?
Re-revisiting the mass privatisation e Mortality debate: A response to Stuckler, King and Mckee
Stuckler, King and McKee (2012) (hereafter, SKM) are unhappy that their bold claims that rapid mass privatisation was an important cause of post-communist mortality fluctuations have been disputed. In defending their claims, SKM variously accuse those questioning them of statistical manipulation and obscurantism , ideological bias, conflicts of interest, misrepresentation, denialism (e.g. as in climate change denial) and data torture. These are uncomfortable associations for any researcher. So, let me be clear, no one is denying the occurrence of unprecedented spikes inmortality and morbidity in the early 1990s across parts of the post-communist world; what is being questioned is the assertion that rapid mass privatisation was “a crucial determinant of differences in adult mortality trends”. This claim, stemming from a cross-national analysis covering 24 countries for up to 14 years, hardly merits associationwith the vast body of impressive scientific theory and evidence underpinning the link between environmental damage and climate change. Nevertheless, the discussion is important because it goes to the heart of what, when and how much can be claimed from small scale, cross-national data analysis. While SKM have raised an interesting question, their claims are far louder and bolder than the data justify and there are lessons, for social scientists and epidemiologists, that can be drawn from this. Moreover, their claim that it is wrong for their critics to explore their findings without putting forward and testing alternative hypotheses is anti-scientific (Popper, 2005, p.39): falsifiability defines the scientific endeavour. In their response to my article (Gerry, 2012) SKM state that I am part of a game of“statistical obscurantism”. They advance this allegation in four stages by: (i) revisiting the key facts and concepts; (ii) repeating discussion of the methodological challenges; (iii) comparing Russia (a mass privatiser) with Belarus (not a mass privatiser); and (iv) providing a set of regressions which they claim demonstrate “data torture”. I will take these in turn.