How are you handling it? Funding my students keeps me awake at night. I dropped my target steady state group size, which actually works better with my management style anyway. While I always want to recruit quality, a mistake in smaller group and with less funding cushion is much, much more painful. I am being VERY careful with who I am taking. In this case, I am attempting to select for enthusiasm, work ethic, and scientific curiosity (better predictors of success than GPA or pedigree, in my experience). Unfortunately, this also means I am only taking students with research experience, because I can't afford for someone to try it out for the first time in my lab and decide it isn't for them.
I am fortunate, because in my field it is possible to do research without lab techs and postdocs (I have exclusively students right now). Postdocs in my field last 1-2 years, so a one year contract with a possibility to renew is the norm, which is helpful in the current funding climate. At ProdigalU, postdocs are still more expensive than students, but students come with a 5-6 year time commitment. This compares poorly to the usual 3 year timeline on grants in my field. It is definitely possible to start a student on a project and then run out of money part-way through the PhD. I worry deeply about this, but so far, I have been able to string together related projects in such a way that my students don't get disrupted.
One may say that there are too many PhDs, and that reducing the number of PhDs is a feature, not a bug of the current funding situation. I don't doubt that this is true in some field and specialties. That said, my students are finding jobs that use their degrees (though it has taken up to a year for some). As is the norm in my area, most of my students are interested in industrial positions, not academia. While I of course think my students are really good, I would think that if there were too many PhDs in my field that some of my students (even if very good) would be unable to find good jobs and would move on to other things. This has not been my experience so far. I actually don't think there is much of a connection between the demand for highly trained workers and support for their training. Aside from the long lag time due to the time to degree, companies can always import trained people from other places if they have unmet needs. And universities will always be able to fill paid student positions as long as the money is there, regardless of whether the students are employable at the end. If there is an actual interest in reducing the overall number of PhDs, I would think that a strategically planned reduction (that targets overpopulated areas) would be much better than random chance, which is what we are getting with the current system.