docs: what sets the praktikant overtime boundary (7,4 t stays)

[renemadsen]

Research only — no code changes. The presets already encode 26640 s, so this confirms current behaviour rather than changing it.

Decision

The daily boundary stays 7,4 t. A planned day shorter than 7,4 t does not lower it.

§ 22 stk. 1 triggers overtime "efter normal daglig arbejdstids ophør" but never defines that figure numerically. Searching the agreement for planlagt arbejdstid, arbejdsplan, vagtplan, aftalt arbejdstid: two do not occur at all, and the rest appear only inside opt-in flexible regimes — never in the ordinary regime § 22 governs.

The result is asymmetric:

Planned day	Plan sets the line?	Authority
Longer than 7,4 t	Yes, but only in an opt-in regime and up to its ceiling	§ 9 stk. 4, § 9 stk. 5 (premium only above 45 t/uge), protokollat stk. 1 (6–9,25 t/dag)
Shorter than 7,4 t	No clause anywhere	—

Two clauses cut against lowering it: § 22 stk. 4 anchors the reckoning to "den normale ugentlige arbejdstid", and § 8 stk. 4 c bars "lønmæssig kompensation, fordi arbejdstiden er kortere end den normale".

For praktikanter it is largely moot anyway — the circular's monthly figures divide out to 160,33 t/md (37 × 52 ÷ 12), so a praktikant is a full-timer at 37 t/uge by definition.

Also recorded

• 296 t / 8 uger is an average, not a weekly cap (§ 8 stk. 2: "i gennemsnit"). A single 40-hour week triggers nothing by itself; 296 shrinks by 7,4 t per søgnehelligdag; and averaging the weekly norm does not license unpaid overrun of the daily one.
• § 22 stk. 4 makes overtime a weekly reckoning — culpable absence is netted off first, with two exceptions that must be modellable per absence or sick workers lose overtime wrongly. Not modelled today; the engine has no weekly context.
• Afspadsering is the trainee's choice, not the employer's (§ 50 stk. 4 c).
• Three defects in GLS-A's own published circular, including an overtime table headed "7 - 12 måneders praktik" that carries the 7–18 rates (proven by arithmetic).

What is deliberately left open

§ 22 stk. 1 never defines the daily figure, so an employer with a fixed written daily schedule could argue that schedule is the normal daily working time. No clause, protokollat or published voldgift resolves it. The doc records the question to put to GLS-A rather than resolving it by guesswork — and notes that Industriens Overenskomst § 13 stk. 2 defines overtime the other way (plan-anchored), which is why the opposite expectation is reasonable even though it does not govern here.

🤖 Generated with Claude Code

[Copilot]

Pull request overview

Adds a research write-up to the existing “Overenskomst Research Findings” document to record the decided interpretation of what constitutes “normal daily working time” for praktikant overtime (default boundary remains 7,4 hours / 26640 seconds), along with supporting citations and related notes.

Changes:

• Documented the decision that a planned day shorter than 7,4 hours does not lower the overtime boundary, including rationale and counter-arguments.
• Added clarifications on “296 timer / 8 uger” being an average (not a weekly cap) and noted the unimplemented weekly netting requirement in § 22 stk. 4.
• Included a verbatim excerpt from GLS-A’s circular plus source links.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.