David R. Henderson  

Piketty's Dodge on Inequality

PRINT
How the interest rate increase... 1896: Immigration and The A...

New York Times economics columnist Eduardo Porter recently interviewed economist Thomas Piketty on his work on income and wealth inequality. Piketty, in case you haven't followed, has been documenting the increase in income and wealth inequality in the richer countries, including the United States.

Porter doesn't ask any questions about world inequality. My impression is that it has decreased as large countries like China and India have gotten wealthier. But, of course, their real incomes could be growing faster than Americans' real incomes and income inequality could still be increasing. Start with a per capita income of $2,000 that is growing at 8% per year in poorer countries and a per capita income of $30,000 that is growing at 1% per year in richer countries. One year hence, per capita income in the poor countries will be 8% higher, which is $160 more. One year hence, per capita income in the poorer countries will be $300 more, which is, obviously $140 more than $160. So the gap will have grown. It will take a number of years before the absolute gap falls.

By the way, there is one quick way to make sure that global inequality of income falls dramatically: for the richer countries to allow in an additional 50 to 100 million immigrants a year.

I would be interested in Piketty's take on both of the above. Apparently, Porter is not.

But I want to focus on the area on which Porter focuses: increasing inequality within a country, specifically the United States.

Porter asks a good question:

Might inequality in the United States be less damaging than it is in Europe because the very rich were not born into wealth, but earned their money by creating new products, services and technologies?

Porter was implicitly, I think, getting at an obvious point: that large rewards for innovation give incentives for innovation. The innovation will help hundreds of millions of people who will never be really wealthy: think of how you gain from a computer and a cell phone.

Here's how Piketty answers:
This is what the winners of the game like to claim. But for the losers this can be the worst of all worlds: They have a diminishing share of income and wealth, and at the same time they are depicted as undeserving.

Do you see what Piketty did? He didn't answer. Not only did he not answer what I think was Porter's implicit point--the large social value of the incentive to innovate--but also Piketty didn't even answer the narrow question asked: is inequality less damaging because many very rich people earned their money by innovating?

What did Piketty do? He made it about what the "winners" "like" to claim, not about whether the claim is true. Had I been Porter, I would have followed with something like the following:

Regardless of what they would "like" to claim, is it true?

Piketty also conflates increasing inequality with "losing." Notice that he calls those with a diminishing share of income and wealth "losers" even though they may well be winners, but not as big winners as those with an increasing share of income and wealth.



COMMENTS (20 to date)
R Richard Schweitzer writes:

In addition, there is a false assertion:

They have a diminishing share of income and wealth, and at the same time they are depicted as undeserving.

Actually, they get more "pie" as the pie gets larger (or as more "pies" are distributed globally). They may not get as much more as those better positioned get from the increase in the size of the "pie," (or in the number distributed).

More bellies are empty less of the time.

Further, as studies going back to Pareto's time seem to indicate, inequalities in a wide variety of economic and social measures seem to remain roughly within the Pareto Paradigm for most, if not all, societies. See, Transformation of Democracy Pareto, 1923

Aaron Zierman writes:

Good post.

I feel that it is disingenuous whenever I see increasing income inequality described in dollars. Inflation alone expands the $ amount difference.

Mr. Econotarian writes:

I know someone who made hundreds of millions of dollars in the first dot-com boom (his company was an important developer of Internet connectivity, and made several beneficial innovations as well as developing the industry).

I don't feel that I "lost" to him. Indeed, he was able to afford to invest in my start-up, so actually I benefitted from his "win". A large number of people he employed in other start-ups and even his household & boat helpers would not have those jobs if he did not "win".

David R. Henderson writes:

@R Richard Schweitzer,
I’ll check out the Pareto cite. Thanks.
Your point is not “in addition.” It’s the very point I made: think about what the noun “share” means.

Becky Hargrove writes:

Piketty has already ruffled my feathers by declaring that human capital does not count as real wealth. The above example illustrates the fact that he wants to move the discussion as far away from aggregate skills capacity as he possibly can. That way, the winners can keep knowledge use to themselves as their own artificially limited wealth. Which further diminishes the value of education, and the focus on divisions of assets will be the fight of the 21st century. Hopefully not, but that seems to be behind the inequality debate thus far.

Noah Carl writes:

According to a paper in the National Tax Journal in 2004, 'Top Wealth Shares in the United States', wealth inequality, unlike income inequality, has not really increased over the last half century. However, there may be more recent evidence on this.

Jameson writes:

The way I read it, there is no evading the question here. "This is what the winners of the game like to claim" seems to be a "No". In fact, taken at face value, Piketty seems to be saying that inequality in the US might be the worst possible kind of inequality, hence more damaging than in Europe.

Roger Koppl writes:

Great analysis, David. Most of the discussion about inequality on "both" sides of the issue is not about what most helps the least among us, but extraneous things like signaling empathy on the one side or disparaging the poor on the other side.

Gary writes:

"even though they may well be winners, but not as big winners as those with an increasing share of income and wealth."

Oh well....they "may be".....well, let's NOT study the data on how pretax wages have declined for the lowest quintile '79-'09 while they rose greatly {along with all other quintiles} '45-'79.

A share in the wealth and income....indeed.

Michael Brennen writes:

From his win/lose casting of inequality, one may be left with the impression that for Piketty economics is a zero sum game; when one wins, another necessarily loses.

In the interview he discussed the rate of return on capital against the rate of output growth, but there was no discussion of why the rate of return on capital is outpacing production growth. Thinking within the bounds of the US (as I know essentially nothing about other countries' tax systems,) with the distortions introduced into the US tax code to benefit very high wealth I wonder how much those codified distortions contribute to the high return on capital. If that is a relevant factor, to focus only on redistribution to correct the effects of distorted fiscal policy cannot but further distort outcomes.

In an Econtalk interview, James Galbraith made a startling assertion: subtract 15 counties from the data, and equality across the US did not change. I've not read his work, but it seems important to better understand that.

Galbraith on Inequality

Thomas Sewell writes:
Oh well....they "may be".....well, let's NOT study the data on how pretax wages have declined for the lowest quintile '79-'09 while they rose greatly {along with all other quintiles} '45-'79.
Gary,

Quintiles contain individuals, but they don't track individuals over time. This is a common fallacy when people start talking about income quintiles. For example, the people in the lowest quintile in 1979 are not the same set of people in the lowest quintile in 2009.

So it's entirely consistent to think that those in the lowest income quintile in 1979 are currently better off... and likely in a different quintile now, or (with that much time) they may be wealthy and retired and back into the lowest income quintile after passing through the others.

John Binder writes:

David Henderson:

Hasn't the following (quoted from your article) already happened with regard to the U. S.:

"By the way, there is one quick way to make sure that global inequality of income falls dramatically: for the richer countries to allow in an additional 50 to 100 million immigrants a year."

That is, illegal immigration has greatly increased over the last 40 years. And since many, unless I'm totally mistaken, of those immigrants were low skilled individuals who did not speak English, their incomes have been/are below the national average. And, of course, some of them have been granted citizenship via periodic amnesties.

And has this not by itself (everything else equal) caused an increase in national income inequality--the variance of the distribution of income?

When people look at a specific statistic and attribute a change in it to one particular thing, they are (implicitly at least) ignoring everything else--acting as if everything else has been constant over time.

Has anyone empirically addressed this issue or at least recognized that it affects the data for the U. S.? And it could for other countries with large influxes of relatively low income people.

Gary writes:

Gary,
Quintiles contain individuals, but they don't track individuals over time. This is a common fallacy when people start talking about income quintiles. For example, the people in the lowest quintile in 1979 are not the same set of people in the lowest quintile in 2009. So it's entirely consistent to think that those in the lowest income quintile in 1979 are currently better off... and likely in a different quintile now, or (with that much time) they may be wealthy and retired and back into the lowest income quintile after passing through the others.


Dearest "Thomas Sewell",

You avoid the point, the point is not tracking individual earners, but to look at the relative difference between groups of earners incomes in 2 time periods. The quintile of top earners in 79 is not the exact same group of people as in 09, so your point is moot. We know why earners in the lowest quintile in 79 are not earning what earners in that quintile did in 09, just as we do for top in the 2 time periods.

The point remains that Henderson ASSUMES that the lowest quintile has had real pretax wage gains, but they have not.

You don't want to discuss inequality of wages or Henderson's error....you want to change the topic to income mobility. Henderson was not discussing income mobility.

Roger McKinney writes:

Readers should chase any reading Piketty with a large shot of McCloskey, especially her Bourgeois Values.

While inequality has changed little over the past two centuries according to Piketty, real per capita income in the West and parts of the East have increased up to 100 fold.

If inequality has "hurt" people, then hurt me some more! I care more about getting filthy rich than worrying who has more!

Thomas Sewell writes:

Gary,

I still think it is you that is missing the point. Why does it matter to compare the income of one group of people in 79 to a different group of people in 09? It's a pretty useless statistic, which doesn't say much about how people have or have not had their lives become better over time.

To exaggerate to the extreme for the sake of illustration, if the bottom income quintile of people in 79 is in 09 in the second to the bottom, with the bottom quintile now filled with Mexican immigrants with improved lives from where they were in 79, you could still have more "inequality" in the U.S. and have every single person better off.

So the level of "inequality" as you're defining it doesn't have a good relationship to whether actual people are better or worse off over time. Why not use statistics about wealth levels of specific families or individuals over time, rather than income quintiles that aren't even very well correlated with wealth while some of the wealthiest are retired without much income.

It's a bad statistic that typically just leads to people drawing poor conclusions from meaningless data. Of course, if your goal is to mislead people, I suppose it may have some use to you. That seems to be how it most generally shows up in news reports, as a way to mislead people who don't understand it.

There isn't a finite pie of wealth out there that needs to be divided by some elite so as to be distributed to the chosen while the elite argue over the distribution percentages to their favored groups. People can be free to work peacefully with each other to create more wealth... and they will if the bureaucrats, politicians and rent-seekers don't continually get in their way.

nobody.really writes:
Porter asks a good question:
Might inequality in the United States be less damaging than it is in Europe because the very rich were not born into wealth, but earned their money by creating new products, services and technologies?

Porter was implicitly, I think, getting at an obvious point: that large rewards for innovation give incentives for innovation....

Here's how Piketty answers:

This is what the winners of the game like to claim. But for the losers this can be the worst of all worlds: They have a diminishing share of income and wealth, and at the same time they are depicted as undeserving.

Do you see what Piketty did? He didn't answer. Not only did he not answer what I think was Porter's implicit point--the large social value of the incentive to innovate--but also Piketty didn't even answer the narrow question asked: is inequality less damaging because many very rich people earned their money by innovating?

Wow. That’s damning. What the hell is wrong with that Piketty guy, dodging the question like that? I guess that just proves he’s a total fraud, huh?

But what if Piketty disagreed with the idea that inequality driven by earned income is less harmful than inequality driven by other reasons? What if Piketty agreed that it was possible that large returns for innovation provided a necessary incentive for innovation – but found that the data simply did not support this thesis? What if the data showed that income inequality accelerated since 1980, but the rate of productivity growth did not? What if Piketty felt that some incentive was important to innovation, but that we had reached the point of diminishing marginal returns?

But if Piketty really felt this way, he should have stood up on his own hind legs and said something like this:

[Some economists argue that inequality, in the United States, at least, is a good thing because it acts as an incentive for entrepreneurs to take risks and innovate, thus driving economic growth.] But in practice you see inequality everywhere, except in the productivity statistics. The rate of productivity growth has not been particularly good in the U.S. since 1980. Inequality is desirable up to a point. But beyond a certain level it is useless. One of the key lessons of the 20th century is that you can have high growth without the inequality of the 19th century.

Gosh, does that text look familiar to anyone? Could it perhaps be the answer Piketty gave to the question immediately prior to the question cited by Henderson?

And could it be that Piketty declined to repeat this answer in response to Porter’s follow-up question not because Piketty was dodging, but because he assumed that the reader could remember the answer he had just given?

Could be. But to find out, we’d have to actually read the interview before we judged it – something, I surmise, people are not inclined to do.

nobody.really writes:

Wait, wait, wait – back up the snark truck. Perhaps *I’m* the one who didn’t read carefully enough.

That is, Piketty forthrightly addresses, and dismissed, the thesis that huge incentives are needed to increase productivity. But he doesn’t specifically address innovation as distinct from other means of increasing productivity.

Piketty claims that since the 1980s we’ve seen more inequality yet no great growth in productivity. But that doesn’t disprove the possibility that 1) people have become more productive via innovation, 2) they wouldn’t have done this in the absence of large incentives, but 3) those productivity gains do not get reflected in aggregate social productivity gains because they’re offset by the loss of other sources of productivity. For example, perhaps we used to increase productivity by working more hours, increasing the labor force participation rate, desegregating the workforce, educating the workforce, and breeding – but now we’re achieving diminishing marginal returns on these strategies. Our primary path toward greater productivity now is innovation – and unlike prior methods of increasing productivity, innovation requires larger incentives.

I’m not persuaded by this thesis, but it’s not crazy either. And when asked about incentives for innovation specifically, Piketty offered a non-sequitur. So the criticism was justified – and my snark was not. My apologies.

David R. Henderson writes:

@nobody.really,
Apology accepted.

James Galbraith writes:

The comment on my work is close, but not quite what we found.

What we found is that if you take a measure of income inequality across counties, then taking 15 of them out of the mix eliminates the rise of inequality between 1993 and 2000. Taking out just 5 removes about half of it, and those five encompass Manhattan, Silicon Valley and Seattle.

We took counties out by replacing their real data with the average income growth for the country in the period.

Between-county inequality is not all inequality. But it does capture the extreme movements due to the information-technology boom in that period, and those movements do account for the close tie between the movement of overall inequality and the NASDAQ at that time.

The most recent paper on this is here.

http://wer.worldeconomicsassociation.org/

JG

David Henderson Author Profile Page writes:

Thanks, Jamie.

Comments for this entry have been closed
Return to top