The Great Divide

March 15, 2009 12:31 PM

British academics Richard Wilkinson and Kate Pickett believe they've discovered the underlying cause of all modern society's ills: inequality. In their book, The Spirit Level: Why More Equal Societies Almost Always Do Better, they explain how health and social problems follow a strikingly similar pattern, being closely correlated with income distribution (pdf). To spread the word, they've founded The Equality Trust
posted by Elizabeth the Thirteenth (99 comments total)

Also, bears shit in the woods.
posted by kldickson at 12:36 PM on March 15 [5 favorites]

From the Times UK Online link: In California in 2004, there were 360 people serving life sentences for shoplifting. California has built only one new college since 1984, but 21 new prisons.

Anyone know if that's true?
posted by ornate insect at 12:44 PM on March 15

First, correlation's not causation. Second, apples aren't oranges. And finally, just because it's comforting doesn't make it so.
posted by MarshallPoe at 12:46 PM on March 15 [3 favorites]

also, three cliches in a row does not a stellar point make.
posted by ornate insect at 12:52 PM on March 15 [38 favorites]

California has built only one new college since 1984, but 21 new prisons.

That's one way to solve the dropout problem.
posted by srboisvert at 12:53 PM on March 15 [3 favorites]

First, correlation's not causation.

The authors are or were both highly qualified epidemological statisticians at major British universities. But I'm sure they'll be grateful for the heads-up anyway!
posted by game warden to the events rhino at 12:55 PM on March 15 [51 favorites]

Okay, half the work is done, now just to fix it.
posted by Krrrlson at 12:57 PM on March 15

From the Times UK Online link: In California in 2004, there were 360 people serving life sentences for shoplifting. California has built only one new college since 1984, but 21 new prisons.

Anyone know if that's true?

I can't say for sure, but it sounds non-crazy. Wikipedia on effects of CA three-strikes law. The Times numbers are high enough that they were probably calculated in a squirrely way, but the point is not crazy: after your 3rd felony offense, you are very likely to get 25-to-life; it doesn't matter whether your 3rd offense was (or all three offenses were) shop-lifting.
posted by grobstein at 1:02 PM on March 15 [1 favorite]

"It became clear," Wilkinson says, "that countries such as the US, the UK and Portugal, where the top 20% earn seven, eight or nine times more than the lowest 20%, scored noticeably higher on all social problems at every level of society than in countries such as Sweden and Japan, where the differential is only two or three times higher at the top."

This is clearly bogus. I'll eat all my shoes for dinner tonight if you can prove to me that the senior executives (much less CEO) at Mitsubishi, Mitsui, Sumitomo, and Yasuda earn only 30% more than the bottom 20% of Japanese society.

The authors are or were both highly qualified epidemological statisticians at major British universities.

Plenty of people have made careers on bogus statistics work.
posted by b1tr0t at 1:02 PM on March 15

Anyone know if that's true?

Not exactly. California has a three-strikes law that requires a sentencing court to impose a 25-to-life prison term on any offender convicted of a third felony. Under California law, shoplifting is a felony, so someone convicted of shoplifting as one of their three felonies could get a life sentence. At least that was true until 2004, when the Ninth Circuit determined that the application of the three-strikes law to shoplifting was unconstitutionally excessive. See Ramirez v. Castro, 365 F.3d 755 (9th Cir. 2004). I don't know where those 360 people who were serving life sentences for shoplifting are, but I expect they had their sentences modified pursuant to the Ninth Circuit's decision. Since 2004, there has been at least one other decision from the Ninth Circuit finding that the application of the three-strikes law to certain nonviolent felonies is unconstitutional. See Gonzales v. Duncan, No. 06-56523 (9th Cir. Dec. 30, 2008).
posted by monju_bosatsu at 1:06 PM on March 15 [4 favorites]

"...Japan, where the differential is only two or three times higher at the top."

This is clearly bogus. I'll eat all my shoes for dinner tonight if you can prove to me that the senior executives (much less CEO) at Mitsubishi, Mitsui, Sumitomo, and Yasuda earn only 30% more than the bottom 20% of Japanese society. [...] Plenty of people have made careers on bogus statistics work.

Yes, but you're clearly a statistical genius.
posted by robcorr at 1:06 PM on March 15 [24 favorites]

This is clearly bogus. I'll eat all my shoes for dinner tonight if you can prove to me that the senior executives (much less CEO) at Mitsubishi, Mitsui, Sumitomo, and Yasuda earn only 30% more than the bottom 20% of Japanese society.

The multipliers for top execs vs. bottom employees is much (orders of magnitude) less in Japan than in US, so it fits the pattern if overall inequality is lower, measured this way. But your problem is that you're using "senior executives" as your mental example of the top 20% of the income distribution; probably they are the top 2%.
posted by grobstein at 1:07 PM on March 15 [1 favorite]

ha oops
posted by grobstein at 1:08 PM on March 15

This is clearly bogus. I'll eat all my shoes for dinner tonight if you can prove to me that the senior executives (much less CEO) at Mitsubishi, Mitsui, Sumitomo, and Yasuda earn only 30% more than the bottom 20% of Japanese society.

But that's not the claim. The claim is that in Japan the top 20% earns two to three times what the bottom 20% earns. How does that match your statement here? And in any case, the senior executives are likely in the top 1%, where the differential is probably much higher.
posted by monju_bosatsu at 1:08 PM on March 15 [1 favorite]

By the way, this is what the WSJ reported:

On average, chief executives at Japanese companies with more than $10 billion in annual revenues are paid about $1.3 million a year, including bonuses and stock-option grants, according to Towers Perrin, a consulting firm, based on data gathered between 2004 and 2006. But chiefs in the U.S. are paid about $12 million, and chiefs in Europe are paid $6 million.

Hint: WSJ does not stand for World Socialist Journal.
posted by robcorr at 1:10 PM on March 15 [9 favorites]

kldickson: Also, bears shit in the woods.

MarshallPoe: And finally, just because it's comforting doesn't make it so.

By the way, I love how this thread has attracted both sides of the I Am Too Jaded To Read This So Will Just Dismiss It With A Sigh Instead brigade - the ones who think it's boringly obvious, and the ones who think it's obviously bogus.
posted by game warden to the events rhino at 1:15 PM on March 15 [8 favorites]

In Japan, the stigma that attaches to mental illnesses causes many of them to go unreported.

Also, the data on income equality for Japan are likely out of date; income disparity is widening and creating a gap between the rich and poor (a social phenomenon known as 格差社会（kakusa shakai).
posted by Gordion Knott at 1:15 PM on March 15

But at least we are divided equally.
posted by TwelveTwo at 1:16 PM on March 15

Incidentally, (in developed countries) equality increases in bad economic times. (Yay?)
posted by grobstein at 1:20 PM on March 15 [1 favorite]

Yes, but you're clearly a statistical genius.

No, just highly skilled at typos.
posted by b1tr0t at 1:21 PM on March 15 [1 favorite]

Also, my lunch of caviar and champagne must be impairing my ability to post coherently.
posted by b1tr0t at 1:22 PM on March 15

My non-scientific view, having lived in UK, US and Nordic countries, is that they're absolutely correct.
posted by Lord_Pall at 1:27 PM on March 15

Its worth specifying exactly what Wilkinson et al. argue and why it is important - and controversial. There are many ways that income inequality may be related to health. Among these are the following two:

(1) Individuals with lower incomes are less healthy. This is almost universally agreed upon, and supported by considerable evidence.

(2) Populations in which income is distributed less equally have poorer average health outcomes than those in which income is distributed more equally. This is hotly contested, with considerable disagreement over whether available evidence supports or refutes it.

Suppose, for the moment, that (2) is true. What are some of the possible explanations? Here are a few [reference]:

(A) Populations with unequal distributions of income will, by virtue of proposition (1) above, also have unequal distribution of health. The poorer health of the people at the bottom of the society will drag the population averages down more than the better health of the people at the top will drag it up, so the overall average will be lower.

(B) Whatever has led to the unequal distribution of income (say, less investment in social programs) will also lead to poorer overall health. So it is not that income inequality --> poor health, but that something else --> income inequality and something else --> poor health.

(C) Income inequality in itself leads to poorer overall health, through lower social cohesion, increased social stress, or some other mechanism.

There are two things to note about this debate. First, proposition (2) is hotly contested. Wilkinson has already published a book defending this thesis, as have others. But a number of careful researchers - many of whom were generally sympathetic to Wilkinson's general political outlook and too finding correlations between inequality and health - found little evidence for this proposition (here is a comprehensive meta-analysis of the evidence). In 2002, a BMJ commentary concluded that "the evidence for a correlation between income inequality and the health of the population is slowly dissipating. There is very little confirmation of such a relation outside the United States."

Second, even if proposition (2) now holds, it is worthwhile to figure out which of the three possible explanations is responsible (or if there is some other mechanism). Wilkinson, et al., like to argue some version of (C). That is, they want to say that inequality is bad for everybody's health, not just those at the bottom of the social hierarchy. But (A) and (B) seem more plausible as explanations, and are better supported by some available evidence. But - and this is the most important point - if either (A) or (B) are the explanation, then we can no longer say that inequality is itself bad for everybody's health. Instead, we can say either that some other social policy is bad for a number of things, or that inequality is especially bad for those at the bottom.

Wilkinson gets a lot of press for making an argument (inequality harms everybody, not just the poor) that appeals to liberals. But so far (I haven't read the new book) the argument has been flimsy at best, and it remains to be seen whether it is convincing to anyone who doesn't already think inequality is a Bad Thing.

The US and many other countries already tolerate extreme income inequality despite its obvious negative effect on crime rates, productivity, and the health and quality of life of large proportions of the population. Will a tenuous argument that it is also correlated with slight gradients in overall health really convince anyone that massive redistribution of wealth is necessary? I doubt it.
posted by googly at 1:37 PM on March 15 [26 favorites]

Make up your own correlations, courtesy of Hans Rosling.
posted by you at 1:38 PM on March 15 [2 favorites]

First, correlation's not causation.

The pdf in the post makes a compelling argument that, in the UK at least, government policy decisions have a large effect on the income inequality gap. If income inequality produces social problems rather than the other way around, then increases in the British Inequality Gap (BII) should be followed by increased disparity in the other social problems measured (literacy, death rates, depression, obesity, teen pregnancies, etc.).

If they really want to nail this argument, I'd like to see some of this time course data. The causation could very easily be going the other way here. Obese, illiterate, pregnant teens probably aren't going to be moving up the financial ladder any time soon. Just sayin'.
posted by Nquire at 1:40 PM on March 15

Will Durant said over and over again that when the gap between rich and poor gets too large one of two things happens: either the wealth is redistributed by legislation or the poverty is redistributed by revolution.
posted by RussHy at 2:08 PM on March 15 [10 favorites]

And finally, just because it's comforting doesn't make it so.

Indeed. The sooner people accept that their station does not make them any more deserving of a comfortable life than those who can't afford one, the better everyone will be.
posted by 0xdeadc0de at 2:12 PM on March 15 [2 favorites]

This is interesting.

What strikes me first is that the authors clearly have a very specific kind of equality in mind, or at least a small few kinds - namely, economic and (to some extent) political equality. And I don't know if you can draw from their data that equality is an absolute good - no matter how much our societies may enshrine it; they show rather that our societies (in the western and developing world) are neither as equal as we think they are nor equal enough to be really just or healthy.

I think that's a pretty good point.

I also have a feeling that there have been several new universities in California since 1984. At the very least, I know UC Merced started in 2005. I somehow doubt that that's the only new one in 25 years - California is a pretty big state with a thriving university system. Also, the "x new colleges compared to x new prisons" comparison is a fallacious one to draw on its face - it would take a lot of work to demonstrate a correlation between the quality of education and the number of new universities being built, even aside from the fact that California probably has more colleges and universities than any other state in the union; there are 133 colleges in the California Community Colleges System and the California State University System alone, and those combined serve over 3 million students. Of all the fifty states, Californians are the fourth most likely to be students of some kind. I think that speaks pretty highly of their educational system.
posted by koeselitz at 2:14 PM on March 15

Proletarier aller Länder, vereinigt euch!
posted by oaf at 2:25 PM on March 15

I Am Too Jaded To Read This So Will Just Dismiss It With A Sigh Instead brigade

Sorry, I don't belong. I've worked on this stuff for some time. And I'm pretty good with stats too--I can even explain Gini coefficients. But who the heck cares. These well-meaning folks have claim to have found a correlation between some measure of inequality and some other measures of things across an incredibly heterogeneous set of instances. So that's enough to call into question the correlation itself, at least in MHO. But even if we put that aside and admit that there could be a correlation between inequality (what kind?) and some other things (which ones?), we still have to come up with a mechanism. They don't, and that's probably because there is none.

And what the heck does the "do better" mean in the title of the book. "Equal Societies Do Better." That's what they say. In terms of income and wealth distribution, the US is and probably has been for decades the most unequal society the Earth has ever known. It has also raised more people out of poverty and into the educated middle class than any society ever known. The US also gave the world what was arguably the the first successful democracy, the first trans-national political culture (very much admired in these days of resurgent nationalism), and countless achievements in science and technology. If inequality produces all that (and of course it doesn't), then the world needs more of it, not less.

Of course we--humanity, that is--experimented with the "solution" proposed by these nice folks: radical equality. That didn't work out very well, as you may remember (cf. "Soviet Union"). And here the mechanism is clear: if you take people's freedom away in order to make them equal, you are both going to impoverish them and make them resentful.

It's very nice to think that the fact that some people have a lot and other people have less is the root of all evil, or even most of it. Too bad it's false.
posted by MarshallPoe at 2:39 PM on March 15 [6 favorites]

googly: thanks for chiming in.

My own view, having lived in both UK and NL, is that a narrower income distribution is in some way related to a higher overall standard of living. However, there are all sorts of cultural factors at work in both societies, so I see it as more of a chicken and egg relationship. A relatively equal society is likely to be better organized, with a stronger infrastructure and better community support networks, while stratified cultures that encourage walled gardens for the fabulously wealthy are also more likely to tolerate urban squalor for the poor.

The problem with this research is as you say: it doesn't necessarily establish cause and effect. I can see that the public health argument could be very tenuous, and since that's the discipline from which it originated, it's very interesting to get the inside track.
posted by Elizabeth the Thirteenth at 2:41 PM on March 15

And it's nice to see Marshall Poe channelling the spirit of Harry Lime.
posted by Elizabeth the Thirteenth at 3:00 PM on March 15

First, correlation's not causation. Second, apples aren't oranges. And finally, just because it's comforting doesn't make it so.

That's not true. Correlation does prove causation, it just doesn't tell you what direction it goes in (or if there is a third hidden cause). People throw around "correlation is not causation" as if correlation disproves causation or something. It's really ridiculous.

Speaking of Japan, did you know the CEO of Toyota makes less then a million dollars a year?
posted by delmoi at 3:07 PM on March 15 [1 favorite]

It has also raised more people out of poverty and into the educated middle class than any society ever known.

They seem to be asserting the opposite, and they have evidence, Marshall ; you can't just reassert the patriotic legends you learnt at your father's knee by way of response.

I agree. though, that more explanation is needed. The assumption seems to be that the main reason for inequality being harmful is that it makes people unhappy, but is that it? It could equally well be that societies where people feel insecure and hence unhappy tend not to establish redistributive social arrangements. Maybe if you feel you're on the edge, you feel you have to look after number one, whereas people in Sweden feel so safe they feel they can afford any amount of social spending. I'm not saying that's true, but there are alternative hypotheses here which are not being investigated.
posted by Phanx at 3:09 PM on March 15 [1 favorite]

First, correlation's not causation. Second, apples aren't oranges. And finally, just because it's comforting doesn't make it so.

The majority rule trumps all that. Besides, you're inaccurate about correlation and causation.
posted by Brian B. at 3:11 PM on March 15

In terms of income and wealth distribution, the US is and probably has been for decades the most unequal society the Earth has ever known.

I'm willing to bet that this award goes to, well, China.
posted by vivelame at 3:17 PM on March 15

Correlation does prove causation

Argh.
posted by Phanx at 3:18 PM on March 15

Elizabeth the Thirteenth: not Harry Lime, but George Orwell.

Which prompts a further thought: one of the signal contributions of the U.S. to the world is the popular acknowledgment that the prosperity of others need not be injurious to you, at least under the right conditions. On the contrary, the prosperity of others can help you--don't take what the prosperous have by force, but learn from them, do what they do, and you too will be prosperous. This ethic has worked awfully well in the case of the U.S. It's opposite, that inequalities are necessarily injurious, didn't pan out now, did it?
posted by MarshallPoe at 3:22 PM on March 15

In terms of income and wealth distribution, the US is and probably has been for decades the most unequal society the Earth has ever known.

It is? Seriously? Even with our leftover socialist programs like Social Security and Medicare? I'm pretty sure there are countries in the third world with worse income inequality, and with no social programs whatsoever. But please feel free to dispute this.

I think I'm more troubled by your implication that it's our inequality that Makes. America. Great. Not, you know, our investment in public education, our democracy, or any of that commie shit like that.
posted by emjaybee at 3:22 PM on March 15

California has built only one new college since 1984

This has to be incorrect unless they're using some restrictive definition that doesn't include community colleges (maybe just the Cal State system?) and UC campuses.. If Folsom Lake College was built before 1984, so was the iPhone, and UC Merced was opened in 2005.

I'd be willing to bet that there have been not only 21, but over 40 total post-secondary institutions built in California since 1984.

the underlying cause of all modern society's ills

All? Seriously?

I'm pretty ready to believe in the basic thesis that inequality can cause some social problems, but that's a pretty extraordinary claim.
posted by weston at 3:24 PM on March 15

I'm willing to bet that this award goes to, well, China.

India, anybody?
posted by Elizabeth the Thirteenth at 3:27 PM on March 15

Marshall Poe: In terms of income and wealth distribution, the US is and probably has been for decades the most unequal society the Earth has ever known

I'm confused..you acknowledged the gini coefficient, but then made this erroneous claim. The United States is by no means the most unequal society on earth. Virtually all of Latin America has gini coefficients significantly higher than the US. (UNDP 2008 Report)

It has also raised more people out of poverty and into the educated middle class than any society ever known

China has seen more people rise out of extreme poverty (albeit into a somewhat different middle class) in the last two decades than the United States has citizens. (World Bank)

inequality (what kind?) and some other things (which ones?).....we still have to come up with a mechanism. They don't, and that's probably because there is none.

Googly did a great job of summing up the suggested mechanisms, even though he was casting some doubt on the idea. Wilkinson's work is very clear on what kinds of inequality we're describing and what the potential outcomes are. It is not nearly as vague as you make it out to be.
posted by Adam_S at 3:30 PM on March 15 [5 favorites]

"Correlation does prove causation, it just doesn't tell you what direction it goes in (or if there is a third hidden cause)"

No it really, really doesn't. IQ and foot size are strongly correlated but one does not cause the other. But maybe you just phrased that really poorly.
posted by 517 at 3:31 PM on March 15

I'm willing to bet that this award goes to, well, China.

It's opposite, that inequalities are necessarily injurious, didn't pan out now, did it?

I'm pretty sure there are countries in the third world with worse income inequality,

Yes, and I've always felt in my heart that Switzerland is a flat, marshy country situated in the Southern Hemisphere, and no amount of data is going to cause me to review that belief.
posted by Phanx at 3:31 PM on March 15

It's very nice to think that the fact that some people have a lot and other people have less is the root of all evil, or even most of it. Too bad it's false.

Which is itself the comforting self-deception employed by the likes of Charles Murray and right-wingers of all stripes.

The United States has the highest incarceration rate of any nation. What is the origin of crime? Does anyone deny that the wealth disparity has anything to do with it? That it is the primary factor?

And what affect has public education had in the rise of the middle class? Strong property rights and the inequality of wealth have existed in many other nations, and yet they haven't given birth to a large and robust, well educated, technically skilled populace. No, the European socialist democracies are our closest neighbors in that regard. What other shared characteristics would contribute to this result?
posted by 0xdeadc0de at 3:36 PM on March 15 [1 favorite]

I think I'm more troubled by your implication that it's our inequality that Makes. America. Great.

You shouldn't be troubled because I explicitly say the contrary in the post: "If inequality produces all that (and of course it doesn't), then the world needs more of it, not less." Inequality didn't produce the greatest engine of human prosperity in history, that is, the economies of Western Europe and United States. Just what did is a long and complicated story. But I'm pretty sure that if the U.S. had fallen under the control of a government that believed (as suggested by this research) that inequality 'makes people unhappy' then we would still be in the kind of pre-modern poverty that seems to strangle much of the Communist and post-Communist world today. Liberty alone doesn't make a country prosperous, but lack of liberty alone will almost certainly impoverish it.
posted by MarshallPoe at 3:38 PM on March 15

I apologise for the intemperate tone of my remarks above. My concentration was affected by the high-pitched whine emerging from the corpse of David Hume after it hit 3,000 rpm.
posted by Phanx at 3:54 PM on March 15 [2 favorites]

Scroll down for a list of prisons opened in California since 1984. I'm still looking for similar stats on public (i.e., state) universities/community colleges opened in the same time period, but my google-fu is weaker here.
posted by rtha at 4:10 PM on March 15

googly, great comment, I'm very much in agreement.

The only thing I could possibly add (or at least reinforce because you did mention it) is poverty has so many potential externalities, and therefore cost to individuals who are not poor but live within the same society. Even on the most crude level - that of 'national effiency' - poverty has long been recognised to affect whole-society outcomes. Because of this, I would lean much more towards accepting the premise that high inequality (and therefore poverty) is a negative factor in the success of a society.
posted by Sova at 4:23 PM on March 15

The figures look convincing at first, but what if the low rates for say mental health or drug convictions are simply a result of some countries having less effective diagnosis/detection rates.
It is notoriously difficult to compare these types of statistics across countries.

Also to follow up googlys comment, the countries at the top of those graphs all have liberal policies for admitting immigrants - it would be interesting to replace the figures for income inequality with immigration figures and see how they compare.

Perhaps the real difference here is that immigrants tend not to choose cold Nordic countries?
posted by Lanark at 4:34 PM on March 15

From the article:

What is it about unequal societies that causes the damage? Wilkinson believes the answer lies in the psycho-social areas of hierarchy and status. The greater the differential between the haves and have-nots, the greater importance everyone places on the material aspects of consumption; what brand of car you drive carries far more meaning in a more hierarchical society than in a flatter one. It's the knock-on effects of this status anxiety that finds socially corrosive expression in crime, ill-health and mistrust.

This seems like an especially poor explanation. As the rich get really super rich, like really extraordinarily rich, their consumption takes on dimensions that the lowest 20% of the economic spectrum are probably not even aware of, frankly. Weekend private flights to Monte Carlo, bespoke shoes from Italy made of the skins of obscure animals, a personal museum full of contemporary art by artists that only museum boards-of-directors have ever heard of.

I think the idea that status anxiety is causing all these problems is actually quite quaint; it hearkens back to something like Gilded-Age New York, when the wealthy had to walk the same streets as (or at least drive past) the hoi polloi, and one's social status could be assessed and compared in terms of clothing and grooming. Now the really super-mega-wealthy never have to be exposed to the commoners, and more to the point the commoners sure as hell don't have access to the really fabulous indicia of their extreme wealth, like garages full of Ferraris, private ice rinks, personal jets, etc.

How would someone making $10,000 a year even become aware of consumption like that? I mean, this is way beyond "Cribs." I suppose there's a Conde Nast publication about it, but if you're really just scraping by you probably can't even afford the magazine.

(By the way, I'm not trying to make the case that "poor people are stupid and ignorant," I'm just trying to call into question the idea that really extreme wealth disparities are necessarily broadcast in such a way as to make status anxiety a reasonable causal mechanism here.)
posted by rkent at 4:36 PM on March 15 [1 favorite]

Anyone know if that's true?

The California State University system has established three new campuses since 1984: San Marcos, Monterey Bay and Channel Islands.
posted by basicchannel at 4:41 PM on March 15

Which prompts a further thought: one of the signal contributions of the U.S. to the world is the popular acknowledgment that the prosperity of others need not be injurious to you, at least under the right conditions. On the contrary, the prosperity of others can help you--don't take what the prosperous have by force, but learn from them, do what they do, and you too will be prosperous. This ethic has worked awfully well in the case of the U.S.

I lol'd.
posted by kittens for breakfast at 4:48 PM on March 15

How would someone making $10,000 a year even become aware of consumption like that?

It goes beyond Cribs, all the way to... Cable TV, and the ubiquitous advertising and media that is modern life. Just because you make $10,000 a year doesn't mean you don't still see the tabloids in the grocery stores.
posted by tybeet at 4:52 PM on March 15

Even on the most crude level - that of 'national effiency' - poverty has long been recognised to affect whole-society outcomes. Because of this, I would lean much more towards accepting the premise that high inequality (and therefore poverty) is a negative factor in the success of a society.

Poverty != Inequality

I suspect that a good argument can be made that poverty is harmful to society as a whole. From that, you could easily conclude that a mechanism which makes it difficult or impossible to fall below a meaningful poverty line would create value for everyone.

Inequality, though, is what you expect to find in a dynamic economy. Even if you manage to bring everyone above the poverty floor, you would expect to see economic inequality if achievement is possible. If you don't have inequality, your economy is probably structured such that achievement is not incentivized.
posted by b1tr0t at 4:53 PM on March 15 [1 favorite]

Justice as Fairness.
posted by voltairemodern at 5:14 PM on March 15

That economic inequality greater than x is undesirable does not imply that less inequality is always more desirable. Perfect economic equality is extremely undesirable; it's the economic equivalent of a frozen iceberg. Nobody will be able to provide to anyone something that the first person does not have. Inequality must exist in order to provoke transactions: A must have something B wants, and vice versa (or at least, A and B must each have something that someone wants), for there to be any economy at all.

On the other hand this comparison, the foundation of the desire to perform transactions, can also cause problems. A sees what B has (a guitar, say), and desires it. Had A never become aware of the guitar's existence, A would have never felt that desire. If it's possible for A, given sufficient effort, to obtain B's guitar, or a guitar from elsewhere, then A has the ability to make a choice to obtain the guitar or not. But if A cannot, by any means whatsoever, obtain a guitar, she suffers an unfulfilled desire. In a sufficiently unequal society there are a vast variety of guitars, of varying desirability (including food and basic shelter), and a vast number of As, with varying desires. Some of these will reach a threshold of willingness to engage in socially damaging activity in order to obtain the thing. The more often this happens, the more socially damaging activity occurs.

There are ways around it, of course. Arbitrarily unequal societies can exist; North Korea is the ideal example, a near-perfect inequality where one person has approximately everything and everyone else has approximately nothing. This regime is maintained with almost all of the effective social mechanisms for doing so: isolation, absence of information on which to form comparisons, zero social mobility, irrelevance of merit, etc.

Less unequal societies use the same mechanisms, for good or ill: Hindu castes, medieval European "birthright", and so on. Smaller sub-societies use them too: the cult brainwashing of Scientology, the union-busting activities of Walmart, the view of priest as path to God in Catholicism, the chain of command in an army. If B can do something A can't, there must be a reason. The less compelling the reason, the unhappier A is. If A desires what B has (actually the same is true for avoidance of unpleasant things that B suffers), fulfilment of that desire can be measured on a spectrum from impossible, difficult, possible, easy, or certain. The more desires A has, the more difficult their fulfilment, and the more arbitrary these barriers to fulfilment seem to A, the more unhappy A will be.
posted by aeschenkarnos at 5:16 PM on March 15 [6 favorites]

I'm not sure why Hume, of all people, would spin in his grave over the claim that correlation proves causation, phanx. Hume, after all, believed that we know nothing of causation at all other than "constant correlation" allied with temporal order.

As for this: No it really, really doesn't. IQ and foot size are strongly correlated but one does not cause the other.

You simply ignore his option of "a third hidden cause." In this case, IQ and foot size would both be produced by the third hidden cause of overall good nutrition.

Where you have a constant correlation, there must be causation. Either the one of the two things causes the other, or they are both caused by a third thing. I'm puzzled that that statement seems to have rocked so many people's worlds.
posted by yoink at 5:26 PM on March 15 [2 favorites]

their consumption takes on dimensions that the lowest 20% of the economic spectrum are probably not even aware of, frankly.

There are whole industries devoted to making sure we understand the stagering wealth of the ultra rich. Lifestyles of the Rich and Famous. Gossip-oriented mags like Dolly! or Women's Weekly. Most car mags focus less on cars ordinary people will ever own and more on multi-million dreammobiles like the Veryon. If you follow football you know all about Abramovich and the other tycoons who buy football teams and pump quite literally billions of dollars into their playthings. Larry Elison's superyachts are hardly a state secret.

Hell, a whole chunk of popular hip-hop is guys bragging about how much they've got now they've made it big.
posted by rodgerd at 5:37 PM on March 15 [1 favorite]

I'm puzzled that that statement seems to have rocked so many people's worlds.

I'm not. Too many people have taken the Internet smart-arse line and mistaken it for profound wisdom.

I'd love to know how the "you aren't draw any conclusions from correlation" crowd get through life. Really. "Hey, when I stick my hand into fire, it hurts every single time. But correleation doesn't imply causation. I'll just leave my hand in there until someone works up a complete model of how fire can create the sensation of pain."
posted by rodgerd at 5:39 PM on March 15 [2 favorites]

I'd love to know how the "you aren't draw any conclusions from correlation" crowd get through life.

Yeah--it's such a great piece of half-smartness. Correlation may not equal causation, but when you consistently find the one, you'd better adopt the other as a working hypothesis.
posted by yoink at 5:45 PM on March 15

MarshallPoe : But I'm pretty sure that if the U.S. had fallen under the control of a government that believed (as suggested by this research) that inequality 'makes people unhappy' then we would still be in the kind of pre-modern poverty that seems to strangle much of the Communist and post-Communist world today.

The problem here is that the alternative to inequality is not communism and that GDP per capita is an inadequate measure of human well-being. Many of the Western European nations using this same 'engine of human prosperity' have veered towards less economic inequality than the United States and have higher standard-of-living measures (life expectancy, education, poverty) virtually across the board. Broad social inequality is not the inevitable consequence of a well-oiled economy; it's the result of endless conscious policy choices that states make.
posted by Adam_S at 5:55 PM on March 15 [1 favorite]

"Correlation does prove causation, it just doesn't tell you what direction it goes in (or if there is a third hidden cause)"

No it really, really doesn't. IQ and foot size are strongly correlated but one does not cause the other. But maybe you just phrased that really poorly.

Correlation does prove causation, it just doesn't tell you what direction it goes in (or if there is a third hidden cause).

Or maybe you just can't read very well. I mentioned the possible third cause right in that comment. If IQ and foot size are actually correlated then there must be some third cause (or network of interrelated causes, which we can think of as a single cause). Just because you can't figure out, or don't know, what that might be doesn't mean it doesn't exist.
posted by delmoi at 6:04 PM on March 15

If it doesn't rule out another cause, then it doesn't prove causation.
posted by robcorr at 6:21 PM on March 15 [1 favorite]

This seems like an especially poor explanation. As the rich get really super rich, like really extraordinarily rich, their consumption takes on dimensions that the lowest 20% of the economic spectrum are probably not even aware of, frankly.

Nice attempt at misdirection. The highest 20% aren't aware of it either, since they are even less likely to pay attention to tabloid style media. Now, the lowest 20% is certainly aware of the relative ease and comfort of the lives of the highest 20%. They're certainly aware of the expensive things that they own. And they are most certainly aware of the power imbalance created by the gradient between "needing" and "having".

Every income level is aware of status afforded by wealth disparities. There is every bit as much a "bling culture" among yuppies as there is among rap aficionados. You talk as is you've never heard the phrase "keeping up with the Jones". Most people consider that behavior a character flaw. It's wasteful, it's tacky, it's ultimately self-destructive, but there are also those who cheerleader it anyway. Those who, while admonishing us for our "class warfare" and "politics of envy", give speeches about how envy is necessary to aspire greatness and fuel the economic engine.

This drive to acquire exorbitant wealth, wealth created out of nothing, wealth based entirely on the assumption that there will be an increase in desire for ever more exclusive, sought-after evidence of wealth, is what has nearly destroyed the world's economy. Social status is a god damn cancer of the human psyche.
posted by 0xdeadc0de at 6:23 PM on March 15 [1 favorite]

Delmoi, correlation, in the case of IQ and foot size, has to do with age. As age increases so does foot size and IQ in a general sense. However, an increase in foot size does not cause an increase in IQ and a decrease in foot size does not cause a decrease in IQ, they do not cause each other as I could lop off someone's foot and it would not alter their IQ if I did it right.

Just because something is strongly correlated with another event does not prove that those two events are related by a third hidden event at all, because it could just be random noise. Like the winning football team that correctly predicts the party of the president that will be elected. There is no directionality or third hidden event there.
posted by 517 at 6:29 PM on March 15

when the gap between rich and poor gets too large one of two things happens: either the wealth is redistributed by legislation or the poverty is redistributed by revolution.

Sadly, most of history is littered with examples where this is simply not the case.
posted by Civil_Disobedient at 6:29 PM on March 15

Hell, a whole chunk of popular hip-hop is guys bragging about how much they've got now they've made it big.

That's actually kind of my point, though - hip-hop bragging focuses on relatively small-potatoes stuff like owning Bentleys and gold jewelry. In the grand scheme of things, most hip hop stars a lot closer to middle class than they are to ultra-wealthy.

I'm a little more convinced by 0xdeadcode's explanation... it does definitely seem to be the case that, no matter how high you get in the wealth/income game, someone is always willing and ready to rub in your face how much more they have.
posted by rkent at 9:17 PM on March 15

Like so many things, this thread, if not the study, suffers from a failure to define terms, and, as usual, means much heat but little light is created while people argue their meandering way further and further from the actual issue at hand.

Man it's tiring sometimes.
posted by stavrosthewonderchicken at 9:29 PM on March 15

Social status is a god damn cancer of the human psyche.

It's also what gets you laid, apparently. So I don't see it going away any time soon.
posted by Crabby Appleton at 10:40 PM on March 15

Social status is a god damn cancer of the human psyche.
The analogy to cancer is a good one; cancer is also an extreme form of a necessary process.
posted by aeschenkarnos at 10:55 PM on March 15

Also to follow up googlys comment, the countries at the top of those graphs all have liberal policies for admitting immigrants - it would be interesting to replace the figures for income inequality with immigration figures and see how they compare.

I don't think you'd find much of a correlation. Looking at the immigrant population as a percentage of total population, we see Switzerland (23%), Australia (20%), and Canada (19%) near the top (leaving out the Israel), followed by New Zealand (15%), the USA (13%), Sweden (12%), Germany (12%), and so on (the UK at 9%, down to Japan at 2%).

Canada's immigrant population is high, yet they end up in the middle of most of the charts. Conversely, the UK has a relatively low immigrant population, but ends up near the top of the charts. The US, Sweden, and Germany have very similar immigrant populations, but the US is at the top of the charts, Germany is somewhere in the middle, and Sweden is right down at the bottom.
posted by ssg at 11:27 PM on March 15 [1 favorite]

xkcd weighs in on correlation vs causation.
posted by ghost of a past number at 6:44 AM on March 16

There is very little confirmation of such a relation outside the United States.

You know, it seems fairly easy to see how this correlation might hold in specific instances and yet not in others. A nation that provides a strong social support net might be providing a lot of the benefits (health care, cheap and efficient public transportation, quality education, public safety, etc.) that only comes with money in other nations.
posted by Durn Bronzefist at 7:00 AM on March 16

Ah, apologies. I've got it now. Inequality is bad for everyone, not just those who would be directly affected by lack of a social safety net. Yeah, extraordinary claims and extraordinary evidence, methinks. Correlation is a fine thing, but you at least want to be able to theorize causal ties. I see only ideologically wishful ones here.
posted by Durn Bronzefist at 7:13 AM on March 16

Delmoi, correlation, in the case of IQ and foot size, has to do with age. As age increases so does foot size and IQ in a general sense.

So, in this reading, age is the "third hidden cause" that Delmoi specified. Is there something about your computer that renders the phrase "third hidden cause" invisible, or something?
posted by yoink at 9:11 AM on March 16

"Is there something about your computer that renders the phrase "third hidden cause" invisible, or something?"

I see that you fell right into that one.
posted by 517 at 10:46 AM on March 16

I see that you fell right into that one.

Because of your subsequent paragraph?:

Just because something is strongly correlated with another event does not prove that those two events are related by a third hidden event at all, because it could just be random noise. Like the winning football team that correctly predicts the party of the president that will be elected. There is no directionality or third hidden event there.

But that's, frankly, silly. What's at stake here is constant correlation. The point about short run coincidences like sports results and political events is that a large enough sample will demonstrate the absence of correlation. If you found a sufficiently large sample that demonstrated genuine correlation, you would have proved a causal link between the two events: either one would cause the other, or both would be caused by some common thing.
posted by yoink at 11:13 AM on March 16

"Because of your subsequent paragraph?"

No, because age, foot size, and IQ don't cause each other, they are just correlated. It comes down to a fine point but the point is still there.

"But that's, frankly, silly. What's at stake here is constant correlation."

That's an argument from a perfect data perspective and moves from statistics into parameters. If you had perfect data, there wouldn't have been a strong correlation in the first place. But no one has perfect data so there is noise. And even if you had perfect data their would still be noise.

Stay with me moment while I talk out of my ass to make a point. I remember reading some study done several years ago that demonstrated a statically significant greater number of very successful athlete were born under the sign of Aries than any other zodiacal sign. Let's just assume it wasn't random noise because the researchers managed to filter the noise out. Now you and I know that being born under the sign of Aries has no impact on athletic ability because the constellation Aries is composed of stars that are hundreds of millions if billions or trillions of mile away from us and don't have any appreciable effect on humans on earth. But the correlation is there, so lets look for some hidden third reason why this might occur. Well, Aries births are in the spring so maybe it's conception time or birth time or mothers that have better genes are hornier in the late summer or...

Now we're looking at things completely unrelated the constellation Aries, so saying that a correlation proves some third hidden cause is basically saying nothing at all. It's like saying that it's true, false, or something else. It's an argument that's unfalsifiable because it covers everything and it will allow you treat noise as evidence.
posted by 517 at 12:04 PM on March 16

Now we're looking at things completely unrelated the constellation Aries, so saying that a correlation proves some third hidden cause is basically saying nothing at all.

Where did the constellation Aries per se ever come into this? You switched the terms in your example. The question at stake was "is there some causal relationship between being born under the sign of Aries and having a certain athletic ability."

Let us say that we determine that the correlation is statistically significant. Does this imply causality at work? Yes, it does. Let us say that the underlying reason turns out to be a matter of being born at a certain time of the year. Well, then the time of year is the cause for the position of the constellation Aries in relationship to the earth's orbit around the sun (i.e., the cause of that birth being "under the sign of Aries") AND the cause of the baby being exposed to whatever it is in that particular seasonal that makes for superior athletic ability.

You keep trying to smuggle back in the simplistic claim that "correlation between A and B proves that A causes B" claim in order to have an easy straw man to defeat. Of course this correlation wouldn't prove that the position of the constellation Aries in relation to the night sky causes superior athletic ability. But equally, it does prove that some third thing is at work that causes both A and B to correlate.

As for the argument about "perfect data": you do realize that Hume proved pretty conclusively that "causality" is something we don't, in fact, understand. In other words, causality is only ever "known" statistically. We never have "perfect" knowledge of causal relationships, we only ever have "historical" knowledge ("A has always caused B in the past, therefore I expect and hope that it will continue to cause B in the future"). Rolling billiard ball A into billiard ball B "causes" billiard ball B to move....until it doesn't. If the laws of the universe cease to operate tomorrow as they have done in the past, we will have nothing to say except "huh...well, that's an interesting addition to the data set."

Oh, and "aaaarrrgh!"
posted by yoink at 12:54 PM on March 16

I tried.
posted by 517 at 1:15 PM on March 16

I tried.

If you mean you tried to consistently misunderstand a pretty straightforward argument, then I'd say you not only tried, you succeeded.
posted by yoink at 1:19 PM on March 16

Okay, now you're not just wrong, you're also being a jerk.
posted by 517 at 1:28 PM on March 16

Perhaps rather than patronizingly sighing "I tried" you'd actually point out what you think is wrong in the argument.
posted by yoink at 1:37 PM on March 16

"I tried."

Wasn't meant to be condescending to you. It was more along the lines of, "I've got other stuff that needs to be done today and I bet you do too. This little point isn't important enough for both of us to spend our time on. Have a good day."
posted by 517 at 1:59 PM on March 16

Athletes are more likely to be born earlier in the year - because children who are the older in their year group are more likely to excel compared to their year group, and then be selected for special training. Malcom Gladwell wrote about this recently.

So the Aries and athletic thing have a common cause - the correlation isn't meaningless.
posted by jb at 3:16 PM on March 16

What a pointless argument over semantics. When someone says "correlation != causation" they are generally correcting someone who erroneously concluded A caused B without considering a third (C) factor -- not saying the third factor doesn't exist. This is indeed a mistake people make often (either from lack of intellectual rigour or honesty), hence the phrase comes up a lot. What else could you say when you observe someone jumping to conclusions about causality that isn't just a long-winded way of saying the same thing?
posted by cj_ at 3:26 PM on March 16

Also, if you want to argue semantics, I think it's worth noting that "correlation does not imply causation" and "correlation is not causation" (what the OP actually said) are two different things. The latter is always true, of course.
posted by cj_ at 3:32 PM on March 16

That's an interesting point, cj. The problem is that the quip misstates the error. It suggests that you simply can't (ever) conclude causation from correlation, whereas in fact you can't conclude causation from anything else (ultimately). The actual error is concluding a particular causal mechanism from a particular correlation, and to show that this is a mistake requires a more specific argument.
posted by grobstein at 3:40 PM on March 16

When someone says "correlation != causation" they are generally correcting someone who erroneously concluded A caused B without considering a third (C) factor -- not saying the third factor doesn't exist.

Not in this case:

saying that a correlation proves some third hidden cause is basically saying nothing at all.
posted by yoink at 4:12 PM on March 16

The problem is that the quip misstates the error.

I'm not sure it does. The quip is just a statement of fact. Causation != correlation is true simply because they are two different words. So, why would someone need to point this out? The implication is that the person they are addressing doesn't know the difference -- i.e. assumed A caused B without considering the alternative because they're an idiot. It's meant to be a dismissive, smart-ass remark and nothing more. It's also saying "I don't know what the causality is myself, but I'm pretty sure you're wrong anyway," otherwise why bother saying it at all when you could instead lay out possible causes or refute that the correlation exists?

Its appropriateness aside, I don't think it's fair to pick the semantics of the phrase apart so as to accuse the speaker of not understanding causality, since they almost certainly do. I think it's safe to say that everyone does, whether they know the terminology and formal logic aspect of it or not, as it's really damn simple. I doubt someone who cannot make the distinction between causality from correlation would be very functional.
posted by cj_ at 4:35 PM on March 16

saying that a correlation proves some third hidden cause is basically saying nothing at all.

This is just what happens when you argue semantics into the ground. :)
posted by cj_ at 4:38 PM on March 16

I agree with most of your exegesis of the "correlation ≠ causation" quip. I just think it's disappointing that so many people want to rush into threads like this one to post a comment that's the logical equivalent of "That's wrong, and you're an idiot."
posted by grobstein at 5:03 PM on March 16

I'm not sure it does. The quip is just a statement of fact. Causation != correlation is true simply because they are two different words.

Well, different words can describe the same thing, of course--and Hume's argument that causality simply is "constant correlation" seems to me to be hard to refute. But to address the more important point you're making: I find that this "correlation isn't causation" tag is used all the time in political discussions to simply dismiss important scientific findings (look at any debate on global warming, for example--and see its first instance in this thread).

I do think that it's become a piece of dogma that people cling to and which actually distorts proper critical thinking. You'll often find someone trot it out as an excuse to simply ignore some striking correlative result, so I do actually think it's worth challenging. "Correlation implies causation, but does not prove it" would actually be a far more accurate and useful version of the saying--and one that would lead people to a more critical engagement with the kinds of claims under discussion here.
posted by yoink at 5:07 PM on March 16

Or, you know, what grobstein said.
posted by yoink at 5:08 PM on March 16

Agreed that it's an intellectually lazy remark meant to be dismissive of a claim you either can't or won't refute. Even if that claim is nonsense, it's not a particularly helpful thing to say.
posted by cj_ at 5:34 PM on March 16

This is really kind of a bummer, since there were a lot of neat ideas in this thread.
posted by Adam_S at 8:25 PM on March 16

There's an important and actually relevant topic somewhere near the top of this thread, yet it's devolved into a semantic wank. Nice.

As discussed in this thread and the linked article "equality" is the buzzword currently twisting in the right wing's reality-distortion field. It seems that they're trying to make "equality" code for suppression of individuality (and, by implication, individual freedom) and income redistribution, because these are easier targets than directly attacking the concepts like sensible regulation, a livable minimum wage, fair taxation, public education, universal health care - things that we lib'ruls think most people would be in favour of if they considered them on their own merits.

By setting up this strawman, distorting the real achievements of European social democracy, and associating both of these with Obama's programs, the right is struggling for another simplistic meme to rally against.

Is my (always naive) understanding correct?
posted by Artful Codger at 6:58 AM on March 17

There Is About To Be A Very Significant Change In Headlines Around The World: The middle classes are revolting!

previously :P
posted by kliuless at 8:49 AM on March 18