The global social insurance movement began with German social insurance legislation of the 1880s. Social insurance is defined as national compulsory systems to insure workers against the possibility of losing their income through industrial injury, sickness, old age, or unemployment. The global movement for social insurance established the historical core of the welfare state as we know it today through the German model of social insurance, associated with Bismarck, and also planted the seeds for the Soviet model of social insurance to be associated with Stalin. This movement proceeded throughout the twentieth century, involving almost all the countries in the world; it was not until the 1980s, with the rise of neo-liberalism, that it reached its peak. From the 1990s, however, the Soviet model collapsed with the socialist systems, whereas the German model remained the dominant form despite the rise of the new global model of privatization.

The worldwide character of the social insurance movement has been known since its inception, but no previous historical studies have traced it at the global level. Historians have tended to address social insurance one nation at a time, though Daniel Rodgers has explored the issue in the USA in connection to the North Atlantic context.Footnote ¹ A few comparative studies on the diffusion of social insurance across nations, sociological and quantitative in nature, take the German model of social insurance as the global norm; they explicitly exclude socialist states.Footnote ²

The global social insurance movement has been a process of diffusion, incorporating more and more countries around the world. For the German model, the movement started in Germany, then first diffused to north-western Europe and European overseas settlements, followed by southern and eastern Europe, South America, and North America, and finally the rest of the world; for the Soviet model, its fundamental ideas were diffused and experimented with through the global socialist movement but the model was implemented beyond the Soviet Union only in the post-war era and within the socialist world.

In addition, the global social insurance movement has been a process of evolution, gradually extending coverage to more and more social risks. For the German model, researchers have shown a consistent pattern: a country first adopts insurance against industrial injury, then sickness and/or old-age insurance, and finally unemployment insurance.Footnote ³ By the end of 2007, accident and old-age systems had been adopted virtually worldwide. Systems of sickness and unemployment insurance were also by then almost universally adopted in the developed countries, although many developing countries still lacked these types of social insurance.Footnote ⁴ For the Soviet model, countries adopted a comprehensive system covering at once all the major social risks of injury, sickness, and old age, but with the explicit absence of unemployment insurance because socialist leaders claimed unemployment was eliminated in their economies.

In the remainder of this article, we will begin by constructing an analytical framework labelled as ‘interactive diffusion’.Footnote ⁵ For the purposes of this study, diffusion is defined as being both hierarchical and non-hierarchical, including diffusion from transnational actors (the International Labour Organization (ILO) and the World Bank) to national actors and diffusion from national actors in one country to those in another.Footnote ⁶ The mechanisms of diffusion are either policy learning or policy emulation. Thus, of the many social and historical factors affecting the worldwide spread of social insurance – changing structures of industry and labour; domestic social and political struggles; wars and regime changes; the rise of international organizations – we have chosen to give particular attention to the circumstances and motives of the policy makers who ultimately set up the legislation and bureaucracy for social insurance programmes. Our narrative of ‘interactive diffusion’ – encompassing innovation, diffusion, evolution, and the balance of emulation and learning in policy – breaks into five periods. These include the origins of national compulsory social insurance (1790s–1880s); the rise of two social insurance models (1880s–1918); global social insurance in formation and wartime transition (1919–45); the great expansion in the post-war era (1945–81); and social insurance in the era of privatization (after 1981).

Interactive diffusion of global models: policy learning vs. policy emulation

The principal literature on the origins and growth of the welfare state consists of qualitative historical studies, which not only focussed on the Western industrial democracies but were also dominated by national analysis. These studies constructed three major theories – centred on industrialization, the working class, and the state – arguing for the decisive role of these national, structural forces in the origins and development of the welfare state. The modernization theory originated in Durkheim’s functionalism and was developed by leading post-war social scientists, who regarded industrialization as the major factor in adopting social insurance.Footnote ⁷ Class-based theories arose from the Marxist tradition, stressing the role of working-class movements or of the working and middle classes in demanding social insurance.Footnote ⁸ Another group of neo-Marxists deployed a functionalist explanation of the adoption of social insurance, arguing that it supported the dominance of capital.Footnote ⁹ State-centred theories originated in Max Weber’s theory of the state, stressing the autonomy of the state from the interests of class, and arguing that the structures of politics (including the state’s administrative capacity) played an important role in establishing and expanding social insurance.Footnote ¹⁰ At the margin of these studies, a few pioneer historical researchers – notably Hugh Heclo and Daniel Rodgers – went beyond individual countries to look for social policy diffusion across national borders.Footnote ¹¹ But these pioneer studies were restricted to the European and the Atlantic world.

A second dimension of the literature consists of quantitative social science studies that focussed on testing these theories by correlating the adoption of social insurance systems across numerous national cases. These studies argued that the national historical theories did not hold true for the worldwide adoption of such programmes: that is, that it was transnational diffusion or ‘a world historical process’ rather than the national economic, political, social, and institutional factors that played the critical role in the adoption of social insurance systems.Footnote ¹² Other social science researchers paid attention to the role of international organizations, especially the ILO, in the diffusion of social insurance systems. These researchers agreed that these international organizations played a positive role in the adoption and growth of such systems.Footnote ¹³ More recently, Mitchell Orenstein has focussed on ‘transnational actors’, especially the World Bank, in the spread of pension privatization since the 1980s.Footnote ¹⁴

A third, more recent, section of the literature develops and applies the notions of ‘policy emulation’ and ‘policy learning’ in its analysis of the new global order based on neo-liberal diffusion.Footnote ¹⁵ Policy emulation describes a process in which the motive of policy makers is reputational: that is, to enhance the images and status of the nation in the international world.Footnote ¹⁶ Thus policy makers often follow global norms before there are any domestic issues to address.Footnote ¹⁷ Policy learning describes a process in which the motive of policy makers is to solve pressing domestic problems. Thus, they analyse the various available solutions and assess the effectiveness of these solutions. In both policy learning and policy emulation, the favoured models are those from countries that are similar to theirs in terms of region, language, religion, and colonial origins. But the most favoured models are those viewed as successful.Footnote ¹⁸ These diffusion studies, however, are cross-sectional rather than historical, ignoring national forces as they emphasize diffusion, especially via emulation. They account for one-way dominance rather than two-way interaction.Footnote ¹⁹

Our own interpretive framework is that of ‘interactive diffusion’. This approach incorporates advances from each of the previous approaches and stresses two-way influence. It argues that national forces determine the timing and specifics of the adoption in cases of both policy learning and policy emulation, while the global forces provide basic ideas without determining the specifics of adoption.Footnote ²⁰ Our framework enables us to propose an analytical narrative of the spread and unfolding of social insurance around the world. It addresses a wider range of national experiences and a fuller set of factors than the previous approaches, yet offers an interpretation that we believe is coherent. There are five main analytical points to our interpretation.

First, we argue that two principal models of social insurance – the German capitalist model associated with Bismarck, and the Soviet socialist model associated with Stalin – diffused around the world. Whereas the German capitalist model was established by the initial social insurance legislation and remained significantly stable in terms of institutional structures, the Soviet socialist model had its roots in the German Social Democrats’ responses to the legislation of the 1880s, took shape with the Lenin principles of social insurance in 1912, and then evolved significantly until the 1930s, when it became the Soviet model associated with Stalin.

Secondly, this article further argues that the developed industrial countries adopted the German model through policy learning. These countries already faced pressing social questions accompanying industrialization and were able to assess the German model carefully before they either directly adopted it or indirectly adopted its variants, such as compulsory workmen’s compensation, means-tested old-age systems, and subsidized sickness funds. Compulsory workmen’s compensation originated in the employers’ liability but abolished the infamous ‘principle of fault’, which required workers to prove in court that the accident was the fault of their employer. Means-tested systems had their roots in the early modern poor-law relief systems, often providing general, tax-financed old-age benefits to those who were defined as needy through means tests. Voluntary state-subsidized systems were subsidized by state tax revenue, often providing sickness benefits to the insured.Footnote ²¹ For instance, Scandinavian countries and British dominions evaluated the German model but eventually adopted means-tested old-age pension systems, in decisions determined by their domestic institutions. The Scandinavian countries also adopted state-subsidized sickness funds, in contrast to German sickness insurance, in the late nineteenth and early twentieth centuries. Continental European countries adopted German old-age and sickness insurance systems from the beginning. In the 1940s, a new variant emerged in Britain, the Beveridge model, providing flat-rate benefits of all kinds, including comprehensive medical care, to achieve universal coverage based on flat-rate contribution or simply on citizenship.Footnote ²²

Despite all the differences, these variant systems shared the same ideology as the German model, in that they all embodied the modern concept of ‘social citizenship’ or ‘social right’ to state social protection. All variant systems were transitional, in that they steadily gave way to more standard German measures: voluntary state-subsidized sickness systems were generally replaced by the German model by the 1940s;Footnote ²³ the means-tested old-age systems were mostly replaced in the late 1940s by universal old-age systems based on the Beveridge principles – the latter, however, could not be fiscally sustained, and thus had to be backed up with the German model.

Thirdly, this article argues that socialist parties and socialist countries adopted the socialist model of social insurance (the Soviet model) through policy emulation. Socialist parties had used the German Social Democrats’ model as a tool to fight against the established political system before they came to power. After they did so, they emulated the Soviet model, as a concomitant of Communist Party rule in eastern Europe and as a process of eager emulation in the poorer socialist countries such as China.

The Soviet model differed from the German model not only in its underlying ideology but also in its institutional structure. It asserted social insurance as a fundamental right, but in the context of a ‘workers’ state’ in which the industrial workers and the governing party were seen as the leading elements of society.Footnote ²⁴ Workers, their trade unions, and their party were thus the administrators of the social insurance system; as a result, the trade unions played a role in disciplining the work force, for which there was little parallel in capitalist societies.Footnote ²⁵ In addition, the Lenin principle of socialist social insurance was supposed to cover all citizens but not capitalists and landlords. It was to be financed solely by the employers and the state, and managed solely by the workers themselves. The Stalin model inherited the coverage from the Lenin model but it abolished unemployment insurance. Most importantly, most of the social insurance benefits were paid through the state budget, financed by general taxation.

Fourthly, we argue that the developing countries adopted the German model through policy emulation, encouraged by the ILO. Upon independence in the post-war era, these predominantly agricultural countries were not initially pressed by the social questions accompanying industrialization. Still, in order to improve their international image and status as modern states – and to maintain the support of small but well-placed wage-earning classes – these former colonies emulated the German model upon independence.

However, the former British colonies in Asia, Africa, and the Caribbean adopted Central Provident Funds (CPFs) for their industrial workers upon independence, providing a lump-sum old-age payment at retirement.Footnote ²⁶ These provident funds, publicly managed, are compulsory individual accounts with contributions by both employees and employers. Such shifts were the result of a diffusion process determined by national forces, which began in India in 1948 for coal miners and in 1952 for industrial workers, later expanded to other South and Southeast Asian and Pacific countries and regions, and reached Africa and the Caribbean from the 1960s. However, these CPFs were designed to be transitional to social insurance.

Finally, a new global model emerged with the privatization of the Chilean pension system in 1981, in an atmosphere of expanded neo-liberal thinking. In the 1990s, this new global model spread widely, promoted increasingly by the World Bank. The Soviet model collapsed with the European communist systems in the early 1990s; it was first replaced by the German model and then partially supplanted by the new global model of privatization. But the German model has remained – and seems certain to remain – the mainstream social protection system around the world in this new century.

The origins of national compulsory social insurance: from private to public, 1790s–1870s

The European idea of compulsory social insurance dates back to the end of the eighteenth century. At that time, some municipal governments in Europe undertook a few experiments with social insurance. In England, Francis Maseres in 1773 and John Acland in 1786 proposed general plans of social insurance for parliamentary discussion.Footnote ²⁷ But the first practical proposals for social insurance, according to Gareth Stedman Jones, were put forward in the 1790s, as a direct result of the American and French revolutions, with the aim ‘not solely to alleviate the lot of the poor but to reproduce on European soil the conditions of existence of a viable commercial republic akin to the United States’.Footnote ²⁸ These proposals included those on social insurance by Thomas Paine and by Antoine-Nicolas de Condorcet in France. As Stedman Jones notes, these proposals did not target social questions or emphasize the hostility of workers to capitalists (perspectives that were not yet developed), but they were inspired by Enlightenment ideas, democratic revolutions, and modern republicanism.Footnote ²⁹

During the 1848 uprising in Germany, the demand for sickness insurance arose explicitly in the liberal middle class, under the leadership of physicians who were well aware of the ‘health hazards created by the new economic order and were convinced that the problems could not be solved without a thorough reform of medical services’.Footnote ³⁰ These physicians put forward the concept of ‘the people’s rights to health’ as the most precious property, and emphasized the duty of the state to protect health.Footnote ³¹ Their campaign did not succeed.

Nevertheless, from the mid nineteenth century, German states began to regulate the existing voluntary insurance funds set up by guilds and miners’ associations.Footnote ³² These voluntary insurance funds were often called ‘provident funds’. The Prussian state issued a series of statutes to establish the rights and duties of the members and employers, and to regulate the coverage, benefits, and administration of these funds, even allowing local authorities to set up compulsory funds directly.Footnote ³³ As a result, these voluntary funds became compulsory for both employees and employers; and, more importantly, employers as well as employees were required to contribute to them. The motives for this legislation wave were to reduce the cost of relief for the poor. As E. P. Hennock puts it, ‘Anxiety for poor law finance runs like a thread through the history of provident-fund legislation.’Footnote ³⁴ The Prussian Industrial Code of 1845 required all journeymen to participate in journeymen provident funds and stipulated that industrial workers who did not belong to any funds should join newly established funds set up by the local authority. In 1854, district level authorities were also required to set up compulsory provident funds. The same year, Prussia issued a law on the creation of miners’ provident funds, requiring both employers and workers to contribute to the funds, which were administered by employers.Footnote ³⁵ In 1876, after the unification of the German empire, a new law was passed, requiring all existing provident funds to be registered with the state; these funds were to provide only sickness and death benefits, avoiding any other purposes such as trade union recruitment.Footnote ³⁶ By the 1880s, the existing programmes, whether voluntary or compulsory, had covered only a small portion of the workers: for instance, at that point about two million German workers (about 25% of the total workforce) were insured against sickness and death by various funds.Footnote ³⁷ Nevertheless, the compulsory provident funds became the model for the social insurance legislation of the 1880s.

The rise of social insurance models: capitalist vs. socialist, 1880s–1918

After unification in 1871, industrialization speeded up in imperial Germany, worsening the existing ‘social questions’, which included labour conditions; public health, education, and housing; and workers’ risks that would lead to income interruption or loss from work injury, unemployment, sickness, old age, and death of breadwinners.Footnote ³⁸ As a response to these social questions and to the inadequate protection of workers, socialist movements gained in influence in Germany and elsewhere in Europe. Thus the threat of the socialist movement became a pressing domestic concern for leaders of the newly established imperial Germany.

Bismarck had long been aware of this issue, commenting in 1849 that ‘The social insecurity of the worker is the real cause of their being a peril to the state.’Footnote ³⁹ During his visits to France in 1855 and 1857 and his stay in Paris as ambassador in 1861, he became extremely interested in the state-subsidized pension system. The French state pension system operated from a national old-age insurance bank established by an Act passed on 18 June 1850, under the Second French Republic; it was voluntary, providing old age and disability pensions to working men through payment of premiums at the post office.Footnote ⁴⁰ Bismarck believed that Emperor Napoleon III gained the support of workers because of this system. As Zollner pointed out, more than twenty years later, in 1889, when discussing the old age and invalidity bill in the Reichstag, Bismarck could not help mentioning his impression of the French state pension system and its social effects. Zollner quoted Bismarck as saying that, ‘I have lived in France long enough to know that the faithfulness of most of the French to their government … is largely connected with the fact that most of the French receive a state pension.’Footnote ⁴¹

While Bismarck learned the idea of the state pension system from France, his decision on compulsory social insurance as the right form was determined by German national forces. With the exception of the extreme liberals and conservatives, other political forces (including conservatives, Catholics, socialists, and big industrialists) all agreed with the idea of compulsory social insurance to deal with the various ‘social questions’. Meanwhile, contemporary German scholars of the historical school also advocated state protection in general and social insurance in particular.Footnote ⁴² Among the various social questions – accidents, sickness, old age, invalidity, death, and unemployment – the most urgent issues for the state to address were the rising accident rate and inadequate protection for workers. For instance, in 1879, only 20% of accidents registered with the court were compensated through the ‘principle of fault’ because most workers could not afford the cost of demonstrating fault.

Three kinds of proposal were submitted to the Reichstag to deal with the issue of accidents. The liberals proposed reform of the existing employers’ liability law, and later supported voluntary insurance. Carl Ferdinand Stumm, an industrialist in mining, iron, and steel in the Saarland, proposed compulsory old-age and invalidity insurance through extending the compulsory provident funds to all industrial workers. Louis Baare, the director general of the coal and steel combine in the Ruhr, proposed compulsory insurance for accidents only. Of these proposals, Bismarck liked Baare’s proposal, and in 1880 formed his own policy of providing imperial social insurance for accidents. In March 1881, Bismarck’s accident insurance bill was submitted to the Reichstag, but it had not been enacted by the end of that legislative period and Bismarck ordered it to be withdrawn. Instead, he wrote a speech for the emperor to support his social insurance legislation, which eventually became the Kaiser’s message delivered at the Reichstag in November 1881. This message was regarded as the founding document of the German welfare state.Footnote ⁴³

Thus, in May 1882, a second accident insurance bill was submitted to the Reichstag, together with a bill on sickness insurance. The sickness insurance bill was much less controversial because it maintained many of the features of the existing sickness funds and simply made them national and compulsory, thus satisfying vested interests. So while the sickness insurance bill became law in June 1883, the second accident bill encountered serious opposition and thus did not become law. It was the third accident bill that finally became law in July 1884, after many significant modifications.Footnote ⁴⁴ Overall, Bismarck succeeded in introducing national compulsory social insurance for sickness in 1883 and for accidents in 1884, laying down the foundation for the legislation on old-age and invalidity insurance adopted in 1889. Later, under the Weimar Republic in 1927, Germany adopted its own first national compulsory unemployment insurance and thus had a comprehensive system covering all the four major social risks.

The socialist model of social insurance was being formulated concurrently with the debate on Bismarck’s proposals for social insurance bills. The German Social Democratic Party accepted the idea of compulsory social insurance but attacked the specifics of Bismarck’s bill.Footnote ⁴⁵ Party leaders argued that Bismarck’s social insurance was only another form of poor relief and complained that workers were required to pay. They called for direct progressive taxation of the wealthy. Secondly, they argued that social insurance should cover the entire working class, regardless of industry or occupation. Thirdly, they believed that workers were entitled to benefit levels that represented an adequate compensation based on the principle of social justice. Fourthly, they denounced waiting periods and other restrictive rules on access to social insurance. Finally, they argued in principle for workers’ self-government and control of the social insurance programmes but, in practice, they insisted on workers’ participation in the administrative bodies.Footnote ⁴⁶ At the end of the nineteenth century, however, the Social Democrats, led by Karl Kautsky, sharply changed their policy, setting aside their revolutionary spirit to support social reforms.Footnote ⁴⁷ Thus they began to participate actively in the social insurance systems founded by Bismarck.

Once the German model came into being, it was diffused within Europe and overseas through international associations to insure workers against such major risks as accident, old age, sickness, and unemployment. For instance, Uruguay in 1914, Chile in 1916, and Japan in 1911 all adopted accident insurance or its equivalent, compulsory workers’ compensation. Three major European-based associations formed: the International Committee on Social Insurance (1889), the International Association for Labour Legislation (1890), and the International Association on Unemployment (shortly after 1910), the last devoted to the diffusion of German social insurance systems in the Atlantic world. These associations provided information and publications through conferences and congresses, spreading the basic ideas of the German social insurance model.Footnote ⁴⁸ However, the results varied: Scandinavian countries adopted voluntary, state-subsidized sickness funds (Sweden in 1891 and Denmark in 1892); British territories adopted means-tested old-age systems (New Zealand in 1898 and Australia in 1908); while the other European countries adopted the German-style sickness insurance (Austria in 1888, Norway in 1909, Britain in 1911, and Russia in 1912) and old-age insurance (Austria and Bohemia in 1906, France in 1910, Luxembourg in 1911, Romania in 1912, and the Netherlands in 1913).

Here, again, the basic idea of national social insurance came from Germany, but the timing and specifics of adoption were determined by national forces. In the case of Japan, the German model was known in academic circles from the 1880s but the first compulsory social insurance law was not adopted until 1922.Footnote ⁴⁹ Some Japanese officials gained practical experience with social insurance: for example, Goto Shinpei (1857–1929) was attracted to such schemes when studying in Germany in 1889 as an official of the Home Ministry of Japan. In his long administrative career, Goto devoted himself to the promotion of social insurance in Japan and its overseas enterprises. In 1907, he set up compulsory mutual aid associations for the workers on the Southern Manchurian Railway.Footnote ⁵⁰ Meanwhile, in 1913, Seki Hajime, a scholar-official and mayor of Osaka from 1914 to 1934, proposed immediate implementation of accident insurance and unemployment relief, but thought the timing for other compulsory insurance was not mature because the government could not then afford the financial requirements.Footnote ⁵¹

In Britain, between 1883 and 1911, according to E. P. Hennock, German-style social insurance was regarded as unsuitable for old-age and sickness insurance. Thus, in 1908, Britain passed an Old Age Pension Act ‘based on a deliberate rejection of that concept’.Footnote ⁵² But only three years later, Britain changed its attitude toward German social insurance when facing a revenue impasse. The German system – financed by contributions from employees and employers – was suddenly regarded as cheaper than means-tested systems financed by general taxation. Thus the German model provided a way out of that impasse. Eventually, Britain adopted its National Insurance Act in 1911, including two independently administrated, German-style social insurance systems: Part I set up a health insurance, and Part II established an unemployment insurance, which was the first compulsory unemployment insurance in the world.

In contrast, the diffusion of the German Social Democrats’ original ideas of social insurance took place through emulation. The German Social Democratic Party, from its inception in 1869, was affiliated with the First International (1869–76) and joined the Second International at its foundation in 1889. The Russian Social-Democratic Labour Party, founded in 1898, affiliated with the Second International. In fact, as early as the 1890s, the German capitalist model attracted attention in Russia, where industry was flourishing, while the Russian socialists emulated the socialist model. In 1903, at its second congress, the Russian Social-Democratic Labour Party issued labour programmes including social insurance demands.Footnote ⁵³ In the middle of the Russo-Japanese War of 1904–05 and the workers’ revolution of 1905, the Russian government made social insurance an urgent issue and set up a commission to revise labour legislation. Subsequently, the Russian government submitted a German-style social insurance bill to the Duma, which became the Health and Accident Act of 1912.Footnote ⁵⁴ While that bill was being debated, Lenin criticized it at the Russian Social Democrats’ sixth All-Russian Conference held in Prague in January 1912, and proposed a set of principles based closely on those of the German Social Democrats of the 1880s.Footnote ⁵⁵ This became the famous Lenin principle of social insurance, frequently mentioned by other communist parties and states.

The Lenin principle was restated in a 1914 Social Democratic bill submitted to the Duma. This bill required universal coverage for all wage and salary earners and landless peasants; capitalists and landlords were excluded. It required comprehensive benefits to be financed through taxes on property, inheritance, and income, and proposed self-administration by the insured workers under a centralized system. The highest administrative organ was to be the All-Russian Workers’ Insurance Conference, which would issue regulations, while local, regional, and all-Russian social insurance councils were to be established for everyday administration. The bill was not adopted. The Lenin principle was further reaffirmed five days after the Bolsheviks came to power in 1917 in a series of social insurance communiqués and in a law on social insurance in 1918. Although the law basically remained on paper, it was the beginning of socialist social insurance legislation.

In the years up to 1919, the spread of both German capitalist social insurance and the socialist alternative remained principally within Europe, but with extensions to South America and the South Pacific. However, people outside Europe were very well aware of the movement and referred to it as worldwide, owing to the efforts of the international associations on social insurance and labour legislation, and those of the socialist networks. Charles Richmond Henderson, in a 1909 article, argued that, a few years earlier, ‘the American mind was empty of knowledge of a world movement’, and ‘industrial insurance in this country has been monopolized by private companies’.Footnote ⁵⁶ There were many reasons for countries not to adopt the German model but the principal one was fear of the financial burden. However, the contemporary opponents also stressed individual responsibility as an excuse for non-adoption, claiming that ‘it weakens the sense of individual responsibility; that it intensifies the cupidity of the masses; that it demoralizes the working-people’.Footnote ⁵⁷ It took until the ILO was established in 1919 for a truly global social insurance project to arise through its promotional campaign.

Global social insurance in formation, and wartime transition: 1919–45

In the three decades after the First World War, both models of social insurance became far more global, covering all the four major social risks, although through different channels. During the Second World War, social insurance programmes not only survived but also expanded among the main combatants. In addition, a British-led discourse emerged, calling for universal social security (including comprehensive medical care) together with full employment as major principles for post-war reconstruction.

In 1919, the ILO was created by the Treaty of Versailles at the Paris Peace Conference to promote post-war social justice and world security. The ILO had two concrete objectives: to collect and distribute information from its member states concerning conditions of labour and industry, social legislation, and labour law administration; and to create international labour legislation in the form of conventions and recommendations, providing models for its member states to follow.Footnote ⁵⁸ Thus, spreading social insurance was one of the ILO’s major tasks. In 1927, the International Social Security Association (ISSA) was established, in affiliation with the ILO, to focus solely on spreading social insurance. As Cedric Guinand put it, ‘without these organizations the development of social insurance institutions at the international level would have been considerably less marked’.Footnote ⁵⁹

As the Soviet Union took shape, laws on sickness insurance, unemployment insurance, and disability and survivorship pension were adopted in 1922 and incorporated into the Labour Code. All these laws inherited the two outstanding features set up in the Lenin principle of 1912: the cost was to be shouldered solely by employers and the programme was to be administered by workers themselves. Otherwise, the laws became much more practical. They covered only wage earners, excluding self-employed peasants, artisans, and professionals. Although the risks covered were still comprehensive, the benefit levels became lower than previous wages, except in the case of temporary disability. Nevertheless, eligibility for benefits was less restricted than in other systems: for instance, for sickness and industrial accidents, there were no work requirements during the 1920s, and no waiting period nor concept of involuntary unemployment. The Commissariat of Labour became responsible for administration in 1924, but trade unions had a supervisory role in administration and had a joint role in selecting the executive personnel.Footnote ⁶⁰

During the 1930s, Soviet social insurance evolved into a system that can be labelled as the Soviet socialist model of social insurance: this was the system to be emulated by other socialist countries in the post-war era. Unemployment insurance was abolished in 1930; occupational funds were established in 1931 to grant special treatment to workers hired in the key industries; the Commissariat of Labour was abolished; and the trade-union central assumed administrative responsibility in 1933, though by this time it was entirely under the control of the Communist Party. Under the Constitution of 1936, a ‘state budget’ was set up to finance medical care, workers’ housing, and old-age, survivorship, and permanent disability pensions. In 1937, the coverage of old-age pensions was extended to white-collar employees.Footnote ⁶¹

In the interwar era, meanwhile, the Lenin principle rather than the Soviet model was widely emulated, through the global communist movement. In China, the Communist Party introduced social insurance systems that were close copies of the Lenin principle in their rural soviets in the 1930s. These systems were drafted under the direct influence of Soviet experts, although the Chinese rural soviets barely had any industry. When they were implemented, it was the workers in the communist-controlled enterprises who received the benefits. Thus, the contemporary Nationalists criticized such systems as privileging the new communist elites.Footnote ⁶²

Meanwhile, from 1919 the German model spread widely to southern and eastern Europe, the Americas, South Africa, and Asia. For instance, Italy and Spain adopted old-age insurance in 1919, and Hungary in 1928. Cuba adopted sickness insurance in 1921, Yugoslavia in 1922, Brazil in 1923, and Chile in 1924. The USA adopted old-age and unemployment insurance in 1935. British Commonwealth and Scandinavian countries continued to adopt means-tested old-age systems, as with Canada in 1927, South Africa in 1928, Norway in 1936, and Finland in 1937.

Poland adopted its first sickness insurance in 1920, providing free medical care and cash benefits for maternity and sickness. Then it adopted accident and unemployment insurance in 1924, and comprehensive social insurance legislation in 1933, which established uniform employees’ insurance, providing sickness, old-age, disability, survivors’, and unemployment benefits. In South America, Uruguay adopted an old-age system in 1914, but only for public utility and mass-transportation workers. It greatly expanded its old-age system in the 1920s, to cover the majority of blue-collar workers and employees in industry, commerce, and various trades and services. In 1934, during the Great Depression, the country adopted unemployment insurance funds for workers in export industries.

Japan adopted its first social insurance against sickness in 1922, though it was not implemented until 1927, after a national crisis caused by the 1923 Great Tokyo Earthquake, at which point it provided sickness, injury, and childbirth benefits to workers.Footnote ⁶³ In the US, the German model of social insurance had long been known but was rejected by many national interests. It was only in the context of the Great Depression that the Roosevelt administration was able to pass the Social Security Act of 1935, based on the German model. Rodgers used the term ‘the intellectual economy of catastrophe’ to describe this origin of the US welfare state.Footnote ⁶⁴

India, while still a colony of Britain, was a good example of the ILO’s important role in the global diffusion of social insurance.Footnote ⁶⁵ In 1921, the ILO adopted its Workmen’s Compensation Convention: India faced increasing industrial accidents, and thus the Indian central government followed the ILO’s recommendation to introduce the Workmen’s Compensation Act in 1923 for Indian factory workers.Footnote ⁶⁶ This turned out to be the only social insurance law that India adopted before its independence in 1947. In the case of sickness insurance, as early as 1928 (after the ILO issued its two draft conventions and one recommendation on sickness insurance in 1927), the Indian government began to discuss the possibility of ratifying the conventions on sickness insurance, but adopted no laws. Instead, after one decade of preparation and discussion, the government concluded in May 1937 that there was ‘an absence of any real demand for health insurance on the part of the Provincial Governments and employers, and indeed of workers if they were to be asked to contribute’.Footnote ⁶⁷

With the outbreak of European war in 1939, ILO activities were interrupted until a Montreal office was opened in 1942. Nevertheless, the social insurance movement continued at the global level, especially in the Americas. For instance, many South American countries followed the German model: Bolivia adopted old-age insurance in 1939; Cuba adopted old-age and unemployment insurance in 1940; Brazil and Venezuela adopted sickness and old-age insurance in 1940; Paraguay approved old-age insurance in 1941; Peru set up sickness and old-age insurance in 1941; and Panama adopted sickness and old-age insurance in 1942.Footnote ⁶⁸

The wartime German regime reorganized the social insurance systems in its occupied territories. In Poland and other German-occupied territories, the previous social insurance systems were essentially replaced by German laws for the purpose of serving German interests at the expense of those living under occupation. Thus, Polish social insurance laws were abolished by decree on 7 March 1940. This decree cancelled all claims and benefits from Polish social insurance, and provided that benefits and services from social insurance were to be replaced by much more limited measures of assistance. This decree was originally applied to the entire population, but soon it only applied to people identified as Poles and Jews, while providing a preferential treatment to German nationals. The funds were still raised by contributions from employers and employees, who were expected to support not only themselves but also the insurance institutes set up by Germany.Footnote ⁶⁹

Japan expanded its social insurance systems under the newly established Welfare Ministry in 1938. Only four months later, this ministry drafted a new health insurance law, then passed by the Diet, which launched a voluntary system of people’s health insurance cooperatives to cover the residents or employees who were not covered by the existing employee health insurance law of 1922. During the Second World War, this system expanded rapidly, with the goal of universal coverage. By 1944, about 95% of Japan’s cities, towns, and villages had health cooperatives, insuring 41.1 million people, while the health insurance system established in 1922 covered about 9 million people. In 1942, the Welfare Ministry sponsored a compulsory pension insurance system for workers, which was extended to cover employees in businesses of five or more in 1944. This pension system covered 3.46 million persons in 1942 and 8.44 million at the end of 1944.Footnote ⁷⁰

Meanwhile, in the early 1940s the Atlantic powers saw the emergence of a new global discourse of universal social security, inspired by wartime developments. For instance, the Atlantic Charter of 1941 referred to ‘the fullest collaboration between all nations in the economic field with the object of securing for all, improved labor standards, economic advancement and social security’. The Montreal office of the ILO published a 1942 book entitled Approaches to social security: an international survey, with the purpose of preparing for the establishment of complete social security programmes as part of post-war reconstruction.Footnote ⁷¹ This survey promoted the concept of social security, defined as consisting of social insurance and social assistance, which were referred to as means-tested systems in this study.

In November 1942, Britain published the famous Beveridge Plan, entitled Social insurance and allied services, which was regarded as a milestone in the development of the welfare state. As Rodgers commented, ‘Of all the welfare state programs and platforms that were to follow, none was to match its influence or the electricity of its reception.’Footnote ⁷² As a result, the US passed the Wagner-Murray-Dingell Bill in 1943 and Canada published its Marsh Plan in 1943. In 1944, India, supported by two ILO experts, considered issuing an Indian Beveridge Plan built on the Adarkar draft of health insurance.Footnote ⁷³ Even the Nationalist government in China issued its outline of four major social policies in 1945, two of which addressed national health and post-war social security programmes, inspired by the Beveridge universal plan.Footnote ⁷⁴

In 1944, the ILO convened its Philadelphia Conference, which restated the aims, purposes, and fundamental principles of the organization. The conference further recognized the obligation of the ILO to advance world programmes that would achieve full employment and social protection. In terms of social security, the Declaration called for ‘the extension of social security measures to provide a basic income to all in need of such protection and comprehensive medical care’.Footnote ⁷⁵ Thus, the ILO affirmed the global norms of full employment and universal social security (including comprehensive medical care) that emerged during the war, and the term ‘social security’ replaced ‘social insurance’ in most of the international documents; meanwhile, the terms ‘sickness insurance’ and ‘health insurance’, which we have used interchangeably, began to include a universal public healthcare system to be developed after the Second World War.

Great expansion: post-war diversity, 1945–1981

With the end of the war, hopes continued to rise for implementing the new global norms. Soon, however, the world divided into two dominant camps: the industrial capitalist world, led by the US, and the socialist world, led by the Soviet Union. Between these centres of power lay numerous developing countries, some long-since independent and others just escaping colonial rule. The industrial capitalist countries extended their social insurance scope to cover their entire population and thus build up the welfare state.Footnote ⁷⁶ The socialist countries in eastern Europe emulated the Soviet socialist model, replacing their previous German-style social insurance, and the developing countries emulated either the Soviet- or German-style social insurance programmes. The ILO continued to play an important role in spreading the new global norms among the non-socialist countries.

Industrial capitalist countries expanded German-style systems to cover the four major social risks immediately after the war: Switzerland adopted its first old-age insurance in 1946 and Japan its first unemployment insurance in 1947. Thus, by the 1950s, most industrial capitalist countries had comprehensive systems for all of the four social risks. After the 1950s, countries such as Germany, Italy, France, the US, Austria, and Japan (among many others) expanded their comprehensive social insurance systems to cover previously uninsured citizens, especially the self-employed. Meanwhile, a tendency emerged in the British Commonwealth and Nordic countries (including Sweden, Britain, and Canada) to replace their means-tested old-age systems with the Beveridge flat-rate universal system. Soon, however, these systems proved too costly to continue. Thus, Canada, Norway, Finland, and Sweden all introduced German-style old-age social insurance systems to back up their universal systems, and in 1978 the UK introduced a German-style old-age scheme, supplementing its existing universal old-age system.Footnote ⁷⁷ In contrast, some countries with German-style old-age systems adopted supplementary flat-rate old-age systems, such as the Netherlands in 1956, Italy in 1965, and even Germany in 1972.Footnote ⁷⁸ In the case of Japan, a universal health insurance system inspired by the Beveridge model was initially proposed by American officials while the country was under American occupation, but was later dropped because of domestic American opposition to such a system in the US.Footnote ⁷⁹ Uruguay continued to expand its old-age system to cover the entire working population by incorporating rural workers and professionals into the system. The country also introduced its first maternity insurance in 1958 and health insurance in the 1960s.

In the socialist world, the Soviet model was extended to the new socialist regimes of eastern Europe, replacing the German-imposed wartime systems. For instance, Poland concentrated on the needs of war victims, repatriates, migrants, and the most destitute in the aftermath of the war. However, in the late 1950s, the same country began to adopt the Soviet-style social insurance systems in which the state remained the major provider of various benefits paid through the state budget.Footnote ⁸⁰ In the Soviet Union, Khrushchev denounced Stalin’s policies in 1956, and carried out reforms in which a new pension law was promulgated: it increased pension benefits and loosened eligibility for benefits but retained the fundamental structure of the Soviet model. By this time, the Soviet Union had established a near-universal social security system for its people except for collective farmers, who relied on separate mutual-aid funds that pooled money from themselves and provided limited benefits for sickness and old age.

By the 1970s, the Soviet Union and its European allies had extended their socialist social insurance systems to their entire population, matching the competing systems in the capitalist world. In the Soviet Union, a state pension system was established in 1964 for collective farmers, providing them with old-age, disability, and survivors’ pensions, and allowances during pregnancy and confinement. These benefits were financed by the national social security fund, supported by the state budget and a contribution of 4% of income from collective farms. This finally made the Soviet social security system universal. In Poland, during the 1970s, independent farmers and their families and other self-employed groups were covered by its inclusive social insurance system. As a result, socialist social insurance programmes also became universal here.Footnote ⁸¹

The poorer socialist countries, such as China and Cuba, introduced Soviet-style social insurance programmes, and thus shared many attributes: workers were not required to contribute, and the state-owned enterprises financed the system. In addition, these systems had the same eligibility for pension benefits, such as the same retirement age of sixty for men and fifty-five for women. Coverage was universal in Cuba, but in China it was heavily weighted towards the urban population.Footnote ⁸² Further, in China’s case, its Soviet-style system did not survive the country’s political and economic constraints and broke down during the Cultural Revolution.Footnote ⁸³

The former French colonies adopted German-style social insurance under the influence of the ILO. As John Dixon commented, ‘Encouraged by the proselytizing technical experts from international agencies, long nurtured in the nuances of social insurance … developing countries in the 1950s and 1960s began replicating the European social insurance model.’Footnote ⁸⁴ In 1952, a Labour Code was established for French overseas territories, allowing African wage earners in the private sector to gain social insurance benefits. However, it was not implemented immediately; it was only from 1955 and 1956 that African wage earners could actually be entitled to the family allowance benefits, and in 1957 to medical care and work compensation benefits.Footnote ⁸⁵ As E. J. Ejuba pointed out, ‘As the territories gained independence, they found themselves administering social security schemes that resembled those of the French social security funds.’Footnote ⁸⁶ By 1960, the colonies under Belgian administration (Burundi, Rwanda, and the Congo) provided old-age, disability, and survivors’ pension schemes. During the 1970s, however, all French-speaking countries in Africa (most of which became independent in 1960) provided old-age social insurance pension systems.Footnote ⁸⁷

Upon their independence, the former British colonies in Asia, Africa, and the Caribbean adopted central provident funds for their old-age protection as a transition to social insurance. India introduced the initial CPF in 1925 for its civil servants and railway workers. After its independence, in 1948, it established a CPF for its coal miners and in 1952 for its industrial workers.Footnote ⁸⁸ Provident funds then spread to former British territories in Southeast Asia, the Pacific, Africa, and the Caribbean. Nigeria was the first in Africa to emulate this fund. As Martin Tracy pointed out, the Nigerian decision to adopt a CPF was largely made by civil servants who were aware that provident fund schemes had been established in India and other former British colonies.Footnote ⁸⁹

Social insurance in the age of privatization: 1981 to the present

With the 1980s, a new set of challenges arose for social insurance. At the national level, the maturation of long-established national social insurance systems brought fiscal challenges as manifested by increasing contribution rates and rising expenditure; the growth of aging populations brought special problems for old-age insurance as shown by the decreasing ratio of workers to pensioners. At the global level, the slowing of global economic growth limited tax revenues. Neo-liberal ideology grew sharply in influence, with its leading centres at US universities, notably the University of Chicago.Footnote ⁹⁰ Since the 1980s, its supporters have campaigned for privatization of social insurance, especially old-age pension systems. These national and global factors interacted with one another and the global neo-liberal forces appeared to be the driving force for the fundamental change towards privatization.

The privatization of the Chilean pension system in 1981clearly opened this new period in the history of social insurance.Footnote ⁹¹ The major policy makers, including the Labour Minister Jose Pinera (a Harvard graduate), were the so-called Chilean Chicago boys, who were trained by neo-liberal economists such as Milton Friedman at leading US universities, such as Chicago, Columbia, and Harvard, with the support of the US government.Footnote ⁹² Beginning in 1962, Friedman had argued against the ‘social security programs’ that were in the middle of their post-war expansion around the world.Footnote ⁹³ These Chicago boys returned to Chile to implement such ideas in economic, social, and political reforms. They first proposed to privatize Chile’s pension system in 1973, shortly after the military coup. Then, under the Pinochet regime, in 1980–81 the Chicago boys proposed replacing the public pension system with a fully funded (individual-accounts) system, with contributions solely by individuals and administered by competing private companies. Thus the reform began. By the end of 1999, the new private system had covered almost the entire insured population, and only 4% of the insured remained in the public system.Footnote ⁹⁴

Meanwhile, the communist systems in eastern Europe and the Soviet Union collapsed from 1989 to 1992. With regime change came major changes in social insurance. In these countries, their Soviet-style social insurance systems were replaced by the German model in the early 1990s: in the Russian Federation, four German-style social security funds were established: the Pension Fund (1991), the Employment Fund (1991), the Medical Insurance Fund (1991), and the Social Insurance Fund (1992).Footnote ⁹⁵ In Poland, a unified German-style social insurance system for all employees was gradually set up from 1991 to 1995.Footnote ⁹⁶ Pressures then grew for privatization.

The growing transnational network of neo-liberal economists endeavoured to make the Chilean reform a global model, as Mitchell A. Orenstein makes clear, by persuading many international organizations such as the World Bank to promote pension privatization around the world.Footnote ⁹⁷ Ultimately, in that process, privatization encountered great opposition both from the ILO and from national governments. Thus, in 1994, the World Bank moderated its strategy of full privatization and recommended partial privatization through setting up a ‘three-pillar’ pension system, publicized in the widely-disseminated Averting the old age crisis: polices to protect the old and promote growth.Footnote ⁹⁸ The World Bank’s three-pillar pension system consisted of ‘a mandatory, publicly managed, unfunded pillar and a mandatory but privately managed funded pillar, as well as supplemental, voluntary, privately funded schemes’.Footnote ⁹⁹ The first pillar is actually a public social insurance system, the second pillar is compulsory individual savings accounts that are managed privately, and the third pillar consists of voluntary and privately managed individual accounts.

The World Bank, because of its strong resource pool, appeared to be replacing the ILO as the major player in worldwide pension reforms in the mid 1990s. Thus, many Latin American countries followed the three-pillar model recommended by the Bank. So did many former socialist countries in eastern Europe, and discussions of privatization diffused around the world, including to some African countries with central provident funds.Footnote ¹⁰⁰ In reality, few states fully accepted the global models provided by either the ILO or the World Bank. As Stanford Ross put it, ‘few countries have systems fully in accord with World Bank, ILO, or IMF [International Monetary Fund] prescriptions’.Footnote ¹⁰¹ His comments confirm the determining role of national forces in the pension process.

In the case of Uruguay, its public pension system was the oldest in Latin America, and the first to mature and encounter financial difficulties. Attempts to reform the public pension system were made throughout the 1980s and early 1990s, with the support of the World Bank and the International Monetary Fund, but none of them were successful, owing to internal opposition from various interest groups. It was only in 1995, when the newly elected president was able to create a pension commission including all political parties with parliamentary representation, that a mixed system following the World Bank’s three-pillar model could be proposed and eventually passed in September 1995; it came into force in March 1996.Footnote ¹⁰² The former socialist countries in eastern Europe then began to adopt such mixed systems. Poland set up a special office for social security reform in 1996, chaired by a World Bank appointed official, and adopted a mixed system in 1998.Footnote ¹⁰³ The Russian Federation espoused a mixed system in 2002.Footnote ¹⁰⁴

From the late 1990s, the ILO and the World Bank seemed to begin a search for consensus, as expressed at the 1998 Stockholm conference that recognized ‘the need to balance social goals and macroeconomic requirements when designing and implementing reforms’.Footnote ¹⁰⁵ As a result, in the early 2000s the ILO began to recommend its own ‘multi-pillar’ pension systems, consisting of four parts: a means-tested social assistance system financed by general revenue, a public social insurance system, compulsory individual savings accounts (which could possibly be managed privately), and voluntary individual savings accounts (also possibly managed privately).Footnote ¹⁰⁶ The ILO’s multi-pillar system is fundamentally different from that of the World Bank: the World Bank stresses privately managed individual savings accounts while the ILO emphasizes public systems, including means-tested assistance and social insurance systems.

The instance of China further demonstrates the limitation of any of the global models. Socialist China in the mid 1980s began to reform its Soviet-style social insurance systems and introduced German social insurance in the late 1980s. From 1993 to 1995, however, China began to experiment with a new system labelled as ‘social pooling combined with individual accounts’. And in 1997, a national, unified, old-age social insurance system was officially promulgated for urban workers. The Chinese government has claimed it as China’s innovation, but, in practice, it can be best described as a creative mixture of all available models, including the ILO’s social insurance, Singapore’s CPF, and the three-pillar model of the World Bank.Footnote ¹⁰⁷

The German model has survived to the present day partially through the persistent efforts of the ILO. Developed countries were able to cope with the financial difficulties until the end of 1990s by increasing contribution rates, decreasing the benefit levels, increasing retirement age, and tightening the eligibility criteria. Then, in the 2000s, with soaring financial burdens, some of the developed countries had to reform their old-age systems by introducing new supplementary measures. For instance, Sweden introduced mandatory individual accounts in 1999; Germany adopted voluntary individual accounts for its private sector employees in 2002;Footnote ¹⁰⁸ Britain adopted mandatory individual accounts for its pension systems as early as the 1980s. All these new individual accounts were supplementary only, existing alongside the public old-age systems. Thus the leading social insurance programmes in western Europe, North America, and elsewhere, stayed public in nature rather than private.

Conclusion

This is a global historical study of large-scale social institutions. It has described the spread and transformation of social insurance, and revealed that the patterns of change were more complex than the local narratives reported in national historical studies or the social science stories of global diffusion. Our framework of ‘interactive diffusion’ of global models emphasizes that the global social insurance movement was the diffusion of two principal models of social insurance – the German capitalist model and Soviet socialist model. We have given particular attention to patterns of adoption that we label as policy learning and policy emulation and to the balance of global and national forces within the social insurance movement.

Bismarck’s insight – that national programmes of social insurance would elicit support for the national government – has certainly been validated across the experience of the twentieth century. Once the initial German system was implemented, experiments with it began not only in European countries but in such distant nations as Japan and Uruguay. Despite the numerous variants attempted, the German system of national and compulsory insurance – with contributions by employees and employers – consistently emerged as the leading system.

Diffusion of social insurance programmes accelerated after the First World War. The ILO, founded to conciliate conflicts between workers and employers, adopted general conventions and conducted consulting trips to all independent countries and to some colonies, working especially with governments but also with trade-union centrals. The rise of the Soviet Union nurtured the socialist model of social insurance, and the socialist model was discussed in many parts of the world through the networks of communist parties and trade unions.

As early as the 1930s, global patterns of development in social insurance were beginning to clarify: programmes expanded from accident insurance to sickness and old-age insurance, and then to unemployment insurance; coverage extended from voluntary to compulsory and from selected industrial workers to wider groups of employees and even to non-employed persons such as family members and pensioners. Remarkably, even during the years of the Second World War, maintenance and expansion of social insurance programmes remained a priority, as warring governments worked to maintain the support of their home populations.

With decolonization and the establishment of numerous socialist governments from the 1940s to the 1970s, many new systems of national social insurance were established, with much national variation within the overall global pattern. The option of central provident funds was widely followed in ex-British territories for old-age insurance, but most of these gradually changed to compulsory national social insurance systems. Poland was unusual in the number of changes in its social insurance system, especially because of regime change (from the initial Russian system to that of independent Poland; the system under German occupation; the immediate post-war system; the socialist system; and then changes within the post-socialist system). Uruguay and Japan, in contrast, have had remarkable continuity in their systems.

From the 1950s to the 1970s one can see fuller patterns of expansion within both capitalist and socialist systems, with extended coverage of new risks and wider populations (especially the self-employed in capitalist countries and agricultural workers in socialist countries). In both capitalist and socialist systems, comprehensive healthcare systems (including both health insurance cash benefits and healthcare services) arose in many countries (we have left the important issue of healthcare services outside our analysis).

The 1980s brought a new trend: privatization of social-insurance-based pension systems. This option, which was different from the private and voluntary social insurance that had earlier prevailed and then fell into decline from the mid nineteenth century, emerged with the expanding neo-liberal refutation of the value of public services and the rapid growth of financial institutions. Within a decade of the rise of privatization, the Soviet model of social insurance collapsed at the same time as most of the regimes that had relied on it.

For many countries, German-style social insurance has remained dominant. The lesson of the past century seems to be that social insurance will continue to be a contested and fluctuating set of institutions, but that it will continue to develop along its past trajectory. Compulsory social insurance systems have regularly shown themselves to be fiscally and financially feasible (despite repeated projections of impending fiscal deficits), and have commonly been governed responsibly in developed countries. Perhaps the dominant factor in their continuing importance is the consistent public demand for dependable systems of social insurance.