The country would need to add 300,000 jobs a month every month for five years to close that gap. If the law allows helps get some of those folks back to work by getting them back in the interview process, that is fine with me.

TORVIK:  Whether or not you think being employed is a "good" tiebreaker, the question is whether it should be an "illegal" tiebreaker. To me, there's a big difference. Generally, we let employers make the call on what's a good tiebreaker, with immoral discrimination being the notable exception. So I think there's a heavy burden on proponents of this law to explain how it will help the economy to constrict how businesses make hiring decisions. Perhaps there's a fundamental temperamental difference that underlies our disagreement about who bears the burden of proof about new laws: should it be the person who is trying to expand our constitutionally limited government or the person who arguing for the status quo? I've always believed that you need to persuade to change the status quo.

I don't think the bill being pointless undercuts my argument. That is my argument: no good plus some harm equals bad law. I guess what you're saying is that a pointless law can't do harm. I disagree. Whether or not there turn out to be many lawsuits filed, there will be lawyers hired to ensure compliance, and there will be marginally less incentive to hire. Those are costs, and those are harms. In any event, even if there is no harm I submit that no good plus no harm still equals bad law. But perhaps that's another example of our different ideologies.

Question: how will this bill create jobs, even theoretically? It seems to me that whether you hire an unemployed person or an employed person, there's still only one person getting a job and only one job has been created. Is the argument that if you hire the unemployed person, the employed person still has a job, so there are two people employed where previously there was only one? But if you hire the employed person, that creates a new job opening where that person was working before. So I don't think this bill can do anything to reduce unemployment, and my worry is that it will actually worsen unemployment by making hiring more expensive. The last thing we want to do is discourage the former employer from hiring a replacement, which is what this bill could do.

Here's another thought on why this bill makes hiring more expensive even if there are no lawsuits. Currently when employers put out a job listing, they get deluged with applications because there are so many people looking for work. This deluge creates administrative cost. One way employers reduce this administrative cost is to add requirements -- such as current employment -- to the job listing. Even when they do not have that explicit requirement, when faced with 10,000 applications for a job, an employer is going to look for rational shortcuts to separate the wheat from the chaff. Current employment status is an obvious way to do that. Making that shortcut illegal -- whether or not it is a good idea from the standpoint of hiring the very best candidate -- clearly adds costs to hiring by making it take longer. 

Another potential cost: token interviews of unemployed people. You note that this bill would encourage employers to interview the unemployed. More interviews means more time spent on the hiring process, which means a more expensive hiring process. Which means less incentive to hire.

I support a federal jobs bill and support government spending to encourage businesses to create jobs -- I just don't support ideas that will add further regulatory burdens on businesses, increase the cost of hiring, and do nothing to create jobs.

GILLETTE:  Mr. Torvik, there are not enough hours in the day to explain all the ways in which you are mistaken. Accordingly, I am going to have to resort to bullet points:

• Re: burden of explanation, you ask “should it be the person who is trying to expand our constitutionally limited government or the person who arguing for the status quo?” To me the employment statistics I mentioned in my last email are proof of the need for the legislation. If the status quo, 14 million unemployed, is acceptable to some, then so be it.

• Re: pointlessness of the bill. You write “Those are costs, and those are harms. In any event, even if there is no harm I submit that no good plus no harm still equals bad law.” I see a good in the bill. As I wrote before the good is that unemployed folks won’t be categorically barred by policies against hiring the unemployed. I also see this good as outweighing the amorphous harm that view as compelling reasons to be against the bill. You, apparently, don’t see this as a good and thus the formulation of your “no good plus no harm equals bad law.”

• Re: whether the bill will create jobs, I imagine other parts of the bill do that. It seems to me that this part is designed to create job opportunities. One might say that the bill is poorly titled because of the inclusion of this particular part of the bill. That is a reasonable position. Since I didn’t name the bill, I don’t feel obliged to defend it. As we both know, lots of bills have stupid, or misleading, titles. Your hypothetical in which the employed new hire creates a job opening by leaving his old position assumes that the old position doesn’t have a “no hiring the unemployed policy.” As the reason the bill exists is that apparently these policies are widespread, I am not sure that is a valid assumption. That said, if the hiring policy at issue isn’t common, then I agree that Congress should not ban the policy.

• Re:  Costs, you mention the added costs involved would be employers having to consider, and interview more people. If I were I more of a wag, I would suggest that the next time I meet an overworked HR person will be the first time. But, assuming that there are overworked HR departments for whom the cost of considering additional applicants is burdensome, this particular cost is outweighed by the previously mentioned 14 million people looking for work. With respect to token interviews, we don’t know how many interviews will be token and how many will be non (un?) token. As such, I don’t have a sense of what this cost would be. Since I assume the new law will result in at least some of the long-term unemployed being hired, I think that benefit outweighs the cost. 

TORVIK:  Well, that seems like enough conversing for one day, at least with you. Since I get the last word this time, let me just end by saying that it seems like we agree that this provision does not fit nicely into the context of a jobs bill. Instead, this is just something that liberals like. This leads me to suspect that its inclusion in the jobs bill is merely a negotiating tactic, and it is something that Obama intends to concede on in exchange for some concession from Republicans.