Piketty Makes Up “Facts” and I’m the Bad Guy
I’m being cheeky in the title, but my irritation is real. Both ideological friends and foes have been chiding me for (in my paper co-authored with Phil Magness) focusing on trivial details, like Piketty getting the dates wrong for tax hikes and minimum wage increases. Big deal, Murphy! Focus on the inequality data.
But Phil and I couldn’t ignore the boneheaded mistakes. Chris Giles of the FT did exactly what my critics are suggesting, and he was in the news for a week. Then Piketty responded (quite inadequately), Piketty’s fans said, “Heh what tools, leave the data to the grownups, kids,” and now progressive bloggers smirk at you if you bring up Giles.
And that is understandable. Giles was a writer for a newspaper, whereas Piketty is an expert in this literature–a guy whom Larry Summers said should get a Nobel for the empirical work. Without taking 2 months to read up on the literature yourself, how could you possibly weigh in on this matter?
So that’s why it’s so important for Phil and me to show you just how unbelievably sloppy Piketty was when it came to things that a high school student should get right in a term paper, and then to see him dismiss them as “typos.”
So then, with that as background, when Phil and I show you that Piketty on some of his Excel cells just adds a number one year, then subtracts it away, with no apparent justification, you don’t just say, “Well, I’m sure the guy has a good reason for doing that.” No, because this is the same guy who just made up an alternate universe of the minimum wage, and then didn’t seem very fazed about it. (Note in his response to FoxNews, he used “etc” to refer to his systematic distortion of the minimum wage history.)
Anyway, frequent commentator (who is always earnest when he often criticizes me) Keshav Srinivasan just wrote the most astounding defense of Piketty yet:
By the way, concerning the FDR and Hoover stuff, I think you’re too quick to assume malice when carelessness is more likely. It’s more plausible that Piketty simply assumed that Hoover must have been the one who cut taxes because he was a Republican, rather than Piketty deliberately falsifying the history in order to make Republicans look bad. (That’s why he called them “typos”, because they were unintentional errors.) Isn’t it best to be charitable to your opponents’ motives?
!!!
We’re not talking about an appearance on the Daily Show, we’re talking about a published book, and moreover a book that is being lauded for its contribution to the historical record. I don’t know what to say at this point. I can’t believe I am being put on the defensive when Piketty is the guy who made up a bunch of “facts.”
“Isn’t it best to be charitable to your opponents’ motives?”
Nobody does this, so I don’t expect Bob to.
I gotta say, if on page 297 he has a 1933 where it should be 1931, and then on page 344 he spells “Roosevelt” with one O, and you guys made a big deal of his sloppy scholarship… yeah, that would be pretty trivial. But that is sinply NOT the magnitude of the errors here. It’s not close. Piketty’s data, to my understanding, is frequently and significantly incorrect. Is it possible to draw meaningful conclusions from systematically flawed data?
Yes. Exactly one: that the author did a lousy job.
I’ve used Microsoft products for a very long time. Clearly when he talks about typos what he really means to say is Excel is auto-correcting his numbers.
Hey Bob, Piketty is a Rockstar economist while you are only singing karaoke in some shady bars, what did you expect to happen if you criticize him?
People said the same thing about Bill Cosby and Hannibal Buress.
Didn’t know about the Cosby Buress story, and what I read now about it, I didn’t find people calling Buress to be the bad guy. Not even to mention that I have no clue if Cosby is guilty of the stuff he has been accused of (didn’t get convicted yet for anything, did he?).
Didn’t know about the Cosby Buress story, and what I read now about it, I didn’t find people calling Buress to be the bad guy. Not even to mention that I have no clue if Cosby is guilty of the stuff he has been accused of (didn’t get convicted yet for anything, did he?).
Bob, you know why you’re being made to be the bad guy there. You’re going against the mega-apologist for all the wreck that governments have done to their economies for nearly a century now: Keynes redux. Even here in Brazil op-ed column writers (marxists, as one should expect of them here) are praising him whenever they can. No one should be allowed to spoil their fun when the IMF-mandated haircuts start happening for good, so people must be well-prepared to accept them by pretending to read the Sloppy Frenchmen’s book, or by believing in writers who also pretend having read them.
Not just “typos”. Nobel “typos”.
like
It’s ‘Type Os’. Like me. I’m O positive.
I accept the “sloppy” narrative, but I don’t get the biased narrative. What does it matter if Coolidge is the tax boogeyman or Hoover? Same with minimum wage. He has plenty of criticisms of leftist government policy in France and it’s well known that he’s often a critic of the Socialist Party there despite his work advising them. Do you really think he has that much invested in US party politics? And whether the minimum wage rises under Clinton and Obama but not Bush I and II or Bush I, Clinton, and Bush II it doesn’t seem to bias in his favor since he’s got inequality increasing through all four of those administrations.
So sloppy, yes, but I don’t get the biased part.
“Do you really think he has that much invested in US party politics?”
I do think that. I am German and i care way more about what happens in the US than what happens in Germany, because lets be honest Germany is not that important. And in my opinion France is even less important.
I also think it does make a difference if Coolidge cut taxes or if Hoover did. Coolidge was the President during “the roaring twenties” Hoover was President during the bedinning of the Great Depression. I am not saying that there is a causal relationship here, but it does help Piketty´s case to pretend that taxes were cut during the worst economic crisis in US history as opposed to in one of the most prosperous times.
Was it intentional? I dont´t know and neither does anyone else (except Piketty himself).
That’s because for all your professed objectivity, you’re just as ideologically blind as the next (statist) guy.
“Do you really think he has that much invested in US party politics?”
I do think that. I am German and i care way more about politics in the US that about German politics, because lets be honest Germany is just not that important. And I think France i even less important.
In my opinion (and im not an authority here, i have not even read the book) it does make a difference if Hoover or Coolidge cut taxes.
Coolidge was President during “the roaring twenties” Hoover during the beginning of the Great Depression.
I am not saying that there is a causal relation here but it does help Piketty´s narrative to pretend that taxes were cut at the beginning of the worst economic crisis of US history as opposed to one of its most prosperous times.
Was it intentional? idk and neither does anyone else (except Piketty himself)
But to pretend that it does not matter who cut taxes is incorrect.
I tried to post that 3 times now and it did not work. so if this shows up 3 times later i apologize
“Do you really think he has that much invested in US party politics?”
I do think that. I am German and i care way more about politics in the US that about German politics, because lets be honest Germany is just not that important. And I think France i even less important.
In my opinion (and im not an authority here, i have not even read the book) it does make a difference if Hoover or Coolidge cut taxes.
Coolidge was President during “the roaring twenties” Hoover during the beginning of the Great Depression.
I am not saying that there is a causal relation here but it does help Piketty´s narrative to pretend that taxes were cut at the beginning of the worst economic crisis of US history as opposed to one of its most prosperous times.
Was it intentional? idk and neither does anyone else (except Piketty himself)
But to pretend that it does not matter who cut taxes is incorrect.
If you post as Anonymous (i.e. forget to fill the Name) it goes into the holding queue and usually Bob lets it out later.
I agree, there is a common narrative around the Great Depression is based on Hoover supporting laissez-faire and this being a big failure, then Roosevelt jumping to save the day with government intervention. Piketty invents false evidence to support his anti-capitalist view.
Bob just explained exactly why it matters, Daniel. If Piketty is sloppy and tilts his narrative politically on multiple basic statements of historical fact, why should we assume that his more complex data presentations – which exhibit multiple errors in their root spreadsheets, all generally operating in the same direction toward his narrative – are free of a similar slant?
So you DO think he is invested in US party politics? I do not.
As I say to Steve who seems to read me similar to you, I have no doubt he has preconceived notions – biases – about US party politics and that that is indeed the source of much of this sloppiness. But as far as I can tell it has no bearing at all on his narrative and therefore I don’t see how he’s biasing the narrative.
I don’t believe that Piketty has strong affections for either US party, although I do believe his preferences lean very heavily toward policies advanced by politicians on the left. This is true both historically and in the present day.
(FWIW Piketty also has a long and documented history of providing consultation and support for partisan electoral endeavors in his native France, and just this week signed on as an adviser to an anti-austerity party in Spain. So while I don’t think this is his main motive in the book, I can state with confidence that partisan electioneering is not at all beneath him)
Please keep in mind, Daniel, that bias does not need to be intentional. Whether his bias is caused by preconceived notions about party politics or real interest in influencing the political landscape, Piketty has still pushed the data and facts in his book to be more in line with his preconceived notions and less in line with reality. Also, if these preconceived notions are influencing his data and facts, they are affecting his conclusions (1) through the bias in the data on which he bases his conclusions and (2) through the preconceived notions that have already influenced his selection of the facts and data (and would therefore most likely influence how he arrives at his final conclusions).
How can he not be interested in US party politics,at least tangentally? When he’s calling for a world-wide wealth tax you cannot ignore the country with the highest GNP.
Plus, some of us give a damn about historical truth. In addition to what Phil has said, supporters of Piketty should care about his errors because they should care about getting the history right, not just whether it serves his argument or not.
re: “supporters of Piketty should care about his errors because they should care about getting the history right, not just whether it serves his argument or not.”
You’ll get precisely zero disagreements on this out of me, Steve.
So let’s have a little competition between friends and start a guessing game.
How many of Piketty’s supporters are going to come out in 2015 and demand the historical record be set straight?
I’ll leave this open to all comers to stake their position on this, but first come first served once a given integer is taken it can’t be taken by anyone else. Fun fun fun!
Sigh….
I guess it’s crickets on the menu…again.
Find me one of his “typos” or data errors that works against a broadly leftist world view and I’ll accept them as sloppy.
Sloppiness should be much more close to randomly distributed in terms of working for and against his overarching world view. When all of the “typos” go in one direction, well isn’t that exactly what the technical notion of “bias” means?
Steve you’re misunderstanding me. I’m sure he’s biased in the sense that he just figured Hoover raised taxes and wrote it up. What I’m saying is that that mistake absolutely doesn’t help his narrative in the book about inequality. How is his narrative helped if it was a Coolidge tax or a Hoover tax?
Show me a case where argument in the book relies on the direction of the mistake and I’ll call it biased, not sloppy. Hint: The argument of the book is not Democrats=good, Republicans=bad.
So you think it makes no difference if taxes were cut during the “roaring twenties” (when Coolidge was president) one of the more prosperous times of U.S. history or during the beginning of the Great Depression (when Hoover was president) the worst economic crises ever?
I’m not saying that there is a causal relationship but the correlation doesn’t help piketty’s case.
It does not influence his documentation of income and wealth inequality but does help him with his call for higher taxes for the rich.
I don’t know if he did that on purpose or not, but to claim that it does not matter if Hoover or Coolidge cut taxes is in my opinion incorrect
aby –
It does not make a difference for Piketty’s argument. It certainly would make a difference for a book on, say, labor supply elasticities for top rates or fiscal multipliers.
Ok. I understand that it does not make a difference for for his documention of inequality.
But he does call for higher taxes for the rich, right?
I’m not saying that it is reasonable to say that the great depression or the 1920 have anything to do with tax policy.
but some people will make a connection (this might be reasonable or not but people do it)
I think that’s pretty tenuous particularly because I don’t think he makes any claim about the impact of his proposed tax on growth.
But some people will draw those conclusions.
I don’t see how it is relevant whether he explicitly said it or not.
If he calls for higher taxes and says the price for that will be more episodes like the great Depression (I’m not saying that’s the case) his policy recommendation will get a lot less support.
and his incorrect claim that taxes were cut at the beginning leads one to believe (at least the layman) that the opposite is the case.
it could be the case that whenever taxes were raised the economy improved except this one time. But I think it does support his policy recommendation to not only ignore this episode but pretend the opposite is true.
he does advocate higher taxes for the rich, doesn’t he? (If he doesn’t everything I wrote is irrelevant)
But some people will draw those conclusions.
I don’t see how it is relevant whether he explicitly said it or not.
If he calls for higher taxes and says the price for that will be more episodes like the great Depression (I’m not saying that’s the case) his policy recommendation will get a lot less support.
and his incorrect claim that taxes were cut at the beginning of the great Depression leads one to believe (at least the layman) that the opposite is the case.
it could be the case that whenever taxes were raised the economy improved except this one time. But I think it does support his policy recommendation to not only ignore this episode but pretend the opposite is true.
he does advocate higher taxes for the rich, doesn’t he? (If he doesn’t everything I wrote is irrelevant)
I have not read the book by the way and I am not an economist so if I misrepresented the facts I apologize.
But many non economist have read the book and especially for them: tax cuts for the rich -> great depression
Or tax cut -> roaring twenties
makes a difference.
“The argument of the book is not Democrats=good, Republicans=bad.”
No Daniel, it isn’t. But the relevant portion of the book to which that specific error pertains *is* making the case that a series of specific historical events in the 1930s should serve as a precedent for Piketty’s own modern proposal of an 80% global wealth tax. A part of that case is the assertion that FDR did something in line with Piketty’s own proposal, whereas other presidents before him who have more of a laissez-faire reputation (deservedly or not, and I would argue not) are directly portrayed as being at odds with that proposal, plus a healthy heaping of insinuation of causal linkage to the Great Depression.
He stresses at length that the United States “pioneered” this level of 80% taxation, and for specific reasons rooted in creating a less unequal society in response to a depression that supposedly revealed the perils of inequality. If his basic historical narrative of that “pioneering” tax system innovation is off – and we have shown that it indisputably is – then one has to wonder if he has correctly interpreted the lesson he seeks to impart from that episode.
Phil –
re: “But the relevant portion of the book to which that specific error pertains *is* making the case that a series of specific historical events in the 1930s should serve as a precedent for Piketty’s own modern proposal of an 80% global wealth tax.”
So NARROWING the range of historical precedent in support of his position is supposed to FAVOR his narrative?
My head is starting to hurt.
No Daniel. But identifying what he purports to be the *specific* historical precedent, and attributing it to the Depression-relief policies of a much-beloved U.S. president, does indeed work in favor of an argument that seeks to revive and expand upon that claimed precedent.
OK so the real problem here is that people don’t like Hoover but they liked FDR so it makes a difference if it’s Hoover’s depression relief rather than FDRs. And therefore it’s biased in favor of his narrative and not just sloppiness likely derived from popular narratives about Hoover.
You are grasping at straws here. I don’t understand why you can’t just say “Piketty is strikingly sloppy throughout the book”. Spinning these more salacious narratives cuts into your credibility. And you need all the credibility you can amass, because as we’ve talked about many times I don’t think your arguments on the wealth and capital data are very strong.
It’s a simple matter of historical interpretation, Daniel. There’s an extensive mythology that FDR enacted progressive/proactive policies to alleviate the many problems of the Great Depression, one of which is commonly asserted to be inequality in the 1920s. There is a similar mythology that claims Herbert Hoover sat by and did nothing, believing that free markets would sort everything out. While specialists do indeed approach these subjects with greater nuance, those mythologies are pervasive in popular perception and even among many scholars who work in other areas besides the Great Depression. The point here is that Piketty used incontrovertibly wrong factual claims in a way that bolstered that mythology. His reason for doing so is not difficult to discern either, as that mythology is a basis for the historical precedent he claims to support his own preferred tax policy.
The counterfactual can be quite illustrative here.
Do you honestly think the average reader would react no differently to the part of the book in question if Piketty had written: “We should turn to the pioneering income tax policies of Herbert Hoover at the outset of the Great Depression as a model for how we may address inequality today.”?
re: “The point here is that Piketty used incontrovertibly wrong factual claims in a way that bolstered that mythology.”
Right. I have said several times now that the bias here is that he’s internalized this narrative about Hoover that’s historically inaccurate (what you call the “mythology”). Obviously repeating that narrative strengthens THAT narrative. But that was never what I was talking about. It does not strengthen *Piketty’s narrative*.
Daniel, they are attempting to pretend that Hoover supported laissez-faire and that Hoover’s depression relief was to do nothing at all. Here’s a typical history (not Piketty):
http://iws.collin.edu/kwilkison/Online1302home/20th%20Century/DepressionNewDeal.html
Now go and look up the actual Roosevelt campaign speeches of 1932 he promised:
* lower taxes
* balanced budgets
* A 25% cut in size of government
He didn’t do any of those things, but he promised them so he had something to contrast to Hoover’s policies. Here is Roosevelt himself explaining the situation in 1932:
The fact that Roosevelt didn’t keep his promises, and immediately changed tack to something completely different once he stepped into office, has been neatly airbrushed from history.
I’m sure he’s biased in the sense that he just figured Hoover raised taxes and wrote it up. What I’m saying is that that mistake absolutely doesn’t help his narrative in the book about inequality. How is his narrative helped if it was a Coolidge tax or a Hoover tax?
Show me a case where argument in the book relies on the direction of the mistake and I’ll call it biased, not sloppy. Hint: The argument of the book is not Democrats=good, Republicans=bad.
The argument of the book is that the market fails as demonstrated by growing “inequality” in the absence of high taxes. Piketty is totally unfamiliar the counter-explanation of the Great Depression in that it was the result of central bank monetary shenanigans beginning in WWI and thereafter. DK knows all about that topic at least as applied to WWI:
http://factsandotherstubbornthings.blogspot.com/2013/01/poor-kid-was-just-couple-years-too.html?showComment=1358998365568#c2813953362544372035
The outrageous mistake Piketty made about Hoover demonstrates that he is also completely unfamiliar with the evidence that Hoover did not impose “laissez faire” or “liquidationism”. Hardly anyone knows that Hoover raised taxes in 1932 and that fact alone would crucial to having the public understand the bigger truth. The left and the statists have eviscerated the US constitutional limitations on the power of Congress to “regulate interstate commerce” and “to coin money” based solely upon their dishonest claim that the Great Depression proved once and for all that market failed and requires government spending and money dilution. That is the story everyone is taught in law school in about 3 minutes the first day of constitutional law class with the warning that you’re not there to argue “original intent”. The impact is that the Feds now have plenary economic power over every molecule of person and property in the nation (and perhaps the world).
The closest is the minimum wage case. Minimum wage is allegedly supposed to reduce inequality (I think this is what he said – I haven’t read that section for a couple months). Inequality is supposed to be increasing steadily through the 90s and 2000s. Multiple periods of delay in raising the minimum wage would be consistent with that. That seems weak to me, though, since the real value of the minimum wage (which I think he presents? again it’s been a little while) has tended to be lower than it was in the 60s even with the corrected minimum wage history.
On the tax history it’s not clear to me at all how that’s biased in favor of his narrative.
http://4.bp.blogspot.com/-pOol7YySfB0/Ug0-6-GdsAI/AAAAAAAAAF4/OklCt1wI0gA/s1600/unemp+and+min+wage.jpg
Worth talking about.
Yep:
Excluding Inferior Workers: Eugenic Influences On Economic Reform In The Progressive Era [PPT]
http://www.princeton.edu/~tleonard/papers/Excluding.ppt
Slide 44
VIII. The Eugenic Benefits of Minimum-Wage Laws
“But eugenically minded progressives advocated minimum wages precisely because binding minimums would cause job losses.
“They argued that minimum-wage-induced job loss was a social benefit because it performed the eugenic service of ridding the labor force of the “unemployable.”
If you ask the average guy (person a) on the street what the worse economic crises was in the U.S. they will say the great Depression. And now you tell them that taxes were but for the rich right at the beginning
and a second person (person b) you tell that taxes were raised during the beginning of the great depression and taxes were cut during the soaring 20s.
so you think this will not cause person a to be more likely to support tax increases and cause person b to be more likely to oppose tax increases.
OR
are you saying piketty advocates tax increases but he has NO interest whether they will be implemented or not
I meant cut and roaring obviously
Another example:
you write a history textbook that says taxes were cut at the beginning of the great Depression.
This will not cause students to support tax hikes more than if you had given them the truth?
sure some kids might like depressions and some will be smart enough to not support nor oppose a tax policy based on a single correlation but I think it would be ridiculous to claim that on net this will not cause this generation to support tax hikes more than he you give them the Coolidge tax cut, Hoover tax hike story.
maybe you need a PhD in econ to understand why I am wrong, who knows
Another example:
you write a history textbook that says taxes were cut at the beginning of the great Depression.
This will not cause students to support tax hikes more than if you had given them the truth?
sure some kids might like depressions and some will be smart enough to not support nor oppose a tax policy based on a single correlation but I think it would be ridiculous to claim that on net this will not cause this generation to support tax hikes more than if you give them the Coolidge tax cut, Hoover tax hike story.
maybe you need a PhD in econ to understand why I am wrong, who knows
Read my second question to you, Daniel.
If Piketty had written “We should turn to the pioneering income tax policies of Herbert Hoover at the outset of the Great Depression as a model for how we may address inequality today” do you honestly think that the enthusiasm for his global wealth tax idea would be the same?
Phil –
Of course there will be a different reaction because lots of people share this mythology and they’ll be confused. If he had written this I don’t think enthusiasm would be any different:
“Hoover was actually more activist than popular mythology suggests and is generally considered to be quintessential of the progressive wing of the Republican Party at the time. We should turn to the pioneering income tax policies enacted by Herbert Hoover and FDR at the outset of the Great Depression as a model for how we may address inequality today”
The difference is entirely due to the prominence of the popular narrative, not what Piketty has to say or propose, in other words. Add a clarifying sentence on that point and I don’t think this changes anything at all, no.
Like I said, I think you’re grasping at straws to make this much more salacious than it really is – what “it really is” being that Piketty is very sloppy in a number of cases and in this case his sloppiness is likely the product of his inculcation of popular Hoover myths.
Daniel – If that’s the point he wished to make, then fidelity to historical accuracy dictates that he should have made it and should have credited both Hoover and FDR.
He didn’t though. And a part of why he didn’t stems from either (a) catering to the popular mythology around Hoover – which makes him biased, and in a way that insulates his general argument from the complicating critique of having to endorse Hoover’s Depression response – or (b) believing in that mythology himself, which makes him an inept historian. Neither is an enviable position for Piketty’s argument.
Yes, precisely. It’s both (a.) and (b.). He’s biased about the Hoover mythology and though he’s done widely recognized work in economic history in his area he is not a competent historian generally and apparently not on Hoover specifically.
That’s all fine. But again this is not the bias I referred to initially. This is what I initially referred to as him being sloppy about the history.
Although I agree your (a) is right, this bit of it isn’t: “and in a way that insulates his general argument from the complicating critique of having to endorse Hoover’s Depression response”
So you think that making the case for “We should implement a policy based on what Hoover did during the Depression” is no more difficult than the case for “We should implement a policy based on what FDR did during the Depression”?
Curious.
Dude, we’ve already done our dueling counterfactuals. Do you really need this re-explained?
There’s no way in the world Piketty is ever going to write that, nor even something vaguely similar to that.
Is anyone here going to put up their hand and say, “Yes I expect this will happen?”
I should say the counter-narrative on Hoover is a little mythical in its own right. The difference between Hoover and FDR is still quite stark despite the fact that Hoover was a progressive. That’s not particularly important here of course.
I make no defense of either Hoover or the counternarrative about him. I only ask that we acknowledge the deficiencies of mythology, and acknowledge that the actual issues at stake are much more complicated than popular mythology permits.
Piketty does neither, which can only mean that he either chose to ignore something that complicates his own interpretation of the same events, or he’s inept with the subject matter.
There’s no “or” about it. There’s no evidence at all to attack him on the former explanation. It’s clearly the latter.
I simply raise the “or” as an alternative possibility. I’m quite content to note that Piketty is an inept historian. In any case though, it undermines his work and exposes some of the biases at play behind his arguments.
Maybe it’s just me, but I don’t trust a person who is so biased that he doesn’t even check his facts a whole lot more than a person that is so biased that he intentionally distorts his facts. The debate isn’t “Is Piketty wrong?” Instead it’s “Was his wrongness intentional or simply negligent?” and I’m not sure why it matters.
But since we’re already debating it, I think it would be wise in this instance to ask, “Which is more plausible?” Given the pattern of distortion, I think it is far more plausible that Piketty intentionally distorted his facts/data to support his narrative. If you disagree with me on this point, please consult the previous paragraph.
Andrew – it only matters if you don’t want to fan a smear campaign. Negligence and bias are both bad, but the latter raises questions of integrity.
We’ve already concluded that Piketty’s work is biased. We are now debating the nature of the bias. On this, the point of consensus seems to be that his bias is at best laughably negligent and at worst intentionally inaccurate in support of his narrative.
Are you saying that you put significantly more stock in a person who is laughably negligent?
Defenses like “typos” should raise questions of integrity.
When I screw up I’m expected to own the problem: point out and admit error if it was not found by someone else, fix it and any downstream problems, and identify the causes of the error so that it can be prevented in the future.
There’s “sloppy” and then there’s failure to act like a professional (while running around the country living off the adulation of ideologues).
I’m sure he’s biased in the sense that he just figured Hoover raised taxes and wrote it up. What I’m saying is that that mistake absolutely doesn’t help his narrative in the book about inequality. How is his narrative helped if it was a Coolidge tax or a Hoover tax?
Show me a case where argument in the book relies on the direction of the mistake and I’ll call it biased, not sloppy. Hint: The argument of the book is not Democrats=good, Republicans=bad.
The argument of the book is that the market fails as demonstrated by growing “inequality” in the absence of high taxes. Piketty is totally unfamiliar the counter-explanation of the Great Depression in that it was the result of central bank monetary shenanigans beginning in WWI and thereafter. DK knows all about that topic at least as applied to WWI:
http://factsandotherstubbornthings.blogspot.com/2013/01/poor-kid-was-just-couple-years-too.html?showComment=1358998365568#c2813953362544372035
The outrageous mistake Piketty made about Hoover demonstrates that he is also completely unfamiliar with the evidence that Hoover did not impose “laissez faire” or “liquidationism”. Hardly anyone knows that Hoover raised taxes in 1932 and that fact alone would be crucial in having the public understand the bigger truth and reject the phony narrative. The left and the statists have eviscerated the US constitutional limitations on the power of Congress to “regulate interstate commerce” and “to coin money” based solely upon their dishonest claim that the Great Depression proved once and for all that market failed and requires government spending and money dilution to avoid a repeat. That is the story everyone is taught in law school in about 3 minutes the first day of constitutional law class with the warning that you’re not there to argue “original intent”. The horrendous impact of this phony narrative is that the US government now has plenary power over every molecule of activity in the nation (and perhaps the world). DK is again attempting to obfuscate this essential point and is again failing.
True, but forgivable, because many people have been fed a particular version of history.
However, to go and invent some facts and figures as false evidence based on ignorance is much more difficult to forgive.
Yes, that is all rather convenient.
Maybe there’s scope for someone to make a lot of money writing a blatantly false history book that fits a lot of preconceived notions and suits the political narrative… then refuse to give interviews or answer any questions but a few years later say, “I was just trolling you guys, thanks for the cash.”
The question is how outrageous can you get without giving the game away? A nom de plume like “Murf Bobby” would be borderline.
Just putting the idea out there.
Parents’ night at [Piketty’s high school:
https://www.flickr.com/photos/bob_roddis/6670091329/in/set-72157600951970959
Best case scenario, Piketty and his editor got lazy. Worst case, he was deliberately falsifying data. Neither look good for a guy trying to convince us to take him seriously.
After all, if we can’t even trust him to get right simple information that can be easily checked via a Google search, how can we trust with with anything else?
Besides, one’s an anomaly, two’s a trend
I’m not a historian nor statistician but it seems Piketty’s data has major flaws and his history is incorrect, both mysteriously in a way that helps his narrative.
I’m not an economist but even I managed to understand Deirdre McCloskey’s basic Demand/Supply criticism just because I’m helping my daughter with her high school junior micro class.
I’m not a mathematical whiz but I managed to go through Daniel Kuehn’s (and others) basic demolition of Piketty’s mathematical “models”.
Preponderance of evidence does count for something.
“Spinning these more salacious narratives cuts into your credibility.”
–Daniel Kuehn, explaining to Phil Magness why Piketty’s salacious narratives about Herbert Hoover do not cut into Piketty’s credibility..
You know what would boost your credibility on that last comment? Quoting me saying that this doesn’t cut into Piketty’s credibility.
“I am not saying it does not cut into his credibility, but I won’t say it does cut into his credibility.” – Daniel Kuehn, on how to say and not say what is difficult to say but is better left unsaid.
MF I know you have a tough time keeping up sometimes, but the whole point in responding to Bob above is that I emphatically agree that it cuts into Piketty’s credibility.
What I don’t agree on is that the mistakes (not typos, outright errors) always buttress his narrative on inequality. Even Phil has come around to my view on that point it seems, claiming only that it buttresses the Hoover mythology he’s absorbed. But that’s practically tautological – yes, repeating the Hoover methodology buttresses the Hoover methodology. Is there a lurking bias to make all the history and numbers go in one direction? If there is nobody’s bothered to point it out yet.
“MF I know you have a tough time keeping up sometimes”
No worries, what you call me not keeping up, is me close to lapping you. It may look like I’m not keeping up, but…
For what it’s worth, Magness’ thesis from both the audio and his initial posts here was clear to me from the start. I’ve just been reading, waiting for this awkward and strained series of “buts” to end.
“but the whole point in responding to Bob above is that I emphatically agree that it cuts into Piketty’s credibility.
What I don’t agree on is that the mistakes (not typos, outright errors) always buttress his narrative on inequality.
You don’t need to believe they *always* do so, notwithstanding the perhaps exaggerated insinuations that they do. It is sufficient that he does so in most instances, in which case the answer is emphatically yes.
“Even Phil has come around to my view on that point it seems, claiming only that it buttresses the Hoover mythology he’s absorbed.”
Claiming only that it does so? Or claiming that it only does so? There’s a difference. You should be clearer on that point.
I don’t see Magness saying that the mistakes *only* buttress the Hoover mythology. On the other hand, if Magness has thus far only mentioned that the mistakes buttress the Hoover mythology, that doesn’t imply that the mistakes only buttress that mythology, and nothing else. He was just using it as an iilustration. He can’t be expected to repeat his entire paper here.
Indeed, if you listen to the audio interview, and if you read the paper, you will notice that the mistakes *also* buttress the “inequality is U-shaped during the 20th century” narrative, which is the important one. Magness has made that very clear. Magness shows that Piketty selected among 6 or 7 studies on inequality and when constructing the trend line, over each decade, chose the study that showed the results that are most consistent with the results being the U-shape. Along the trend, meaning going from the early 20th century to the late 20th century, Piketty used one particular study, then dropped it for another, then went back to the previous study. He literally cherry picked data out of a handful of different studies that used different methodologies, in order to make the whole trend shaped the way that fits the inequality narrative/mythology.
“But that’s practically tautological – yes, repeating the Hoover methodology buttresses the Hoover methodology. Is there a lurking bias to make all the history and numbers go in one direction? If there is nobody’s bothered to point it out yet.”
I don’t like these tangents. Too hand wavey. At any rate, what you said is not true, because that’s one thing Magness (and Murphy) are talking about!
OK Daniel you’re right (I just thought my line was funnier that way). A more accurate (but less funny) way to handle your charge to Phil would go like this:
“Daniel, it seems that you think Phil has some kind of agenda when he invents this salacious narratives about the historical episode of Piketty writing his book. Why can’t you just say Phil was sloppy, and leave it at that? I don’t see what Phil would gain by impugning Piketty, as you seem to suggest. Do you honestly think Phil cares about Piketty’s motivations?”
I don’t know, Bob. He seemed quite sure in the comments above that Piketty’s spreading of salacious narratives about Hoover did nothing to advance Piketty’s case. While not quite the same as asserting that Piketty emerges from such antics unscathed, it also appears that the harm he finds in Piketty’s antics is minimal.
But then again, we’re talking about a person who also seems to believe that Piketty has made a fairly watertight empirical case for rising wealth inequality 😉
I don’t know the literature well enough to say how watertight his case on wealth inequality is. I will say:
1. The points you’ve raised are not very convincing.
2. He definitely overstates how watertight his findings are, but
3. His findings don’t appear to be obviously wrong either. If they’re wrong it’s clearly on more detailed assumptions and estimations that I don’t know enough to rule on.
I know you want to always see my view in black and white, but that’s actually how I see things.
“I don’t know the literature well enough to say how watertight his case on wealth inequality is.”
Followed by a list of musings in which you proceed to do exactly that.
If you do not know that literature and do not feel qualified to comment upon it, Daniel, then you should exercise your own advice and abstain from asserting things like “His findings don’t appear to be obviously wrong either.”
If you think Piketty has largely – even if imperfectly – demonstrated his case that inequality is on the rise, then as someone who has read the book you should be able to at least articulate how this is so. And if you think that my counterargument – that he has not proven his case – falls short, you should also articulate your reasons why you think this is so. The answer needn’t be black and white, but it should entail something more than a strongly asserted but superficial opinion, followed by an immediate retreat into feigned humility.
None of the three things I’ve listed exhibits confidence that his results are watertight. What exactly do you take “watertight” to mean? I take it to mean something like the best out there and fairly irrefutable. I don’t claim that and none of what I listed suggests that.
So take the wealth numbers that you were opining to Bob he just gives you nothing and codes the numbers in. Well it took all of two minutes for me to figure out he doesn’t “give you nothing”. There are two important adjustments to the HMRC wealth data: mortality multipliers and the adjustment for the share of the population with probates. Add to those two whatever adjustment is made to smooth or harmonize the two series together. That’s three adjustments where I don’t know the literature well enough to rule on Piketty. But I certainly know the literature well enough and have read enough to say that:
1. It’s not obviously wrong that these adjustments are problematic – if they are wrong (which they may be) it’s on some detail I don’t know much about, and
2. When you go around saying he just coded these numbers in or (one of your favored phrases) “pulled it out of thin air”, you’re not offering a convincing argument. This is what much of the more damning claims you’d provided amount to: “look it’s REALLY different and we don’t know exactly what he did”.
He is sloppy. I’m not really sure if he has an agenda or not. His sloppiness always seems to come in the form of loving to tell people outside his field how little they know.
Look Phil clearly doesn’t like me and to be honest the feeling is mutual. That’s not the important part. I raise the question because I want you to be clear that I’m definitely not trying to defend or shield Piketty – I’m critical of him and I want to get the criticism right. I don’t have any particular hope Phil is ever going to realize that about me, but I hope you would.
Murphy, you are up against a very entrenched worldview that mainstream academics, journalists, authors, and political pundits have taken for granted as gospel.
When you criticize Piketty’s data, your arguments are not, as unfair as this is, viewed as intended to spread truth. They believe they have the truth, period. You cannot be on the side of truth finding. You are necessarily someone motivated by sinister intentions. Malice, jealousy, hatred, and power hungry. Yes, you are power hungry because you are criticizing an author whose data is supposed to imply a moral value in economic equality.
Hardly anyone is even questioning whether wealth inequality is an evil at all. Piketty’s on the moral high ground because he is preaching wealth equality. His data must be protected and defended against all attacks, where leeway is given only to the most egregious “typos”.
Yes, it is unfair. We are living in an age where economic equality is still the dominant moral high ground. At the same time, it is very difficult to defend wealth inequality on empirical grounds, because we also happen to be living in an age where a significant enough number of people are acquiring wealth through violence, coercion, deceit and deception that arguing the morality of wealth inequality on the basis of merit, is becoming increasingly difficult to defend with diminishing real world examples to refer to.
One of the worst things about socialism is that the longer and more pronounced it is practised, the less people are able to know (in our positivist age) that free markets are a necessary requirement to many of the problems they experience.
Daniel – You do realize that saying a proposition is watertight is very different from saying *how* watertight it happens to be, don’t you? I hold no belief that you view Piketty’s argument as completely watertight. But your comments do strongly suggest that you accept it as largely valid, with a few peripheral caveats. My point is that asserting Piketty is largely correct, followed by a retreat into feigned humility about the subject, is insufficient to validate your assessment.
Now you raise a particular point about Piketty and the HMRC data: “There are two important adjustments to the HMRC wealth data: mortality multipliers and the adjustment for the share of the population with probates.”
That’s an interesting theory, and indeed a plausible one. But it also has a big problem: Piketty’s source tables do not reveal how he performs any of these alleged adjustments. He only gives us the post-adjustment results of whatever modification he performed, and they’re only decennial averages at that so we can’t even confirm if he handled the raw data correctly.
Now maybe he did adjust the numbers in some systematic and methodologically justified way. But I’d like to see it, and I’d like to know more about his calculations than can be ascertained from the various hypothesized but as of yet unsubstantiated explanations offered up in his defense by a Piketty internet fanboy. And considering that Piketty *does* have a demonstrable history of making up numbers out of thin air in his other charts, this request about his UK “adjustments” is entirely reasonable for me to make.
Yes, when a scientist goes through a career of peer review and professional collaboration to produce a body of results I rely on the institutions of the republic of science and assume basic levels of validity. Clearly there are a range of results out there and some are stronger than others on various margins. To make that sort of judgement call I need to know more.
re: “That’s an interesting theory, and indeed a plausible one. But it also has a big problem: Piketty’s source tables do not reveal how he performs any of these alleged adjustments. He only gives us the post-adjustment results of whatever modification he performed, and they’re only decennial averages at that so we can’t even confirm if he handled the raw data correctly.”
Well it’s not a theory it’s the footnote. Yes – before I made accusations like you have I’d try to figure out exactly what adjustments were made.
re: “But I’d like to see it, and I’d like to know more about his calculations”
Clearly you aren’t THAT interested in seeing it if you’re already jumping on this in an article without investigating!
Out of curiosity – would you be willing at all to make your referee reports for your paper available? I’m kind of curious where they pushed back.
DK wrote:
Clearly you aren’t THAT interested in seeing it if you’re already jumping on this in an article without investigating!
Just to be clear, we didn’t talk about UK data in the paper. You’re here referring to Phil’s musings on his blog and Facebook, right?
Yep – I’ve only skimmed the article and I knew it was a compilation of the issues you’ve gone over (which I had read in detail). Didn’t realize UK was out of it.
But the point remains – this data appendix provides way more details than you’d typically get. Yes if you want to make the case that he botched the analysis you DO have to do more work. Reproducing results is very arduous work. You have to actually ask for do-files or SAS programs or whatever he uses and it takes a ton of effort.
So I’m a bit skeptical when Phil says he’s “interested” in this, or at least skeptical of how interested he is, particularly when he says that Piketty gives him nothing when he clearly gives him at least several components of the answer in a footnote.
So there we have it – after lambasting me for expressing skepticism over what Daniel imputes to a far more opaque and deficient footnote in Piketty, and after carrying on about his supposed concern for the “republic of science” and the need to be charitable – nay deferential – to published work that has emerged from its processes, it emerges that he’s criticizing me over a point that I did not make in a paper he now he admits he has not read! As usual, the complete lack of self awareness prevails.
I never claimed to have read it! You did make the point, though. You’re getting awfully jumpy over a misattribution to your paper.
What’s so opaque about the footnote?
We should neither be charitable or deferential. Note what I said carefully – it wasn’t that.
Daniel – I’m simply noting the curious juxtaposition of your call for a basic level of process-induced deference to Piketty, posted immediately along side of your own offering of criticism over a point we did not make in a paper you did not read. It’s also hardly “jumpy” to simply note as much in a response to something you’ve been blathering on about – erroneously – for several posts prior.
Since you bring up the footnote though, let’s look at what Piketty actually says in his footnote about that chart…
“UK: 1910-1980 series based upon Atkinson-Harrison 1978 pp.139 et 159 and Atkinson et al 1989 table 1; 1990-2010 using IRS data; (Note: no top 0.1% estimate)
1810-1870: Lindert 2000 table 2 (see also Waldenstrom 2009 Tables 4.A1-A3)
1990-2010: official HMRC estimates using estate tax data; see Waldenstrom 2009 Tables A.A1-A3 and official HMRC publications for recent years, in particular “Personal Wealth Statistics 2001-03 to 2005-07” (June 2011) and “Personal Wealth Statistics 2008 to 2010″ (September 2012).
(Note: HMRC estimates are expressed relatively to population with probates, i.e. only 31% of adults in 2008-10; so top 30% in MRC Tables 13.8 corresponds to top 9.3% of total population)”
The supplementary document he released after the FT critique adds almost nothing substantive to this note, so it seems as if what we have then is a single opaque reference to probates. So – are you asserting from that tiny speck of a tea leaf, that you can divine how Piketty adjusted and reconciled the HMRC data to the older Atkinson series?
I ask because (1) Piketty actually does not claim to be doing what you attribute to him at anywhere near the level of specificity you claim, and (2) even if he did perform the adjustment you speculate to have occurred, it does not appear in the data or formulas of his excel file from whence the chart is purportedly derived.
In short, what Piketty derives for his UK wealth data is not presently replicable from the sources he lists, the annotation he provides, or the formulas and data sheets he shares. And with that in mind I am entirely justified in calling it opaque.
re: ” So – are you asserting from that tiny speck of a tea leaf, that you can divine how Piketty adjusted and reconciled the HMRC data to the older Atkinson series?”
Ummm… no I’m not asserting that. That’s my whole point Phil. I am not asserting that I can assess how watertight this all is, only that I’d know where I would start if I were interested in exploring it further. I’m not even 100% sure where the adjustments on the mortality multipliers and probates were made (with Piketty or with the HMRC – the note refers to the HMRC data he used but I’m not clear on where)
I forwarded Bob the Atkinson paper btw, if you or he are interested in doing more work on this. I’m sure Piketty has programs you can request too.
My point is the idea that your point that he “gives you nothing” is absurd. These data notes alone provide you with the material to get started on an extensive data replication project. That’s hardly “giving you nothing”!!!
re: “In short, what Piketty derives for his UK wealth data is not presently replicable from the sources he lists, the annotation he provides, or the formulas and data sheets he shares. And with that in mind I am entirely justified in calling it opaque.”
No, it just requires more work than you seem willing to put into it.
It’s amazing to me that you are trying to portray Piketty as being lazy or withholding of information on this unbecoming of an academic. A good comparison is another HUP book that Piketty criticizes in Capital (unfairly I think) by Goldin & Katz (The Race Between Education and Technology). Their technical appendix is considerably less detailed than Piketty’s and much of it refers the reader to methods described in other papers (much like Piketty’s). This is pretty standard and very reasonable. People who actually want to reproduce results have to make some extra effort of emailing and asking for programs, downloading the raw data, etc. But Piketty gives you much more detail even than that. There’s nothing out of the ordinary about what he provides.
You just need to be willing to actually do some work, which you don’t appear willing to do. You can’t just open his excel file with his figures from the book and expect that to answer every question you have.
Major journals are starting to require that do-files and (when it’s possible) data be posted online with the article, but this is relatively new. There’s not some big conspiracy because you might have to actually ask someone for their programs and put in a little work. Result reproduction is a lot of work. “It’s not immediately obvious to me” is not the same thing as “it’s not reproducible”.
I have to say, Daniel, your perspective on this discussion is, well…odd.
You seem to be of the belief that the burden of demonstrating the validity empirical work falls not on the author of that work, but on his critics. As does, apparently, the task of deciphering exactly what that author did from footnotes and data files that are demonstrably insufficient to permit even basic replication.
I don’t mean to sound immodest here, but I’ll go ahead and venture that I’ve probably spent more time digging through the specific data file in question for the UK numbers than 99.9% of all *academic* readers of Piketty’s book, which doesn’t even dip into his generally-oblivious popular readers. I similarly obtained both Atkinson sets that the chart is supposedly derived from, as well as the supplemental data from HMRC, and after reassembling them into a consistent series, attempted a number of different averaging techniques to get the decennial numbers that Piketty reports. As Piketty did not share those calculations, it is necessarily guesswork based on very limited and opaque annotation practices. But I will say that even on the ostensibly “clean” part of the series – Atkinson from ca. 1960-1980 – Piketty’s numbers still routinely fall a couple percentage points off from what a conventional average yields.
The lesson to take away from all that is actually the exact opposite of your insinuation that I need to do “more work” to figure out what Piketty in his UK charts did before criticizing them. Rather, it is to note that it is simply impossible at this point to reconstruct those same charts from the limited, opaque annotation and data files he has presented for scrutiny. But that is his burden to overcome, not mine. If he can’t even meet the necessary minimums to allow basic replication of his data presentation, then his quantitative work in that instance has not been sustained on any scientific basis. The truly odd thing about it all though is that you seem to be fine with that outcome – that, as Bob said from the outset, Piketty gets to make stuff up and somehow I’m the bad guy.
Of course you’re welcome to disprove anything I have just said by trying your hand at reproducing Piketty’s chart based on the limited material he publicly provides. If you were to successfully do so, I’d even be grateful as it would allow a deeper analysis of whether the discretionary calls he made in his assemblage were the correct ones. But it seems from the above and from your comments elsewhere that it is in fact *you*, Daniel, who remains uninterested in those larger questions. So instead you try to impose burdens of proof that were never mine to meet, but Piketty’s, only to retreat into feigned humility and the cover of claimed unfamiliarity with the subject matter whenever pressed on the absurdity of your expectations. Argumentum ad Kuehnum indeed.
Phil has spent more time arguing with Daniel than 99.9% of academics.
We all have our hobbies, Bob =)
I don’t mean to be saying much of any of that Phil.
I’m saying you have not raised serious questions by pointing out that data notes on their own are insufficient to validate the analysis.
I’m saying these analyses take pages and pages of code and while a few journals require co-publication of that code it’s not suspect that pages and pages of code don’t come with the data notes. He tells us how he adjusts it – he uses the methods in Atkinson et al. Anyone with Google can find what they are. It may, of course, require more than Google to understand them. It’s not a matter of using different averaging techniques on the HMRC data. There were adjustments made to the data before you even average them.
re: “But it seems from the above and from your comments elsewhere that it is in fact *you*, Daniel, who remains uninterested in those larger questions. ”
Right I’m not interested. I am way out of my depths. I’ll follow the literature in a general way as it comes out. I’m sure there will be a very productive literature on this. I find it *interesting* but I am not interested in pursuing it. It’s well outside my field.
“I’m saying you have not raised serious questions by pointing out that data notes on their own are insufficient to validate the analysis.”
That isn’t what Magness indicated to you. He indicated that the data notes, the data, and the adjustments are not sufficiently explained to replicate what Piketty has done.
Wow.
I gave you his full footnote above, Daniel. What little it actually contains is far removed from what you keep claiming it contains, or what you evidently imagine it contains. Just like your prior attempt to excuse away Piketty’s odd Soviet adjustments, you’ve speculated an explanation that Piketty himself does not actually offer, elevated it to a strange new reality that simply isn’t so in Piketty’s text, and proceeded to defend Piketty as if it were a settled matter of his contribution.
It may be the case that Piketty performs some sort of adjustment before his average. It may be the case he adjusts after his average. Or it may be the case that he simply massages his numbers upward or even makes things up to fill in the gaps, as he has been found to do elsewhere in the book. We don’t know, because he didn’t bother telling us in his annotation. And we certainly don’t know if he performed any of these adjustments correctly, because they aren’t there in his source files. But we do know for certain that he did not annotate his data with sufficient detail to permit even a basic reconstruction of whatever it is that he did.
Now that could only mean three things about your proposed explanation: (1) You’re supplying an argument on Piketty’s behalf that Piketty himself did not make, in which case all that I’ve said above applies in criticism of your position. (2) You’re genuinely imagining all manner of detail in Piketty’s text & notes that simply aren’t there, in which case you might want to get that checked out. Or (3) you’re being a less than honest interlocutor when you raise the point. None of these is an enviable position.
This Phil: “UK: 1910-1980 series based upon Atkinson-Harrison 1978 pp.139 et 159 and Atkinson et al 1989 table 1; 1990-2010 using IRS data”
And see the technical appendix where he also discusses using Atkinson’s method.
Yes, Daniel. Nobody disputes that he traces his source data for 1910-1980 from Atkinson, or that he ostensibly mirrors Atkinson’s findings for those periods.
I might also add that Piketty nonetheless deviates from Atkinson at a number of points in his results, including the creation of a 1950s data point for the top 10% where Atkinson indicates that the available data was not sufficient to permit such an estimate. I might also point out that Piketty merged multiple distinct series in Atkinson including an early partial estimate for 1913, a 1930s series for only England and Wales, and a post 1950s series for all of Great Britain, after Atkinson specifically stressed the implications of viewing each distinctly. But the source is indeed Atkinson.
What Piketty does not do, however, is specify how he adapted any of the corrections in Atkinson’s series to the post 1980 data or reveal his implementation of *any* of those corrections. You’ve certainly imagined that for him and credited it to him, but it simply isn’t there in the note or spreadsheet file. So once again we’re back to my original point: Piketty’s annotation on that chart is entirely deficient and opaque. There is also no more of a reason to believe your theory occurred – much less check his work since he gives us no way to check it – than to postulate that it’s yet another example of Piketty’s documented habit of simply massaging the numbers to get the results he expects.