11.6.3. Progressiveness of funding

The previous sections address how much money is being spent and how resources are collected. This section considers how the funding burden is distributed within the population. Payments are progressive if higher income groups pay disproportionately more than those on lower income (De Graeve and Van Ourti, 2003). The distribution of the financial burden and the degree of progressiveness and regressiveness differs across funding sources. .

The extent to which a funding system will redistribute income from higher to the lower income groups depends on both the progressiveness of revenue collection and the incidence of public spending. In order to achieve the same redistributive effect as a progressive system, a proportional system must unequally distribute benefits (Ervik 1998). It could be argued that a less progressive system in which public spending benefits the lower income groups disproportionately may create a better situation for low-income people than in a more progressive system, but with less public spending for the poor. However, public spending on healthcare may be difficult to separate from overall public spending which may also redistribute revenues. A longitudinal perspective is needed to understand the redistributive effect of a healthcare system in order to account for redistribution between periods of wealth and periods of poverty over a lifetime. While most redistribution studies focus on one point in time, the few longitudinal studies generally show a redistribution from ‘lifetime richest’ to ‘lifetime poorest’, although the redistribution is relatively flat (Mossialos and Dixon, 2002).

Among the different types of taxation, there are different degrees of progressiveness. One study of healthcare financing in OECD countries found the UK and Italy to be the most progressive (Wagstaff et al, 1999). The study also disaggregated the funding sources to examine their contribution to overall progressiveness. Direct taxes are progressive in all countries, while indirect taxes were regressive in all countries except Spain in 1980 (this may result from higher value-added taxes on luxury goods). Moreover, among the EU member States of the time, direct taxes are progressively distributed while indirect taxes are regressive according to Kakwani indices: indirect taxes constitute a larger proportion of income of poor people than wealthier people. The regressive effect of indirect taxes can be seen in the UK. Lower income households pay a greater proportion of their income on indirect taxes (32%) than higher income households (11.3%) (Glennerster, 1997). In examining the tax system as a whole, in 1998-1999 the UK tax system was found to be slightly regressive due to indirect taxes, with the lowest income quintile paying 40% of income in taxes, while the highest income quintile paid 36% of their income on taxes (Commission on Taxation and Citizenship, 2000).

There has been an increasing reliance on indirect taxation in many Member States since the mid-1980s. This not only increases income inequality but also reduces the progressiveness of healthcare finance. The share of value added tax as a proportion of total taxation increased from an average of 15.4% in 1980 in OECD countries to 19% in 2005, along with a slight decline in the share of direct (personal and corporate) income tax (OECD 2007, as cited in Thomson, Foubister and Mossiaos 2008).

The extent of progressiveness of income taxation also depends on the number and rates of marginal tax bands, where fewer tax bands and low marginal tax rates will create a regressive system. Income tax in France, Germany, the Netherlands, Sweden and the UK appears to be progressive, with income being transferred from the highest income quintile to the rest of the population (Zandvakili, 1994). Progressive taxation also depends on the relative role of national and local tax collection. National taxation has been found to be a more progressive system of financing than local taxation. For example, in Finland, an increase in the average rate of local income taxes led to a decline in progressiveness in the early 1990s (Klavus and Hakkinen, 1998). Moreover, tax-funded systems are more progressive than countries relying more on social and private insurance like the Netherlands, Germany, Switzerland and the US (Wagstaff et al, 1999).

Within social health insurance systems, the degree of fairness depends on whether or not the contributions are mandatory. It also depends on the existence of ceilings for contribution rates; payments are progressive up to a ceiling, and then regressive. In cases where individuals are either not allowed - as in the Netherlands up to 2006 - or are not obliged - as in Germany - to stay in the public system, the payments become regressive (De Graeve and Van Ourti, 2003). In Germany, about 21% of the population can choose to opt out. However, only 7% choose to be fully covered with private insurance. As a result, the social health insurance systems in Germany and the Netherlands have been found to be regressive (Wagstaff et al, 1999). Conversely, in France, with the recent extension of insurance coverage for the costs of user charges and expanded contribution basis to include total income of employees, there is a higher degree of risk pooling across the population, with a positive impact on equity.

In terms of the redistributive effect of tax and benefit systems, there is considerable variation across countries. Comparing eight countries in the 1990s, the funding system in Sweden redistributed the most, reducing income inequality by 50%, followed by Denmark and Germany, reducing inequalities by more than 40% and the UK redistributed the least, with a 35% reduction in inequalities. However, it is likely that the observed redistribution resulted more from social transfers than from taxation (Ervik, 1998).

The redistributive effect of social insurance funding has been studied in greater depth in Germany. Three types of interpersonal redistribution are seen in Germany: (1) due to varying health risks, there is considerable horizontal and vertical redistribution with the renunciation of experience rating; (2) dependents are insured, suggesting redistribution from single people and couples to people in large families. In addition, all insured people are equally entitled to healthcare services, regardless of previous contributions; (3) intergenerational redistribution between employed and retired people (Hinrichs, 1997). However, it is argued that redistribution is more effective in a tax-funded system due to limited income equalization through social health insurance and negative economic effects of linking insurance contributions to earnings (Lutz and Schneider, 1998).

Contrary to taxation and social insurance where people contribute on the basis of their ability to pay, private funding increases inequity because it shifts the funding burden away from such population-based risk-pooling arrangements towards out-of-pocket payments by individuals and households with a pro-rich distributive impact (Evans and Barer, 1995; Creese, 1997). Furthermore, international comparisons of progressiveness in healthcare funding reveal that healthcare systems that are largely privately funded are more regressive than those in which funding is predominantly public (Wagstaff et al, 1999). Importantly, with private contribution mechanisms access to health care depends on the ability to pay and risk of ill-health as opposed to the need for healthcare, in contrast with public contributions.

However, while private health insurance has been found to be regressive in France, Ireland and Spain, contributions appear to be proportional to income in Finland and even progressive in Denmark, Germany, Italy, the Netherlands, Portugal and the UK (Wagstaff et al, 1999). Some argue that by encouraging (or mandating) high income individuals to purchase private health insurance, this will make the financing system more progressive, since the rich will pay proportionately more than the poor. However, private health insurance may also skew the provision of services to favour the higher income groups.

While studies to date remain cross-sectional, and no study has been conducted to measure the change in progressiveness of financing in the transition countries of central and Eastern Europe, it can be estimated that health financing in this region has become less equitable because of the increasing reliance on out-of-pocket expenditure, payroll taxes with a ceiling and direct consumption taxes.

The WHO World Health Report 2000 devised an alternative measure of fairness of healthcare financing. This formula is based on the goal that healthcare payments should not be linked to consumption and that there should be a proportional relationship between the ability to pay and healthcare payments. Thus, a fair system of financing would be one where the ratio of total health contribution to total non-food spending is identical for all households, regardless of income, health status and utilization; the index would take a value of 1. This method was criticized for being insensitive and unreliable, along with the use of estimation rather than explicit calculation of the values for each country (Wagstaff, 2002) (De Graeve and Van Ourti, 2003; Musgrove, 2003). According to data from the 1993-2000 period, the country with the fairest financing system appears to be Slovakia, followed by the UK, Sweden, and Denmark, with the most unfair in Latvia, followed by Portugal, Greece, Bulgaria, and Croatia. Similarly, when examining the proportion of households with catastrophic payments (defined as representing more than 40% of their total disposable income), countries faring particularly poorly are Latvia, Portugal, Greece and Bulgaria with between 3%-4% of households reporting catastrophic payments. Note that comparisons between the Wagstaff et al and WHO studies cannot be made because - among other aspects - they use different methodologies (the former is based on household income data, the latter on aggregated expenditures at country level) and measure different things (proportionality versus progressiveness).