Terry Pratchett – Boots theory of poverty

Cafe Society
42

As I take it from debating Gemini in Deep Research mode, the general consensus of sociologist and microeconomists is that the phenomenon is real and viewed by a sufficiently large segment of the group to corrupt Gemini’s output as evidence of the unfairness of the economy.

However, when I asked it to model the lives of four people:

  1. An average person living in poverty, working a dead end job, in an average American location.
  2. A perfectly rational actor with the above limitations but who does things like splitting costs with roommates, avoids drugs and alcohol, doesn’t procreate prior to being financially stable, and maintains a healthy diet centered on bulk purchases of rice and beans.
  3. The same as #2 but forced to buy a car as a necessity of occupation.
  4. An average person (like #1) but who makes the one change of living with a roommate and splitting costs.

Gemini found that all three of the latter were able to afford good boots and work their way out of poverty, while still contending with health and car emergencies, relative to the risk profile of their lifestyle choices, and periodic and unexpected loss of job due to gig economies and the like.

All I did was force Gemini to provide a mathematical analysis and ignore outcomes that didn’t match what the math said.

So I’d take it that the phenomenon is real, and that most people will choose that course, but that’s a bad choice not an issue of there being too little income to support a different outcome.

But, obviously, I had to browbeat the AI to get it to ignore a majority of science in favor of math.

43

I liked your post but have issue with something in no way limited to you:

“Does not procreate until financially stable”

Leaving aside the case where a person or couple procreates BEFORE a catastrophe that catapults them into poverty, there are financial costs to NOT procreating that I am not at all sure is calculated in the above. Even if a poor person/couple uses birth control no method is perfect and a child can still result. At which point there is the cost of abortion if that is even possible or, if that is not possible, there is now a child to support. But said person/couple has made a “poor decision” even though they did everything “right” but had a undesired outcome. Even more so - the cost of birth control is higher for women than men, and if the default “average poor person” is considered to be male it, again, might not even be considered.

Some conservatives might argue that single poor people should simply not have sex (in practice that doesn’t seem to work so well). Well, what about poor married couples? Shouldn’t they be allowed intimacy? And even if a poor, single woman remains abstinent there remains the possibility of sexual assault and a potential pregnancy and child happening entirely against her will.

OK, getting off my soapbox now. You may now return to your regularly scheduled discussion.

44

Assume a spherical cow? And “ignore a majority of science in favor of math”? Then it is “a bad choice not an issue of there being too little income to support a different outcome …” and not “evidence of the unfairness of the economy”?

Of all that I will focus on one: “bad” and “unfair” are not a mathematical values. “Perfectly rational” even isn’t.

So amazingly you can browbeat an AI to give you an answer you want to hear. Huzzah.

Not an economist but I never understood the point of The Boots Theory to be anything other than an effective illustration of a factor that significantly stacks the deck against climbing out of poverty, not one that claims that it is impossible for a spherical cow to do so.

Your illustration with weight loss is cogent here. When “the math” conflicts with what is observed in reality, the science looks to why. The why in this case is that the assumptions of the physiology are incorrect. When and if the assumptions reflect actual dynamic adaptive physiology, how the body adapts to calorie reductions and weight loss over time (without ongoing exogenous hormonal medication inputs), the math is correct.

45

Eternal socks ! I can’t believe I’m the first to follow up on this by asking “how do I get some?”. Who makes those and what are they called ?

46

Am I right that the top quote is AI, and the bottom is Sage_Rat?

The AI had it right. It’s trivially correct and functionally useless to say that a “perfectly rational actor” will succeed in clawing out of poverty if they live like a hermit for an undetermined length of time.

It’s not a “bad choice”, it’s a real choice, made by real people who have normal human desires to have the things that make life worth living.

47

Even then the math is utterly suspect.

Posit someone poor who’s doing absolutely perfect play on their expense side. And given their income, they’re at breakeven. Now something changes. 10% CPI inflation w 0% wage increase. Trips over a crack in the sidewalk and turns an ankle and is out of work for 2 weeks. Catches the flu or COVID.

Suddenly the math doesn’t work.

There is no “mathematical proof” that income can always exceed expenses. It beggars belief.

48

And given that we are not dealing with only one person but with millions it’s absolutely guaranteed that it will happen to many of them.
And some will be lucky and everything will go their way, and they will be absolutely convinced that if they could do it it means that the others weren’t applying themselves and are just whiny losers.
Survivorship bias is a curse upon mankind.

49

It’s more of a cashflow problem than a total income problem. Meaning that while the poor can afford to spend more on doing laundry at the laundromat than the $600 it would cost for a new washing machine, they don’t have the cash flow that would allow them to actually save for and/or buy that washing machine and a jug of detergent.

I wouldn’t call it a “poor tax”, in that it’s not anything punitive or specific to being poor. Any more than middle class people not seeing corporate/government scale price advantages because we don’t buy stuff in quantities of thousands at a time.

I know it’s a semantic distinction, but it’s more the idea that this is done specifically to be onerous to the poor that bothers me. It’s not- it’s more economics than anything else.

50

Even if you can afford a $600 washing machine, you’re still going to need the drain and water supply lines installed. Many live in apartments or houses where this isn’t already in place.

51

The perfectly rational person ends up with a surplus of roughly $76,200 per year, on average (50th percentile). The cost of an abortion is $1000 and a bus ride to the next state is probably less than $100.

Emergency medical procedures are (supposedly - who knows how much work the AI is actually performing beyond roughing in a guesstimate) considered and affect the range of outcome probabilities.

The least lucky (10th percentile) perfectly rational person ends up with a surplus of $58,300 after a decade.

52

No, but you can run the range of probabilities given known statistics of health hazards and costs attributed to various causes.

For the poor, the biggest hazards - beyond what a middle class person would have - are work hazards, work toxins, raised risk of diabetes due to poor diet, raised risk of accident due to alcohol or drug abuse, raised risk of asthma and COPD due to alcohol and drug use, raised probability of encountering a medical emergency due to dependents. The other basket are car breakdowns, etc.

The perfectly rational person suffers the workplace risks but skips the lifestyle ones. He encounters the car maintenance issues, in the case where he’s forced to get a car.

The average person, who lives with roommates still encounters all those lifestyle/car risks.

On his budget, living in a regular American location (Kansas City, Missouri - apparently), earning $22k a year, he pulls ahead even in the least lucky 10th percentile, ending up with a surplus of $8500 after a decade. The average case is a surplus of $28,100.

53

So, a person who makes perfectly rational decisions and lives a deeply frugal lifestyle can anticipate, after an entire decade of perfection and frugality, to have saved up enough money to buy a nice-ish used car, if they’re unlucky, they can buy a dumpy used car.

Of course, they never would do that because they would likely hold onto that money as a lifeline in case of emergency.

On the plus side, they won’t have to worry about paying for college, no kids!

54

I include that information so you can evaluate the result.

It calculated a series of budgets that disproved Vimes’ theory but then would say, “Which proves Vimes’ theory because it’s impossible to form a budget that allows a person to escape poverty given evidence from financial diaries from those in poverty.”

But then you go through the things that make it impossible - the high price of rent, living on your own, the high probability of needing to seek out medical care due to diabetes and dependents, etc. - and that’s all the stuff that we’d worked to recalculate the costs and probabilities of, depending on the choices that a person had made.

The AI wants to give the findings of the research and, in the research, everything is calling for government intervention on the basis that this is how poor people “are”. They are people who eat bad food, they are people who get pregnant, they are people who drink too much, etc. And, with that being true, there’s no way to budget for success.

But, is that base assumption of a lack of personal agency true? To be pessimistic, we might agree with the research and agree that these are, in essence, people who live without personal agency. But, personally, even if I take this view that these are people without agency, I think that a lot of the choices of a person are driven by culture. In our culture, we push people to live alone and be self-sufficient. We make multi-tenant residences between people who aren’t related illegal through local ordinances. Perhaps, at one time, there was enough available land to allow for such laws but today that’s not true.

In general, if we look at other countries, they have lower levels of poverty. And when we do look at the poor, they’re making different choices than our poor. They might have big families but they’re not suffering from diabetes. They’re gardening or buying food that is healthy, because the culture pushes people to buy and eat real, healthy food and the economy has adapted to that.

There are some choices that are distinctly “poverty centric” but there are others that are less universal, and which afflict American poor, specifically, and many of those are cultural.

But, as I told the AI this isn’t a question of what choices a person will make. Vimes’ theory is that the math don’t math. That has nothing to do with human foibles.

A pure look at the math says that there’s every opportunity there.

If Vimes’ theory was that the poor tend to live a life of trying to live unthinkingly under the influence of McDonald’s ads, visions of how cool you look smoking, and wanting to have passionate, impromptu sex at every turn…and subsequently can’t afford to buy good boots then, yes, that agrees with the research.

In my read, though, that’s not the theory as stated. That’s a different theory - possibly one that’s more useful, but we don’t evaluate theories on the merit of unstated assumptions. We evaluate them on the basis of what was written, and nothing further.

Maybe that other theory should exist - I’d posit that it does. But the people who write it aren’t willing to put it in those terms since it makes them sound classist and dismissive of poor people.

55

Replace “makes them sound” with “is” and you have it.

56

That one kinda cuts across class lines, though…

57

And so if we take it that people do have complete agency and choice over their activities, then…well, each and every one could budget their way out; even with the jobs and wages that are currently available.

58

In a significant number of cases, budgeting your way out means “you could spend a decade of your life scrimping and saving, eschewing romance, family, and virtually all luxuries, in order to budget your way to have just enough savings to weather one car accident.”

In the ‘average’ case it means “you could spend a decade of your life scrimping and saving, eschewing romance, family, and virtually all luxuries, in order to budget your way to have just enough savings to take one two week long vacation without setting you back to where you started.”

Yes, a person could do this immense amount of work for such a limited benefit. They would work a full time job, then go home and work AGAIN to budget and manage their lives with perfect rationality, and continue to do this for 20% of the time they have remaining on this Earth, just to get enough cash socked away to take a breath. Not enough to retire on, not close to enough to retire on. If they want to actually retire someday, they’d have to live like this for 90% of their time left on this planet, just to be allowed to skip the full time job part of their lifestyle, they’d still be living the frugal life until the end.

But you think their problem is McDonalds and looking cool.

59

Darn Tough.

60

“The poor deserve it” is nothing new and certainly not more persuasive coming from a machine.

61

You have really worked very hard from a conclusion to create a response that you interpret as supporting your hypothesis.

It simply is not a fair assumption for analysis, at every definition of “fair”, to assume a “perfectly rational actor.” Maybe there is someone out there who makes all decisions in what those models would assume are “perfectly rational” ways, but I doubt it. They are definitely rare among the “haves”, who often use wealth for to satisfy desire for social signaling without economic return, who inadequately save for retirement, who invest more frequently on the basis of fear (balanced between fear of missing out vs fear of loss) and more. It is of course less costly (in terms of fraction of income and effort) for them to have healthier lifestyles, even down to pollution and stress related to the neighborhood, but I do not see any superior personal agency among the haves.

And of course the poor have other sets of fears that result in decisions that deviate from “perfect rational actor” choices. And structural/environmental obstacles in contrast.

The field of behavioral economics investigates and attempts to mathematically model how different circumstances result in how the inputs and possible results are valued in decision making.

I honestly would be more interested in a discussion that begins with the reasonable assumption that no actors are perfectly ration and delineates the differing flavors of irrational behaviors and the divergent costs of them.

Can you give some specific examples? Most other countries that significantly hit those “different choices” IME also have stronger social safety nets, less wealth inequality, and less structural and institutional aspects to their various -isms, than we do.