Oh, please. Yes, Hamermesh & Parker's study shows that faculty rated more physically attractive receive higher end-of-term instructional ratings, and they try to control for other factors associated with ratings of attractiveness to isolate the effects of physical beauty. But they write: "For example, what if students simply pay more attention to good-looking instructors and learn more from them? We would argue that this is a productivity effect—we would claim that the instructors are better teachers. Others might (we think incorrectly) claim that the higher productivity arises from students’ (society’s) treating them differently from their worse looking colleagues and is evidence of discrimination." Setting aside the endless debates about whether course evaluations say anything about "productivity," what's the implicit definition of teaching on which Hamermesh & Parker rely? I prefer to think of teaching as an intentional and moral activity. Students paying attention, and even learning more, simply because a teacher is hot -- that's not good teaching.

Of course teaching is an intentional and moral activity. But a unique feature of teaching as a profession is that it is a two-way street. Effective teaching requires active participation on the students' end. Much of what defines a good teacher is an ability to cultivate that participation.

Take two teachers with exactly the same degree of intention and morality (and knowledge, communication skill, etc..). If the more attractive of the two gets better results from students simply because of his/her looks, then he/she is more productive. That's actually the definition of productivity (output per units of input). Sure, the less attractive of the two can (and often does) surpass the more attractive by other means. But in that case inputs differ.

In their paper (which you can tell in their writing they write tongue-in-cheek), they cite the example of TV journalists. There's a reason that (even reputable) TV news programs choose attractive anchors. Two TV anchors may have exactly the same skill and knowledge of current events, but if the more attractive is better able to capture the audiences attention--which is an important component of the anchor's job--he/she has an edge.

An unfortunate proposition, but not at all unreasonable...

P.S. All that said, teaching evaluations are definitely not the best metric of teaching effectiveness, and I think Hammermesh and Parket admit as much.

Sorry, dumb economist, I don't buy it -- "it" here being the equation of productivity and good teaching. Yes, an attractive teacher may get better results due to his or her looks, and we can label that as being more productive in the output-per-unit-of-input sense, since looks aren't viewed as a relevant input in your basic teacher value-added model. But if that greater productivity is due just to looks, and not to the intentional practice of the teacher, it's not due to better teaching.
Let me tackle the discrimination storyline, too. Suppose that students are more likely to pay attention to a white teacher than a black teacher who is otherwise equally skilled, and that the white teacher's results are therefore better. Would you really be comfortable calling the white teacher a better teacher just because students are more likely to pay attention to him based on his race?

Skoolboy - I'm glad you picked this example of white vs. black teachers, as a pretty solid group of studies has found that same-race teachers can make a difference for students' academic outcomes. In fact, this finding (or the suspicion that teacher race matters) is one reason why there has been such a push for more minority teachers in the classroom.

Here I wouldn't say a black teacher is a "better" teacher in the sense of intentional practice, but if she is better able to connect with students as a result of her race or cultural background, then by definition she is more effective than an equally capable white teacher.

If race can matter in the classroom (and no one pushing for more minority teachers would call this discrimination), then it's not a big stretch to think that other personal characteristics not directly tied to practice--good humor, even looks--can matter.

Hmm. Apparently I'm not being successful in conveying what I see as the complexities in the white-v.-black teachers example. (Probably because these dumb economists can't see how attractive skoolboy is.) More minority teachers is like apple pie, motherhood, and the flag. But let's suppose that a majority-white school district declines to hire a minority teacher for a high-paying teaching position, on the grounds that she wouldn't be able to connect as well with the predominantly white students as an equally capable white teacher. Isn't that a problem?