Kalle Moene, Department of Economics, University of Oslo
There are huge differences in the pre-tax and transfer distribution of wages across countries. Why don't countries with the same level of GDP per capita have the same distribution of income? There are even larger differences in the inequality of final incomes than in the inequality of wages after taxes and transfers.
Thus redistribution through social policies amplifies rather than mitigates cross-national differences in wage inequality.
The most important institutional factors in explaining wage inequality (before taxes and transfers) are
- whether wages are set locally, at the industry level or at the level of the economy as a whole,
- the extent to which wage setting is dominated by a few large unions who are able to coordinate informally, and
- the extent to which union contracts cover the workforce.
The more wages and salaries are set in a centralized manner the more egalitarian the distribution of wages and salaries.
The political support for welfare benefits to those without income declines as inequality rises. This kind of support should be understood as social insurance rather than pure redistribution. Insurance is a normal good in the sense that demand goes up with income (while redistribution is an inferior good). This leads us to the theoretical prediction that both as a share of GDP and as a share of government spending
the support to those without earnings goes down as inequality increases.
This prediction is also supported by our empirical investigations:
OECD countries with more unequal distribution of wages and salaries spend less, not more on welfare programs that pay benefits to those without earnings.
According to our arguments and findings, we should also find that countries with centralized wage setting institutions should be associated with low levels of poverty. These countries should have a smaller share of workers with low incomes and a higher support for social insurance. Applying the US poverty line (of 40 per cent of the median income level) and implementing this absolute poverty threshold in other OECD countries, we find exactly that: Countries with little wage inequality (and centralized wage setting) have low levels of poverty also when we control for cross-national differences in average incomes.
Thus, labour market institutions have a direct effect on the distribution of wages and an indirect effect on the political support for social insurance.
The more centralized the wage setting process, the more concentrated the unions and the higher the level of coverage, the more compressed the wage scale.
The more compressed the wage scale the greater the support for social insurance policies.
The more compressed the wage scale and the higher the level of spending on social insurance, the lower the share of households below the poverty line.
These findings are elaborated in several articles by Michael Wallerstein and myself. See for instance the two articles listed below.
Moene, Karl and Michael Wallerstein (2000): “Institutions, inequality and social policy”, Sosialøkonomen, 54.
Moene, Karl and Michael Wallerstein (2001):"Inequality, Social Insurance and Redistribution”, American Political Science Review, December.