What Is the Average Restaurant Staff Turnover Rate?

In the United States, accommodation and food services had an average monthly quits rate of 4.2% in 2025, about twice the 2.0% for all industries (BLS JOLTS). Singapore's food and beverage services averaged 2.2% resignations a month against 1.2% overall (MOM). In Japan, 23.7% of accommodation and food service workers left their employer during 2025, against 13.7% across all industries (MHLW). The UK publishes no official hospitality turnover rate.

The official benchmarks

These are the only national figures we could trace to a statistics agency. They measure different things, so compare each one with your own figure in the same unit, not with each other.

The direction is the same in the two markets with a time series. The US rate peaked at 5.8% in 2021 and 2022 and has fallen to 4.2%. Singapore's F&B rate fell from 2.9% in 2022 to 2.2% in 2025. Turnover is still about double the national average, but it is well off the peak.

Seniority matters more than sector. Within Singapore F&B, rank and file staff resign at 2.8% a month and professionals and managers at 1.2%. A head chef on a written career path is a different risk from a line cook.

Any UK "hospitality turnover" percentage you see quoted comes from a private survey, not from ONS. Ask for the sample before you rely on it.

How to read a monthly rate

BLS defines its annual figure as the sum of twelve months of quits divided by the sum of twelve months of employment. It is an average monthly rate. Multiplying 4.2% by twelve gives about 50 quits for every 100 jobs in a year, and you will see that quoted as "half the staff leave every year".

That reading overstates it. Quits are counted as events, not people. One line cook post that turns over three times in a year produces three quits, while the ten people beside them stay. The share of individual workers who leave is lower than 50%, and BLS does not publish a person based figure for this industry. So the fair statement is: somewhere below 50 quits per 100 jobs per year, concentrated in a minority of high churn posts, before dismissals and layoffs.

Japan's 23.7% is closer to a person count, because MHLW divides the year's separations by the number of employees at the start of the year. It includes every reason for leaving, not only resignations.

The other half of the picture is how hard you are to replace. MHLW's ratio of job openings to applicants for food preparation workers (飲食物調理従事者, including part time) was 2.18 in July 2026, against 1.05 for all occupations; excluding part time it was 1.83. A chef in Japan has more offers than a typical worker, which is why a replacement from Japan takes months, not weeks.

Work out your own rate

Your own figure is worth more than any national one. Fill this in from payroll for the last twelve months, kitchen only.

If E is shorter than your probation period, the problem is hiring or the first weeks, not pay. If one station produces most of A, fix that post before anything else.

What one leaver costs

We could not find a citable cost per hire for restaurants in any country. The last authoritative Japanese figure is all industry and six years old: Recruit's 就職白書2020 put the average cost of a mid career hire at ¥1,033,000 in FY2019, and Recruit stopped publishing the measure after that edition. Use it as a sense of scale, not as a benchmark.

Build your own from lines you can count:

• Advertising, agency or referral fees for the replacement.
• Management hours spent screening, interviewing and running trials, at a real hourly cost.
• Overtime and agency cover while the post is empty (line G above).
• Senior staff time training the new person, and the weeks before they work at full speed.
• For a sponsored chef: government fees already paid and not refundable, plus the same fees again for the replacement, flights and relocation.

Worked example: a sponsored chef resigns in month five

Take an Australian restaurant with turnover under A$10 million that sponsored a chef on a four year Skills in Demand (subclass 482) visa. The official fees are as published for 2026 to 27; the firm specific lines are left for you to fill in.

Two points decide the size of the loss. First, the refund: Home Affairs guidance allows a refund of the unused full years of the levy when a sponsored worker leaves the employer within the first twelve months of employment on a visa granted for more than twelve months. We found this rule in the department's guidance written for the TSS visa that the Skills in Demand visa replaced; confirm it applies to your nomination before you count on it. The refund is not automatic. Second, the calendar: the replacement nomination needs labour market testing, two advertisements running at least four weeks within the four months before you nominate, before any processing starts. Budget for the post to be empty or covered for at least that long.

Singapore works differently. The S Pass levy is S$650 a month, so a chef who leaves in month five has cost S$3,250 in levy that bought five months of work; nothing further is lost, but nothing is recovered either. The real cost is the vacancy. In the UK, a sponsored chef who leaves often cannot be replaced from overseas at all, because Chefs (SOC 5434) are on neither the Immigration Salary List nor the Temporary Shortage List; see UK hiring without sponsorship.

What this means for your budget

• Carry a replacement reserve. Take the "paid again" column plus your firm lines and hold it as a contingency for every sponsored post in its first year. The first year is where early departures cluster, and where the Australian refund window sits.
• Spend on the first weeks, not on recruitment. If your average leaver tenure (line E) is under six months, money moved from advertising into arrival support will usually return more. See retaining Japanese chefs.
• Hire senior where you sponsor. The Singapore split between rank and file and managers is the clearest evidence here: people with authority stay longer.
• Start the next search early. If a chef tells you they plan to go home next year, begin advertising now. An international replacement is a matter of months.

Figures are as published by the agencies named, for the periods stated. Where no official statistic exists we say so rather than substituting one. The worked example is an illustration of the published fees, not advice on your nomination.