Rates of spending

What accounts for the wide differences in rates of social security spending among industrial countries, ranging all the way from 28.6 percent of gross domestic product (GDP) in Sweden to 7.8 percent in Japan in 1980? This is a question that has attracted the attention of a number of investigators, with results that are generally consistent from one study to another.

In Table 2.1, we have assembled data including a number of variables that have been found to be associated with social security expenditures as a percentage of GDP. The countries with the oldest social security programs have been found to be the largest social security spenders in several studies. In an analysis of data for 22 industrial countries, Aaron (1967, 25) found that the “maturity of the system” was the most important of any of the variables he tested in explaining the percentage of national income devoted to social security expenditures (see also Pechman, Aaron, and Taussig, 1968, appendix D). I had earlier developed a related finding, that the longer a country had had an old-age benefit program, the higher its average old-age benefit was likely to be in relation to per capita income and earnings levels (M. S. Gordon, 1963a, 455–6). In a more complex analysis, Wilensky (1975, 24) found that economic level and social security effort were strongly related, but that this relationship was “largely mediated through the proportion of the aged in the population … and, second, through both age of the population and age of the system … .”