Here’s a diagram from a recent paper (pdf) on the Swedish distribution of income. It shows that the kinds of income distribution through time that Andrew Leigh and Anthony Atkinson came up with for Australia and NZ are typical of most western countries. It also shows that you can lean pretty successfully against the wind – as the Swedes have. And also that you can lean pretty ineffectively against the wind – like the French have. I’ve asked the authors of the paper for a better resolution picture – but I think it’s fairly clear even in it’s somewhat foggy state.
Share of income going to the highest percentile of income earners.
Two interesting comments in the referenced paper:
1. the large drop in upper percentile of incone was before the welfare state came into being.
2.the recent increse in Sweden has been due to capital gains.
It would be interesting to see how the distribution of wealth (rather than income) varied over the same time period. Also for the same period the distribution of wealth between genders. I have seen statistics that purport to show that the percentage of wealth owned by women far exceeds their overall percentage of the population.
Nicholas, the figures you cite are pre-tax (and may even be pre-social transfers, I don’t know. I am no great fan of French employment policies but in fairness they have much less inequality of disposable incomes (a lower relative poverty rate) than the English-speaking countries. France rates quite well for “equity effort” – percent of poor households removed from poverty by fiscal policy.
It’s an interesting measure of inequality you’ve chosen, though. The paper uses tax data – as I’ve said elsewhere, changes in the tax base over time (ie what’s defined as assessed income) makes drawing historical trends and doing international comparisons very treacherous.
Where tax policymakers pursued a deliberate “broaden the base and lower the rate” strategy, as many Anglo ones did in the 70s and 80s, then you will record an apparent increase in inequality in taxable income even if in fact there had been no change in the real income distribution at all. The broader definition of income along with the reduced penalty for honesty (IIRC in the UK, the top marginal rate went from 90% to 30%) meant that a lot of rich people who would in the 1960s not have been assessed as having high incomes now are so assessed (BTW, this is why I believe the estimates of the effect on income of tax cuts that supply-siders like Feldstein draw from the Reagan tax cuts are biased upwards).
It’s an effect that differentially hits the top percentile, who have access to tax advice, are subject to the top income tax rate and/or have income sources (capital) that are inherently harder to assess. I wouldn’t draw these sort of strong conclusions without other data sources.
Yes, fair point. The broad similarity of all countries is striking though, and though changes in their tax systems were also broadly similar, I doubt they were as well syncrhonised as the curve shows – ie there’s a real effect there.
Btw, I didn’t ‘choose’ this measure – which is probably not one of the measures of inequality that means most to me. I just found the diagram looking for other stuff, and thought it was worth posting.