These are arguably identical statements. I think Milton Friedman's "Theory" was descriptive, not prescriptive: ultimately shareholders can sway a corporation, a

These are arguably identical statements. I think Milton Friedman's "Theory" was descriptive, not prescriptive: ultimately shareholders can sway a corporation, and thus the corporations' "duty" is "to" them. If public opinion moves in a way that makes it so a corporation would become unpopular and unprofitable by harming customers and "stakeholders", then they are effectively "harming" their shareholders, who would be the ones that then say "we should take these concerns into consideration". This may seem perhaps a pedantic or semantic way of looking at it, but I do think it somewhat gets to the heart of many of Friedman's beliefs regarding incentive structures. By understanding that corporations are ultimately essentially accountable to their shareholders, you can develop a model where the shareholders goals are aligned with the customers'. Some would argue that this is always the case given enough time, others would argue that it isn't and you thus need additional laws to create the alignment of goals (if you go to jail for harming your customers, then your shareholders all of a sudden have the goal of not harming your customers). The main distinction I'm trying to make is that this is a description of where agency lies, vs. pretending corporations have some "decision" as to who they are responsible to. I promise you that if tomorrow they decided they're all actually accountable to me, the CEOs would get fired and that would reverse fairly quickly. The most interesting place this comes up IMO is in the emergent behavior of collective shareholders. Many people may believe they would tell Phillip Morris to behave a certain way, and yet, without knowing it, they actually indirectly tell them the opposite. If their retirement plan unbeknownst to them is invested in Phillip Morris, and Phillip Morris starts making decisions that are good for customers but lower profits, and thus random people get angry at their retirement fund managers and those managers pull money out of Phillip Morris and into some other venture that is more profitable, without knowing it they have sent a clear signal that "Phillip Morris is ACTUALLY accountable to shareholders, and in particular the money they make". This is the default case without actively doing any extra work. It is absolutely the case that if enough education is employed that people start choosing to tell their retirement funds to not invest in Phillip Morris because they are profitably harming customers, then you can have the more socially desirable effect of forcing them to behave in a way that doesn't harm customers. But again, the goal is to understand that "steady state" of this relationship.