The lack of exclusionary ads demonstrates this point. Tuesday's New York Times article on the proposal by Robert Pear says:
The Equal Employment Opportunity Commission has received reports of such advertisements but has no data to show how common they are.
If such ads were common, then such data would be readily available.*
Will This Proposal Create Jobs? (No.)
But let's pretend, just for argument's sake, that discrimination against the unemployed really does occur with modest frequency. Let's say that one out of ten employers discriminates against the unemployed -- which is almost certainly a much higher percentage than the portion that really does. Let's think about how this might work.
So imagine that Joe is unemployed and applies for a job at XYZ Corp. It turns out XYZ likes to discriminate against the unemployed, so a current employee at ABC Corp, Sheila, is hired instead. Now for this to be legitimate discrimination, we must assume that Joe was at least equally qualified for the job that Sheila got. So they must have similar experience levels. Now that Sheila has left her job, ABC will need to hire someone. Joe will be a prime candidate for Sheila's old job, and it's statistically very unlikely that Sheila's previous employer is also a discriminator.
This story of labor market turnover shows that, even if this sort of discrimination does occur sometimes, it shouldn't cause unemployment to be higher or the duration of unemployment to be much longer than it would have been anyway. At worst, it causes slightly slower wage growth for unemployed Americans. Whenever a new job opening is filled by a current employee, another opening is created, which can be filled by an unemployed person. This discrimination does not actually kill jobs. At most, it shuffles the labor market outcome.
Will This Proposal Kill Jobs? (Probably.)
Instead, the rule to ban discrimination against unemployed Americans could actually slow hiring. Imagine that you're an employer that is considering bringing on another employee. At this point, you figure you have the money to hire someone now, but you could probably wait six months or a year -- consumer demand in the short-term looks weak.
But you hear of a new law that will forbid you to discriminate by not hiring unemployed people. You would never discriminate anyway, but now you worry about the get a huge influx of unemployed applicants you'll get -- after all, millions of Americans are unemployed. And there's definitely a chance that, if you don't hire one of them because a better already-employed candidate is out there, then you'll get sued.
A potential lawsuit is not something you want to deal with. So you figure you will hold off on hiring until unemployment declines a little further and the risk of litigation declines.
Did you see what just happened there? By creating a law to forbid a practice that only a small minority of firms take part in anyway, everybody else may change their behavior. The decision to create a job just took on additional risk -- risk that's much higher when the number of unemployed Americans is very great. As a result, firms now have a disincentive to hire when unemployment is high. Obviously, this is precisely what we don't want.