We get this question quite a lot and this is one way of answering it.
In the beginning of the evolution of a circle, there is only the circle’s aim, which is policy, and operations that support the aim. For instance, a garden circle grows food in a garden. The garden is the domain, a set of garden-workers (or otherwise involved) are the circle members, and the aim is growing food. Buying seeds, planting, watering, harvesting, these are all operations. So where does policy come in? In the beginning, everyone works enthusiastically, and everyone is doing what (in their minds) supports the aim. Without policy, everyone can do anything within the circle’s domain. (Domain means: garden circle cannot make decisions about the basketball court right next to the garden.) But everyone doing their own thing typically won’t go for long. For instance, members of that garden circle could be planting whatever they want in a garden. Then this leads to issues and the circle decides (by consent) that in wintertime they will make a plan and decide which row will have what kind of plants in that year. Now circle members can put those plants in according to policy but they still decide how far spread out they put in plants in carrying out the policy (i.e. in the concrete act of planting). Policy reduced the options the individuals had. Policy specifies and therefore constrainst the options. Not everything and anything can be done anymore.
Now the circle members are happily planting tomatos in the tomato row, and carrots in the carrot plot. In the herb garden, there is still free choice and everyone is fine with that. Let’s imagine the spacing of individual tomato plants leads to issues because planting tomato plants to close makes it easier for diseases to spread. Even though this is known, not everyone pays attention, so now the circle might decide to make policy around that. From now on, tomato plants need at least 2.5 feet spacing, and every circle member planting tomatoes will follow that. (Again, remember that the circle members constrainted themselves/each other by choice and by consent because they assume that the policy will make their work more effective. After all, they want their tomato plants to do well.) Are we going to make policy about everything? No. We have to hit the sweet spot where everything that will likely lead to issues or has led to issues in the past is framed by policy, without micromanaging the work. For instance, the circle member will still decide how much water to use after planting, and which individual tomato plant goes where.
How much policy the circle members want to give themselves is highly context-dependent. That is why it makes sense for circles to each decide for themselves. For instance, how much water to give a newly planted tomato plant might not need policy in a garden circle of a community (to me, personally, that would be micromanaging me to tell me that every tomato plant needs 2 gallons of water and I’d have to measure it out!). In a plant nursery, there might be policy around that, and that would be appropriate for that circle/aim because their context is different.
Policy decisions are only made where there is a reason to do so. We don’t start out defining every possibility, we start from a place of trust wherever possible and safe and start from there. Working together is the focus, and making policy that supports our working together. It is important to see that operational decisions are made by (sets of) individuals but they have to be made within the frame of the circle: any operational decision has to be compatible with the circle’s policy, supportive of the circle’s aim, and within the circle’s domain. Any policy decision in sociocracy has to be made by consent in the circle that has this decision in its domain. Sociocracy aims to balance the individual’s freedom with the supervision of and equality within the circle.
Another example: deciding that a circle meets every Tuesday and has a secretary who sends meeting reminders are both policy decisions. The act of “now I will send out a meeting reminder” and the act of sending it, that’s operations and the impulse to do so at a certain time is an operational decision. If for some reason it is important that reminders be sent out at a particular time and that did not happen and has led to issues, there might be some need for policy. Every circle has policy and operational parts to it. This generic statement is true because just deciding to have a facilitator is policy, and even just sending out minutes is operation – from the shop floor to the board level, there is policy and there are operations.