Thanks, Mark. As you note, the debate on whether active manager alpha exists is still raging--but that's not actually the question I was asking. The existence of active manager alpha is a necessary but not sufficient condition for expenditures on manager search to be worthwhile: that is, if active managers can produce alpha, then it still has to be shown that manager selection can identify, ex ante and after internal as well as external fees and expenses, those managers who will produce alpha in the future.
The main research I have seen on this question is the paper "Picking Winners? Investment Consultants' Recommendations of Fund Managers" by Jenkinson, Jones & Martinez (http://onlinelibrary.wiley.com/doi/10.1111/jofi.12289/full), which is slated for publication later this year in Journal of Finance. The authors conclude "we find no evidence that these recommendations add value, suggesting that the search for winners, encouraged and guided by investment consultants, is fruitless."
That's pretty damning for the entire manager selection exercise, regardless of whether active management alpha exists.