2 ms·
Really depends on what you’re trying to use the number for. If you’re trying to study wealthy people, it makes sense to use stats that are counting people filte
by fshbbdssbbgdd 5y ago
Really depends on what you’re trying to use the number for. If you’re trying to study wealthy people, it makes sense to use stats that are counting people filtered by wealth. If you are trying to study wealth itself, it makes sense to use stats that are weighted by dollar.
- jxramos 5y agothat's what I figured, the people metric seems more pertinent to ideas of economic mobility across the population as a whole, we could think of a population percentile, like say jumping into the notorious 1%. But the other metric is interesting too, something like an age x wealth product, the longer the wealth was held the higher the score, and the more wealth held the higher the score too. One could imagine like an area under net worth curve and create a histogram with decades of time as buckets, say across a 300 year span maybe?