Does income inequality increase mortality?

In his recent Times column, Paul Krugman writes:

Rising inequality has obvious economical costs: stagnant wages despite ascent productivity, ascent debt that makes us to a greater extent than vulnerable to fiscal crisis. It too has large social too human costs. There is, for example, rigid show that high inequality leads to worse health too higher mortality.

The links are from the online version of Paul's column.  I followed the second link to an interesting article by Angus Deaton.  Angus writes the next (emphasis added):

Darren Lubotsky too I ⁷ receive got investigated the human relationship betwixt income inequality, race, too mortality at both the province too metropolitan statistical expanse level. In both the province too the metropolis data, mortality is positively too significantly correlated with almost whatever mensurate of income inequality. Because whites receive got higher incomes too lower mortality rates than blacks, places where the population has a large fraction of blacks are too places where both mortality too income inequality are relatively high. However, the human relationship is robust to controlling for average income (or poverty rates) too too holds, albeit less strongly, for dark too white mortality separately. Nevertheless, it turns out that race is indeed the crucial omitted variable. In states, cities, too counties with a higher fraction of African-Americans, white incomes are higher too dark incomes are lower, too therefore that income inequality (through its interracial component) is higher inwards places with a high fraction black. It is too truthful that both white too dark mortality rates are higher inwards places with a higher fraction dark too that, once nosotros command for the fraction black, income inequality has no effect on mortality rates, a resultant that has been replicated yesteryear Victor Fuchs, Mark McClellan, too Jonathan Skinner⁹ using the Medicare records data. This resultant is consistent with the lack of whatever human relationship betwixt income inequality too mortality across Canadian or Australian provinces, where race does non receive got the same salience. Our finding is robust; it holds for a broad gain of inequality measures; it holds for men too women separately; it holds when nosotros command for average education; too it holds in 1 lawsuit nosotros abandon age-adjusted mortality too facial expression at mortality at specific ages. None of this tells us why the correlation exists, too what it is close cities with substantial dark populations that causes both whites too blacks to travel sooner.

In a review of the literature on inequality too health, I regime annotation that Wilkinson's master copy evidence, which was (and inwards many quarters is still) widely accepted showed a negative cross-country human relationship betwixt life expectancy too income inequality, non solely inwards levels but also, too to a greater extent than impressively, inwards changes. But subsequent travel has shown that these findings were the resultant of the purpose of unreliable too outdated information on income inequality, too that they practice non look if recent, high lineament information are used. There are immediately too a large release of private degree studies exploring the wellness consequences of ambient income inequality too none of these render whatever convincing show that inequality is a wellness hazard. Indeed, the solely robust correlations look to live those with United States cities too states (discussed above) which, every bit nosotros receive got seen, vanish in 1 lawsuit nosotros command for racial composition. I advise that inequality may indeed live of import for health, but that income inequality is less of import than other dimensions, such every bit political or sex inequality.¹⁰

Is Angus's article actually back upwards for Paul's claim?  It seems to me that it is to a greater extent than the opposite.
Sumber http://gregmankiw.blogspot.com/