What does a "mandatory curve" actually require?
Most ABA-accredited law schools impose a grade normalization rule on required first-year courses and large upper-level courses. The rule typically has two parts. First, a required mean or median: the section's grades must average to a specified number, often 3.0, 3.2, or 3.3 on a 4.0 scale. Second, distribution bands: caps and sometimes floors on how many students can receive each letter. A common structure allows no more than 10-15% A range grades, requires roughly 25-35% B+ grades, places the bulk in the B/B- band, and requires or permits some percentage below B-.
The practical effect is that your professor cannot give everyone a B+ even if everyone wrote a competent exam. If forty students take the class and the cap is six A-range grades, six students get A-range grades. The seventh-best exam gets a B+ regardless of its absolute quality. This is what people mean when they say law school grading is zero-sum.
Some schools use a soft curve: a target mean with a tolerance band (say, 3.15 to 3.35) and no hard caps on individual letters. Others use a hard curve with rigid percentages. A minority use no curve at all in upper-level courses. Your registrar's academic rules document — not the student rumor mill — is the authoritative source.
How does a professor get from raw exam scores to curved letters?
Nearly all law schools use blind grading: you write under an anonymous exam number, and the professor does not learn whose exam is whose until grades are submitted. The professor reads and scores every exam, usually against a rubric or issue checklist, producing a raw numerical score. Then the exams get rank-ordered from highest to lowest raw score.
The curve is applied to that rank order. The professor draws lines: exams 1 through 5 get A, 6 through 14 get A-, and so on, until the distribution matches the required bands and hits the required median. This means the gap between an A- and a B+ is often two or three raw points out of a hundred — sometimes one. Small differences in coverage, organization, or how many issues you spotted move you several grade steps.
After grades are locked, most schools allow limited discretionary adjustment for class participation, but only within narrow limits (often one grade step, and only for a small number of students). A few schools forbid it entirely. This is why 'she likes me' is not a grading strategy.
Why does one exam decide the whole grade?
The single-exam model is a holdover from a century of law school practice, and it persists because it is administratively cheap and because faculty believe a three-hour issue-spotter tests the synthesis they care about. ABA accreditation standards now require both formative assessment (feedback during the term) and summative assessment (the graded evaluation), so many schools have added midterms, ungraded practice exams, or short assignments. But at most schools the final still carries 70-100% of the grade in doctrinal courses.
The consequence for you is that exam-taking technique is a separately trainable skill from knowing the doctrine. Two students with identical mastery of promissory estoppel can land three grade steps apart because one wrote in IRAC-organized, fact-linked paragraphs and the other wrote an essay about the law in the abstract. Under a forced curve, technique is what separates the top of the pile from the middle.
Which classes escape the curve?
Curve rules almost always contain exemptions. The most common ones:
- Small classes. Many schools exempt courses under a threshold enrollment — often 20, 25, or 30 students — because a forced distribution is statistically meaningless in a small group. Seminars therefore tend to run high.
- Paper courses and seminars. Where the grade rests on a research paper rather than a timed exam, the curve is frequently relaxed or waived.
- Clinics, externships, moot court, and journal credits. These are commonly graded credit/no-credit or on a separate, more generous scale.
- Legal writing. Some schools curve legal research and writing to a higher median than doctrinal courses; others exempt it entirely.
- Pass/fail electives. Many schools let you take a limited number of upper-level courses pass/fail, which removes them from your GPA.
How do curves differ across schools?
Variation is large and it is driven by two competing pressures: faculty desire for grade differentiation, and administrative desire to make graduates look competitive to employers who compare raw GPAs across schools. Around 2009-2011, a number of schools raised their curves in response to the recession-era job market, some retroactively. That is why medians cluster higher today than they did in the 1990s.
A handful of schools opt out of numerical ranking altogether. Yale uses Honors/Pass/Low Pass/Fail with a credit/fail first term. Stanford uses Honors/Pass/Restricted Credit/Fail with a cap on Honors. Berkeley uses a High Honors/Honors/Pass system with a percentage cap at the top. Northeastern uses narrative evaluations instead of grades. These systems still ration the top designation, so competition does not disappear — it just gets less granular.
Also check whether your school ranks students at all. Some publish exact class rank, some publish only decile or quartile bands, some release rank only to students in the top percentages, and some release nothing. This affects how you can describe yourself on a resume.
What does the curve mean for rank, honors, and hiring?
Because the curve compresses everyone into a narrow GPA band, tiny GPA differences translate into large rank differences. At a school with a 3.3 median, the difference between a 3.35 and a 3.55 can be twenty or thirty percentile points. This is why students obsess over fractions of a grade point that would be noise in an undergraduate transcript.
Cutoffs that matter: Order of the Coif is limited to the top 10% of the graduating class at member schools. Latin honors cutoffs are set by each school, usually as a fixed percentage rather than a fixed GPA. Law review membership at grade-on schools typically draws from the top of the 1L class. Large-firm on-campus interviewing screens by GPA cutoff, and those cutoffs move with the school's rank and the market.
One structural point students miss: sections matter. If your school assigns 1Ls to sections and each section is curved separately, you are competing against roughly 60-90 classmates, not the whole class. A strong section can cost you grades. Nothing you can do about it, but it explains outcomes that otherwise look arbitrary.
How should the curve change how you prepare?
Study for relative advantage, not for comfort. Everyone in your section will know the black-letter rules by December; reading the casebook harder does not distinguish you. What distinguishes you is producing a well-organized, issue-dense answer under time pressure. Take old exams from your actual professor, timed, closed-book if the exam is closed-book, and compare against model answers or sample A exams if the school releases them.
Build your outline to be usable in an exam, not comprehensive. A 90-page outline you cannot navigate in ninety seconds is worse than a 25-page attack outline with rule statements you can drop into an answer verbatim. Under the curve, points come from spotting and analyzing more issues, not from writing more beautifully about the same issue.
Finally, go to office hours and ask what a strong answer looks like in that course. Professors differ sharply on whether they reward policy discussion, counterarguments, or crisp rule application. Since the professor also draws the curve lines, that preference is the scoring rubric.