« Opportunistic Federalism | Main | Althouse? »

Friday, May 13, 2011

Law School Curves

I am interested in the number of grades below B- your school requires you to give in classes with mandatory curves.  At Hastings, we recently changed our curve to require professors to award 12% to 17% of the class grades below B-.  I am curious if this is out of step with what your school does.  Please let me know in the comments: tell me your school (assuming you have a mandatory curve) and what percentage of students in a class must receive a grade lower than a B-.  If your school has gotten rid of the C, I'd be curious about that as well.

Posted by Ethan Leib on May 13, 2011 at 12:09 PM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Law School Curves:


" It would be ridiculous, of course, to claim that a "C" means anything special or absolute."

In GU's defense, I think the letters A, B, C and so on obtain some meaning apart from whatever grading systems they fall under. If for example a school offer a curve with a max grade of F+ and a low grade of F-, I think people would still look askance at straight F's, even if the small print on the back of the transcript explained that straight F's reflected a decent grade. These letters of course should not be interpreted without regard to the interpretive guidelines, but most readers will associate the letter with something bad.

Additionally, some law schools (UChicago, I think) have recognized that letters are "loaded" with meaning and thus use weird grading systems. And some have suggested grading systems of A* through A*****. Although these suggestions were of course made tongue in cheek, they reflect that the letters "A" and so on do carry some independent meaning in some quarters, even though they probably shouldn't. To consider an even sillier hypo, if a school awarded the grade of "Genius" to the top 50% students and the grade of "F$%king Idiot!" to the next 50%, I think it'd be hard to ignore the general meaning of the chosen phrase, even if a special definition were offered that showed that the person was not, in fact, an idiot.

On a side note, I was reading an autobiography of someone who attended Harvard law school in the 1930s. The author explained that in those days, one did not apply in advance. Instead, the author appeared on campus with his diploma from an Ohio college and immediately enrolled in classes starting the next week.

Posted by: andy | May 16, 2011 1:52:35 AM

"GU": grade medians have absolutely nothing to do with how things used to be graded, nor was my "Remember when C used to be average" meant to convey something like "Law students nowadays have it so easy." It would be ridiculous, of course, to claim that a "C" means anything special or absolute. As grading curves recognize, grades are all relative.

However, just because the overall quality has increased doesn't mean that there's no longer any distribution of quality. If 60% of the students make it through with an "A" average, that isn't a problem because there couldn't be that many students who are "that good," but because it's almost certainly a fact that there's an upwards distribution there that's being truncated by the top of the scale.

Thus, grade inflation makes the more average students look the same as the high-achieving students, because there's little-to-no way for the high achieving students to outperform their classmates.

Posted by: Andrew MacKie-Mason | May 15, 2011 7:42:31 PM

Grades serve a second function, at least at non-elite schools--they are predictive of success on the Bar, which is another measure that law schools need to worry about. A curve, particularly a curve with mandatory low grades, thus needs to serve a weed-out function. A student who slides through three years with a 2.01 GPA because professors (especially in upper-level classes) do not want to give C's--and really does not want to give out anything below a C--is statistically less likely to pass the Bar. A curve with mandatory highs and lows enables a school to help identify its top students while also weeding out students who are unlikely to pass the Bar.

Posted by: Howard Wasserman | May 14, 2011 10:31:07 PM

Let me presage this by saying it breaks my heart every time I have to assign a C to a student's work, and I can understand why one would never want to do it under any circumstances. But should one give in to that generous instinct?

A lot depends on what grades are intended to do. If you only need to signal to the world who your top 10 percent are, you can hand out 10% As and 90% Bs. Students will be happier, and employers looking for the top 10% can find them. Of course, employers hoping to distinguish between the 20th percentile and the 80th percentile will be hard-pressed to do so. I'll hazard a guess that there aren't many schools where there is no real difference between students in the 20th and the 80th percentiles, but your experience may tell you otherwise.

You thus might need to--or employers might prefer that you--identify students in some bottom tier, however broadly or narrowly defined. This might also help you identify students who need extra help preparing for the bar, or to determine whether there might be groups of students systematically left out in the cold by current exam procedures, or to give legitimate signals to a certain percentage of the class that they might want to think seriously about another line of work. (It may seem laughable that a law school would ever encourage a student not to pay for those extra two years of education, but that can never change if you lose your mechanism for identifying those students.) The 'C', that most cruel and dastardly grade, might just send one or more of those valuable signals.

Of course, if it behooves you to make sure that employers can't distinguish between your 20th and 80th percentile, you can give everyone a 'P' and call it a day. Or an LW for Lake Woebegone - perhaps we've finally reached that glorious day when all law students really are above average.

Posted by: Jake Linford | May 14, 2011 3:31:00 PM

@Andrew MacKie-Mason

Law school admissions are incredibly competitive these days. Most alumni would not be admitted to their alma mater today (or at least would've had to work a lot harder than they did to be admitted, but still, some simply wouldn't be admitted). "When I was at Harvard, x% of the class got C's" is beside the point; you probably wouldn't get into Harvard today.

Ignoring ugrad grades to control for grade inflation, LSAT scores are undeniably higher now than in the past (and percentage of applicants admitted is lower). I'm not inferring causality, i.e. I'm not arguing that better students are the reason for grade inflation. But it is not unreasonable for objectively better students to get better grades.

Everyone wants to pile on the "soft" students these days who never receive a bad grade in their life due to grade inflation. How many of these older critics could handle the crushing burden of perfection foisted upon kids today?* Did you know that scoring in the 95th percentile of the LSAT gives you virtually no chance of being admitted to the top 6 schools and iffy chances at schools 7-18 (unless you're black, hispanic, or native American)? It's kind of crazy that being in the top 5% could mean relegation to a "strong" regional school.

The pool of quality students goes much deeper than it did in the good old days of "real" grades. Your typical UCLA grad would've been at Penn or maybe NYU 10 or 15 years ago, and the Hastings grad would've been at UCLA. Rising medians, therefore, make some sense.

*I know that MacKie-Mason is young. Most grade inflation critics are not.

Posted by: GU | May 14, 2011 2:54:50 PM

Remember how C used to be average?

Honestly, I'm still not sure that MK's comment isn't satire. I hope it is, at least.

Grade inflation is only a disservice to those students who actually deserve - and work hard - to do well.

Posted by: Andrew MacKie-Mason | May 13, 2011 7:36:23 PM

To MK's comment: In my reasonably extensive experience, including at elite institutions, there are meaningful differences among the papers and exams marked at B- and below -- and it is not excessively difficult to avoid the basement for three years. Absent required distributions, it is evidently difficult for faculty to avoid a tremendous clumping around B+ that disguises very substantial differences in quality. The illusion that one can eliminate "this part of the curve" without repercussions for students is convenient, but still an illusion.

Posted by: Ani | May 13, 2011 5:31:18 PM

At Illinois, we have a strongly encouraged mean of 3.2 to 3.4 for upper-level classes and 3.2 for first-year classes, where at least 20% are strongly encouraged to get some type of A. There are no other distribution requirements, but the full grading scale is available.

Posted by: Christine Hurt | May 13, 2011 4:28:54 PM

These curves are brutal. Anything below a B- is such a terrible stain on the poor student's record, at least for the most coveted jobs. (It shouldn't be, but it is.) And at 17%, such grades will be hard to avoid for 3 years. Ouch. Still, I suppose I've seen students make up for it and get stellar jobs regardless. But I'm against this part of the curve. Limiting As and A-s makes sense, but requiring so many Cs in today's day and age? Eesh.

Posted by: MK | May 13, 2011 2:36:30 PM

At FIU, our curve is as follows:

Foundation Classes: 10-15% C- or below, 50 % B- or below.

Upper-level Classes: 5-15 % C- or below; 45-55% B- or below.

Posted by: Howard Wasserman | May 13, 2011 1:39:04 PM


That's the curve at GULC. 5-10% B- and below, no mandatory Cs.

Posted by: jsl | May 13, 2011 1:33:50 PM

The comments to this entry are closed.