I thought it was reasonably good session in class today, but I hurried a bit at the end. So here are a few things to consider that I didn't get too.
(1) One question is the frequency at which performance is measured. You might consider that from the point of view of how the blogging is graded in our course. An alternative to how we are doing that would be to have each post get a grade. Is that alternative better or worse? Can you explain why you feel that way?
(2) A second question, related to the first, is whether the evaluation of performance has a direct impact on performance (the incentive effect definitely exists but lets call that an indirect effect). The direct effect is due to the person feeling "on stage" and therefore less relaxed. So thinking about it this way, a question is whether performance is better when the person is relaxed or better when the person feels on stage. (The answer may depend on the person.)
(3) A third question comes from noting that pay for performance provides extrinsic reward or punishment. When there is intrinsic reward, meaning the person finds the work engaging for itself, what is the consequence of adding extrinsic reward on top of that?
These three questions are considered implicitly in this entertaining video - an animation with the voice of Daniel Pink. I encourage you to watch it. Then read my critique, because I don't think he gets the story quite right.
(4) There is then the question when you have many people doing the same sort of work in an organization if you want to use performance pay as a way to differentiate them by their productivity, or if you instead want to pay them equally because most of the observed productivity differences can be attributed to measurement error and equal pay promotes an esprit de corps among these employees that itself provides a productivity boost.
This makes the consideration of pay for performance considerably more complex than the way I left it in class, but also considerably more interesting.