In an example apple orchard in the Yakima Valley, in Washington State, four of the eight year-round workers left in one year. The owner guessed $3,000 for each departure and wrote down $12,000. The last replacement, added up from payslips and invoices, cost $4,500, so the real bill was $18,000: $6,000 that the guess left out. These are example values, not data from a real orchard. To know what turnover costs your farm, list each departure, price your last replacement from your own paper, and write down the crew size you divide by.
Worker turnover has two results. The turnover rate is the share of the crew that left in the year: the people who left divided by the size of the crew. The cost of turnover is the departures that left a job empty times what one departure costs you. The turnover calculator on this site does both sums in a minute. Three choices move the answers: the months you count, the crew size you divide by and the cost per departure. A turnover record, one page that holds those choices, lets you set next year’s figures beside this year’s.
At the end of a season the crew is back to its usual size, so few farms count how many came and went in between. The December payroll looks like the January one. Each departure that left a job empty still cost something: the days the work went undone and the months a new hire needed to learn it.
Which crew size should you divide the turnover rate by?
Any of three, as long as you write down which one and use the same one next year. You can divide by the crew of one month, by the average of the start and end counts, or by the average of the monthly counts. On a farm with a harvest crew, the three give very different rates.
Zaballa and colleagues published a method for analyzing staff turnover in a peer-reviewed journal in 2021. Their average crew is the sum of the values at the start and at the end of the period, divided by two, and the calculator asks for that average under “Average workforce over the period”. Another of their formulas adds the head count at the start of each month and divides by the number of months. The method was written for any workplace, not for farms, but the arithmetic is the same on any payroll.
The example orchard keeps eight people all year and hires thirty-six pickers for a four-month harvest, so the crew peaks at forty-four. Forty people left in the year: the thirty-six pickers, when picking ended, and four of the eight year-round workers, who were replaced. Every figure in this orchard is made up to show the arithmetic.
Nothing changed in the orchard between 500%, 200% and 91%. Only the crew size it was divided by changed. None of the three is the correct one. Two rates compare only when they divide by the same kind of crew size, so write beside the result which one you used, with its dates, and use it again next year.
The months you count matter for the same reason. If you count January to December, pickers hired in November and let go in February show up as hires in one year and as departures in the next. Write down the first and the last month you counted, and count the same months next year.
Which turnover rate answers your question, and is 200% a lot?
There are three rates, and each answers a different question. Zaballa and colleagues give a formula for each. The calculator gives the first two, and you work out the third by hand.
The first, which the calculator calls the classic rate, averages hires and departures and divides that by the crew. The authors use it to measure the turnover index for the purposes of human resources planning: on a farm, when you plan next season’s hiring.
The second, the separation rate, counts only departures. The authors use it to analyze the losses of people and their causes, and with it you see how much of the crew you had to replace.
The third counts only resignations, the departures the worker chose, and leaves out the contracts you ended. The authors use it to study the departures that come solely from the attitudes and conduct of the staff. The calculator has no box for who decided each departure, so it cannot give this rate.
In the orchard, the separation rate is 200%, which makes it look like a place that loses people all year. Thirty-six of the forty departures were pickers whose contracts ended on the date written on them. Four were year-round workers who resigned. The resignation rate, 4 ÷ 20, is 20%, and that is the figure that tells the owner whether people are walking away.
To work out the third rate, add one column to your list of who left, and mark on each line whether the person chose to leave or you ended the job. Count only the ones who chose to leave and divide by the same crew size.
No published figure says whether 200% is a lot for a farm. A study of nurses in a Brazilian hospital emergency unit worked the rate out month by month. It notes that there is no consensus on an acceptable rate and that some authors put it at about 3%. That is a monthly rate from another kind of work, and it cannot be set beside a yearly rate.
The rate to compare yours against is your own rate from last year, worked out over the same months with the same crew size. A rate puts a price on turnover and does not show what to change. For that, an article on which crew number shows what to change suggests tracking how long the people who left had been with you.
Why does the month someone leaves change what it costs?
Because a new hire takes months to reach full output. Jiří Duda and a colleague, management researchers at Mendel University in Brno, in the Czech Republic, wrote a paper on the costs of employee turnover in business, not on farms. In it a new hire works at 25% of normal output in the first month, 50% in the second and 75% in the third.
Output on a farm is not worth the same all year. One of the orchard’s four resignations was the sprayer operator, who left in early August, three weeks before picking began. His replacement works at a quarter of normal output in the first month, and that month falls in the weeks when the work cannot wait. Had he left in January, the same learning months would have fallen in winter, when the work can wait.
On the payroll the two departures look identical, and in the turnover rate each counts as one. The August one can cost far more.
The same paper counts the output lost while the replacement learns as an important part of what a departure costs, next to the cost of covering the job in the meantime and the manager’s time. It leaves no invoice, so write it down yourself: the month of each departure, the days the job sat empty and who covered the work.
Where does the cost of one departure come from, and which departures do you price?
From your own records: the wages, overtime and invoices of your last replacement. The cost you use is multiplied by every departure you price, so an error in it repeats each time. The orchard’s owner guessed $3,000. The payslips and invoices for the last replacement came to $4,500.
The orchard prices the four resignations and leaves out the thirty-six pickers. Their contracts ended on the date written on them, and nobody was hired to replace them. Duda and Žůrková say it matters whether the leaving was planned, because an unplanned leaving adds extra costs while the job sits empty without a replacement. Write down which departures you left out of the bill, and why.
The calculator multiplies every departure you enter by one cost. Run it once with all departures for the rate, and once with only the priced ones for the cost. To see which departures a different decision might have prevented, the article Many short absences or a few long ones shows how to read a crew’s absence record.
If you have no replacement on record, the nearest published figure is in an issue brief by Kate Bahn and a colleague for the Washington Center for Equitable Growth, a United States economic research group. It pools thirty-one case studies of jobs in healthcare, education, hospitality, finance, retail, transportation and manufacturing; farming is not among them. Replacing a worker cost on average 39.6% of the position’s annual wage, with a median of 23.5%. Borrow the median: half the case studies found a lower cost and half a higher one, while a few expensive jobs pull the average up.
For the cost lines of a replacement, see the article on what goes into the bill for replacing a worker.
What do you write first, and in how long?
Set aside an hour with last year’s payroll, the crew list and the paperwork from the last replacement, before the next hiring season opens.
The most useful lines of the record are the departures a different decision might have prevented, each with that decision beside it. With them, the talk with your foreman before the next season can be about whether people left over pay, schedule or treatment, instead of about whether 200% is a lot. For the other records and decisions that go into running a farm, start at the farm management guide on this site; the people library has more on hiring, pay and keeping people.