I'm hoping that my method will actually ameliorate much of this. By starting the person part-time, perhaps their familiarity with the project builds "faster" over a unit-work-time than it would otherwise (I've personally found my comfort level with a project to be in some part a function of calendar time). In this way, I'm also not spending a lot on the candidate search itself.
The idea is to spend less time on much-higher-yield methods of finding people who are better suited to the job crossed with people who are happier with the environment of the job. It is about avoiding turn-over, but not in this perverse way of "ignore the problem because the alternative is worse", but by trying to avoid the hiring of a bad worker in the first place.
In this system, you'll never see a job notice from me. You won't know I have a need for an employee unless I ask you to come work for/with me. You'll only ever get a job offer from me if A) I know you're not an idiot, and B) I know you're available. Yes, that means I have to know the person very well already. If I have an opening that needs to be filled and I don't currently know of someone that fills it, then it is likely to be a long time before it gets filled, if I do nothing.
But then again, it's not like sifting through resumes is providing quality applicants. The resume and job interview are so poor of a selection process that I'm willing to try just about anything else.
There is also an element of respect for the worker involved in what I'm doing. I don't think I have a right to monopolize a person's time. I think people should have the time to be engaged in many pursuits. I want it to be up to them to choose how much they work. I want them to have the ability to find other work as well. The idea being, people who are happier in their work-life balance will be more productive.
Of course, we'll see.