Income and wealth inequality data: Nonsense on stilts

I keep running across blog posts showing the inequality of income and wealth in America.  In a recent post I already discussed one reason why this data is fatally flawed, capital income is nothing like wage income—rather it is deferred consumption.  Counting capital income and wage income is actually counting the same income twice.  Here I’d like to discuss some other problems with the data:

1.  Life cycle effects:  I dragged out my annual Social Security data that I get in the mail, and it shows how much I earned during each year.  I tried to do a rough adjustment for cost of living changes, to make things fairer (otherwise my income looks extremely unequal.)  I am pretty sure my five income “quintiles” are roughly as follows:   3%, 13%, 22%, 27%, and 35%.  In other words during my worst 7 years I made 3% of my total real lifetime income, and during my best 5 years about 35%.  Some people have a more equal profile, whereas others have a far more unequal profile.  I think I’m probably not that atypical.  The point is that if we had 100% lifetime equality in earnings, but wages that rose with age and experience, then that’s the sort of income inequality we might observe in America.  The actual income inequality is greater, because inequality is not just due to life-cycle effects.

2.  Inflation:  Suppose you are the richest guy in the world, owning $100 billion in Microsoft stock.  You cash out and decide to live off the interest.  To avoid inflation risk, you put it all in 10-year indexed TIPS.  You would earn $650,000,000 per year in interest.  Unfortunately your tax liability would be more than $650,000,000.    The government would report your income as about $2 billion.  So the person who might well have the highest reported income in the entire country according to official data, might not have any real income at all.  Now obviously most rich people don’t put all their money in TIPS.  They take bigger risks and get positive rates of return.  But risk implies the possibility of loss.  Some do much worse than the hypothetical I gave you.  Of course even with no income a person this rich has a fabulous lifestyle, which is what I would argue is exactly the point.  Look at consumption inequality, not income inequality.

3.  I am pretty sure that a lot of the wealth inequality data is incomplete.  I recall reading that it often ignores structures (which is much of the wealth for average homeowners and shopkeepers.)  It may ignore pensions.  Many retired public employees have defined benefit pensions that would be hard to replicate with a 401k holding a million dollars.  It ignores human capital, making it impossible to compare human capital-rich brain surgeons and lawyers, with physical capital-rich farmers and landlords.

4.  Income inequality data is often collected at the household level, implying that a doctor making $250,000 with a stay at home spouse is no better off that a Boston cop making $150,000 (including lots of overtime) married to a nurse making $100,000 (including lots of overtime.)  But the two income couple might have to spend money on child care, and have very stressful lives doing household chores on top of their paid jobs.  This isn’t a major bias, but many people who naively think of the top quintile as being “rich” would be shocked at how many working class couples in their 40s or 50s who are dual income and live in high cost areas like NYC and Boston actually fall into that category.  I’d guess two married people each making $55,000 would make it the top quintile, and I’d guess a couple who each make $75,000 would make the top decile.  Those aren’t gaudy incomes around here.

I am not trying to argue the upper middle class can’t afford to pay more taxes.  (I don’t want to get mauled like that poor U of C professor.)  Indeed, I think most Americans could afford to pay much more taxes, as we’ve become used to having lots of stuff we really don’t need.  A small Hyundai will get us from A to B just as well as a Lexus.  So that’s not the point.

Instead, my point is that we should ignore all the official data, and use our eyes.  Travel around the country.  Go into poor people’s houses.  In the 1970s I recall staying with a rural family of six in a small house who had running water for only a couple hours of the day.  Over the course of my life I’ve seen lots of poor urban and rural neighborhoods.  And I’ve driven around affluent areas like Newport Beach and Wellesey, and very rich areas like Beverly Hills.  I don’t really know what’s it’s like to be poor in a cultural sense, or not be able to afford food for my family, but I think I do have a rough sense of the different sorts of consumption bundles purchased by different classes of people.  For what it’s worth, here’s my impressions:

1.  All classes in America are better off than in the 1960s.  But the gains are most noticeable for the poor and rich.  Especially the poor in the rural South.

2.  The US has more inequality than other developed countries.

3.  Some income redistribution should occur, via a progressive consumption tax and subsidies for the poor (wages subsidies, disability subsidies, HSA subsidies, education vouchers, etc.)

So I don’t have any objection to policies of redistribution, which is what motivates all these comparisons.  But it annoys me that people are making arguments using worthless data.  Here’s an example of a graph I found in a Matt Yglesias post:

It shows wealth inequality by quintile, using data that is utterly meaningless, then shows wealth inequality as perceived by the average American.  Here you are supposed to laugh at the stupidity of Americans.  But their view may actually be closer to the truth than the official data.  Then the graph shows the inequality that Americans think would be fair.  Even the Bush voters opt for a wealth distribution far more equal than what we actually have.

Here’s the problem with this entire enterprise.  Let’s work  with the wealth definition that was probably used in this table, that is, only easy to measure financial assets.  Assume this data is correct.  How much income equality would we need to get things as equal as the Bush voters want?  I’m going to claim that even 100% income equality would not be enough.  That’s right, if you paid a 16 year old boy with pimples at McDonalds exactly the same income as a brain surgeon at Mass General, measured wealth in America would still be far more unequal that what Bush voters say we should aim for.  That makes Bush voters to the left of Mao, almost at Pol Pot levels of egalitarianism.  And the reason is simple.  Even with exactly equal incomes, people will vary greatly in how much they save, and how well they invest what they do save.  So even with equal incomes, some would become very wealthy, and some would save almost nothing.

With apologies to Bentham, income inequality data is nonsense, and wealth inequality data is nonsense on stilts.  It’s all about consumption.

Yglesias’ post was entitled “Poor People Are Much Poorer Than You Think.”  Actually, if you are an observant person, and ignore the data, economic inequality in America is exactly what you’d think.



58 Responses to “Income and wealth inequality data: Nonsense on stilts”

  1. Gravatar of wcw wcw
    29. September 2010 at 05:49

    If you are an observant person, economic inequality in America is exactly what you’d think — but you have to get out. Lots of people don’t. If you take public transit in an urban setting *and* travel a good deal *and* keep your eyes open, sure. Otherwise, it is extremely easy to live a life cloistered in your immediate class, and that makes the US look flat, instead of like a power law.

    Oh, and income data are not nonsense, but I don’t feel like arguing.

  2. Gravatar of Upandaway Upandaway
    29. September 2010 at 05:51

    “1. All classes in America are better off than in the 1960s. But the gains are most noticeable for the poor and rich. Especially the poor in the rural South.”

    But in what way? If, as you say, “It’s all about consumption”, then yeah, thats true. If you instead consider the power, prestige and security that money buys then it’s not. Or the fact that they are (on average) much less likely to move to the top during their lifetimes (income mobility).

    And the example of the Mc:d 16 year old and the surgeon is mostly a straw man. That’s not the problem at all, rather that the bottom 40% have *no* buffer against misfortune what-so-ever.

    The real beef I have with the graph you display is that it doesn’t display the upper 1%, 0.1% and 0,01% wealth, who possess some eye-popping amounts of wealth.

    But hey, if you really want an aristocracy that bad, who am I to tell you not to? After all, if they have all that money, they must really smart and deserving. And if you have no money, then you just have to work harder!

  3. Gravatar of Keith Keith
    29. September 2010 at 06:10

    Furthermore, isn’t capital accumulation a good thing? After all, it is the accumulation of capital that drives economic growth (including human capital and the technological progress associated with each new generation of capital).

    When we create policies that hinder capital accumulation, we shoot ourselves in the foot: rich and poor alike.

  4. Gravatar of William William
    29. September 2010 at 06:29

    Just a quick defense of Dan Ariely (the author of the paper in question): I’m sure that the point of the graph was not to laugh at the stupidity of Americans. The point was to illustrate how far we are from what his survey respondents said they consider “ideal” (and how much closer Sweden is).

    That said, Dan Ariely is wrong about inequality and you’re right. Using even a little bit of economics completely smashes these measures of “Wealth inequality.” If you leave out human capital, you’re leaving out almost all of the wealth that young people have. We’re closer to the “ideal” than Ariely thinks.

  5. Gravatar of Indy Indy
    29. September 2010 at 06:31

    I think we’ve lost all sense of perspective when we talk about the experience of poverty in American in 2010 – internationally, and historically. For the left, the “distressed condition of the poor” serves as the fundamental motivation and justification for their program of societal reform that will remedy this suffering and achieve “social justice”.

    If one were to admit, say, that the condition of the “American poor” today is so distant from the poor of yesterday, or the poor abroad, that it is hardly comparable, then one loses that fundamental pressing argument for urgent and dramatic political change.

    This is why the modern progressives *can not let this go* and be satisfied with the amazing success that has been achieved (largely by technological and productivity progress, I would argue, though partially by government intervention), in the last century.

    They will still be harping about the inexcusably awful condition of “the poor” in 2050 too, when the poor will have things that would make Bill Gates envious today. “A rising tide lifts all boats” is true, but ineffective as a politically satisfying doctrine that justifies aggressive action, and so the “relative position” of different classes in an inexhaustible fountain of malice, even if the absolute position of the poor is incredibly improved.

    An objective measure of the physical manifestations of true destitution – as judged from the perspective of an early progressive of, say, 90 years ago, would yield a conclusion that their view of the problem is largely solved, and, in fact, they would simply be stunned and in awe at the many ubiquitous and cheap technological luxuries that our “poor” routinely enjoy and that would have made any Emperor or ultra-rich capitalist of even 75 years ago completely envious. I’m positive that air conditioning would be an an incredibly happy bonus for anyone with my thermal preferences.

    This is experience of new immigrants who come from genuinely poor countries and it was certainly that of my immigrant grandparents who thought the poor of even 1948 America were actually just “less rich”, at least compared to the poverty they had experienced in their own youth. Though, today, even the poor in poor countries carry cheap cell phones around (which I saw in Afghanistan, for example), which would have been a complete miracle-technology of usefulness in the developed world only a generation ago.

    When a 1920’s progressive spoke about the poor, they were talking about people who were not just uneducated but largely illiterate, malnourished or starving, ill-clothed, cramped in awful unsanitary shelters with dirt floors, no plumbing, no electricity or appliances, no ventilation, outhouses, no refuse or sanitation services, and sparse, make-do furniture, (we would barely agree to call their sleeping arrangements “beds”) and no ability to afford almost any convenient mode of transportation. Think “How the other half lives” which was about New York slums of the 1880’s.

    The question, I suppose, is that if you were to put the average material circumstances of today’s bottom quintile in front of someone from the middle quintile of 50 years years ago, would they want to switch? Would they see themselves as being better off and having more utility? I think the answer is clearly yes.

    And if you’ve raised the material condition of your “poor” such that they exceed those of the “middle class” of a generation or two prior – you’ve won, and you are now allowed to stop exaggerating the scope of their utter depravity, even though your personal vision of Utopia wasn’t quite fully achieved.

  6. Gravatar of Contemplationist Contemplationist
    29. September 2010 at 06:34

    Scott you maybe like Tino Sanandaji. He has done a few posts on this topic (and comparisons to Europe).

    HERE he compares the productivity and income (gdp/capita) b/w US and Europe and tries to normalize for things like ethnicity, etc.

    HERE he compares the income distribution is egalitarian Sweden vs inegalitarian US.

  7. Gravatar of Master of None Master of None
    29. September 2010 at 06:41

    Well done, both in terms of content and writing style (much more concise than usual!)

    You have changed the way I will talk about ‘inequality’ forever, for the better.

  8. Gravatar of woupiestek woupiestek
    29. September 2010 at 07:05

    There are of course advantages to concentrating wealth and power in the hands of a few capable persons. Capitalism works because of wealth inequality. The market chooses competent people to guide its invisible hand. O.K. I don’t really have that much faith in the market, but I can imagine that when in the optimal situation the 20% most competent (and altruistic?) people should have 80% of all wealth.

    Of topic: a while ago you wondered why we (Europeans) started to do so many explorations at the beginning of the renaissance. The formal ending of the middle ages is the fall of Constantinople in 1453. Form then on the Turks could block our trade with the East. So I don’t think it is a coincidence that in the decades thereafter, we started looking for new routes to the East.

  9. Gravatar of ChacoKevy ChacoKevy
    29. September 2010 at 07:23

    Poor U of C professor? C’monnnnnn…. From the Dixie Chicks to Jake DeSantis, he should have seen this coming. I too am saddened he was threatened, but his surprise to it leads me to think him quite naive… and terrible with personal finance.

  10. Gravatar of StatsGuy StatsGuy
    29. September 2010 at 07:31


    “In a recent post I already discussed one reason why this data is fatally flawed, capital income is nothing like wage income””rather it is deferred consumption. Counting capital income and wage income is actually counting the same income twice.”

    This is half true, half false.

    Risk adjusted investment returns in excess of the inflation rate (whatever that is) are not delayed consumption. They are return on investment – PAYMENT for delaying consumption. As such, they are income. Since the entire post begins with this premise, I’m not sure what to make of the rest of it.

    If risk adjusted returns are lower than inflation, that it’s a PENALTY for delaying consumption. If this is widely true, then it’s an expression of the market’s belief that the future will involve lower supply (per person, at least) than the present.

    FYI, one of the reasons I like your NGDP targeting regime is that it’s a way to prevent excess persistent wealth aggregation. In times of decreasing supply (perhaps due to 30 years of insufficient technology investment, or some sort of revenge of Malthus), an NGDP target prevents accumulated capital from claiming an increasing share of distributional outcomes into the future. In essence, it’s a tax on wealth that prevents feudalism. It also creates some distortions – notably, a rush to buy and store commodities to claim future consumption. Then again, if the future really is a harsh place with low supply, this is a good thing, since it decreases commodity consumption now, encourages storage, and encourages a search for new supply and alternatives.

    An NGDP target is an incredibly progressive policy, IF it can be committed to. One of my great fears would be that an NGDP target is liked by capital when real growth is high, and disliked when real growth is low – causing financial interests to try to defect from a commitment to an NGDP target (or even an inflation target) in times of low growth (or retirement bubbles). This could be compounded by some countries seeking to keep currencies high to become safe banking havens, although empirically it seems that export and BoP interests have kept even countries like Switz from seeking too high a currency.

  11. Gravatar of ssumner ssumner
    29. September 2010 at 08:00

    wcw, Regardng income data, check out the post I link to at the beginning of my post. It simply adds wage and capital income as if they are two different things, rather than the same income measured at two points in time.

    upandaway, I’m afraid you didn’t understan my post at all.

    1. I argued for redistribution, not against.

    2. I never claimed the income differences between kids and surgeons are a problem, I claimed if their were no income differences between the two, MEASURED wealth would still be highly unequal.

    3. The wealth inequality data you cite is worthless, as it excludes the sort of wealth most people rely on.

    You said;

    “But hey, if you really want an aristocracy that bad, who am I to tell you not to? After all, if they have all that money, they must really smart and deserving. And if you have no money, then you just have to work harder!”

    Some of your comments are debateable, but not related to any assertions I made in the post. In general, I don’t like people who use emotional arguments, not logic. They think anyone who questions the accuracy of their data must be a really selfish person. Those seem to be your views.

    Keith, Exactly. This is why we should tax consumption, not saving.

    William, Good point, I shouldn’t have implied he thought Americans were stupid, I suppose what I meant is that most people reading that table would assume Americans are really uninformed about economic inequality, and I see no evidence they are. Every American I’ve ever met knows there are lots of poor people in the inner cities, Appalachia, Indian reservations, etc. They know those people own very little in the way of stocks and bonds. That also know about wealthy areas of the country. There’s no big mystery here.

    Indy, You are right that the poor have a much higher standard of living than in the past. Of course the left complains about inequality relative to our current livings standards. I prefer to get past that sterile debate by looking for small government models that also help the poor–such as forced saving schemes, where the middle class self-insures and the government need spend relatively little money on social services. It makes it easy to then subsidize the poor.

    Contemplationist, Thanks for those links. BTW, I could have made this post far more right wing. For instance, I said that America has more inequality that Europe. Instead, I could have mentioned that many of those who are worst off in Europe (Gypsies, Arabs, etc) are far better off when living in America.

    Thanks Master of None. Some of my posts desperately need an editor.

    Woupiestek, That’s a good point. Given that people who are super-rich generally don’t consume all that much of their wealth, I am content to let them allocate capital, and then give most of their fortune to charity when the die (or before.)

    I knew that about the silk route being blocked, but I forgot it when writing that post. That’s for reminding me.

    ChacoKevy. You misunderstood me, I was using “poor” in a slightly sarcastic way. On the other hand I didn’t know he’d been threatened. If he has then I think you are wrong, I do sympathize with him.

  12. Gravatar of Benjamin Cole Benjamin Cole
    29. September 2010 at 08:20

    Excellent post.

    Here in Los Angeles, there has been a widening of the economic spectrum in the last 50 years.

    Huge middle-class swaths–cities like Long Baech, South Gate, Compton, Bell Gardens, where people made good incomes in auto, steel, aluminium, rubber, tire and glass plants, have become more poor, though better in last 15 years.

    In the last 15-20 years, immigrants seemed to have replaced previous residents in the poor parts of town, upgrading neighborhoods, and reinvigorating small businesses, such as markets, auto repair, small cabinet shops etc. Still, large swaths of Los Angeles were more prosperous the 1950s-70s.

    Meanwhile, the wealthier parts of Los Angeles have exploded with jumbo-sized mansions, and cars that are so fancy I am not sure what model or brand they are.

    My back-of-the-envelope take is that typical or average Beverly Hills residents used to be three times as wealthy as those of Altadena, where I grew up. A home in BH was 100k; you might buy one in Altadena for 30k, back in the 1960-70s. A Cadillac was $10k and a Ford Falcon was $3k.

    Now I estimate the typical BH resident is about 10 times as wealthy as the typical Altadena resident today.

    Maybe my youth was naive (and certainly very middle class) but it seems to me the second and third vacation home is a newish trend, but standard among the wealthy now.

    Also, if you watch films/TV shows from the 1950s-60s, the representations of what is “rich” seem modest by today’s standards.

    The top tax rate was 90 percent throughout the 1950-60s.

    I would say the poor are better off, but the wealthy live on another planet now.

  13. Gravatar of Benjamin Cole Benjamin Cole
    29. September 2010 at 08:24

    I enjoyed your post. Remember, the WWII draft found men who were on average two uniform sizes smaller than those of today. Many men were 4F due to malnutrition. Bad teeth, feet, weak bones, etc. Rickets.

    Now the typical American is fat, whatever the income class.

  14. Gravatar of Morgan Warstler Morgan Warstler
    29. September 2010 at 08:52

    1. I said this at TP: what people “expect” means nothing. Give them the choice of modifying their behavior, or deciding reality is just fine, and they will choose to accept reality as harsh as the real graph looks.

    The FACT is people don’t think they should have to give up anything else, and they don’t THINK the poor deserve more than they have.

    Likely voters are comfy with things as they are.

    2. Disparity means nothing. This is Rawls. I made a nice graph for everyone to understand this:

    The American system is best because our poorest have more than the poorest in any other country.

    3. Finally, it is all about the toilets.

  15. Gravatar of Joe Joe
    29. September 2010 at 09:27


    Here’s an article by Terry Fitzgerald at the Minneapolis Fed. This is the data you’re looking for….

    â–  The U.S. Census Bureau reports that median household
    income stagnated from 1976 to 2006, growing by only 18
    percent. In contrast, data from the Bureau of Economic
    Analysis indicate that income per person was up 80 percent.

    â–  Three data issues adversely impact reported median
    household income gains: the choice of price index, a
    change in the mix of household types and the measure of
    income used.

    â–  After adjusting the Census data for these three issues,
    inflation-adjusted median household income for most
    household types is seen to have increased by 44 percent
    to 62 percent from 1976 to 2006.

    And here’s an awesome article by Robert Rector going through teh actual material conditions of the bottom 20%…



  16. Gravatar of Lee Kelly Lee Kelly
    29. September 2010 at 09:47

    In the interminable long-run, income and consumption are equal. If returns on investments are income, then two people who spend equal incomes differently can derive unequal consumption in the long-run. Reductio ad absurdum: investment returns are not income.

  17. Gravatar of Nuveen Nuveen
    29. September 2010 at 09:56

    Interesting post. I myself am guilty of blog posts with graphs like these, and I never considered life-cycle effects. Some questions:

    1. I’m not sure of the source of Norton and Ariely’s data, but other common graphs that float around the internet are from the CBPP and Emmanuel Saez. Do you know if their data make any attempt to account for your concerns?

    2. Has there been empirical investigation to measure how great a difference things like life-cycle affects make on the data? I ask because it seems like a lot of the top-earners start out in the upper quintile, like, for instance an young invesment banking analyst who will later become a hedge fund manager or a CFO. What they earn at 23 is what plenty of people earn after 30 years on the job. Life-cycle affects skew the data, but perhaps not by that much.

    3. I agree with you when you say that the poor are better off today than they were in 1960, but can you say the same of the middle class? What has filled the void that used to be filled by skilled labor? Maybe the Lorenz curve now has a convex kink?

    4. Are there any drawbacks to focusing on consumption equality?

    5. What do you have to say about the graphs showing that real wages have fallen for everyone but the top quintile since 1979? Presumably those reflect what can be consumed.

  18. Gravatar of Paul Zrimsek Paul Zrimsek
    29. September 2010 at 10:12

    Even with exactly equal incomes, people will vary greatly in how much they save, and how well they invest what they do save.

    And even if they didn’t, you’d still get massive inequality in wealth simply because older people have been accumulating it for longer. Mistaking life-cycle changes for class divides is fatally easy even when looking at income data; looking at wealth only aggravates the problem.

  19. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    29. September 2010 at 10:17

    ‘Risk adjusted investment returns in excess of the inflation rate (whatever that is) are not delayed consumption. They are return on investment – PAYMENT for delaying consumption. As such, they are income.’

    So, those who don’t delay consumption don’t get paid for so doing. What’s the gripe then?

  20. Gravatar of abb1 abb1
    29. September 2010 at 10:19

    1. Life cycle effects:

    Those quintiles are probably households, not individuals, and when you had that very low income of yours, you probably belonged to a relatively well-to-do household. Counting households has to significantly weaken this “life cycle effect”.

    2. Inflation

    If all you wanted was to preserve your savings, you could, of course, invest in some tax-exempt muni-bonds or something. If you didn’t, I have to conclude that your goal wasn’t hedging against inflation, but you were actually trying to use your money to make more money. Fair enough? And that’s fine, but in that case your gains are clearly qualified as income and should be taxed as any other income – IMO at a much higher rate than earned income, but that’s just my opinion.

    3. I am pretty sure that a lot of the wealth inequality data is incomplete.

    Yes, me too. In reality it’s probably even worse.

    4. Income inequality data is often collected at the household level

    Well, good. Your $250K doctors’ household will probably come out much wealthier than your cop+nurse household. Makes sense.

    Travel around the country. Go into poor people’s houses.

    There are plenty of places in the country where you can’t travel and go into people’s houses. You can’t travel to urban ghettos and go people’s houses there, because it’s not safe; in fact, it’s probably very dangerous. And that’s where poor people live.

  21. Gravatar of JeffreyY JeffreyY
    29. September 2010 at 10:47

    Steve Waldman had an interesting post arguing against the use of consumption inequality:

    You say, “I recall reading that [wealth] often ignores structures (which is much of the wealth for average homeowners and shopkeepers.)”. But the U.S. has only a 67% homeownership rate (which includes people who are still deeply in debt on their houses), which means nearly all of the bottom two quintiles don’t have that sort of wealth. If they also haven’t been public employees, they don’t have a pension plan (and most public employees make above the second quintile). So, where’s their wealth? Why do you think it’s 10% of total U.S. wealth?

    You also ask “How much income equality would we need to get things as equal as the Bush voters want?”, but says that’s as equal as Sweden actually is, so the answer is “Sweden’s income equality and redistribution mechanisms.”

  22. Gravatar of Ted Ted
    29. September 2010 at 10:52

    I agree that consumption is a better measure of welfare, which is what we should care about, but I still think income inequality data (at least micro data) helps us to better understand the dynamics of economic inequality.

    On a related note, if you are interesting in better measurements of economic inequality – including consumption inequality – the best paper I’ve seen on the subject is Heathcote, Perri, and Violante’s “Unequal We Stand”:

    It’s extremely thorough. It covers inequality in individual wages and earnings, household wages and earnings, disposable income inequality, pre-transfer inequality, and, most importantly consumption inequality. It’s also the only paper I’ve seen that really address financial asset income, life-cycle effects, and what not.

    Some of the results for consumption inequality they found are actually interesting. As well as confirming your intuition that wage and income inequality are bad measures of the relevant variable, consumption inequality, they specifically find:

    i) Inequality in consumption rises by less than half as much as disposable income over 1967-2006 so measures of income inequality are likely vastly overstating relevant welfare inequality.

    ii) Income inequality played a very little role in consumption inequality up until about 2000 and thereafter it was the behavior of the top 5% of households that were driving consumption inequality.

    iii) There is very little difference in consumption inequality between the median 50 household and the bottom 10 households, though there is a sizable difference between the top 90 households and the median and bottom households. What I take this to mean is that the bottom households are much better off in terms of welfare than income inequality data would suggest. The authors suggest this may mean that income inequality data is largely driven by transient components, such as unemployment, and this is why it doesn’t show up in consumption data. Something I find plausible.

    iv) Consumption inequality rises until about age 50 and then stabilizes

    So, it appears that income inequality probably vastly overstates “economic inequality”, especially for the bottom bracket.

  23. Gravatar of Norman Norman
    29. September 2010 at 11:01

    Norton and Ariely don’t specify the source of their wealth data in the paper. The definition they offer is net worth, which they say is basically the sum of all financial holdings and physical property, less any debt or mortgages. This is the definition they gave people when asking about the survey. The “actual data” they offer is only comparable if they followed suit, in which case I’m really not sure how they got that information. Could have been part of the survey for all we know.

    Speaking of the survey, they also don’t give any information about response rates and selection, and a lot of other information I really hoped would help me interpret the results were not included. So I’m tempted to say it’s not just the first bar that’s meaningless; their entire dataset sounds sketchy.

  24. Gravatar of abb1 abb1
    29. September 2010 at 12:16

    …one more thing, about the money, wealth, and consumption.

    Money is not just some pieces of paper that allow you to consume stuff. Money is power. The Medici family motto: “money to get power, power to guard the money.” Money is the means of control, domination, corruption. And that alone is a damn good reason to prevent excessive concentration of it.

  25. Gravatar of Lord Lord
    29. September 2010 at 12:19

    You really need to do a little research before blasting away. I believe this is based on Fed data which does attempt to value all these different assets, real property, pensions, investable assets, business interests, life insurance. Not that it is always successful. Defined benefit pensions are problematic because vesting is so delayed for example. But this is far closer to the truth than those naive estimates.

  26. Gravatar of Scott Sumner Scott Sumner
    29. September 2010 at 12:48

    statsguy, You said;

    “Risk adjusted investment returns in excess of the inflation rate (whatever that is) are not delayed consumption. They are return on investment – PAYMENT for delaying consumption. As such, they are income. Since the entire post begins with this premise, I’m not sure what to make of the rest of it.”

    This is a common misconception. Suppose the real interest rate is 3.5%, so your investment doubles in 20 years. Also suppose the relative price of blueberries (to other goods) is constant. That means the price of a pound of blueberries in 20 years is only one half a pound bought today. It is the interest rate, not the inflation rate, that correcly shows the relative price of current and future consumption. You are free to call investment gains “income” but then it is a pretty meaningless concept. The value of what you get back in 20 years is just equal to the foregone consumption today (in present value terms, or utils) So calling it “income” makes it seem like those with a stronger preference for future consumption are economically “better off” than someone with the same resources, who prefers to use those resources on current consumption. Someone who eats two pounds of blueberries in 20 years, is eating something of identical value to one pound of blueberries today.

    I’d like to agree with your view that NGDP is progressive, but since I view money as being neutral in the long run, I can’t quite accept that view. I wish it were true. It’s possible that there might be a slight non-neutrality around zero inflation, or below. But that would be an argument for a higher NGDP or inflation target, vs. a lower NGDP of inflation target.

    More to come . . .

  27. Gravatar of abb1 abb1
    29. September 2010 at 13:11

    t is the interest rate, not the inflation rate, that correcly shows the relative price of current and future consumption.

    What’s “the interest rate”? If you lend money to a drug dealer, the interest rate is 50%/week, and if you buy a T-bill it’s about the same as the inflation rate.

  28. Gravatar of Benjamin Cole Benjamin Cole
    29. September 2010 at 13:54

    OT but topical:

    Philly Fed’s Plosser Says Economy May Be Tanking But New Quantitative Easing Won’t Help
    Gregory White | Sep. 29, 2010, 2:42 PM | 839 | 7

    Philadelphia Federal Reserve President Charles Plosser opposes any more asset purchases by the Fed, and also sees the economy getting worse, according to Reuters.

    Plosser doesn’t want the Fed to lose public confidence, and doesn’t actually think quantitative easing II will have much of an impact.

    Plosser said today that there are a “very limited amount of things we can do at this point,” according Zero Hedge.”

    Hoo-boy. This is the Japan Wing on steroids. Plosser is assertively impotent. He doesn’t want the public to lose confidence in the Fed, but he states the Fed is impotent now, in the face of gathering deflation and a possible double-dip.

    Oh, why would I lose confidence in institutions that declare their impotency?

  29. Gravatar of r.d. r.d.
    29. September 2010 at 14:47

    Excellent post.

  30. Gravatar of Matt W. Matt W.
    29. September 2010 at 15:18


    I agree with you about income inequality data. Inequality is necessary in a functioning free market economy, as capital goes to people who produce more.

    But I wonder about the affect of moral hazard on the run-up of inequality during the 00’s. Most of the run-up was from people in the financial sector. In turn, their higher returns were usually not from producing more but from using more leverage. Borrowing short, lending long, and earning big returns as long as the market kept going up.

    A similar thing happened during the last big run-up in income inequality during the 20’s. Banks borrowed short with deposits and invested in irresponsible loans to companies or in equity. After the crash, income inequality came back down, but that has not happened in a big way after the 2008 crash.

    So, my question is, what role does moral hazard play in increasing inequality? Is the implicit government guarantee for bank bonds and money market funds keeping income inequality at an unsustainable level?

  31. Gravatar of Jason Jason
    29. September 2010 at 15:29

    Isn’t consumption proportional to income? That’s at least what I see in most charts I can find online. E.g.

    You have particular examples where, using a particular financial transaction, income is not necessarily proportional to consumption. But the data appears to show the two are pretty well proportional to each other. Is there data that demonstrates this is false?

    Are people frequently living well outside (or inside) their means or undertaking these large financial transactions in such a way that income is not proportional to consumption? The precautionary desired savings rate only varies by about a factor of 2 over all quintiles and it is the lower end who have a higher desired rate that is more likely not to be fulfilled. Everyone above about $60k/year income has pretty much the same rate.

    As long as income is proportional to consumption, income inequality and consumption inequality, looking at the full distribution should come out the same.

    Is there data that shows consumption is not proportional to income?

  32. Gravatar of Tomasz Wegrzanowski Tomasz Wegrzanowski
    29. September 2010 at 15:36

    For every cherry-picked arguments suggesting that inequality is somehow lower than measured, you can find another suggesting that it’s higher than measured.

    If you start from desired self-serving conclusion that inequality is not a big deal, it’s certainly not hard to find some supporting data.

    The only reason you rarely see people concerned about inequality making such excuses is that they don’t have to – raw data is bad enough as it is.

    I’m sure in Sweden people blog explaining how inequality is actually higher than it seems, and those who thing it’s lower than measured just don’t bother writing.

  33. Gravatar of Matt W. Matt W.
    29. September 2010 at 16:01


    That’s pretty lazy intellectually, isn’t it? Every single non-obvious issue has the problem you mention, that there’s “cherry-picked” arguments for both sides.

    In general, the truth is somewhere in between and it’s worth trying to honestly seek out the truth. If his arguments are wrong, let Scott know and post your counter-arguments.

  34. Gravatar of Jim Glass Jim Glass
    29. September 2010 at 16:44

    I wrote some professional (non-academic) articles on this subject a few years back. Some thoughts that quickly pop back into my deteriorating memory:

    Yes, consumption clearly is the superior measure of economic welfare compared to income and wealth. After all, the purpose of income is to be able to consume, and wealth is only the discounted to present value measure of expected future income.

    [] The consumption distribution is much more equal than the income/wealth distributions. E.g., quintile data in the BLS Consumer expenditure survey for 2008 puts the top-to-bottom quintile ratios as…

    Income: 15.4 to 1 ($158,652 to $10,263)
    Expenditures: 4.3 to 1 ($97,003 to $22,304) — a non-trivial difference.

    Fundamental point: Professionals and the well-informed (do Yglesias, Noah, Ariely & Norton qualify?) who know these fact but harp on solely about income/wealth inequality without even mentioning consumption (“admittedly, another measure of welfare, consumption, is not anywhere near so unequal, nevertheless…”) are knowingly dishon^h^h, er, disingenuous.

    [] Over lifetimes, average expenditure inequality is even less than that 4.3x, because everybody’s average average is below the average peak.

    [] Capital gain — the big driver of income inequality at the top — is not economic income. Look at the BEA national income accounts and you won’t see “capital gain” in there. In the early days of the income tax, the US Supreme Court ruled four times(!) that capital gain is not income. E.g.:

    “Enrichment through increase in value of capital investment is not income in any proper meaning of the term.”
    – Eisner v Macomber (1920).

    This was, of course, following the advice of the top economists of the day. Then Congress used its power to redefine a pig as a bird for tax purposes, and re-defined capital gain as income.

    [] Wealth tables rhetorically greatly exaggerate the welfare difference between wealth levels to the naive, because wealth is only the discounted present value of income.

    E.g.: A has wealth of $50k and B has wealth of $250k. Wow, B is *five times richer*. How unequal!

    But using the same 6% long-term risk-free interest rate used by the SS Trustees, the $200k difference in wealth represents a $12k flow of income. If both A & B have $57k of income (such as from a job) to start with, B is only 20% better off, not anything like 5 times so. (The difference between 20% better off and 400% better off is a 20x exaggeration!)

    [] Many people at the very bottom of the income tables are not poor at all: Business owners and investors having a loss year. Young persons in medical/law school, etc. Retirees with savings.

    My mother is an 80+ year old widow living alone with zero income but for Social Security — the very person facing a diet of cat food the left so goes on about. Except for her $500,000-value paid off home, my father’s life savings she inherited (paying no income because of today’ rates, and consumption of savings is not income), Medicare, etc.

    [] “I am pretty sure that a lot of the wealth inequality data is incomplete”. Of course. E.g., Medicare and SS represent a package worth about $400,000 at retirement on average, not counted in wealth (and Medicare isn’t counted in income).

    [] “you’d still get massive inequality in wealth simply because older people have been accumulating it for longer.” And beyond that, in an economy that is steadily getting richer “income inequality by decile” grows as a statistical artifact over time, even if it is divided in an absolutely 100% egalitarian manner.

    Take a colony of termites, each lives 10 years, is equally productive, receives income equal to its production. Initially each produces 1 unit and receives 1, inequality is zero. Then the termites evolve learning, and become 3% more productive each year of life after their first-year productivity of 1, still receiving income equal to what they produce.

    Now after nine years the “highest income decile” of termites receive 30% more income than the bottom decile. Exploiters! Even though if learning stops there all receive exactly the same lifetime income, zero inequality.

    If learning compounds, the top decile will move up from the bottom forever (the rich getting richer and richer, relentless exploiters!) — even though there is zero inequality within generations, with each generation getting exactly what it produces … as egalitarian as termites.

    So the common “decile income inequality” measures are significantly biased to exaggerate inequality.

    [] the best paper I’ve seen on the subject is Heathcote, Perri, and Violante’s “Unequal We Stand”

    A very good popular write-up of this was in the New York Times, of all places. (Even so, Krugman never gave it or the issues mentioned in it even the slightest mention in all his going on and on about income inequality. I must conclude that PK either does not read the Times and is not up on the literature, or is dis… disingenuous.)

    [] Another more recent paper, by Robert Gordon (of the NBER business cycle dating committee and CBO, hardly a right winger). NBER Working Paper No. 15351:

    “The rise in American inequality has been exaggerated both in magnitude and timing … This paper shows that a conceptually consistent measure of this growth gap over 1979 to 2007 is only one-tenth of the conventional measure.

    “Further, the timing of the rise of inequality is often misunderstood. By some measures inequality stopped growing after 2000 and by others inequality has not grown since 1993…”

  35. Gravatar of scott sumner scott sumner
    29. September 2010 at 17:24

    Benjamin, I wonder how much of that is a function of LA. I live in an affluent area of the East Coast, but the culture here doesn’t flaunt wealth as much. A lot of wealthy people in my town drive Camrys, indeed I’d say that is the most popular car by far for Newton millionaires. I suppose LA is more showy.

    I agree that the wealthy have gained much more than the middle class, as I said. I think the poor are also much better off, especially in the rural South where many used to live in shacks.

    Joe, The Heritage piece is interesting, but it also misses some of the other flaws in the data. There are many people who are “poor” in a cultural sense, who don’t show up as poor in US data. They earn more than the poverty line, but struggle to get by in high cost areas like NYC. In contrast, I was probably technically “poor” for 8 years of my adult life, but wasn’t poor in a cultural sense. Poverty isn’t really about lack of “stuff” in American, it’s about the stressful condition of living on the edge, paycheck to paycheck, having trouble paying bills, temporarily being out of work, etc. In any case, its a very complex problem, and the income distribution data greatly oversimplify the problem.

    Morgan, Yes, even after WWII many Americans lacked indoor plumbing.

    Lee Kelly, I’d simply say investment income isn’t “income” as most people understand the term.

    Nuveen, Interesting comments:

    1. I’m not sure what sort of wealth was measured in this particular survey. Because the bottom 40% have essentially nothing, I assume that at least human capital was overlooked.

    2. I agree that lots of Wall Street types make 6 figure incomes right out of their MBA programs. But over the course of their lives their incomes will be very volatile. indeed much more volatile than my income. The only difference is that all their annual incomes will be much higher.

    4. One drawback of consumption inequality is that it doesns’t really show the level of inequality produced by the market economy, rather it shows consumption levels that reflect both earned income and transfer payments.

    3 and 5. I strongly believe the middle class today have higher living standards than when I was young in the 1960s and 70s. The college students I see today seem much more affluent than we were (and I grew up in a relatively affluent midwestern city (Madison). They live in much bigger and nicer houses, take vacations by jet airplane, have lots of fancy electronic gear, nicer cars than we had, etc. College have had to extensively remodel dorms because today’s student simply won’t put up with the spartan dorms we had. I know that’s all very subjective, but I still think it’s true. I think the median income numbers are hurt by stagnation at the lower working class level, where real wages haven’t done well. But there has been huge turnover in those jobs, with immigrants from Latin American taking jobs in areas like meatpacking that Americans no longer do. There’s where real wages haven’t done well. But those Latin Americans are often better off than their parents in Mexico, and the kids of the white meatpackers of the 1960s have gone to college and moved on. So my sense is that median data is very misleading. It’s also distorted by families fragmenting, which leads to many low-paid single young people and old people. Middle age intact families are clearly better off than in the 1960s.

    Paul Zrimsek, Excellent point that I forgot to mention.

    Patrick, Yes, my complaint is that they are viewed as better off than someone with an equal wage income, who was free to do the same thing.

    abb1, You said;

    “Those quintiles are probably households, not individuals, and when you had that very low income of yours, you probably belonged to a relatively well-to-do household. Counting households has to significantly weaken this “life cycle effect”.”

    I never belonged to a “well-to-do” household (unless you mean middle class.) And during all those years I cited I was not a member of any other household. I was a single adult living alone, without support from my parents. The government treated me as a household. I lived on about $3000 a year, which would be like $10,000 a year today. So I know how to scrimp by on a low income. I cut my own hair, wore clothes from Goodwill, ate hot dogs and carrots almost every night, no car, no TV, etc. (I can see people reading this with tears in their eyes.) But seriously, I’m not claiming I know what its like to be poor, because I think poverty cannot be defined solely by income. I knew I’d be middle class by my late 20s.

    2. Investment income should not be taxed at all, check out my post from a few days ago. There is no respectable intellectual argument for taxing investment income. That’s why smart progressives like Yglesias favor a progressive consumption tax.

    It’s a myth that most poor people live in inner cities, they live all over the country. I knew lots of different kinds of people when I was younger. I can’t say whether others had that opportunity.

    JeffreyY, I am not claiming the bottom 40% have 10% of the wealth, I have no idea how much they have. My point is that the public might be closer to the truth than that data, not that the public was precisely right.

    The Slate assertion is almost certainly wrong, I’d guess they confused wealth and income inequality in Sweden.

    Ted, Thanks, that’s very useful data.

    abb1, You said;

    “Money is the means of control, domination, corruption. And that alone is a damn good reason to prevent excessive concentration of it.”

    I’d say the average poor or middle class person is more corrupt than Bill Gates or Warren Buffett. I do think the reverse is a problem, however. I think countries with a lot of corruption tend to be more unequal because of the corruption. But I think the corruption occurs for other reasons.

    Lord, I don’t think I “blasted away” (I said I didn’t know how they collected their data) but any survey that shows 40% of the public having essentially no wealth is not measuring wealth as an economist would measure it.

    More to come . . .

  36. Gravatar of scott sumner scott sumner
    29. September 2010 at 17:50

    abb1, You asked;

    “What’s “the interest rate”? If you lend money to a drug dealer, the interest rate is 50%/week, and if you buy a T-bill it’s about the same as the inflation rate.”

    Yes, people may face different interest rates, just as they may pay different prices for gasoline, depending on where they live. But capital markets tend to equalize rates for a given level of risk. Generally a higher expected rate of return is associated with a higher level of risk.

    Benjamin, That Plosser quotation is pretty frustrating.

    Thanks r.d.

    Matt W. That’s a good question, unfortunately neither I nor anyone else really knows. It’s certainly a problem, and as you may know I’d like to see TBTF ended, and FDIC severely reformed. But in all honesty, I think Wall Street salaries would look absurdly high in even a undistorted economy. My hunch is that when we went from old-line manufacturing like GM to newer companies with ultra-low marginal costs like Microsoft and Google, the productivity of people who are good at allocating capital skyrocketed. I know it doesn’t seem fair to people, but I don’t see any alternative. You just have to hope that they give most of it away to charity.

    Jason, I believe I read that consumption is far more equal than income, but don’t have the data in front of me. It seems unlikely that billionaires spend the same percentage of their incomes as poor people. (Check out Jim Glass’s comment at the end)

    In a total life-cycle sense, wealth should be proportional to consumption plus charity.

    Tomasz, If you’d actually read my post, you would have noticed that I claimed inequality is a problem. I also argued that we shouldn’t use totally meaningless data when making public policy decisions. Some people don’t care whether they’re spreading the truth or lies, as long as it supports their pre-determined political beliefs.

    The “raw data” isn’t good or bad, it’s just meaningless. You can’t measure economic inequality using income data. One might as well use the number of square feet in people’s houses, in which case the policy conclusions would be exactly the opposite of what the left claims. Now the poor in America would be better off than the poor in Europe. Hence they cherry pick phony data to support policies they have already made up their minds about.

    Matt W. Exactly.

    Jim Glass, Great comment. Indeed much better than my post. “Capital gains” are another complete travesty in the tax code. I hope everyone will read your comment, as it shows just how unreliable the income data is. If Robert Gordon says that there has been almost no increase in inequality, that ought to make progressives take a second look at the issue. He is certainly no right-winger.

  37. Gravatar of Philo Philo
    29. September 2010 at 18:25

    “Some income redistribution should occur . . . .” Just within the U.S., or internationally?

  38. Gravatar of abb1 abb1
    30. September 2010 at 00:30

    But capital markets tend to equalize rates for a given level of risk.

    Yes, but if you’re taking risk you’re not trying to preserve your savings against inflation – which was your rationalization for declaring taxing capital gains (and, I imagine, interest and dividend income) unjust – no, if you’re taking risk you’re trying to make income.

    You certainly can find a way to preserve your savings by getting tax-free and (virtually) risk-free interest roughly equal to the inflation rate.

  39. Gravatar of scott sumner scott sumner
    30. September 2010 at 04:50

    Philo, Yes, international as well.

    abb1, Check out the post I link to, there is no justification for taxing capital income regardless of whether there is any inflation or not.

  40. Gravatar of StatsGuy StatsGuy
    30. September 2010 at 05:30

    “I’d like to agree with your view that NGDP is progressive, but since I view money as being neutral in the long run, I can’t quite accept that view. I wish it were true. It’s possible that there might be a slight non-neutrality around zero inflation, or below.”

    First, yes around zero bound. Feudalism was a near zero growth economy, with near finite monetary supply.

    Second, I think you need to rely on one of the stronger versions of Money Neutrality to argue that nominal variables have no impact on the _distribution_ of wealth in the long term. This would depend on asset distribution by wealth tier. (In fact, we supposedly have empirical data that high inflation is worse for the poor due to a tendency to hold more wealth in liquid assets, right?)

    Third, an NGDP target is fundamentally different from an inflation target with regards to money neutrality. In the presence of a fixed inflation target (that the CB actually hits, instead of NOT hitting), creditors have no inflation risk and are only exposed to growth rate risk through default rates. In an NGDP target, creditors face inflation risk to the extent real growth underperforms. Risk is not costless (just look at CDS premia). You could argue that creditors would pass this through to borrowers, but this is not so – if they try to pass it through by withdrawing from credit markets, the central bank will step forward and supply credit since lack of credit will cause the economy to not hit NGDP targets. In essence, the threat of competition from the CB forces participation or loss of real wealth due to inflation.

    That, my friend, is why bondholders hate an NGDP target. Unfortunately, they’re now questioning whether the US can structurally support its debt load at a <2% fixed inflation target, and there's a lot of hedging going on that the US will be forced from that target in the long term precisely because it's trying to hard to hit it in the short term.

  41. Gravatar of ssumner ssumner
    30. September 2010 at 06:20

    Statsguy, To me, it seems like you are confusing money and credit here. The Fed doesn’t control credit, they control money.

    I don’t think I am relying on any “strong form” of money neutrality, just the standard version. I don’t see any reason to abandon long run money neutrality, it is the core concept in macro.

    Since we are implictly comparing inflation and NGDP targets that would produce the same long run inflation rate, superneutrality doesn’t come into the picture. And regular money neutrality is very absolute–no real long term effects at all.

  42. Gravatar of StatsGuy StatsGuy
    30. September 2010 at 08:48

    “The Fed doesn’t control credit, they control money.”

    If that’s true, certain Fed members seem to disagree:

    “The FOMC could provide stimulus by saying that it predicts that low rates are likely to be warranted for an even longer period of time than an “extended period.” This kind of stronger language should lead to a decline in medium-term and long-term interest rates.”

    One can argue the Fed doesn’t control credit directly (although, right now, the Fed is buying tens of billions a _week_ in treasuries as it rolls out of MBS), but their objective is to influence credit. So the Fed seems to disagree with the bright line between money/credit. In fact, Kocherlakota says:

    “In this sense, bank reserves held with the Fed are licenses for banks to create a certain amount of money. By giving out more licenses, the FOMC is allowing banks to create more money.”

    Did he mean credit? Cause he said money.

    On neutrality: I certainly could be wrong, so let’s check. The _distributional_ argument simply requires that IN EQUILIBRIUM in the long term, poor households will hold a relatively greater percentage of wealth in cash than wealthy households. All this requires is that poor households require more physical cash for transactional purposes than wealthy households. This argument does not require any frictional lags on the way to long run equillibrium. It’s an argument about the differential opportunity costs for holding M for different households. Let’s assume long term neutrality holds in aggregate (maybe it does, maybe it doesn’t). Even if it holds in aggregate, where did it require parity in opportunity costs for holding money across rich/poor households? Nor do I see how breaking distributional parity in opportunity costs violates the assumptions of long term neutrality.

    Could you correct the flaws here? Does the regular form of long term money neutrality apply at the individual household level even with heterogenous households? If it does, guess I misread it. I’d thought the point of Friedman’s optimal quantity thoery was to get rid of that opportunity cost.

  43. Gravatar of Philo Philo
    30. September 2010 at 11:08

    Wouldn’t the amount of international income redistribution be so much greater than the domestic amount that the latter would be negligible?

  44. Gravatar of Lorenzo from Oz Lorenzo from Oz
    30. September 2010 at 12:50

    Sources of income inequality (leaving aside data issues).

    (1) Increased higher education — more poor students in their 20s who are high-income professionals in their 40s.

    (2) Increased female participation in paid work — high income women tend to marry high income men, low income women tend to marry low income men (or not marry at all because their pool of available men fail to meet their standards).

    (3) Importing lots of developing world migrants whose skill/human capital levels are below the domestic average.

    (4) Rapid technological change creating new skill premiums

    (5) Highly unequal access to income generated from increased prosperity in developing world: the sort of increased prosperity that has seen massive decreases in global poverty.

    (1) to (3) are all from policies progressive folk support. None of the above strike me as things which we would particularly want to undermine or stop. Even (3) acts to increase the income of poor folk, which is why they do it. If the drivers are understandable and legitimate, the fuss is, exactly?

  45. Gravatar of Floccina Floccina
    30. September 2010 at 16:57

    Yes I notice that when people calculate wealth they leave off the forms of wealth that the bottom people have. (Like accrued SS, tools, cars etc.)

  46. Gravatar of scott sumner scott sumner
    30. September 2010 at 18:25

    Statsguy, I think we are talking past each other. Monetary policy is only effective on real variables (including credit) in the short run. But I thought we were debating a long run issue. Your arguments seem more relevant for a super-neutrality debate, as when high inflation hurts the poor because they hold a larger share of their wealth as cash. But I am assuming a given long run inflation rate, and merely comparing NGDP targeting and inflation targeting that produce the same long run inflation rate.

    Philo, That raises a lot of questions that don’t have easy answers.

    1. What about the ease of redistribution?
    2. Are you making pragmatic arguments for improving policy in an imperfect world? (There’s almost no chance of the public suddenly treating foreigner welfare equally with our own, not matter how desirable that is. I’m just trying to move us toward the Singapore model, Once we get there we can try to convince the public to replace domestic welfare programs with international aid, but it won’t be easy.

    Lorenzo, Good points, and don’t progressives also support #4, at least in most cases?

    Floccina. I agree.

  47. Gravatar of StatsGuy StatsGuy
    1. October 2010 at 06:52

    “merely comparing NGDP targeting and inflation targeting that produce the same long run inflation rate.”

    OK, that helps. To get back to the main point, then, if you have two policies:

    1) Inflation target that generates X% long term inflation rate
    2) NGDP target that generates X% long term inflation rate

    The primary difference (other than zero bound behavior, for instance), is that in an NGDP target regime, bondholders absorb medium term inflation risk resulting from variation in real growth. In an inflation target regime, they generally don’t since the inflation target keeps inflation stable in medium term as well.

    I would also note that it’s a big assumption that NGDP hits X% inflation long term – that implies we know with certainty that the real growth rate long term is NGDP% – X%. In reality, we don’t know this, and so bondholders also absorb greater inflation risk in the long term as well.

  48. Gravatar of ssumner ssumner
    1. October 2010 at 18:27

    Statsguy, Where I disgree is that I don’t think the actual risk is inflation risk, I think it is NGDP risk.

    Suppose you start in an economy where inflation and real growth are both 2%, and assume nominal interest rates are 3.8%. Now assume you lend money at 3.8% for 30 years. Suddenly they begin neoliberal reforms and real growth rises to 8%. (Think India) Nominal interest rates rise to say 8.5%, but inflation stays at 2%. You as a lender have not suffered any real loss if you look at inflation. But if you look at NGDP, you’ve falied to share in your country’s rapid growth. After 30 years your interest payments will be tiny compared to the prevailing wage at the time. That’s an example of why I think it is NGDP risk that really matters. People don’t care about real income, they care about income relative to their neighbors.

  49. Gravatar of Greg Ransom Greg Ransom
    1. October 2010 at 20:53

    Great post.

  50. Gravatar of Robert Wiblin Robert Wiblin
    1. October 2010 at 21:37

    “To avoid inflation risk, you put it all in 10-year indexed TIPS. You would earn $650,000,000 per year in interest. Unfortunately your tax liability would be more than $650,000,000. The government would report your income as about $2 billion.”


  51. Gravatar of Robert Wiblin Robert Wiblin
    1. October 2010 at 21:37

    “To avoid inflation risk, you put it all in 10-year indexed TIPS. You would earn $650,000,000 per year in interest. Unfortunately your tax liability would be more than $650,000,000. The government would report your income as about $2 billion.”


  52. Gravatar of ssumner ssumner
    2. October 2010 at 08:20

    Thanks Greg.

    Robert, Believe it or not the federal government does not just tax the real interest on the TIPS, but also the inflation adjustment (even though it is not paid out as part of the annual interest.) Thus if inflation is say 1.3%, and the real rate is 0.65%, then you are taxed as if you earned 1.95%. If you are in the 33.3% tax pracket, you’d pay 1/3 of the 1.95%, or 0.65% in taxes. In other words your entire real return would be paid in taxes. The same is true of regular bonds, but it seems less noticeable as your inflation compensation is paid out annually as part of the interest payment.

  53. Gravatar of Lorenzo from Oz Lorenzo from Oz
    3. October 2010 at 23:32

    Scott: yes, I guess they mostly do support (4) as well.

  54. Gravatar of Gene Callahan Gene Callahan
    26. October 2010 at 00:43

    “There are plenty of places in the country where you can’t travel and go into people’s houses. You can’t travel to urban ghettos and go people’s houses there, because it’s not safe; in fact, it’s probably very dangerous.”

    I bet you’re quivering in the suburbs just thinking about it! But some of us have actually been in housing projects many times and not been killed!

  55. Gravatar of ssumner ssumner
    26. October 2010 at 06:02

    Gene, That was my thought as well.

  56. Gravatar of JLF’s John Hood: What Shouldn’t Worry Us JLF’s John Hood: What Shouldn’t Worry Us
    6. February 2012 at 03:55

    […] are all very real improvements in the standard of living that are not captured by many of the usual statistics trotted out by the usual suspects to make the usual political points. Furthermore, careful analysis of just the recent income data […]

  57. Gravatar of Cutting Poverty Is Super Easy | MattBruenig | Politics Cutting Poverty Is Super Easy | MattBruenig | Politics
    29. September 2014 at 15:51

    […] of the more glaring versions of this creep is his armchair commenting on wealth inequality. Again and again, he has called wealth inequality data “nonsense on stilts” because it […]

  58. Gravatar of The Longest Post Ever on Piketty | laeffyblog The Longest Post Ever on Piketty | laeffyblog
    25. May 2016 at 01:50

    […] believe the welfare of people has declined since the 1960s, when income inequality was much lower? Look around you. Or, if you’re too young, like me, ask your parents. They will tell you about all the things […]

Leave a Reply