I presume that those without jobs are significantly more likely to accept a job offer than those who are comparing it against their current job given that all other factors are the same. Given this, I wonder whether HR department's disinterest in the unemployed is rational. Let's formulate the problem:
Presume the following:
1. The cost of assessing the qualifications of both the employed and the unemployed is identical and that the company can do this with perfect precision. Calls this cost I.
2. The position is unique in the organization, so only one job offer may be extended at any given time.
3. There is a known, positive cost C for each day the position goes un-filled.
4. An unemployed person takes N days to evaluate a job offer letter and an employed person takes M. I would presume that M > N given that the opportunity loss of an employed person is greater.
5. The probability that a candidate accepts an offer is X if unemployed and Y if employed. Obviously, I would presume that X > Y
6. The probability that a candidate is found to be qualified after the interview process is A for the unemployed and B for the employed. For the sake of argument, presume A < B.
7. It is known with certainty and at the start whether any given candidate is currently employed.
So, for some set of costs I and C and probabilities (N,M), (X,Y), and (A,B) I think one could compute whether it is more rational to prefer the employed or the unemployed. Of course, I am ignoring all goodwill-esque costs (bad reputation as an employer, etc.). After writing this I realize that (1) I don't even know the proper nomenclature for expressing this and (2) I can't do the math anymore to solve this (there's a dynamic programming aspect, right?).
Presume the following:
1. The cost of assessing the qualifications of both the employed and the unemployed is identical and that the company can do this with perfect precision. Calls this cost I.
2. The position is unique in the organization, so only one job offer may be extended at any given time.
3. There is a known, positive cost C for each day the position goes un-filled.
4. An unemployed person takes N days to evaluate a job offer letter and an employed person takes M. I would presume that M > N given that the opportunity loss of an employed person is greater.
5. The probability that a candidate accepts an offer is X if unemployed and Y if employed. Obviously, I would presume that X > Y
6. The probability that a candidate is found to be qualified after the interview process is A for the unemployed and B for the employed. For the sake of argument, presume A < B.
7. It is known with certainty and at the start whether any given candidate is currently employed.
So, for some set of costs I and C and probabilities (N,M), (X,Y), and (A,B) I think one could compute whether it is more rational to prefer the employed or the unemployed. Of course, I am ignoring all goodwill-esque costs (bad reputation as an employer, etc.). After writing this I realize that (1) I don't even know the proper nomenclature for expressing this and (2) I can't do the math anymore to solve this (there's a dynamic programming aspect, right?).