It is a simplified toy economy. The whole point is that only bananas are produced and demanded. Given that assumption that they both spend all their time growing bananas and they wish to eat as many bananas as possible, then what I describe holds true.

The point here is to provide a simple narrowly defined example to explain how a salary increase for one party can cause lower purchasing power for another group.

I described salaries as zero sum games under specific conditions:

An increase in salary for one group doesn't case more optimal allocation of resources or higher efficiency in parts of the economy.

For instance the economy could have a misallocation: Too many people spend time making things in low demand while too few are making things in high demand. A salary shift can help cause workers to move over to the production which is in higher demand.

The other case in which a salary increase is warranted is if a group of workers has increased efficiency of production. If you produce twice as much (assumming wages are the only expense) then twice the salary is theoretically possible without negatively affecting other people in the economy.

Yet, as I pointed out, not all salary increase is related to any increase in production or more optimal resource allocation. I am simply trying to point out that in the cases where salary increases isn't related to more optimal resource allocation or higher productivity it will function as a zero-sum game.

A lot of people, such as yourself, try eagerly to obscure this fact. And a lot of people are very effective at obsuring this fact because they throw in numerous variables that hide the fact that salaries are very often a zero sum game.

OECD has research showing that inequality lowers economic output, hence when salary differences become larger they are not actually contributing to higher producitity and efficiency. No, they are actually working as a zero-sum game