The mid-season quit is the expensive one. A worker who leaves in December can be replaced from the same applicant pool you just hired from. A worker who leaves in February is replaced, if at all, from whoever is still in town and available, at a point when you've already absorbed the full cost of recruiting, onboarding, training and housing them, and with two months of peak revenue still to run.
Four things move that number in practice: the housing, what you told them before they came, when they started, and how you structured the money at the end. Only one of those is about pay.
February is when the math gets real
There's no public dataset that breaks seasonal quits out by month, so treat anyone quoting one with suspicion. What operators in ski and park-gateway towns describe is consistent enough to plan around.
The novelty is gone by February. Holiday overtime and holiday tips have ended, so the paycheck that made the housing deduction tolerable in December has shrunk. The worker now has enough information to price the rest of the season accurately, including the shoulder, when most businesses in Jackson and Summit County cut hours or close outright and there is no income at all. And the decision window for next season's plans opens right about then.
None of that is a compensation problem you can solve in February. It's an expectations-and-conditions problem you had to solve in September.
Housing is the lever, and the rules around it are stricter than most operators think
In markets where a lift op cannot rent a room on a lift op's wage, employer housing is often what makes the job possible at all. That gives you leverage. It also gives you a specific set of legal obligations that get overlooked because the housing sits with facilities rather than with HR.
Start with the honest part. Employer housing is not automatically a retention win. Summit County's housing survey found employer-owned housing less desirable to workers, partly because they feel watched in it, as KUNC reported in 2023. Tying the bed to the job also means that losing the job means losing the home, which workers understand perfectly well and which raises the stakes on every disagreement with a supervisor. Operators who treat staff housing as a control mechanism get the retention outcome that implies.
The improvements that actually register are unglamorous and cheap relative to a build: a room that locks, somewhere warm to dry gear, working laundry, reliable internet, and enforced quiet hours for people on opposite shifts. Tahoe Donner's 2024 workforce housing study, covering an association that runs roughly 100 year-round staff and up to 425 seasonal and J-1 workers at peak, took the same approach worth copying: survey current and former employees about how they're actually meeting their housing needs before deciding what to build.
Now the compliance side, which is where the money is. If you count housing toward wages or deduct its cost from pay, you're using the FLSA section 3(m) credit, and Field Assistance Bulletin 2015-1 sets out what that requires:
- Reasonable cost means actual cost. It cannot include a profit to the employer. A charge benchmarked to local market rent is not a reasonable cost, it's a market rate, and the two are rarely the same number in a resort town.
- Acceptance must be voluntary and uncoerced, under 29 CFR 531.30.
- You have to keep the records. Mortgage or rent payments, utilities, and the wage calculations showing the additions or deductions.
- No credit at all if the lodging violates any federal, state or local ordinance. This is the one with teeth. A dorm over the local occupancy limit, or a converted space without the required permits, doesn't just create a code problem. It can void the wage credit, which can convert a compliant payroll into a minimum wage shortfall across your entire housed crew.
Whether a specific arrangement qualifies is fact-dependent, and state wage laws sometimes impose stricter limits on deductions than federal law does. Worth an hour with employment counsel before the housing agreement goes out with the offer letter.
Realistic job previews work, modestly, and only if they're honest
Telling candidates the hard parts up front is one of the better-studied interventions in hiring. Jean Phillips' meta-analysis of 40 realistic job preview studies found RJPs associated with lower voluntary turnover, lower turnover overall, and higher job performance.
The size of that effect is worth stating plainly, because most retention content oversells it. The correlations are small. Later work summarized by the Quality Improvement Center for Workforce Development found the primary mechanism is enhanced perceptions of organizational honesty, with role clarity second. That finding is the useful one: the preview works because it signals you're being straight with people, not because of the information itself. A recruiting video with one tasteful caveat about "mountain weather" doesn't produce the effect, because it doesn't produce the perception.
What belongs in a preview for a ski-town or park-gateway season:
- Actual season start and end dates, including whether the job ends before the pass does.
- Hours you can guarantee versus hours you expect, and what the schedule looks like in a low-snow January.
- Housing specifics: cost, deduction mechanics, how many people per room, whether it's shared with the opposite shift.
- The commute in February, named. Hoosier Pass from Leadville into Summit County and Teton Pass from Victor into Jackson are both routine and both occasionally closed.
- The shoulder. When the hours stop, what closes, and what people typically do about it.
That last one costs you a percentage of applicants. It is also the single item most likely to prevent a February quit, because it removes the surprise that causes them.
Staggered starts break the all-at-once cycle
Seasonal towns hire and lose everyone in unison, which is why the housing market spikes and collapses on the same two weeks. Summit County housing officials described that synchronized arrival and departure to Summit Daily as the hardest part of the timing problem back in 2019, and nothing about it has improved since.
Splitting your intake into two or three waves, two to three weeks apart, does several useful things at once. Your trainers aren't onboarding two hundred people in four days, which means the training is better and the early-departure rate drops. Housing turns over in manageable batches. Your first wave has enough tenure to mentor the second, which is the cheapest onboarding you will ever buy. And when February arrives, the people hitting their personal wall hit it at different times rather than all in the same week.
The cost is real: staggered starts complicate scheduling, and your earliest hires carry more of the pre-season load. For operations running above roughly 50 seasonal hires, that trade has usually been worth making.
Completion bonuses and the overtime trap
The season-completion bonus is the standard retention tool, and it is where wage-and-hour exposure quietly accumulates.
A bonus you announce in advance to induce people to stay is not discretionary. 29 CFR 778.211(c) is explicit that bonuses promised at hiring, announced to induce employees to remain with the firm, or contingent on continuing in employment until payment is made must be included in the regular rate. The label doesn't matter; the regulation says so directly.
That means 29 CFR 778.209 applies when you pay it. Once the amount is known, it gets apportioned back across the workweeks in which it was earned, and for every one of those weeks in which the employee worked statutory overtime, you owe an additional half-time premium on the bonus-adjusted rate. If the bonus can't be allocated week by week, a reasonable and equitable method is required, commonly an equal amount per week or per hour.
Worked through: a $1,000 completion bonus over a 20-week season allocates to $50 per week. In a week where the employee worked 48 hours, that adds about $1.04 to the regular rate, so the additional half-time premium is roughly $0.52 across 8 overtime hours, or about $4.17. If that employee worked overtime in 10 weeks, you owe them roughly $42 more than you paid. Across 150 seasonal staff, one bonus program, one season, that's about $6,250 sitting unpaid, and back wages don't expire quietly. DOL Fact Sheet 56C works additional examples.
There is a genuine exception that may swallow this for some operators, and it deserves care rather than assumption. Under FLSA section 13(a)(3), employees of an amusement or recreational establishment are exempt from both minimum wage and overtime if the establishment either does not operate more than seven months in any calendar year, or meets a receipts test comparing its six lowest-receipt months against its six highest. Plenty of single-season resorts, camps and outfitters look like candidates.
Three things complicate it. The exemption is analyzed establishment by establishment, and "establishment" means a distinct physical place of business. There is a carve-out for private entities providing services or facilities in a national park, national forest, or on National Wildlife Refuge land under contract with the Secretary of the Interior or Agriculture, which reaches a lot of gateway-town concessioners. And DOL's own fact sheet warns that some state wage laws don't recognize the exemption at all, in which case the stricter state rule governs. Do not decide this one from a blog post, including this one. It's a counsel question, and getting it wrong exposes the same payroll in both directions.
What to change before September
Pick the two cheapest items on this list and do them for the coming season: write an honest preview with real shoulder dates in it, and have someone who isn't in facilities read your housing agreement against the section 3(m) requirements. Both are a week of work. Both pay back in February.
If you're building next season's crew, you can post roles on TurnSeasons, list staff housing alongside them, and see what other operators in your town are advertising. Listings that name the housing terms up front tend to attract the people who were going to stay anyway.