http://tvhe.wordpress.com/2007/10/31/was-jesus-an-early-applied-economist/

Huh??? I never expected a Catholic to acknowledge constraints on God’s creative powers 😛

Indeed, but by increasing its effectiveness you also increase the use of the policy, and its ability as a signaling device – which is where I think there is the potential for movement to a suboptimal outcome.

“Matt, I don’t doubt you can model a situation where compulsory probation is bad compared to “perfect information, negotiable probation”, but I really think the question here is whether it is bad compared to “imperfect information, no probation”, which on the basis of Nigel’s comment, is the status quo.”

Indeed I completely agree. However, as I have said it is not obvious which one of these outcomes is better – it depends on what one is further away from the 1st best outcome. Remember, we are in a second best world here 😉

This does away with all of that, but only for the first 90 days. That is the principal advantage.

Yes, definitely.

“I realize that assessment of this policy hinges on a number of aspects for which their is no quantitative data yet, but my instinct is that workers hiding their type (hard-worker or shirker) is a bigger inefficiency than employers hiding their type (good or bad employer)”

Agreed – however, my comparison many depends on employees with multiple dimension that depend on the existence of probation – rather than an employee and employer dimension. I definitely do not assume bad employers.

As Rauparaha says my idea needs a lot of work – however, I do get the feeling that we can have “excess probation, a feeling I just can’t shake. As a result, at some point I must model it 😛

I realize that assessment of this policy hinges on a number of aspects for which their is no quantitative data yet, but my instinct is that workers hiding their type (hard-worker or shirker) is a bigger inefficiency than employers hiding their type (good or bad employer). I base this on the fact that the returns to securing a good job through hiding your quality seem a lot higher than the gains from constantly firing and rehiring (given that new workers normally cost the firm in net terms).

In the law they are passing it is illegal to have a rolling probation contract – you can’t hire the same person with a probation clause twice in some space of time.

“I thin we really need an endogenous explanation of why they dont’ like the probationary period in order to make this model fly”

Agreed – I am just assuming that “preferences differ” – which is of course a cardinal sin in economics. However, for practical purposes its just how I feel on the inside when looking at this policy ya know 😛

It’s illegal to hire someone on a fixed term contract if you don’t have a good reason for the fixed term. I can imagine that there are benefits to having people on rolling probationary contracts since it gives you a lot more flexibility in dealing with your workforce. That probably has to be balanced against the specific investment you’ve made in the worker, as you point out.

“Averse to probation to not averse.”

The question is, why are they averse to probation? Why would hard working types not want to credibly signal their hard working nature? Admittedly, there is probably uncertainty over whether they keep the job, regardless of their effort. But that’s an issue that can be sorted out with the right contractual incentives.

I think we really need an endogenous explanation of why they don’t like the probationary period in order to make this model fly. I’m not panning your model, I just think it needs further development before you publish it in the AER 😉

Lazy to Hard working and,

Averse to probation to not averse.

As firms can’t observe the characteristics, they may presume that agents who don’t want a probation period are lazy – when in fact some of them are just averse to probation. In this sense, people who are hard working by averse to probation have to choose between taking a contract with probation or not at all, even though net welfare would be maximised if no probation clause existed.

In the extreme, we can compare a situation where everyone is on probation to a situation where no-one is. I don’t think it is clear which situation is socially optimal.

Because of signaling, it is possible that allowing probation could lead to an equilibrium where everyone has to accept probation in their contract, even though with full information the employer and employee would not agree on a contract with probation.

As a result, if the benefit to employees is sufficient small and the cost to employees in the right range, it could be socially optimal to ban probation altogether.