Your Evaluation Platform Is Also a Staff-Monitoring System — and the Works Council Gets a Vote

Bottom line up front: In several European jurisdictions, a course-evaluation platform that reports results at the level of the individual teacher is not, in the eyes of labour law, "just a survey." It is a technical system suitable for monitoring the conduct or performance of employees — and that classification hands your works council a formal say over whether, and how, you may deploy it. In Germany the trigger is a co-determination right under Section 87(1) No. 6 of the Works Constitution Act (Betriebsverfassungsgesetz, BetrVG). In the Netherlands it is a consent right under Article 27 of the Works Councils Act (Wet op de ondernemingsraden, WOR). Crucially, neither right is discharged by GDPR compliance. Your platform can be flawlessly lawful under data-protection law and still be un-deployable until the staff representatives agree.

For procurement teams who have learned to run the GDPR checklist, this is the second gate they keep walking into at rollout. This piece explains why teaching evaluation trips the labour-law wire, what the case law actually says, and how to design an evaluation programme that clears co-determination instead of colliding with it.

Why a course evaluation counts as "monitoring"

The intuition in most quality offices is that course evaluation is about the course, not the lecturer — a diagnostic on the module, not a verdict on a person. Labour law does not read it that way, and the reason is instructive.

Under Section 87(1) No. 6 BetrVG, a German works council has an equal-footing co-determination right over the introduction and use of technical devices designed to monitor the conduct or performance of employees. The decisive word in practice is not designed. The Federal Labour Court (Bundesarbeitsgericht, BAG) has held in long-settled case law that it is enough for the system to be objectively suitable for monitoring; the employer's subjective intention is irrelevant. A system that records information capable of being attributed to an identifiable employee and used to assess their behaviour or performance falls within the section, whether or not anyone plans to use it that way.

A per-instructor evaluation report is a near-textbook example. It links quantified performance signals — scores, rankings, trend lines — to a named member of teaching staff. That it was introduced to improve courses is beside the point; it is suitable to appraise the teacher, so the co-determination right attaches.

The Dutch route arrives at the same destination by a slightly different road. Article 27(1) WOR gives the works council a consent right (the strongest instrument in its toolkit) over an employer decision to introduce, amend or withdraw a system that monitors employees' behaviour or performance, and over regulations governing the processing of personnel data. The Dutch data-protection authority (Autoriteit Persoonsgegevens) is explicit that staff-tracking and personnel-assessment systems fall here. If the council withholds consent, the employer may not, in principle, use the system — though it can ask the sub-district court for substitute permission under Article 27(4).

This is not the same obligation as GDPR

It is tempting to fold all of this into the data-protection work you have already done. Resist that. GDPR and works-council law protect different interests and are enforced by different actors.

GDPR asks whether you have a lawful basis, a proportionate purpose, and adequate safeguards for personal data. Co-determination asks whether the collective workforce, through its elected representatives, has agreed to a system that can be used to appraise them. Article 88 GDPR expressly leaves room for this second track: it permits member states to lay down more specific rules for processing employees' data, "including" through collective agreements. In other words, the GDPR itself anticipates that a works-council agreement (a Betriebsvereinbarung) may become part of the legal architecture governing an evaluation tool — not a substitute for it.

The practical upshot: a data protection officer can sign off your DPIA and you can still be stopped by a works council that was never consulted. The two approvals are cumulative, not alternative.

"But surely evaluation isn't surveillance?" — the strongest objection

The most serious counter-argument is that treating teaching evaluation as "monitoring" cheapens the word and chills a legitimate quality process. Students have a right to be heard; institutions have a duty to assure teaching quality; calling the resulting report "surveillance" seems to weaponise labour law against accountability.

Two things should be conceded to this objection, and one should not. Concede, first, that the purpose of most evaluation programmes is genuinely developmental, not disciplinary — and that this matters enormously to how the tool should be designed and governed. Concede, second, that co-determination is not a veto on evaluating teaching; it is a right to negotiate the terms. German works councils cannot forbid quality assurance, and Dutch employers who are refused consent have a judicial route.

What should not be conceded is that intention settles the legal question. The BAG's "suitability" test exists precisely because systems built for benign purposes acquire disciplinary uses over time — a low score cited in a probation meeting, a trend line produced in a promotion case. The labour-law right is a safeguard against drift, and the honest response is not to deny the risk but to design it out.

Where Koji fits

The design choices that keep an evaluation programme on the right side of co-determination are the same ones that make the data more valid — a convenient alignment.

• Report at the level you actually act on. Koji's programme- and institution-level reporting lets you surface course- and programme-level themes without foregrounding individual-instructor league tables. Aggregated, purpose-limited reporting is both better methodology (it resists small-sample noise and paradoxical reversals) and a materially weaker "monitoring" footprint to negotiate.
• Make the purpose legible and standing. Koji's structured question types and AI-moderated interviews are built around diagnosis — what worked, what to change — rather than ranking people. A tool whose outputs are formative and thematic is far easier to write into a works-council agreement than an opaque scoring engine.
• Bias-aware, standardized moderation. Because the AI moderator applies the same probing consistently, you avoid the human-moderator inconsistency that makes performance data contestable — and you can document the safeguard for staff representatives.
• Formative, mid-cycle collection. Evaluation used to help a course while it is still running reads as development, not appraisal, and belongs naturally in the "improvement, not monitoring" framing that co-determination negotiations reward.

None of this removes the co-determination right — nor should it. It makes the agreement easy to reach. The same conversational interview engine underpins the general-purpose research platform at koji.so, where the employee-consultation angle recurs whenever HR or people-analytics data is involved.

What to do before you sign the contract

Engage staff representatives before selecting a vendor, not after. Put purpose limitation, reporting granularity, retention, and the ban on using scores for standalone disciplinary decisions into a written agreement. Treat the works-council agreement as a deliverable of the procurement, alongside the DPIA. Institutions that do this ship on time; institutions that treat consultation as a formality discover that Article 27 and Section 87 have teeth.

Frequently asked questions

Does GDPR compliance mean we don't need works-council approval? No. They are separate legal tracks protecting different interests. GDPR governs the processing of personal data; co-determination (BetrVG Section 87(1) No. 6) and the Dutch consent right (WOR Article 27) govern whether the workforce has agreed to a system capable of appraising them. Article 88 GDPR expressly leaves room for collective agreements. You may need both approvals.

Our evaluation is only ever used to improve courses, never to discipline staff. Does the right still apply? In Germany, yes. The Federal Labour Court's settled test is whether the system is objectively suitable for monitoring performance, not whether the employer intends to monitor. A per-instructor report is suitable, so the co-determination right attaches regardless of purpose. The right way to reflect your benign intent is to write it into a binding works-council agreement.

Which countries does this affect? Most directly Germany (BetrVG) and the Netherlands (WOR), and by extension public-sector staff-representation regimes such as Germany's Personalvertretungsrecht. Many other EU states have works-council or employee-consultation rights that engage when performance-relevant systems are introduced. Always check the local regime; do not assume the German position is universal.

Can a works council simply block our evaluation platform forever? No. Co-determination is a right to negotiate terms, not an absolute veto on quality assurance. In the Netherlands, an employer refused consent can seek substitute permission from the sub-district court (WOR Article 27(4)). In Germany, disputes go to a conciliation board. The realistic outcome is a negotiated agreement, not a permanent block.

How does reporting granularity change our exposure? The more your reporting foregrounds individual-instructor scores and rankings, the stronger the "monitoring" character and the harder the negotiation. Aggregated, programme-level reporting reduces that footprint while also improving statistical validity. Design granularity deliberately.

Does anonymising student responses solve the problem? It helps with GDPR but not with co-determination. The labour-law right concerns data about the teacher, not the student. Even perfectly anonymous student feedback can produce a report that appraises an identifiable employee — which is exactly what triggers the right.

---

Koji for Education gives European institutions an AI-native evaluation platform designed for programme-level insight and formative use — the reporting posture that makes co-determination agreements straightforward rather than fraught. Book a walkthrough.