I began by observing that “crisis” talk, however misleading, is universal. But has the action of governments matched the rhetoric of antispending campaigns? What cutbacks have actually occurred in the period of austerity after 1975 or 1980 when economic growth and productivity growth slowed down? With some exceptions, the core programs of the welfare state – pensions, disability insurance, and national health insurance, programs that have generally outpaced GDP growth – have proved most resistant to real cuts in benefits per capita or even in their GDP shares. Most vulnerable to real cuts or at least spending restraint have been education, social assistance, unemployment compensation, and family allowances (but not child care or parental leave).

With few exceptions, there are five main reasons for this pattern of growth and restraint. First, demography, although it is not destiny, does count. Declines in education spending reflect declines in school-age populations. The “young” countries with a school-age bulge cut education expenditures per child while still raising such spending as a share of GDP, but as the school population declined, the GDP share leveled off or declined. The “older” countries spent more on pensions both per capita and in GDP share but at a diminishing rate, eventually leveling off (Lindert, 2004: vol. 1). Aging, as we have seen, also increased health and disability spending, especially as the “old-old” increased their share of the population. Second, after universal coverage is achieved, various measures to control costs or restructure programs have some effect, especially in health care (as in German reforms of the 1980s and early 1990s). Third, programs where abuses are obvious and widespread (sick pay, disability insurance) have evoked substantial government reform efforts with varying success; disability cutbacks have encountered especially fierce resistance (Wilensky, 2002: ch. 15 and table 15.3). Fourth, the rate of economic growth has an automatic effect on these numbers: below-average growth will automatically increase the expenditure ratio (SS/GDP) as the denominator levels off or decreases while social spending continues upward or grows slowly. We see this automatic effect in the current economic downturn. Finally, the interaction of three forces – a very large clientele (all pensioners, all the health insured), strong political organization or influence, and great mass popularity – means that welfare-state leaders have already achieved generosity of benefits; their citizens now have entrenched interests and strong sentiments for maintaining the status quo.