Background
Ritchie White, a former Sydney Trains signaller and later an area controller, brought a representative proceeding on behalf of signallers and area controllers employed between 1 May 2018 and 26 November 2025. He alleged that Sydney Trains had underpaid entitlements under two enterprise agreements, including overtime loadings, shift-related entitlements, weekend and public-holiday penalties, a one-off payment and back pay.
The proceeding also alleged failures to maintain proper employment records, pay wages in full at least monthly and provide required breaks. The approximately 614-person class ultimately had 392 members registered to receive settlement distributions, including five late registrants and one person who was permitted to withdraw an earlier opt-out notice.
Sydney Trains agreed to pay $18.1 million in exchange for releases covering the claims advanced in the proceeding. Unregistered class members who had not opted out would receive no distribution but would nevertheless be bound by the settlement and barred from pursuing claims concerning the proceeding’s subject matter.
The Court’s Holding
Justice Perram approved the settlement under s 33V of the Federal Court of Australia Act 1976 (Cth), finding it fair and reasonable. Although incomplete payroll and timesheet records required modelling and extrapolation, the Court considered the methodology a practical and acceptable basis for settlement. The settlement reflected a significant but appropriate discount from the claims’ headline value, given the risk that White would not succeed on every construction issue and the costs and delay avoided by resolving the case before evidence and trial.
The Court also approved the settlement distribution scheme and deductions of $1,997,760 for legal costs and disbursements and $302,240 for administration. Those deductions totalled $2.3 million, or 12.7% of the settlement sum, and were supported by expert costs evidence.
The Court allowed Paul Ninness to withdraw his opt-out notice because he had misunderstood the proceeding and had been accounted for during settlement negotiations. It also admitted five late registrants, finding that their inclusion would not materially prejudice other registered members; their addition reduced the average estimated share by about $600.
Key Takeaways
- A class settlement must be independently scrutinised as fair and reasonable because the class’s interests may diverge from those funding or conducting the litigation.
- Imperfect data, averaging and extrapolation may be acceptable for settlement approval when more precise analysis would cause disproportionate cost and delay, even if the same evidence might be inadequate at trial.
- Late registration or withdrawal of an opt-out notice may be permitted where adequately explained and where participation will not materially prejudice other group members.
Why It Matters
The decision illustrates the Federal Court’s pragmatic approach to resolving employment class actions when incomplete records prevent precise calculation of every worker’s claim. It also underscores that approval of a compromise does not endorse the modelling as proof of loss at trial.
The judgment further highlights the consequences of soft class closure: class members who neither opted out nor registered were bound by the releases but received no settlement distribution, provided they had been adequately notified of that result.