Higher education is infatuated with numbers. Educators assign students a mark for an assessment, combine those marks into a final unit grade, and then institutions combine grades from multiple units into a Grade Point Average (GPA). The result is a wonderfully convenient numerical value. Years of learning ultimately represented by a number that can be recorded, sorted, compared and ranked, not only internally in the institution, but also with other institutions globally.
There is something very attractive about that simplicity. A 78 can be compared with a 72. A GPA of 3.7 can be compared with one of 3.4. We can establish cut-off points, award scholarships, determine progression and make decisions about entry into further study. The number gives us something contained and apparently comparable.
But that is also the problem… Learning is not nearly as neat as the number we use to represent it might suggest.
We Like Things We Can Count
It is easy to understand why grades persist. Higher education operates at enormous scale, and universities need systems that can record the performance of thousands of students across thousands of assessments. Grades make that complexity manageable. A number can be entered into a database, averaged, used to determine whether someone has reached a threshold, and compared with another number.
Describing a student’s learning journey is more difficult to interpret. Perhaps our infatuation with grades tells us much more about what higher education systems need than about representing a student’s learning itself. The danger is that, because the number is useful, higher education has gradually allowed it to take on greater meaning. A mark initially produced as a judgement about a particular piece of student work starts to become a representation of the student and their learning.
The administrative convenience of a grade has gradually been mistaken for educational meaning.
The Number Looks More Certain Than the Judgement Behind It
When educators assign a grade, they are making a professional judgement about evidence of learning. That judgement may be supported by assessment criteria, moderation and a carefully developed rubric, but it remains a judgement. Two educators teaching the same unit can assess the same piece of work and arrive at different marks. They might agree broadly about its quality but disagree about whether it should receive 68, 72, or 75.
Even the same educator may not necessarily make precisely the same judgement about equivalent work across different semesters. Their experience develops, their interpretation of the task changes, expectations shift, and their understanding of the discipline and of what students can produce continues to develop. None of this necessarily represents poor assessment practice. Professional judgement is part of assessment.
The answer is not to simply create an even more detailed rubric. The limitations and possibilities of rubrics are a different conversation. What interests me here is what happens once that judgement becomes a number.
Once an educator’s professional judgement becomes 73, much of the uncertainty seems to fade away. This score is entered into a system and perceived by many- students, institutions, and future employers- as an exact figure. It can be compared with another student’s 74 and combined with other grades to derive statistics such as GPA. Although we recognise the role of judgement in assigning a grade, we often treat the final number as objective once it is recorded.
The number also masks the reality that students begin from different starting points and under diverse conditions. They have different prior knowledge, varying amounts of time for study, and different work and caregiving duties. While this doesn’t imply that grades should be modified for every situation, it highlights that identical grades can reflect very different learning experiences.
When the Number Starts Losing Its Meaning
Recent developments at Harvard provide an interesting example of how dependent higher education is on grades maintaining a shared meaning. By 2025, more than 60 per cent of undergraduate grades at Harvard were As, compared with around one-third in 2010. Over the same period, the median cumulative GPA at graduation rose from 3.56 to 3.83.
In May 2026, Harvard’s Faculty of Arts and Sciences decided to cap the percentage of A grades at 20 percent of undergraduates in each course, plus four additional students. This policy will be implemented starting Fall 2027. Additionally, Harvard will consider using an internal average percentile rank instead of GPA for certain awards and honors decisions.
The Harvard discussion is primarily about grade inflation, and there are important debates about the response it has chosen. But another question sits underneath it. If the meaning and distribution of an A can change so significantly over time that an institution feels the need to recalibrate the system, how stable was that grade as a representation in the first place?
And if GPA becomes less useful once grades become compressed, what exactly was the GPA representing before?
Perhaps this is another reminder that the apparent certainty of grades depends on institutional convention, context and human judgement. If institutions need to manage the distribution of grades partly so those grades continue to distinguish between students, are we simply measuring learning, or are we also maintaining a system of classification?
Then We Average Everything
This becomes even more interesting when we move from an individual grade to GPA. Across a degree, a student completes different units, assessment types and learning experiences. Different educators assess their work, each exercising professional judgement at different points in time of the student’s learning journey.
We then calculate the average of those grades, which could be 3.42. Now, years of learning are now represented by a single, concise number. This averaging also does something else: it flattens time.
A student might struggle at the beginning of their degree. Perhaps university is unfamiliar. Perhaps they entered with limited prior knowledge. Perhaps they were working significant hours or managing family commitments. Their early results might be around a low Credit.
Then something changes. They learn how to navigate higher education. Their disciplinary knowledge develops. They become more confident. Their external circumstances might improve. Perhaps they change their work commitments or their children become older. Over the next two years, they consistently improve, eventually completing their degree with Distinction and High Distinction-level work.
This is a substantial story of growth and learning, yet their GPA still reflects who they were at the start. It does not provide insight into the progress they achieved throughout their learning journey, which could be helpful for future study applications. The initial Credit is still mathematically included in the total used to represent the graduate they have become.
GPA Does Not Show the Direction of Travel
Consider two students graduating with the same GPA. One began their degree strongly, but their results gradually declined. The other struggled initially but improved consistently and finished with their strongest work. Their GPAs could be identical, yet their learning journeys are clearly not.
If education focuses on development, this should inform our approach. Keeping a student’s previous grades isn’t problematic; it reflects what the student showed at specific times. The key question is whether averaging these moments offers a worthwhile overview of the student’s overall growth.
A GPA tells us something about the collection of grades a student received. It tells us considerably less about their development. The student who graduates is not the same learner who commenced the degree, yet the GPA treats every graded point along the way as a contribution to a single final number.
The Number Is Comparable. Is the Learning?
This is perhaps where our attachment to grades becomes most apparent. The number is wonderfully easy to compare, while the learning behind it is not.
A grade may have been awarded by a different educator, through a different assessment task, within a different discipline, during a different semester, and under completely different circumstances. Yet once we reduce each experience to a number, those differences become much less visible, or at least less front of mind.
Comparing 74 with 76 or a 3.2 GPA with a 3.5 is straightforward mathematically. However, assessing the educational value is much more complex. Grades do not eliminate the complexity of learning; rather, they conceal much of it. Perhaps that is precisely why some find grades so useful.
What Does the Number Actually Tell Us?
This does not mean grades are meaningless. Perhaps we simply need to make a more modest claim about what they represent.
A grade reflects the quality of evidence a student demonstrates through specific assessments, judged by certain criteria at a particular moment for a particular educator. While this information is useful, it doesn’t reveal the student’s initial level, the progress they’ve made, their responses to feedback, how their thinking has developed, or how they might perform in different contexts. It also doesn’t show what kind of student they are or whether they are suitable for further study.
A GPA faces even greater limitations when used to represent a student’s learning over an entire degree. The main issue might not be the use of grades in higher education but rather the amount of significance we attach to that number.
Chipping Away at Our Infatuation
None of this is going to change overnight. Grades are deeply rooted in higher education and linked to progression, admissions, scholarships, accreditation, graduate recruitment, and institutional systems. There’s no quick alternative that can instantly handle all these functions.
But that does not mean the only options are to retain the existing system unquestioningly or abolish grades tomorrow. We can chip away at our dependence on them. We can question whether small numerical distinctions really tell us anything meaningful, reconsider when GPA should be used, and make development and progression more visible.
We can offer more detailed descriptions of achievement, enable feedback that accompanies students throughout their learning process, and utilize portfolios and demonstrations of skills more extensively along with grades. Most importantly, we should stop expecting the number to convey more than it genuinely can.
None of these alternatives match the elegantly contained number that higher education has come to expect. That’s partly the point. Learning is naturally chaotic, evolving gradually. Students start from various points and learn in different contexts. Their progress is assessed through professional judgement, and their abilities shift over the course of a degree.
Perhaps any attempt to condense all of that into a single number should make us slightly uncomfortable. We do not need to abolish grades tomorrow, but perhaps we can become a little less infatuated with them.
After that, a new question arises: Could there be stages in a student’s learning process where the grade isn’t necessary?
Questions for Reflection
- Why does higher education place so much trust in numerical grades as representations of learning?
- How meaningful is the difference between a mark of 72 and 75 when professional judgement sits behind both?
- If two graduates share the same GPA but have followed very different paths during their degrees, what does that GPA truly indicate?
- Are there decisions in higher education where we rely on grades primarily because they offer an easy way to sort and compare students?
- If we accept that grades cannot capture the full learning journey, where could we begin to rely on them less?
