مکمل های سیاسی در دولت رفاه: بهداشت و درمان و تامین اجتماعی

All OECD countries target a large majority of their welfare spending to the elderly, through public pensions and health care programs. Spending in both programs has largely increased in the past decades — often more than the share of elderly in the population. We suggest that these phenomena may be due to political complementarities between these two transfer programs. We show that these two programs may coexist, because public health care may increase the political constituency in favor of social security, and vice-versa. Specifically, public health decreases the absolute longevity differential between low and high-income individuals, therefore rising the retirement period and the total pension benefits of the former relatively to the latter. This effect increases the political support for social security among the low-income young. We show that in a political equilibrium of a two-dimensional majoritarian election, a voting majority of low-income young and retirees supports a large welfare state; the composition between public health and social security is determined by intermediate (median) income types, who favor the contemporaneous existence of these two programs, since public health increases their longevity enough to make social security more attractive. Technological improvements in health care strengthens this complementarity and lead to more welfare spending.

In all OECD countries, the two major welfare state programs – public pensions and health care – target mainly the elderly. Unfunded pension systems use current contributions from the workers to finance current pension benefits to the retirees. Public health care programs may instead be financed through general taxation or earmarked labor taxes, but provide most of their benefits to the elderly. As a result, in the US, people aged 65 to 74 receive five times more benefits than individuals with less than 64 years (see Hagist and Kotlikoff, 2005). Similar, yet less extreme, patterns apply to other OECD countries. Furthermore, in the last few decades social security and public health care spending has dramatically increased in all developed economies. Aging has typically be identified as the major culprit of this spectacular raise in public spending, due to the increase in the share of beneficiary from these programs. Yet, social security and health care spending has often increased more than the share of elderly in the population, thereby suggesting that pension and health care benefits per each elderly individual have also increased. Public provision of health care and pension transfers have traditionally been justified because of inefficiency in some relevant markets – such as annuity and private health care – due to asymmetric information. Recently, the health care literature has emphasized the relevance of technological progress, and the adoption of new, more expensive medical treatments (see Newhouse, 1992) to explain the raise in health spending. Yet, why are these new technologies so massively used in spite of their cost? In a recent paper, Hall and Jones (2004) argue that health care is a superior good: as individuals get richer they choose to spend a larger proportion of their income on health care. Policy-makers are thus merely adjusting public health spending to individual preferences. Analyzing the political support to social security and medicare, Bohn (1999) forecasts a further increase in spending driven by the change in the age profile of the electors. Galasso and Profeta (2004) concentrate on social security, and reach as similar conclusions: aging will lead to a further increase in the share of social security spending over GDP.1 This paper presents an additional explanation of the contemporaneous existence of these two large welfare state programs – health care and social security – mainly targeted to the elderly, based on a notion of political complementarity among welfare programs. We show that more health care may increase the political constituency in favor of social security, and viceversa. There may be several channels of political complementarities, as one program may modify some relevant characteristics of the voters — thereby changing their preferences over the other welfare program. Here, we concentrate on how health care policies may affect the redistributiveness of social security. The seed of this intuition was in Philipson and Becker (1998), who argued that social security induces the elderly to increase their private investment in health care, because the existence of an annuity – the old age pension – raises the value of longevity. Here, we identify a new link that goes from (public) health care to social security. Expenditure in public health care increases longevity in a non-linear way, since its effect tends to be larger among low-income individuals than among well-off people. However, richer individuals tend to live longer. Thus, for a given income distribution, expenditure in public health contributes to decrease the longevity differential between rich and poor individuals. As a result, the retirement period, and thus the total pension benefits, increases more for low-income than for high-income individuals, therefore rising the returns on social security for the low-income workers, as opposed to high-income ones.2 The main contribution of the paper is to show that, for a sensible – yet stylized – representation of the two separate programs, some political complementarity between social security and public health care emerges. This political complementarity justifies the use of two welfare programs to transfer resources to the elderly and helps to explain the large government spending in health care and social security. Social security and public health care are sustained as a politico-economic equilibrium outcome of a majoritarian voting game. A voting majority of low-income young and all retirees supports a large welfare state, as in Tabellini (2000) and in Conde-Ruiz and Galasso (2005). Its composition between public health care and social security is determined by intermediate (median) income types, who favor a combination of the two programs, since public health care increases their longevity enough to make social security more attractive. Additionally, we show that an improvement in the health care technology that increases the effectiveness of public health care in raising longevity strengthens this political complementarity and thus increases welfare spending. Our theoretical model is built around three crucial elements. First, income has a protective effect on health — hence, high-income individuals live longer. This effect gives raise to a health “gradient”: income inequality is associated with health inequality. Second, health care improves the quality of life and increases longevity. In particular, the increase in longevity due to public health care is stronger for low-income than for high-income individuals — thereby reducing the degree of health inequality due to income. Third, both social security and health care entail an element of intragenerational redistribution, since contributions to both programs are proportional to labor income, while benefits – i.e., pension transfers and health care services – are assumed to be constant. The combination of these three elements is critical to explain the political support for health care and social security among low-income individuals. The empirical counterpart of these building blocks of the theoretical model are presented in detail in Section 2. We introduce a dynamically efficient overlapping generation economy with storage technology. Individuals differ in their income, and therefore in their longevity. Agents value their old age consumption and total health care, which is provided publicly and privately. Private health care is more efficient in increasing the quality of life, and therefore in providing direct utility. Public health care is less efficient in rising the quality of life, but it affects longevity. This effect on longevity is non-linear, and is stronger for low-income agents. The welfare state collects a proportional income tax on the young, which finances public health care expenditure to the old and social security transfers. Public health care is available in equal amount to every elderly person at the beginning of her old age, whereas the unfunded social security system pays out a lump sum pension during the entire retirement period, i.e., an annuity. The size of the welfare state and its composition between the two systems are determined in a two-dimensional majority voting game by all agents alive at every election. These types of voting games display two critical features. First, because of the multidimensionality of the issue space, the existence of a Condorcet winner of the majority voting game is not guaranteed. Second, if an equilibrium exists, in absence of a commitment device over future policies, young voters have no incentive to support any intergenerational transfer scheme.3 To deal with these characteristics of the game, as Conde-Ruiz and Galasso, 2003 and Conde-Ruiz and Galasso, 2005, we combine the concept of structure induced equilibrium, see Shepsle (1979) and Persson and Tabellini (2000), with the notion of subgame perfection. The paper proceeds as follows: Section 2 presents some empirical evidence in support of the crucial assumptions underlying the theoretical model, which is then described in Section 3. Section 4 discusses the voting game, and the equilibrium concept, while Section 5 characterizes the politico-economic equilibria. Section 6 analyzes the effects on the welfare spending of an improvement in the health care technology. Section 7 concludes. All proofs are in the Appendix.

In most OECD countries, the welfare state makes contemporaneous use of two programs – public health and social security – to generate a flow of resources from workers to retirees. Both social expenditures have been increasing in the last few decades. Demographic dynamics are certainly responsible, since the number of recipients from these programs – the elderly – has largely increased, but often so have also the per capita resources. This paper provides a new explanation for the contemporaneous existence – and increase – of health care and social security spending based on a political complementarity between these two programs. Political complementarity implies that the existence of health care increases the political support in favor of social security, and viceversa. Philipson and Becker (1998) emphasize the link from social security to public health. The existence of an annuity – the old age pension – increases the value of longevity and, hence, the demand for public health. We focus on the opposite direction, from public health care to social security. We consider a model in which public health reduces the longevity differential between high and low-income agents, thereby increasing the value of the pension annuity to low-income individuals. This effect strengthens the within cohort redistributive component of social security, and increases the political support to this program among the low-income individuals. In a two-dimensional voting model, in which voters determine the size and the composition of the welfare state, we show that this political complementarity may lead to the adoption of a large welfare system, composed of both programs, in which the public health component is large, relatively to social security. Our theoretical model carries an interesting testable implication. As health care becomes more effective in raising longevity and in reducing the longevity differential – due to more public spending and/or to technological improvements – the return from social security for low-income agents should increase, relatively to those for high-income individuals. Fig. 9 provides supporting evidence in this direction by reporting the differences in the internal rate of return (IRR) from social security for subsequent generations of US retirees according to socio-economic status – namely income – as provided by Gokhale and Kotlikoff (2002). Using a dynamic microsimulation model, Gokhale and Kotlikoff (2002) estimate the social security’s treatment for several cohorts of US individuals, from those born in 1945–49 till those born in 1995–2000. Computing IRR requires data on social security contributions and benefits. While contributions' histories for the early cohorts may be available, data on future contributions and benefits need to be forecasted. Fig. 9 summarizes the difference in internal rate of returns between individuals in the lowest and the highest income quintile for three cohorts, 1945–49, 1970–74 and 1995–2000, under different scenarios. “Current rules” refers to a scenario with current contribution rates and pension benefits. In this case, the IRR differential is expected to increase slightly, from 4.9% to 5.1%. Yet, due to population aging, the current level of contribution rate may not be sufficient to finance future pension benefits at the current generosity level. The remaining sets of columns in Fig. 9 presents the IRR differentials under five alternative reform scenarios. With the exception of the reform featuring a tax rate increase (around the year 2000), in which case the IRR differential increases only temporarily (i.e., for the 1970–74 cohort), in all other reform scenarios the difference in IRR between low and high-income individuals is forecasted to increase. The magnitude of this change depends on reform option considered.These estimates hence provide supporting evidence to the relevance of this notion of political complementarity between health care and social security in the debate on the contemporaneous raise in health care and social security spending. Further research in this area should follow to help evaluating quantitatively how much of the increase in social security could plausibly be explained by lower mortality differentials, and to incorporate this political element in the forecast of future spending. Appendix A.