College Grade Inflation

Apparently the news of the week is that the letter grade "A" is now the most common.  Mark Perry has more on college grade inflation.

I am actually a fan of the grading system at Harvard Business School when I was there.   15% of the students in each course get the top grade (category I) -- no more, no less.  10% get the bottom grade (category III) -- again by rule, no more and no less.  All the rest are in the middle.  It effectively acknowledges that for most folks, the point is to demonstrate you have satisfactorily learned the course material, while still allowing folks to distinguish themselves on both ends.  Budding young executives who complain that it is unfair to automatically "fail" the bottom 10% of each course are reminded that this is exactly how many Fortune 500 companies run their HR systems, seeking to constantly weed out the bottom 10%.

Update:  The argument usually is that students need high grades to compete with other kids from grade-inflated schools in the marketplace.  I just don't think this is true.  Colleges themselves deal with this all the time in admissions.  When they get a high school transcript, attached to that transcript is a fact sheet about the high school that gives its distribution of grades.  That way the recipient can discount the GPA as appropriate.  Every company doing hiring should demand the same of colleges.

Here is a personal anecdote.  My son Nic's school grades hard.  Something like 2 kids over the last 2 decades have graduated with a 4.0.  One could argue my son's grades could have been higher at another school, but knowledgeable consumers of high school GPA's know how our school works and we have never felt he somehow was at a loss due to the school's grading policies (but Oh God can type A parents fret about this incessantly among themselves).   [edit:  took out brag about my son.  Nothing more boring than other people bragging on their kids.]

  • SB7

    What's the average class size for HBS? I think I would like this a lot in courses with >50 people, but I think it would be a disaster for a 15 person seminar.

  • dsfan


    At least now (I was class of 2011), first year classes are about 95 people, and they still use the same distribution (although I think it's 20-70-10 now). Second year classes are more variable, but I believe seminars under 25 people are not required to give any "3s" for the reason you describe.

  • pegr

    Ah, but since the grades are arbitrary, at least between schools, they are effectively meaningless as a basis for comparison. In fact, you could have learned all the material, but since you didn't stay after school to clean the chalkboard, your grade is lower than your intellectual equal.

    School isn't about learning. It's about submission to authority. Learning something is a happy coincidence. As the Old Man put it to me, the purpose of school is to get a degree. No more, no less. It's the very antithesis to learning.

    It's amusing to me to interact with recent college grads, as they think they know everything. It takes a year or so for them to learn that they know nothing, at least for the smart ones.

  • Ted Rado

    Grade inflation is nothing new. There was a piece on the subject in the Wall Street Journal several decades ago.

    I commented many years ago to the HR head that most of the young engineers sent to me for interview had a 3.5 GPA or higher and asked what was going on. He acknowledged that grade inflation was rampant.

    When I was in engine school at Ohio State in the late 40's, the all-university men's average was (as I recall) 2.3. In engineering it was 2.1. The engineering mechanics dept was 1.9. (The dept of "international studies" was 3.75 - so much for studying something wortwhile). I took a course called "advanced strength of materials" as an elective and got a C, which was the second highest grade in the class. Almost everyone got a D or F. The course was very difficult (lots of differential equations, and a thorough understanding of engineering mechanics was a must). If you passed, you were supposed to understand the subject, not just parrot the problems.

    Giving grades on the 15-75-10 basis in nonsense. If everyone in the class truly masters the subject matter, give them all an "A". If nobody understands, flunk them all.

    Many try for grades rather than learning the subject matter. They memorize every problem in the book and get an "A". I knew many who went this route. I never encountered a problem in industry that read right on a problem in the text books. One had to understand the underlying principle and apply it to a unique situation.

    There should be a comprehensive exam at the end of the study program which tests one's true understanding rather that their ability to memorize. The current grade system doesn't do that.

    This idea of giving a certain percentage a high grade persists into industry. Raises are often passed out accordingly. If you fortuitously have an exceptionally good group of engineers, firing the bottom 10% is absurd. Same story if you have a bunch of weak people. You don't give the top 10% a big raise. Management is supposed to use their head, not rely on some arbitrary measure to rate people.

  • Will

    All I can say is in comp sci there weren't any give gimmes where I went. Some courses are harder than others, so some courses may have a higher percentage of A's than others. The hardcore math courses like advanced analysis of algorithms or cryptography were very hard (at least for me), and to pull off a C+ took lots of studying.

    Forcing the curving of classes simply doesn't make much sense, as you get further along in the program to 3rd and 4th year courses you've already culled out much of the herd that will never graduate and your left with mostly motivated and good students. Failing 10 or 15% of them just because doesn't make much sense.

    If a course is properly designed, it should be perfectly acceptable for everyone to receive A’s or F’s. It should be based purely on performance instead of the need to always fail a certain percentage.

    It seems kind of funny that many people would oppose quotas for hiring minorities, but would find it perfectly acceptable to have quotas for grades in a course. Many of the arguments against quotas for minority hiring would work perfectly well for arguing against mandatory curving of courses.

  • DoctorT

    College grades have steadily inflated for the past thirty years. I taught a course at Old Dominion University in Virgina nine years ago, and GPAs in every major but one had risen toward or exceeded 3.5 on a 4.0 scale. (The exception was the Medical Techology program I taught in that had maintained an average GPA of 2.7 to 2.9 for decades.)

    Our colleges have moved towards the left-wing "progressive" goal of equalization: 70% of high school grads go to college, college has become an extension of high school (with fewer classroom hours and more socialization), and the combination of watered-down courses and grade inflation means that most graduates will have GPAs above 3.

  • LTMG

    Grades and GPAs are interesting, but ...

    As an operations manager at various levels for 31 years, I have interviewed hundreds of soon-to-be college graduates, experienced individual contributors, and managers. What do I really look for in candidates?

    - Is the candidate's education relevant to the job requirements?
    - Has the candidate completed any project courses that had a required deliverable that he or she can describe in detail?
    - For an inexperienced and less experienced candidate, how has he or she used their education in practical ways?
    - Does the candidate consistently deliver quantifiable results important to whomever his or her internal and external customers are? This questions is certainly relevant for inexperienced candidates.
    - For the inexperienced or lesser experienced candidate, who paid for his or her college education? If the candidate worked to earn money for college education, how heavy was their work load? Is it possible to explain a lower GPA if the candidate was working and developing useful skills for working as a professional? "Work" in this context can include unpaid volunteer work or lowly paid internships.
    - Does the candidate have internal motivation and drive, or "fire in the belly"?

    The typical internal or external recruiter doing screening interviews is unable to assess these traits. Resume screening software is also unable to assess these.

    Grades and GPAs correlate well with intelligence. Intelligence is a very useful trait, but given a hard choice between hiring a very intelligent candidate or hiring a candidate with wisdom and lesser intelligence (lower grades and GPAs), I prefer the latter.

  • Russ R.

    That's why standardized tests exist.

    The SAT, LSAT, GMAT, MCAT and GRE are all pretty reliable indicators that aren't subject to inflation.

  • NL_

    Among law schools, only a handful try to move to a pass-fail-honors system. Even at some of the best law schools, most of the first year classes are curved. The reason is to provide employers with a way to differentiate graduates. Whereas most law firms would be happy with most graduates of Yale, Harvard or Stanford, they want to filter out some graduates of Georgetown or UCLA. Even the best schools outside the top 3 have to have some way to approximately rank their graduates. Most law schools have an explicit rank for their students (though mine does not, employers have a sense of the distribution).

    The worst schools have the strictest curves, with the worst students failed out so they won't fail the bar and tarnish the school's record. They also use curves to push some scholarship students out of qualification for their scholarships.

    Law schools are themselves a sorting mechanism, and the GPAs are a further sorting metric for employers. A given top-shelf law firm would be happy to hire a Yale graduate from the lower part of the class, so they can roll the dice based on the limited knowledge they glean from the . That same firm might be fine with a graduate of Duke or Penn who graduated in the top third, or UCLA in the top fifth, or WashU in the top tenth. Without some way to distinguish students, law firms find it harder to know who is a top third student and who is a bottom third student. If one school adopts a weak sorting system (blurring the lines between bottom third and top third) while the others maintain ranking, it's less of a gamble for employers to find skilled graduates at the curved school.

  • NL_

    Also, Russ R., it's interesting that at least the LSAT has been subject to inflation and variability over time. Among the problem types, the Reading questions have gotten harder since the 90s and the Logic Games have gotten easier. I know because I prepped for the LSAT by taking 40-something practice tests, some of them first administered well back in the 90s. Most others had the same impressions on the evolution of the LSAT.

    Apparently the LSAT has also experienced inflation, as scores in the 170s (out of 180) have gotten more common. In order to get into the 99th percentile, the cutoff used to be around 170, then 172, and when I took it the cutoff was 173 to get 99th percentile.

    But it still works because the important thing is that it provides a good benchmark against the other recent test-takers. Scores older than a few years are invalid and LSAC won't provide them, so the inflation is mitigated by the expiration of old scores. Whereas old college grades don't expire, so a GPA from 1964 is presented alongside a 2012 GPA as though they are comparable.

  • Ted Rado

    The emphasis on grades and the school is overblown. As a young engineer, I thought that a person from a prestigious university and an advanced degree was automatically a better engineer. Wrong!

    Every graduate from a recognozed engineering school has the basic training for their profession. What they do with it afterward is the issue. Some really dive into their work, make every assignment into a study project, and steadily improve their skills. After twenty years, they are way ahead of a person with a PhD from MIT who rested on his oars. Interest in one's profession, initiative, imagination, study, and hard work eclipse the more obvious attributes over time.

    Having said all this, it is still true that the smarter guy will do better, other things being equal. The problem is that the smarter guy does not always follow through in later life. Ideally, a person would attain the highest level of education that he is capable of, and then apply himself to his work with enthusiasm. Unfortunately, this does not always happen. Many smart people choose not to go to grad school, and many that do do not apply themselves afterward.

  • Sam L.

    For bragging on sons, I like the way Sippican Cottage does it.

  • DoctorT

    Russ R. said: "... The SAT, LSAT, GMAT, MCAT and GRE are all pretty reliable indicators that aren’t subject to inflation..."


    The SAT was "renormalized" in 1995 for two reasons: 1. The increased numbers of students taking the exam lowed its average score and 2. If one compared the top few hundred thousand exam takers from each year in the early-1990s to the same number from 1970s and 1980s, the scores were lower. The post-1995 SAT average combined scores are about 100 points higher than the pre-1995 scores.

    In 2002 the GRE discontinued the separate Analytic Ability section. That section had the highest correlation with success in graduate school.

    The MCAT exams use scaled scores based on the previous year's results. The mean scaled score for each part is 8 plus/minus 0.3. The Association of American Medical Colleges does not release information on what it does to keep the mean scores near 8. My strong suspicion is that the exam is easier now than a generation ago.

    The LSAT normalizes every exam. We do not know if today's LSAT-takers are doing better or worse than in the past. I suspect worse.

  • IGotBupkis, Poking Fun At President Downgrade For 4 Years and Counting...

    Yeah, well, years ago, I signed up for an Archery course during the Winter term (Florida, not as bad as it sounds). There were only five other students in the class, so we all got lots of time and attention from the teacher. By the official grading specs in the syllabus -- test results, etc. -- Everyone should have gotten an "A" for their accomplishments.

    The teacher informed me that, though I'd earned an "A", she had to give me a "B+" because they'd scream at her if she gave everyone in the class an "A", and I had the lowest "A" grade.

    In another class, I got points taken off on trivia on the final which dropped me to a B+ from an A -- I went to the teacher, and they acked that they knew -- from my class participation -- that I inarguably knew the points in question they'd taken from me.

    They denied me the boost anyway because "they didn't think I deserved it" -- translated: they didn't like me, and didn't WANT to give me an "A".

    This sort of crap is ephed, it's been ephed for decades, if not always. Teachers take grades up and down based on who they like and dislike all the time.

  • IGotBupkis, Poking Fun At President Downgrade For 4 Years and Counting...

    >>> As a young engineer, I thought that a person from a prestigious university and an advanced degree was automatically a better engineer. Wrong!

    Indeed. I had a friend a year ahead of me that went to UF, a much more prestigious school than UCF, in Orlando, where I'd gone, in the same major, IT.

    I can state categorically that, though he and I were about the same intelligence, I knew far more at the end of my Freshman year at UCF than he knew at the end of his JUNIOR year. Part of that was the teaching structure at UF being pretty much brain damaged, as you took your BS courses before your major courses, while UCF had no such structure. That, however, was not the sole reason by any means -- for various reasons, I transferred to UF for my Senior year. Although UF had as much or more stuff than UCF did to learn from, resource-wise, a lot more of it was sequestered away for the use of Teacher 'X's assistants and so forth, while at UCF, if you went looking for it, and they had it, it was pretty easy to get access. One of the benefits of "less prestigious" schools, along with the softer class-taking structure.

    As an example, some years later UF was GIVEN a LISA after it was introduced. It immediately got locked in a room and only grad students got to use it... one of whom promptly set the floppy containing the drivers for the graphics tablet that had been provided with it on the tablet itself, inadvertently wiping it in the tablet's magnetic field... LOL.

  • IGotBupkis, Poking Fun At President Downgrade For 4 Years and Counting...

    That, by the way, brings me to a short rant about the idiocy called a "block course" system, which offers lots of benefits to registrars and departmental schedulers, but none whatsoever to the learning process.

    UF's system was that you take all the non-major courses the first two years, getting what is basically an AA degree. THEN you take your major courses only after that.

    The problem with that is, it means that you're at the end of your freakin' junior year before you find out that your major isn't half as interesting as you thought it would be, and you're practically screwed, it's not easy to change your major since you might even change to an entirely different track which has little in common with the one you were planning on, even down to the non-major courses (UF's Engineering degrees are very different from their "Liberal Arts and Sciences" degrees, for example).

    I started out in Physics/Math, found out about midway through my sophomore year that I didn't like physics that much but did like the IT stuff I'd taken as a part of the physics degree -- so I switched to IT/Math instead, with plenty of time.

    Another aspect of such idiocy is that you can leaven the hard major courses with notably easier non-major courses, and thus actually take MORE courses involving your major than you could if you were cramming all the damned hard courses into a single series of 4-odd semesters... and generally get better grades in both the major and non-major courses, as well as have more time for fooling around -- be that partying OR geek activities like hacking computers and stuff.

    The current edumacational model is remarkably lame, and far more concerned with issuing official paper bits than it is with anyone actually learning jack or his smelly companion.

  • IGotBupkis, Poking Fun At President Downgrade For 4 Years and Counting...

    LTMG: The companies out there are no longer looking for "can do" people, they pretty much only want to hire the people they already have. Seriously.

    This from a recruiter I've been corresponding with during my current job search:
    I used to recruit to fill permanent positions –and my clients were always hired people who ‘could do’ the job—of course they’d have required skills, but they’d consider the whole person and would hire someone who only met ½ the requirements if they had the right aptitude and attitude.

    Now days, you not only have to hit every required skill, you also have to have used the exact same tools, in the exact same domain—they won’t hire ‘can do’ …only ‘have done’

    She also noted this:
    The resume is another case of ridiculousness. In the past, managers used the resume as a tool to see 1) what the candidate had done in the past and 2) how well the candidate communicated. Now they use computer programs to do word and phrase matches for the job at hand and if there is not at least a 60% match up—you won’t get an interview. Even if you come through that screen, if the manager doesn’t see reference to the skills needed within the top third of the resume, you won’t get an interview. So instead of getting a true picture of what a candidate has to offer, they’d rather read their own words back to themselves…

    And then you wonder how it is that so many companies whine about "lack of suitable talent"...?

    The talent is there, the idiot lazy POS HR managers won't do their ^@#@$^$#&^ jobs.

    I've got 30+ years of experience with computers doing just about every job there is, from low-level hardware programming to internal and external tech support functions. It's a varied enough set that it's pretty clear my learning ability hasn't atrophied. I'm about as good a software tester as you could hope to find, and pretty damned good with SQL as well. I have excellent logistical and detail skills, my language skills are top notch, and I have both high intelligence AND at least pretty decent wisdom/common sense abilities.

    Yet it's damned near impossible to find a job even though there are literally dozens listed in my general location that I could almost certainly be doing better than 90% of other potential hires after eight weeks... but few of them will even give me a phone interview.

    It's ephing ridiculous. This country is getting more and more run by lazy, incompetent morons who've become the "flappers" for everything. Sooner or later it's going to bite us in the ass.