Is it better to lay off 10 more people and save the money or is it better to keep more people around to deal with the uncertainty of the next 6 months? How do you want to protect the team? With extra people or with extra money?

Depends on how many billions you spent on stock buybacks that year..

Two direct reports aren't getting along. Their work styles are both valuable but clash with each other. You need to ask one of them to adjust to the other's expectations.

There's plenty of situations where you can do things right, but it will be expensive, slow or otherwise inconvenient, or you can screw someone over (that someone may be the taxpayers, or local community whose environment you pollute), and make the problem go away.

Tradeoffs become multi-arm as you go up the chain. For instance, complying with the letter of the law at great engineering, operational and product UX expense, vs cutting corners to maintain a smooth UX, low operational overhead, and straightforward maintainable systems. In todays world no one is able to be experts in all those considerations. Leadership must make apples-to-oranges tradeoffs all day long, and they must rely on communication and maturity to navigate the cascading inputs of these choices. What is possible to execute on cross-functionally, in what timeline, and what are the externalities of different tradeoffs? These are fiendishly difficult to navigate, so generally financial metrics will serve as the tiebreaker, and even then it's based on a lot of guesswork and proposals made up by self-interested internal parties over-representing an uncertain outcome. It's no surprise that the really big ethical concerns get lost in the shuffle—there's simply no system of checks and balances on ethics with anywhere near the effectiveness of that imparted by the incentives corporate capitalism bolstered by individual self-interest (note I avoid the word greed here since the incentives aare chained all the way down to individuals which no one would consider well-off enough to be labeled "greedy").

> to individuals which no one would consider well-off enough to be labeled "greedy")

Greed doesn't have anything to do with how much money you have, it's about how much do you want "more" and what are you willing to do to get it. You can be content with nothing and greedy with nothing; content as a billionaire or greedy as a billionaire.

Sure, fair point, but where you draw the line isn’t really material to my argument. In fact no one needs to be particularly greedy in order for large corporations to do very bad things. It just happens organically through local incentives and diffusion of responsibility.

not business but I'll give one from my own work in academia.

Do you change a requirement for a BS degree to only require 6 instead of 9 credits of research?

Upside - students might graduate faster (this is one of many constraints on graduation) which saves them money and time.

downsides - faculty may have less students to help with research (which helps faculty succeed via research productivity)

complicating - students might do the research anyways, they can still get credit it just doesn't officially count for anything.

context: the department is responsible to both...the actual impact of both the upside and downside are functionally impossible to predict because they are part of a broader complex partially social system that will anneal itself in ways you don't expect, metrics of success for students and faculty are important to everyone, albeit through different arms of an org chart which results in them each advocating for their metric at the expense of all others, because their success is partially evaluated on that metric.

solution (this is sarcasm): remove a math course as 'required' that most students get transfer credit for and have to have as a pre-requisite for other classes. Only effect is on the 2% of students who actually have to take the course.

result: everyone sees change, which they want, without perceiving a risk of harmful side effects, and with no actual change to the status quo.