A complaint channel is a route the crew uses to raise a problem the day it comes up, with a person named to receive it and a date for the answer. The unsigned survey is one page of short questions that each person answers alone, with no name. To hear of a crew problem before a good hand quits, give the crew a route that skips whoever the complaint is about, and run the survey twice a year. Labor experts often put the cost of one departure at about 150% of the worker’s annual wages. For a hand paid C$50,000 a year, that is C$75,000, an example built on a rule of thumb from outside farming. Without the route, the complaint about the foreman reaches you on the day the hand asks for final pay.
Does the crew leave over the boss or over pay?
Pay comes first in both of these farm studies. A 2026 study in the journal One Health asked workers with a TN visa on US pig farms why they might change employer. 36.7% named poor compensation and benefits and 30.0% named poor management. The TN visa is a US work permit for Canadian and Mexican workers with a degree, and these workers came predominantly from Mexico.
In a 2009 survey of past jobs held by dairy workers in three California counties, 29% of departures were over compensation and benefits and 8% over relations with management. The boss ranks lower in the California count than in the pig-farm study, and the two measure different things: departures that happened, and thoughts of leaving.
The California workers reported not getting along with “either their direct supervisor or the dairy owner (sometimes one and the same person)”. Pay is its own question. A complaint about the boss needs another way out, because a complaint about the foreman has to pass through the foreman, and a complaint about the owner has nobody further up. Telling the crew your door is open leaves the worker to say it to the boss’s face.
Without a route, the boss shows up in your turnover count only after the worker has left.
Who can take a complaint about the boss?
It depends on who runs the crew. On a small farm that person is often the owner, and then the owner cannot be the route around the boss. For 2015, the 2016 census of agriculture counted 9,565 farms in Alberta, 23.5% of all farms, with 33,498 year-round, seasonal or temporary employees. Dividing the two figures gives about 3.5 employees on each of those farms, and 50.3% of the employees were seasonal or temporary.
If a foreman runs the crew between you and them, you can be the route around the foreman. If you hired a farm manager, write down what the manager may decide without asking you, as in the article on what your manager decides alone. A complaint about one of those decisions then reaches you without passing through the manager. If you run the crew yourself, the route has to sit outside the farm’s management, such as a relative who is nobody’s boss, or the phone number of an outside authority.
GLOBALG.A.P. is a farm certification that many buyers ask their suppliers for, and its working-conditions module, GRASP, binds only the farms that seek it. Its rules still make a handy checklist. GRASP requires “one or more well publicized places to file complaints, at least one of which should be independent of the supervisory staff”. It counts the foreman, the manager and the crew supervisor as supervisory staff.
Every place has to offer an “option to file anonymously”. When the crew works far from the farm buildings, one place to file a complaint has to be “a designated person of the crew at the field”. A complaint made to that person is not anonymous, which is why the unsigned survey runs alongside the channel.
Post the routes and the deadline in the crew’s language, or in pictures for anyone who cannot read, as GRASP asks. They belong on the list of what a new hire gets on the first day, so people know the channel before they need it.
A complaint of harassment, or of someone breaking a rule of conduct, needs a consequence settled beforehand, which is what a code of conduct each person signed is for. This channel takes the everyday problem, such as a short paycheck, the work roster or the way a foreman talks.
How soon must you answer a complaint?
By a set date, because the crew judges the channel by what happened to the last complaint, not by the box on the wall. GRASP defines a timely answer as “in general at most 30 days after filing or before the last day of employment”, whichever comes first. A complaint counts as resolved with “a positive or negative answer, or no resolution with explanation of why there was no resolution”, so an honest no closes it. For an anonymous complaint, “the answer shall be included as a general notice posted on workers notice boards”.
The Fair Food Program covers tomato and other US farms. It reports that workers who complained to supervisors about missing pay or unsafe conditions were frequently subject to retaliation. Those workers often concluded that raising complaints at work was not possible. Over the life of the program, 52% of cases were resolved in less than two weeks and 79% in less than a month. The program also has audits and buyers who can drop a farm, so its numbers do not show what a deadline alone does. The 30 days are a written standard, not a measured effect.
Each complaint gets one line in the log: the date it came in, what it is about, who handles it, the date the answer is due, and the answer with its date. The due date is the day the complaint came in plus 30 days. If the worker’s last day comes first, the answer is due before it, so a seasonal hand hired for the harvest gets an answer before leaving. The log below belongs to an invented grain farm in Alberta with a crew of four.
| Came in | What it is about | Who handles it | Answer due | Answer, and when |
|---|---|---|---|---|
| Sept 2 | The foreman shouts at new hires (brought straight to the owner) | Owner | Oct 2 | Talked with the foreman; told the worker, Sept 9 |
| Sept 5 | Two days of overtime missing from pay (seasonal hand, last day Sept 20) | Office | Before Sept 20 | Paid with the next check; told the worker, Sept 7 |
| Sept 6 | Anonymous, box: no drinking water at the far field after 10 a.m. | Owner | Oct 6 | Water tank moved; notice posted, Sept 8 |
| Sept 12 | Wants every Saturday off | Owner | Oct 12 | No, at harvest the weekend roster needs two combine operators; told the worker, Sept 14 |
When the same complaint comes in a third time, the first two answers did not fix it, and the problem may sit in how the work is set up.
What can you learn from an unsigned survey that the channel misses?
You learn the problem that many people share and nobody raises alone. Asked face to face, people tend to answer what is safe. In the pig-farm study, 211 workers answered in person and 50 answered online, and the in-person group was more likely to report higher job satisfaction and less likely to consider a job change. The authors give two possible reasons: the in-person respondents may have feared retaliation and given the answers that look good, or the online respondents may have come from companies with poorer conditions. The study cannot tell which.
A 2013 study of eight dairies in four US states handed each worker the questions on paper in their own language. The worker then called a bilingual interviewer and gave the answers by phone. Of the eligible employees, 78% were interviewed. Of those who answered, 47% said training was offered never, or only when they started. The authors write that employers underestimate how much their employees want to learn.
The same workers rated their relationship with the supervisor at 4.43 out of 5 and with co-workers at 3.76. The count tells you which subjects weigh on the crew, and a low result is never read as a verdict on one person.
Ask about five or six things that weigh on staying, such as pay, training, the chance to move up and how people are treated. Write each as a short statement, such as “My pay comes on time and correct.” Give everyone the same choices (yes, no, not sure) and add one open question: the one thing they would change. Collect the sheets through someone who is nobody’s boss, or have people drop them folded into a sealed box. On a crew of six, even an unsigned count can point to who answered, so do not split it by age, language or job.
Count each question on its own, because one total for the whole crew lets a good result on pay hide a bad one on training. Then tell the crew what you counted and what you will do about the three biggest complaints, or they have less reason to answer the next survey.
Who do you name first, and by when?
An afternoon to write the routes, the deadline and the survey page, and one talk with the crew.
Six months in, lay the survey count next to the complaint log. When half the crew marks a complaint on the survey and nobody brings it to the channel, the route is not reaching them. The next step is a route held by someone outside the crew’s chain of command.
More on hiring and leading a crew is in the library’s section on people, and more on the rest of running a farm in farm management.