Average inequality and extreme income inequality are two very different concepts

There are two types of indicators to measure income inequality: A measurement of average income inequality, concerning the entire population (the GINI index of income inequality before or after redistribution is then generally used); A measurement of extreme income inequality, concerning individuals with the highest income (the percentage of national income taken by the 1% of persons with the highest income, for example, is then used). We first calculate, based on the sample of OECD countries, the correlation between average income inequality and extreme income inequality: extreme income inequality is correlated with average income inequality after redistribution (but not before redistribution). We then seek to determine what average or extreme income inequality is associated to : innovation, the employment rate, labour force skills, changes in income distribution . We see that: Average income inequality is negatively correlated to innovation, the employment rate and labour force skills; Extreme income inequality is positively correlated to innovation and the employment rate and negatively correlated to labour force skills; Increasing labour force skills therefore reduces all forms of inequality (average and extreme). But increasing innovation or the employment rate reduces average inequality but increases extreme inequality. This means, for example, that innovation makes it possible to increase low wages (productivity and disposable income increase ), but leads to the emergence of very rich persons.