52% Passed the HSC. Why Does One Exam Still Decide So Much?

Published: 17 November 2026, 02:06 PM
Representational Graphic
Representational Graphic © TDC

On 31 October, Bangladesh's education boards reported that 52.12 percent of the 1.27 million HSC candidates had passed. The rate was 58.83 percent in 2025 and 77.78 percent in 2024, all on the same 11 boards.

The 2024 figure needs context. That year's HSC began on 30 June, and the student uprising halted it after seven exams. The government cancelled the remaining six after students protested, and it filled the gap with "subject mapping", which gave each student marks in those subjects based on their earlier SSC results. The chairman of the education boards' coordination committee said at the time that this method was one reason more students received top GPA-5 grades.

So the 25-point fall compares an unusual year with a normal one. The pass rate was already 78.64 percent in 2023, though, so results had been high for some time. A Dhaka University professor told The Daily Campus that grace marks and lenient grading had pushed results up for years. Weak English results and errors in two physics questions also affected 2026, and no public data separates these causes.

Whatever the exact mix, a result that swings this far raises a fair question: how much can one exam tell us about what a student knows? A high-stakes exam is a test where one set of papers, taken over a few days, decides a student's next step, such as a university seat. Its strength is simplicity: every student sits the same paper, and an examiner who has never met them marks it.

That consistency matters in Bangladesh, where families need to believe the process is fair. A national exam gives them something visible to trust. Systems that rely on teachers' marks have to answer a harder question: how does a family know those marks are fair?

The cost shows up in the classroom. When one exam decides everything, students and teachers study for the exam and little else. A 2023 survey by the education research group CAMPE found that more than three-quarters of students use private tutors or coaching. Families with money can buy more preparation than families without, so the exam that looks equal on paper can favour the better off.

Finland offers a contrast. It runs no national exam for all students in primary or lower secondary school, and teachers grade their own students until the matriculation exam at the end of upper secondary. A review of 159 studies, published in April 2025 by researchers at the University of Jyväskylä, found no convincing evidence that national exams of this kind raise learning. Finland is not a perfect model, though.

Its scores have fallen in recent years, and in June 2026 its parliament approved national assessment points for grades 5, 8 and 10. The details are still being set, so it is unclear how much testing that will add. The lesson is that even a country known for trusting teachers keeps looking for a way to check results across schools.

Changing the exam does not remove the competition behind it. Singapore ended mid-year exams at all levels by 2023 and changed how it scores its primary school leaving exam, to ease pressure on children. Over the same period, spending on private tuition rose from S$1.4 billion in 2018 to S$1.8 billion in 2023. When places in good schools are limited, families look for an edge wherever they can find one, and Bangladeshi HSC candidates also compete for a limited number of university seats.

Alternatives to exams can measure things a paper cannot. Classroom projects, oral presentations and long pieces of writing show whether a student can explain an idea, solve a problem over several days or build an argument over several pages. A three-hour paper has no room for any of that. Those tasks depend on who marks them, and that is where the risks sit.

England's experience in 2020 shows the problem. Exams were cancelled, and a computer formula set the first A-level grades, downgrading 39.1 percent of teachers' own predictions. After public anger, the government switched back to teachers' grades, which were higher overall. England's exam regulator, Ofqual, also reviewed the research and found that teachers' judgments tended to disadvantage poorer pupils, although it said the evidence was limited.

Queensland, in Australia, tries to reduce that risk. Its schools set and mark the tasks, but the state authority approves them in advance and rechecks a sample of marked work afterward. An evaluation found that school marks and the authority's marks agreed more than 90 percent of the time. The authority paid for that evaluation, and figures on the extra work it creates for teachers are hard to find.

Bangladesh has already tried a version of this change. From 2023, it ended the primary and junior school exams and moved to classroom assessment under the 2022 curriculum. A 2024 survey by the national curriculum board found that just 37.6 percent of 253,000 teachers had a working grasp of the new assessment system. The interim government returned to the 2012 curriculum and exam-based assessment, and the survey points to teacher preparation as a major obstacle.

Policy is still moving. A secondary education committee proposed in February 2026 that national exams cover five subjects alone and that the Class 5 and Class 8 exams end. In June, the education minister announced one national question paper for the SSC and, later, the HSC. A Class 5 scholarship exam came back in April 2026, so the steps pull in different directions.

Does less exam pressure improve learning? The evidence is mixed. The Education Endowment Foundation, a British research charity, estimates that regular feedback during the school year adds about six months of progress, though some studies found no gain. Economist Ludger Woessmann found that countries with national exit exams tend to score higher in the subjects those exams cover, but that comparison cannot prove cause.

Read together, these cases show that no system has proved that dropping exams raises learning, and none has proved that keeping them does. The ones that work share two conditions: families trust the marking, and teachers know how to assess. The 2027 plan for a new Class 6 curriculum will test both. The ministry will need to say who marks the work that isn't an exam, how those marks will be checked, and how teachers will be trained before the change begins, or the 2023 experience could repeat.