International Variation In The Size Of The Welfare State: A Preference-Based Explanation

In 1980, government transfers as a fraction of GNP in OECD countries varied from 30.4% in Belgium to 11.1% in Greece. In this paper, we seek a political-economic explanation of this variation. In particular, we shall attempt to explain the variation as due to a variation in preferences of the citizenry's. Voters are modeled as caring about their personal welfare, the welfare of the unemployed, and the size of government. The political issue is the tax rate, where taxes are sued to support those who live on transfer payments. The first part of the paper demonstrates that, with the above formulation, utility functions are single-peaked in the marginal tax rate (the single political dimension); hence a Condorcet winner exists. We identify the Condorcet winner with the political equilibrium in a country. Although tax functions in the OECD countries are not linear, we may, given the data, compute what linear tax system would have been 'equivalent' to the existing one in each country, in respect of generating the same level of transfers to the unemployed that exist in that country. We calculate these tax systems, for each country. Next, assuming a specific parametric form for the distribution of voter traits for each country, we can compute a 'preference parameter locus' for each country, which tells us something about the degree to which voters are 'other regarding' or 'self-interested,' and their degree of antipathy towards big government. We then can order countries in terms of their degree of 'social orientation.' The results essentially conform to conventional wisdom about the various countries.