From an economics standpoint, local "optimum" can be easily evaluated by looking at the cost of labor. Labor is expensive in the US, significantly more than her

From an economics standpoint, local "optimum" can be easily evaluated by looking at the cost of labor. Labor is expensive in the US, significantly more than herbicides, and weeding is extremely labor intensive. However, that ignores the long term costs, which may be what you are alluding to. Farms don't operate at huge profit margins, so they cannot compete with each other if they started hiring significantly more labor. One primary feature of governance is passing on long-term expected costs (e.g. ecological collapse and unemployment) to citizens/businesses optimizing for short-term profitability. In this case, you would want the government to somehow identify each farm's contribution to the ecological cost, and figure out how to pass that cost along. One possibility would be a pollution style tax. If herbicides all of a sudden cost significantly more due to taxes, then other solutions (like labor) will become more common. Our society also benefits from an increase in the number of jobs, which may have also been what you meant by "optimum". Creating meaningless jobs (e.g. picking/weeding when you could use a machine for less) does not add any long term economic value itself, though it does increase stability. It would be more efficient to just tax the economic base, though, and use the money for something economically more productive than weeding.