David R. Henderson  

Confusion about Income Inequality

Keynesian confirmation bias... How to measure the influence o...

My Econlog co-blogger, Scott Sumner, on his own blog, The Money Illusion, writes that G.I.'s (he thinks this is Greg Ip) post on The Economist blog is a "wonderful post."

My view: it's good in some ways and not in others. What I liked was his paragraph about minimum wages, Costco, etc.

But I think he was off on income inequality. GI writes:

Economists can offer explanations for why inequality has risen and what might reverse it, but they cannot advance positive reasons for what the right level of inequality is; this is [a] function of social preferences outside the realm of empirical or even theoretical economics. (At least, they can't make the case on microeconomic grounds. There are macroeconomic reasons to fret over higher inequality because shifting income from the high consuming poor to the high saving rich reduces aggregate demand.) Fundamentally, economists are troubled by inequality for the same reason non-economists are: it doesn't seem right.

Notice that GI thinks that higher inequality means "shifting income from the high consuming poor to the high saving rich." He confuses income with wealth. Although there is a high correlation between income and wealth, we can't say someone is rich or poor based on income alone.

But put that aside. Let's say we could say that someone with high income is rich and someone with low income is poor. How does an increase in inequality mean that income was shifted from poor to rich? It doesn't. An increase in inequality tells us nothing about income shifts. You could have everyone's income growing and yet inequality increasing, as, in fact, has happened over the last 30 years. So there's no "income shift" at all.

This is further evidence for my view, that I laid out briefly on NPR yesterday, that we shouldn't worry about income inequality, but, instead, should worry about everyone being better off.

And, of course, that means that GI should not have said "economists are troubled by inequality." He could have said "some economists." He might even have said "most economists." But "economists" seems to imply "economists in general." That's simply not true. Interestingly, also, most non-economists don't seem that troubled by increases in inequality.

Comments and Sharing

CATEGORIES: Income Distribution

COMMENTS (28 to date)
Bostonian writes:

I suggest that libertarians and conservatives challenge the use of term "inequality" unless the *kind* of inequality is specified. Otherwise you accept the leftist idea that there is a single quantity, income, that every person's utility depends on and ignore trade-offs.

Suppose, for example, there are two married couples, in which all spouses are earning $50K annually. After having children, one wife chooses to stay at home and the other continues to work. Looking at income alone, "inequality" has increased, since one couple now has double the income of the other, but the couple with the housewife is not worse off overall than the other couple, since they have decided that the wife being at home is worth more than $50K.

Here is another example of ignored trade-offs. University professors decry inequality of income in part because they envy the higher salaries of businessmen, but they never mention their greater job security and freedom to publish their work. Academics trade off income for security and intellectual freedom.

Bill writes:

Suppose income WERE shifted from those with high MPCs to those with low MPCs. If the shifted income is saved - becomes loanable funds - and invested, how does this reduce aggregate demand from what it would have been had the income been spent on consumption?

Brian E. writes:
How does an increase in inequality mean that income was shifted from poor to rich? It doesn't. An increase in inequality tells us nothing about income shifts.

This is what I don't understand. Why do so many of the income inequality commentators think the rich are getting richer by preying on poor people, rather than by preying on other rich people? Why would someone wanting to get rich, or someone who is already rich and wants to stay rich, spend their time going after people who, by definition, have no money?

Technology is one of the best places to see this. The first televisions were expensive, and by today's standards, extremely poor quality. But rich people bought them, and over time the quality improved while the price fell. Now even poor people have televisions. I wasn't there to witness it myself, but I'm willing to bet the first guy to market an extremely expensive, 6" black and white television wasn't thinking "Just think of all the poor people that will line up to throw their hard-earned money away on this." while twirling his nefarious black mustache and adjusting his monocle.

Pick any appliance or technology produced in the last 60 years and you will see this. Novel and expensive inventions aren't marketed to poor people, though poor people may eventually own them once rich people have invested enough of their money into being early adopters.

Thomas Lee writes:

In my mind the current fretting over inequality is setting the stage for a huge power and wealth grab by governments. The justification will be that 1) taxes will be increased on the rich (or high income); 2) those funds will be re-distributed to the poor (or low income); 3) inequality will be reduced.

What is likely to happen is that 1) will be accomplished, but that 2) and 3) will not. What is to stop the U.S. government, for example, from starting new wars with the new revenue? Or to ramp up incarceration, which tends to target the poor? Even increased spending on medical care may not help the worse off as much as expected.

Effem writes:

What if psychological well-being is driven more by relative status than absolute status? Isn't it then possible to make society worse off as a whole by making everyone better off in a very unequal way? I see lots of evidence that confirms this. When 99% feel inferior to 1% (and further often feel that the 1% have benefited from hidden subsidies) it is possible that well-being has declined. Perhaps this is why all sorts of confidence measures are far down from previous levels despite everyone being monetarily "better off."

Vivian Darkbloom writes:

"Notice that GI thinks that higher inequality means "shifting income from the high consuming poor to the high saving rich." He confuses income with wealth. Although there is a high correlation between income and wealth, we can't say someone is rich or poor based on income alone."

Some generous reading of the quoted passage from G.I. might enable an interpretation that the "shift" is not a *transfer* from poor to rich as such, but merely that the "shift" represents a statistical phenomenon. That is, the "shift" might be due to the greater earnings differential (not caused by income transfer as such but simply growing disparity caused by other factors) combined with the greater propensity of those with the higher incomes/wealth to save.

I know this is generous, particularly in light of the use of the preposition "from", but can one rule it out?

Brian writes:

"This is what I don't understand. Why do so many of the income inequality commentators think the rich are getting richer by preying on poor people, rather than by preying on other rich people?"

Brian E.,

People think this because old stereotypes die hard. Prior to the rise of free-market capitalism, one of the few ways to get rich was to take from the poor. Capitalism made this approach unnecessary and unattractive, but most people have a hard time recognizing the difference. Of course, this confusion is strongly linked to the fact that some people can profit politically by perpetuating that stereotype, so it continues to be accepted.

A similar stereotype exists with regard to the term "poor." To most people this term implies economic destitution, but based on this standard almost no one in the U.S. is poor. Since this stereotype is useful for certain political purposes, it continues to be propagated.


Your point about income shifts is the right one to make. If "inequality" is increasing because the rich are growing their incomes faster than the poor, then it may very well be a good thing. After all, doesn't this mean that everyone is doing better?

And like Bill, I'm not sure I agree with GI's point about aggregate demand. When the wealthy "save" their income, they are usually investing it. And where do invested funds go? To pay people to create something new. The money isn't disappearing from circulation, as if someone were hiding it in a closet, so there should be no drop in aggregate demand. I don't get this argument at all.

Pajser writes:

Inequality really doesn't seem right; there are few possible reasons for that. One is diminishing value of wealth; $1 helps to poor more than it helps to rich people. Also, human psyche makes people on the bottom unhappy even if they are not poor. However, the main reason for egalitarianism could be that the value of act is relative to moral effort. Poor man gave $100 for good case; billionaire gave a million. People know that poor man is better man; that he deserved better reward; Catholics believe that poor man is closer to Heaven. But people also know that if they want to make that mentioned good case, they must show more gratitude to the billionaire.

Arthur_500 writes:

Saul Alinsky said we should focus on income inequality as a means to increase social unrest. That is what I see happening in the political arena while the media and other uninformed people talk about it because it "feels good" that people should have more money.

This really isn't an issue at all.

We don't live in a zero sum world so the income you have does not come at my expense. As we have seen in the data with inequality of income between men and women, much of the inequality comes from the types of jobs women gravitate towards.

When you have the poor living in a home with electricity and heat and food on the table and a color television and cel phone it is hard to worry about inequality. It seems the focus on income inequality is just a political guise to convince the public that we need to tax the "rich". Unfortunately, many of the supporters of that idea one day wake up and find out that they have been deemed rich.

Dan writes:
An increase in inequality tells us nothing about income shifts. You could have everyone's income growing and yet inequality increasing, as, in fact, has happened over the last 30 years. So there's no "income shift" at all.

The last part of this statement is not correct. There could have been an income shift over the past 30 years. If a larger portion of the new income generated goes to the top, everyone's income would have gone up even though there has been a shift.

michael Moran writes:

The income inequality debate is one I hate. But to start with a quote from a Ronald Reagan, "you can make the rich poor, but you cannot make the poor rich." This is another way of saying it is not a zero sum game, and policies designed to reduce income inequality will reduce the income of the rich (and middle class) but not increase the income of the poor.
Compare Sweden and the US. In an article written by the Coyote blog writer, he does a chart showing on a ppp basis what % of the population of Sweden (with its extensive policies designed to reduce income inequality) is better off than the US population. Basically it is the bottom 5%. So the top 95% of the US population is better off than the top 95% of the population of Sweden. Now let's think, do Sweden's schools and students rank higher on international charts than those in the US (yes), does more of the population of Sweden exhibit math and language skills at high levels (yes), do they waste less money on defense than does the US (yes), are they less corrupt than the US (yes), and do Swedes in the US have much higher income than average Americans and those left behind in Sweden (yes). So with the same policies it is likely that 100% of Swedes would be better off than Americans. But with their policies focusing on income inequality, they have made their top 95% poorer so that the top 95% of Americans are better off.

Julien Couvreur writes:

Pajser, repeat with me: I will not commit the error of interpersonal utility comparison.

I know it's tempting, but rigor is critical in economic analysis. Utility is subjective and not quantifiable.

Pajser writes:

Julien Couvreur, you're radical. I'm moderate. It is great if I can quantify phenomenons, but rarely I can, and if I cannot, i could still know something.

Thomas Boyle writes:


Avoiding interpersonal utility comparisons is standard economics. It is not "radical".

Simple example: A has an easy, emotionally-rewarding job. B works his ass off in an unpleasant one. They have the SAME income and wealth. It is plausible that, at the same income and wealth level, a marginal dollar is worth more to B, because he works harder for it. In fact, if B had double the income, plausibly the marginal dollar might be worth more to him, because he works harder for it.

Of course, we don't KNOW that. We surmise. We cannot know all the factors. We don't really know how much they value the dollar.

When you assume that dollars have less utility to those who earn more of them, stop and ask yourself: if the dollars are worth less, why do they earn more of them? Could it be that they have other factors that raise their marginal utility?

You should avoid interpersonal utility comparisons.

Aaron Zierman writes:

A lot of truly excellent comments above. Cheers to Henderson for addressing this topic, which can be difficult and frustrating as so many typically view this as an issue of "fairness". It is vital to keep pointing out that unequal outcomes are not simply a matter of unfairness. Those on the left all too often use income inequality to declare the failure of free markets (which to me makes no sense).

Roger Koppl writes:


Minor point: Vivian Darkbloom is right about the word "shift." I think we should give "GI" the benefit of the doubt on that score.

Main point: I think liberals err terribly to discount the importance of income inequality. I say so here:

FWIW, Matt Zwolinski agrees:

If liberalism (in our sense) most about the least among us, as I believe, then we should grab onto the income inequality issue and make it *our* issue. Inherently it is our issue.

David R. Henderson writes:

@Roger Koppl,
I read your piece. Excellent. BTW, is there a verb missing from your second-last sentence above?
So here’s how I would say it in the future, now that I’ve read your piece:
I don’t care about income inequality per se. What I oppose is the various government programs that increase inequality, such as the drug war, the para-militarization of police forces, lousy government schools, and occupational licensing.
What do you think?

Roger Koppl writes:

Hi David,

Adam Smith defined substance income as the level consistent with “common humanity.” That's a moving target: The subsistence level is higher the richer we are. Once we let subsistence depend on “common humanity," I think, we are really saying that we do have *some* interest in the *distribution* of income. Glen Whitman and I went around on this issue in the comments section of the ThinkMarkets post, with Glen taking the position you just articulated. I confess, I don't see it. If you cede the point that "subsistence" depends on how rich we are in general, then it seems to me that logic prevents you from saying that you don't care about distribution "per se."

MikeP writes:

I second the recommendation for Roger Koppl's excellent piece. But why stop with the short list?

You should add bigger fish such as Medicare, Social Security, the SEC, and the Fed.

The Fed benefits the rich as an accident of how it operates. If it could helicopter in new money, that would be better. But then there would be the problem of helicoptering it out when it wants money to tighten. The SEC, however, is simply a pure sop that keeps the richest on Wall Street rich by making sure they get to trade on new information first.

But even these are small potatoes compared to Medicare and Social Security. By not being severely means tested, these do nothing but take 15 cents of every dollar the poorest worker makes to protect the savings of the the richest age cohort until it is inherited by the second richest age cohort.

Frankly, anything government does that isn't specifically targeted towards the poor will necessarily increase inequality because it either benefits directly or flows through the hands of those whose incomes are above median. This includes even legitimate government functions such as defense and courts.

If you want more income equality, you want (a) more turnover at the top and (b) less funneling of money toward the top. For both, you need smaller government.

Roger Koppl writes:

Which is not to discount, BTW, your really fine performance with Krueger on PBS. Well done.

Pajser writes:

Thomas Boyle, your examples are valid and interesting. However, the claim "statements like '$1 helps to poor more than it helps to rich people' should be avoided" is indeed very radical; it is some kind of taboo. It is hard to justify that. You should show that "$1 helps to poor more ..." is necessarily wrong or meaningless or unprovable. I think you haven't that. Also, you asked "if the dollars are worth less, why do rich earn more of them?" There are many reasons. Luck, talent, effort, incentives, unscrupulousness? Where are we going with that?

David Summers writes:


One is diminishing value of wealth; $1 helps to poor more than it helps to rich people.

This is simply not true; a simple disproving case is this:

1) There are 10 coconuts on the island, 1 millionaire, and 2 guys with $1 each. The millionaire gets all the coconuts he wants (6, say), and the guys split the rest (2 each).
2) You then give the poor guys $10. The millionaire still gets all the coconuts he wants (6, say), and the guys split the rest (2 each).

What people seem to miss when they say this is that having an extra $1 only helps a poor person if no other poor person gets a dollar! Poor people do not compete with rich people for resources, they compete with other poor people. The rich get what they want, until they are "satisfied" - I'd say that is the definition of being rich. Whatever is left over is divided amongst the poor.

Taking from the rich and giving to the poor can only devalue the currency and decrease production. The poor (again, basically by definition) are not investing money, so investment disappears, and the GDP of the economy goes down. So there is less stuff available for purchase. The poor all have more money, so the settling price must rise to counter the increase in money (average price[of poor people stuff] must equal total money[of poor people] / total production[of stuff desired by poor people].

So taking from the rich and giving to the poor makes everyone worse off.

Admittedly, the psychological argument has a little more teeth to it, assuming the psychology holds. I could just as easily say that Bill Gates gives us all something to aspire too, though...

Curtis L writes:


I don't disagree with the majority of the post, except this sentence is definitely inaccurate: "You could have everyone's income growing and yet inequality increasing, as, in fact, has happened over the last 30 years."

No, everyone's income has NOT grown over the last 30 years. For many working class Americans, that number is stagnant at best and dropping at worst. Add to that the twenty- and thirty-somethings who don't even show up if you go back to 1984 but only the past 15-20 years and the income numbers are even worse.

I could easily be guilty of nitpicking, except this sort of level of factual detail is what you bring to GI's article (which I have no problem with doing by the way).

Pajser writes:

David Summers, what, $1 doesn't mean more to those on the bottom because nothing can be bought for $1? It is untypical situation. I can reformulate the claim "$1 helps to poor more than it helps to rich people. " to cover that particular situation, but the formulation will be less intuitive: let us denote minimal amount of money poor man needs to buy something useful with $Δ. $Δ helps to poor more than it helps to rich..

There are various indirect effect of redistribution, some bad, some good. Reduced investments, reduced crime, depression, suicides. Poor people have less children and their children are better educated. More jobs for teachers and professors. Less for lawyers. Increase of bureaucracy. Who knows what. All that should be taken into consideration.

Brian writes:

"There are various indirect effect of redistribution, some bad, some good."


David Summers is right. Let's put it this way. Redistribution has the effect of giving money to people who have done nothing for it. They have not created additional value nor increased productivity. How do we know? If they had, someone would have been willing to pay them for it. When people receive money through the free market, it's because the provided something of value. When the money is redistributed by the government, the transfer occurs without the production. This is a net loss to society and, frankly, a net loss to everyone, including the poor.

Pajser writes:

Brian, let us assume that work and goods are initially distributed by free market. Additional redistribution organized by state occurs without the production and that is the cost for whole society; I agree about that. However, I do not see the conclusion that it is, as you said, frankly, net loss to poor people. You're gone too far with that. It seems obvious that poor people are better off after redistribution.

I would also agree that distribution would be net loss for society as whole, if there is no diminishing value of wealth; the phenomenon that $1 is worth on the bottom more than on the top. Although that idea was criticized on few ways - it still sounds true to me.

Brian writes:

"It seems obvious that poor people are better off after redistribution."


Yes, they seem better off in the short run because they have more money. But in the long run it doesn't make them fit for anything besides being on the receiving side of unearned largesse. If they can get by simply by being, AND since they risk losing the payments (in some cases) by earning too much, they have little motivation or confidence for improving their lot. This is terribly destructive of their dignity and potential. So yes, it's a net loss for poor people.

Regarding the value of a dollar, I think it's correct to say that the marginal value of a dollar is higher for a poor person than for a rich one. But the value to society as a whole is the same--society doesn't care who spends the dollar as long as it's spent. So there's no gain to society based on redistribution alone. Even worse, a high proportion of the money spent by the poor goes for consumption, which is a wealth-destroying activity. The wealthy spend a relatively large proportion of their money on investment, which is a wealth-creating activity. It seems, then, that society benefits greatly by keeping money in the hands of those who create wealth. This is exactly what the free market does, provided the government has the sense to keep out of it, beyond providing protection for property rights, of course.

Pajser writes:

Brian, you're not wrong about possible influence of redistribution, but you jumped on conclusion that it will cause so much harm that it will be net loss for poor people. We don't know that. There are different kinds of poor people, different reasons for poverty, different ways redistribution can be organized.

Value for society is probably something like sum of the values for individuals.

And you're right that in capitalism, wealthy people invest more than poor people. It is good argument against redistribution. The state should counteract that.

Comments for this entry have been closed
Return to top