Fixing a staffing crisis with the levers that actually hold people

Six night-shift operators gone in two years is not a personnel problem. It is a budget decision. Every time your board leaves a raise on the table because it looks bigger than the alternative, they are choosing the more expensive option without knowing it.

You know the pattern already: post the opening, interview whoever answers, train them for months, and watch them leave the day a neighboring system offers a better schedule or fifty cents more an hour. Then you cover the empty shift yourself, on overtime, while the next hire studies for the license the last one walked out the door with.

Still rung 2, running the system. It treats turnover the way your finance training treats a rate case: waiting costs more than acting, and the bill lands on you either way.

Why operators actually walk

Ask around and four reasons keep coming up, roughly in this order: pay that fell behind the market, on-call that never really ends, disrespect from the people above them, and no path to more money or more responsibility.

Pay lag works like a stale rate schedule. Hold a wage flat while the cost of living keeps moving, and you are not looking at a modest correction anymore. You are looking at a resignation letter.

Operators compare notes with people at other plants constantly, and the same line comes up again and again when they explain a move: they switched employers to chase better pay, and some admit afterward it was not always worth it. That does not mean the new job was better. It means the raise that would have kept them was smaller than the one it took to make them leave.

On-call burnout is the driver boards understand least, because it rarely shows up as its own line item. Disrespect covers a wide range, from a toxic senior operator nobody corrects, to a supervisor who treats every question as a complaint. And no ladder means an operator can see exactly how far this job goes, and it is not far enough.

What on-call is actually costing you

On-call time is unpaid, but it still owns your operator's evening. How utilities compensate for it varies enormously from one system to the next. Reported stipends run from about $22 to $25 a day on the low end, up past four figures a week at the best-paying shops, plus a two-hour overtime minimum once a call actually comes in. The full range, and how deep a rotation needs to be, gets its own guide: the on-call rotation guide.

Whatever the number, put a real response window in policy. Build it around where your crew can actually live, not a distance measured on a map.

Treat every figure as a reported range, not a rate to copy. Before you set or change on-call pay, check it against your union contract and your local policy, the same places these terms actually get locked in.

One reframe is worth borrowing for the negotiation itself: on-call is not a favor you are asking for, it is time you are buying. If the job controls what an operator can do on their own time, that time has a price. That argument tends to land better than a pure dollar figure, and the version that sticks gets written into a contract, not just promised out loud.

The levers that actually hold people

Four things show up again and again in how operators describe staying somewhere on purpose.

• Tie certifications to pay. Stack licenses so that earning the next grade is also

• Put raises on autopilot: tenure step-ups plus annual cost-of-living increases, so pace

• Make evaluations move pay. A review that never changes a paycheck teaches people that

• Build a documented ladder, operator to lead to supervisor, with what each step

earning the next pay step, not just a new line on a resume. with inflation isn't a yearly fight. performance does not matter here. requires and what it pays, so someone can see the whole staircase before taking the first step.

A promotion is not automatically a win. Moving a union operator into a superintendent role for modest extra money can cost them their union representation and protections, along with some vacation time. If you offer someone a step up, get the new role's terms in writing before they say yes, and make sure they know what they are trading away.

The knowledge that leaves with them

Every departure costs more than the empty seat. A veteran who has run your system for decades knows things no manual wrote down: how to coax a particular valve through a freeze, which readings actually mean trouble at your plant versus a normal quirk. When that person retires, the knowledge leaves with them unless someone captured it first. That risk alone is worth treating retention as urgent, separate from the dollar math. It gets a guide of its own: capturing knowledge before it retires.

Integrity sits right next to this. Whoever signs a report is liable for it. A coworker who logs a sample they never ran is not taking a shortcut. They are ending a career and inviting criminal exposure, for themselves and for whoever signed above them. A system running short-staffed and short on senior people is exactly where that pressure shows up, because nobody experienced is left in the room to say no.

Putting the cost of churn in front of your board

Your board already knows the argument for small, steady rate increases over one shocking catch-up jump. Both paths land at the same rate eventually. The difference is that the steady one collects money the whole way and never asks for a vote that makes the room flinch.

Turnover runs on the identical logic. A modest raise held every year, on schedule, costs less than the overtime, the recruiting, and the training it takes to replace the person who leaves without it.

There is also a documented link between what you charge and who you can keep. Systems charging $15 a month post SDWA health-compliance and management-compliance rates of 93.9 percent and 68.4 percent. Systems charging $75 a month post 98.9 percent and 85.9 percent. Underpriced service does not only starve capital projects. It starves payroll too, and low rates correlate directly with higher staff turnover. If your board resists a raise for operators, ask them to look at the rate schedule first. The staffing problem and the rate problem are often the same conversation held in two different rooms.

When you bring a raise request to the board, frame it the way you would frame a rate case. Name the cost of waiting, not just the cost of acting: what the next six months of overtime costs to cover the shift empty, what the next hiring and training cycle costs in your own time, and what license walks out the door if the next departure is your most senior operator. A board that would never approve a surprise 31 percent rate hike should not be surprised by the staffing version of the same shock.

A decision rule for the next raise request

• If the raise costs less than six months of overtime to cover the shift empty, it is

• If the person holds a license or certification the system needs for compliance,

• If on-call pay has not been checked against your contract recently, check it before

• If you cannot point to a documented ladder for this role, build one before the next

very likely the cheaper option. treat the request as a compliance cost, not a personnel favor. the next negotiation, not during it. opening. Its absence is already pushing people toward the door.

Next on the ladder: controlling what the plant takes in.