increased spending through deficits, change in levels of debt will be positively associated with
spending growth, but levels of debt should always be negatively associated with spending growth.
Inflation (t-1). Though the extent of cost-of-living adjustments varies across countries and over
time, welfare benefits are somehow indexed to the cost of living in most countries. Hence we
expect the rate of inflation (consumer prices) to be positively associated with growth of social
spending per capita. As with the aforementioned variables, we lag this variable by one year. The
justification for the one-year lag is very straightforward in this case: indexation typically implies
that benefits levels in any given year are determined, in part, by the rate of inflation in the
previous year.
Voter turnout (in previous election). The standard argument concerning the effects of voter
turnout proceeds from the observation that class (or income) differences in voter turnout decline
as turnout rises. The higher voter turnout is, the greater the gap between the income of the
median voter and the average income. Consequently, the median voter will be more likely to
favor redistributive policies. On these grounds, we expect voter turnout to be positively
associated with social spending growth (cf. Meltzer and Richard 1981, Hicks and Swank 1992,
and Iversen and Cusack 2000).
Dependency ratio (t-[t-1]). Spending on the elderly accounts for a very large portion of total
social spending in all the OECD countries and many welfare states also target children. Holding
welfare-state generosity constant, an increase in these targeted groups’ share of the total
population will automatically translate in higher social spending per capita. An increase in the
dependency ratio (defined as the share of the population below the age of 15 and above the age of
64) should thus be associated with higher social spending growth.
18
Convention