In most developed countries, the government is the dominant financer of health care services. Over the past few decades, however, the role of the private health sector and private health finance has increased (Maarse Reference Maarse2004; Propper and Green Reference Propper and Green2001). Although the fiscal crisis in the 1970s was an important catalyst for welfare state reform, market-oriented reforms in health care went far beyond being mere instruments of budget constraints and cutbacks by focusing instead on the established institutional arrangements that regulated the provision and consumption of health care services (Blomqvist and Rothstein Reference Blomqvist and Rothstein2005; Gingrich Reference Gingrich2007; Helderman Reference Helderman2007). Privatisation, choice and regulated competition became buzzwords in the contemporary debate on public health care in addition to being guiding concepts of the market-oriented reforms. This article is concerned with one particular aspect of this change: the rise of Private Health Insurance (PHI). The article focuses on the Danish case and compares it with similar cases in other industrialised countries.
In the past decade the number of people taking out PHI, which duplicates existing public universal coverage, has increased dramatically in Denmark. In 2001 PHI was held by a small exclusive group of people, whilst in 2007 more than 30 per cent of the labour force had some form of PHI. Along with the rise in PHI, a private hospital sector has sprung up. In the 1990s a few troubled private hospitals existed, whilst by 2007 at least 178 private hospitals and clinics had been established. This development is somewhat surprising, given that Denmark is a typical Scandinavian welfare state with a universal health care system of relatively high quality, which in general has enjoyed wide public support. The aim of this article is to explain what happened and consider the consequences of this development.
Private health insurance on the rise in Denmark 2001–2007
The argument presented here is that the development of PHI in Denmark can not be explained by citizens opting-out of the public health care system, whether due to low satisfaction with the service or long waiting lists, as suggested by economic theory. Neither is it explained by institutional approaches, suggesting path-dependency (Pierson Reference Pierson2000) or the resurrection of a dormant path (Schneiberg Reference Schneiberg2007). Nor is it the result of a public debate for and against private health care or the deliberation of MPs in parliament. Rather, it is the result of complex dynamics, institutional constraints and both political action and non-action. The development is a gradual institutional change which has resulted in the addition of a new institutional layer to the Danish health care system. Further, the development of PHI has had unintended consequences, regarding both equality and public finance.
The process is an example of ‘policymaking without policy choice’ – a process where the final outcome was not what was chosen or expected by the policymakers who initially launched the programme. It does not mean that no choices were made during the policymaking process. Rather these choices, often small and apparently insignificant, can expand through unforeseen dynamics to an outcome that is far from expected; Big changes are not always the result of big causes, but can be attributed to the accumulated effect of rather small initial reconfigurations of the rules. While a particular government or its opposition may regard changes in a particular rule to be marginal and its short-term impact to be limited, the accumulated and long term consequences can be significant. If only minor rules regarding PHI are altered, while changes takes place in the broader economy and in the labour market, over time, PHI may become popular with a significant proportion of the labour force. Such developments may be further re-enforced because of pressures on policy actors, such as public employers or unions, to support such a development, as shown in the case below.
PHI covers a range of meanings (OECD 2004). It can refer to supplements to public health care or to the public finance of health care, while other PHIs replace national health care insurance. Another type of PHI is considered be an alternative source of funding to increase the capacity of a health system. In Denmark there are different schemes, but what is being explained here is a rise in the number of people who hold private health care insurance giving access to treatment in private hospitals or clinics. These insurance schemes duplicate existing public universal coverage, offering a private alternative. Danes holding PHI also continue, however, to rely upon publicly financed hospitals because PHI does not cover all types of medical care – a common feature in many duplicate markets. Private hospitals focus on a limited range of elective services, leaving the responsibility for more expensive services (or populations) to the public health system (OECD 2004). Other examples of large duplicate markets are Ireland, Australia and New Zealand (OECD 2007), and countries such as Greece and France have also experienced rises in the uptake of PHI in the past decades (Saliba and Ventelou Reference Saliba and Ventelou2007; Tountas, Karnaki et al. Reference Tountas, Karnaki, Pavi and Souliotis2005). In Australia, for example, duplicate PHI covers a much larger part of the population (up to 50 per cent) and has done so both before and after the introduction of universal health care in 1975. There have also been relatively important changes over time, with the period 1985–1999 having seen a large decline in PHI coverage, while in 1999–2002 numbers rose again (Colombo and Tapay Reference Colombo and Tapay2003).
The remainder of the article is organized as follows. The next section reviews different theoretical explanations of institutional change and suggests that the rise of supplementary PHI constitutes a new institutional layer to the Danish health care system. Section three briefly presents the history of health care insurance in Denmark relevant to the understanding of the analysis. Section four analyses the policy process towards PHI and considers the unintended consequences. Section five concludes and discusses the theoretical and empirical implications.
1. Possible explanations
The academic literature on PHI primarily stresses two factors which explain growth or decline in PHI purchase in countries with a national health insurance. The first factor is people's perception of the public health care system (often proxied by the length of waiting lists in public hospitals) and the second is the price of the insurance relative to income. The exiting literature departs methodologically from micro-economics and the assumption of economic calculationsFootnote 1; A patient who needs non-urgent medical treatment is often placed on a waiting list in the public health care system prior to seeing a consultant or receiving treatment. Economic theory suggests that private insurance purchase is sensible to this problem since, in the private sector, insurance holders are guaranteed (almost) instant access to treatment if necessary. We could therefore expect that a rise in PHI would be closely related to rising waiting lists. Besley et al. (Reference Besley, Hall and Preston1999) find that in Britain, where 14–17 per cent of the population hold PHI, the popularity of PHI is a result of dissatisfaction with the National Health Service and, in particular, the length of waiting lists. The study also reveals that household characteristics (income) matter, and the typical PHI-holder is relatively well-off, middle-aged and belongs to a low-risk group with regards to medical needs. Similar results are found in Spain (Costa and Garcia Reference Costa and Garcia2003) and in France (Saliba and Ventelou Reference Saliba and Ventelou2007). These explanations do not provide a satisfying explanation for the Danish case, however. Waiting lists increased in the 1980s and 1990s and there was much public debate surrounding this about waiting lists and the quality of public health care. Waiting lists became an important political issue at the beginning of the 1990s and attempts to reduce them were made, however, without much success (Christiansen, Enemark et al. Reference Christiansen, Enemark, Clausen, Poulsen and Mossialos1999; Pedersen, Christiansen et al. Reference Pedersen, Christiansen and Bech2005; Vrangbæk, Reference Vrangbæk2004). At that time PHI schemes existed, but almost no-one purchased them. In the last decade, a period in which PHI has risen steeply, waiting lists have been reduced steadily for a wide range of the treatments for which there was most demand (Ministry of Health and Internal Affairs 2007). Furthermore, in the same period (from 2002–2007) there was a guaranteed maximum 2 months waiting time in public hospitals. Although waiting lists may play a role in explaining the rise in uptake of PHI from 2001–2007, it seems that it cannot alone explain the rise; waiting lists were also present in the 1990s (and were possibly even longer at that time) and PHI was available from the late 1980s. Considering citizens support more broadly, the large majority supported a public sector health care system at the end of the 1990s (J. G. Andersen Reference Andersen2000). In short: in the period where PHI became more popular, waiting lists had been declining, there was support for the public health care system and the price of PHI to individual consumers remained unchanged. Although certain changes in tax-policy (as explained below) have lowered the real cost of PHI in particular circumstances, the individually based rational choice explanation seems insufficient to explain the growth of PHI in Denmark. The explanation is rather to be found in institutional theory and changes in the institutional setting in Danish health care.
Institutional explanations
The new institutionalism has developed a large literature on the stability of institutions and the constraints institutions have on preferences and policies. Institutional theory has focused on the consistent nature of institutions and on finding order and stability. The basic claim is that welfare state institutions i.e. the existing structures of social provision, health care and education, constrain the possibilities of policy actors. The outcome of reform efforts is therefore to a large extent the result of a particular institutional set-up in a given country at a given time – thereby privileging structure over agency. For the same reasons, revolutions in health care and health insurance are rare, because health care systems are surrounded by a range of institutions and political actors (Immergut Reference Immergut1992). National health insurance programmes are usually not accepted automatically as a part of the welfare state: they are often politically contested. Health care insurance, whether public or private, is an issue of principle, which stirs passions and mobilizes politicians and interest groups on a massive scale (e.g. Hacker Reference Hacker2002, ch. 5). As Immergut (Reference Immergut1992, p. 1) states, ‘National health insurance symbolizes the great divide between liberalism and socialism, between the free market and the planned economy. Doctors, unions, employers, and other interest groups actively engage in national health insurance conflicts. Political parties look to national health insurance programs as a vivid expression of their distinctive ideological profiles and as an effective means of getting votes’. Although the issue at hand is not whether to establish a national health service or not, the establishment of a duplicate private health care insurance system is an issue we could reasonably expect to engender political contestation and protest by opponents. The institutional perspective has given great insight into how systems are stable and reinforced by policy feedback: the influence of existing welfare state institutions shapes future developments and creates a tendency for path-dependence in welfare state reform (Pierson Reference Pierson2000; Steinmo, Thelen et al. Reference Steinmo, Thelen and Longstreth1992), which would suggest that change outside of the path of the public health care system is unlikely in Denmark. In a similar vein Esping-Andersen argues that welfare state change towards privatisation and retrenchment of services is less likely in universal and well established welfare states, not so much because of their size but because of the structure of provision (Esping-Andersen Reference Esping-Andersen1990, p. 33). From such a perspective, privatisation in the Danish (and Scandinavian) setting of public, integrated health care systems with a strong degree of decentralization is an unlikely policy change since it challenges a number of existing structures and institutions. As recently as 2002, an academic evaluation of the Danish health care system concluded that ‘The market thrust in other countries appears rather alien to the whole Danish health service and indeed to Danish culture more generally’ (Mooney Reference Mooney2002, p. 170). However, the path-dependency argument fails to explain the changes in Danish health care – which has deviated from the traditional path of public provision and financing.
Within historical institutionalism the question of change has been dealt with in two ways: One strand has argued that changes are incremental and small scale, and will work within the set framework and ultimately not change anything profoundly. The other strand was the original historical institutionalist framework, where change was to come about through punctuated equilibria (Steinmo, Thelen et al. Reference Steinmo, Thelen and Longstreth1992). Only at these junctures would path-breaking change come about. However, there have not been any such ground-breaking events in Danish health care politics that can explain the rise of a private sector health care insurance to supplement the public health service. With its main focus on explaining stability, the historical institutionalist framework comes short of explaining the gradual institutional changes which have led to the increased uptake of PHI in Denmark.
As a response to the impasse of the historical institutional approach in explaining gradual institutional change in the absence of external shocks, Marc Schneiberg (Reference Schneiberg2007) has suggested that institutional change can be path-breaking without having to give up the institutionalist framework. What Schneiberg suggests is that institutional change will draw on already existing ideas or ‘paths not taken’. That is, instances of change, which may seem new, are often in fact drawing on existing structures. Schneiberg argues that what is defined as an institutional ‘path’ theoretically is often not as uniform and pure as we might like to believe. Instead, paths may contain within them ‘ambiguities, multiple layers, potentially decomposable components or competing logics which actors can use as vehicles for experimentation, conversion, recombination and transformation.’ (Schneiberg Reference Schneiberg2007, p. 52). Schneiberg's argument sustains an internalist approach to institutional change, rather than change coming from outside the existing institutional setting. Sometimes changes occur when fragments of institutional ideas from the less dominant – or even dormant – paths resurrect and combine with elements of the dominant path to create something new.
The health care insurance company ‘danmark’ has for a long time been part of the Danish health care ‘path’. It was established in 1973, when a system of ‘sick funds’ (sygekasser) was abolished and public universal health care insurance was installed. ‘danmark’ did not initially offer duplicate insurance, but instead offered coverage of parts of the co-payments (in particular for medicine and dentistry) for its members. The company has over 2 million members and therefore covers around 36 per cent of all Danes (in 2007). Following Schneiberg's argument we would expect that people would choose this existing insurance system to get coverage for private hospital care, which ‘danmark’ also offers. However, as shown in Table 1, the number of members who have coverage for private hospital care has been relatively stable – it has only risen by 27 thousand persons – over the past decade.
Members of ‘danmark’ (000's) with coverage for private hospital care/operations 2001–2007
Source: Data provided by ‘danmark’.
Thus, the suggestion that institutional change will tend to draw on ‘what is already on the path’ does not seem to be adequate in explaining the institutional changes in relation to private health care insurance in Denmark.
Finally, Thelen (Reference Thelen2004) has argued that one model of institutional change can be labelled institutional layering. Layering is the grafting of new elements onto an otherwise stable institutional framework. Such amendments can alter the trajectory of an institution's development (Thelen Reference Thelen2004, pp. 35–36). This concept relies on the framework of historical institutionalism but offers a more nuanced tool for the analysis of institutional change compared to the ‘punctuated equilibrium’-model. Without disregarding the idea that institutional change can be a highly discontinuous matter, and the consequence of exogenous shocks or punctuated equilibriums, the idea of institutional layering is different in its essence. It suggests that even in ‘settled’ times there is contestation and renegotiation going on, which over time, or due to small changes, adds up to significant changes, through the addition of different systems or layers to those which already exist. Whilst institutional analysis has suggested that institutional reproduction and institutional change are distinct analytical problems, Thelen suggests that these two elements should be studied together. Large parts of the health care system may prove resilient to change, while more subtle and smaller scale changes add a new layer to the health care system, which may be very different in nature from the overall institutional arrangement. I argue that PHI constitutes a new layer in Danish health care and show how small changes in policies interact with particular economic and labour-market circumstances and (accidentally) add up to a dramatic rise in the number of PHI holders.
The article employs a broad focus on how the public policy process played itself out by tracing the reform process. A broader analytical scope allows the analysis to capture the complexity of institutional change (P. A. Hall Reference Hall and Mahoney2003) and account for a significant policy change which happened incrementally without dramatic disruptions (Streeck and Thelen Reference Streeck and Thelen2005). The following sections consider the institutional constraints, the development of policies and the various stands of political actorsFootnote 2.
2. The Danish health care system
The Danish public health care system has been the principal source of medical care for the vast majority of Danish residents since the consolidation of national health insurance in 1973 (Vallgårda and Krasnik Reference Vallgårda and Krasnik2007). Real health care expenditure has increased over the past 15 years (OECD 2006; Vallgårda and Krasnik Reference Vallgårda and Krasnik2007) and in international comparison, Danish health care expenditure lies close to the OECD average (OECD 2006). The health care system is decentralised to counties (amter), which until 2007 had the right to collect taxesFootnote 3. The main source of income for the health care system is general tax revenue, although user charges apply to a range of services such as prescriptions, physiotherapy and dental treatment.
In parallel with most other Western health care systems, the Danish health care system came under pressure in the 1980s. People in the industrialised nations demanded quality health care, delivered efficiently, equitably and with a focus on the specific needs of the patient. At the same time technological development made more and more sophisticated operations and treatments possible. As western countries all have health care systems that are largely publicly funded, governments were met with enormous pressure to raise funding to meet demands (Mossialos and Grand Reference Mossialos and Grand1999). Moreover, governments and academic analysts started to believe that the best way to understand health care financing was Wildavsky's Law of Medical Money: ‘medical costs will rise to equal the sum of all private insurance and government subsidy’. In other words, the demand for health care spending is endless and health care systems are not self-stabilizing (Wildavsky Reference Wildavsky1977). After the health care crisis of the 1970s (of ever rising expenditure) the Danish health care sector was therefore met with cost-containment measures, primarily by introducing fixed annual budgets with a ceiling on overall expenditure for municipalities and counties and global fixed budgets for single hospitals (Christiansen, Enemark et al. Reference Christiansen, Enemark, Clausen, Poulsen and Mossialos1999). The strict financial limits on health care imposed in the 1980s led to an increase in the length of rising waiting lists – a phenomenon also observed in other OECD countries (Besley, Hall et al. Reference Besley, Hall and Preston1999; Hurst and Siciliani Reference Hurst and Siciliani2003).
3. Why PHI in Denmark?
Until 2001, PHI schemes offering treatment at private hospitals were very limited. Around 43,000 people held such policies in 2001, corresponding to less than 1 per cent of the population and approximately 1.5 per cent of the labour force. PHI was not on the political agenda, with supply and demand limited. However, since 2001 PHI has exploded in popularity. In 2004, 11 per cent of the labour force held PHI schemes and in 2007 that figure was 30 per cent, with almost 850,000 Danes holding PHI. Insurance providers expected (in 2007) growth to continue and that 1 million Danes will hold PHI schemes within a year or two, which would then include more than 35 per cent of the labour force (Hebsgaard Reference Hebsgaard2007).
In the broadest institutional perspective Denmark is a classic Scandinavian welfare state, in which the state provides a range of universal services from childcare to education, health care and elderly care. These public services were created with the aspiration of being of such a high quality that there would be no need for a private market (the public services would crowd out the private alternative), and hence, through a ‘one size fits all’ service-model, a high degree of equality would be achieved. Denmark has a universal public health care system based on the fundamental principle of free and equal access for all. In other words, the Danish public health care system is a part of an institutional package of public services, which in general enjoys both political and public support. Such institutional arrangements are often considered fairly robust and resistant to change, unless something radical occurs (Pierson Reference Pierson1994; Steinmo, Thelen et al. Reference Steinmo, Thelen and Longstreth1992). From such a perspective it would seem unlikely that PHI would be able to obtain broad popular support. So what happened?
Falling trust in the public health care system
In most European countries people basically agree that ‘taking care of the ill’ should be a primary public responsibility (Taylor-Gooby Reference Taylor-Gooby and Jowel1999). The same goes for the Danish case, where 94 per cent of the population in 1994 believed that health care was definitely a public sector responsibility (J. G. Andersen Reference Andersen2000). Throughout the 1990s there was much public debate about the waiting lists in – and the quality of – the Danish public health care system. These debates affected the population's trust in public health care. Combined with a more general discussion of the more long-term problems of public sector finance, this contributed to a popular sense that a satisfying public health care system might not exist in the future. In 1999, 30 per cent of Danes had the impression that the public hospitals did not offer an adequate service, however the majority (69 per cent) of the population more or less agreed that they would be able to get a fully satisfactory treatment in the public health care system (J. G. Andersen Reference Andersen2000, p. 11). In a survey from 1998, hospital patients generally showed high levels of satisfaction and a majority of citizens were satisfied with the hospitals (Finansministeriet 1998). But when it came to people's perception of private vs. public hospitals, the picture was somewhat different, as shown in Table 2.
Where is hospital treatment preferred? Perceived quality in public and private health care
Source: For the survey in 1999: (J. G. Andersen Reference Andersen2000). For the survey in 2008: Survey conducted by Zapera.com for Mandag Morgen, spring 2008.
In 1999, 44 per cent of Danes would prefer a private hospital if they had the choice without costs and only 25 per cent had a public hospital as their first priority. The public hospitals came out even worse with regards to quality assessment. Almost no-one believed that quality of care was best at public hospitals (J. G. Andersen Reference Andersen2000). Even though the public health care system enjoys public support at the broadest level, the perception that the private alternative was superior to the public system was widespread in 1999. By 2008 the clear preference for private hospital treatment had faded (see Table 2).
Andersen (Reference Andersen2000) shows that in 1999 there was also little trust in there being sufficient economic resources in the public health care system in the future. 72 per cent of the population expected there to be more user-payment in the future; 74 per cent believed they would have to pay themselves if they wanted to avoid long waiting lists; 53 per cent thought that people would have to purchase PHI if they wanted to be sure to get the best possible treatment. However, in 1999, only 34 per cent of the population were interested in having PHI as a labour market fringe, while 52 per cent were not interested in PHI at all. On the other hand 70 per cent agreed that expenditure for the public health care system should be increased, which could indicate a more general support for the public system. Andersen (Reference Andersen2000) concludes that the main reason for the interest in PHI and private health care solutions is mistrust and low expectations of the quality and financing of the public health care system. Yet, very few people had purchased PHI in the 1990s. In 1999 PHI schemes were not on people's minds as a solution to the perceived problems in the quality of public health care. But that was soon to change, due to changes in tax-policy.
Changes in tax-policy: the genie out of the bottle
The Danish tax structure gives limited incentives to offer labour market fringe benefits to employees, because fringe-benefits (e.g. a telephone or a car) are not free to the employee, but taxed according to the general tax-law which states that all income is taxed, regardless of the nature of the income (whether in cash or other valuable assets). However, in 1998 the Social Democratic government proposed a break with this principle, suggesting that employees who were offered alcohol rehabilitation by their employer should not be taxed for that specific good. In the preface to the law the government stated: ‘A range of goods should not necessarily be regarded as a part of the salary, but as a natural part of the company's employee-policy. For example, a company can have a general alcohol-policy, which also includes rehabilitation to employees who have an alcohol problem. There can be a need to [re-]consider the [tax-rules on] fringe-benefits that employers offer their employees, apart from the salary, to avoid tax-rules standing in the way of companies' possibilities of taking a social responsibility’ (L149, 1999, author's translation). A particular Social Democratic feature of the law was that employers were required, for equality reasons, to insure all employees in a company (i.e. not only top-executives) in order to achieve tax-exemption for the receivers. The law was enacted and gave tax-exemption for alcohol rehabilitation which employees received either through direct payment by the employer or through the coverage of an insurance scheme. During the parliamentary debate of this proposal, the right wing opposition was eager to extend the proposal to include more broader health care insurance schemes for tax-exemption. The Social Democratic government, however, refused, arguing that such an extension ‘could evolve into an attack on the public health care system’ (L149, 1999). The social democrats had nevertheless planted the seed for a new institutional layer to grow.
Extending tax-exemption to private health insurance
When a new government of right-wing parties took over government in 2001 they pursued their wish to extend tax-exemption to include health care insurance. The government proposed a revision of the tax law (‘ligningsloven’) in 2002, extending the former government's law on tax-exemption for alcohol rehabilitation to also grant tax-exemption for employees' private health care insurance. The law had several aims: (1) to encourage companies to show social responsibility, by caring about their employees' health, (2) to ensure that employees would receive quicker treatment when covered by private insurance schemes, (3) for employers to have fewer employees on sick-leave and to reduce the economic and organisational costs related to employee absence and (4) to bring about a positive effect on the broader economics of health care (the state saves money when insured persons are paid for by private insurance and waiting lists in the public hospitals will be reduced for those who do not have private insurance).
The law had (in its final draft) the same premise for tax-exemption as the former: that all employees in a company had to be insured in order for the insurance to be tax-exempted. The law was enacted with the votes of the government parties (Denmark's Liberal Party and the Conservative Party) and their supporting party (the Danish People's Party). The Social Democrats voted against the law, arguing that they could not support a law which would break the fundamental principle of free and equal access to health care and contribute to creating a two-tiered health care system. In particular, the Social Democrats argued that the law would contribute to inequality by excluding all those people, who were not on the labour market, from tax-free PHIFootnote 4 (L97, 2002).
The tax-exemption law coincides with the rise of PHI in 2002 and kick-started this development. Tax incentives for PHI is a policy tool also used in Australia and Greece to boost PHI uptake (Colombo and Tapay Reference Colombo and Tapay2003; Tountas, Karnaki et al. Reference Tountas, Karnaki, Pavi and Souliotis2005). But to conclude that this is the sole explanation of the extraordinary rise would be flawed. There are other tax-exempted fringe benefits, but that does not mean that they are held by 30 per cent of the labour force. For example the second most popular tax free fringe benefit is a ‘home-PC’ held by approximately 8.7 per cent of the labour force (Ministry of Taxation 2007). With PHI being both tax-deductible for companies and tax-exempted for the receiver, the real cost to purchasers (employers) of PHI became very low, and free to the receivers. As a fringe benefit PHI became a valued asset for employees, paid for, by and large, by the common tax pool.
Tax-exemption and equality
In 2002, when the government announced the introduction of the law on tax-exemption of PHI, the ‘equality-premise’ was not in the initial proposal. In the first draft of the law companies could choose who, among their employees, would be offered tax-exempted PHI. The Confederation of Danish Industries (one of the largest employer organisations) was pleased with these less strict rules and wrote in their comments on the law: ‘we have noted with particular satisfaction that tax-exemption is not contingent upon whether all employees are offered the same health insurance’ (L97, 2002, app. 1, author's translation). However, the Social Democrats worked hard to convince the government to maintain the ‘equality-premise’ in the law. Further, the government-supporting party (the Danish People's Party) decided only to support the law if tax-exemption was contingent upon companies covering all employees. Paradoxically, this clause sought by the Social Democrats and the Danish People's Party, which was not the intention of the right wing government, has contributed significantly to driving the number of people holding PHI. By way of a counterfactual: had the law not included this equality-premise, companies would most likely have chosen to offer PHI to exactly those employees who they considered most valuable, CEOs and other leading employees, who were unlikely to amount to significant numbers.
Despite this equality-enhancing premise, evidence shows that PHI is not equally distributed among the population. First there is an insider/outsider split: the unemployed, the elderly (once retired people do not uphold private coverage) and low-paid workers with unstable jobs do not hold PHI. Second, surveys also indicate that, among the working population, there are also large differences in who is covered by PHI. Whether a person has a PHI scheme is closely related to income-level as shown in figure 2. People earning more than the average wage are much more likely to hold PHI than people with below-average or average salaries. Job-positions also matter: 65 per cent of the high position privately employed hold PHI, while only 17 per cent of low position public employees hold this insurance (Olsen Reference Olsen2007). Further PHI is more widespread among white-collar workers than among blue-collar workers (Iversen Reference Iversen2007).
Income brackets of holders of PHI, 2008
In a peculiar way the ‘equality-clause’ of the tax-exemption law in 2002 has not resulted in equality among PHI holders, which was the aim; rather it may – independently – have had the unintended consequence of boosting the success of PHI. Analyses on the Danish case (Glavind Reference Glavind2008), thus, confirm the findings from other countries: PHI is held mainly by well-off, well-educated, middle-aged persons. Since around 90 per cent of PHI-holders get it through their workplace, most are also active on the labour market and most often in white-collar jobs. In other words, it is the low-risk groups that hold PHI, like in the UK and Spain (Besley, Hall et al. Reference Besley, Hall and Preston1999; Costa and Garcia Reference Costa and Garcia2003).
The cost of private health care insurance
With PHI tax-deductible for companies and tax-exempted for the employee, PHI represents a loss of tax-revenue for the state. In the comments on the tax-law it was estimated that in total 150,000 people would be covered by PHI reducing tax-revenue by approximately 90 million Kr. with an additional 10–20 million Kr. for the treatments paid directly by the employer. The estimate was based on the assumption that a relatively limited group of employers would offer employees PHI. However, considering that PHI would reduce other public expenditure, as mentioned above, the net-revenue loss was estimated to be only 35 million Kr. (L97, 2002). PHI is also tax-deductible for companies (in the same way that most other company expenses are) but no estimate of the tax-revenue loss of tax-deduction of PHI exists. The exact value of the tax-deduction of PHI is unknown, as no registers are kept and employers are not obliged to report figures on PHI. While the tax revenue loss was calculated in 2002 to be approximately 90 million Kr. pr year, in 2007 the revenue loss amounted to 645 million Kr. – a rise of 716 per cent in only 6 years. The accumulated loss of tax-revenue between 2002 and 2007 amounted to 1.9 billion Kr.Footnote 5. A report from The Ministry of Taxation in 2007 showed that PHI is now the fringe benefit held by most employees. PHI represents the second largest fringe benefit by value (fewer people have company cars, but they represent a larger value), and PHI is the fastest growing fringe benefit and the rise in PHI is likely to continue (Ministry of Taxation 2007, pp. 10,12). This implies that a large part of the financing of PHI (and indirectly private hospitals) comes from the general tax-pool.
The steep increase in the loss of tax-revenue up until 2007, further indicates that the estimates of PHI in the 2002 tax-law were wrong. The Economic Council of the Labour Movement objected to the economic estimates made in the original law-proposal, arguing that the net-cost would not be 35 million Kr, but rather 80–90 mill. Kr (L97, 2002, appendix 6). But no-one at the time of the proposal in 2002 was able to predict, or expected, that it would rise to the level it had risen to in 2007. The report from the Ministry of Taxation (2007) warned that the rapid growth in fringe benefits (created by the rise in PHI) might pose a threat to public finances.
As argued by Immergut (Reference Immergut1992), health care politics has many actors and public institutions that are likely to affect policymaking. This is particularly true in corporatist Denmark. The remainder of this section considers the positions of employers, unions and the insurance industry.
The employers' position
One of the arguments for making PHI tax-exempt was that companies could save money by purchasing PHI because it would save them the costs related to the absence of employees who were stuck on waiting lists in the public health care system and therefore not able to work. As noted earlier the employees holding PHI (particularly in white-collar sectors) are not a high-risk group. The risk of the need of surgery is low compared to other groups (e.g. blue-collar workers or the elderly). Further, the assumption that employees will report sick and be absent while on a waiting list for medical attention or an operation lacks empirical evidence. On the contrary, evidence shows, with some uncertainty, that few employees (3–8 per cent) report sick while they are on waiting-lists (Petersen Reference Petersen2005, ch. 9). Most likely there are other reasons for why companies offer PHI as a fringe benefit to their employees.
The employer's organisations were positive towards PHI from the start and in particular favoured tax-exemption for employees. Tax-exemption gave employers the possibility to offer employees a fringe benefit which employees believed had relatively high value, but with low real costs. At the same time it offered employers the possibility of deploying an employee-policy which signalled social responsibility, which was important in the competition for labour. PHI is dominant in, but not exclusive to, private sector companiesFootnote 6.
2003–2007 were times of relatively high growth in the Danish economy and in 2007 the economy was growing at such a pace that there was fear of overheating. Employment was very high and unemployment in 2007 was the lowest (3.2 per cent) in more than 30 years. Therefore, the pressure on the labour market was high – there was a shortage of labour supply, which led to an increase in salaries of up to 5 per cent in 2007 (DEC 2007). With a heated economic situation on the labour market, fringe benefits became a competition parameter for companies in the attempt to attract and maintain employees and PHI as a fringe benefit displayed a domino-effect throughout the labour market. A survey conducted in 2007, showed that 81 per cent of the population expected that employers would offer PHI as a part of their salary (Rechnagel Reference Rechnagel2007). The suggestion that PHI is a labour market fringe benefit, highly dependent on the health of the economy is not exclusive to Denmark; it is also well recognised in Britain, where PHI rose in times of economic prosperity, but stagnated in the recession of the early 1990s (Timmins Reference Timmins1995). While it is not surprising that employers were content with PHI as an almost free gift to the employees in times of short labour supply, the absence of loud protest from the unions or the medical profession is more puzzling.
Unions left in a dilemma
As Vennesson (Reference Vennesson2007) argues, process tracing should also consider ‘non-action’ when appropriate. In this case, the (non-) reaction from unions can be considered as such. Unions are traditionally strong actors in the Danish highly institutionalised corporatist political system. The expectation would be that unions would object to the rise of PHI and seek to protect the universal health care system they themselves had initially helped to create on the premise of large risk-pooling and free and equal access for all citizens. Although objecting to the tax-exemption of PHI, unions have been somewhat paralysed on the issue and find themselves caught in a dilemma with regard to PHI. On the one hand they are defending the free and equal access to the universal public health system which they took part in creating throughout the 20th century. On the other hand unions face rising demand from their members to include PHI in collective bargainingFootnote 7 (I. H. Andersen and Madsen Reference Andersen and Madsen2008). The Danish Confederation of Trade Unions (LO) were ‘strongly dissatisfied’ with the tax-exemption of private health care insurance and fear that tax-exemption of private health care insurance is a step towards ‘American conditions’ in health care (LO 2003). The unions argue that private health care is unfair and will create a health-divide, both between those who are offered PHI and those who are not (mainly low-paid), but more importantly between those who are in the labour market and those who are outside of it. The General Workers Union in Denmark, which mainly organised un-skilled, low-skilled and skilled workers, proposed in 2001 that members should be offered a PHI, which was met with harsh criticism from most other unions (P.N., 2001). However, as early as 2002 a Gallup survey showed that 63 per cent of unskilled workers were very interested in private health insurance (UgebrevetA4, 2002). By the end of 2007 a survey showed that 20 per cent of the members of the same union, now called Working in Denmark, held some form of PHI (3F, 2007), indicating the pressure on unions from their members on this issue (3F, 2006). By 2009 some unions include PHI in the collective bargaining process and even a few public municipalities offer PHI as a part of the fringe package to employees.
The dilemma facing Danish unions has similarities (but also differences) to the questions faced by American unions in 1940s and 1950s. The American unions had argued for a universal health care insurance, but realising that it would not succeed in that objective and seeing that employee health insurance had spread through the workforce, they changed positions. These developments reduced the interest of unions in covering workers in a government program of protection, and increased their interest in collectively bargained private health benefits (Hacker Reference Hacker2002, pp. 221–242). As PHI has risen in popularity among Danish workers, a similar dynamic has come into play.
The medical professions are often powerful interest groups in health care questions (Immergut Reference Immergut1992). They enjoy a recognised monopoly on medical practice and doctors should have been able to affect or even block the rise in PHI, as it is the general position of the Danish Medical Association (DMA) that the rise in PHI could potentially threaten the free and equal access to the public health care system. Needs (by professional judgement) ought to determine who gets medical care first, not the size of one's wallet, argues the DMA. The DMA has generally opposed the tax-exemption of PHI (Jensen Reference Jensen2007) and was not involved when the tax-law was changed in 2002. Since PHI has become popular among citizens, it seems that the government can comfortably ignore the DMA's arguments and protests.
Insurers: Skip the queue!
Finally, there are also the interests and actions of the suppliers of PHI. Insurance companies had long been eager to widen the extent of PHI in Denmark. Health insurance was one of the only ‘blank spots’ on the Danish insurance market in the 1980s and 1990s, while it had existed – and been a profitable business – for a long time in many other countries. When PHI came on the agenda and became tax-exempted in 2002, insurance companies saw new sales opportunities and launched a large commercial offensive. Prime time television commercials aimed at individuals played on the notions of long waiting lists in the public sector and the security of having direct access to private health care. Further, PHI often includes access to psychological therapy (e.g. due to stress) and physiotherapy, which also gives PHI the broader image of ‘well-being’ insurance (e.g. massages etc.). Further more, as the CEO of the largest PHI supplier (Danica) explained in 2002, the commercial campaign was also aimed to encourage companies to make use of the new law on tax-exemption of PHI (Birkemose Reference Birkemose2002). Around 90 per cent of PHIs are purchased by employersFootnote 8.
4. Conclusion
The paths welfare states end up taking are often not the result of carefully considered policy choices. Relatively small changes in public policy can create a process of institutional change, where one step leads to another and institutions and actors come to support (or at least not resist) developments which accumulate to form a significant large scale change. Such a process is one of ‘policymaking without policy choice’. This does not imply that no policy choices were made. But at the point of decision, no involved actor expected the resulting outcome. Indeed, policy makers had no idea of the subsequent dynamics that would be put into play with PHI and unions did not expect their members to demand PHI and nor would anyone a decade ago have expected that a social-democratic municipality would offer PHI to all its employees.
Policymaking without policy choice is not a situation of stasis – it is one where a development is set in motion and continues, creating new dynamics as it moves along. The concept of policymaking without policy choice has similarities to what Rose and Karran (Reference Rose and Karran1987) have called Inertia change, where the cumulative effect of small changes add up to have a large impact. Rose and Karran also lay out the political logic of such types of developments; once citizens are given certain advantages (e.g. tax-free PHI) the political cost of reversing it can be high. Equally, the political options for reversing the PHI-path seem somewhat limited in the Danish case: PHI is now a very popular fringe benefit both among employers and employees and the political cost of attempting to reverse the development could be high, as more and more employees get access to it and value it. Further, as Rose and Karran suggest, such processes often involve unforeseen and unintended consequences, which was also the case in Denmark: stated polemically, PHI is an add-on to the health care system for the well-off, which is financed through general taxes paid by all. The ‘equality-premise’ has boosted the scale of PHI to an unexpectedly high extent. It has become a highly demanded commodity and because of this, is becoming a drain on public finances. Such self-enforcing and unintended institutional dynamics are hard to predict, but can be important drivers of further change in welfare state policy.
PHI has become a part of Danish health care politics in a selective fashion (Kjær and Pedersen Reference Kjær, Pedersen and Campbell2001). There were no radical reforms of public policy, the process was gradual and through conjunctural causations led to private sector growth. There was no withdrawal of the state; public health care finance is increasing and in fact the state indirectly finances a large part of the private sector growth. Nevertheless, PHI constitutes a new institutional layer in Danish health care and is the result of minor policy changes placed in a specific economic context. This analysis shows how this new institutional layer came into existence in the absence of a punctuated equilibrium, but rather through a process of gradual institutional change. As such, a process of ‘policymaking without policy choice’ can lead to path-breaking outcomes through a slow evolutionary process, without any critical junctures or exogenous shocks assumed in the traditional institutional literature (for a discussion, see Thelen Reference Thelen2004; pp.23–31).
Why were the existing political institutions not constraining this development? While PHI constitutes a new institutional layer in Danish health care, it is not a frontal attack on the existing public sector provision, but rather the slow introduction of a competing system, where private alternatives are offered alongside the universal public health care system. This new alternative sets new political and economic dynamics in motion. Since these new layers do not directly undermine existing institutions, they have not initially provoked much counter mobilisation (Rothstein Reference Rothstein1998; Streeck and Thelen Reference Streeck and Thelen2005). In 2008, 99.1 per cent of the population believed that it is the government's responsibility to provide health care for the sick and 80.9 per cent thought that the government should spend more money on health careFootnote 9. Thus, the rise in PHI has not corroded the support for the traditional public service model in Denmark.
