Re-evaluating self-evaluation. A commentary on Jen, Jones, and Johnston (68:4, 2009) Anna Barford a, * , Danny Dorling a , Kate Pickett b a Department of Geography, University of Sheffield, Sheffield, S10 2TN, United Kingdom b Department of Health Sciences, University of York, York, YO10 5DD, United Kingdom article info Article history: Available online 26 November 2009 Keywords: Inequality Self-evaluated health International comparison Introduction In a paper recently published in Social Science & Medicine, ‘‘Global variations in health: Evaluating Wilkinson’s income inequality hypothesis using the World Values Survey’’, and one in another journal Jen, Jones and Johnston provided analyses of self- rated health, using data from the World Values Survey (Jen, Jones, & Johnston, 2009a, 2009b). They showed that income is significantly associated with self-rated health; higher incomes being related to better health. Although they found that self-rated health varied between countries, they did not find an independent effect of income inequality on self-rated health. The authors concluded that their analyses provided a test, and refutation, of what they described as the ‘‘Wilkinson hypothesis’’. Briefly stated, the ‘‘Wilkinson hypothesis’’ is that population health in rich countries tends to be better in societies where income is more equally distributed. There are now more than 200 analyses of the nature of this relationship, and reviews of these studies have come to conflicting interpretations of the evidence, with researchers disagreeing over methodological issues, such as the scale at which inequality is measured, whether or not various control variables should be interpreted as confounders or media- tors, and the appropriateness of various statistical models (Lynch et al., 2004, Macinko, Shi, Starfield, & Wulu, 2003, Subramanian & Kawachi, 2004,Wilkinson & Pickett, 2006). In the context of this controversy, do Jen et al’s (2009a, 2009b) analyses put a final nail in the coffin of the ‘‘Wilkinson hypothesis’’? In this paper, we consider the relationship between ‘health’ (the outcome in the Wilkinson hypothesises to be affected by income inequality) and ‘self-rated health’ (the outcome analysed by Jen and colleagues). We argue that ‘health’ and ‘self-rated health’ cannot always be assumed to be proxies for one another. We also suggest that if, as Jen et al. find, average levels of self-rated health tend to be higher in more unequal societies, that this has something to tell us about the psychosocial effects of living in unequal societies. Rating our health What are we actually measuring, when we ask people to rate their own health? Self-rated health might reflect a ‘spontaneous assessment’, which incorporates many elements of a person’s health into an overall or global statement about health status. Alternatively self-rated health might be thought of as reflecting a person’s ‘enduring self-concept’ which remains stable over time, often with independence from changes in health (Bailis, Segall, & Chipperfield, 2003, pp. 205–207). The data on self-rated health collected in the World Values Survey fits into the first typology, a global assessment of health that amalgamates all aspects of one’s health into a single statement. The question to which respondents replied was: ‘‘All in all, how would you describe your state of health these days? Would you say it is excellent, very good, good, fair, poor or don’t know?’’ (Jen et al., 2009b, p. 645). In a review of 27 studies, Idler and Benyamini (1997) found that self-rated health was generally predictive of mortality (Idler & * Corresponding author. University of Sheffield, Department of Geography, Winter Street, Sheffield S10 2TN, United Kingdom. Tel.: þ44 7866 759 301. E-mail address: Anna.Barford@sheffield.ac.uk (A. Barford). Contents lists available at ScienceDirect Social Science & Medicine journal homepage: www.elsevier.com/locate/socscimed 0277-9536/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.socscimed.2009.10.059 Social Science & Medicine 70 (2010) 496–497