After the 'Golden Age': Convergence Towards the Supply Side Model, Dualization and Individualization of Outcome Risks
Size and Instruments of Welfare States
In stark contrast to the predictions of a ‘race to the bottom’ between countries towards a residual model aiming at low taxes and contributions, welfare states today play a more dominant role in the composition of public spending and total economic activity than ever before. Figure 2.1
Fig. 2.1 Total public social expenditure, 1980-2014. Notes: As a percentage of GDP (left axis) and coefficient of variation (right axis). Source: OECD Social Expenditure Database, OECD.stats
shows total public social expenditure and its coefficient of variation for 21 OECD countries as well as the selected cases of Germany, Sweden and the US: the average social expenditure share has risen by eight percentage points in the last three decades, from 17 per cent of gross domestic product (GDP) in 1980 to 25 per cent in 2012.
At least four functional drivers of such an upward convergence potentially play a role: first, changes in legislation and implementation increase the generosity of existing programmes and expand into new areas—with the most genuine and deliberate expansion of welfare effort being driven by political decision-making. Second, demographic ageing and rising prices caused by ‘Baumol’s cost disease’ increase spending on social services, even in the absence of benefit expansion. Third, changes in socioeconomic conditions, and especially drastic recessions as experienced in the recent financial crisis, drive up the demand for transfers and services, increasing the number of beneficiaries without necessarily changing generosity in the form of eligibility, benefit levels or duration. Finally, recessions that drive up demand for transfers and services at the same time suppress GDP, which is the denominator of the commonly used social expenditure measure. The economic downturns in the early 1990s and after 2007 reveal the last two mechanisms of rising social expenditure shares most markedly, each period accounting for about 3 percentage points of social expenditure increase.
Even more striking than this increase is the strong convergence between OECD countries in terms of social expenditure within the same time period (see Schmitt and Starke 2011; Starke et al 2008 for more detailed analyses), with the coefficient of variation being almost halved from 0.30 to 0.16. This convergence is in line with the conclusion of many scholars that socioeconomic conditions have been the dominant driver of recent welfare effort trends, muting the differential influence of political factors such as party composition of a government, which played a more prominent role in previous decades (Huber and Stephens 2001; Kittel and Obinger 2003; Kwon and Pontusson 2010).
Given the strong effect of socioeconomic conditions on expenditure data, drawing conclusions on the content of welfare state change requires analyzing social policy instruments and specific policy fields. Such analyses suggest a more nuanced picture compared to the expenditure trajectory: objectives, instruments and core policy fields of welfare state activity have been shifted or extended. A particularly dominant trend in social policy instruments is the shift from passive cash transfers towards in-kind benefits, which predominantly consist of social services such as healthcare, active labour market policy or public childcare. While transfers are by no means disappearing, Figs. 2.2(a)-(d) show that replacement rates of core cash transfers have consolidated or somewhat retrenched as well as converged in advanced democracies during recent decades.
As opposed to cash transfers, expanding social services form one building block of what has been coined the ‘social investment state’ (Esping- Andersen et al. 2002; Hemerijck 2012) or ‘active social policy’ (Bonoli 2013) and what we more broadly describe as the ‘supply side model’. Compared to unconditional cash transfers, modern social services tend
Fig. 2.2 Average net replacement rates (left axis) and coefficient of variation (right axis) of selected programs, 1971-2011: (a) unemployment, (b) sickness, (c) minimum pension, (d) standard pension. Notes: Net income replacement rate data is based on hypothetical model households. For unemployment and sickness replacement rates, the mean of the replacement rates for a single household and that of a household with one full-time average earner, a dependent spouse and two children was calculated. For pensions, the average is based on the mean of a single pensioner and a married couple. Source: Comparative Welfare Entitlements Dataset 2 (Scruggs et al. 2014) to be driven more by the aim to improve the capacities of citizens to supply their labour to the market than to insure against income loss. Figure 2.3 shows that the share of in-kind benefits as a percentage of total public social expenditure has risen by about 10 percentage points across OECD countries since the early 1980s. Looking at Sweden, Germany and the US as representatives of different welfare regime types, we see that the latter two have experienced a catch-up process while Sweden is back at the same level it was at in 1980—a phenomenon that fits in well with the overall lower level of variation between OECD countries and is part of the convergence towards a supply side-oriented welfare state.
-  This term refers to the phenomenon that wages in labour-intense services are increasing in linewith overall wage development, even though productivity in these jobs does not rise accordingly.