I've noticed a few times how surprisingly easy it is to be in the upper echelon of some narrow area with a relatively small amount of expenditure (for an upper middle class American professional). This is easy to see in various entertainment hobbies- an American professional adult who puts, say, 10% of his salary into Legos will have a massive collection by the standards of most people who own Legos. Similarly, putting 10% of a professional's salary into buying gadgets means that you would be buying a new one or two every month.

I recently came across an article on political donations and saw the same effect- to be in the top .01% of American political donors, it only takes about $11k an election cycle (more in presidential years, less in legislative only years). Again, at 10% of income, that only takes an income of ~$55k a year (since the cycles occur every two years), which is comparable to the median American salary (and lower than the starting salaries for most of my friends who graduated with STEM bachelor's degrees).

It's not clear to me what percentage of people do this. It's the sort of thing that you could only do for a few narrow niches, since buying a ton of Legos impedes your ability to buy a bunch of gadgets, and it seems like most people go for broad niches instead of narrow niches. If you spend 10% of your income on clothes, say, then if most people spend 10% of their income on clothes you need to be in the top 1% of income-earners to be in the top 1% of clothes-buyers.

I know a handful of people in the LW sphere give a startlingly high percentage of their income to MIRI and are near the top of MIRI supporters. They probably also end up in the top percentile of charitable givers, but I don't have numbers on hand for that.

I'm curious if this is a worthwhile pattern to emulate. I currently do this for art collection in a narrow subfield, and noticed the benefits of being at the top percentage of expenditure mostly by accident, but don't have a good sense of how those benefits compare to marginal value comparisons between different potential hobbies. (Actually, now that I think about this, this might just be a special case of the general "specialization pays off" heuristic, where it may be better to have one extreme hobby than dabble in twenty things, but this may not be obvious when moving from twenty hobbies to nineteen hobbies.)