The link between inequality and social problems is made clear by this graph from the book The Spirit Level which I posted on last week. The straight diagonal line indicates a direct correlation: the more inequality, the worse your social problems.
The vertical axis shows how well countries do against an index of social indicators that include life expectancy, infant mortality, obesity, imprisonment, teenage births, trust (or lack of it), social mobility, mental illness (including drug and alcohol addiction), homicides, maths and literacy.
Any thoughts?
These things dont just occur overnight. Hmm , who was in charge the last NINE years ?????
Phil in your view, where should NZ be positioned on this graph? Is where we are OK or acceptable? Judging by these parameters, you can still have high inequality but fewer social problems. It just that no country is listed in that area.
Th numbers only show that very very homogeneous populations have less social problems. Doesnt Japan also have a very high proportion of elderly?
So how come Denmark comes out of this worse than Spain? Given Denmark’s low rating for income inequality, shouldn’t it score lower on the social problems scale?
Just a question – (given that copyright has been a subject on this forum a few times)
Are the graphs you are using subject to copyright? Do you have permission from the author to reproduce their work?
That’s a line of best fit. You can put a line of best fit on any graph – it’s not necessarily correlated (although that looks strong).
I would also remind you that correlation does not imply causation.
Maybe the issue isn’t that high levels of inequality lead to mental illness and murder and low literacy rates but the other way around? Maybe if you have a society with low literacy, it is more unequal; maybe if you have a society with high levels of mental illness, it is more unequal? In fact, isn’t that common sense? People with low levels of literacy will struggle, if you’ve a bunch of such people isn’t society clearly going to be more unequal because of this?
Use of copyrighted material is allowed when its very minor and its attributed. For books this is well established.
Trademarks and music seem to be handled differently
Where are the data labels on the graph?
Has any transformations been done to the data?
Fit seems to be too perfect.
Which countries have been left out?
NZ has one of the highest literacy rates in the world and also one with the most inequality.
Is there a reason why the United States isnt on that graph? Where would they sit?
Also – while the inference you draw may be correct – this graph doesn’t prove it. There are plenty of other indicators for either axis which would show a similar prima facie example. For instance, some of the factors in the index of problems are explained in a statistical sense by some of the other factors – the correlation between (say) obesity and life expectancy or drug addiction and imprisonment are high – probably higher than any of those with income inequality. The right sort of analysis to show the effect is a factor analysis – income inequality is unlikely to be the primary explanation of many of those social ills – many other issues which may or may not fall into the catch all category of “income inequality”. It is a very poor measure as it is a relative not an absolute measure. For instance I would rather be poor in an “unequal” country like NZ, than middle class in a more egalitarian society like North Korea.
Look at it another way – lets say in 1990 NZ had a low income of $100, and a high of $110 – for a 10% “inequality measure.” Now in 2009 lets say we have a low of $200 and a high of $300 – inequality measure is now 50%, Headline shrieks “INCOME INEQUALITY BLOWS OUT BY 400%. GOVERNMENT UNCONCERNED BY INCREASING POVERTY IN NZ!”
Yet clearly the poor are much better off in absolute terms.
Graphs like this also tend to confuse correlation with causality, but thats so obvious it doesn’t need observation.
New Zealand’s social problems are entirely attributable to the Labour electorate.
The Labour electorate is incorrigible, it simply cannot be convinced to contribute productively to society or to obey society’s laws, no matter how much welfare decent Kiwis throw at it.
Labour talks a lot about responsibilities and entitlements and there is always a clear demarcation;
Decent, law-abiding, productive taxpayers have responsibilities e.g. “social responsibilities.”
The unproductive and the criminal of the Labour electorate have entitlements e.g. “welfare entitlements.”
When our society performs an about-turn and our political class appreciates that the productive and the law abiding have an entitlement to enjoy the fruits of their own labours and to live in a safe society, whilst the Labour electorate has a responsibility to contribute positively to society and obey it’s laws, our social problems will diminish.
Lets assume the correlation holds true, what do we do to sustainably reduce the inequality? I’ll give you a clue (this is a good clue, it will help you win elections in the future too) – we don’t fix it by increasing welfare dependency.
“gomango says:
Is there a reason why the United States isnt on that graph? Where would they sit?”
They would probably sit in the very top right corner!
Also note how the English speaking countries are clustered around the top.
This is also why NZ needs to stop looking exclusively to the Anglosphere. New Zealanders and the politicians seem to think that the world consists entirely of Australia, the US, and UK. There are plenty of people doing things differently, and we can learn from them, adopting things that can work here.
Phil, some interesting work last year by Robert Wade, currently at the LSE, shows that during the seven year economic upswing of the Clinton administration the top 1 percent of households captured 45 percent of the total growth in pre-tax income, and in the four year upswing of the Bush administration, the top 1 percent captured no less than 73 percent of the total growth. Meanwhile, the average income of the bottom 90 percent actually fell and household debt relative to personal disposable income more than doubled. So, has the great Reaganomics experiment delivered higher living standards to most people as promised? Looks not. Will more of the same do the trick? Surely not. Is there a better way? Let’s see. That’s up to you guys. What is certain is that the ACT formula is a dog, with fleas, and why people still swallow their garbage is beyond me.
Wow look at all the squirming from the right-wingers. “What about the copyright!” “Oh look here’s an exception!” “It’s Labour’s fault!” (even though inequality reduced in the most recent 9 years – be interesting to see if the “factors” improved too).
Great graph Phil, it makes complete sense. Although I do see the point that Graeme makes that causation could also occur in the opposite direction. In fact I think it probably does – the point being that causation probably works both ways, and it seems more straightforward to fix inequality than the other various outcomes.
So what to do to reduce inequality? Well a more progressive tax system including (critically) a tax-free threshold. A capital gains tax to reduce home-ownership inequality and a greater role of the government in basics like housing sound like a good start.
Michael says:
“Also note how the English speaking countries are clustered around the top.”
Well the answer is clear. We should all go learn Japaneese and every thing will be right with the world.
It’s interesting that, by this measure the ‘best’ performing country is one with an incredibly small set of policies we might call ’social welfare’ – it’s virtually non-existent in Japan.
Then the next three are those from the complete end of the social welfare spectrum…
“Any thoughts?”
Actually, yeah, one more:
Rule number one in The Politicians Guide to Graph Making* states: Never trust a graph that doesn’t have properly marked axis. It’s the surest sign the creator is hiding something for their own agenda.
* OK, I made that up on the spot.
Ok so tell us something we don’t know- you treat people like sh@ts, they behave like sh@ts.Desperation drives despair and dehumanizes.
The real issue is how do you fix it? Highlighting the issue is laudable, but surely it carries an obligation to deliver a solution?
Phil, would be interested to hear your thoughts? What is Labour’s strategy to reduce the gap?
I am responding in particular to Simon: firstly, when NZ had a more equal society it also had a much lower crime rate. Secondly, Labour did what was possible within the economic realities of modern NZ to redress the imbalance by prioritising employment and setting up Kiwisaver, so as to turn us into a nation of savers rather than borrowers. If you are a saving nation you are able to address inequalities because you are not at the mercy of the lenders. Note also that Labour did not have to coerce people into working, they were delighted to have jobs, which gives lie to the ongoing bleat about career beneficiaries, etc.
Come on guys, engage brain before blogging, and read the book, or at least visit the website: You really are hilarious reacting the way you do, without seemingly wanting to do the homework.
You’ll see a lot of your questions answered, and hopefully a whole lot more posed. You can look at more of the graphs from the book (graphs I hope Phil will post on) on http://www.equalitytrust.org.uk/
Do go back too and see Phil’s earlier post looking at where NZ sits on the rising inequality scale.
They are on the graph, right in the top right hand corner.
No, it doesn’t. The rest of the data and the analysis in the book does.
I suspect that amount of inequality would register as zero on the Gini scale.
I am curious to know how China, Vietnam or Cuba would rate on these scales. More because I am interested in what sort of inequality is seen there and if the different ideology there would move them away from the graph.
This would help understand how other societal impacts affect the data being collected.
Working back from the bottom (sorry for not responding earlier, been away on the ‘Naki caucus trip):
@ Geek – The Spirit Level data only includes OECD so you won’t find China, Vietnam or Cuba there. It would be fascinating I agree, to see to what extent the growing inequalities in China and Vietnam are having an effect.
@ dave rutherford – I agree we have to start talking about solutions, but hey this conversation is just getting started. Will post in future on what kinds of policies would tackle inequality.
@ Phil Anderson – Japan is interesting in this context. Very low expenditure on social provision but its employment market delivers low income inequality. Just goes to show the welfare state is not the only way to skin this cat.
@ Jennifer – do you have a reference for the Robert Wade data? I’d be interested to see it.
@ Labrat – One person’s dependency is another’s helping hand, but putting that aside, there are ways of tackling inequality other than through provision of benefits. One of the ways is investing in wealth creation opportunities for lower socioeconomic groups (better education and assistance into post-secondary education, trade training, seed capital for new business ventures, better early childhood interventions in health and education, etc).
@ gomango, Graeme – see the discussion at http://blog.labour.org.nz/index.php/2009/08/03/mind-the-gap/comment-page-1/#comment-5157 re causality and correlation. Better still read the book. It makes the case convincingly.
@ indiana – NZ is far too far up the high end on both axes for my liking.
& starboard – Have a look at the graph in http://blog.labour.org.nz/index.php/2009/08/07/on-inequality/ to see that this problem has been around for a lot longer than the last nine years.
I found it very interesting on their web site when they described how Japans inequality developed. Definitely shows there are ways outside of state redistribution to handle it although it would probably fall into the too hard basket as it would be difficult to force high income earners to lower their own wages.
I am far from a socialist so am a bit sceptical of the fact that the site contained no counter arguments or contrary evidence. I am curious Phil, is the book a bit more balanced?
@ Geek – The book advocates a case but it backs it with quite a bit of considered argument and data, and it does look at contrary evidence. As David Craig has noted in comments on earlier posts, this stuff is well established in epidemiological and public health circles – no one is disputing correlation or causation. What is interesting about The Spirit Level is that they take the arguments squarely into the public domain.
Someone will always dispute whether or not there is a case to dispute =)