Inequality and debt: the soft bigotry of low expectations

"The poor don't have enough income to save, and can't help going into debt to the rich. Debt is caused by inequality".

That statement is wrong on many levels. It's wrong theoretically. It's wrong empirically. But most of all, it's wrong because it might make inequality worse. It's the soft bigotry of low expectations. Providence is especially important to the poor. Saying that the poor can't help but be improvident is the last thing they need to hear.

(This post will piss some people off. "Nick's blaming the poor; of course they can't help either their poverty or their lack of saving!")

If we want a policy that will promote equality of income we should do it because it's a good policy in its own right; not because of some spurious argument about reducing debt. And the sort of policy that would help attain that objective in the long run is one which encourages a higher savings rate by the poor than by the rich. For example: a policy that increased the after-tax rate of return on savings for the poor, relative to the rich. Canada's Tax Free Savings Account (everybody is allowed to save $5,000 per year tax-free) might be such a policy.

Let's start with theory.

A lot of debt is used to finance investment. In physical capital, houses, education, etc. But let's set that aside, because it's not what I am talking about here. I am talking about the distribution across the population of consumption and savings, not investment. Whether those savings are lent to others or used to finance one's own investment is a separate question.

Economic theory of the consumption/saving decision is based on intertemporal utility maximisation. Milton Friedman's Permanent Income Hypothesis is a simplified version of this. If you consume less today you can consume more tomorrow. You compare the marginal utility of present consumption to the marginal utility of future consumption. You compare your personal subjective rate of time preference (your degree of patience or subjective discount rate) to the market rate of interest at which you can borrow or lend.

This theory says that people who have a low rate of time preference (who are more patient) will tend to save, while those who have a high rate of time preference will dissave (have negative saving).

This theory says that people who have temporarily high income will tend to save, while those who have temporarily low income will dissave. Because if they didn't, and just consumed their current income every period, their marginal utility of consumption in good times would be lower than their marginal utility of consumption in bad times. We save and dissave to smooth consumption over time.

This theory does not say that people who have permanently high income will tend to save, while those who have permanently low income will tend to dissave. Consumption depends on permanent income. But the proportion of your permanent income you consume (and hence the proportion you save) need have no relation with the level of your permanent income. Depending on the particular utility function, the relation could go either way. If the intertemporal preferences simply scale up (which is the standard assumption), then consumption and savings scale up with permanent income, and the savings rate (the percentage of income saved) is independent of permanent income.

Start in one equilibrium. Double both income and consumption both today and tomorrow. If the marginal utility of consumption falls by the same percentage both today and tomorrow when consumption doubles (which it must for example if today's and tomorrow's consumption are equal), then you are still in equilibrium. If the ratio of today's to tomorrow's marginal utility of consumption stays the same, then everything scales, and you save the same fraction of your income regardless of whether income is permanently high or permanently low.

Sure, the poor really need to consume today. But they really need to consume tomorrow too. So their need to save is as strong as the rich.

1. Inequality in permanent income will not affect the level of debt.

2. Inequality of transitory income will affect the level of debt. Those having temporarily low income today will borrow from those with temporarily high income today. And will repay the loans in future when the transitory shock reverses. (And it must reverse, because it wouldn't be transitory if it didn't.)

3. Inequality of rates of time preference will affect the level of debt. Those with high rates of time preference will borrow from those with low rates of time preference, until they are credit constrained and cannot borrow more.

That third inequality is what is driving debt in the Gauti Eggertsson and Paul Krugman paper (pdf). Debt is caused by inequality of patience, not inequality of income.

The second and third inequalities are what are driving debt in Matteo Iacoviello's paper (pdf) (HT goldilocksisableachblond). Despite what it says in the abstract of that paper, it is not income inequality driving debt in his model. Debt is not caused by some people being permanently rich and others being permanently poor. Debt is caused by fluctuations of income over time, where one person's fluctuations are imperfectly correlated with others' fluctuations.

A rich or poor person, having a bad year, will borrow from a poor or rich person, having a good year. An impatient person will borrow from a patient person.

Being poor means having low permanent income. Being rich means having high permanent income. Debt is not caused by inequality in permanent income. Debt is not caused by inequality between rich and poor.

The only way in which a change in permanent income can cause a change in savings is if it causes a change in the pattern of transitory income. For example, as people have become richer over time, the wealth effect on labour supply means they spend a smaller fraction of their lifetimes working, and retire well before dying. Retirement is a time of negative transitory income (current income is below permanent income). So people save for their retirement. And the longer their retirement, the greater the percentage of their income they will save while working.

Here's the nightmare scenario of standard theory (thanks commenter Greg). Take a very simple model with infinitely-lived agents and no investment. Assume all agents have an exogenous flow of income that is constant over time. Assume all agents are identical, except some have low and some have high discount rates. The equilibrium rate of interest will be between the high and low discount rates. The impatient borrow from the patient, for consumption loans. The impatient will initially consume more than their income, and the patient will initially consume less than their income. But over time, the consumption of the impatient will fall and asymptote towards zero (or starvation-level). And the consumption of the patient will rise and asymptote towards their consuming all the labour income of both sets of agents. Debt, and inequality of wealth, rise inexorably over time.

Transfers from rich to poor (i.e. from patient savers to impatient dissavers) will not solve this problem. All they do is reduce the after-tax rate of return to both sets of agents, by an equal amount. The implied tax rate on saving increases the rate of interest, and the patient still save more than the impatient, net of taxes and transfers. It only works if people don't see it coming. But of course they will.

A policy that would help solve this problem is one that gives the poor a greater after-tax return on saving than the rich. That works in two ways. First, it gives those with low savings a greater incentive to save than those with high savings. Second, it means that a given level of saving will accumulate faster if the stock of savings is low than if it is high. One way to implement such a policy would be to put a limit on the amount of saving that could accumulate tax-free. I think that Canada's Tax Free Savings Accounts might do this. They let every person save $5,000 per year without paying tax on the interest earned.

That's what standard theory says.

Does that standard theory make sense? Well, OK, if you are in danger of starvation today it doesn't. Because if you don't consume today the marginal utility of your future consumption is irrelevant. But our ancestors, whose saving and investment created the income we have today, and who were much poorer than we are today, would laugh at us if we gave this excuse for not saving. (Mine, I think, would have been more disgusted than amused).

Does it make sense empirically? Does it fit the facts? I don't know for sure. But the opposite theory, which says that the rich save and the poor dissave, is contradicted by some widely known facts. And big facts at that.

Our grandparents, and great grandparents, despite being massively poorer than us, saved. That's how we have the capital and technology that we have today, that makes us so much richer than them. Real interest rates have not been massively declining, and savings rates have not been massively rising, over the last 200 years as our incomes have massively risen.

China is still much poorer than the rich countries, and yet has a much higher saving rate.

And it's no good just looking at a cross-section of people today and seeing if the savings rate rises with income. Current income is the sum of permanent income plus transitory income. And theory says that transitory income will be saved.

Moreover, current income will be correlated with patience. This is not to say that the poor are impatient. But it is to say that the impatient will tend to become poor. Because the impatient will save and invest less.

The casual assertion that the poor can't afford to save is what bugs me. It bugs me because it ignores theory. It bugs me because it ignores the evidence of our grandparents. But mostly it bugs me because it sends the wrong message.

I understand this only through vague ancestral memories. In the rest of this post I'm talking myth, not history. (I am bad at history).

It doesn't matter that much if rich dissolute celebrities blow their wealth by consuming the lot. (Except they set a bad example for those who can't afford to blow what little wealth they have). In the past, similar rich dissolute aristocrats were the celebrities of the day, and many set a bad example too. But opposing that example was a cultural movement (centred around a religious movement) which promoted a very different set of virtues, and especially for the poor.

Father said that Grandfather paid the farm men's wages on Saturday, to be met by some of their wives on Monday, asking for an advance on next week's wages, because their husbands had already spent the lot on drink. And how the village was split between Tory/Church of England/drinkers and Liberal/Non-Conformist/abstainers.

Commenter Determinant tells me it all changed in the 1930's, for the United Church in Canada anyway, when they shifted from promoting the old virtues towards something more like socialism. I think that was a dead end.

Nowadays the alliances seem to have shifted. The old virtues would now be seen as cultural conservatism, and are more associated with the right than the left of the political spectrum. Which is a pity. It's the poor that need them most.

(This post is not as well argued and evidenced as I want it to be. But I'm aware of the limits of my comparative advantage, so I'm going to stop here and post it as is. And give Frances something to respond to.)

Comments

[vimothy]

"Here is my basic assumption: everybody has the same preferences (until shown otherwise). If you took a poor person and gave them a rich person's (permanent) income, they would consume and save like a rich person, and vice versa. Everything scales."

These assumptions are crazy.

[Phil Koop]

Nick: oops, you are right. That was careless reading by me.

[Nick Rowe]

vimothy: "These assumptions are crazy."

Maybe. They are the standard assumptions of (e.g.) nearly all New Keynesian models. (The Egertsson/Krugman paper is an exception).

Phil: thanks. Thank God. Someone understands!

[vimothy]

Nick,

But that's a non sequitur. The issue isn't whether your assumptions are assumptions, it's whether those assumptions hold in the world outside the model.

It's crazy to think that, in reality, everyone has the same utility function, and that it is stable over time and location.

[vimothy]

Nick,

You wrote that “theory says that people who have a low rate of time preference (who are more patient) will tend to save, while those who have a high rate of time preference will dissave (have negative saving).”

People who go into debt have a high rate of time preference. Therefore, poor people who go into debt have a high rate of time preference.

Then you wrote that,

“Moreover, current income will be correlated with patience. This is not to say that the poor are impatient. But it is to say that the impatient will tend to become poor. Because the impatient will save and invest less.”

I.e., impatience (high rate of time preference) causes poverty—though you note that other things cause poverty as well.

[vimothy]

Sorry--I think I just italicised the comments.

[Patrick]

I see. Innate. Well, I still wouldn't dump on you. I think 'lefties' would dump on you because they'd perceive it as a moral judgement: the poor are born losers, or inferior, or whatever. Personally, I'd be glad to know what the root cause really was so we could devise effective policy.

FWIW, my view is that given how poverty breeds poverty and how hard it is for people to get out of poverty, that there MUST be something intrinsic or innate (or a lot of somethings) at work. Heck, if it really is just impatience causing poverty that'd be great! At least we'd know what do to tackle it. Meditation classes for all and we're done. Or maybe sitting around singing a little Kumbaya would do it.

:)

[Greg]

Nick:

I’m wondering if C = cY is just what is necessary (and sufficient?) for a stable distribution of income, in an economy. Then the parameter a in C = a + bY would be non-zero in a economy where the income distribution was undergoing change. Thus zero when Friedman was gathering his evidence, but positive now, where we know, at least in the US recently, income inequality is growing.

So one goal of government policy could be manipulating the a, making it positive to increase inequality, or (effectively?!) negative to flatten the income distribution. A non-zero, and more likely positive, a could also be an unintended consequence of inept government policy.

Of course, why settle for that, when you have a whole polynomial to play with: cY^2, which seems to be something like the motive behind progressive income taxation, etc.

[Oliver Davey]

I noticed my link above didn't work.

copied from the reviews (haven't got the book on me right now):

If pains and troubles are high enough, extra pain and trouble just isn't so bad. You hardly notice it.

paying the first bill in a stack of overdue bills does little to relieve a guilty conscience.

Let’s say you have 5 children. In a large house, where each child has his or her own room, a child leaving the house to go out into the world gives you diminishing marginal utility. The first room turns into an entertainment center, the second into a hobby room, and the third just sits empty. But if you are in a cramped, small 2 bedroom place for all of you, the first child leaving might only make a slight bit a difference compared to the second child leaving. By the time the 5th child leaves the home, you get the most marginal enjoyment of having your small place less cramped. Karelis point is that this inflection point is where we should be thinking about poverty, because as the token example above mentions, normal policy mechanisms based on neoclassical microeconomic theory won’t necessarily hold.

...Here’s the normal story. Picture you are in a room with 10 people. Each of them has a slice of cake. How much you are willing to pay for a slice of the cake is the ‘marginal utility’ of having it, and the more cake you have the less any more cake is worth to you. You’d be willing to pay a $1 for the first slice of cake, but you’d only be will to pay 90 cents for the second slice. You’d only be willing to pay 10 cents for the 9th slice, and a penny for the 10th slice. Eating the 10th slice of cake in that room would probably make you sick, hence you want it a lot less than the first slice, which is delicious. That’s declining marginal utility.

Now picture you are in a room with 10 people screaming. You hate it when people scream, and you can pay a person to get them to stop screaming. Would you pay in a similar way to the cake example? Would you pay a $1 to get the first person to stop screaming, and a penny for the 10th person to stop screaming?

No. Getting one person to stop screaming would make very little difference in how much you dislike being in the room. Modern psychology tells us you might not even notice it. You’d probably only pay a penny to get that first guy to stop screaming. However getting the second guy to stop screaming might be worth 10 cents. And the last guy, the difference between some screaming and no screaming, might be worth the full dollar to you. The more quiet it got, the more a marginal difference in how quiet it is would be worth to you. There’s increasing returns to this good; the 10th guy not screaming is worth more than the first guy not screaming, which is the exact opposite dynamic of the 10th cake being less delicious than the first.

For those not involved with economic theory this might just elicit a shrug, but this mechanism turns everything on its head. Let’s say that instead of money, you are given 20 tokens to be used over 4 days, and each token gets you one slice of cake in room #1, and one person to stop screaming in room #2. In the cake room, the optimal decision is to consumption smooth – eat five slices of cake each day, so you use the tokens {5,5,5,5}. In the screaming room, all the enjoyment is not in getting a room with half screaming but in getting a quiet room, and instead of consumption smoothing the optimal choice is to binge – pay 10 people to stop screaming the first two days, and deal with a loud room the last two days – {10,10,0,0}. This will hold even with ‘nudges’, say offering two extra tokens if you have people consumption smooth, since the marginal utility isn’t increasing that much. The utility of {10,10,0,0} is greater than that of {5,5,5,7}.

[reason]

Nick,
I don't have to assume that people are different, (observing that the wealthy become even more wealthy), but the consumption is a non-linear function of income. When Friedman make his hypothesis real wages for the unskilled were higher and tax rates on the rich were higher. You were looking at a smaller range of incomes where non-linearities are less clear.

[Charles Karelis]

Nice summary of the central point of my book on poverty by Oliver Davey. Charles Kareli

[vimothy]

Oliver,

Wow, that is a really excellent example of what I’m talking about. Where is it from? It feels familiar somehow.

Of course, there are countless reasons why we wouldn’t want to assume that everyone has the same utility function. We wouldn’t even want to assume that the same person has the same utility function in different times or different locations.

For instance, a common assumption is state irrelevance (i.e. only the payoff counts). In reality, this is unlikely to hold in many situations, and the fact that we have assumed it doesn’t tell you anything about whether it does or not. After all, you can assume anything that you want—go wild, make those indifference curves spell out your name in joined up handwriting if that’s what gets you off. But the whole point of making assumptions explicit is that it allows us to see what preconditions must be met for our results to be relevant.

"The best safeguard against overestimation of the range of applicability of economic propositions is a careful spelling out of the premises on which they rest. Precision and rigor in the statement of premises and proofs can be expected to have a sobering effect on our beliefs about the reach of the propositions we have developed."

—Koopmans

Isn’t it? I don’t think I’m going out on a limb here. It really is theoretically impossible to compare utility across individuals—by construction. Utility is subjective. That’s what I’ve been taught anyway.

Reason,

But people are different. Here’s what Nick should have said:

“Look at the paper! The workers and "investors" have totally different utility functions! They are totally different people! Given the exact same opportunities, they will behave in totally different ways!”

Indeed—they have, they are and they will. How can you possibly understand inequality if you don’t recognise this? Come back Karl, all is forgiven.

[Oliver Davey]

Well, Vimothy,

I think the point from the book by Karelis and also Nick's point, if I understand correctly, is that the utility function changes mainly with the level of poverty. Of course there are also outliers within every group and someone who becomes rich or poor very quickly may have difficulty adapting. But by and large it is situations that determine behaviour that is then internalised and passed on as 'culture' and not vice versa. Karelis' main point is that there is an inflection point of the utility function which is good indicator for finding and defining the proverty level in wealthy nations and should make us think about how we go about alleviating it. The title, for the fourth and last time (no, I'm not on the take ;-)) is 'The Persistence of Poverty: Why the Economics of the Well-Off Can't Help the Poor'.

[vimothy]

Oliver,

I thought I recognised it—it's on my Amazon wishlist! Apologies for making you write that out a fourth time. I see that Karelis even appeared in the comments directly above my post to think you for your summary. [/Reading comprehension FAIL/]

Unfortunately, you are quite wrong about Nick’s point—what he’s claiming is actually the opposite of that: “everybody has the same preferences (until shown otherwise)”.

But not only does everyone not have the same preferences, but they don’t even have the same preferences as themselves:

“One never steps in the same river twice and the comparison between a man’s utility now and his utility yesterday stands on precisely the same footing as the comparison of the utilities of two different men.”

[Fisher, Franklin M. and Karl Shell. (1968). "Taste and Quality Change in the Pure Theory of the True Cost-of-Living Index". Value, Capital, and Growth: Papers in Honour of Sir John Hicks. J. N. Wolfe. Edinburgh, University of Edinburgh Press: 97-139.]

That is, comparing the same person’s utility in different times or locations is as impossible as comparing the utility of different people. The problem of noncomparability is the reason why neoclassical price index theory cannot construct a genuine cost of living index—it just doesn’t make sense conceptually. Instead it looks at a hypothetical consumer fixed in time and space facing two different price systems (from the same paper):

“Given an indifference map, we compare two hypothetical situations, A and B. We ask how much income the consumer in B would require to make him just indifferent between facing B’s prices and facing A’s prices with a stated income. Note that the question of whether the consumer has the same utility in A as B never arises”.

It never arises, because it’s been ruled out of the analysis. That’s what consumer choice theory tells you about inequality: √FA.

[Oliver Davey]

There seems to be an intertemporarily high utility in reading miscomprehension :-).

[RSJ]

Vimothy,

I'm not sure how important the cardinal/ordinal distinction is. In studies in which people were offered baskets of various goods, they found it extremely difficult to choose between baskets, often requiring 20 or 30 minutes of hard thinking in trying to weigh which basket they preferred. It was an frustrating experience, and it's difficult to assume that people go through this experience each time they choose which of the thousands of goods and services they want to purchase. When they were given the baskets in different orders, they selected inconsistent (non-transitive) preferences.

In general, the notion that there is some mathematical utility function is itself an enormous simplification. I have no problem with making such a simplification, but once you do, worrying about whether it is cardinal or ordinal seems like a minor distinction. It's a bit like, after assuming that everyone is a sphere, you get into some heated debate about whether the sphere is transparent or opaque. The big leap of faith is mathematical utility maximization in the first place. Just go ahead and assume that its a function -- you would need that in order to define aggregate demand anyways, as you can't aggregate demand without assuming cardinal utility. Utility-maximizing macro *requires* cardinal utility. And I say go ahead and define a general consumer price level, too. It's no less realistic that assuming utility maximization in the first place.

What this has to do with poverty is beyond me. Debt doesn't have anything to do with poverty. The way I think about it, there are so many slots. People compete for the slots. The least competitive end up with the lowest income slots. Sure, you can tell them to shape up and be smarter, more patient, but then someone else gets the slot. Macro *should* be about the distribution of slots, and not about which characteristics happen to determine who ends up in which slot.

This may not be the whole truth. Some people are sufficiently entrepreneurial/lucky/etc to make their own slots. But few are, and when we see growing inequality or a shrinking middle class, it's not because character has changed, but because the institutional or emergent macro factors have eliminated some of the middle class slots and replaced them with others. This, too, is a great simplification, but again no less realistic than utility maximization.

[vimothy]

RSJ,

The distinction between ordinal and cardinal utility isn't important at all! Are you even reading my posts, or just Nick's responses?

Look:

“I think you’ve misunderstood me. I wasn’t bringing up cardinalism because I think that the distinction between cardinal and ordinal utility relates directly to the issue of interpersonal comparisons of utility. It doesn’t—it’s impossible in either case.”

worthwhile.typepad.com/worthwhile_canadian_initi/2011/01/inequality-and-debt-the-soft-bigotry-of-low-expectations.html?cid=6a00d83451688169e20147e200e586970b#comment-6a00d83451688169e20147e200e586970b

You’re making criticisms that I’ve already made upthread. Modern neoclassical consumer choice theory doesn’t really have anything to say about inequality. Like the square root of a negative number in the real number system, inequality is undefined. You ask your question, and the theory just returns an error message. That’s why it doesn’t make sense to draw on it, as Nick does in the original post, as some kind of proof that debt is not caused by inequality. Given certain assumptions, you get certain results. But since these assumptions are not realistic, the results are not relevant--because you can only compare utility across actors, times and locations if you make assumptions that are so incredible that they take what you're doing away from social science and redeposit it firmly in the realm of obscure theological disputes.

The cardinalists had other things going for them, such as an actual commitment to understanding and measuring inequality. But I don’t think it’s helpful to discuss them any more—it’s obviously just confusing the issue.

[RSJ]

No, I am reading everything -- maybe I misunderstood, but it seemed to me that you were arguing, particularly in the 1:01 post, against cardinal utility as part of the "war on assumptions".

But for me, some assumptions don't matter too much, and others are very important. IMO, cardinal vs. ordinal doesn't matter. What matters is the assumption that there is some endogenous "endowment" that is traded independently of the actions of others, or independent of the structure of the economy.

No one is productive outside the production system, and that production system is outside of their control. The CEO of Cisco, outside of the institution of Cisco, is just another old man whose income earning powers are limited. Whereas the current models assume that Chambers already has a few hundred million dollars of labor to supply (each year), and the firms are just bidding for this executive labor. But outside of the institution of the firm, the intrinsic labor supply of the CEO is undefined. His real endowment is the position he holds, which perhaps he got because of some character traits, but the position itself exists because of institutions.

If you get the poor to be more patient, then we get more patient poor people. But we don't get fewer poor people until we shuffle the institutions a bit so that our production system has more middle class slots and fewer superstar slots.

The distribution of these slots is not God-given, it's the result of power struggles, of active measures that firms take to reduce their dependence on some types of labor and increase their dependence on others. Firms actively take steps to reduce the number of value creating opportunities and concentrate the value creating potential into a few top slots. Government needs to actively fight these tendencies.

Then people compete for those slots. Asking people to compete harder isn't going to change the distribution of income earning opportunities in the economy. It's not going to do anything for equality.

So with any model, there is a whole host of assumptions that you can criticize as being unrealistic, but my argument is that the utility assumptions are not the ones to challenge. I would challenge the assumption that the supply of goods is separate from the demand for goods, that the supply of money is separate from the demand for money, and that the productivity of workers is independent of how firms structure themselves. If firms structure themselves in such a way as to rely on large numbers of low-value add jobs and a few superstar jobs, then we will have growing income inequality. If firms structure themselves in the opposite way, then we will have shrinking inequality. None of this has anything to do with the character issues that may determine who gets which job.

[vimothy]

I'm not arguing that there is anything wrong with making assumptions or with mainstream theory. I'm arguing for their correct application.

[Oliver Davey]

RSJ is right to keep the macro focus on a macro problem. Nevertheless, seeing also as inequality is here to stay in one form or another, it can be illuminating to search for better micro analyses. A goal which I think Karelis achieves.

[Nick Rowe]

I've been thinking about the Charles Karelis' point. It's interesting. It's plausible. I've finally got my head clear on it, I think. But I disagree with his conclusion.

Assume that the MU of C first increases, then decreases. You do not in that case consumption-smooth. You do lumpy consumption. You consume nothing, save up your income, then consume a big lump by dissaving. Then repeat.

And if you think about it, almost all goods have that property. I eat apples, but I do not spend every waking minute slowly nibbling an apple. Same with beer. Same with my own quiet time (canoeing). At a high enough frequency, our consumption may be less smooth than our income. You have to look at lower frequencies to see consumption-smoothing.

[reason]

Nick,
I think of it this way, I remember reading some sociology books from the 60s where they made a clear distinction between a working class culture and a middle class culture. The main distinction was the planning horizon. The working class culture didn't have one, they really did live day for day. If they had money, they partied. If they didn't they went hungry or scrounged. But hey they were used to it. And from their point of view it made sense. They didn't make enough to ever get anything substantial together - not enough to make a difference. The occasional party was worth the hard times that followed. At least they had some highs. And so if they had a couple of dollars left over at the end of the week, it went on lottery tickets.

Nick, does that make sense?

[Nick Rowe]

reason: It makes sense. But I'm not sure I would agree with that "culture of poverty" view. Some people just have bad luck, or skills that nobody happens to want much, and so they are poor. Doesn't mean they will or should buy into that culture. And others may have good luck, or skills that lots of people want a lot, and so could be rich. But if they happen to buy into that culture they won't stay rich, and may not even get rich.

Some people win big in life's lottery, and some win small. Some people blow their winnings and some don't. My baseline assumption is that the two dimensions are orthogonal.

[Oliver]

This is actually not Karelis' point, his being more along the lines of: often observed behaviour of the poor, such as drug abuse, violent crime, low academic performance etc., although not beneficial in the long run, is both rational and also utility maxising, just not in the common sense that is assumed in the consumer choice theory. I was umprecise above when I said culture of poverty above as Karelis consciously circumvents that discussion with his main theory. In fact the theory replaces it somewhat.

It remains though, that statistically it is far more likely for a person who is born poor than for another person, to remain poor, no matter what exogenous or endogenous factors he / she encounters along the way. Life is far less random than your description above suggests. You can put two equally intelligent kids in the same class but the family background will have the strongest influence on outcomes of the two later in life.

[Too Much Fed]

I believe this link is appropriate:

Inequality, leverage and crises

http://www.voxeu.org/index.php?q=node/6075