gemini://gemini.circumlunar.space/users/emptyhallway/halfbaking/rubric-resource-redistribution.gmi

Here are some scattered thoughts about how to approach resource redistribution. I'm not an economist, and this isn't a framework for any public policy. It's just some thoughts that I wanted to write down.

The goal of this rubric is to help define a lower bound and upper bound of individuals in a population who have "unfair" amounts of a resource—either too much or not enough—relative to the rest of the population. I imagined using it in the context of income inequality, but I've tried to write about it in terms of general abstract "resources", because I like to do that sort of thing.

Step 4 deserves a little more detail. The maximum boundary point is defined relative to the minimum boundary point and the median point. Take the ratio of the median to minimum and apply that ratio to the median. For example, if the minimum is 200 and the median is 500, then the ratio is 5/2. Apply this to 500 and the result is 1,250, which becomes the maximum boundary point.

The point of all this is just to get a rough rule of thumb for what a sensible range of acceptable values might look like.

In the following sections, I'll describe what I mean in a little more detail. Most of the steps are self-explanatory, and I only include them for clarity.

For example, you might want to consider the annual salaries for all employees within a company. Or your might want to investigate the total wealth for households within the United States.

Decide what a reasonable cutoff is at the lower end of your resource range. For example, if you are looking at employee salaries, maybe you want your bottom cutoff to be the livable wage for a single person for your state or region. Or if you are looking at annual per capita income within a country, you might use the poverty level as the minimum boundary.

In principle, this step should probably rely on some source outside your organization, and the decision should, in theory, be made without considering the actual data points for the population defined in step 1. That is, you should make a sincere effort to decide what a reasonable cutoff is without checking how that will affect the results of this exercise. In reality, I understand that this level of honesty and separation of concerns is probably impractical.

I realize that the stated intent of this rubric is to identify the range that defines "fair" and "unfair" quantities, so having one step of the rubric be "just pick one of those boundaries" does undermine the exercise a bit. I think this reflects a failure to properly present the idea more than anything else.

Do the research and the math to find out the median value for your chosen resource in your chosen population.

If your median is lower than the cutoff defined in step 2, then you have a problem. Your resource distribution is too unfair to be represented by this rubric. Hopefully that won't happen.

This is the defining step of this rubric. Basically, you are just finding a maximum boundary that has the same proportion to the median as the median value has to the minimum boundary.

To repeat the example from earlier, if you choose 200 as your minimum boundary, and your median is 500, then your ratio is 5/2. Apply this to 500 and you get a maximum boundary of 1250.

Now that you have an upper bound and a lower bound, you can try to reduce resources for individuals who have more than the maximum and increase resources for individuals who have less than the minimum.

For example, if you are in charge of salaries at your company, your action might look something like this:

First, cut salaries on anyone earning more than the maximum so that they are at or below the maximum boundary.

Use the savings from this action to increase salaries on anyone earning less than the minimum until they are earning at least the minimum.

If there are still individuals below the minimum, I guess you need to keep cutting salaries at the top, but the details are up to you.

If there is excess savings (after cutting salaries above the maximum and raising salaries below the minimum), I guess increase some salaries wherever you like, while keeping everyone below the maximum.

After you've made some changes, repeat the process to find out what effect your changes have had. If everyone in the population is between the minimum and maximum boundaries, then congratulations! You have moved beyond the need for this rubric.

Of course, having a hard line where everything on one side is Okay and everything on the other side is Not Okay is often too simplistic. If you want, you can pick two minimums: a "worst case" and a "less worse case", for example. Calculate the maximum relative to each minimum, and you'll end up with a range within another range. When you take action in step 5, you can prioritize the individuals outside the worst case boundaries, or however you want to approach the problem.

In this variation, you use the same population for both cases, which means that you will calculate the same median. One range will be fully contained by the other.

Any individual will belong to multiple overlapping populations, and sometimes you might want to investigate these overlaps.

For example, an individual might live in a city or county, which is located in a state, which is located within the United States. Each of these geographic regions contains a different population, with different median incomes and different average livable incomes. It might be useful to understand where an individual exists on each of these scales.

So just do the same steps with each of the populations you're interested in and see what you find, I guess?

This is the kind of example that I had in mind when I started thinking about this rubric. What is a "reasonable" range of income for an individual or family? If I want a rule of thumb for redistributing my own resources, how much should I be giving away?

For my population, I'm starting with the United States. I want to look at per capita income.

For my minimum boundary, I'd like to use the livable wage. There is a Livable Wage Calculator maintained by Amy Glasmeier at MIT which looks very thorough.

Unfortunately, it doesn't seem to provide an average living wage for the entire country. If you read some of the articles and notes, it seems that there is a good reason not to report at this level, and that an average would probably be more misleading than helpful. Regardless, I'm just trying to get a rule of thumb, so I'm not really worried about accuracy right now.

In an article from May 2020, they do mention an average US living wage of $68,808 per year per adult for a family of four (two working adults, two children). That's a total of $137,616 for the family, which works out to $34,404 per person. I'm going to round this up to $36,000 ($3,000/month) and call it close enough.

I had trouble finding the median income measurement that I wanted. The US Census web site lists median income by household.

But I want a median income by individuals within the household. For example, if a household had an income of $100,000 and contained 4 people, then for my purpose, each of those four people would be treated as if they had an income of $25,000, regardless of who actually earned the income within the household. This kind of breakdown doesn't seem to exist (or at least I haven't been able to find it).

Anyway, I don't think I'm going to find the data I want, so I'm going to make some guesses for the sake of this exercise.

The US Census report lists a median income of $67,521 per household. Elsewhere on the US Census web site, I see that the average household size is 2.61, so I'm going to just estimate a per capita income of $25,870. I realize that this is a bad approach, but I don't have the data to do any better.

Sadly, we have already failed in our first attempt to follow the rubric. The median income is below our desired minimum target, which means that (according to this rubric), the population has serious inequality problems (accurate) and anyone earning above the minimum should be redistributing their resources aggressively (true).

For the sake of this exercise, I'm going to just pick a different minimum boundary. In the US Census report, I see that the weighted average poverty threshold is $13,171 for one person. Let's use that as our minimum boundary and see what we come up with.

Comparing our median per capita income of $25,870 to our minimum boundary of $13,171, we get a ratio of 1.964. Multiply our median of $25,870 by 1.964 and we get a maximum boundary of about $50,809.

So, according to this rubric, any household earning more than about $51,000 per capita is above our threshold, and should be redistributing any income above that amount. Note that this is a per captia number, combined for the entire household, so in a family of four (two adults and two children, for example), the total family maximum would be about $204,000.

As a quick follow-up, I ran some numbers on global income. According to some search results, the global poverty level is $1.90 per day (which I understand to be $693.50 per person per year) and the median annual income is $850 (per person?). Applying our rubric, that give us a maximum boundary of about $1,042 per person per year. In principle, anyone earning above that amount should be aggressively redistributing their wealth. This analysis doesn't take into account local cost of living, so it may not be a useful result. I suspect that a more rigorous approach would still give a shamefully low maximum, though.

What did I learn from this? Was it a useful rubric? I don't know. I like the simplicity of the relationships between a minimum, median and maximum. Collecting accurate data seems difficult, though. Probably that's a challenge with any analysis, though.

I imagine that I will keep this in my mind for a long time. Maybe one day it will prove useful.

A rubric for resource redistribution

The process in detail

1. Define a population and a resource to measure.

2. Choose a minimum boundary point.

3. Calculate the median point for the population.

4. Calculate a maximum boundary point.

5. Take action to redistribute resources.

Nesting rubrics

Overlapping rubrics

Example: Annual per capita income in the United States

Conculsion