This chart and its commentary not only create the false impression of a new confirmation of The Spirit Level's (TSL) hypothesis that health and welfare in the world's richest countries, unlike amongst poorer countries, is crucially positively unrelated to their average income, but is instead strongly negatively related to their income inequality. But they also falsely imply it was originally confirmed by the international statistics just for the one year 2002 presented in that mistakenly celebrated book. But in fact possibly neither of these alleged confirmations are scientifically valid, at least by virtue of invalidating cherry-picking of the testbed data. And to the best of my knowledge no valid confirmations of it have ever been published, contrary to Wilkinson & Pickett's many claims of voluminous confirmations of it reported in the literature.

On the contrary, rather it has many valid strong refutations, including for those sets of rich countries selected by TSL's fourfold testbed criteria of (i) those countries out of the richest 50 that have data for both (ii) GDP per head and (iii) income inequality, and also (iv) with populations more than 3 million. For using UN Human Development Report (UN HDR) data as TSL does, these refutations even include a strong one for the very same year 2002 falsely proclaimed to confirm it. For example on health, for those 30 of the 50 richest countries in 2002 that satisfied TSL's selection criteria, their populations' longevity had an almost strong 65% positive correlation with average income rather than none, but only a very weak 18% negative correlation with income inequality rather than a strong one, thus refuting the book's hypothesis.

But whilst crucially omitting to state the correlation of longevity with average income, TSL claimed there was a 44% negative correlation of longevity with income inequality, which is anyway only a weak correlation at best, explaining less than 20% of the international variation in longevity, and thus surely a refutation of the TSL hypothesis of a strong one rather than a confirmation of it as the book most risibly perversely claimed it to be. Yet even this weak correlation was only achieved by deselecting 7 of those 30 countries that satisfied its selection criteria from the testbed, which therefore only included 23 rich countries, but without any scientifically valid justification for their omission. These deselected countries were Argentina, the Czech Republic, Hong Kong, Hungary, Poland, Slovakia and South Korea. This was thus surely quite outrageus and desperate invalidating data cherry-picking that omitted almost a quarter of the designated testbed countries, but even then to little avail.

Again on health, the UN HDR data for 2005 entails longevity had a 73% positive correlation with average income but only a 5% negative correlation with income inequality for those 32 rich countries satisfying the TSL testbed criteria for that year, and the data for 2009 when TSL was first published shows corresponding correlations of 68% positive and 3% positive for the satisfying 33 rich countries, again both clear refutations of TSL's international health and incomes hypothesis. Altogether scientifically valid statistical tests to date show this hypothesis for the world's richest countries is robustly refuted by UN HDR data for at least 14 different years from 1985 to 2009. And it will most likely also be well refuted for the years 2010 and 2011. Apparently unlike the media journalists who have mistakenly accepted TSL's claim as proven, anybody with a spreadsheet can easily and quickly test the relevant correlations for these two years by inputting the relevant data from the UN HDR website when available, and indeed for any other years.

Certainly this New Statesman chart for just a paltry 6 rich countries proves nothing and may well be yet another example, and indeed most extreme case, of the notorious pseudoscientific malpractice of invalidating testbed data cherry-picking to try and create the false impression of the confirmation of a hypothesis. Finding some choice subset of the world's richest countries to make a testbed that favours it does not thereby scientifically confirm it if some rich countries have been unjustifiably omitted when their inclusion would refute the hypothesis. This would surely be nothing less than scientific fraud in yet another exemplar of "lies, damned lies and statistics".

no surprise USA would be the worst....its b/c of religion.

Inequality in a society reflects the innate character of the population and culture of that society and its diversity, the more diverse a society is culturally the more unequal it is likely to be and it is liberals who see diversity in a society as a characteristic which is to be encouraged and ironically who complain most about inequality.

Strange how the British have a better road safety record than most but generally the trends have been maintained. I hope NS will give those supporting less inequalities more support as a key issue now growth seems to of disappeared from he real agenda.