Bartending doesn't launder a manager's cut of the tip pool. That's the entire lesson of a recent Department of Labor Opinion Letter.

Here's the setup. A restaurant has servers "tip out" a percentage of sales to bartenders, hosts, and bussers. One employee—titled "shift supervisor"—periodically works bartending shifts. While bartending, he also sets schedules, decides when shifts end, and handles other management functions. He collects a tip out from the servers. He also grabs a slice of the tips meant for hosts and bussers when he pitches in to help them.

Can he keep any of it? The answer is a clear, "No."

If this employee's duties satisfy the executive exemption test—primary duty of management, customarily directs two or more full-time employees, and has real hiring/firing/promotion authority—he's a "manager or supervisor" under the Fair Labor Standards Act. Full stop.

It doesn't matter that he also tends bar. It doesn't matter that he's helping out servers who are shorthanded. Once you meet the duties test, you're locked out of other people's tips, period.

Title doesn't control this. Duties do. A sometimes-bartender with no manager title who schedules staff, orders inventory, and helps hire people is still a manager for tip purposes, and still can't touch the tip jar.

There is one important exception. A manager can keep tips a customer hands her directly, for service she personally and solely provided. If she tends bar and a customer drops a five in her personal jar or in her hand, she can keep it. But the second the tip get pooled and split with the other bartenders and workers on shift, she is out—because now it's impossible to say the tip was for her alone.

1. Stop looking at the title. Look at the duties. "Shift supervisor," "lead," "keyholder"—none of that matters to WHD. What matters is whether the person is directing staff, setting schedules, and carrying real hiring and firing weight. If yes, they're out of the tip pool no matter what the badge says.

2. Segregate any tips a manager legitimately earns solo. If a manager works a shift doing hands-on service work, make sure any tips attributable only to that manager stay separate from the pooled tips. Mixed in with everyone else's, they're contaminated—and now the manager can't keep any of it.

3. Know what it costs you if you get this wrong. This isn't just a "give the money back" violation. Improperly letting a manager keep tips also blows the tip credit for every tipped employee affected. That turns a tip-pool mistake into a minimum-wage violation, with all the back-pay exposure that comes with it.

The math is simple: manage, and you don't get tips. Pick one.

The Department of Labor just answered a question a lot of employers have been wondering about: if an employee splits her day between home and office, who pays for the drive in between?

TL;DR: Splitting the workday between home and office doesn't mean paying for the commute.

Start with what's never been in dispute: the drive to the office in the morning and home at night has never been compensated. That's true no matter how far away the employee lives, how bad the traffic is, or how many hours she works once she gets there. It's an ordinary commute, and ordinary commutes have always been on the employee's own time.

The harder question is what happens when that commute moves to the middle of the workday. A new Wage and Hour Division Opinion Letter works through three real scenarios.

One employee dodges rush hour by working from home in the morning, driving in around 10, working until 3:30, driving home before traffic hits again, and completing their work again from home. Another wants to knock out a volunteer project at home before her regular shift, instead of coming in early. A third rides the city bus, can't finish before the last one leaves, and asks to take the rest of his work home instead of staying late.

In each case, the employee travels between home and office in the middle of the workday. Under the continuous workday rule, once an employee starts working for the day, "travel from job site to job site during the workday" is paid. For that reason, some employers have assumed that the mid-shift commutes described above are compensated.

They aren't. The continuous workday rule does not come into play here, because, according to the DOL, it's all still just a commute.

WHD draws a clean line: an "ordinary" home-to-work commute isn't work whether it happens at 7 a.m., 6 p.m., or noon. It goes a step further, recognizing a mid-day commute as its own category of unpaid time during the workday, alongside bona fide meal breaks and off-duty time.

The test was never the clock. It's who benefits. If the mid-day trip is the employee's idea, taken to beat traffic, catch a bus, or carve out home time, it "primarily benefits the employee." That makes it ordinary, even if it cuts her total drive time in half.

Two things haven't changed, though.

1. Work is still work. The hours the employee actually spends working at home, finishing that volunteer project or wrapping up the day's assignment, get paid in full. Only the drive itself is free.

2. Job-site-to-job-site travel still counts. A repair tech who drives from the office to a customer's building and back is on the clock the whole time. This letter is about commuting home, not bouncing between worksites.

If you're building a hybrid or split-shift option, make it genuinely optional, let employees pick their own timing, and keep timekeeping honest about hours actually worked away from the office. Do that, and the commute in between belongs to the employee. Not your payroll.

For those who are unfamiliar with my podcasting side hustle, for the past five years I've co-hosted a show with my daughter, Norah. We just launched season 5 of The Norah and Dad Show, and for the first time it's transatlantic.

Norah recorded from Aix-en-Provence, France, where she's spending the semester in a French immersion program. While her audio quality may have suffered a bit without her usual recording rig, the conversation most definitely did not.

On this episode we check in on her first week living in France. We'll be back with new episodes every other Tuesday with a new episode.

Here is short clip to whet your appetite.

You can find this week's episode full episode on Apple Podcasts, Spotify, YouTube, in your browser, and everywhere else you get your podcasts. If your not yet subscribed, please consider doing so.

---

Here's what I read this week that you should read, too.

Most employers worry about the thief outside the building. Yet, if someone's stealing from you it's probably an employee inside it.

Jennifer O'Neal worked as a program specialist for a Georgia nonprofit, CASA of Polk & Haralson Inc., which advocates for abused and neglected children. From 2018 to 2022, prosecutors say she used her access to the organization's bank accounts to cut payments to herself, disguised as legitimate reimbursements.

The total: $96,700.

The money went to her power bill. Her water and sewage bill. Her cellphone bill. A Netflix subscription. Softball gear. Lingerie. Home theater equipment. A trip to Six Flags.

O'Neal pled guilty this week to theft of federal program funds.

U.S. Attorney Theodore S. Hertzberg said O'Neal violated CASA's "core values" by stealing money meant for the "most vulnerable members of our community." That's the moral lesson of the story. Here's the operational one: nobody was watching, and for four years, that was enough.

You don't need a U.S. Attorney or a forensic accountant to stop this. You need controls that don't rely on trust alone.

1. Split the job in two. The person who requests or initiates a payment should never be the same person who approves or issues it. O'Neal could do both. That's not a personality flaw. That's a design flaw, and it's on you to fix it.

2. Demand real backup for every reimbursement. No receipt, no verifiable business purpose, no payment. Period. Trusted employees get the same scrutiny as everyone else, because trust is exactly what gets exploited.

3. Audit on a schedule, not a hunch. Someone with no stake in the transactions should be reviewing the books regularly, whether or not anything looks wrong. Fraud isn't caught by intuition. It's caught by someone actually checking.

4. Force people to take time off. The employee who never takes a vacation and guards their process isn't dedicated. They're unsupervised. Make someone else cover the role while they're out, and let a fresh set of eyes look at what's been happening.

5. Reconcile every payee against a real vendor list. A personal name or a residential address where a vendor should be isn't a clerical error. It's a red flag with a bow on it.

6. Require dual signatures above a set dollar threshold. And have a person outside accounts payable actually read the bank statements every month. Not skim them, actually read them.

None of this is complicated. None of it's expensive. It just requires believing, before the fact, that someone you trust might not deserve it.

Trust is not a control. Verification is.

Two days.

That's the entire gap between John Shoemaker asking McKesson Corp. for FMLA leave and McKesson telling him he was fired. 

A federal judge just said that two-day gap alone survives summary judgment.

Shoemaker was a Senior Director overseeing McKesson's contact center. In the spring of 2024, McKesson started planning a reorg. The company says that it locked in the decision to eliminate Shoemaker's position by December 2024, at least a month prior to anyone at the company knowing that he was even thinking about taking paternity leave. Typically, that's enough. Business decisions made for business reasons aren't retaliation just because a protected request comes along later.

But then it got messy. In January 2025, Shoemaker started quietly talking to HR about paternity leave. On March 3, he formally told his boss he'd be taking it that spring. McKesson says that in early March it picked April 11 as Shoemaker's termination date, but couldn't pin down exactly when, or produce anything showing that it did so before March 3.

Then, on April 9, Shoemaker formally requested FMLA leave. Two days later McKesson fired him.

Shoemaker was the only employee terminated on April 11. Everyone else caught in the reorg wasn't let go until July.

That's not just bad timing. That's a company that can tell you what it decided, but not when. And in a case like this one, the "when" is the whole ballgame.

Retaliation claims still require more than pure timing to survive summary judgment, and temporal proximity alone doesn't prove pretext. But in this case, Shoemaker had something more. He said his work didn't stop when he left. A coworker backed him up, testifying his job functions "did not cease." And unlike everyone else swept up in the RIF, he alone got cut two days after invoking the FMLA and months before the rest.

The court didn't need much more than that. Between the near-immediate proximity and the factual fight over why only Shoemaker's termination was accelerated, summary judgment was denied.

So what should you take away from this case?

1. Contemporaneous documentation matters. If your defense is "we already decided to fire this person before the protected activity," you need a document with a date on it, not a witness remembering "sometime in early March." If it's not on paper, it didn't happen.

2. If you jump someone ahead in line, know why. When one employee gets cut days after protected activity and everyone else in the same RIF survives another three months, you have to be able to explain that gap.

3. The line for what qualifies as a adverse action is a thin one. You don't get credit for "we were going to fire him anyway" if the record shows you fired him sooner because he took leave. Acceleration counts. 

Employers, plan your RIF all you want. Just make sure you can back up not just why a decision was made, but also when.

Big firm ability. Small firm agility. Reprehensible conduct, forgettable price tag The fist inside the velvet glove Mathing the overtime calculation for bonuses and incentive comp WIRTW #809 (the 'senioritis' edition)