Most unintended consequences trace to who got to decide, and to whether anyone was honest enough to say the decision had gone wrong.
A company I worked with once put in a simple rule: any refund over a hundred dollars needed a manager's sign-off. The intent was obvious and easy to defend. Someone had been burned by a bad call, and this made sure it could not happen twice. Eighteen months on, that same rule was steadily costing them their best customers, who now waited two days for a manager to bless a refund a frontline rep could see in ten seconds was legitimate. The rule worked exactly as designed. What it got wrong was who it took the decision away from.
We tend to file an outcome like that under bad luck, or an unforeseeable side effect, or a good rule that aged badly. I have come to think most unintended consequences are none of those. They are a decision-rights problem wearing a disguise. Somebody put a decision in the wrong hands, and the wrong hands did exactly what you would expect them to.
The cleanest framing I know is a rule of thumb worth memorizing. Codify the generalizable, and reserve genuine discretion for the accountable owner. Anything predictable enough to write down as a standard, write down and push out wide, because a rule scales and a judgment call does not. But the moment a situation needs real judgment, the decision belongs to the person who will answer for how it turns out, and to no one else. Most bad outcomes come from breaking that in one of two directions. You codify something that needed discretion, like the refund rule, and build a machine that cannot tell a good exception from a bad one. Or you hand real discretion to someone who carries none of the consequences, and you get choices made by people with nothing at stake in whether they were right.
You can get that allocation perfect on the org chart and still walk straight into the outcome you were trying to avoid, because reserving a decision for an accountable owner only pays off if that owner can do the hardest thing the arrangement asks of them. They have to find out they were wrong while there is still time to fix it. And that is less a structural problem than a human one.
Valorie Burton has a rule for exactly this in Rules of Resilience, one I keep handing to leaders. She calls it do not pretend, do not defend. Do not pretend a decision is working when it is not, and do not defend a call once the evidence has turned against it. Both are the natural reflexes of a person who owns an outcome and has staked their competence on it, which is precisely the person we just handed the discretion to. Pretending buries the consequence for another quarter. Defending kills the feedback that would have caught it. An owner who cannot manage Burton's two moves turns the discretion you granted from an asset into the most expensive kind of liability, the kind no one is allowed to name.
Which is where the people around that owner decide everything. Kim Scott's Radical Candor rests on two things done at the same time, caring personally and challenging directly. Drop either one and you land in a failure mode with a name. Care without challenge is what Scott calls ruinous empathy, the warm and conflict-averse silence of people who like their boss too much to tell him the refund policy is bleeding customers. It is the most common failure in otherwise decent organizations, and it is the one that lets a reserved decision rot in place. The accountable owner cannot see the consequence from where they sit. The people who can see it have decided, kindly, to say nothing. I wrote this week about executives backing away from empathy over an objection nobody in the room had actually raised. This is that same machine from the other side. The challenge that would have saved the decision was in the building the whole time. It simply was not sayable.
Ruinous empathy is the kind version of that failure. There is a colder one. Sometimes a person curates what the decision-maker gets to know on purpose, handing over a narrow and carefully chosen slice of the truth so that a choice made on partial information comes out where they wanted it, and they never have to be seen steering it. Scott has a name for this corner of her map too, manipulative insincerity, and in a lot of boardrooms it has quit being a failure and become a technique. We have a whole genre of prestige television built on the move, executives lying to one another for position, and the reason those stories land is that we recognize it. What the shows almost always get right is the ending, where the person who won by controlling the information finds out it cost more than it bought. This was never only a boardroom habit, or even a work one. Any time one person decides what another is allowed to know and then lets them act on it, they have taken a decision right that was never handed to them, and the consequences come back around to everyone, the curator included.
So the three have to travel together, and almost nowhere do we teach them as one subject. Decision rights put the choice in the hands of the person who will answer for it. Burton's rule keeps that person honest with the result instead of invested in defending it. Scott's candor gets them the truth, from the people who can see what they cannot, early enough to matter. Pull any one of the three and the other two cannot hold. The decision drifts to someone with nothing at stake, or the owner protects the mistake, or the room protects the owner, and the sound choice becomes the surprise everyone later agrees no one could have seen. It gets filed under unforeseeable at the postmortem, which is usually the last untruth in a fairly long chain of them.
The consequence was almost never unforeseeable. Somebody in the building foresaw it. The question worth asking is why it stayed unspeakable, and that answer is rarely about the org chart. So about your own last expensive surprise, three questions. Was the decision in the right hands. Could the person holding it admit out loud that it was going sideways. And had you built somewhere that a person who saw it coming could say so in time. If any of the three was missing, it was not luck. It was a design, and a design is a thing you can change.
