Join 3,417 readers in helping fund MetaFilter (Hide)

Tags:

Piketty findings undercut by errors
May 23, 2014 12:33 PM   Subscribe

Piketty findings undercut by errors -- The Financial Times analyzes the data and calculations that Thomas Piketty used to support his result in his book "Capital in the 21st Century", and discovers that some of the data was "adjusted" and some of the formulas contain fudge-factors. When those things are removed, the calculations show that there hasn't been any significant change in wealth inequality in Europe since 1970.

Quote:
Prof Piketty, 43, provides detailed sourcing for his estimates of wealth inequality in Europe and the US over the past 200 years. In his spreadsheets, however, there are transcription errors from the original sources and incorrect formulas. It also appears that some of the data are cherry-picked or constructed without an original source.

For example, once the FT cleaned up and simplified the data, the European numbers do not show any tendency towards rising wealth inequality after 1970. An independent specialist in measuring inequality shared the FT’s concerns.

Contacted by the FT, Prof Piketty said he had used “a very diverse and heterogeneous set of data sources ... [on which] one needs to make a number of adjustments to the raw data sources.
posted by Chocolate Pickle (128 comments total) 24 users marked this as a favorite

 
I wonder the degree to which this disproves his thesis? It sounds pretty damning to this layperson (isn't rising inequality the whole argument?), but for all I know it's just statistical carping.

Also very curious how long it will take for the FT's claim to be verified or disproven.
posted by Oxydude at 12:37 PM on May 23


Giles' blog post; and Piketty's response. For a non-paywalled version, here's Matt Yglesias' post, which if correct, make Giles' findings rather less than meets the eye. With the exception of the UK, the differences between Piketty's and Giles' data aren't that great.
posted by Cash4Lead at 12:38 PM on May 23 [13 favorites]


Since FT isn't publishing its findings (at least to the public) and there isn't a non-FT-vetted independent evaluation of FT's findings, maybe skepticism of its claims is warranted.
posted by Blazecock Pileon at 12:41 PM on May 23 [36 favorites]


Why are there so many other studies, then, that reach the same conclusions? This is not even remotely the only study to have found major increases in inequality over the time period in question. It's just one of the biggest data sets. Even official reports from government and international bodies have been finding the same result. For many years now.
posted by saulgoodman at 12:42 PM on May 23 [25 favorites]


I'm all for methodological rigor, but the inaccessibility of the data (and the obvious glee Giles is taking in the whole enterprise) makes me leery. How far is "cleaning up and simplifying the data" from the ludicrous 2012 "poll unskewing"?
posted by Bromius at 12:45 PM on May 23 [10 favorites]


Like Voltaire's God, if there had been no flaws in Pilketty's data or argument, it would have been necessary to invent some.
posted by edheil at 12:46 PM on May 23 [40 favorites]


I'm not sure that things remaining merely really unequal instead of having become incredibly unequal is especially comforting anyhow.
posted by dng at 12:46 PM on May 23 [3 favorites]


If the graph in the Yglesias link is correct, it doesn't look like Piketty's own numbers suggest a very significant change since 1970 either. Maybe 4 or 5 percent? Though I don't know his thesis.
posted by Hoopo at 12:48 PM on May 23


Why are there so many other studies, then, that reach the same conclusions? This is not even remotely the only study to have found major increases in inequality over the time period in question. It's just one of the biggest data sets. Even official reports from government and international bodies have been finding the same result. For many years now.

Piketty and his collaborators are pretty much the godfathers of this field of study, however. It's largely due to him that governments began paying attention to the issue at all, and his data and research on the issue is considered to be some of the most rigorous and thorough. If there's big fuck ups in it it's a problem.
posted by Diablevert at 12:52 PM on May 23 [2 favorites]


Annnnnd here it is, right on schedule - the Big Lie.

Obama is a Kenyan, Global Warming is a Hoax, Piketty's numbers are Fudged. Expect to see this FW:FW:FW:FW:FW'd to you and plastered all over Facebook by the crazy republican uncles in your life.
posted by Slap*Happy at 12:55 PM on May 23 [47 favorites]


News flash: people predisposed to hate Piketty's conclusions are convinced. People predisposed to agree with Piketty's conclusions are not convinced. Each will investigate further. No one will become convinced as a result.
posted by norm at 12:57 PM on May 23 [14 favorites]


oh come on, theres truth here and we are going to find it.
posted by MisantropicPainforest at 12:58 PM on May 23 [3 favorites]


Ahh, man, I just got the book :-<
posted by Pendragon at 12:59 PM on May 23


True. My attitude is not as hopeless as my prior comment. I pledge to move my view on political issues as the science comes in. I just despair because I feel like no one else does that. They believe what they believe and conform reality to their views based on apologists and lies about inconvenient truths.
posted by norm at 1:00 PM on May 23 [1 favorite]


The FT article says, "Once the data are cleaned and simplified the European results do not show any tendency towards rising wealth inequality after 1970."

But the FT doesn't give ANY evidence for this claim. Why should we believe it?
posted by MisantropicPainforest at 1:01 PM on May 23 [12 favorites]


That's nothing, you should see the things that my secret data that I'm not going to show you proves! Man, would your mind be blown, let me tell ya.
posted by enn at 1:03 PM on May 23 [7 favorites]


Well, then FT should have no trouble proving its claims with specific examples and providing opportunity for a formal response. In fact, I'd think before running with such a major story, they'd want to make sure they had given the authors the opportunity to clear up any methodological confusion first. Did they?

Also:
For example, once the FT cleaned up and simplified the data, the European numbers do not show any tendency towards rising wealth inequality after 1970.
Exactly what simplifications and clean-up operations are we talking about here? Data's fussy. Small tweaks can skew summary level results a lot. They should publish all their adjustments and give the authors an opportunity to respond with clarification on their own methodology if they're really interested in getting at the truth as opposed to just getting a headline out quickly to create a particular impression of the story in the public imagination (since, as other studies that no one argues with have demonstrated that impressions formed from initial headlines can be very hard to correct when new information arises).
posted by saulgoodman at 1:04 PM on May 23 [7 favorites]


Watch the video with Borat and noted data analyst Chris Giles. Giles doesn't just say that there are some inconsistencies in the data, he comes out and says that Piketty deliberately fudged the numbers in order to support his conclusion. That is a HUGE charge to make and suggests bad faith on the part of the FT.
posted by MisantropicPainforest at 1:08 PM on May 23 [5 favorites]


Every mediocre right-leaning pundit and his mother are trying to discredit and slime Piketty right now. The previous attempts have not been so impressive, and I'm not holding my breath for this one to hold up either. #nicetry
posted by Noisy Pink Bubbles at 1:08 PM on May 23


Really annoying that the source data for this argument is behind the paywalled blog post. The article that is the subject of the FPP is just conclusory. A post that gave analysis instead of conclusions might be an okay post. This post is not that post.
posted by MoonOrb at 1:08 PM on May 23 [3 favorites]


DON'T FORGET POLAND!
posted by srboisvert at 1:09 PM on May 23 [1 favorite]


Even if Piketty's methodology was flawed, there's some credit due for actually publishing all the underlying data; a lot of authors don't do that, and if the guy ends up being pilloried it's going to make other people even more reluctant to show their work.

And FT should at least be as transparent as Piketty was to begin with, in showing their work.
posted by Kadin2048 at 1:10 PM on May 23 [12 favorites]


... calculations show that there hasn't been any significant change in wealth inequality in Europe since 1970.

In 1970 half of Europe was behind the Iron Curtain.
posted by jamjam at 1:10 PM on May 23 [11 favorites]


I understood that one of the things that set Picketty's work apart was the fact that he used data as far back as la belle époque? Just because the data hasn't moved since 1970 doesn't necessarily negate Picketty's overall thesis, surely?
posted by LN at 1:13 PM on May 23


If you go to this Google search result, and click the first article link, you can read the blog post without registering.
posted by smackfu at 1:13 PM on May 23


Really annoying that the source data for this argument is behind the paywalled blog post

Its not behind a paywall.
posted by MisantropicPainforest at 1:14 PM on May 23 [1 favorite]


I'm pretty sure most of the global austerity going on right now was based on a report that had a big ol' error right there in the Excel spreadsheets. That error got a lot of press on the left but was outright ignored by the people who staked their political reputations and futures on disembowelling Western social welfare states for profit.

I wonder if they'll ignore this apparent error?

Who the fuck am I kidding, of course they won't. Even if it turns out not to be true, or wildly distorted, they'll bring this up anytime anyone even breathes a word about inequality.
posted by Happy Dave at 1:14 PM on May 23 [12 favorites]


MisantropicPainforest: It's behind kind-of-a-paywall. FT works so that it has a softer sort of paywall where you can look at like 5 articles a month. To circumvent, Google the article and click on the link.
posted by curuinor at 1:15 PM on May 23


Mistakes in data generally are rampant. Something like source control for data should really be a thing, but it would get used in science as much as source control for actual source. That is, not. The attitude common in good developer shops, where someone will kick your ass if you don't use source control, should be prevalent in science.

I mean, source control for data exists already, but people need to use it. Git is actually fine for the datasets we're talking about here, where people can do things by hand.
posted by curuinor at 1:19 PM on May 23 [1 favorite]


The thing that kills me about how desperately certain parties seem to want to whitewash the bad economic trends with PR spin, etc., is that it won't work. This stuff isn't theoretical for the majority of people anymore.

People in the US at least are living the reality behind these statistics every day and it's reached a point where it's not even possible to ignore it anymore. People can personally see the evidence of these broader economic trends from just about any viewpoint in the country except the box seats. Or more likely, they can even see it from up there, but just don't have any reflexive responses to problems that don't involve trying to manage the perceptions around the problem first, because that's just the way we roll now.
posted by saulgoodman at 1:20 PM on May 23 [12 favorites]


c) Averaging

Prof. Piketty constructs time-series of wealth inequality relative for three European countries: France, Sweden and the UK. He then combines them to obtain a single European estimate. To do so, he uses a simple average. This decision (shown in the screen grab below) is questionable, as it gives every Swedish person roughly seven times the weight of every French or British person. Using an average weighted by population appears more sensible.


But Sweden was less unequal than the UK, so overrrepresenting Sweden would underestimate unequality. Nice try Giles.
posted by MisantropicPainforest at 1:20 PM on May 23 [15 favorites]


Frankly, given the political slant of the Financial Times, I am not at all surprised to see them trying to discredit Piketty. I think it would be a bad mistake to accept their criticism of Piketty's numbers, without other more established experts checking them out.

(And besides, the idea that there hasn't been any significant change in wealth inequality in Europe since 1970 seems to defy all reason.)
posted by markkraft at 1:23 PM on May 23 [2 favorites]


oh come on, theres truth here and we are going to find it.

We will, if by "we" you mean academics. The Financial Times, however, is probably far more interested in the pravda than it is the truth.
posted by ROU_Xenophobe at 1:32 PM on May 23 [1 favorite]


Did Piketty Get His Math Wrong? NYT
posted by MisantropicPainforest at 1:33 PM on May 23 [1 favorite]


We've seen this same tactic used by the right in arguing against theories of climate change.
posted by humanfont at 1:37 PM on May 23 [1 favorite]


If anyone would like access to the paywalled peer reviewed literature on this, please feel free to memail me with the paper you want, an email address I can send a PDF to, and a promise not to digitally distribute that PDF any further. If you would then come back to the thread with either questions or conclusions from what you read that would make me embarrassingly happy, but if not no worries.
posted by Blasdelb at 1:40 PM on May 23 [3 favorites]


Its not behind a paywall.

I wasn't very precise. I should have said that it's annoying that the actual analysis, or data, or whatever it is that supports the conclusions in the single-link FPP is something I can't easily access by clinking the link in the article. I find that really annoying. I used "paywall" as shorthand for this idea and I see how that was maybe not a very good way of saying that.
posted by MoonOrb at 1:41 PM on May 23


The Financial Times would like you to know that the greedy, fat bastards it represents are not, in fact, greedy, fat bastards, but are actually big boned, and have a rare glandular disorder. Now don't you feel ashamed for even bringing it up?!
posted by markkraft at 1:43 PM on May 23 [5 favorites]


Vox.com weighs in, says the actual discrepancy is minor except re: the UK
posted by Aubergine at 1:43 PM on May 23


Just because sometimes people make bad arguments against progressive viewpoints does not mean every argument against a progressive viewpoint is bad. That's a fallacy. Progressives should not try to defend themselves the same way that tea party folks might try to, through accusations of conspiracy and bad faith rather than rigorous examination of the claims.

The truth, no matter how inconvenient to the partisan argument, should matter most.
posted by shivohum at 1:45 PM on May 23 [8 favorites]


It's the Financial Times, of course they're lying, pay no attention to it, pay no attention to the sensible liberals who are now going go and say, weelll, both sides made errors yadda yadda yadda.
posted by MartinWisse at 1:46 PM on May 23 [2 favorites]


yeah and re the vox article, the only discrepancy is inequality in the UK between 1980 and 2000. The FTs graph and data agrees completely with Piketty's from 2000-2010.
posted by MisantropicPainforest at 1:46 PM on May 23


Just because sometimes people make bad arguments against progressive viewpoints does not mean every argument against a progressive viewpoint is bad.

Welcome to the thread. Please read some of the comments we are making. I'm combing this data and am doing nothing of the sort. And neither is anyone here.
posted by MisantropicPainforest at 1:47 PM on May 23 [4 favorites]


There's a link in the link MisantropicPainforest posted up here that looks like it should go to a detailed response from Pikkety (it's labeled "Mr. Piketty’s detailed response") but it only leads to the original article posted in the FPP, which doesn't provide detailed point-by-point responses to the criticisms. Annoying.
posted by saulgoodman at 1:49 PM on May 23


Here is the Picketty response.
posted by shothotbot at 1:51 PM on May 23


shivohum: "Just because sometimes people make bad arguments against progressive viewpoints does not mean every argument against a progressive viewpoint is bad. That's a fallacy. Progressives should not try to defend themselves the same way that tea party folks might try to, through accusations of conspiracy and bad faith rather than rigorous examination of the claims."

That right there is the crux of the issue though. Progressive thought and politics by it's nature seeks to treat things fairly, hedge things you're not sure of, justify justify justify.

But the right learned long ago that if you keep shouting the same thing, over and over again, loud and long, without reference to facts or reality, eventually people will believe you, or simply give up entirely. They don't play by the rules of rational, informed and reasoned debate and they keep winning precisely because they don't.
posted by Happy Dave at 1:53 PM on May 23 [11 favorites]


I'm surprised it took this long to start getting these articles into the MSM. All I've seen until now is stuff like the "Piketty is an unamerican Marxist Communist Fascist" 1-star reviews on amazon that only serve as in-group signaling and have no PR effect.
posted by MillMan at 1:54 PM on May 23 [1 favorite]


Just because sometimes people make bad arguments against progressive viewpoints does not mean every argument against a progressive viewpoint is bad. That's a fallacy. Progressives should not try to defend themselves the same way that tea party folks might try to, through accusations of conspiracy and bad faith rather than rigorous examination of the claims.

The truth, no matter how inconvenient to the partisan argument, should matter most.


The difference here is that the bulk of macroeconomic research was telling the tea party that what they wanted to hear was wrong, and they chose not to listen. That doesn't seem to be the case here, as multiple people in this thread and elsewhere have pointed out.
posted by zombieflanders at 1:57 PM on May 23 [2 favorites]


Here is the Picketty response.

Ach, now I see. It's a pay-wall thing. FT keeps kicking me back to the original article when I try to access the response because it's behind the pay-wall.
posted by saulgoodman at 2:00 PM on May 23 [1 favorite]


It seems to me that Piketty's "errors" that I have seen are largely minor, cherry-picked matters of inevitable interpretation. It doesn't appear to discredit the much larger, more voluminously researched issue at hand.

Piketty maybe made a few minor errors in some charts, but he isn't telling the big lie here. Indeed, there isn't meaningful evidence that he tried to fudge the numbers.

However, when the FT says that the calculations show that there hasn't been any significant change in wealth inequality in Europe since 1970... well, that's a big lie, and it's one that would've been simplistically easy for them to verify, to the point where their inability to do so defies credibility.
posted by markkraft at 2:01 PM on May 23 [5 favorites]


1) seriously people, excel is not for grown ups.
2) No actual economists have chimed in on this yet other than to note its existence. We will see over the weekend.
3) pikitteh.
posted by shothotbot at 2:02 PM on May 23 [5 favorites]


Anybody who has lived through the period they are referencing, has at least one eye, and half a brain already knows that Piketty is waaaaay more right than wrong.

This is the typical attack plan from the right - find (or create) some picayune error or inconsistency and then blow it out of all proportion, hoping that the noise will overtake the whole conversation.
posted by Benny Andajetz at 2:20 PM on May 23 [6 favorites]



1) seriously people, excel is not for grown ups.

Like, maybe, Reinhart and Rogoff?
posted by carping demon at 2:20 PM on May 23


Like, maybe, Reinhart and Rogoff?

Exactly! It's just too fragile. Reproducible Research with human readable data files FTW.
posted by shothotbot at 2:23 PM on May 23 [1 favorite]


Something like source control for data should really be a thing

Data is increasingly captured in structured ways, one of the most common being "laboratory information management systems". This includes stuff like pharmaceutical QA/QC, environmental, pollution and population data. However, it's not at all trivial to do: the SAP implementation for my organization has taken much less time and about an order of magnitude less money compared to getting our LIMS and common GIS platforms running.

To drag this back to the context of the post, a version controlled repository would add modest value to the discussion. It improves data traceability a bit, but that is possible in tabular data in excel with a small bit of effort as well. Version control is most important in contexts where the source data might be changed (by repeat measurements or samples simply being omitted and not reported). In cases like this where the source data is relatively easily traceable back to a public historical record, version control would not solve many of the concerns raised by the FT.

What we need to see is here is the FT's analysis. They need to do is put their cards on the table and make their methods and datasets available.
posted by bonehead at 2:28 PM on May 23 [1 favorite]


Now I have an enlightened reason justifying my failure to read his book.
posted by vorpal bunny at 2:36 PM on May 23 [1 favorite]


excel is not for grown ups

Show me a better interface for (modest amounts of) tabular data.

It's got many faults: its graphics dreadful, its numerics still questionable (though apparently improved), its data formats ugly, the program full of cruft. Still though, nobody has yet come up with a better, clearer way to present and work with tables of numbers than a spreadsheet.

Used with care, excel is not a problem. I'd bet a very large fraction of all publications using tabular data goes through excel at some point.
posted by bonehead at 2:40 PM on May 23 [1 favorite]


I understood that one of the things that set Picketty's work apart was the fact that he used data as far back as la belle époque? Just because the data hasn't moved since 1970 doesn't necessarily negate Picketty's overall thesis, surely?

Picketty's thesis is that capitalism naturally tends toward inequality (since the return on capital tends to exceed the rate of growth of the rest of the economy). His data set allows him to show that return on capital exceeds growth for most of the period from 1800 on.

For Picketty, the destruction of the Great Depression and the two world wars briefly allowed the economy to grow faster than the return on capital, producing lower inequality in the decades after WWII. It is fairly crucial to his thesis and call for action that inequality be increasing again as the world reverts to "normal". The data post-1970 is therefore crucial.

I can't imagine that the modern data is in worse shape than the old data though. As I understand it, Picketty's major contribution was to turn the old historical records into something useful (if still rather uncertain).
posted by sedna17 at 2:45 PM on May 23


I work in economic research for a central bank, I'll offer my educated (read: not authoritative, just educated) and brief opinion:

Macro and historical analysis works with data that is reduced from thousands or millions (or uncountable) dimensions, to a few dimensions. As it turns out, scaling those dimensions down actually retains a lot of information, but it also loses a lot. There also has been lots of variation in what data means, how it is collected, and how is computed. I've done lots of work with US census data from the 1920s, and it's really unclear how certain it is.

The problem is, for economic history, you must go back further.

The further back you go, the murkier the waters of uncertainty become. You have to make choices about how to impute and deal with data issues, and often this is where the perception of bias/priors can creep into argument.

With Reinhart and Rogoff, and now Piketty, I don't think many of my peers or economists will think less of his findings. That's because we already assumed stuff like this existed. So we have noticed a few calculation errors? So what? I guarantee those errors are still small compared to methodological and omitted variable bias. So it remains a reasonable theory, given the massive constraints of economic history and past-data analysis (hell, even with current data we have loads of issues!).

At the same time, as any academic knows, research is half the battle. The other half is marketing. Any scientist who markets with scientific restraint is destined to never make it to the policy world, where they CRAVE certainty. And feigned certainty counts in this game! So to that end, Piketty has been damaged. But those of us in real research see that for the game it is.
posted by jjmoney at 2:51 PM on May 23 [21 favorites]


Excel is a very good tool for producing modelled data which can be easily distributed and quickly modified. What's more it is well known, allowing people to check your figures using automated audit tools. Also it is very powerful (sometimes maybe a bit too powerful) and very flexible.

Even if you have masses of data you might well do your first pass processing in something faster and more powerful and then bring final stage calculations in Excel.

Also you can (and we do) source control your excel models.

Basically my point being, don't discount serious research usung excel just because you can also use it to do your household budget.
posted by Just this guy, y'know at 2:53 PM on May 23 [1 favorite]


I'm reading Capital now and I'm finding it really interesting and readable. I've been impressed by how it seems Piketty goes out of his way to explain exactly where the data is from and why he chose to interpret it a certain way rather than another way. And the data is there for anyone to see. It seems to me he is very transparent.

If people want to argue in good faith that's fine, but saying that a 700 page thesis is completely wrong but we're not going to tell you why seems a little suspect.
posted by maggiemaggie at 2:59 PM on May 23 [3 favorites]


"Why would we be­lieve him? · Be­cause data. He and his col­leagues have in­vest­ed years in build­ing the World Top In­comes Database  —  WTID. Piket­ty is ob­vi­ous­ly a first-class da­ta geek. To the ex­tent that it’s amus­ing: He is unique among hu­mans in re­al­ly lik­ing in­come tax, be­cause income-tax record­s, as a side-effect, feed his database. The WTID has way more sources than tax records though: Pro­bate , university-endowment-fund per­for­mance, lots and lots and lot­s. ¶
His scorn for second-rate da­ta  —  for ex­am­ple Forbes’ lightweight, anec­do­tal Rich List  —  is fierce. And when, at the end of the book, he of­fers pro­pos­als for a new tax regime, he’s at pains to point out that one ben­e­fit would be com­pre­hen­sive da­ta gath­er­ing about wealth-related as­pects of the econ­o­my.
This mat­ter­s, in gen­er­al, be­cause you can’t man­age what you can’t mea­sure."

I am confident he is substantially and technically correct--but lets not overlook that many of the preceding posts on this topic were quite defensive and dismissive because of the source. Sounded a bit like those dismissing the 'Liberal/Leftist Media"
posted by rmhsinc at 2:59 PM on May 23 [2 favorites]


Basically my point being, don't discount serious research usung excel just because you can also use it to do your household budget.

No, Excel is awful for data analysis. The problem is that its interface is exactly the opposite of what it should be. Data analysis is about transformations of data, not the data itself. Excel puts the data front and center, and allows it to be easily changed/replaced/deleted etc. And then, it hides the critical things - the transformations - by either making them invisible (like when you find/replace) or hard to understand (spaghetti of formula dependencies, each formula only visible one at a time, with no evident structure).

People that care about doing data analysis right should be using a script language like R or python. Excel is awful.
posted by Philosopher Dirtbike at 3:03 PM on May 23 [12 favorites]


He is unique among hu­mans in re­al­ly lik­ing in­come tax

I, too, like income tax. I am happy to pay for civilization and at least this one, by being progressive, agrees with my values.
posted by shothotbot at 3:05 PM on May 23 [11 favorites]


It would be extremely naive to read the FT's criticisms as neutral and disinterested in the reception of Piketty's work. Let's not pretend: they have a direct stake in the reception of what Piketty has accomplished, because his thesis challenges prevailing normative attitudes and formal policies that the unbelievable wealth and success of the global and national financial elites is predicated on. That's pretty obvious, but it needs to be said. And the photo that Vox used of Dr. Piketty, where he happens to be scratching his head, is a really scummy and prejudicial choice.

As for their substantive critiques:

For example, once the FT cleaned up and simplified the data, the European numbers do not show any tendency towards rising wealth inequality after 1970.

This is utterly fatuous. "Cleaning" and "simplifying" are not technical terms with precise meanings, so they're actually saying that once they played around with the data, they got different results. If they can show that their clean-up and simplification is not tendentious, publish it. I mean, really. This is beyond weak. It's disingenuous and doesn't deserve to be taken seriously by anyone, especially folks who want to present themselves as non-ideological and faithful to data.

Prof. Piketty constructs time-series of wealth inequality relative for three European countries: France, Sweden and the UK. He then combines them to obtain a single European estimate. To do so, he uses a simple average. This decision (shown in the screen grab below) is questionable, as it gives every Swedish person roughly seven times the weight of every French or British person. Using an average weighted by population appears more sensible.

This criticism seems confused, to me. Giles seems to be arguing that Piketty essentially weighted the data by not weighting the data properly, but as MisanthropicPainforest pointed out, the "more sensible" analysis doesn't contradict the central thesis anyway, so what's the point? It's pretty clearly an insubstantial criticism, and I'm inclined to guess that the idea is not to locate fatal inconsistencies or contradictions but instead to point out unimportant data choices that might be rhetorically effective in conveying the impression that there are substantive issues.

Bottom line, I don't see any real evidence that the thesis is incorrect. If I were going to interpret the significance of the FT critiques, I would say that Giles is trying to break Piketty down one in small corner here and one little bit there, not because Piketty's work isn't correct by some reasonable standard, but because the financial elite are panicky about how correct it is. They are thus interested in weakening the public perception of the correct parts by what amounts to immaterial nit-picking; people will then later say "Well, there were some very serious-person-style concerns with Piketty's analysis, hohoho, let us now move on to a topic besides this silly inequality stuff".
posted by clockzero at 3:07 PM on May 23 [27 favorites]


The only reason that FT and others are even checking the work is because it is in Excel. If he instead released some .csv files and 10K lines of python code, there would be a much smaller pool of people able and willing to look under the hood.
posted by paper chromatographologist at 3:10 PM on May 23 [5 favorites]


Basically my point being, don't discount serious research usung excel just because you can also use it to do your household budget.
We have applied the methodology outlined by McCullough (1998) to assess the reliability of Excel in three areas: estimation, random number generation, and statistical distributions. Excel has been found inadequate in all three areas. We also note that Microsoft did not correct flaws noted by Sawitzki (1994b). We advise that Excel not be used for statistical calculations.
Granted, that was written back in 1999, but what is the likelihood that Microsoft has fixed these issues, when the majority of the work since then has consisted of adding a ribbon UI and implementing a file format that breaks backwards compatibility?
posted by Blazecock Pileon at 3:13 PM on May 23 [1 favorite]


Granted, that was written back in 1999, but what is the likelihood that Microsoft has fixed these issues, when the majority of the work since then has consisted of adding a ribbon UI and implementing a file format that breaks backwards compatibility?

You understand that Microsoft isn't Apple and actually works on things that aren't UIs, right?
posted by Pope Guilty at 3:22 PM on May 23 [1 favorite]


It's already been said here, but my take is that this is an "There's absolutely nothing wrong with the economy" analogue to "There's absolutely nothing wrong with the atmosphere."
posted by ob1quixote at 3:23 PM on May 23 [11 favorites]


clockzero--I happen to agree with almost all of your comments--however--this one is questionable "his thesis challenges prevailing normative attitudes and formal policies that the unbelievable wealth and success of the global and national financial elites is predicated on. I am not sure what those normative attitudes and formal policies are but we all know--and the wealthy know it even better--wealth begets wealth and capital begets capital. Picketty has provided data to support what every wealthy investor knows when they first realize they can not spend fast enough to use their principle (even in down times). The media, the Right, Republicans, Conservatives, Tories all know that once you accumulate a certain amount of wealth its conservation is barely an issue. it is just a matter of how much more you want to accumulate.
posted by rmhsinc at 3:25 PM on May 23


Heh. I just got a new doctor and we digressed pretty wildly in my first visit, ending on a conversation about Piketty after she had said something about favoring a flat tax to reduce inequality (something she walked back by granting it was significantly more complex), and we ended with her promising to read Piketty. Hope this doesn't derail her.

And, given the long waits for appointments, I'm likely to both get and finish the book before I see her again ;)
posted by klangklangston at 3:47 PM on May 23 [1 favorite]


Picketty has essentially stated the obvious. Just like scientists working on global warming have discovered the obvious. Whenever it becomes TRANSPARENTLY obvious that a ruling elite has gained unfair advantage, their "expert gremlins" go to work to undercut truth. Examples? Tobacco; environment; medicine; nutrition; wages differentials; health care etc. etc.
posted by Vibrissae at 3:53 PM on May 23 [7 favorites]


Why are there so many other studies, then, that reach the same conclusions? This is not even remotely the only study to have found major increases in inequality over the time period in question. It's just one of the biggest data sets. Even official reports from government and international bodies have been finding the same result. For many years now.

Wealth inequality, not income inequality. Piketty's doing some fairly original data collection on the former, not the latter.

Also, this is why economists should always produce a data appendix accompanying their work as well as make available the programs and the data on their websites. It's scary to do that, but the reasons it's scary is exactly why it should be the norm -- no one likes to be embarrassed. Piketty has probably spent his adult life on the things that led to this 600 page book. The time he spent accumulating the datasets (which were not available before he and Saez began this enterprise -- or at least that's my understanding) is not fun to contemplate if it all goes up in flames over something trivial like how he chose to weight the data or whatever.
posted by scunning at 4:03 PM on May 23 [1 favorite]


For those wanting a break down, I found this article to at least be doing that. I keep running into FT paywalls and haven't been making a lot of progress otherwise really understanding what is and is not the case.
posted by scunning at 4:12 PM on May 23


Well ignore what I said - Piketty did post his excel files online. That's how FT found them.

I have a sneaking suspicion this won't amount to anything at all, but because it's going to involve reading some actual stuff about data, most people will not read it, and believe whatever they want to believe.
posted by scunning at 4:18 PM on May 23


Picketty has essentially stated the obvious.

What? Piketty's argument is really novel.
posted by MisantropicPainforest at 4:22 PM on May 23


Really? I mean, I understood his book to be the boldest, most comprehensive formulation of his thesis, but I was under the impression that a lot of it had been known for a while — not least because Piketty's got a long career of working on this issue.
posted by klangklangston at 4:31 PM on May 23


Depends what part of his argument we're talking about. the inevitably part (where r > g or whatever) is pretty new from what I know.
posted by MisantropicPainforest at 4:35 PM on May 23


So, I work in academic publishing in this field; it’s something I have a real personal stake in. While research is included in faculty salaries, and reviews aren’t done for pay, there’s a lot of labor that goes into coordinating all that.

Just a few of the things that go into bringing a piece of research from the author to the page are proofreading, copyediting (few authors are really able to do these two steps by themselves); identifying not only the right reviewers, but the reviewers who have the free time to do it at the moment; securing copyrights; maintaining the double-blind process; making sure content editors are paid; making sure everybody gets credit so they’ll work with you again when you need them; and a million other things that I’m not thinking of because this is just off the top of my head. Basically, just making sure that what you’re getting is actual quality merchandise that you can read.

Nobody’s more aware than we are that the publishing world, and academic publishing in particular, is undergoing a sea change, and we’re scrambling to keep up with it. But I’m not sure that anyone really wants the mess that an all-volunteer system would result in. Right now, journal subscriptions directly put a roof over my head and put food in my stomach – that gives me a pretty strong motivation to make sure that everything we publish is in the best damn shape we can humanly make it.

MetaFilter is where I first encountered Harlan Ellison's Pay the Writer, so I'm a little surprised to see so many people complaining about research being behind paywalls. We do our best to give good prices on site licences to libraries, and they bend over backwards to make as much material as possible available to as many people as possible.

I have no problem with the kind of sharing Blasdelb is offering; it sounds ideal to me because of the terms under which he's offering it. I'm just not sure that everybody who's angry that the paywall even exists have considered that there are workers besides the author, who have put in a lot of labor and need to get paid, too.
posted by The Underpants Monster at 4:41 PM on May 23 [1 favorite]


the thing is that Piketty's thesis (that rate of return > growth in income leads to growth of total capital income) doesn't really challenge the conservatarian message. The whole point of "supply side" economics is that accumulating large amounts of capital is good for the economy. The fact that income from capital is growing rapidly is just a sign that the capital has been put to good use.
posted by ennui.bz at 5:28 PM on May 23


It's an interesting world, isn't it? Money-making-money makes more money than
money-rewarding-achievement. A lot more. Exponentially more. Grotesquely more.
What will be the end game to this little scenario? Will there be one? I mean, I don't
believe anyone could've anticipated the ameliorative effects of world-wide depressions
and world wars, for instance. I mean in terms of encouraging or enabling them. Now? What rough beast, etc.? Or are we just in for Brazil on top of Brazil on top of Brazil, etc., all the way down?
posted by Chitownfats at 5:58 PM on May 23 [5 favorites]


Chitownfats--I hope your post is not prophetic but you may be right. Or some cataclysmic leveling/equilibrating event, or apprehension of it, may happen. I hope it can happen through decades of planned policy changes--I guess there is no reason to expect to to change with an imperative for change.
posted by rmhsinc at 6:09 PM on May 23


clockzero--I happen to agree with almost all of your comments--however--this one is questionable

"his thesis challenges prevailing normative attitudes and formal policies that the unbelievable wealth and success of the global and national financial elites is predicated on."

I am not sure what those normative attitudes and formal policies are but we all know--and the wealthy know it even better--wealth begets wealth and capital begets capital. Picketty has provided data to support what every wealthy investor knows when they first realize they can not spend fast enough to use their principle (even in down times). The media, the Right, Republicans, Conservatives, Tories all know that once you accumulate a certain amount of wealth its conservation is barely an issue. it is just a matter of how much more you want to accumulate.


The elite-to-everyone-else discourse about how there's nothing basically wrong with our distributive systems, in the US and elsewhere. Legal trends of tax decreases for the wealthy and corporations. Ideological trends of identifying the redistributive functions of government as both technocratically wrong and morally suspect.

Basically, the open contempt expressed for the poor and working class and middle class by the rich and powerful, the ongoing denial that the "game" is hopelessly stacked against people who are not already wealthy. The make-believe that politicians play when they say that there's an economic recovery. But also the ongoing deregulation and privatization of pretty much everything, which just makes it easier to incorporate all economic activity directly into the regimes of value controlled by the wealthy. That's what I mean.
posted by clockzero at 6:32 PM on May 23 [11 favorites]


Just in case anyone is interested, a json api of Piketty's datasets
posted by maggiemaggie at 12:05 AM on May 24 [1 favorite]


the thing is that Piketty's thesis (that rate of return > growth in income leads to growth of total capital income) doesn't really challenge the conservatarian message. The whole point of "supply side" economics is that accumulating large amounts of capital is good for the economy. The fact that income from capital is growing rapidly is just a sign that the capital has been put to good use.

That rather depends on just how cynically dishonest you think the language of "a rising tide lifts all boats" is. In any case, I think you'd be hard pressed to find a conservative politician who would openly endorse a vision of an economic system in which wealth tends inexorably to accumulate in the hands of a self-replicating elite and in which the have-nots face increasingly higher obstacles in any attempt to improve their material conditions. Again, they may or may not be cynically dishonest when they talk about fostering the "opportunity society" and so forth, but Piketty is undoubtedly a challenge to the explicit political "message" that conservative politicians send out.
posted by yoink at 6:50 AM on May 24 [2 favorites]


"I pledge to move my view on political issues as the science comes in. I just despair because I feel like no one else does that." -Norm

Recent NYT article on that-
posted by Mr. Crowley at 7:33 AM on May 24 [1 favorite]


What I found most interesting in the FT article were graphs of UK wealth inequality dating back to 1810. Good times peaked around WW1, after which trends began to disfavor the rich for most of the 20th century. Damn Labour governments! Damn Depression! Damn wars!

The FT has a comparable chart for the US which, alas, I cannot find on line, but which shows a similar trend, albeit with a less dramatic post WW1 slide. Worth looking up if you're into bigger history.

Worth pondering what happened in the 1970s....
posted by IndigoJones at 8:47 AM on May 24


re: obviousness...
viz. The Daily Show: Timothy Geithner Extended Interview
cf. How to Build a Better Bailout

also fwiw...
-The Triumph of the Rentier?
-Thomas Piketty and the End of Our Peaceful Coexistence With Inequality: "Economic disparities aren't new. But from Piketty to the pope, talking about them is."

oh and btw...
-Studying the Rich: Thomas Piketty and his Critics
-The Poverty of Right-Wing Piketty Criticism
-What Piketty's Conservative Critics Get Wrong
-What Piketty's Neoliberal Critics Get Wrong
-What Larry Summers Gets Wrong On Piketty's 'Capital'
-Larry Summers on Piketty (Mouthbreathing Machiavellis Dream Of A Silicon Reich)*
-pmarca boils it down: "Does Piketty think we'll succeed (r > g due to productivity boom) or fail (r > g due to new tech oligarchy)?"

more broadly...
-The problem of distribution:
In a comment here, Nuno Ornelas Martins says: "the central problem of economics is the distribution of the surplus rather than the allocation of scarce resources."

This, of course, flatly contradicts the standard view that scarcity is the problem of economics. However, in one context at least, he is right. J.W. Mason points out that, in the US, companies have (net) long ceased to raise money from financial markets. A similar thing is true in the UK; for years, companies' retained profits have exceeded capital spending - something which the OBR expects to continue.

It used to be the case that finance was scarce. The purpose of the joint-stock company was to overcome this problem, by allowing firms to raise small amounts of money from many dispersed investors.

But this is no longer so. From the point of view of existing companies, finance is abundant. Their problem is no longer one of the scarcity of finance, but instead of the distribution of surplus...
-Begging the inequality question:
Tim Worstall says Labour's idea of predistribution is inefficient because it is "cocking up markets". However, Labour defends predistribution - for example the freeze on energy prices - by claiming that markets are broken already. If they're right, such a freeze might be efficient in second-best terms.

What Tim's doing here is - literally - begging the question. He's assuming what must be proved - that we have an tolerably well-working market, intervention in which would be inefficient.
-Keeping the rich rich: "The bailouts, though, are only a small part of the story here. There are countless other ways in which the state helps keep the rich rich..."

back to the immediate topic at hand...
-Housing vs financial wealth: "should we conflate housing and financial wealth as Chris and Thomas both do? Perhaps not, because housing wealth might not have as much 'wealthiness' as financial wealth, in four senses:"
  1. Is housing net wealth at all?
  2. Does housing do as good a job as financial wealth in cushioning us from shocks?
  3. Does housing wealth give us as much freedom as financial wealth?
  4. Does housing give us political power?
-What can governments do?
I suspect there has long been an excessive division on the left, the latter-day manifestation of which is between Blairites and the far left. On the one hand are those who are too happy to compromise with capital, and on the other are those who are over-optimistic about the prospects for radical change. There is, though, a third way: we can recognise that governments can only do a little to help the worst off, whilst at the same time deploring this fact and regretting the power that capital has.
-Ten ways to fight inequality without Piketty's wealth tax:
  1. Open the borders
  2. Get rid of some intellectual property protections
  3. Cut taxes
  4. Crack down on offshore tax havens
  5. Open up the city
  6. Wait for China’s labor costs to catch up
  7. Unleash antitrust
  8. Punish the financial sector
  9. Create more sovereign-wealth funds
  10. Massive social upheaval and bloody conflict
-4 ways to stop the U.S. from becoming a Piketty-style oligarchy:
  1. Increase growth (first, by letting more people live here).
  2. Less rent-seeking and more bargaining power for workers. And more houses.
  3. It's time to tax inequality.
  4. Capital for the rest of us: the 99 percent need to own more stock.
that's all?
posted by kliuless at 9:41 AM on May 24 [12 favorites]


By the way, the top Google results for "Piketty" are now all about his math being wrong, his data being riddled with errors, etc. This is not about proving or disproving anything, it's about muddying the waters.
posted by clockzero at 11:35 AM on May 24 [13 favorites]


I don't quite understand this discussion.

If you compare a rich person from a hundred years ago (say, the average person from the top 20%, so at the 90th percentile) to a poor person from today (say, the average person from the bottom 20%), it turns out that today's poor person leads a much more affluent life. They have better options in food, transportation and education. They have their TVs and a world of information at their fingertips. They have access to health care that was just about unimaginable 100 years ago. (Well, perhaps not in a third-world country such as the US, but definitely in the socialist hell that is France, where Piketty is from...)

So why does it matter if the distribution of wealth is not equal? Isn't the real question the affluence of the bottom (or possibly the medium) X%?
posted by sour cream at 3:14 PM on May 24


So why does it matter if the distribution of wealth is not equal? Isn't the real question the affluence of the bottom (or possibly the medium) X%?

No. Money is power. Having vastly unequal wealth means having unequal power; in particular, an asymmetric situation where a small number of people have power over everyone else. This is incompatible with democracy. Are you really going to put TV or the internet over self-determination?
posted by Philosopher Dirtbike at 3:40 PM on May 24 [6 favorites]


sour cream,

The distribution of wealth matters because wealth is power. The distribution of power is the fundamental question of politics--who gets what when how.
posted by MisantropicPainforest at 3:48 PM on May 24


I don't quite understand this discussion.

If you compare a rich person from a hundred years ago (say, the average person from the top 20%, so at the 90th percentile) to a poor person from today (say, the average person from the bottom 20%), it turns out that today's poor person leads a much more affluent life. They have better options in food, transportation and education. They have their TVs and a world of information at their fingertips. They have access to health care that was just about unimaginable 100 years ago. (Well, perhaps not in a third-world country such as the US, but definitely in the socialist hell that is France, where Piketty is from...)

So why does it matter if the distribution of wealth is not equal?


Ah. So it's not that you don't understand, per se, you just think that we're not talking about what's really important.

And apparently the really important thing is that, hey, poor people have it pretty good these days, what with the refrigerators and the in-house toilets and the computers and whatnot.

Personally, I think it would be a derail to address such a fatuous, insulting, and absurd argument. Suffice to say, it's totally irrelevant to this discussion, which is about a critique of the work of an economist named Thomas Piketty.
posted by clockzero at 4:03 PM on May 24 [4 favorites]


Paul Krugman weighs in:
Giles finds a few clear errors, although they don’t seem to matter much; more important, he questions some of the assumptions and imputations Piketty uses to deal with gaps in the data and the way he switches sources. Neil Irwin and Justin Wolfers have good discussions of the complaints; Piketty will have to answer these questions in detail, and we’ll see how well he does it.

... [all the other evidence is that wealth in the US in particular has become markedly more concentrated in the last 40 years] ...

None of this absolves Piketty from the need to respond to each of the individual questions. But anyone imagining that the whole notion of rising wealth inequality has been refuted is almost surely going to be disappointed.
posted by shothotbot at 6:50 PM on May 24 [3 favorites]


Here is another review of the review, from The Economist blogger Ryan Avent
The fourth question is whether the book's conclusions are called into question by Mr Giles's analysis. If the work that has been presented by Mr Giles represents the full extent of the problems, then the answer is a definitive no, for three reasons. First, the book rests on much more than wealth-inequality figures. Second, the differences in the wealth-inequality figures are, with the exception of Britain, too minor to alter the picture. And third, as Mr Piketty notes in his response, Chapter 10 is not the only analysis of wealth inequality out there, and forthcoming work by other economists (some conclusions of which can be seen here) suggests that Mr Piketty's figures actually understate the true extent of growth in the concentration of wealth.
Who then goes on to cheerfully argue that the mistakes being minor and on a secondary point of his book (the bulk of the historical digging is about income) will have no bearing on the transcription errors being waived like a bloody shirt to cow critics and "refute" the entire work.
posted by shothotbot at 8:03 PM on May 24 [4 favorites]


So why does it matter if the distribution of wealth is not equal? Isn't the real question the affluence of the bottom (or possibly the medium) X%?

This is an excellent and highly relevant question, contrary to what some others on this thread might arrogantly pronounce. Some answers:

-unequal wealth in modern democracies leads to unequal political power, which is contrary to the spirit of a democracy and leads to social instability
-social science research shows well-being and measures of health are often linked to relative wealth rather than absolute living standards
-the inequalities are often gained through grossly unfair governmental favoritism and market failures: so the inequality is unjust and that galls people
-the inequalities have a tendency to build on each other over time, leading to a "caste system" over time -- that could again lead to social instability
-inequality means that a lot of human potential goes to waste as people who could be productive are not given as many opportunities as they could be and instead sit on the sidelines or even go to jail. The best people do not necessarily go to the positions where they could do the most good.
posted by shivohum at 9:09 PM on May 24 [5 favorites]


"So why does it matter if the distribution of wealth is not equal? Isn't the real question the affluence of the bottom (or possibly the medium) X%?"

This is an excellent and highly relevant question, contrary to what some others on this thread might arrogantly pronounce. Some answers:

-unequal wealth in modern democracies leads to unequal political power, which is contrary to the spirit of a democracy and leads to social instability
-social science research shows well-being and measures of health are often linked to relative wealth rather than absolute living standards
-the inequalities are often gained through grossly unfair governmental favoritism and market failures: so the inequality is unjust and that galls people
-the inequalities have a tendency to build on each other over time, leading to a "caste system" over time -- that could again lead to social instability
-inequality means that a lot of human potential goes to waste as people who could be productive are not given as many opportunities as they could be and instead sit on the sidelines or even go to jail. The best people do not necessarily go to the positions where they could do the most good.


I wouldn't dispute the empirical validity of these consequences. But is that why wealth inequality matters, exactly? Those are all bad outcomes, in other words, but they're perhaps not what makes wealth inequality important. Putting it somewhat simply, I'd say what makes it deserve our attention is that we can right a wrong that doesn't need to happen, and the fact that the people enacting that wrong to their own narrow benefit are trying to deny the existence of the problem even as their efforts cause greater pain and suffering for the rest of us. I feel income inequality is wrong because it reifies privilege, it makes it official that a very few people are just categorically more important than billions of others. Its own ideology is a big part of what's wrong with it.
posted by clockzero at 12:39 AM on May 25 [2 favorites]


I agree with shivohum, I think it's actually crucially important to discuss why relative poverty is important. The entire conservative schtick in North America is based on the idea that the word poverty should only mean absolute poverty. It's an idea that a lot of regular people have bought into.

The main thing is that there is no such thing as democracy in a country with high inequality. So what if you can vote? The only parties that can actually run a campaign, and actually get anything accomplished if they are elected, are going to be the ones with huge financial backing and the support of the wealthiest. Which means that if you have anything less than a seven figure salary, your interests will be consistently steamrolled despite being in the vast majority.

At the heart of it, inequality is simply unjust. Exorbitant executive pay does not make any sense. We live in this bizarre world where people think fast food workers are greedy for wanting high enough wages to pay for rent and gas, but think it's fair that executives collude to give each each other seven and eight figure salaries. The resulting power inequality will only escalate if you don't actively work against income inequality.

I think Piketty is hugely important. But I would tweak his policy recommendations a bit. For me, one of the big takeaway points from Piketty is that wages/salary income is very different from capital gains income. So to my mind, imposing a punitive income tax just mixes up these two very different forms of income. I would put a punitive tax on high salaries. Then, completely separate from your total income, I would put a single flat tax on all capital gains, and I would adjust this rate based on GDP versus average income growth. So when GDP grows faster than average incomes, tax capital gains at a higher rate to balance it out and slow down the compounding effect. Incomes growing faster? Ease back on the capital gains tax. This means that if your income is based solely on capital gains, it is actually in your immediate interest that average wages and salaries are rising.
posted by molecicco at 12:42 AM on May 25 [3 favorites]


-unequal wealth in modern democracies leads to unequal political power, which is contrary to the spirit of a democracy and leads to social instability

Does it? So why are societies in which the wealth and the power is concentrated in the hands of a few individuals (Saudi Arabia, Gulf states ...) much more stable than those where most people are equally poor (Gaza, Egypt, just to stick to the same region/culture)?
In any case, it seems that most developed countries are a long shot away from any uprisings, at least as most people have their flat-screen TVs and a sufficient supply of beer.

-social science research shows well-being and measures of health are often linked to relative wealth rather than absolute living standards

Probably because we measure our self-worth and well-being by comparing ourselves to others.
And yet the life expectancy of a poor person today is higher than that of a rich person 150 years ago, so it's clearly not just relative wealth.

-the inequalities are often gained through grossly unfair governmental favoritism and market failures: so the inequality is unjust and that galls people

But that means that the inequality is a symptom, not a cause. And if that is true, then redistribution of wealth will not cure the cause. (If it can be cured at all... Maybe markets are always imperfect and there will always be some favoritism, could be just human nature.)

-inequality means that a lot of human potential goes to waste as people who could be productive are not given as many opportunities as they could be and instead sit on the sidelines or even go to jail. The best people do not necessarily go to the positions where they could do the most good.

And what solutions are there to solve this problem?
Making everyone equally rich/poor has been tried before e.g. in the Soviet Union and its satellite states and didn't work out very well. It turned out that the political power still went to just a few individuals and it also wasn't very productive. On the contrary. It can be argued that a little unequality is necessary to make people productive.

Anyway, we already have a mechanism in place that redistributes money from the wealthy to the needy. It's called taxes. It might not work very well in the US or third-world countries, but it does work quite well in most Western European countries, for example. Also, it is certainly true that money buys power, and that requires checks and balances. Again, this might not work very well in the US, where money seems to be the deciding factor in elections, but appears to work reasonably well in most other Western countries (don't know enough about democracies in Asia, so can't include them in the comparison).
posted by sour cream at 1:09 AM on May 25


Sour cream, the options are not absolute equality versus extreme inequality. That is the worst kind of straw man argument. Also, your stable but unequal examples are not democratic, which kind of proves or point for us: democracy cannot in any way exist when there is extreme inequality.
posted by molecicco at 2:47 AM on May 25


(A bit late to the party, but posting anyway)

The comments by Krugman and the Economist blogger quoted above are as good a summary as any (though I notice Giles's obvious bad faith is politely overlooked): the only point that looks serious at first glance is the discrepancy in the UK series after 1980. As Krugman explains, it is a question of how Piketty compensated for the gaps and discontinuities between the various data sources available.

So the problem is this: in Giles's UK wealth inequality graph, there is a discrepancy between his forest of red lines (the "raw data") and Piketty's series (the blue line), starting around 1980. The reason for this is that, as Giles himself is at pains to point out, the various data sources covering various time periods cannot just be stuck end-to-end, so Piketty had to harmonise them and fill in the gaps somehow ("imputation"). Unfortunately, Piketty's appendix (pdf, p. 56, follow link to spreadsheet) does not go into much detail.(*)

It's possible that a closer reading of the Atkinson and Waldenström papers he cites would be enough to reconstruct his methods. Without going that far, though, I feel safe in dismissing Giles's point, for the following reasons.

First, obviously, considering that this data series is not central to the point of the book (though of course of great interest to anyone nervous about wealth concentration in Britain), and that the rest of Piketty's work, including his openness in publishing his data sets, points to exemplary professionalism, I have no trouble believing, until proof of the contrary, that his imputed numbers are no less (and no more) justifiable than those any other economist could come up with in good faith. Note that Giles does not give us any more detail about how he gets to his series.

Second, even in Giles's graph, Piketty's series joins up nicely with the "raw" Inland Revenue statistics by 2010, showing 70.5% of wealth in the hands of the top 10% of individuals (Tables 13.8 here; you have to sum the top three classes to get close to 10% of the whole population). The number from the Office of National Statistics household wealth survey is 44%. To agree with Giles that the latter data point is more convincing, we would have to believe that the extent to which the IRS numbers overestimate concentration (because they do not count estates too small to go through probate), is likely to be greater than the extent to which the household survey underreports large estates. This strikes me as highly unlikely.(**)

(*) In this case, some annotated scripts with a record of his computations and judgment calls, rather than an Excel spreadsheet with the end result and a laconic footnote, would actually have been more helpful, pace skeptics above. I agree that computer code may be a barrier to interested laypeople going further (while disputing that Excel is not), but on the other hand I can't say I really see polemicists on daily-newspaper deadlines as a key constituency for open research.

(**) There are a few reasons for this, depending on the survey: questionnaires often code wealth in bands topping out at ‘X and above’; the very wealthy are a tiny group that consequently tends to be underrepresented in random samples; perhaps most importantly, people tend to underreport their total wealth in surveys, especially the better-off.

The shortcomings of household wealth surveys, in fact (underreporting at the top + absence of retrospective data), are why Piketty and his team use tax data in preference to surveys, whenever available. (The fact that he appears not to have succeeded in obtaining tax data for the US series in chapter 10 is the second point that Krugman agrees needs clarification.)

Finally, even Piketty's series is likely underestimating wealth concentration, for the obvious reason that very large estates expend great effort on obfuscating ownership through offshore trusts, non-profit foundations, and other schemes before they are assessed for inheritance tax. Which again, the beneficiaries are unlikely to report to a passing survey taker.
posted by ormon nekas at 3:41 AM on May 25 [4 favorites]


Even later to the party, but likewise. Invigorating discussion - this might prompt more attention to Piketty rather than less, which would be a good thing.

I'm not sure if I've understood this correctly, but it seems that we don't have a good idea of the wealth (as opposed to income) distribution in developed Western economies. I'm assuming that's wrong, because if it isn't, it would be a remarkable confirmation of how deeply politically structured the economics discourse is. Can it be that perhaps the most important piece of information from a leftist point of view isn't systematically collected and used? The very fact that this argument is on the level of cobbled-together data sources is a pretty damning indictment of the power/knowledge regime that underpins political argumentation.

Politically, i'd agree that, intrinsically, inequality is wrong - although much depends on the means and extent to which you go about responding to it. Instrumentally, I'd have thought that the greater the inequality the less the marginal utility of that last buck placed on top of some rich guy's teetering pile, and therefore more equality gives more utility, but I'm not an economist and I assume people have looked at this. I believe I've seen evidence correlating improved health with greater equality, but i couldn't find it again.
posted by YouRebelScum at 5:44 AM on May 25


well - if you look at the data that gets collected that's changed quite a bit over the years. WRT to the top of the distribution, ormon nekas point about the sample is spot on. When the polling was constructed no one really considered if knowing the wealth of the 99.95% of the distribution was going to be important. A lot of what was done here was imputing that number based on what we do know about other points in the distribution.

The whole idea of studying wealth inequality is quite new. Piketty and Saez are both in their early 40's and the sorta created the area.
posted by JPD at 6:28 AM on May 25 [1 favorite]


-unequal wealth in modern democracies leads to unequal political power, which is contrary to the spirit of a democracy and leads to social instability

Does it? So why are societies in which the wealth and the power is concentrated in the hands of a few individuals (Saudi Arabia, Gulf states ...) much more stable than those where most people are equally poor (Gaza, Egypt, just to stick to the same region/culture)?


Which are those are modern democracies?

And what solutions are there to solve this problem?...Anyway, we already have a mechanism in place that redistributes money from the wealthy to the needy. It's called taxes.

Yeah, which is exactly what Piketty is proscribing. What did you think we were saying should happen?
posted by MisantropicPainforest at 6:30 AM on May 25 [1 favorite]


Also, another reason why extreme inequality is injust is because the poor and middle class aren't being justly rewarded for what they produce. Wages in the US have stagnated but the incomes at the top have skyrocketed
posted by MisantropicPainforest at 6:32 AM on May 25 [2 favorites]


We know about income distribution because most income needs to be declared to tax or regulatory authorities somehow. For wealth, that is the case only in countries with a wealth tax. Even that isn't foolproof. France has a wealth tax and so we have a fair idea of the wealth distribution in France, but even that is not entirely accurate because part of the wealth is held in undeclared offshore accounts (to say nothing of more sophisticated structures) – even a socialist finance minister, no billionaire by any stretch, has been found red-handed hiding money in Switzerland.

Piketty argues that even a tiny (say, 0.1%) wealth tax would be a net gain for democracy if it could be enforced worldwide (with compulsory automatic sharing of customer information by banks), because we would at least know who owns what where, and we'd be able to set policies in consequence.

Right now I am in Japan where the government deficit is being fought by raising the consumption tax, which all economists know hits poorer people hardest. And yet I've heard it argued in academic seminars that tallying wealth, let alone taxing it, would be unpopular because household wealth is a private matter and people are attached to their privacy. This kind of argument will hold until there is broad agreement that inequality is a matter for public concern.

On preview: the big scoop is indeed how much of the national wealth the top 1%, 0.1%, 0.01% actually hold in modern economies. Our political systems are completely unprepared for this.
posted by ormon nekas at 6:42 AM on May 25


Anyway, we already have a mechanism in place that redistributes money from the wealthy to the needy. It's called taxes.

Congratulations; you agree entirely with Piketty as to the best solution to the problem.
posted by yoink at 6:46 AM on May 25 [3 favorites]


YouRebelScum: there's a lot of literature on the consequences of inequality. Wilkinson and Pickett's The Spirit Level (review) is a popular book by two epidemiologists that argues for the advantages of more equal societies. Maybe that's what you remember.
posted by ormon nekas at 6:52 AM on May 25 [2 favorites]


I guess what I'm saying is that value statements about wealth distribution have been politicised and of interest since, well, politics. That this is a new idea for economics I find so extraordinary as to be unbelievable. I must be missing something. Surely?

Ormon nekas, brilliant, thanks for the ref. Re you're earlier comment, it's the data/policy cycle that I find so interesting: the data seems based on tax documentation, which is founded on a policy regime, which is reliant on a discourse, which is shaped in terms of the existing data, etc.. Basically the Foucualt power/knowledge critique. It's to Piketty's credit that he's forging this field.
posted by YouRebelScum at 7:03 AM on May 25 [1 favorite]


There seems to have been a kind of hiatus in mainstream economics' interest in wealth distribution starting around 1945. Part of it was the (now disproven) optimism of the Kuznets curve, part of it the use of mathematical models that assimilated all actors to a single "representative" firm/consumer/worker, some part of it may well have been politics.
posted by ormon nekas at 7:17 AM on May 25 [2 favorites]


Anyway, we already have a mechanism in place that redistributes money from the wealthy to the needy. It's called taxes.

Congratulations; you agree entirely with Piketty as to the best solution to the problem.


Never said I disagree with that conclusion. It seems to be pretty much in place already in many European countries.
What I don't really see is the social upheaval that is supposed to be a logical consequence of inequality, at least in modern (or unmodern) democracies. That might have been the case in the French revolution, but these days, flatscreen TVs and the internet keep the masses much too occupied, and with rising affluence, it will take much larger inequality for any protests to occur.

Perhaps there is a critical level of (median) affluence beyond which serious protests simply do not occur and maybe we (in the developed parts of the world) have already reached that level of affluence.
posted by sour cream at 8:08 AM on May 25


What I don't really see is the social upheaval that is supposed to be a logical consequence of inequality

But it's not a situation where there's either barricades on the streets or everyone's happy. Piketty isn't saying that bloody revolution is imminent and that without sudden and drastic redistributive measures we're doomed. We can talk meaningfully about increases in social disaffection and discontent without having to invoke the specter of a mass uprising.

these days, flatscreen TVs and the internet keep the masses much too occupied


People have been making some version of that claim since it was panem et circenses that was supposed to be keeping the masses quiescent. It's not an interesting or new aspect of the social equation. Historically, political conflict and upheaval does not track nicely with economic well-being.
posted by yoink at 8:18 AM on May 25 [1 favorite]


Well, panem et circensis (along with massive wealth inequality) allowed the Republic to lapse into Empire, and as has been noted by historians of Late Antiquity, the Roman civilization then delegated itself to death, one mercenary warlord at a time.
posted by Slap*Happy at 6:53 PM on May 25


Thomas Piketty accuses Financial Times of dishonest criticism: Author of bestselling economics book says paper is ridiculous to suggest his thesis on rising inequality is incorrect (The Guardian)
posted by ormon nekas at 6:48 AM on May 26 [2 favorites]


Well, panem et circensis (along with massive wealth inequality) allowed the Republic to lapse into Empire, and as has been noted by historians of Late Antiquity, the Roman civilization then delegated itself to death, one mercenary warlord at a time.
210 Reasons for the decline of the Roman Empire
posted by Blasdelb at 7:29 AM on May 26 [3 favorites]


Piketty's data sets are here, btw.

Concerning the bunfight above about tools: as far as I can tell from a quick skim of his technical appendix and sampling a few files, he does appear to have done everything in excel. But, horses for courses, the data values are what mattered to him and his math doesn't appear to be very complicated.
posted by bonehead at 8:22 AM on May 26


Paul Mason weighs in: Piketty asks the question that mainstream economics doesn't want to answer: do we want a society based on work and ingenuity or on rent?
posted by YouRebelScum at 12:26 AM on May 27


sour cream: You have thrice mentioned flat screen TVs in this thread as though they constitute some kind of argument. Do you have anything else to say?

Personally, I think the fight is over. Democracy lost, we lost, money won. The only thing to decide now is what to do about it. I think the only real answer in the centuries to come will be to create a parallel economy; to just withdraw from the system that is so obviously stacked against us and create our own.
posted by Steely-eyed Missile Man at 10:57 AM on May 27


Political economy is political.
posted by shothotbot at 8:23 AM on May 28


Piketty has formally rebutted:

http://piketty.pse.ens.fr/files/capital21c/en/Piketty2014TechnicalAppendixResponsetoFT.pdf
posted by MisantropicPainforest at 9:45 AM on May 29 [4 favorites]


Choice quotes:

On Sweden:

"The FT argues however that my choice of years from raw data sources is not entirely
clear. For instance, they point out that raw data for year "1908" for year "1910", year
"1935" for year "1930", and so on. These issues are already explained in the book
and in the technical appendix, but they probably need to be clarified. Generally
speaking, when I present series on wealth-income ratios and wealth inequality (and
also for some figures on income inequality), I usually choose to present decennial
averages rather than yearly series. This is because wealth series often display a lot
of short-run volatility (in particular due to sharp movements in asset prices)."

"The FT also suggests that I made a transcription error by using the estimate for 1908
for the top 1% wealth share (namely, 53.8% of total wealth) for year 1920 (instead of
the correct raw estimate for that year, namely 51.5% of total wealth). In fact, this
adjustment was intended to correct for the fact that there is a break in a data sources
in 1908: pre-1908 series use estate tax data, while post-1908 use wealth tax data,
resulting into somewhat lower top wealth"
posted by MisantropicPainforest at 9:47 AM on May 29 [2 favorites]


“Not Another Piketty Symposium”, Mike Beggs, Jacobin, 28 May 2014

“Piketty’s Fair-Weather Friends”, Seth Ackerman, Jacobin, 29 May 2014
posted by ob1quixote at 10:46 AM on May 29


Piketty Answers His Critics at the Financial Times
posted by scody at 3:41 PM on May 29 [2 favorites]


NYT blurb on the response. More importantly: Bloomberg Businessweek gets Pikettymania!
posted by whir at 5:42 PM on May 29


FT pretty clearly overreached in their rush to proclaim that the left's latest idol has feet of clay. I think that overall Piketty's going to come out of this with his reputation enhanced, if anything.
posted by yoink at 5:55 PM on May 29


I am in favor of wealth taxation, international cooperation and automatic exchange of bank information [because]...this would be a way to develop more financial transparency and more reliable sources of information on wealth dynamics (even if the tax was charged at very low rates, which everybody could agree with. (emphasis mine)

Does he really believe this? There is no rate other than zero that the Raving Right in America would "agree with".
posted by Steely-eyed Missile Man at 2:46 PM on June 1


Piketty on The Colbert Report.
posted by homunculus at 8:53 PM on June 3


Omnivore: Piketty fever
posted by homunculus at 9:03 PM on June 3


« Older "This began my quest to find as many of the first ...  |  A meteor shower that has never... Newer »


This thread has been archived and is closed to new comments