any of the measures established by legislation to maintain individual or family income or to provide income when some or all sources of income are disrupted or terminated or when exceptionally heavy expenditures have to be incurred (e.g., in bringing up children or paying for health care). Thus social security may provide cash benefits to persons faced with sickness and disability, unemployment, crop failure, loss of the marital partner, maternity, responsibility for the care of young children, or retirement from work. Social security benefits may be provided in cash or kind for medical need, rehabilitation, domestic help during illness at home, legal aid, or funeral expenses. Social security may be provided by court order (e.g., to compensate accident victims), by employers (sometimes using insurance companies), by central or local government departments, or by semipublic or autonomous agencies.

The International Labour Organisation (International Labour Organization) (ILO) uses three criteria to define a social security system. First, the objective of the system must be to grant curative or preventive medical care, to maintain income in case of involuntary loss of earnings or of an important part of earnings, or to grant a supplementary income to persons having family responsibilities. Second, the system must have been set up by legislation that attributes specified individual rights to, or that imposes specified obligations on, a public, semipublic, or autonomous body. And third, the system should be administered by a public, semipublic, or autonomous body.

In its statistics the ILO includes provisions according to which the liability for the compensation of employment injuries is imposed directly on the employer, although such schemes do not strictly meet the third criterion above. For this reason employer liability is included here.

An alternative but wider term for social security in the countries that are members of the European Union is social protection, which includes voluntary schemes not set up under legislation. In some countries the term social security is used in a narrower sense. For example, in the United Kingdom only statutory benefits in cash are regarded as social security. The term social services is used to cover social security; health, education, and housing services; and provisions for social work and social welfare. In the United States the term social security is restricted to the federal social insurance system (OASDI) as distinct from state benefits and “welfare,” which in Europe would be called social assistance. In some countries (for example, Denmark and the United Kingdom) the reduction of poverty historically has been a central aim of social security policy, and the concept of maintaining income has been grafted on at a later stage. In other countries, such as France, measures to deal with poverty have been seen as quite separate from the income maintenance aims of social security.

A report published in 1984, prepared by 10 international experts appointed by the director of the ILO, set out the ultimate aims of social security.

Approximately 140 countries have some type of social security scheme. Nearly all of these countries have schemes covering work-related injury and old-age and survivors' pensions. Well over half have provisions for sickness, and nearly half have provisions for family allowances. The least commonly provided schemes are for unemployment, though at least 40 countries have them.

Because general social security schemes based on compulsory insurance did not come into being until the last two decades of the 19th century, it has often been argued that social security in its modern form has been a response to industrialization, which caused large numbers of people to become dependent for their security solely on earnings from employment. Indeed many families became dependent on one male earner and thus on his capacity to find work, to undertake it, and to remain in it. Moreover, industrialization led to the migration of people toward centres of work, thus separating them from the support given by the wider family. In addition, the development of compulsory education prolonged the period during which children were dependent on their parents; later the system of enforced retirement created dependency at the other end of life. This situation is contrasted with an often idealized image of the extended rural family with access to land, on which both husband and wife worked, children started work early, and old people continued to work until they became too frail or disabled to do so. On the basis of this oversimplification, some theorists have proposed that social security developed out of a need peculiar to industrial societies and that there is less need or no need for social security programs in the rural areas of developing countries today.

It is true that support from the extended family, often enforced by local custom and religious beliefs, contributes to the survival of peasant societies. But by no means do all the rural populations of developing countries have access to land, and many people work for wages in agricultural estates and mines. Moreover, peasant farmers are subject to formidable risks of crop failure, quite apart from the risks associated with the shorter average life span that characterizes developing countries. Although there is a need for social security in rural societies, the importance of specific risks may vary from region to region. Moreover, the irregular incomes in cash and kind emanating from agriculture do not lend themselves to the payment of regular social insurance contributions. Thus, what may be lacking in rural societies is the economic and administrative base for providing such security. Furthermore, provision for sickness and old age is not generally seen as the highest priority by peasant farmers overwhelmed by problems of weather and debt.

While the advent of industrialization has undoubtedly added to the need for social security by breaking up the extended family and leading to urban poverty, it is by no means the sole reason why the system evolved. Two of the first three countries to make provision for old-age pensions were primarily agricultural societies—Denmark in 1891 and New Zealand in 1898. The Danish scheme was clearly an attempt to alleviate rural rather than urban poverty. And it is notable that the first province in Canada to develop compulsory health insurance (1962) was Saskatchewan, which was overwhelmingly agricultural. These cases indicate that statutory social security may evolve for a variety of reasons. Moreover, it depends to a considerable degree on the economic level attained by the groups that might be covered and the administrative capacity of the country to operate such a scheme. It is certainly the case that, as countries become wealthier, there is greater willingness to defer consumption by paying insurance contributions or taxes.

In many societies charity has been the traditional way in which provision was made for the poor. Charitable giving has been encouraged by many different religions, and in many parts of the world religious agencies have long collected charitable donations and distributed help to those in need.

The imposition of obligations on communities to pay taxes in order to provide for the poor can be traced back for hundreds of years in a number of different societies. For example, part of the function of the Christian tithe or the Islāmic zakat was to provide for the poor. Town poor laws were passed in Germany from 1520 onward, and a law passed in 1530 clearly placed on towns and communities the obligation of sustaining the poor. In 1794 the Prussian states assumed the responsibility of providing food and lodgings for those citizens who were unable to support and fend for themselves. From the 16th century it became recognized in England that there were people who could not find work, and legislation was passed to provide work for the poor and houses of correction for rogues and idlers. From 1598 a clear obligation was placed on parishes to levy local taxes and appoint overseers of the poor in order to give relief to those who could not work and to provide work for those who could. This formed was the essence of the Elizabethan Poor Laws (Poor Law), an early provision of social assistance.

The Elizabethan Poor Laws were poorly enforced in the 17th century but widely used and liberalized by the end of the 18th century. A new Poor Law enacted in 1834, and reflecting a harsh moral view of poverty, required the poor persons to be admitted to the workhouse so as to receive relief only in kind, with occasional exceptions, but this again was by no means uniformly enforced, though it added greatly to the unpopularity of the Poor Laws. Some U.S. states copied the Elizabethan Poor Laws but exempted recent immigrants. The English Poor Laws were also introduced in Jamaica in 1682 for destitute European immigrants and much later in Mauritius (1902) and Trinidad (1931). In Latin America the Spanish colonists, instead of establishing a public relief agency, gave grants to charities to provide “hospitals” for the poor (beneficencias), and the Portuguese promoted lay brotherhoods such as the Misericórdia.

The first general social insurance scheme was introduced in Germany in 1883. The scheme drew upon three types of precedent. The first was the ancient system of guild collection boxes—funds to which each member of a particular trade was required to contribute at regular intervals; such funds were originally used for hospital and funeral expenses and for food and lodging for aged and disabled members. By the middle of the 14th century these arrangements were covered by statutes and regulations. Relief funds were later established by associations of miners. The second precedent was a Prussian ordinance of 1810 that placed on masters a duty to ensure that their servants were given medical attention in case of illness. From 1849 communities could make bylaws requiring both employers and employees to contribute to relief funds, and a law of 1854 introduced compulsory health and accident insurance for miners. The third precedent was the employer's legal liability to pay damages for accidents caused by negligence. As a result of this liability, which was widened in 1871, many employers took out private insurance. The system did not work well because the burden of proof lay with the worker, who normally had to incur high legal costs and delay before he could hope to obtain lump-sum compensation.

Chancellor Otto von Bismarck's (Bismarck, Otto von) 1883 sickness insurance law provided to employees in defined types of industry both medical care and cash benefits during a period of sickness, to be paid for out of contributions from both employees and employers. This was followed by a law of 1884 making accident insurance compulsory. The schemes were operated by numerous funds controlled by the insured and their employers. Finally a law establishing a pension for all workers in trade, industry, and agriculture from the age of 70 was passed in 1889. This was directly administered by the Imperial Insurance Office. Austria followed part of the German example in 1888, Italy in 1893, and both Sweden and The Netherlands in 1901.

Bismarck's political aim in introducing social insurance had been to address the legitimate grievances of workers so as to check the growth of socialism and avert revolution. A proportion of previous earnings were to be paid in cases of sickness, injury, widowhood, and old age. Employers and employees were to work together in implementing the scheme. In Austria part of the driving force was the Christian Socialists' aim of improving the worker's position. Although Britain had been the first country to industrialize, the developments in Germany and Austria originally attracted little British interest because of an aversion to state intervention, an apparently lesser likelihood of revolution, and the slower development of British socialism. In Britain self-help through friendly societies (friendly society) and savings banks was seen as the solution. The friendly societies were run by skilled workers with no employer participation and provided flat-rate cash benefits for sickness as well as treatment by the society's doctor, who was normally paid a flat rate per member insured—a so-called capitation payment. By 1870 membership had grown to 1,250,000 and by the early 20th century to 7,000,000. Apart from the regulation of friendly societies, the only social security legislation passed in the United Kingdom during the 19th century was to widen the liability of employers to compensate workers for personal injury arising out of work. By a law of 1897, compensation could be obtained whether or not the employer had been negligent.

Further action arose in the United Kingdom out of social concern about poverty, which was systematically investigated both in London and in York. In 1899 the government carried out an inquiry into the incomes of 12,000 elderly people. The influential precedents for action were those of New Zealand and Denmark, which had made provision for old age without establishing social insurance schemes, in contrast with Germany, where the scheme was based on insurance. In 1908 in Britain, pensions at age 70 were introduced in a noncontributory, income-tested basis, partly because such a scheme could bring immediate relief to the aged poor, as opposed to a contributory scheme, which could only pay pensions to those who had paid contributions. The social insurance approach was, however, applied to sickness and also to unemployment in certain occupations three years later. This compulsory scheme, including the first state scheme of unemployment insurance, again reflected Britain's concern to address the main causes of poverty. Benefits and contributions for sickness and unemployment insurance were flat-rate, building on the precedents established by the friendly societies and ensuring the maximum impact on the living standards of low earners. From 1925 the social insurance approach began to be extended to provide for widowhood and old age.

Unemployment insurance was subsequently introduced in Austria and Belgium (1920), Switzerland (1924), Germany (1927), and Sweden (1940). In the case of health insurance, Denmark, Norway, and Sweden promoted voluntary health insurance before making such schemes compulsory, much later than in Britain or Germany. In France voluntary insurance had long been less developed, and mutual insurance societies had long been regarded by government with suspicion, and therefore suppressed. When they ultimately were allowed to expand, around the end of the 19th century, the bulk of their membership was middle class. During the second half of the 19th century larger employers established their own pension and welfare institutions. An employers' liability law was passed in 1898 for accidents at work irrespective of negligence, and in 1910 modest contributory pensions were introduced for industrial and agricultural workers. This law met with limited success, owing to opposition on the part of workers, noncompliance among employers, the loss of rights on change of job or bankruptcy of the employer, and the erosion of the value of pensions during inflation. Health insurance, though provided for in a law of 1920, did not come into effect until 1930, owing to the opposition of the medical profession.

A major innovation came in Belgium (1930) and France (1932) with the introduction of family allowances, although New Zealand had introduced a limited means-tested scheme in 1927. These derived from the ideas of social Christianity regarding “the just wage” and had originally been introduced by Christian employers on a private basis; special funds were later set up to equalize financial burdens among employers. Family allowances became relatively generous in France, partly because of concern to increase the birthrate after the heavy loss of men in World War I. (There is, however, no clear evidence that family allowances have any impact on birthrates.) France later introduced family allowances in many of its colonies during the 1950s.

During the interwar period social insurance schemes were introduced in more and more countries in Europe and Latin America (Latin America, history of). The most common model was that established in Germany—autonomous funds paying earnings-related benefits. The first group to benefit in Latin America was civil servants, followed by those working in railways and public utilities. There were separate schemes for hospital personnel in Argentina (1921), shipbuilders in Uruguay (1922), merchant seamen in Chile (1925), and dockworkers in Peru (1934). Thus the foundations were laid for the complex social security schemes in Latin-American countries that later reformers tried to amalgamate. The first comprehensive scheme for industrial workers was established in Chile in 1924. In African colonies many schemes of social security were originally introduced only for expatriate Europeans.

The Great Depression of the 1930s finally overcame opposition in the United States to federal intervention in social security. Earlier government activity had consisted of piecemeal initiatives at the local or state level. The Social Security Act of 1935 not only provided federal grants for state public assistance to the aged, blind, disabled, and dependent children but also established a federal old-age insurance scheme and federal financial backing for state unemployment insurance plans that met federal guidelines. Provision for survivors was added four years later and for disability later still. A quite different approach was taken in New Zealand, which introduced in 1938 the first universal non-means-tested pension from age 65, available only on a test of residence and financed in part from a special social security tax on income.

A major influence on world developments was the British government's report by Sir William (later Lord) Beveridge in 1942, which argued for the maintenance of full employment as a responsibility of government, family allowances for all children after the first, comprehensive health care for the whole population, and a unified national scheme of social insurance run by the state with the safety net of a unified national scheme of social assistance. The aim was to eliminate want or poverty. By 1948 the scheme had been introduced in the United Kingdom with some compromises and modifications. A drive, inspired by Pierre Laroque, to unify social insurance in France after World War II was less successful.

During the period of rapid world economic growth from 1945 to 1973 there was a further major expansion of social insurance to more countries, covering higher percentages of population and wider risks. The expansion was particularly notable in Latin America and in certain French colonies in Africa, where comprehensive social insurance schemes were introduced following the original schemes for family allowances. In the British colonies a different approach was taken: provident funds (see below) were widely developed for particular categories of workers. Discrimination on racial grounds was widely prohibited but still persisted in South Africa.

The major innovations in social insurance after World War II were the protection of pensions by linking them to the inflation rate; the development of dynamic pension formulas that indexed past pension contributions to the level of earnings at the time of retirement; the introduction of flexible retirement providing for part pension and part-time earnings in the last few years before full retirement; the movement toward equal rights for men and women; attempts to provide for all disabled people on the basis of the degree rather than the cause of disability (i.e., whether or not work-related); the growing recognition of extra needs arising from disability and of the needs of persons caring for the disabled; special provisions for one-parent families; the development of parental allowances in addition to family allowances; the integration of child tax allowances with family allowances; and the extension of the same health-care rights to all citizens.

Many countries that once held employers themselves legally responsible for compensating victims of work accidents and for paying for their medical care have now adopted state schemes of compulsory insurance. From the point of view of the worker the problems with the former system include the delays and costs of going to court and the possibility that the employer may be uninsured, unable to pay, or bankrupt by the time the case is heard. Moreover, a lump sum awarded by a court cannot be invested so as to provide a secure inflation-protected income for life. And when the employer is privately insured, the insurance company is in a position to offer the worker a small lump sum soon after the accident, knowing that the worker may well accept it rather than incur the delay, costs, and uncertainty of a court case to obtain the full value of the claim. From the national point of view such a system is wasteful because of the legal costs and the high administrative costs incurred by the insurer and passed along to the insured by way of higher premiums. The argument in favour of this approach is that insurers quote premiums for individual employers according to their experience of risk, which provides financial incentives for industrial safety. But insofar as such incentives are effective, premiums for a national program of accident insurance can also be risk-rated.

In some countries, when a statutory insurance scheme of occupational injury has been introduced, the right of the employee to sue the employer for negligence is removed. In other countries the employee is free to supplement industrial injury benefits by making a claim for negligence.

The legal liability approach is still used in many developing countries for the general provision of medical care. Thus large employers or employers of labour in mines or specific agricultural estates (e.g., sugar, tea, and rubber) are required to provide clinics and hospitals for their employees and dependents. This is one way of ensuring that health services are provided to people working far from the main urban health services. It is, however, difficult to ensure that employers comply with the spirit of the law. Moreover, employees may suspect that the doctors and nurses working in such services owe their primary loyalty to the employer and thus tend to economize on the treatment or are reluctant to certify time off for sickness. A further problem is that it is uneconomic to provide government services in these areas for the remainder of the population who are not employed by the major local employer and is difficult to integrate employers' services with government services.

In several countries employers are required to provide defined levels of cash benefits during short periods of sickness (e.g., six to eight weeks). This avoids the administrative complexity of a social insurance benefit paid by a national scheme or a sick fund supplemented by an employer's scheme. Provision may be made to protect the workers' rights if the employer goes out of business.

While social insurance is preferred to the employer liability approach by social security experts because it can give better protection, employer liability is still widely used in developing countries not only for employment injury but also for sickness and maternity benefits and employer severance payments.

Many developing countries require certain employers to contribute to a provident scheme providing a lump-sum payment in the event of death or disability or on retirement. Such a scheme differs from a social insurance scheme in that each worker usually has his own personal account from which he or she can draw if certain contingencies arise; there is no pooling of risks among members as there is in a social insurance scheme. Such schemes, which avoid the administrative complexity of paying a regular cash benefit, may be a step toward a full-fledged social insurance scheme. There are three disadvantages of such schemes from the point of view of the beneficiary. First, provision is inadequate for risks occurring early in working life. Second, the funds are generally invested in government stock with a rate of interest fixed in money terms that may be below market rates; the real value of the accumulated savings may thus be substantially eroded by inflation by the time of retirement. Third, a lump sum once received cannot normally be securely invested to provide an income protected against inflation. Moreover, it may be frittered away or unwisely invested. From the point of view of governments, however, such schemes are attractive in that they generate forced savings that can be used to finance national development plans.

The use of compulsory insurance as a mechanism to provide medical benefits and cash benefits in the case of sickness, disability, widowhood, and old age became acceptable to legislative bodies fearful of accepting extended state intervention that would require higher taxes to finance pensions or other benefits. In societies where self-help by voluntary insurance had been widely supported, the further step of compulsory insurance was seen as a means of making workers “good” by legislation. Because the schemes were financed by contributions levied on both employers and employees with, in some cases, modest state subsidies, unacceptable levels of national taxation were avoided; indeed, as such schemes reduced the need for social assistance or poor relief, the burdens on local taxation were reduced.

Compulsory insurance contributions are essentially a tax on earned income. Employers try—and probably succeed in most circumstances—to shift the burden of their share of the contribution either to consumers in higher prices or more probably, in the long run, to their employees by paying them less in cash. Thus employers' contributions are in most cases not paid at the expense of profits. However, the fact that the worker is told that the employer has to pay a proportion of the total contribution helps to make such schemes acceptable to employees, quite apart from the clearly defined benefits that flow from paying their share. Compared with the complexities of an income tax, a social insurance tax is a simple one to collect. But if the level of contributions is high, it creates incentives for workers to become self-employed in what has come to be called the “black,” or “underground,” economy and for employers to avoid contribution liability by employing contract labour rather than full-time staff.

In terms of meeting social needs or reducing poverty the social insurance method of provision has a number of disadvantages. Over the years many countries have tried to find means of countering these. First, the analogy with private insurance, which made such schemes politically salable, carries with it the social disadvantage that benefits should be paid to those who have contributed. Thus such schemes cannot provide benefits to persons who have never worked, for example, persons who have become disabled before reaching the age to enter employment, those incurring risks very soon after entering employment, and women (or men) who do not enter the labour force because of family responsibilities. Second, the expectation that benefits should be related to the amount paid in discriminates against individuals, usually women, who because of family responsibilities have fewer years in paid employment. Moreover, workers with dependent spouses and children have greater needs than single persons, though the assumption of marital responsibilities—or the converse assumption of marital dependency—is not strictly speaking an insurable risk. Third, where contributions are related to earnings, the benefit will be low for low earners, thus failing to protect them from poverty. The alternative approach, which some countries have adopted, of flat-rate contributions and flat-rate benefits can impose heavy burdens on low earners with family responsibilities. Fourth, it is difficult to bring the self-employed and those working for small employers (e.g., agriculture or domestic work) into such a scheme.

Over the years many countries that started with a purist insurance approach have modified their schemes to try to overcome many of these disadvantages. For example, extra benefits have been provided to persons with dependents. Contributions have been credited to persons outside the labour force for reasons of family responsibility, sickness, or disability. Minimum benefits have been introduced above those strictly warranted by low earnings-related contributions, or the benefit formula has been weighted in favour of lower earners. And some countries have made contributions earnings-related or integrated them with income tax while still paying flat-rate benefits.

Because of the disadvantages of the social-insurance approach, some countries have made certain benefits available to all residents and financed them out of taxation. When the benefit is paid on the basis of age it is sometimes called a demogrant. The most common benefit selected for this approach is the family allowance. The underlying philosophy is that provision for children should not depend on whether the parent is or has been in paid employment. Some countries have adopted this approach for pensions or at least for a minimum pension. In some cases this evolved from an earlier provision of an income-tested pension. In other cases this was the only way forward for governments in which the power to levy social insurance contributions did not rest at the federal level. Some countries have more recently applied this approach to provision for the disabled in the form of a minimum benefit based only on the extent of disability. It is increasingly applied to medical benefits on the grounds that all citizens have a right to health care.

The development of social insurance and demogrants has not removed the need for social assistance to fill gaps in provision in advanced societies. Social assistance is based on need and thus requires declarations of income, family size, and other circumstances. Thus it is provided on the basis of a means test that takes into account not only income but also capital; persons with a specific level of savings may be ineligible. Alternatively it may be only income-tested, the income from capital being assessed in the same way as other income. Often those who have been given the task of operating the scheme (e.g., social workers) have been allowed considerable discretion in deciding whether to give assistance and how much to give in certain types of cases. Not all basic rules are known to claimants. The tendency in industrialized countries has been to try to transform assistance into a right with published scales and regulations and opportunities for appeal. With codification has often come standardization and the unfortunate removal of some of the flexibility available under discretionary systems.

In some countries social assistance plays a residual role, providing a less favourable level of support than is normally available from social insurance benefits. In other countries (e.g., the United Kingdom) social assistance plays a considerable role in supplementing social insurance benefits for those without other sources of income such as sick pay or employers' pension schemes as well as providing for those without rights to benefits (e.g., one-parent families other than widows) or those whose benefits have run out because they are paid only for a specific number of months (e.g., unemployment benefits).

There are disadvantages of the social assistance approach. First, it penalizes saving and earning because income from any source is normally deducted from the assistance that would be payable, and persons with a certain level of savings may be ineligible until they have used them up. Second, it tends to stigmatize the recipient; and third, partly for this reason and partly because of the difficulty of knowing detailed rules of entitlement, there are considerable numbers of people who would be eligible but do not make claims. Partly because of this problem of stigma, social assistance programs are called by a variety of different names in the hope that they will be more acceptable to applicants. For example, the term used is supplementary benefit in the United Kingdom and GAIN (guaranteed income) in British Columbia. Eligibility rules differ considerably from country to country and are usually determined locally rather than centrally. Moreover, schemes are generally financed wholly from taxes—often local taxes. In the United Kingdom, where rules are determined centrally, persons in full-time work are not eligible. In the United States only households headed by a single parent are eligible for the Aid to Families with Dependent Children program, which creates incentives for desertion or fictitious desertion. There are, however, further programs for the blind, the disabled, and the aged.

The United States uses what is essentially the social assistance approach for meeting the medical care needs of low-income persons under the Medicaid program. Ireland operates a scheme by which persons with low income can apply for a medical card that gives them more extensive rights to free health care than are available to other income groups. Those with low incomes in South Korea can also apply for cards giving rights to free or nearly free health care.

A number of countries in Europe have developed separate income-tested provisions to help persons with low incomes meet the cost of rent or property taxes. Such housing allowances are available to persons whether in work or not and take account of family composition as well as rent payable.

Partly because of the stigma attached to social assistance, the difficulty the potential beneficiaries have in understanding eligibility, and their reluctance to apply, it is often proposed that the information provided to the state from income tax returns should be used by the state to determine the need for cash payments to persons with low incomes. The ability to do this depends on persons' being required to make income tax declarations by a certain date however low their incomes, which is not the practice in every country. Canada has a program to supplement on the basis of this information the incomes of persons drawing pensions. This approach is much less appropriate for younger people whose financial circumstances change considerably from year to year and month to month due to sickness, unemployment, job changes, marriage breakdown, remarriage, and so on. People need money when poverty strikes, not after the end of the income tax year.

Provisions for cash benefits change from time to time in all countries. Thus no description can be fully up-to-date. The information presented here is chiefly based on the returns made by 140 countries to the Social Security Administration of the United States and published in 1985 as Social Security Programs Throughout the World.

Three basic types of state pension schemes predominate. The first is a flat-rate pension with no income test. This may be available on a test of residence only or with the stipulation that the person has been employed for some specific period and has paid requisite contributions. This approach is found mainly in Scandinavia and the Commonwealth countries. The second is an income-tested pension. The third, and most common, type is a pension related in some way to earnings during working life. A further complication is that most countries with a flat-rate pension later developed a second tier of pension rights based on earnings during working life. In other words, the first and third principles are combined.

Several countries in Scandinavia abandoned an early means-tested pension in favour of a flat-rate pension after World War II because of the unpopularity, complexity, and discouragement to savings of the means test. Later the level of the pension was regarded as inadequate for all except the low-paid, and an earnings-related tier was established on top. Thus flat-rate pensions are provided on tests of residence in Denmark, Finland, Norway, and Sweden. In three of these schemes a married couple receives substantially less than two single persons. In each case the schemes are supplemented by earnings-related pensions. Canada followed a pattern of development similar to those of the Scandinavian countries. The United Kingdom also gradually moved over to a two-tiered pension, but rights to both tiers depend on contributions paid with credits for absence from work for approved reasons; employers' schemes can be used to provide a specified minimum upper tier of pension.

The Scandinavian and Canadian two-tiered approaches have a number of advantages. First, non-means-tested basic pensions can be provided to persons without a contribution record—including the disabled, those who have not worked because of family responsibilities, and divorced or separated wives. There is a similar advantage in New Zealand's scheme. But, in addition, those who have had higher earnings and thus paid higher contributions receive higher pensions with the value of these guaranteed in terms of purchasing power by the government. This reduces the scope for employers' pension schemes in which the purchasing power of the pension finally paid depends on how far the yield of investments has managed to keep pace with inflation.

Contributory pension schemes, when they were first established, were run on much the same basis as private pension schemes. The level of contributions was calculated by an actuary, and a capital fund was built out of which the pensions could be paid. Even if there were no further contributions, the money was intended to be available to pay out pensions to contributors from the accumulated value of their contributions. This arrangement is known as capitalization or fully funded financing. The first scheme in Germany, enacted in 1889 and based on capitalization, covered most employed persons with earnings up to a specified level. The earnings-related contributions were equal for employees and employers, and there was a subsidy from the state to provide the low-paid with somewhat higher pensions than their contributions warranted. A breach with the principles of private insurance was made to allow workers close to retirement when the scheme went into effect to receive higher pensions than their contributions had earned them. This system of “blanketing in” older workers has frequently been used in other countries when new pension schemes have been established.

It was the experience of rapid inflation after World War II that led to a fundamental change in the financial basis of pensions. Instead of the contribution level's being sufficient to build up a large capital fund, it was calculated according to the expected cost of pensions due to be paid over the next few years. This pay-as-you-go method of financing statutory pension schemes, which became the normal arrangement, contrasts sharply with private pension schemes. The latter still have to accumulate capital funds because, unlike state schemes, they have no power to compel future generations to join them. Thus state pension schemes are essentially a “compact between generations.” Those at work are compelled to pay to the pensioners of today in expectation, written into the law, that their pensions will be paid by the next generation of workers.

A second major development in pensions began in the late 1950s in response to rapid economic growth. It became recognized that, if pensions were paid out on the basis of the money value of contributions paid in over a working life during which real earnings had been growing rapidly, pensions would amount to a low proportion of earnings at the time of retirement and a still lower proportion of what those at work would be earning 10 or 20 years later. Thus complex formulas were introduced to adjust pensions to the general level of earnings at the time of retirement. West Germany set the pattern in 1957 and was followed by several other European countries—for example, Austria, Switzerland, and the United Kingdom. An alternative approach (e.g., in Italy and some eastern European countries) is to base the pension on the last few years of earnings. As this can be unfavourable to workers whose earnings decline in the later years of working life, some countries (e.g., France) base pensions on the best few years of earnings. The former Soviet Union offered an option of the last earning year or the best 5 consecutive years out of the last 10.

The practice of giving low-paid workers higher pensions than were earned by their contributions and those of their employers, which was built into the original German scheme but later abandoned, has been copied in later schemes (e.g., that of the United States). An alternative or further way of helping low-paid workers is to provide a minimum pension, as in Germany or the United States (though in 1981 the U.S. provision was removed for people not yet retired). This particularly helps women, whose average annual income, despite legal inroads against discrimination, remains well below that of men and whose pension contributions are now likely to be interrupted by leave from work for family responsibilities. A much more common provision is an income-tested social pension as, for example, in Belgium or France.

Another development mainly of the period after World War II has been the automatic adjustment of pensions according to an index of prices or in some cases to the average level of earnings, or whichever is more favourable. Some countries have postponed adjustments or modified their formulas, particularly when prices were increasing faster than earnings.

The age at which full pension can normally be drawn varies considerably between countries. In Europe the normal age for men can be as high as 67 and as low as 60 and for women as high as 66 and as low as 55. Some developing countries have still lower pension ages. To some extent pension age tends to reflect the expectation of life in the particular country. The pension age for women, however, is often lower than for men; one reason often cited for this is that husbands tend to be older than their wives, and so the disparity in pension ages permits simultaneous retirement. The arrangement, however, is disadvantageous for women who are retired compulsorily at the lower age after having had less time to accumulate a record of contributions. There is, therefore, a trend to equalize pension ages between the sexes. To do this by lowering the male pension age is expensive; it is for this reason that the European Union has not made this binding on member states in its directive on equal rights to social security.

Some countries have long had provisions allowing the pension to be drawn a few years earlier than the stipulated age of retirement with an actuarially calculated reduction in the pension paid. Such provisions are suited to the more generous earnings-related schemes in which a reduced pension would not normally cause poverty. Ill health is a common reason for early retirement, though many choose this option in order to enjoy retirement while still in good health. There commonly are also provisions by which people who wish to postpone their retirement and continue to contribute can draw a larger pension. In some cases these arrangements have been introduced in the hope of encouraging later retirement, thus modifying the deterioration of the ratio between the employed population and the retired population, which necessitates higher levels of contribution by those still working as the proportion of the pensioned population increases.

Despite the logic of raising the normal pension age in line with an improved expectation of life, changes in schemes of industrial countries in the 1960s tended to lower the age. This trend has continued as the level of unemployment has increased, despite the financial burden this places on the schemes, particularly in the long run. The political objectives of reducing the number of persons recorded as unemployed and creating jobs for younger people have taken priority. Thus a wide variety of complex provisions have been written into pension schemes defining the circumstances in which full pensions can be drawn a few years earlier than otherwise stipulated. This may be allowed to those with many years of insurance (e.g., 35), to those who have been unemployed for a substantial period (e.g., a year), to those who are disabled, to those with arduous or unhealthy occupations, and to those whose jobs are being released for younger persons. In some countries the pension is income-tested below the normal age. Contrary to this trend for earlier pensions, the United States has raised the future pension age in two steps from 65 to 67 in response to the long-run financial prospects for the pension scheme.

A development pioneered by Norway in 1972, and since followed by more and more countries, was to allow persons aged 67 to 69 to reduce their working hours and receive at the same time a partial pension. This enabled older people to make a gradual transition between work and retirement. The change was made when the pension age was lowered to 67. Sweden followed in 1976 with a provision for those aged 60 to 64. Partial pensions have also been introduced in Spain and, on a much more restricted basis, in the United Kingdom.

In most schemes in industrialized societies there is a limit on what the pensioner can earn without leading to a reduction in pension. Some schemes specifically require the pensioner to leave his or her job on receipt of the pension. These provisions add to the wider pressures leading to the steady fall in the proportion of persons in full-time work above the normal pension age. But the main reason for this trend is the increasing generosity of pensions, both public and private.

Early pension schemes made extra provision for a dependent wife, and more did so between World Wars I and II. This can mean that women's contributions are “wasted” in the sense that the pensions they earn are less than they have a right to as a dependent wife. The greater frequency of divorce and cohabitation has meant that more women are wholly dependent on the pensions they earn in their own right. Moreover, some pension schemes make no provision for a dependent wife (e.g., in Germany, Austria, and Italy). The issue of women's rights to pensions is particularly important in the context of poverty, as women on average live longer than men.

A few countries allow housewives to contribute to pension schemes on a voluntary basis, but few women do so in practice. Others have adopted provisions for the dividing of pension credits between spouses. In the United Kingdom a man or woman can be credited with a full year's pension rights for each year up to a maximum of 20 during the whole of which he or she is caring for a dependent child or disabled relative. These rights are based on the individual's previous record of contributions.

In the early schemes widows over pension age were entitled to a proportion of the pension of their husbands. More and more schemes have been amended to make similar provisions for widows and widowers. Usually survivors can choose between their own personal rights and a proportion of the rights of their deceased spouse, but in Sweden widows can receive both earnings-related benefits on top of one flat-rate pension. This right is available, up to a maximum, to a widower as well as to a widow in the United Kingdom.

In most countries provision for occupational injury is the oldest form of social security. The original German law of 1884 provided for workers to receive half pay for four weeks followed by two-thirds pay during temporary disability. In cases of permanent disability two-thirds of earnings from the year preceding the accident were paid out, with a proportion of this pension paid in cases of partial incapacity. Extra provision could be made for persons needing constant attention. The scheme was wholly financed by the employer, who paid insurance contributions, assessed on the degree of risk involved in the employee's occupation, to statutorily established associations. The associations then paid out any benefits.

The British law of 1897 made employers liable for compensation but did not require employers to insure against the risk. Compensation was half the basic pay for up to six months, at which point the claim could be settled by a lump sum. These two very different precedents influenced developments in other countries. Continental European countries tended to follow the German model and the Commonwealth and the United States that of the United Kingdom. An act modeled on the British law of 1897 was passed in India in 1923, though the coverage was small. Moreover, the Belgians, Dutch, and French as well as the British preferred to introduce in their colonies laws imposing liability on employers rather than funded insurance schemes. By the end of World War II most colonies had such laws. These laws were often later augmented or replaced by insurance schemes. Some Scandinavian countries require the employer to insure but allow him to choose his own insurer.

Under an act of 1946 the United Kingdom introduced compulsory insurance through a state scheme with the same rate of premium for all employers. Benefits for incapacity were at a flat rate followed by a disablement pension based on degree of disability to which were added other allowances depending on the situation of the pensioner, including the loss of earnings and need for attendance.

Insurance is now compulsory in most industrialized countries, but the use of private insurance continues in a few countries (e.g., Denmark and Finland) and the majority of U.S. states, while some countries give the employer the right to choose between a public or private insurer. Work-related injuries and an increasing number of occupational diseases lead in nearly all countries to higher benefits and more generous provision than are paid for sickness or injury not arising from work. For example, some countries in western and eastern Europe provide 100 percent of previous earnings as a temporary disability benefit. These benefits normally continue until recovery or the award of a long-term benefit. In most countries loss of earning capacity is a major consideration in the assessment of long-term benefits, and partial disability is more generously treated than in other social insurance programs. The long-term benefits in some countries can also amount to 100 percent of earnings. But the procedure of seeking lump-sum settlements from the courts still remains in some countries and some states of the United States, with all the associated costs, delays, and uncertainties and the difficulty of turning a lump sum into a secure income.

There are three special features of most occupational injury schemes that reflect their historical origins in the employer's liability. First, schemes are normally financed solely by employers' contributions. Second, the right to benefits operates from the very first day of employment. Third, a cash benefit is seen as compensation rather than income maintenance. For this reason dependents' benefits are not normally provided, but there is provision for surviving dependents. In addition, a compensatory benefit may sometimes be paid in addition to earnings or pensions. These features are found only in provisions for sickness or disability that are of occupational origin.

Both employers and employees normally contribute when the scheme is based on social insurance. Some minimum period or amount of contribution is generally required before there is entitlement to benefit. The amount of the benefit may depend on how long contributions have been paid, as for pensions, which is disadvantageous for those disabled early in working life. The main benefit is intended for income maintenance and thus cannot be drawn at the same time as other benefits or pensions with the same purpose. Finally, there is more likely to be provision for a dependent spouse.

There has been a tendency in the period since World War II to bring occupational injury schemes into a closer relation to other social security schemes. Switzerland has always covered work-related and other accidents in the same scheme, established in 1911; New Zealand later adopted the same practice. The separate provisions for occupational injury and other disability raise difficult problems in specifying the distinction. Occupational injury normally has to “arise out of and in the course of employment.” Some schemes allow travel to and from work to be considered within the “course of employment,” while others do not. There are considerable difficulties in identifying whether certain disabilities (e.g., deafness or arthritis) arise from work, and there are instances in which an injury is only partially attributable to the work situation. Part of the justification for combining the provisions for occupational injury and other disability is to eliminate such ambiguities. In addition there is the social argument that it is wrong to pay different benefits to different people, all of whom have the same degree of disability no matter how or when the respective conditions were caused. The Netherlands is the only country that has responded to this argument. From 1976, unified provision for disability has been made irrespective of cause. Costs of such a program can be substantial if all disability coverage is raised to a level approaching that of the previous, often considerably higher, occupational injury coverage.

In the case of sickness that is not associated with any occupational factors, most industrial countries pay a short-term benefit followed by a long-term pension after periods varying from about six months or less to a year or more. Some countries, such as Austria, Belgium, Germany, and the United Kingdom, place responsibility for paying a benefit on the employer for the early weeks of sickness (though he may be reimbursed), after which the social security fund assumes payment. In some countries benefits may not be payable for an instance of illness lasting less than, for example, three days. In longer periods payment for the first three “waiting” days may be included in the benefit. Doctors' certificates may not be required for short spells of sickness. Benefits may be as high as 100 percent of earnings (e.g., Austria and Belgium) in the early weeks of sickness or subject to a maximum, as in Norway, or for the full 52 weeks, as in Luxembourg. Or the rate may be 90 percent, subject to a maximum (Sweden and Denmark). Other countries normally pay only 50 or 60 percent (e.g., France, Canada, and Greece). The benefit is flat-rate with extra for dependents in Ireland and the United Kingdom (after the eight weeks paid by the employer). The benefit is also paid on this basis in Australia and New Zealand, but it is means-tested. The United States is the exception among highly industrialized societies in that in most states there is no provision for short-term sickness apart from a special scheme for railway employees and social assistance, or welfare. In practice, provision is left for bargaining between employers' and employees' representatives.

Long-term invalidity pensions were included in the original 1889 German pension law (which was the first piece of legislation of this kind) for those who had lost two-thirds of their earning capacity. Many countries followed this model as part of (or as a later development of) their pension laws. In European countries invalidity pensions became payable after full short-term sickness benefit rights had been received. After World War II provisions were made in some countries for those who had considerable partial invalidity. Some countries require persons to have been insured for five or more years in order to be eligible for an invalidity pension, though generally there are means-tested pensions in industrialized countries for those who do not meet these requirements. In Australia and New Zealand those who meet residence requirements are eligible for income-tested pensions.

In countries with earnings-related pension schemes the invalidity pension is often calculated in the same way as the old-age pension. This means that the level depends on the number of years of contribution, though some countries have special concessions to enhance the pensions of those drawing them early in working life. Invalidity pensions may be supplemented by allowances for dependents and for constant attendance and other special needs.

Some countries make special provision for housewives who lack the contribution records that would qualify them for an invalidity pension. One such country is the United Kingdom, though the benefit is low and flat-rate. In Denmark a housewife can receive a substantial income-tested pension in her own right. Another group for which some countries have begun to make special provision is those who have been disabled from birth or before entering the labour market. These groups are provided for in the unified scheme of The Netherlands.

Most countries, however, are far from the position in The Netherlands, where all disabled persons are treated on a similar basis irrespective of the cause of disability. Those whose disability arises out of the work situation are generally most favourably treated; some countries provide full compensation for loss of earnings plus special allowances when required. Those who have paid contributions are generally treated better than those who have not, and often benefits depend on how long contributions have been paid.

While sickness and disability are actuarial risks in that the incidence does not vary greatly from year to year, this is not the case with unemployment. It is partly for this reason that the duration for which unemployment benefits can be paid is limited in most countries or that benefits are reduced after a designated period. A further reason is to induce the unemployed to seek and accept work after benefits end or when they fall, although such work may be less well paid than the individual's earlier work and may provide an income that is lower than the unemployment benefit that has ceased.

The payment of contributions plays a critical role in policing eligibility for unemployment benefits; as a result the benefit is not payable to all persons who are involuntarily unemployed. The school dropout who has never had a job or has held one only for a short period is normally ineligible for unemployment benefits. Women seeking to return to work after child-rearing are also ineligible, even though contributions were paid before leaving work. A prospective recipient must normally have held a job from which he has been released immediately before the benefit is claimed, and the individual must establish that he is available for work by registering at an employment office. Normally anyone who has voluntarily left a job or been discharged for misconduct is denied benefits or is penalized.

In some countries the level of unemployment benefits is deliberately set at the same rate as the benefit for short-term sickness (e.g., Canada, Denmark, and The Netherlands) so as to create no incentive for the beneficiary to try to establish eligibility for the higher benefit. In other countries the benefit for unemployment is at a lower level than the benefit for short-term sickness (e.g., Germany, Greece, and Hungary). Some countries that pay an earnings-related benefit for sickness pay a flat rate for unemployment (e.g., Bulgaria and Italy). In Australia and New Zealand unemployment benefits, like sickness benefits, are subject to a test of income. The duration of the benefit varies from 13 weeks in Bulgaria to six months in Hungary, Italy, and The Netherlands and a year in France, Germany, Luxembourg, and the United Kingdom; in Belgium benefits can be continued indefinitely. In many, but not all, countries the unemployed can claim social assistance after their unemployment benefits cease. In several countries in northern Europe unemployment benefit schemes are operated by trade unions, though with substantial government subsidy.

While only a few countries had family allowances before World War II and several of the schemes covered employed persons only, with financing by the employer, there was a rapid extension of schemes in the 1940s and '50s. The extension was in large part attributable to the influence of the Beveridge Report in the United Kingdom. Following the British example most of the new schemes in Europe, Canada, and Australasia included all resident children. A second influence was France, which introduced flat-rate family allowances for all children of employed persons in its African colonies—a system also introduced in some Latin-American countries (e.g., Bolivia, Brazil, and Chile). The majority of schemes cover only employed persons, but a minority, particularly to be found in industrialized countries, pay allowances to all residents. The United States is exceptional among the latter countries in making no provision at all except in aid to dependent children paid on a means-tested basis.

Some systems of family allowances are intended to reduce poverty in large families or, particularly in eastern European countries, to increase the birth rate; the rate paid per child increases with the number of dependent children, reaching a maximum rate with the fifth or sixth child and subsequent children in, for example, Australia, Belgium, France, Ireland, and Norway, or the eighth child, as in The Netherlands. In the former Soviet Union, family allowances began with the fourth child and reached the maximum rate at the 11th. Some systems seem to suggest a desired maximum family size insofar as the rate falls for subsequent children once there are three (e.g., Bulgaria, the Czech Republic, and Slovakia) or two (e.g., Greece and Hungary), or allowances may be payable for a maximum of six children (e.g., Morocco). Finland recognizes that a mother is less likely to go to work if a child is under three years of age and therefore pays a supplement. On the other hand, Austria pays higher rates for older children because they are more expensive to maintain. Entitlement to family allowances ceases when a child reaches a particular age—in most cases the age when compulsory education ceases, though allowances may be continued when a child continues in full-time education or is disabled.

During the 1970s a number of countries decided to abolish income tax allowances for children and make a corresponding increase in the level of their family allowances. It was recognized that the largest beneficiaries from the tax allowances were high-income families with high marginal tax rates, and it was decided that this indirect benefit for children should be fairly shared among all families so as to increase the efficacy of family allowances in reducing poverty. Changes of this kind were made in Australia, Canada, Denmark, West Germany, Israel, New Zealand, and the United Kingdom. Denmark has gone one stage further and removed family allowances from the higher income groups by means of an income test. The United Kingdom has an additional income-tested allowance called the family income supplement which gives further help to low-income families.

It is the general practice for schemes that provide sickness benefits also to provide a maternity allowance starting before the birth of a child and extending for a number of weeks afterward. In some cases the rate of benefit is the same as for a sickness benefit, but in many cases the rate is higher—66 to 100 percent of previous earnings. Sweden has pioneered a parental allowance that can be drawn by the father as well as the mother to encourage fathers to take their turn in staying at home to look after the young child. In some cases a lump sum is also paid on the birth of a child to help pay for nursery goods and clothing.

During the 1970s there was a concerted effort in eastern Europe to try to increase the birth rate by increasing the period for which a maternity benefit was paid and by giving credits in the social insurance scheme to mothers who stayed at home to look after a young child. Similar credits are provided in the United Kingdom for persons who stay at home to care for a child or a disabled relative, but the motive in this case is to increase the personal pension rights of those, particularly women, who have accepted family responsibilities.

Provision is normally made for a widow below pension age left with a dependent child. Where pensions are earnings-related, the pension for a widow typically amounts to one-half to three-quarters of her husband's pension rights. In some countries the benefit is income-tested or time-limited (e.g., three years in France). Other schemes vary considerably in the extent to which provision is made for widows. Some countries pay benefits providing widows are of a certain age when their husbands die. The age may vary between 40 (The Netherlands) and 55 (France). Some countries only pay the benefit providing the marriage has lasted for a specified period (six months in Greece; two years in France). Other countries pay the benefit to any widow who is disabled or to widows of any age for a short period or indefinitely. Widows' benefits normally cease on remarriage. A widower may be able to claim rights similar to those of a widow if he was dependent on his wife. Some countries extend widows' rights to divorced women. Increasingly, long-term provision for widows without dependent children is being questioned in societies where the trend has been for more and more married women to engage in paid work.

Provision for single parents other than widows is normally left to social assistance where such a scheme exists. Some countries have a special income-tested benefit. In Australia this is at the same level as an old-age pension for a person aged at least 65 but less than 70. In New Zealand it is less than half this rate for a single parent with one child. The problem with either of these arrangements is that a less skilled woman is unlikely to be able to improve her position by taking paid work because earnings lead to a reduction of the benefit or assistance. Denmark pays an extra family allowance higher than the normal rate per child for a single parent. Norway pays an extra allowance as if for one more child. The United Kingdom pays an extra allowance at just over half the level of child benefit.

All of the industrialized countries have social insurance schemes, and nearly all of them cover the main contingencies discussed above. The United States is exceptional in not providing family allowances, in not providing short-term sickness benefits in the vast majority of states, and in having no general scheme of national health insurance other than for the aged and the poor. The extent of provision in developing countries varies between those that still make provision by employers' liability and those that make provision by social insurance.

Systems of organizing health services or health insurance systems and of paying providers are changed occasionally but less frequently than the detailed provisions for cash benefits.

The first national compulsory health insurance scheme, introduced in Germany under Bismarck's law in 1883, built upon precedents going back many years in the separate German states. Health insurance had developed mainly on an occupational basis and was a requirement for that occupation. The feudal obligation of the employer to his workers was given legislative substance in a society developing national markets, in which the employer without an obligation to pay to a sick fund might undercut the employer who had such an obligation. But the main reason for the scheme, as mentioned earlier, was to try to contain socialist tendencies.

The administration of compulsory health insurance was left in the hands of numerous local sick funds operating under legislative regulations. They became jointly controlled by employers and employees and made their own contracts with particular doctors and hospitals for the provision of services. All lower-earning workers were eventually required to be members of a fund. Doctors were paid in a variety of different ways, including salary and capitation. In the course of time there were major protests from doctors excluded from contracts with the funds, and the profession demanded the right for any doctor to undertake health insurance work. The substitution of payment per case and later fee-for-service payment, which the German medical profession fought for and eventually won, was a means of establishing open competition between all doctors wishing to take part in the scheme.

Health insurance was enacted in Austria in 1888 and Hungary in 1891 on a similar basis. A bill to introduce such a scheme in Switzerland was, however, decisively rejected by a plebiscite in 1900. The British Radical politician David Lloyd George (Lloyd George, David) visited Germany in 1908 to see the scheme firsthand and subsequently introduced compulsory health insurance for persons with earnings below an upper limit in Britain by a law of 1911. However, the scheme provided only for the services of the general practitioner and the drugs he prescribed; hospital benefits were excluded except for some provision for tuberculosis, partly so as not to disturb the charitable hospitals that provided free care to those in need. Moreover, as a result of pressure from the medical profession, the benefit in kind was administered by statutory committees for each area, which enabled every general practitioner to participate who wished to do so, rather than by the large number of friendly societies that had previously provided medical benefits under voluntary insurance and had made their own contracts with particular doctors. Payment was on a capitation basis, as in the previous friendly society schemes.

The introduction of compulsory health insurance was considered in Sweden in 1884 and Denmark in 1885, but both countries decided instead to encourage voluntary insurance by government subsidy. Whereas Norway introduced compulsory health insurance in 1911, Denmark did not follow until 1933, and Sweden not until 1955. In these countries public hospitals were well developed and heavily subsidized. A compulsory health insurance law was passed in France in 1920 but, as mentioned earlier, did not come into effect until 1930 owing to disagreement about the local control of the scheme and a dispute with the doctors about the method of payment. Fee-for-service payment was finally substituted for the capitation system originally proposed. Moreover, as the doctors refused to accept the intrusion of any third party between them and their patients, the scheme operated on a reimbursement basis; the patient paid the fee and claimed a refund for the major part of it from the relevant insurance fund. This reimbursement system was adopted later in Sweden, Finland, and Australia (under subsidized voluntary insurance).

When health insurance was established in Russia in 1912, insurance doctors were paid salaries and practiced from government-owned premises. This was the pattern adopted in Chile in 1924. Payment of doctors on a part-time salaried basis for work performed on premises owned by the sick fund became the general pattern in Latin-American countries and in Spain, Portugal, and Greece. In most of Europe and Australasia, however, existing hospitals began accepting insured patients when hospital care became covered by insurance; special hospitals for insured persons were built in Spain, some parts of Italy, and a number of countries in Latin America.

The next major step in the evolution of medical benefits was for countries to make them available to the entire resident population, financed wholly by taxation or financed in part by social insurance contributions. This step was taken first by Hungary in 1920 and then by the Soviet Union in 1937. New Zealand implemented similar coverage in a series of steps—free inpatient treatment for the whole population in 1939 and outpatient treatment, free pharmaceuticals, and part payment of general practitioners' bills in 1941, with further steps later on. The United Kingdom established its National Health Service in 1946. Norway made services available to all residents in 1956, Sweden in 1962, Denmark in 1973, Portugal in 1979, and Italy in 1980. By the 1980s more than 20 countries had adopted this system. This does not necessarily mean that all services are free at the time of use. Nor does it necessarily follow that all services are government-owned. Of the eastern European countries, some (Bulgaria, the Czech Republic, Slovakia, Hungary, and Romania) have adopted this approach; others (such as Poland) have not. About half of all countries retain an element of financing by social security contributions after adopting this approach.

Some countries in Europe have succeeded in securing high coverage of the population under compulsory health insurance without switching to a service available to all residents. Schemes cover the employed, the self-employed, and all social security beneficiaries and their spouses and dependent children: this can amount to 99 percent of the population. Other European countries (Germany and The Netherlands), nearly all of which have some system of private insurance, exclude the higher income groups from statutory health insurance. Alternatively, some benefits (e.g., hospital care) are available to the whole population, while higher income groups must make their own arrangements for certain other benefits (Ireland).

Apart from Cuba, which has a national health service, only three countries in Latin America (Argentina, Brazil, and Costa Rica) have managed to cover 80 percent or more of the population by health insurance. Moreover, coverage may not necessarily mean that services are equally available. Coverage extends to about half the population in Mexico, Panama, and Uruguay, and more than a quarter in Bolivia and Venezuela. In the remaining countries coverage is 10 percent or less. By no means do all of these countries extend the same rights to the spouse and children of the insured person. Several provide only maternity care and pediatric care for dependents. Coverage is more easily provided for the employees of larger establishments, which tend to be concentrated in urban areas. Even in urban areas those excluded tend to be the self-employed, domestic servants, and itinerant workers. The obstacles to expanding rural coverage include the much lower levels of earnings, the geographic dispersion, the less formal employment conditions, and more extensive self-employment and seasonal employment. Most important of all, some schemes have become too costly to extend on the same basis with tax subsidy to cover the whole population. Thus the remaining population must depend on poorly financed and staffed services provided by ministries of health. Health insurance is, therefore, increasingly criticized for exacerbating inequality in health care by outbidding government health services for trained manpower and for creating a heavy emphasis on sophisticated and expensive curative services in urban areas while the main health need is for preventive services to cut the incidence of infectious diseases in both urban and rural areas.

Japan has managed to avoid the worst of these effects and to achieve high coverage. India, conscious of the damage that health insurance could do to government services, has developed health insurance slowly as resources have become available for doing so in particular states. South Korea has introduced health insurance for the urban employed population and also has provided rights to those with low incomes in urban areas; the problem of covering the remaining half of the population in rural areas remains to be solved.

Many developing countries, particularly those that were previously British colonies, have made health services available to the whole population, providing free or nearly free services. This is the pattern, for example, in the West Indies, Kenya, Zimbabwe, India, Sri Lanka, Malaysia, and many Arab states in the Middle East. In most cases services were originally developed for the expatriate colonialists and extended in the course of time to local residents. The services tend for this reason to be heavily concentrated in urban areas, with little or no coverage of the rural population. With their limited resources, these countries are striving, as part of the World Health Organization's program Health for All, to extend rural coverage with primary health care to all areas by the year 2000.

Among the various national health schemes, benefits are provided in three ways. First is the direct service approach in which the government or insurance fund owns the facilities (hospitals and clinics), pays for supplies, and remunerates the staff on a full- or part-time basis. This is the approach used in the United Kingdom for hospitals and community services and in Scandinavia, where local authorities provide hospitals and clinics, though there may also be a parallel system of doctors working from their own offices. It is also generally used in eastern Europe, in Greece, Spain, and Portugal, in most countries in Latin America, and in most other developing countries. The hospital system in Canada is exceptional; the scheme determines budgets for general hospitals that remain in the hands of not-for-profit agencies.

The second method is the indirect contract with providers. The providers may be private entities (hospitals or practitioners) or public hospitals, but the health insurance scheme makes a contract with the provider and pays each provider for services used according to rates established in a negotiated contract. This is the system used for all services in such countries as Belgium, Germany, Luxembourg, and The Netherlands.

The third method is reimbursement, in which the patient pays the bill and applies for reimbursement. The provider may be public or private. This approach is widely used in France, some northern European countries for the parallel system using practitioners in the private sector, and to some extent in Australia and Sweden. The patient may be left to pay part of the bill, as, for example, in France. A fee schedule may be established for rates of reimbursement, but, unless strong measures are taken to prevent it, some practitioners may charge more than the established fee.

In practice many countries use a combination of these systems. Thus, for example, the National Health Service in the United Kingdom, with its direct service provision of hospitals and community services, uses indirect contracts for general practitioners, community pharmacists, opticians, and most dentists. Moreover, where private hospitals are used they are paid under contract, as is also the case in Greece, Italy, and Portugal. In a number of countries in Latin America health insurers use the direct service approach in urban areas but service dispersed populations in rural areas by using indirect contracts.

Health insurance schemes vary in the method by which providers are paid, and this can have a substantial impact on costs. Where doctors and dentists are paid on a fee-for-service basis this provides incentives for the provision of further services—even in France where the patient has to pay a proportion of the cost. In the Common Market countries about twice as many prescriptions are issued to patients when the doctor is paid on a fee-for-service basis as when he is paid on a capitation basis. More surgery is performed where doctors receive fees rather than salaries. Moreover, the patient normally has direct access to specialists and can visit several different doctors in the course of one illness; this also adds to costs. When hospitals are paid on the basis of an itemized bill, more items are often provided. Where hospitals are paid per day of care, there are incentives for the hospital to keep patients for longer than necessary. For this reason, some countries in Europe (Belgium, France, and The Netherlands) have required hospitals paid on this basis to adhere to a predetermined budget. Where hospitals are given a budget from the local or central government, costs are kept under control. Financial incentives for the provision of further services are avoided where doctors are paid on a salary or capitation basis (The Netherlands and the United Kingdom). But this can lead to delays in receiving treatment both for an inpatient and for an outpatient. A provision permitting access to specialists, normally only on the basis of referral by a general practitioner, can be enforced where the patient normally has access to only one practitioner; this helps to limit costs. The system of paying doctors part-time salaries, leaving the doctor free to undertake practice, as in Greece, Portugal, Spain, and most countries in Latin America, can lead to what patients see as poor quality in services—a lack of courtesy and limitation of time devoted to the consultation. For this reason many countries are beginning to offer full-time salaries without rights for the doctor to undertake private practice.

The right to free medical treatment was included in the original German scheme for industrial injury, and provision for rehabilitation was added in 1925. In the course of time more and more emphasis came to be placed on efforts to restore working capacity, and specialized institutions were created for this purpose. Many countries have copied the German example and developed highly specialized institutions owned by sick funds or under the control of the agency responsible for national health insurance for both physical and vocational rehabilitation.

Countries vary considerably in the extent to which their social security apparatus is centralized and unified. A high degree of centralization obtains in the Commonwealth countries and Scandinavia (except for health care and social assistance, which are decentralized to lower levels of government). A centralized scheme may be administered by a ministry or by a semiautonomous agency. In other countries schemes are more often run by separate occupational funds or by funds providing for different risks, as tends to be the pattern in continental Europe and Latin America. The control may rest with boards composed equally of employers and employees. Or it may be tripartite, with the government participating as the third party. In the United States responsibility for social security is divided between federal and state agencies. There have been attempts in some countries to secure greater unification, but such efforts have often encountered strong resistance from particular occupational groups with better benefits or lower contributions attributable to lower risks.

Social security regulations have become extremely complex and difficult to understand. Where there are separate funds, each may have a national office, with no branch offices to which the public has access. Disputes often arise over which fund is responsible for paying benefits to particular claimants. It is, therefore, not necessarily the case that all claimants obtain what they are entitled to receive, and substantial delays can occur while entitlements are sorted out. Problems of this kind are not, however, unique to the public sector. Some private insurance companies are resistant to paying out claims. Unified social security systems with local offices are more accessible to the public, but the offices are not always adequately staffed to give the public prompt and efficient service.

Social assistance regulations are inevitably even more complex to operate than other parts of the social security system. Moreover, they frequently contain a considerable element of discretion. Where schemes are administered by social workers there can be what beneficiaries see as potential coercion; failure to follow the social worker's advice may be thought to lead to the reduction or removal of benefits. Some have argued that all social assistance regulations should be published so that claimants can know their rights and thus be in a position to appeal against decisions to refuse benefits or extra allowances. Adoption of this approach has led in some cases to regulations that are too complex for the staff to operate efficiently, or in others to regulations that have been streamlined at the expense of former provisions for discretion. Particularly contentious is the question of cohabitation. If an unemployed married woman living with her wage-earning husband is not entitled to social assistance, it would seem at first sight only fair that an unemployed woman cohabiting with an employed man should be treated in the same way. But cohabitation may not be accompanied by maintenance and is anyway extremely hard to define. The borderline between a lodger and a cohabitant is by no means clear-cut in all cases nor readily established by any outside agency. Attempts to do so can involve considerable invasion of personal privacy.

In most countries the major part of the cost of social security is paid for by proportional contributions of earnings from employers and employees. The contributions may be divided equally between employers and employees, except for the whole cost of the occupational injuries scheme, which falls to the employer. Alternatively the employer may pay about twice the amount falling to the employee. There is usually a “ceiling,” or level of earnings, beyond which the contribution becomes flat-rate at the level of contribution due on this maximum of earnings, though this is not the case in either Sweden or Switzerland. The maximum varies from around 50 percent above average earnings (e.g., France, Ireland, and Italy) to twice average earnings (e.g., Germany, the United Kingdom, and the United States) or higher (Norway). The reason for this may be to prevent insurance contributions from overlapping with high marginal rates of income tax or to leave the replacement of high earnings to the private sector. Some countries also exempt very low earners from contributions or make the employer pay them instead of the employee.

Usually some portion of costs is left to be met from taxation. At the very least the government will stand by to meet any deficit between benefits and contribution income. During the 1970s there was a trend in most countries in western Europe for costs to be shifted away from employers and onto taxes (e.g., Denmark, Ireland, Italy, The Netherlands, Portugal, and the United Kingdom) or to employees (Austria, France, and Germany). One reason for the trend toward tax financing was the growth of unemployment financed by social assistance payments.

Countries in which no costs at all fall on taxes include the small schemes in Burundi and Ethiopia and the wider schemes in Malaysia, the Philippines, and Singapore. At the other extreme, however, countries where contributions play a very small role and by far the bulk of costs is covered by taxation are Australia, Denmark, and New Zealand. In the United Kingdom, where the national health service is primarily financed from taxes and social assistance plays a major role, roughly half of the costs are borne by taxes and half by contributions. Several eastern European countries have no employee contributions; instead, their schemes are mainly financed by employers.

The relative merits of financing by contributions or taxes have long been debated. In favour of contributions it is argued that making beneficiaries pay prevents irresponsible increases in benefits and, where there are separate funds, encourages participation by both employees and employers. The payment of contributions also helps to ensure that commitments are honoured. Contributions are administratively easy to collect since the employee has an interest in securing compliance by the employer. The benefits to the employee of paying are clearly identified, while the cost falling on employers may create some incentive to prevent certain occupational risks from arising. Finally, only by earmarking contributions can earnings-related benefits be justified.

The critics of contributions argue that where they are flat-rate or where earnings-related contributions are only payable up to a low ceiling of income they are regressive and constitute a heavy burden on the poor; progressive taxes on income would be preferable, as they vary according to ability to pay and are also levied on investment income. It is also argued that tax-financing enables governments to judge priorities among all fields of public expenditure, and, where it leads to administration by government, this secures closer coordination between social security and other services. In addition, high contributions lead to the growth of the black, or underground, economy. This is a major problem in France and Italy with their high employers' contributions and leads to a widespread lack of social insurance coverage.

An argument that became more strongly pressed when levels of unemployment rose in the 1970s was that high employers' contributions made products uncompetitive in world markets, particularly in the case of labour-intensive industries, compared with products from Third World countries where social security is less developed. This was said to sharpen the recession and aggravate unemployment in highly industrialized countries. While it is true that employers might gain a short-term advantage if contributions were lowered, it is much less certain that this gain would be sustained in the long run. What was gained in lower contributions might sooner or later have to be conceded in higher wages and salaries or in other wage costs. If the argument were valid, such countries as Australia, Denmark, or New Zealand, which make little use of employers' contributions, would be seen to be cornering a heavy share of world trade. The fact that this has not happened reinforces the argument that it is total labour costs, of which social security contributions are only a part, that affect competitiveness.

It has been claimed that high employers' contributions particularly damage labour-intensive firms and encourage the replacement of labour with capital. In examining this assertion it is relevant first to remember that firms making capital goods also have to pay the same high employers' contributions and that capital-intensive firms pay them indirectly on raw materials, facilities and equipment, and energy. Second, high employers' contributions may well cause cash wages to be lower than would otherwise be the case so that total labour costs are not, in fact, increased by employers' contributions. Third, insofar as high employers' contributions encourage all firms to use more capital-intensive methods of production, this applies to labour-intensive firms as well. This encouragement of investment may lead to production at lower cost and thus a more competitive position in world markets in the longer run.

While there is a lack of convincing evidence that employers' contributions are bad for employment, a low ceiling on contributions may itself damage employment. It may discourage offers of part-time work and lead employers to prefer offering overtime to taking on additional workers. This was the view of international experts appointed by the International Labour Office, who therefore recommended in their report of 1984 that contribution ceilings be abolished.

The substitution of taxes for contributions may not relieve poorer workers if the extra taxes come from goods such as tobacco that are consumed more heavily by those with low incomes than those with high incomes in industrialized countries. There is no guarantee that governments would raise the extra revenue from progressive taxes; they may, for example, lower the threshold at which income tax is paid.

The strongest case for contributions is that they justify earnings-related benefits. The strongest case for taxes is that they are used in many countries to make benefits available to all residents—whether the benefits be health care, family allowances, or minimum flat-rate pensions. Solutions to the problem of persons not currently covered or inadequately covered by social insurance programs normally require a greater element of tax financing. This has been the trend in many countries.

The cost of social security rose substantially in the period after World War II both in real terms and as a proportion of rising gross domestic product. While social security spending amounted to less than 10 percent of the gross national product in nearly all countries in 1950, it had risen to 20 to 30 percent or more in many European countries by 1980. Among the reasons were the extension of the coverage of social security, the widening of the risks covered, the indexing of benefits, and the greater generosity of benefits, which moved up to or near 100 percent replacement of earnings for certain contingencies in some countries. But also of major importance was the maturing of pension schemes. Many of them were recast in the 1940s and '50s, and therefore it was not until the 1980s that people had had the opportunity to contribute on the new basis for all or most of their working lives and thus could draw pensions approaching or reaching the maximum for which these schemes provided. Three further factors were the increasing proportion of aged persons in the population, the decline in pension ages, and the lower proportion of working population.

The costs of health care also rose sharply after World War II. Several reasons contributed to this trend. First, the higher proportion of elderly in the population influenced health care costs as well as the costs of cash benefits. Persons over pension age require two to three times more health care than persons of working age, and the difference is still greater for those over 75, the fastest growing age group. A second factor was the decline in working hours, which meant that more persons (e.g., nurses) were needed in order to staff 24-hour services. A third factor was the continuous development of medical technology, such as new equipment and labour-intensive procedures. Instead of replacing labour, as in industry, innovations in health care normally required more labour for their operation. A further reason was the removal of supply restraints with the provision of more doctors and dentists, a major growth of medical auxiliaries, and the construction of new hospitals, which were more expensive to run. A fifth reason was the financial incentives to supply more services, which underlay many of the systems of paying providers under health insurance.

The final and critical factor that destabilized the finances of social security schemes was the rapid growth of unemployment beginning in the 1970s. In those countries that included unemployment benefits in their social insurance schemes, this phenomenon created both unpredicted higher costs for benefit payments and a loss of revenue from those who were unemployed. The burdens on social assistance programs were also substantial in some countries, coming at a time when unemployed persons were no longer in a position to contribute to tax revenue.

The rapid growth of social security expenditure attracted little attention during the period of rapid economic growth up to 1973. It began to cause concern after the steep rise in oil prices checked economic growth in oil-importing countries. The revenue that financed social security ceased to be buoyant at the same time as new major demands were made on the system. From the late 1970s there was talk of a crisis in social security financing.

By 1980 social security expenditure amounted to 32 percent of the gross national product in Sweden, between 25 and 30 percent in Belgium, Denmark, France, and The Netherlands, and between 20 and 25 percent in Austria, West Germany, Ireland, Luxembourg, and Norway. These figures were much higher than for Australia (12 percent), Canada (15 percent), Japan (11 percent), New Zealand (14 percent), the United States (13 percent), or the United Kingdom (18 percent). The cost was much lower in developing countries. High costs are associated with high levels of social security benefits and also with costly systems of providing health care. Some countries, such as Sweden, have allowed health care costs to continue to rise because of the capacity of this service sector of the economy to provide further jobs and thus avoid high rates of unemployment.

The aim in many industrialized market countries came to be the containment of the costs of social security. This requires that program costs not grow faster than the yield of contributions. Various devices were introduced to help secure this result. Systems of indexing benefits and pensions to prices or earnings were revised downward, or adjustments were made less frequently. Pensioners were made to pay contributions toward health-care benefits. In France tax income was brought in to supplement the yield of contributions. In the United Kingdom the earnings-related additions to short-term benefits were abolished.

A series of measures was introduced to limit the cost of health care. Charges and copayments were increased or new charges were introduced. Payment for drugs was introduced in West Germany (1977), Italy (1975), and Portugal (1982). Portugal and Luxembourg joined France and Belgium in charging for consultations with doctors. Charges for hospital care were introduced or extended in Belgium, West Germany, Portugal, and France. By 1984 there was no country in western Europe that provided free care to all its insured population.

Payment systems under health insurance were revised to reduce incentives for overservicing. The aim in West Germany was to pay the doctor more for the consultation and less for medical procedures. Payments for diagnostic tests were sharply reduced in Belgium. As part of the introduction of a national health service in Italy, payment to all general practitioners was changed from fee-for-service to capitation, and the bulk of specialists began to receive full-time or part-time salaries. Budgets for each hospital were introduced in Belgium, France, and The Netherlands, in part to discourage unnecessary retention of patients paying per day of care. Countries in which hospitals were already paid on a budget basis reduced the budgets. In the United States hospitals began to be paid under Medicare and Medicaid according to a schedule of costs for various groups of diagnoses.

Countries maintained strong controls over new hospital construction or expansions, and incentives were created in a number of countries to transfer beds from general use to the care of the long-term sick. Several countries took measures to develop alternatives to hospital care, such as outpatient surgery, outpatient hospitals, nursing homes, residential homes, and home care by domiciliary teams. The United Kingdom closed some 400 hospitals over a period of 10 years. Restrictions on the installation of major new medical equipment went into effect in Belgium and France. By 1955, 10 of the 12 countries of the European Economic Community had instituted quotas for medical schools. In Denmark, France, Ireland, Portugal, and Spain the number of medical students was cut substantially.

Most countries in western Europe introduced restrictions as to what medications a doctor could prescribe under the health service or health insurance system. Most of these countries exercised tight control over pharmaceutical prices and pharmacists' margins. New measures were introduced in the effort to control overprescribing.

Social security spending tends to vary between countries in direct proportion to their respective standards of living; in other words, the more affluent a country is, the more it is likely to spend on social security. Spending also tends to vary according to the proportion of elderly people in the population. Third, it varies according to the year in which the first legislation was adopted: countries with older social security programs tend to spend more. There are, of course, exceptions to this pattern. For example, the United States and Japan are low spenders both for their standard of living and for their proportion of elderly, and New Zealand is a low spender for a country that introduced pensions as early as the end of the 19th century.

This type of analysis has been criticized, however, for ignoring private arrangements, particularly employers' provisions established as part of collective bargaining. Thus, for example, the large role of fringe benefits in Japan helps to explain the relative lack of development of statutory social security. Similarly, the large role of occupational pensions and health insurance negotiated between employers and employees helps to explain the underdevelopment of statutory social security in the United States. Hence it is argued that private and public social security must both be taken into account in any comparison of national programs. In federal countries such as Australia, Canada, Switzerland, and the United States there were constitutional obstacles to adopting social security that led to the private sector's playing a larger role.

Political orientation also plays a role in explaining the extent to which social security has been developed in the public sector. After some initial opposition, political parties drawing substantial support from the working classes and the trade unions have promoted the expansion of social security. This includes European Catholic Workers' movements. Extensions of social security may be introduced by coalition governments with a conservative majority as the price needed to keep the coalition together. The high spending in Scandinavia can be explained by the strong influence of social democratic parties in the period following World War II. Trade unions have had less influence in this direction in Australia and New Zealand. The absence of a working-class party in the United States is part of the explanation of the relative underdevelopment of its social security program.

Some of the trends leading to increased costs are bound to continue. While the number of aged persons in most highly industrialized societies is likely to stabilize during the later years of the 20th century and the early years of the 21st century, the proportion within it of those over 75 will continue to increase substantially. This has major implications for further increases in the cost of health care. Moreover, pension schemes are still maturing, and there are pressures for further improvements of benefits, particularly to provide sex equality, lower pension ages, and better assistance for persons, particularly women, inadequately provided for previously. On top of all this, costly developments in medical technology continue. If the trend to shorter working hours continues this will also have a further major impact on the cost of health services.

Looking further ahead, the proportion of aged in the population is expected to start to increase substantially in the second and third decades of the 21st century as the increase in births after World War II becomes reflected in an increase in pensioners. It is this prospect that has led the United States to plan for increases in pension ages and the United Kingdom to decide to scale down its second tier earnings-related pension scheme.

The level of contributions and taxes needed to sustain present plans for social security cannot be predicted. While the continuing trend toward a higher number of aged in the population can be safely predicted, the birthrate is much harder to forecast. Of vital importance is the level of unemployment because of its impact on both sides of the balance sheet; reduced unemployment would add to contributions and tax income as well as lower the cost of benefits. Nevertheless, the prospect of a substantial increase in pensioners in the 21st century has led to fears in some quarters that the “compact between generations” may not perpetually be honoured. Hence it is argued that the pay-as-you-go method of pension contribution should be replaced by the capitalization method used in early pension schemes and in the private sector. Alternatively it is argued that the privatization of social security pensions would lead to higher savings and investment out of which future pensions could be paid. The disadvantages of either of these approaches are that there would need to be an immediate increase in contributions to provide the planned level of pensions. This could lead to pressure for higher cash earnings. Moreover, the level of pensions would no longer be indexed but would depend on the yield of investments.

It has been argued that the high cost of social security is in part responsible for the low levels of economic growth in industrialized societies since 1973. The argument takes three forms. First, it is said that high levels of unemployment benefits reduce the incentives to take paid work. Second, resistance to the payment of taxes and contributions leads to wage demands, inflation, and government deficits. Third, it is argued that because people have rights to social security benefits they are less likely to save; this lowers investment and thus economic growth. For all these reasons social security is said to have contributed to or even to have been responsible for not only low growth but also for high levels of unemployment.

In response to these criticisms it has been pointed out that empirical investigations lend little support to the contention that people prefer benefits to work, though the availability of benefits may make them less willing to take low-paying jobs. Second, it is argued that tax resistance would apply whether pensions were provided in the private or in the public sector. Indeed, if pensions were provided in the private sector they would have to be capitalized, which would require higher contributions and therefore lower cash earnings, leading to still greater pressures for higher pay. Third, the evidence that social security reduces savings is by no means conclusive; indeed, in many countries there has been a boom in the variety of different forms of saving following the establishment of pay-as-you-go systems of financing social security. Moreover, investment is limited by the availability of profitable investment opportunities rather than by any shortage of savings. And if low savings does limit investment, governments can generate a budget surplus out of which investment can be financed.

Some critics argue that social insurance benefits should be replaced by a negative income tax. As countries get richer, it is argued, an increasing proportion of the population is in a position to take out private insurance against the risks for which social security provides. If social security were concentrated exclusively on the lower income groups, provision could be more generous and the burden of public provision could be reduced. The administrative and other problems of using annual tax returns as the basis of making cash payments to individuals whose circumstances are constantly changing as they go in and out of employment, marriage, and cohabitation are considerable. But in the context of saving, it was because income-tested pensions were thought to be damaging to thrift that many social insurance programs were established in the first place. Low earners are unlikely to save if the yield of their savings leads to a dollar-for-dollar reduction in pension. Moreover, many countries in Europe already have income-related housing allowance schemes that serve much the same function even though they are separate from income taxes. Where this is the case, there is no room for a further negative income tax or other income-tested scheme without imposing extremely high tax rates on increased earnings. Most important of all, it is by no means clear that the economically securer members of the population would be willing to accept anything like the existing level of contributions and taxes if they stood to gain nothing from social security. As a result, provision for the poor might be no better or, more probably, it might even be worse than it is as part of a scheme to which all contribute and on which all are in a position to make claims. Historically, services for the poor have always tended to be poor services.

Empirical studies have shown some small association between higher levels of social security spending and lower rates of economic growth. But it is not clear that one necessarily causes the other. Many other factors are at work. Countries that had a high proportion of the population in agriculture were in a favourable position to achieve high growth in the postwar period as their agricultural populations declined. Moreover, countries that have had relatively low rates of economic growth, such as the United Kingdom and the United States, are relatively low spenders on social security, while countries that have had high rates of economic growth, such as Belgium, Denmark, and The Netherlands, are high social security spenders.

Critics of social security are not confined to those concerned about its effects on the economy or about personal freedom regarding the extent to which and methods by which individuals provide for their security. There are also those concerned about the “target inefficiency” of social security, or its limited redistribution to the poor. This attack is generally directed at earnings-related benefits. Proposals have been made to use the yield of social security contributions, supplemented by taxes, to provide everyone with a minimum income on which they could live at a modest level supplemented by earnings if they wished to take paid work. Social dividend schemes of this kind are seen not only as a way of redistributing income but also of reducing unemployment. A major problem with such schemes is the high level of contributions and taxes needed to finance the minimum income if it is to go to those with jobs as well as to those without them and be sufficient to live on, if only at a minimum level.

It is true, however, that high spending on social security has failed to solve the problem of relative poverty in industrialized societies. Yet abolishing poverty was never the original intention, or at least it was not in many countries. Social security was seen as a system of maintaining income by redistribution from the well to the sick, from the young to the old, and from those with jobs to those without them. This involves substantial redistribution but not necessarily redistribution from rich to poor. For instance, while there is more illness and unemployment among low earners, higher earners tend to live longer and thus draw pensions for a longer period.

A study financed by the European Economic Community showed the extent to which poverty remained in the Common Market countries around the year 1975. Poverty was defined as half the average level of living for each country. The study showed that the greatest success in combating poverty was achieved in The Netherlands and, second, in the United Kingdom, though Belgium and West Germany came close to the United Kingdom. Both of the first two countries have primarily flat-rate benefits. Poverty, however, was at its greatest in Ireland and Italy—both countries with substantial agricultural populations—though Ireland also relies on flat-rate benefits.

A variety of reasons explain why poverty persists in industrialized countries despite their elaborate provisions for social security; the precise reasons and their relative importance differ from one country to another. One reason may be that the arbitrarily selected poverty line is just above the level of living provided by social assistance. A second may be the failure of all those entitled to benefits to claim them. A third may be that certain categories of people in some countries are not entitled to claim social assistance (e.g., the long-term unemployed). A fourth reason may be that earnings-related benefits do not secure a sufficient income for low earners to rise above the poverty line or that their record of contributions is insufficient to achieve this result. But considerable poverty may also persist among families headed by a full-time worker. This is more likely to occur in one-parent families when the earner is a female with limited skills and low earnings. But it may also occur in two-parent families with one earner and several children, where family allowances fall below the cost of maintaining a child at the minimum level or the cost of rent is considerable.

In the case of developing countries social security is criticized for reinforcing the dichotomies between urban and rural populations in general and between employed and unemployed persons in the urban sector. Social insurance contributions, which are in effect taxes specifically tied to providing benefits to the members of the schemes, cannot be used by governments for the benefit of the community as a whole. This limits the capacity of governments to raise tax revenues for broader purposes. Moreover, different health benefits as well as cash benefits may be provided for different occupational groups, thus perpetuating and accentuating inequalities. Those who defend social security argue in reply that requiring part of earnings to be put into an insurance fund is robbing no one outside the scheme, since only by providing the benefits could the particular taxes be justified and compliance with paying them be secured. Criticisms regarding inequality should be directed at the pattern of original earnings, not at social security, which mobilizes part of them for good purposes.