In this age of metrics everywhere, evidence-based everything, big data...
I think what this story actually is, is a reminder that just because something is based on quantitative data, that doesn't make it 'objective' or automatically 'truth'.
Quantitative analysis rests on so many choices, as to how to measure, what measurements to use, what statistical formula to use, how to interpret what they say. Each of those choices can be mistaken -- _or_, even more troubling to the worldview that quantitative==objective truth, be subject to debate among reasonable and well-intentioned people about the best choices to make and the implications thereof.
I'm not saying it means there is no 'truth', and all research conclusions are equally valid. I'm saying that research conclusions based on quantitative data, no less than those based on qualitative information, are subject to debate and argument, not physical objective material reality simply because there were measurements and numbers involved.
An excellent book on this topic is "A Social History of Truth: Civility and Science in Seventeenth-Century England." It examines what "truth" means in a scientific context by going back to its roots as a discipline.
The tl;dr version is that knowledge (even "hard" scientific knowledge) cannot be meaningfully acquired by a single person independent of a community that makes common assumptions, since going to very first principles for even the simplest analysis is totally unwieldy. Take something like simple genetics — we rely on the testimony of a lot of people just to accept the utility of a basic Punnett square.
I much preferred Summer's assessment of Piketty to FT's: he was measured and responsible in his criticism, and correctly noted that it'll take years for serious academics to sift through the merit of the book. Alas, responsible discussion is hard to come by when it comes to data that challenges peoples' closely-held values.
In this age of metrics everywhere and big data, the data on inequality, a very important social issue, took Piketty and his collaborators, all professional economists, many years of research to assemble and publish.
This data is not even big: to paraphrase a joke I read on HN recently, you could do the analysis in Excel and it wouldn't even crash.
There is something to be said for systematically gathering a lot more data than governments collect now, especially income and wealth data in this case. Not because it benefits anyone economically right now, but because without collecting it now, it is that much harder to reconstruct it later.
The primary issue here is how you can define "wealth" or "rich" and "poor." Income's one measure, sure. But what about the trust fund kids? What about college kids?
The real problem with collecting data like this is that there's no real good source of it. Most of what you get has selection bias, and the other stuff you can get is riddled with holes. And then you try to do it across countries, where definitions change and governments college disparate data at differing granularities and it just gets...frustrating.
An economist's toughest job is finding good data. The analysis is the easy part!
And even then most people don't have a choice--they have to use what data they are "stuck with" and often times that data is provided by people with their own agenda. Public-disclosure requirements is highly politicised and fought-over by various well-financed special interests. And that is not even counting the lawyers, accountants, and advisors paid to dress things up even further after a law or rule is created.
I'm interested to know where you think the debate should lie. Do you believe that it should be debated empirically among economic journals or in the court of public opinion? Because the book is a compilation of the former and we are only seeing doubt in the latter.
Usefulness to whom, and in what context? In the case of the climate change "debate", which has significant parallels to this one, denial is useful to entrenched interests at the expense of everyone else — including, ultimately, those same interests. How "useful" is that, exactly?
Unfortunately, ultimately, the research, data and analysis isn't going to matter much.
Consider the debate on climate change, 98% of climate scientists agree that humans contribute significantly to climate change, but large numbers of Americans (including some of the most powerful politicians and talk-radio hosts) either deny climate change or claim that humans don't do much to cause climate change.
I expect the inequality debate to play out in a similar fashion. Many economists may agree with Piketty. However, the FT has thrown sufficient "doubt" on his findings (it doesn't matter that they compare past tax data with today's "survey" data etc).
This "doubt" provides sufficient ammunition for those who'd like to deny that inequality is on the increase. So the debate will end up as another one of those "he said, he said" topics in the media and people are likely to take sides based on which side of the ideological divide they are one.
The FT data objections are far from the strongest objections to Pinketty's work. Here are a few criticisms seem much more serious than nitpicking the data:
In brief, his theoretical mechanisms that supposedly predict and explain greater wealth accumulation are actually unsound, and if anything this seems to be the evolving consensus among economists who study the area.
It is even harder when there is an ideological block. In the first case, you get stonewalled, and ultimately ignored. In the case of ideology, you get shouted down, mobbed, and subject to various preemptive attacks intended to discredit and demoralize you before you even get to make your case.
Climate change and macroeconomics are really similar in this regard. I think the main reason there is so much controversy is because the science is so entangled with the politics, and that probably applies to the experts just as much as it applies to the average onlooker.
Climate research is nothing like economics. Politics is involved in how we should react to the scientific results, but that doesn't mean the science itself is "entangled" with politics.
You might just as well suggest that paleontology is entangled with religion because its findings offend some people's religious views.
It sounds like you're saying they shouldn't be entangled, which is reasonable, but under what definition of "entangled" are they not currently entangled?
If the analogy with climate science works, then the inequality debate will be even worse than the climate change "debate". The straws on which climate change denialism rests are so small because the level of evidence one can reach with physics is so particularly high. Economics probably won't ever reach the same standard of evidence.
To this day, people in politics are still having arguments about the relationship between inflation and fed interest rates, two of the most easily-observed quantities in the economy. Imagine what the inequality debate will be like in comparison.
Physics and weather forecasting are not similar endeavours. Nor is the relationship with fed interest rates and inflation transparent. FYI "people in politics" describes every economist. their businesss model relies on someone listening to them. Those people pay them money. And those are all "people in politics".
Weather forecasting is applied physics. The systems are big, some interactions aren't known precisely, and the boundary conditions are more complex than most physicists usually work with, but weather forecasting is definitely physics.
It's apparently all 100% ideological. I have enjoyed reading economics but past some point, when it starts to be policy oriented, the thinking just stops. They pick a team and sing the team song.
The thing about inequality is that it's so shockingly close* to a lognormnal/Pareto distribution that my feeble brain says "biology!" and I stop thinking of it as a problem. It starts to feel to me like when I used to argue with algebra if I don't.
*I cut & pasted some data set from the Internet and the r-squared was way above 0.9
And apparently, if you read enough semantics that aren't really there into history, it looks like the best cure for inequality is to simply stop everything and have World War II again. Cures Depressions, too!
We could not engineer a proper monetary policy, so we sacrificed one in 25 of the world's population instead. And this worked!
>>I expect the inequality debate to play out in a similar fashion. Many economists may agree with Piketty.
Whilst "many" economists most surely do agree with Piketty's data, it is closer to 50-50 than the 98% found in climate science.
Additionally, the number of economists who agree with Piketty's data but disagree with him over it's implications (that's the important part) will be smaller still.
Fortunately (or unfortunately) inequality and poverty have significant, damaging effects on the voting public. It's easy to buy into the idea that global warming is a conspiracy when you walk outside on a cool day. It's not so easy to accept the lies about inequality when you and all your friends and relatives start losing their jobs and have their homes foreclosed.
> It's not so easy to accept the lies about inequality when you and all your friends and relatives start losing their jobs and have their homes foreclosed.
Except when the man on the talk radio blames it on "inept Big Government" or "excessive taxation". Then it's super-easy.
Given the amount we spend on government, the inefficient way it's run, and the value the taxpayer gets for the dollar (compared to e.g. Sweden), it's a huge, brightly painted target. It is not all of the problem, but it is a problem.
This is a bit like blaming the murder on the bullet, not on the criminal who pulled the trigger.
The US government would be in a very different shape indeed, if it wasn't owned by the highest bidder. It might be inefficiently run for you; but for the special interests who own all the lobbyists in D.C., and who pay all the politicians - for them it's just fine as-is, thank you.
Several Asian countries like Singapore and Hong Kong have extremely high inequality rates yet far lower unemployment rates than the US. Where is your proof that inequality causes job losses?
Job losses as the mechanism of inequality would very little nothing to do with Piketty's claims, which are about concentration of wealth at the very top due to intrinsic properties of returns on capital. Piketty's theory doesn't involve this either causing or being caused by job losses. So seeing unemployed people and taking it as confirmation of his work would be incorrect.
(In any case if you are correct about direction of causation, then surely we should worry about how to reduce job losses than how to take wealth away from the rich.)
Dr. Tol was appointed to both the Copenhagen Consensus and the IPCC, is an accomplished and respected economist, and very often is a voice of reason for both sides of the debate.
I hope you can appreciate the meta-irony of declaring a blog post "ranty" to invalidate its findings and then going on to accuse the parent of proving a point about partisan ideology.
First, the blog post does not rise to calling the initial survey a hoax. So that addition of jstalin's, is well, completely ridiculous and inappropriate.
Second, the blog post is ranty (and many if not most scientists are ranty, in case you hadn't noticed). The associated comment manuscript isn't evidence of a hoax. I'm not surprised it wasn't accepted anywhere. There's no meta-irony in pointing that out.
The post seems like strong evidence that the finding of the paper is wrong, or at least not justified by the study performed. "Hoax" taken literally would imply deliberate malfeasance, and that doesn't seem to be proven
Where are you getting the 98% figure for belief that humans contribute significantly to climate change? Can you provide a citation? What exactly does 'significant' mean? Can you provide a citation for this?
a 44% wealth share for the top 10% (and a 12.5% wealth share for the top 1%, according to the FT) would mean that Britain is currently one the most egalitarian countries in history in terms of wealth distribution; in particular this would mean that Britain is a lot more equal that Sweden, and in fact a lot more equal than what Sweden as ever been (including in the 1980s). This does not look particularly plausible
Having lived both in Sweden and in Britain you don't even need data to know this isn't very plausible. That said, Sweden is getting a bigger gap between the highest earners and the lowest at a fast rate. "the most equal country in the world is becoming less so." -The Economist
http://www.economist.com/node/21564412
You're confusing stocks with flows. Wealth inequality and income inequality are completely different things, and Sweden being more unequal than the UK does not seem at all implausible. Here's one paper from a few years ago showing exactly that result: http://darp.lse.ac.uk/papersdb/Cowell_(Hills_Wealth_UK).pdf
Particularly the "12.5% for the top 1%" is ridiculous and obviously too low for any modern, capital-intensive, non-socialist society. The FT should know this from their Economics 101, but it seems their main objective here was to discredit Piketty...
The point is that the FT switched from one type of data to another, as per usual truth is probably somewhere in the middle, which would probably still back up Mr Picketty's analysis.
As many people have said, is the UK one of the most equal societies in the world? This seems very unlikely
Yeap... They should've read his bio a little better. This guy's from the most hardcore math school of the country. Less than a few hundreds accepted every year for the whole country ( and only based on a contest in which most top science students participate), and he got inside 2 years younger than normal. He isn't the typical economist. Attacking him on raw data analysis was probably the most daring thing to do.
Burden of proof is on you to prove that FT's data is shoddy since you made that claim. What specifically are you referring to and what sources are you citing?
There is a 10 page rebuttal .. but if you insist I précis the killer points for you then:
The first of two substantive changes the FT notes that greatly changes Pikettys European wealth trend would be the wealth estimate for Britain.
The FT choose a Fig of 44% they have got from a wealth survey carried out by the ONS (Britains govt statistical office) - compared to Pikettys 71%.
The 71% is from estate tax data from Britains tax authorities (HMRC) and is the same method used for the whole time-series.
The FT take that historical time-series then drop the 44% ONS figure in for the most recent time points.
So according to FT methods - Britain unlike the US and the rest of Europe has had declining inequality throughout its period of Thatcherite financial deregulation.
This is just stupid.
The second substantive difference the FT note is that Piketty should weight the 4 European series Britain, Sweden, France, Germany by the size of their economies rather than a straight average.
He could but it only makes difference to the overall European trend if you accept their crazy estimate of declining British inequality. Otherwise Briatin is similar to the other 3 and it make no difference.
Perhaps you would care to point to the part of the PDF that indicates the data are unreliable, and to comment on the plausibility argument, i.e. that the FT's analysis indicates that "Britain is currently one the most egalitarian countries in history in terms of wealth distribution" and in particular that it is far more equal than Sweden is or has ever been?
For me, the money quote comes from Economist's article on the subject:
Scott Winship, a scholar at the Manhattan Institute who disputes Mr Piketty's overarching narrative about inequality wrote on Twitter last night:
"I’ve spent time with Piketty U.S. wealth ineq[uality] spreadsheet and LOTS of time with his income data. He’s not up to funny business."
>He says the newspaper’s analysis rests on apples-to-oranges comparisons of past data from tax returns mixed with current data from surveys, which makes the conclusions they reach deeply flawed, and contrary to what a wide range of other studies have found.
1. I don't think that I or anyone else in this thread have to "prove" that the FT piece was shoddy in order to express that opinion. That said, I do agree that some support is needed.
2. I don't know that FT's data were shoddy, exactly, but Piketty's response certainly suggests that their review was less than thorough and their claims overblown.
That out of the way, Piketty points out a number of things about the FT analysis that, if true, would indeed suggest that the FT analysis was not very good. It is possible that Piketty's statements aren't true -- I'm not an economist, I'm only part of the way through Piketty's book, and I have not looked at the technical papers and datasets exhaustively (though I have looked at them) -- but it all sounds plausible to me, and it is all easily verifiable.
For those who do not want to wade through everything, the basic theme is this: at, it seems, every turn, FT notices an anomaly in Piketty's data and assumes that it reflects an error when, in fact, the anomalies are reflections of conscious methodological choices. FT then makes a big deal out of every "error" even though only one or two of them would make a meaningful difference, even if FT were correct in identifying them. More seriously, in most of these cases, FT could well have answered its own concerns by reviewing the copious technical notes that Piketty has made public. But, it would seem, FT either could not be bothered to consult them, or did consult them but chose not to disclose these explanations to its readers. FT then generates its own results based on its dubious "corrections" of Piketty's "errors" and does not bother to sanity check them. It fails to notice, e.g., that its analysis shows that Britain is more egalitarian than Sweden (hard to believe) and, moreover, that it is one of the most egalitarian societies in the history of the world (even harder to believe). It's not impossible, of course, that Britain actually is that equal, but a responsible researcher, one would expect, would have noticed that feature in its results and made an effort to come to grips with it. FT's failure to approach its own results with that sort of critical eye (the eye with which it read Piketty's results) to me suggests sloppiness or bias.
Some quotes:
...I compute decennial averages on the basis of the closest years available. This is clearly explained in the chapter 10 excel file (see sheet "TS10.1"). For instance, "1870" is computed as the average for years "1873-1877", "1910" as the average "1907-1908", and so on. These choices can be discussed and improved, but they are reasonably transparent (they are explicitly mentioned in the excel table, which apparently the FT did not notice), and as one can check they have negligible impact on long run evolutions.
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 any case, whatever adjustment one chooses to make to deal with this break in series is again going to have a negligible impact on long-run patterns.
The FT argues however that no explanation is given for some of the data construction....This is a surprising statement, because all necessary explanations are actually given in the technical research paper on which these series are based.
What it find somewhat puzzling in this controversy is the following: (i) the FT journalists evidently did not read carefully the technical research papers and excel files that I have put on-line; (ii) whatever adjustment one makes to correct for differential mortality (and I certainly agree that there are uncertainties left regarding this complex and important issue), it should be clear to everyone that this really has a relatively small impact on the long-run trends in wealth inequality. This looks a little bit like criticism for the sake of criticism.
What is troubling about the FT methodological choices is that they use the estimates based upon estate tax statistics for the older decades (until the 1980s), and then they shift to the survey based estimates for the more recent period. This is problematic because we know that in every country wealth surveys tend to underestimate top wealth shares as compared to estimates based upon administrative fiscal data. Therefore such a methodological choice is bound to bias the results in the direction of declining inequality.
Also note that a 44% wealth share for the top 10% (and a 12.5% wealth share for the top 1%, according to the FT) would mean that Britain is currently one the most egalitarian countries in history in terms of wealth distribution; in particular this would mean that Britain is a lot more equal that Sweden, and in fact a lot more equal than what Sweden as ever been (including in the 1980s). This does not look particularly plausible.
maybe you should read Piketty's rebuttal? or any of the other dozens of rebuttals that have been written. burden of proof is on YOU to not be ignorant.
Most data is released to the public in ways to make the data as use-less as possible. This is particularly true with economic data and SEC financial information. So, none of this type of "noise" or its analogue "noise cancellation" should really be taken at face value. The threshold for even disclosure of such mistakes is typically based on "material adversity", and this is the type of standard which people should look to in this case. The question is simply does it move the needle? Its not clear from reading these multiple threads that it does.
I think what this story actually is, is a reminder that just because something is based on quantitative data, that doesn't make it 'objective' or automatically 'truth'.
Quantitative analysis rests on so many choices, as to how to measure, what measurements to use, what statistical formula to use, how to interpret what they say. Each of those choices can be mistaken -- _or_, even more troubling to the worldview that quantitative==objective truth, be subject to debate among reasonable and well-intentioned people about the best choices to make and the implications thereof.
I'm not saying it means there is no 'truth', and all research conclusions are equally valid. I'm saying that research conclusions based on quantitative data, no less than those based on qualitative information, are subject to debate and argument, not physical objective material reality simply because there were measurements and numbers involved.