In the long term? A good hard look at how we determine management capability and effectiveness, and an even harder look at what work we actually expect of each position.
Right now, I see that the most important quality for selection to management is the ability to project confidence, whether or not it's warranted, and the ability to make connections above and around one's current manager. It shouldn't need to be said that those are suboptimal criteria. We'd do better selecting for people who give honest assessments (i.e., "This is unlikely to succeed because of XYZ"), have the ability to say "I don't know, but I will find out" (and then actually do it), and who have the good opinion of peers and their direct reports (if any).
Because once someone is promoted? Nothing. No support for learning to do the people-management aspect of the job, minimal support for the administrivia parts. We pick the wrong people, and we don't even teach them how to do the job.
Potentially relatedly, no one seems clear on what that job actually is. I literally have no idea what my boss does other than attend useless meetings (which are then 'summarized' innsuch a way as to leave no time for context or questions), issue contradictory demands (and occasional self-congratulatory statements), and generally create an atmosphere of paranoia over who will be the next sacrifice to be presented as someone who Boss was able to help "turn their career around". This boss has, admittedly, done a great job bringing the team together- through what amounts to collective disgust.