A number of bloggers (for example Mark Perry, Catherine Rampell, Greg Mankiw) have picked up Bar and Zussman's recent paper on Partisan Grading. (Downloadable here, forthcoming here)
Bar and Zussman take data on student grades, student SAT scores, and professor political affiliation, and find that:
…student grades are linked to the political orientation of professors: relative to their Democratic colleagues, Republican professors are associated with a less egalitarian distribution of grades and with lower grades
There are some problems with the study – it is based on just one university, only 44 percent of professors at that university could be identified on local voter registration lists, and of those, just 27 were Republicans, so two or three curmudgeonly professors teaching large undergraduate classes could skew the results. Also, political affiliations were correlated with subject taught, with more Republicans being found in the natural sciences, but there were too few Republicans to completely control for subject/grade distribution interactions.
What I find remarkable, though, is how few people have picked up on the study's conclusion:
To the extent that the application of objective standards is an important university goal, policy makers should consider limiting the discretion professors enjoy when it comes to grading and making it more difficult for them to use student characteristics as factors in the grading process.
But if control over grading is taken out of the hands of professors, it will be put into the hands of someone else. And that someone else will also have political leanings.
Consider, for example, standardized exams.
Those aiming to boost student achievement can set easy standardized exams. Those aiming to separate out students will set an exam with a high variance. (That can be done by having questions at several different levels of difficulty – some questions only A students could be expected to answer right, some questions for B students, and so on.) "Back to basics" rhetoric has no substance. The past does not provide us with absolute standards we can use to design exams, because what students need to know shifts as the world changes.
Another way of limiting professorial discretion is to impose grading norms, such as 20% As, 30% Bs, etc.
Again, the person who sets those norms will be either liberal or conservative, and – if Bar and Zussman are right – their views will be reflected in the grading norms that they set.
(Grading norms also introduce opportunities for collusion. If all students can implicitly agree to ratchet back their work effort, they can achieve the same grades with less work.)
Another serious objection to fixed grading norms is that we don't know what the optimal distribution of grades looks like.
Human capital – a theory advocated and developed by conservative thinkers such as Gary Becker - argues that education teaches people useful skills. Educated people are more productive because of what they learned in school. Driver training is an example of a good example of human capital type education: students take the class, and become better drivers.
Now, if a driving school taught all its students to be excellent drivers, and each one achieved a perfect score on their driving test, the school would be doing a good job.
University is not driving school, and grades lower than an A are useful in creating incentives for students to learn the material – to invest in their human capital – and to mark out students who have not fully acquired the human capital the course was designed to impart.
But, according to human capital theory, as long as every student has complete command over the course material, there is nothing wrong with giving every student an A.
So why has Bar and Zussman's work had such resonance in the blogsophere?
Last month the New York magazine wrote "The notion that a college degree is essentially worthless has become one of the year’s most fashionable ideas". People are talking about the book Academically Adrift (technical version downloadable here), which argues students critical reasoning and other skills are improved little by university studies. And then there's Charles Koch's arrangement with Florida State University, that will allow his representatives to screen hires for a new program promoting "political economy and free enterprise."
It doesn't matter whether or not Bar and Zussman's findings are robust or generalizable. Right now, they fit the zeitgeist.
(Tuesday's Economy Lab post looks at Bar and Zussman's findings in the Ontario context).
Worthwhile Canadian Initiative 
Here is a list of 150 economists who quite clearly can’t distinguish between their personal political views, and economics. If, as they claim, a high level of public spending does an economy harm, they need to explain why various European countries (Sweden in particular) have for a long time had a much higher level of public spending relative to GDP than the U.S., but despite this, have lower debt and deficit levels, and comparable standards of living.
But of course they can’t explain.
150 economists: http://www.scribd.com/doc/48737188/150-Economists-Call-for-Spending-Cuts-to-Boost-the-Economy
Re Sweden, see http://www.sweden.gov.se/sb/d/3212/a/139225
Good point about the study’s policy conclusions.
I would have thought that the obviously tempting conclusion (at first glance) would be to bell up the grades of righty profs, and bell down the grades of lefty profs. Though that conclusion too would create difficulties. Lefties and righties pretending to be righties and lefties. Or righties grading even tougher, and lefties even easier, to compensate.
It also fits into the meme of universities being one big lefty cathedral, with lefties fudging standards to promote their students (and colleagues). (Not a meme I would immediately dismiss 100%, even though I think it is more commonly false than true.)
Ralph, it’s interesting to look at these names. Looking at them made me wonder – are Republicans more likely to teach 1st and 2nd year courses? I can think of plausible reasons why they would be. At least in economics, the conservative account of the economy (competitive markets, pareto efficiency) is more suited to Econ 1000 than the liberal account of the economy (asymmetric information, social choice theory, complex dynamic equilibria). A Republican who enters the Liberal world of the elite Ivy leagues must have to really enjoy his job – i.e. like teaching. Also, to the extent that the conservative world view is more compatible with the exercise of authority (“Be quiet or leave”) we might expect to see more Republican profs in lower level courses.
The authors included 0-1 dummies for instructor gender, subject, and course level. However their crucial finding was that the slope of the grade distribution was steeper with Republican profs. So to really check for robustness, what you would need to do is run separate regressions, e.g. by course level (1st, 2nd, 3rd, 4th year, graduate), or do a course levelrepublicanSAT score interaction. It could be that the authors results are driven by the steep slope of the grade distribution in 1000 level courses.
Nick, “even though I think it is more commonly false than true” – speaking of lefties and righties pretending to be righties and lefties…
i always thought that many things were, officially or not “graded to a bell curve” – whereas it sounds like republicans here grade to a barbell curve, or a flat line.
there was some controversy in ontario last year in that pass/fail rates for drivers licenses varied widely – high failure rates in cities, very low failure rates in rural test centres – there was outrage at this, and of course, many people flocked to the places where it is easier to pass, even if it means a couple of hours driving there and back.
btg: “whereas it sounds like republicans here grade to a barbell curve, or a flat line”
If you look at the lowest SAT score students, they’re getting Cs and Bs from Republican profs, and mostly Bs from Democrat profs. (Remember this is an elite university, however, so even these low SAT score students are probably still pretty bright). If you look at the highest SAT score students, they’re getting Bs and As from the Democrat profs, but mostly As from the Republican profs. I’m not sure if that’s what you mean by a barbell curve or not.
As an aside, I just booked a driving test for someone, and we definitely had a conversation along the lines of “book me in at test centre A, the failure rate is much lower than in test centre B”. But the Ontario testing service has been privatized, so this must be the outcome of an efficient, competitive market process, no?
One of the profs I respect most taught Econ History (presumably because when he was younger it was Econ Current Affairs). We were a small class and by mid-October he made the pronouncement that we, unfortunately, were likely the least competent group he had ever taught. He didn’t think any of us would get an A and probably no B+’s either. He didn’t say this to discourage or challenge us, but rather as a matter-of-fact statement and somewhat appologetically. We were being graded against the material he was teaching relative to every student he had ever taught.
As for his political leanings, I think he took the view that promoting a specific government policy was only as effective as the government’s ability to understand and properly implement it.