# If a revolution killed the rich at COVID rates, how much $ would we get?

August 2, 2020 5:23 PM Subscribe

Brazil is approaching the 100K deaths milestone and it got me thinking: If a revolution killed the rich at the same rate, starting from the top, how much money would be expropriated? I got some data inside, can you help me come up with a spreadsheet formula?

According to the Credit Suisse 2019 global wealth report (pdf), the wealth pattern within Brazil (page 117) is:

I'm not sure if the values by range are actually useful for this, seems that just the number of adults, mean wealth and Gini index would be enough, but I'm having a hard time formatting this in my head.

This is not for work or anything (hi CIA, I'm not actually plotting a revolution!), just idle quarantine daydreaming. But if there are any nerd comrades out there, this could make a nifty Twitter bot or something, I guess.

According to the Credit Suisse 2019 global wealth report (pdf), the wealth pattern within Brazil (page 117) is:

Adults (thousands) 150,089 Mean wealth per adult USD 23,550 Median wealth per adult USD 5,031 Distribution of adults (%) by wealth range (USD) under 10,000 70.2 10,000 - 100,000 27.1 100,000 - 1 million 2.5 over 1 million 0.2 Total 100 Gini % 84.9I'm trying to come up with something in Google Sheets where you plug in these values for any country (the US and world values are on page 120) and input the number of dead, say 100 thousand dead, and it outputs how much money would be expropriated if a revolution killed the richest 100 thousand people in a country.

I'm not sure if the values by range are actually useful for this, seems that just the number of adults, mean wealth and Gini index would be enough, but I'm having a hard time formatting this in my head.

This is not for work or anything (hi CIA, I'm not actually plotting a revolution!), just idle quarantine daydreaming. But if there are any nerd comrades out there, this could make a nifty Twitter bot or something, I guess.

(Also just to clarify the GINI index doesn't give you what you need--it's an aggregate measure, when what you care about is the you'd need to see the curve, or underlying numbers. To calculate exactly you need the exact numbers for whatever your definition of "rich" is.)

posted by mark k at 6:25 PM on August 2, 2020 [1 favorite]

posted by mark k at 6:25 PM on August 2, 2020 [1 favorite]

Your problem is that unless you're looking at truly titanic death tolls you need much more granular data on wealth distribution within the very wealthiest.

For Brazil for instance, the smallest group you have is 0.2% of the adult population. For the 150m adults in Brazil that is 300k people, so until you get up to a number near 300k you don't know.

Also, that category is everyone above $1m but at the very least you need a mean wealth for people in that category. That's because wealth distribution follows a power law distribution all the way up.

0.2% of Brazil's adults have a net worth of $1m+ but the vast majority of them will have only slightly more and a small subset which have much, much more and have most of the group's wealth.

The data set that you really need is a set of much smaller population groups AND the mean wealth of each group.

So if you knew that the mean net worth of the top 0.01% (about 15k people) of adults in Brazil was $10m, then when you reached 15k dead you would have $150Bn. However that does not mean that if you had 5k dead you would have $50bn because the richest 5k have way more money per person than the next 10k.

I do not know if this data exists in sufficient granularity for any country in the world.

posted by atrazine at 4:00 AM on August 3, 2020 [2 favorites]

For Brazil for instance, the smallest group you have is 0.2% of the adult population. For the 150m adults in Brazil that is 300k people, so until you get up to a number near 300k you don't know.

Also, that category is everyone above $1m but at the very least you need a mean wealth for people in that category. That's because wealth distribution follows a power law distribution all the way up.

0.2% of Brazil's adults have a net worth of $1m+ but the vast majority of them will have only slightly more and a small subset which have much, much more and have most of the group's wealth.

The data set that you really need is a set of much smaller population groups AND the mean wealth of each group.

So if you knew that the mean net worth of the top 0.01% (about 15k people) of adults in Brazil was $10m, then when you reached 15k dead you would have $150Bn. However that does not mean that if you had 5k dead you would have $50bn because the richest 5k have way more money per person than the next 10k.

I do not know if this data exists in sufficient granularity for any country in the world.

posted by atrazine at 4:00 AM on August 3, 2020 [2 favorites]

Unless I'm missing something about a massive estate-tax in Brazil, wouldn't the answer be zero? The wealth just moves down the rung to the next generation in the same family, no?

posted by tiamat at 8:50 AM on August 3, 2020 [3 favorites]

posted by tiamat at 8:50 AM on August 3, 2020 [3 favorites]

It is true that I didn't see numbers available to calculate for only "millionaires and up" (as opposed to the top 1%) but this could be reasonably estimated by just assuming the 0.8% of the population in the top 1%, but not the top 0.2%, have $500,000 each on average. You could do high and low bounds to by changing that estimate.

Note that the numbers will be of necessity

Unless by "dying at the same rate" you meant the same absolute numbers (ie, 100,000 but all from the set of millionaires.)

The hypothetical revolution would, I assume, be confiscating the estates of the wealthy.

posted by mark k at 1:46 PM on August 3, 2020 [1 favorite]

Note that the numbers will be of necessity

*smaller*for the millionaires and up club.Unless by "dying at the same rate" you meant the same absolute numbers (ie, 100,000 but all from the set of millionaires.)

*Unless I'm missing something about a massive estate-tax in Brazil, wouldn't the answer be zero? The wealth just moves down the rung to the next generation in the same family, no?*The hypothetical revolution would, I assume, be confiscating the estates of the wealthy.

posted by mark k at 1:46 PM on August 3, 2020 [1 favorite]

*> Unless by "dying at the same rate" you meant the same absolute numbers (ie, 100,000 but all from the set of millionaires.)*

What I meant was, the revolution kills people at the same

*speed*as covid, and it works its way down the wealth concentration pyramid. The first person to die would be the richest Brazilian, next would be the second richest, and so on. So the 100K dead would be the 100K richest Brazilians.

The formula doesn't have to be 100% accurate, a rough ballpark like you're saying would be nice!

*> The hypothetical revolution would, I assume, be confiscating the estates of the wealthy.*

That's what revolutions are for! :-)

posted by Tom-B at 9:20 PM on August 3, 2020

Ah, I see.

100,000 people is a little under .1% of the adult population, so everyone would have something over 1 million dollars in net worth. The range of aggregate wealth is plausibly between 200 billion and one trillion dollars, or at the high end $6000 per capita (adult). (My other posts should make it clear how I'm coming up with this.)

Your attempt to use the GINI makes more sense to me now. Still, I can't offer a generic formula unless you know the amount of money those people have, in which case the formula is trivial.

(The pedant in me insists this is better described as "dying in the same numbers" rather than "at the same rate" but it's really not important.)

posted by mark k at 9:51 PM on August 3, 2020 [1 favorite]

100,000 people is a little under .1% of the adult population, so everyone would have something over 1 million dollars in net worth. The range of aggregate wealth is plausibly between 200 billion and one trillion dollars, or at the high end $6000 per capita (adult). (My other posts should make it clear how I'm coming up with this.)

Your attempt to use the GINI makes more sense to me now. Still, I can't offer a generic formula unless you know the amount of money those people have, in which case the formula is trivial.

(The pedant in me insists this is better described as "dying in the same numbers" rather than "at the same rate" but it's really not important.)

posted by mark k at 9:51 PM on August 3, 2020 [1 favorite]

*> (The pedant in me insists this is better described as "dying in the same numbers" rather than "at the same rate" but it's really not important.)*

Sorry, English is not my first language!

posted by Tom-B at 2:29 PM on August 7, 2020

« Older Will i get used to reversing with backup camera? | How to be Election Judge in a swing state? Newer »

You are not logged in, either login or create an account to post comments

The money freed up then is merely A * W, where A the portion of people you are taking wealth from, and W is the aggregate wealth.

Since I'm curious now, from your link, in Brazil the top 1% owns half the wealth (I'm eyeballing a bar chart they have, the exact number could be somewhere else.) Using the mean / total numbers you have here,

W = 150,000,000 people * 23,550 dollars per people * 0.5 = 1.8 trillion dollars.

A = 100,000 / 150,000,000 = 1/1500 = 0.07%

The total wealth owned by that fraction of the top 1% is (0.07 / 100) * 1.8 trillion = 1.3 billion dollars. Or a bit under $10 per person.

Did this pretty quickly and I wouldn't be

shockedif I missed a decimal point somewhere, I would appreciate corrections!But as a sanity check, the maximum you could extract from the top 1% would be about $11,000 per capita (if you took all their wealth) for scenarios where 99% of them hang on to it you're in the $100 per capita range.

posted by mark k at 6:22 PM on August 2, 2020