Will the 2026 baccalaureate resemble the 2025 baccalaureate? The Ministry of National Education announced at the start of the school year its intention to adjust the current methods of continuous assessment, sparking new debates surrounding an exam that has been revised several times since 2018. But why is it so difficult in France to determine the ideal structure of the baccalaureate?

Should we rethink the modalities of continuous assessment in high schools? All grades from the first and final years of high school would no longer be counted towards the baccalaureate , it was announced at the Ministry of Education’s back-to-school press conference. The announcement raised many questions, and even caused exasperation, among those involved in the field.

These reactions are understandable, as there has been no shortage of changes to the baccalaureate assessment process in recent years. Following the adoption of the “Blanquer” reform in 2018, which provides for 40% continuous assessment, the exam’s architecture has undergone constant changes.

The years 2018, 2019 and 2020 were years of trial and error. The introduction of specific continuous assessment tests (the E3C) caused significant turbulence in 2019 and 2020. Due to a markers’ strike, simple continuous assessment (as the teaching activity progresses) made a surprise appearance in 2019, with recourse to the marks in the school report . In 2020, the Covid pandemic even forced the acceptance of 100% simple continuous assessment!

The common exams will be abandoned in 2021. In 2022, annual averages will be taken into account, and the final exams for the options will be eliminated. Since 2019, only two sessions have taken place under the same conditions (2024 and 2025). The stakeholders feel a dual need: clarity and stability.

Why so many fluctuations? In a context where grading is becoming a real obsession for students and their parents , and with the success rate reaching 91.8% in 2025, including 96.4% in the general baccalaureate, a “large-scale reflection” is required. Certainly, when it comes to details, the “devil” has gotten involved. We had to take into account pandemics, political games, social movements. But undoubtedly the essential lies hidden deep down.

To see this clearly, we can use a “plate tectonics” type model, assuming that the effects of friction between underground “plates” are manifested on the surface, which confront each other due to deep dynamics that need to be identified, beyond the jolts and tensions.

A tension between support and certification

In the long term, the assessment scene is first and foremost the place of a conflicting coexistence between two requirements, which are part of different, although complementary, designs. The requirement to support and facilitate development, in a process of building knowledge and skills (requirement of formative assessment). And the requirement to socially certify the possession of this knowledge and skills (requirement of certification assessment). Consolidating learning vs. socially attesting to a level.

In the first case, we could talk about the “GPS” function (telling whether we are on the right road); and, in the second, the “issuing a license” function (socially certifying a skill acquired at the end of a training course).

The challenge here is to articulate the two functions of assessment , ensuring that the requirements of certification do not stifle the support effort. Hence the idea, welcome from this point of view, of introducing a significant part of continuous assessment. The first difficulty is then to situate the right level of importance. The second, to find a relevant form of continuous assessment – ​​simple continuous assessment, or specific tests?

A tension between localization and centralization

The second “plates” are those of a localization movement, and a centralization movement. Their confrontation takes the form of a struggle between the desire to give more responsibilities to local assessment actors, with the goal of simple continuous assessment. And the desire to impose a centralizing framework and model, with formalized national tests, identical in all examination centers.

It is often spontaneously thought that the second formula is preferable, because it would better guarantee the objectivity of the grading. As far as the evaluators are concerned, a conflict arises between two legitimacies. A local legitimacy: that of the teacher teaching one or more disciplines within an establishment. And an expert legitimacy: that of an external evaluator, evaluating, according to imposed standards, the work carried out in an officially defined test, common to all, and respecting guarantees of anonymity.

Continuous assessment, the problem is to know if, and how, we can escape the marking “according to the client’s head”. Note that in the oral exam, whatever the examiner, the “client” will always be present, with his “head”. And that the baccalaureate markers are, moreover, all teachers, who will not magically free themselves from what can bias their action as markers, and affect their way of marking.

This second tension then presents a second challenge: that the centralizing movement should temper the questionable effects of localism, without imposing a rigid formalism; without stifling the inventiveness and dynamism of the local.

A tension between a fight for titles and a fight for places

The third “plates” are those of the exam, and the competitive examination. A conflict arises between two issues. An issue of acquiring a school/university qualification (exam logic: the baccalaureate is the gateway to higher education). And an issue of occupying a place (competitive logic: the baccalaureate opens – or does not open – the door to access training disciplines in tension, where places are “expensive”, because quantitatively limited and socially desired.

Comparison between students for places in higher education becomes the real issue for candidates. It is then, in fact, Parcoursup that becomes predominant. And we could, in this respect, speak of a “parcoursupisation” of the baccalaureate. But this moves from an examination issue (having or not having a diploma), to a competition issue (in sectors under pressure, obtaining, or not, the coveted places). In the exam, you have to succeed, to obtain a title. In a competition, you have to do better than the others, to obtain a place that others also covet .

Should we then distinguish between evaluative grades (for internal use, to assess and track student progress) and certification grades (for external use, to certify achievements and take them into account in Parcoursup )? And go so far as to imagine two different reports, one, purely informative, containing all the results; the other, containing only the results deemed representative, and counting alone for the validation of the baccalaureate and the Parcoursup file?

The competitive examination issue gives assessment a dominant sorting function. This is the central task of Parcoursup: to match, as much as possible, demand and supply. To bring together a given plurality of individual training requests, and a set of differentiated places in training courses with limited capacity.

The final challenge is to create the conditions for fair sorting. This will require that the assessment, whether continuous or final, makes the reality of the candidates’ skills and their level visible! The whole question is whether, and how, an assessment can objectively assess (say) a skill, and make a decision… without being too wrong!

Thus, the baccalaureate, as an assessment system, must face three challenges, the importance and difficulty of which allow us to understand that, due to the persistence and weight of the frictional “plates”, the laboriously established balances will always be fragile. The examination formulas chosen, always more or less “cobbled together” in a given context, will always be questionable. This promises many more misunderstandings, exasperations, and tensions…

Author Bio: Charles Hadji is Honorary Professor (Educational Sciences) at the University of Grenoble Alpes (UGA)