I take it for granted the desire for some degree of inequality is rooted very deep in the human physiology (I mean, the desire to be on the higher end of the ladder, of course), just as well as offsetting tendancies for altruism and equity (not the same as equality), which make most primates "social species" and not just lone predators.
May be your question should be rephrased as "does GDP growth implies income inequality growth ?"
Stricly considering the span of all incomes, I think yes (the rock bottom will always earn absolutely nothing, whatever the wealth of a nation, whatever the social system, there are always dropouts - meanwhile, if the total wealth grows, the top earners will probably earn more even if their share in the total erodes).
With more subtle statistical definitions of inequality, like variance, percentiles etc... I think history says no: Jérôme has shown us many graphs with the rich getting much richer in recent years, but also that in many other periods of recent history inequalities were stable or decreasing, although western economies were still experiencing growth during these periods (including most of the cold war, viet nam war, etc..)
May be we should look into other predictors of change in inequality, I think we could find something better with combinations of:
I think I mean what I said: GDP growth correlated with income inequality.
Where are the data for this parametric fitting? Those whom the Gods wish to destroy They first make mad. -- Euripides
Well, I know where to get inflation, growth series for France (e.g. www.insee.fr). But inequality is much harder to summarize with a single indicator. INSEE does regular studies in France, but the frequency is far lower than yearly (like, every 5 years) and the indicators do not all remain the same making comparison very hard over 10, 15 years. Generational accounting is only coming into the spot lights now, and we certainly won't find any retrospective series on this at this point (and probably not anytime soon in France).
For other countries, which we would need to make correlations significant , the same series that are easy for France would be easy too. But I am even more clueless as to how to get the others. I remember one graph from Jérôme with the income of the top US percentile, we could lay it in a table of approximate numbers. But that is still only one country.
At this point I'm thinking, correlation with such simple series as inflation and growth have certainly been run already... I'm gonna go googling ! Pierre
As I read it, the countries that did relatively well did not change their income inequality much but those who did change it fared worse. But to which direction we cannot tell. Orthodoxy is not a religion.
http://www.aeconf.net/Articles/May2002/aef030105.pdf
Very high in google, not sure it really warrants of its worth. Although observing that the debtor-creditor scenario can make inflation beneficial to the middle class (which has access to credit), global empirical data suggest that inflation is bad for inequalities (because the really poor don't have access to credit to take advantage of the inflation ? and because the rich move faster to adapt to it ?), even filtering out data points where it's actually hyper-inflation wrecking an entire country.
There is also this: http://www.eldis.org/static/DOC4760.htm from the IMF, so probably heavily loaded also (it says max growth = good for the poor, whatever the rest), but it gets wrung at: http://delong.typepad.com/sdj/2005/06/is_inequality_a.html (again, don't know exactly what it's worth).
So there are many references to well formatted data series in these papers, but eventually the conclusion after critical reading seems to be that ... there is no general, worldwide, significant link between anything.
OK, I go back to sleep ! Pierre
