> The problem here is not including people in your “system”. Absolutely. All too common. And once you do include them each is not merely a new variable capable

> The problem here is not including people in your “system”. Absolutely. All too common. And once you do include them each is not merely a new variable capable of assuming wildly different values, but a whole system in itself capable of interacting with every other such system within your system. That's why reductionists like to try factoring them out as interchangeable cogs. Pretty much the entire edifice of modern industrial economics since Adam Smith and Henry Ford is built on that model simplification/efficiency.