Productive head drives others to raise their game

One solution is to appoint heads of department that the academics approve of. It is unlikely that a set of colleagues who are doing good research will be asking for a mediocre achiever to play the role - what would he or she know about research? The other real issue is to start moving away from any type of assessment that depends on putting numbers on the uncountable. To use a different example, many academic editors of PLOS ONE received a friendly email a couple of days ago informing them about how many papers they edited last year, how many reviewers that they had contacted accepted to complete their review, how timely they managed the process, what percentage of articles they accepted, etc. and they could see how they "rank" against the average academic editor but also against the journal's "optimal policy", based on "best practice". I suppose a natural reaction is to start comparing one's own numbers to what is happening, and to reflect upon what the numbers mean. What stroke me just a few moments later in this thought process, is that the most difficult aspect of editorial work, the agonising to treat the papers received with due care, the finding of the best reviewers, the evaluation of the authors responses and - on controversial occasions the most difficult aspect - the final decision, were not actually being evaluated (not possible to do so with numbers, or perhaps I am just not good enough in mathematics).