Three decisions issued within three weeks of each other this summer have fundamentally changed how California employers and their counsel should evaluate arbitration agreements covering delivery drivers, couriers, yard workers, and anyone else in the chain of moving goods. The cases are Betanco v. Living Spaces Furniture, LLC, Doss v. Tesla, Inc., and In re Rebecca Orr. Taken together, they establish that a significant category of California workers may be entirely outside the reach of the Federal Arbitration Act — and that courts must resolve that question before ordering anyone to arbitrate anything.

Two decisions dropped this month that, taken together, make the same point from opposite directions: in California wage and hour litigation, the numbers matter — and so does who ran them.

If you defend California employers, you don't need a report to tell you the exposure is real. But the numbers are still worth putting in front of clients who think a single missed break policy or an outdated timekeeping system is a minor administrative gap rather than active litigation risk.

If your payroll system was still running January's rate table this week, you're already behind.

Eleven California cities and counties raised local minimum wage rates effective July 1, 2026: Alameda, Berkeley, Emeryville, Fremont, Los Angeles, unincorporated Los Angeles County, Malibu, Milpitas, Oakland, Pasadena, San Francisco, and Santa Monica. Six of those jurisdictions layered separate hotel, airport, or event-center rates on top. The statewide floor stays at $16.90/hour — but where a local ordinance sets a higher number, the local number controls, full stop.

That's the easy part. The part that actually creates exposure is timing.

The most common ways plaintiff's PAGA penalty calculations are inflated — denominator errors, initial vs. subsequent penalty conflation, improper stacking — and how a record-level forensic audit produces a defensible counter.

When timekeeping data enters discovery in a wage and hour class action or PAGA matter, the format and completeness of the records often matter as much as the content. A checklist of what to request, what to look for, and what the data cannot tell you without operational context.