That article cites the Williams Group, which appears to be some sort of consultancy with the tagline "we prepare heirs."

This isn't the first time I've heard the 70% number, and I've never been able to find a single scientific study supporting it. There's also never any clarification on what "rich" means, or what qualifies as "lose their wealth."

How about this by Gary Becker

https://en.wikipedia.org/wiki/Gary_Becker

Becker, Gary S & Tomes, Nigel, 1986. "Human Capital and the Rise and Fall of Families," Journal of Labor Economics, University of Chicago Press, vol. 4(3), pages 1-39, July.

"The authors examine a number of empirical studies for different countries. Regression to the mean in earnings in rich countries appears to be rapid. Almost all the earnings advantages or disadvantages of ancestors are wiped out in three generations."

> The authors examine a number of empirical studies for different countries. Regression to the mean in earnings in rich countries appears to be rapid. Almost all the earnings advantages or disadvantages of ancestors are wiped out in three generations.

That's deeply at odds with what I know about the world (think about the existence of the "middle class", and how middle class parents tend to have middle class children). That, and the fact that the article was written by Gary Becker (he of "all crime is a rational cost / benefit tradeoff" fame), and published at a time when economists did very little emperical work, make me more or less immediately dismiss the claim.

But if anyone could give me the gist of the argument they make, I'd love to be proven wrong.

My guess is that you can put the bar for "wealthy" as high as necessary to get any percentage you want. Illustrating by reductio ad absurdum, you can make the argument that 100% of "wealthy" parents have no wealthy children, if your definition of "wealthy" is only the single richest person in the world. Expand the definition incrementally, and you can probably get nearly any percentage statistic you want between 10% and 100%. So the fact is "not even wrong" without more context.

There is a massive difference between earnings and wealth.

you see how that's completely different though right?

As a counterpoint:

“The richest families in Florence in 1427 are still the richest families in Florence”

https://qz.com/694340/the-richest-families-in-florence-in-14...

And the catch was it's not the same family.

People sharing a last name today can't be traced back to specific family which existed in past.

I think that's an anecdote for someone coming up with a system that works and not deviating. Obviously most people are very bad at that.

Maybe if they were able to put superior genes in their offspring they'd be able to move that metric.

That implies that it has to do with genes in the first place. Perhaps there's an argument there. However, I would also consider that the size of an estate may grow, but not exponentially (at least not indefinitely). Yet, the size of a family tree does grow exponentially.

Genghis Khan would like a word with you.

How much is higher intelligence corelated to being able to manage wealth well?

You are assuming all is in the genes.

I assume that at least part of it is in the genes.