Estate & Legacy Planning
Andrew Carnegie sold his steel company in 1901 and then spent eighteen years trying to get rid of the proceeds on purpose. He funded 2,509 libraries. He was not being sentimental. He had written the argument twelve years earlier and he was executing his own instructions.
The essay was called Wealth. The line everyone remembers is the hardest one in it: the man who dies thus rich dies disgraced.
Read that as an operator, not a moralist. Carnegie was not making a case against money. He made more of it than almost anyone alive. He was making a case about timing.
Capital you deploy while you are alive gets your judgment attached to it. Capital you leave behind gets someone else’s.
An estate is a set of instructions executed by people who cannot ask you a follow-up question. Every ambiguity you leave in it becomes a guess, and the guess gets made by someone under emotional load, on a deadline, usually in a room with a lawyer billing hourly.
You can write the tightest document in the world and it still only moves assets. It does not move the ability to run them.
Three things a will cannot transfer
Judgment. Nobody inherits pattern recognition. It is built by making calls under real stakes and living with the outcome. If the next operator has never made a consequential decision while you were still in the building, they will make their first one at the worst possible time.
Relationships. A contact list is not a relationship. The banker, the contractor, the broker, the person who actually returns your call: those transfer by introduction and repetition, not by paperwork. That handoff takes years and has to start while you are healthy.
The operating manual. How the thing actually runs, including the parts that are in your head and nowhere else. Most founders assume this is documented. It is documented right up until the person reading it hits the first exception, which is where all the real knowledge lives.
A trust tells people what they own. It does not tell them what to do on the Tuesday everything breaks.
The library model
The interesting part of Carnegie’s giving was not the amount, it was the structure. He would fund the building. The town had to supply the site and commit public money to run it, year after year.
He was not writing cheques. He was requiring counterparties to take on an obligation, and then funding the part they could not do alone. Every library came with a permanent local liability attached to it.
That distinction holds up everywhere. Giving that creates an obligation compounds. Giving that creates a dependency decays, and usually takes the relationship with it. This is as true inside a family as it is between an industrialist and a town council.
Which is the practical translation for anyone thinking about handing a business to their kids: hand them a decision, not a distribution. Give them a real call to make, with real consequences, while you are still around to watch how they make it. That is the only version of succession planning that tests anything.
The legal structures matter and they are worth paying a real advisor to get right, because they vary by jurisdiction and by what you hold. But the structure is the container. It has never once determined whether the person opening it knew what to do next.
Carnegie understood he was better at directing his own capital than his executors would be, so he made sure there was as little left for them to direct as possible. That is not generosity. That is an operator refusing to delegate the last decision.
Deploy it while your judgment is still attached to it. What you leave behind is not a legacy, it is a set of instructions for people who cannot ask you what you meant.