This Master Thesis provides a cross-country comparison of the German and Italian pension reform trajectories between 1989 and 2008 against the backdrop of the Open Method of Coordination in the field of old-age social security. Starting from the observation that, at the end of the 1980s, Italy and Germany faced similar challenges related to the long-term financial sustainability of their respective pension system, this study highlights the social implications of the reform strategy adopted in each country. More precisely, the issue of pension reform is approached form a distributive angle, building on the work of both Camila Arza on the new distributional principles of contemporary pension policies and of John Myles on the issue of intergenerational fairness with respect to old-age social security reform.
The Italian and German pension reform trajectories are dealt with in separate chapters, focusing on the way in which the subsequent reform packages affected the social distribution of rights, resources and risks associated with old-age as well as on the way in which the costs of the reform are apportioned between generations. The conclusion provides for a comparison between the two reform trajectories studied. It highlights some remarkable similarities at the level of the architectural design of the two pension systems as well as with respect to the introduction of self-adjusting mechanisms within the annuity formula but also stresses the obvious differences between the reform strategies implemented over the last two decades, in
particular with regard to the issue of collective burden sharing between generations.
Contents
Introduction
Chapter 1: Distributional effects of Italy’s pension reform trajectory (1989 – 2009)
1. Essential characteristics of the Italian Pension System at the end of the 1980s
2. Challenges and pressures for reform
3. The direction of Reform
3.1. Towards a multipillar system and a clear separation of functions
3.2. Towards a harmonization and tightening of eligibility criteria
3.4. Towards reduced collective risk pooling
4. The Costs of Reform
Chapter 2: Distributional effects of Germany’s pension reform trajectory (1989 – 2009)
1. Essential characteristics of the German Pension System at the end of the 1980s
2. Challenges and pressures for reform
3. The direction of Reform
3.1. Towards a multipillar system
3.2. Recalibrating the eligibility criteria
3.3. Towards a self-regulating benefit formula
4. The costs of Reform
Conclusions
Bibliography
Introduction
Pension reform has been the topic of a hot, European-wide debate for at least twenty years. Although European coordination in this field has only been recently institutionalized under the banner of the so-called “Open Method of Coordination” (OMC), in many Member States (MS) incisive pension policy reforms have been implemented since the early 1990s. As indicated by its title, this Master’s thesis looks at the issue from a cross-country perspective, focusing on the reforms enacted in Germany and Italy. Whereas our choice for these two countries is in part related to our own linguistic background (we felt more secure choosing two countries whose languages we master sufficiently in order to have easy access to the necessary primary and secondary sources), it is not entirely arbitrary. There are indeed a number of prima facie observations which led us to believe that the reform trajectories of Italy and Germany would provide interesting material for a cross-country comparison.
First, it can be argued that, at the end of the 1980s, Italy and Germany faced remarkably similar challenges with respect to the financial sustainability of their respective pension systems, both in terms of their size and in terms of their underlying causes. According to a study performed in 1986 at the request of the IMF, Germany and Italy ranked first and second respectively in the list of the major industrial countries of that time when judged by their total pension expenditure as a percentage of GDP. In 1980, they spent 13,3% and 10,5% of GDP on old-age social security, whereas other developed countries such as the UK, the US and Japan spent 5,8%, 6,3% and 4,2% respectively. Since both countries were confronted with an already high dependency ratio[1] and unfavourable demographic projections, under unchanged circumstances, by 2025, this ratio was expected to further increase to 21% and 20% respectively, by far the highest numbers of all industrialized countries (Heller et alii 1986: 1-11). Given these common challenges, at first sight, it comes as no surprise that both countries have travelled along similar paths, albeit at a different speed. Over the past twenty years, both Germany and Italy have switched from a monopillar to a multipillar pension system and both have radically cut the overall benefit level by introducing a self-adjusting annuity formula. However – and this is a second reason why we believe a comparison between Germany and Italy might prove interesting – these similarities at the level of the newly implemented pension systems conceal some remarkable differences at the level of the redistributive effects of the enacted reform strategies.
This brings us to the scope of the research we conducted in the context of this thesis. Given its intrinsic complexity, the topic of pension reform can be approached from various angles. In developing our own understanding of the issues at stake, we decided to concentrate on the social aspects of the pension reform policies carried out in Italy and Germany. This implies that we will be less concerned, in the following, by the issues of financial sustainability that triggered the reforms, although we will give the issue its fair amount of attention. The focus of our own research, however, will be on the various ways in which the reforms implemented since the early 1990s affected the distributional logic underlying the Italian and German pension systems as well as on the way in which the cost of reform is allocated between generations.
More specifically, we will study the reform trajectories of Italy and German by means of two sets of complementary research questions. The first is mainly inspired by the work of Arza on the distributional principles of contemporary pension policies (Arza 2006a; Arza 2006b). Like Arza, we conceive of pension policies as systems of redistribution with both a social and a temporal dimension, apportioning rights, resources and risks across the population and throughout the individual’s lifecycle. Applying this conceptual model to the cases of Italy and Germany, we will analyze the way in which the successive pension reforms in both countries have altered the distributional patterns underlying their old-age social security systems at the levels of 1) eligibility criteria 2) distribution of costs and benefits and 3) pooling of demographic, economic and financial risks within and between generations. As will become apparent throughout the analysis, despite the many differences in speed and applied reform strategy, Italy’s and Germany’s respective pension reform trajectories since the 1990s display some striking similarities as they both move in the same overall direction towards a tighter eligibility criteria, a more actuarial distribution of resources and reduced collective risk sharing.
Our second set of research questions is mainly inspired by John Myles’ work on intergenerational justice (Myles 2002). It focuses not so much on the way the pension systems of Italy and Germany redistribute rights, resources and risks before and after the successive reforms but on the way in which the rising costs associated with population ageing are allocated among generations. Although it seems commonly accepted since the 1990s that, in order to cope with the rising costs of an ageing population, the pension systems of most EU 15 countries had to be drastically reformed, the question how the costs of these reforms were to be apportioned between generations remained. As we will demonstrate throughout the analysis, the pension reform trajectories of Italy and Germany provide very different answers to this question.
Chapter 1: Distributional effects of Italy’s pension reform trajectory (1989 – 2009)
Compared to most other EU 15 MS, Italy only recently started to take measures in order to alleviate the pressures of its ageing population on the country’s public pension system. Whereas from the late 1980s onwards, almost all MS enacted more or less drastic reforms aimed at reducing retirement spending, Italy continued to enhance the generosity of its public pension system for specific occupational and social-economic categories until the early 1990s. As a result, pension expenditures as a percentage of GDP rapidly increased throughout the 1980s while giving rise to marked disparities between occupational categories (Ferrera & Gualmini 2000; Franco 2000).
Both the rapid growth of total pension expenditure and the increasing fragmentation of the old-age social security system were brought to a halt during the 1990s by means of a series of reforms, which not only gave way to a more transparency and homogeneity but also initiated a number of radical shifts at the level of the overall design of the system. In what follows, we will study these successive reforms in more detail. We will start our analysis by a brief description of the Italian public pension system of the late 1980s. Next, we will study the social consequences of the Amato, the Dini and the Berlusconi reforms of 1992, 1995 and 2004 respectively, focusing on the way in which these successive reforms altered the patterns of redistribution of rights, resources and risks as well as on the way in which their costs are allocated across generations. Instead of devoting a separate subsection to each of the subsequent reforms, we opt for a thematic approach, highlighting the main trends throughout the entire period.
1. Essential characteristics of the Italian Pension System at the end of the 1980s
What holds true for the overall architecture of the Italian welfare state holds true as well for its pension system: although it is closely related to the pension systems of other continental European states, it is important not to overlook the distinguishing characteristics which set it apart from that of neighbouring countries.
The similarities are most readily apparent at the level of the financing mechanism underlying the Italian public pension system. Like that of most EU-15 countries, the Italian system of the 1980s is contribution-based and operates on a pay-as-you-go basis (PAYG). Whereas so-called “fully-funded” or “capitalized” pension schemes are invested in the capital markets, PAYG pension schemes are funded on the basis of “repartition”. Instead of being invested, the social security contributions paid by the working population at a given moment in time “t” are immediately used to fund the pension benefits of the retired population at that same moment “t”. In exchange, the working population gains the right to future pension benefits, payable at the moment t+1 and funded by the social security contributions of the working population of that time. Thus, the pension benefits enjoyed by each generation are financed directly by the social-security contributions of the next generation. Since, in this kind of system, the equilibrium between social security contributions and pension benefits is directly dependent upon the ratio between the working-age and the retirement-age population, PAYG pension systems are particularly exposed to the pressure of population ageing. As we will see in the next paragraph, however, this is only one of the challenges the Italian pension system was facing at the end of the 1980s.
The second important feature the Italian pension system of the 1980s shared with most other continental European countries pertains to the formula used to calculate retirement benefits. Like the majority of continental European pension systems, the Italian old-age social security system in the 1980s was of the “defined-benefit” type (DB). Under such an arrangement, the worker is granted a monthly benefit on retirement that is predetermined by a formula based on such variables as earnings history, tenure of service and age. Whereas in a “defined contribution” pension system (DC), the old-age social security contribution is fixed and the benefits received are dependent upon the rate of return of the invested contributions, in a pension system of the DB type, it is the monthly pension benefit that is predefined in the sense that the formula for computing it is known in advance. Usually, the pension benefit level is determined on the basis of the so-called “reference earnings”, which can be calculated in a number of different ways, as an adjusted average of a predefined number of career years or as an (adjusted) career-long average. Theses “reference earnings” are then multiplied by the number of contributory years and a, more or less generous, accrual factor indicating the percentage of assessed income that enters into the annuity formula. In some countries, the DB formula also includes predetermined credits for non-contributory periods spend raising children, performing one’s military service or obtaining an educational degree. In most cases a DB formula looks similar to the one below:
Pension annuity = RE * n * af
With RE = Reference earnings
n = period of assessed contributions
af = accrual factor
Despite these two common features between the Italian pension system of the 1980s and those of most other continental European countries it is important not to lose sights of the distinguishing characteristics that set it apart. These become readily apparent when one compares the historical evolution of the Italian pension system to that of other European MS. Like that of its continental European counterparts, the post-war Italian pension system is often said to be “Bismarckian” in nature, a label which underlines its status-based character and points to the fact that it provides specific pension benefit schemes to specific occupational groups, whereas the pension systems of the “Beveridgean” kind grant universal benefit entitlements to every citizen reaching the retirement age (independent of their individual contributions to social security, working-life histories or income). Although the distinction between “Bismarckian” and “Beveridgean” pension systems is still often used in the current literature, the reality it refers to has substantially evolved. Since the Second World War, most originally “Bismarckian” pension systems have gradually generalized pension rights to (almost) all occupational categories. During the same period, they have also become less internally fragmented, partially harmonizing both the eligibility criteria and the annuity formulas across the various occupational categories. The Italian pension system of the 1980s, however, still maintained its status-based character to a larger extent. More than most other European pension systems, it presented a highly fragmented picture distributing pension rights and resources in an uneven way across occupational categories (Jessoula & Ferrera 2006: 62).
Thus, for instance, pension entitlements were distributed unevenly across the population, with different requirements as to the retirement age and the minimum number of contribution years for different categories of the population. The same applies to the eligibility criteria giving access to the so-called “pensione d’anzianità” or seniority pensions, allowing many workers to retire ahead of the statutory retirement age after a pre-defined number of contributory years. Reinforcing the fragmented nature of the overall system, the contributory period required to enjoy the “pensione d’anzianità” also varied greatly across occupational groups. Finally, the reference earnings period and the formula used to calculate the level of pension benefits further increased the divide between “privileged” and “underprivileged” occupational categories. The table on the next page summarizes the main features of the Italian public pension system at the end of the 1990s. It highlights the status-based nature of the system which privileges specific occupational groups over others. Among the privileged occupational categories, both the public servants and the self-employed stand out:
- Public servants did not only enjoy a generous annuity formula (based on their last salary and resulting in an average replacement rate[2] of 95%[3] ), the “seniority pensions” system also offered them the possibility of early retirement, independent of their age, after only 20 years of contributions, a settlement which came to be known by the telling name of “baby pensions”.
- As underlined by Natali, the self-employed enjoyed a favourable contribution/benefit ratio. For many years, benefits were established at disproportionally generous minimum levels when compared to the contributions made, resulting in a permanent deficit of the self-employed pension funds throughout the 1980s. In 1990, the self-employed were integrated in the overall “sistema retributivo” applicable to both public servants and private sector employees, a reform measure which, given the very low contribution rate of 12%, even further enhanced their “privileged” status. (
illustration not visible in this excerpt
As the table also makes clear, private sector employees were considerably less well treated by the system, facing tighter eligibility criteria for the seniority pensions than public servants and a less favourable contribution/benefit ratio than both public sector employees and the self-employed. However, treatment of this third category was far from homogeneous. As benefits were calculated on the reference earnings of only the last five years, the system clearly favoured employees with a steep career development over those whose earnings remained more or less constant over their entire life-cycle.
2. Challenges and pressures for reform
By the end of the 1980s, the Italian economy was no longer one of the star performers among continental European countries. After decades of mainly export-led growth, in the early 1970s the pace of economic development began to slow down. When the second oil shock hit the Western world in 1979, even high inflation rates failed to protect the country from further decline in growth rates and a steady rise in unemployment. Despite an average annual inflation rate of 11,8% Italy’s growth rate steadily declined to the level of 2,5%.[4] While this was still relatively high compared to some neighbouring countries, the Italian economy suffered from increasing unemployment reaching 9,7% in the second half of the 1980s, with very big differences between different age, gender and geographical groups (Bertola and Garibaldi 2002).
At the same time, Italy’s life expectancy gradually increased, from 74 to 76,9 between 1980 and 1989 whereas the fertility rated dropped to 1,2 children per woman. As a result of these demographical evolutions, the old-age dependency ratio steadily increased throughout the entire decade from 13,1% to 14,9%, i.e. well above the OECD average of that time (11,7%).
The most immediate challenges faced by the Italian welfare state of the early 1990s, however, relate to the dire state of the country’s public finances. In 1990 the government’s gross financial liabilities amounted to 97,6% of GDP whereas the annual deficit amounted to 11,4%. These figures were not only much higher than the EU average at that time (57% and 4,3% respectively) they were also far removed from the targets set in Maastricht for the accession to EMU Italy was keen on joining.
One of the most important factors contributing to these worrisome figures is directly related to the pension policy adopted by the subsequent Italian governments since the Second World War. The reforms implemented between the 1950s and the 1970s were all carried out without any prior investigation of their long-term impact, leading to an “impetuous expansion” of the system (Jessoula & Ferrera, 2006: 62). Looking back, it seemed as if the governments of the 1960s and 1970s were unconcerned about the rising costs of the pension system or at least unwilling to counter the rising cost by unpopular expenditure cuts. As Natali puts it:
[To the Italian authorities of the 1960s and 1970s] the progressive expansion of the pension system [looked like] a positive sum game, in which everybody won something and nobody lost. In reality, the hidden costs […] were (implicitly) transmitted to the next generations by means of deficit spending and a growing public debt ( 2007: 107, my translation)
Early 1992, however, the eruption of the wide-scale corruption scandals that became known under the name of “Tagentopoli” and the subsequent “Mani Pulite” investigations that brought down the First Republic opened a window of opportunity for radical reform. In addition, pressures from the European level to rapidly bring the public finances in line with the Maastricht criteria for accession to EMU, helped in quieting possible sources of resistance. When in that same year 1992, the lira was forced out of the EMS after a humiliating devaluation, the government pushed through the first of a series of reforms which, in subsequent years, will thoroughly change the Italian pension system.
3. The direction of Reform
Between 1992 and 2004, Italy embarked on a series of four successive pension reforms, three of which were finally adopted by Parliament. In order to discuss the overall direction of the subsequent reforms, in the following we have chosen to adopt a thematic instead of a historical, narrative approach. We will subsequently analyze the changes in the overall architectural design of the pension system (3.1.); the tightening and harmonization of the eligibility criteria (3.2.); the move towards a more actuarial model of resource distribution (3.3.) and the implications of the new pension system at the level of risk pooling across socio-economic actors (3.4.). Complementing this analysis, the final paragraph of this first chapter will be devoted to the issue of intergenerational fairness (4.).
3.1. Towards a multipillar system and a clear separation of functions
At the level of the architectural design of the pension system, the reforms of the 1990s produced two marked changes, the first one being very similar to a change observed in other EU 15 countries, whereas the second is more typical for the Italian situation:
- Like many other European countries[5], during the 1990s, Italy moved from a monopillar pension system operating entirely on a PAYG basis to a multipillar system, adding two fully-funded complementary pension pillars to the existing first pillar which continued to operate on the basis of repartition
- Like most “Bismarckian” pension systems, the Italian pension system of the 1980s combined the functions of “poverty prevention” and “income maintenance” within the same institutional set-up. The reforms of the 1990s led to the separation of both functions through the replacement of the contributed-based “Pensione sociale” by the tax-financed “Assegno Sociale”
The move towards a mixed multipillar system was initiated in 1993, one year after the reform of the public “first pillar”. Since this recent reform implied a (comparatively modest) cut in the future replacement rates for most categories of workers (cf. infra), the aim of the “decreto legislativo 124/1993” was to allow future retirees to compensate for the loss in future income by means of an additional occupational pension scheme. The decree created a legislative framework for two types of occupational pensions: the so-called “closed” and “open” funds. As the names they were given imply, the main difference between these two types of schemes resides in their accessibility for the individual and in the role played by the social partner organisations in creating and controlling them. Whereas the so-called “closed funds” are established by collective agreement between social partner organisations and are accessible only to the members of the union which co-initiated the fund, the “open funds” are created by institutions operating on the financial market, independent of any collective sector agreement and accessible to everyone, regardless of their employment relationship. Initially, the decree conceived of the “open” funds as individual social insurance schemes, but nowadays collective adhesion by employees of one and the same company or working in the same industrial sector is also possible, making them more akin to the traditional “second pillar” occupational pension schemes observed in other EU 15 countries. (Jessoula & Ferrera, 2006: 91-92)
Seven years after the “riforma Amato” of 1992, the final step in the transition towards the current multipillar system was taken by reinforcing the third private pension pillar through the introduction of the “polizze pensionsitiche individuali”. Together with the existing “fondi aperti” to which one adheres on an individual basis and the convertible life insurance contracts, they constitute an additional individual means of old-age social security for those wishing to complement their basic public and occupational pensions. (Jessoula & Ferrera, 2006: 92)
Given the tight budgetary constraints illustrated above, the transition from a monopillar to a three-tier pension system did not succeed without any difficulties. The main problem for the government consisted in finding an elegant solution to the so-called “double payment” problem. As indicated above, in a pension system funded on a PAYG basis, workers already pay for the pension benefits of the retired generation by means of their own social security contributions. Erecting a second and/or third fully-funded pillar to complement the first one therefore requires the working generation to reserve a second portion of their income to save for their own “additional” pensions, which, in effect, amounts to making them pay twice. In most countries which performed the transition towards a multipillar pension system, the government usually has tried to solve this “double payment” problem by providing attractive fiscal incentives to those willing to contribute to the fully-funded second or third pillars. In Italy, however, this solution was far more difficult to implement given the budgetary constraints the government was already facing. The Amato government found an alternative solution, using the so-called “Trattamento di Fine Rapporto” (TFR) to provide for the necessary funding to get the second pillar of the ground. Introduced in the early 1980s, the TFR is a type of severance pay to which all Italian dependent workers are entitled when their employment contract is terminated by their employer. The solution to the “double payment” problem consisted in obliging all the private employees entering the labour market after 1993 and willing to subscribe to an occupational pension scheme to transfer their TFR to the relevant fund.
A second major change at the level of architectural design is related to the double function of the Italian pension system. As most “Bismarckian” systems, the Italian old-age social security system of the 1980s served two objectives, integrating them into one single institutional framework: the same social security contributions were used 1) to reduce the risk of old-age poverty among those categories of workers unable to set aside a portion of their income for retirement (by guaranteeing a minimum “Pensione Sociale”) and 2) to ensure relative income replacement for the rest of the population. With the “riforma Dini” of 1995, however, both objectives of the public pension system were clearly separated from one another and ascribed to a separate layer within the system. The contribution-based “Pensione sociale” was abolished and replaced by the so-called “Assegno sociale”. Set aside from the contribution-financed “sistema contributivo” introduced by the same “riforma Dini” and which entirely focuses on the goal of income replacement (cf. infra), the “Assegno sociale” is financed by general taxation and allocates a basic residual pension on a means-tested basis.[6]
3.2. Towards a harmonization and tightening of eligibility criteria
As indicated above, the Italian pension system of the 1980s presented a fragmented picture, distributing entitlements and resources in an uneven way across occupational groups. One of the intended effects of the reforms was precisely to close these gaps between the “overprivileged” and the “underprivileged”.
As far as the distribution of pension entitlements is concerned, the successive reforms display a double trend towards universal tightening and partial harmonization. In chronological order, the most important measures that contributed to these trends were the following:
The “riforma Amato” of 1992
- increased the statutory retirement age for private sector employees from 55 to 60 years for women and from 60 to 65 years for men, thereby bringing the requirements for a full pension closer to those applicable for public sector employees
- Abolished the much criticized “baby pensions” enjoyed by public sector workers and increased the minimum contributory period for seniority pensions to the level of 35 years (i.e., the level that was applicable for private sector employees).
- extended the minimum contributory period required for seniority pensions from 15 to 20 years for both private sector employees and the self-employed
The “riforma Dini” of 1995
- abolished the different retirement ages for men and women, introducing a flexible retirement age between 57 and 65 in an attempt to reduce the incentive for early retirement. Within this system, a full pension is awarded only to those who retire at 65; below this threshold actuarial deductions form the full pension are applicable
- abolished the “pensioni d’anzianità” for younger workers and tightened the eligibility criteria for middle-aged and older workers (57 years of age plus 35 years of contributions or 40 years of contributions)
The amendments to the “riforma Dini” introduced by the Prodi government
- gradually increased the contribution rates for the self-employed from 12% to 19%
At first sight, these various measures aimed at tightening the eligibility criteria for both old-age and seniority pensions might seem drastic. However, given the financial strain on Italy’s public finances during the 1990s they are by no means exaggerated when compared to the measures that were implemented by other EU MS around the same time. The most radical innovation brought about by the reforms of the 1990s lies elsewhere, namely at the level of the redistribution of resources.
[...]
[1] The old-age dependency ratio represents the amount of people aged 65 and over divided by the amount of people aged 15 to 64.
[2] The replacement rate is the ratio between first monthly pension and the last salary. It is commonly used as a measure of the generosity of the average pension.
[3] Compared to 80% on average for private sector employees (Natali 2007: 168)
[4] Except were indicated otherwise, all the macroeconomic data quoted in this paragraph is based on the StatExtracts made available online by the OECD via http://stats.oecd.org/Index.aspx?DataSetCode=EO77_MAIN#
[5] On the spread of the so-called “Multipillar paradigm” in Europe, see Bönker 2004.
[6] Although, the move from the “Pensione Sociale” to the “Assegno Sociale” does not seem to have brought about any significant change in the eligibility criteria and has not significantly altered the level of minimum pensions, it did have implications for the way in which economic risks are defined and distributed among the population. We will return to this below.