Hacker News new | ask | show | jobs
Piketty's Findings Undercut by Errors (ft.com)
66 points by wcgortel 4416 days ago
8 comments

In Piketty's credit, he did make this easy.

Providing data for reproducibility and replicability is relatively new in economics.

There are, however, two kinds of claims here. One is that mistakes were made, which is grounds for schadenfreude for those (like me!) who don't like the slight jumps in reasoning from "inequality is rising" to "something must be done about inequality" to "this is what must be done". For all I know, the key injustice in modern economies is topocracy:

https://encrypted.google.com/url?sa=t&rct=j&q=&esrc=s&source...

But secondly, it is claimed that he cherry-picks. Economists (we) do that. A lot. And it has to be called out.

"But secondly, it is claimed that he cherry-picks. Economists (we) do that. A lot. And it has to be called out."

The total effect of his cherry picking can't even be discerned from this article. With a lot more work (more than a single blogger probably has time for), it might be shown to be much worse. For example, Piketty chose the U.K., France and Sweden for his list of European countries (averaging them all into "Europe" without adjusting for their populations, as the article pointed out). But why those particular countries? Would the results have been completely different if he would have also included Germany, one of the wealthiest and most populous European countries? We'll probably never know how he came up with his set of countries, and whether he deliberately omitted other countries whose data was available because they would have caused his results to be not significant enough to publish (or even contradicted his thesis).

This reminded me a lot of one of the original studies linking cholesterol to heart disease, which chose a carefully picked subset of European countries even though results were available for several others. Subsequent research showed that including the data from the missing countries would have yielded very different results.

Related, useful concept:

https://en.wikipedia.org/wiki/Simpson%27s_paradox

See in particular the Berkeley Gender Bias case if you don't want to hack the math apart.

The specific way you slice and dice data matters.

> But why those particular countries

Because they had coherent archives of several proxies that Piketty uses for a long period of time. He opened the door to people doing the same work he did and contradicting him with other countries, but I would not bash him for not doing the same work on every country on the planet.

I liked your point, which is similar to my reaction to the book. Part 1: "Relative inequality is rising" Step 2: "?" Step 3: "Policy proposals"

Most notably, Piketty never made a case why his policy proposals are the best way to improve the lives of the bottom 50% on an absolute basis.

Have you read the book? It is clearly not his focus (or area of expertise!) but the book makes its premise clear: extreme inequality is inconsistent with the meritocratic ideals upon which modern capitalist democracies are founded, and can lead to social unrest. This is close to a direct quotation from the book's introduction --0 I would quote it exactly if I had the book in front of me right now. Maybe you don't agree with this premise, but I don't think you can fault Piketty for taking it as an assumption (explicitly!), particularly given that he is not a sociologist or political philosopher.

And in either case, it strikes me a significant contribution in itself to move the debate from "does capitalism and rising productivity raise all boats?" (answer: no) to "ok then...is extreme inequality really such a bad thing?"

How is the answer to raise all boats 'no'?
You're right, I was imprecise. Piketty does not claim (I don't think) that capitalism doesn't raise all boats, only that capitalism and increased productivity do not raise all boats equally.
Lawrence Summers' review/criticism of Piketty's tome is worth a read, even for non Summers' fans (in general I mildly disfavor Summers, but I appreciated this piece).

"Piketty’s timing may be impeccable, and his easily understandable but slightly exotic accent perfectly suited to today’s media; but make no mistake, his work richly deserves all the attention it is receiving. This is not to say, however, that all of its conclusions will stand up to scholarly criticism from his fellow economists in the short run or to the test of history in the long run. Nor is it to suggest that his policy recommendations are either realistic or close to complete as a menu for addressing inequality.

...

Books that represent the last word on a topic are important. Books that represent one of the first words are even more important. By focusing attention on what has happened to a fortunate few among us, and by opening up for debate issues around the long-run functioning of our market system, Capital in the Twenty-First Century has made a profoundly important contribution."

http://www.democracyjournal.org/32/the-inequality-puzzle.php...

The focus on the balance sheet and not just the P&L is important. Unless these are incredibly flawed analyses--rather than minor nitpicks--the main contributions seem to be seminal none-the-less.
The more substantive article on ft.com is here: http://blogs.ft.com/money-supply/2014/05/23/data-problems-wi...

This shows the actual errors, correcting for the errors, and the new results.

If you get an unauthorized popup, search on Google and click the top link: https://www.google.com/search?q=data-problems-with-capital-i...

Thanks a lot. It is annoying that we can't see Piketty's answer without this popup.

I seriously hate the way this has been reported. No one seems to care about the facts or numbers. All news seem to be "The commie was wrong!" without even trying to explain what the criticism is about.

When I see the FT)reconstructed charts, I have a hard time accepting the notion that his errors were significant. He himself explains that he tries to track a quantity that is hidden from public view by using several indirect indicators. It is doomed to be approximate, but I don't see anything like a contradiction in this article. Just a mere correction.

Thank you! The linked bit is so devoid of information it's almost silly.
FT's accusation sounds dubious.

For example, Giles claimed:

"Piketty appears to have added random numbers to certain formula to bend the data toward his hypothesis." (from BI) and Giles: "A 2 is added because the number wasn't high enough — it didn't seem to fit what he wanted to show in his charts, so he just added 2 to it..." Source: http://static1.businessinsider.com/image/537fa370eab8ea427aa...

First off, it's not a random number. That 2 is an estimation from the two actual numbers from (Wolff 1994) Table 4. There was neither 1960 nor 1970 numbers available and only 1962 numbers (25.9% & 7.5%) available from the original paper. Hence, for the difference of that 2 and 8 years, Piketty first estimated the 1960 Top 1% wealth share number (31.4%) by (25.9 + 7.5 - 2). Then, the 1970 number is calculated based on the 1960 number and the ratio of "top 0.1% wealth share of 1960 and 1970, with the addition of that "2" which just took off for the 1960 number (2+31.4*10.4/8.7). I would say "2" is reasonable, even though it's arbitrary, to make the whole data series more smooth. Of course someone could use other estimate number for that two years, but it only makes the whole series more bizarre, and it does not change the pattern of that data series.

Second, FT's own conclusion even shows the patterns are almost the same with FT's claimed "correct" numbers or Piketty's. Piketty even pointed out other researches from Saez and Zucman published after his book also confirms his finding in the book.

All in all, FT just sounds like trying to sell more paper/subscription.

Posted links to piketty's response and more detailed explication of the issues on my blog: http://wcgortel.com/pikettys-problem
Can you get the data out of the website which has Piketty's response/a rundown of the errors? The FT site is not letting people who are not registered read those. And personally, I don't want to create another account just to read that information. (I imagine a good number of people here are similar)
I just updated with a couple more grabs. I don't want to pull the full article, but I added in the FT's long time series chart of income inequality and conclusion though.
Thanks!
If anyone is getting stopped by FT's registration page, you can log in as:

foo25@mailinator.com / Bugmenot

There are a lot of problems, they have been pointed by economists on every side of the spectrum right now, few months after the book has been published. More errors will be pointed during the years, but all the criticisms will be ignored, Piketty will be remembered as a genius for the time being and his ideas will be teached in schools and universities as if they were the God's words.

Just like happened with Keynes.

Or with Reinhart & Rogoff, who made a "mistake" in Excel that helped push disastrous austerity all over the world with a 90% debt-to-GDP cliff that didn't actually exist...
Reinhart & Rogoff are just technocrats that will be forgotten in two or three years.
if he's wrong in 50 years, and still talked about, he will still matter for a simple reason: if a bad idea is popular, you need to implement it and make it fail to prove it's a bad idea. you just learn by your errors. It's better than to not experiment at all.

economics and policy are influenced by history. sometimes you have to find arguments in economics that manage to change politics even if it's not an entirely good idea. that's how history unfold.

Well, Keynes is still wrong, and still talked about, and it was tested many times, failed them all, but, as economics is not testable, nobody can ever be convinced of anything, anyone can interpret the reality in the way he likes most and say it is proving your point.
that doesn't mean we should not try anything.
And Hitler. Let's not forget to bring up parallels to Hitler when trying to smear up someone. Godwin's law must be fulfilled, damn it!

I don't see "criticisms being ignored." The author is dialoguing with his critics. His response was published in this very article.

Keynes also dialogued. I'm not saying the criticisms will be ignored by the author, but by the "specialists" of the future.