World Labour Report overview

Income security is of long-standing concern to the ILO. The ideas expressed 75 years ago in one of its earliest publications remain entirely relevant to income security now, although they would be worded differently. In parts of the developing world many working people and their families live in extreme poverty. In all developing countries considered together, nearly one-third of the population lives on less than US$1 a day,30 per cent of adults are illiterate, 30 per cent have no access to safe drinking water, 60 per cent have no access to sanitation, and 30 per cent of children under the age of five are underweight.² And even in some middle-income countries an unequal distribution of income leaves many workers and their families extremely poor. Poverty is the absence of material necessities. It is associated with low health status and inadequate public health services, insufficient funds to obtain health care, low expectancy of life, minimal or no education, illiteracy, bad housing, dirty water and sanitation, and lack of public infrastructure. Contingencies such as ill health and old age which those in better circumstances are able to cope with become magnified and potentially catastrophic to the poor. In particular, poverty is frequently associated with an absence of political and democratic empowerment and a voice in community affairs. Ultimately, these features can condemn workers and their families to a continuation of poverty into the next generation and beyond.

Poverty is not distributed evenly or at random. Some groups of the population are more likely to be poor than others. Children, women and the elderly are more exposed to the risk of poverty than adult men of working age. And workers in the informal sector of an economy — those without regular jobs — are more

likely to be poor than workers in the formal sector. Poverty is particularly relevant to children since it conditions their experience during their formative years and sets them on a path from which they might never escape. Child poverty is, in human terms, the bottom line of income insecurity. This report shows that, even in rich countries, millions experience material deprivation in childhood. And, primarily because of poverty, millions of others, particularly in the developing world, are drawn into child labour. Such a wretched start in life means insecurity in adult life for all too many. For women poverty means excessive dependence on a male breadwinner (who himself may be poor) or else an individual life of extreme hardship, all the more so if the woman is also the head of a household which includes children. Those who are growing old, even if they are marginally above the poverty line while they are still working, face an impoverished old age during which they must depend on their children and may impose a burden on them — made more difficult if the children themselves have had to move to a distant city in search of a job.

Income security is not only about the adequacy, but also about the regularity of income. It is a subject that is of concern not only to the poor, or to those who are on the verge of poverty, but to all who earn an income from their work, whether they are wage or salary earners, derive an income from self-employed activities, or any of the variants which are to be found in the rural and urban informal sectors, or are simply destitute, often as a result of an incapacity. But fluctuations in income bear particularly hard on those who previously had very little income.

It is easy to obtain agreement that these problems must be remedied. It is less easy to reach a consensus on the measures to be taken. High levels of — productive, well paid and secure — employment are clearly a major part of the answer, and the promotion of employment is clearly a main thrust of the ILO’s objectives. Achieving this goal was a central theme of the Director-General’s report to the 1999 International Labour Conference, Decent work. This will not only generate income but will lead to greater solidarity between income groups and a fairer deal for the poor in those societies where poverty is, at least in part, a consequence of the distribution of income.

This report describes the measures of social protection which can be put in place both to alleviate poverty and to minimize the risk that, through no fault of their own, individuals fall into poverty. Better education and training are crucial — a subject which was covered extensively in the World Employment Report 1998-99 and is therefore not dealt with in the present report.

Income security is derived from many different sources. In developed countries the proportion of GDP which is transferred to those who need it has grown rapidly over the past 50 years and now (including expenditure on health care) absorbs almost 25 per cent of GDP. It has been enormously successful in alleviating the poverty which might otherwise have been associated with the risks of ill health, old age, disability and unemployment. It has been associated with largescale proactive public policies of training and education which have encouraged

individuals to regain employment once they have lost it. It has coincided with a massive entry of women into employment, and hence into their own individual income entitlement. It has been based on mandatory social security contributions as well as general taxes, for both of which there exists a strong consensus. And it has been matched by private schemes and personal provisions which have supplemented the public ones. The health status of the population has greatly improved. Life expectancy has increased. Gender inequalities have been reduced. Poverty in old age has largely disappeared in many countries. Unemployment is no longer synonymous with poverty. Invalidity is cared for.

For the most part these developments rest on the basis not only of economic growth but on good governance, full — or nearly full — employment, well-regulated labour markets, a reasonably equitable distribution of incomes, high levels of compliance with the laws concerning both taxes and social security contributions, and almost complete coverage of the social security programmes, as far as both benefits and contributions are concerned.

No such success story can be applied to developing countries nor to many middle-income countries. A large segment of their labour force works in the informal sector and many workers are underemployed rather than unemployed. In contrast to earlier expectations, the scale of the informal sector has grown rather than diminished. Mechanisms of governance are weak and compliance with either tax or social security laws low, reducing the base of tax revenues or social security contributions. The distribution of income is highly skewed, making it difficult to achieve a consensus on redistribution. Gender inequalities remain pervasive, and the empowerment of the poor, in spite of recent moves towards democratization, is negligible. As a result many developing countries possess neither the tax base with which to finance social assistance schemes nor the collection base with which to finance contributory ones. Hopes that in due course the developed country models might spread to the developing world have been disappointed and long delayed.

This dilemma must be addressed; its resolution can no longer wait upon the general drift of economic growth, the development of formal employment, a more equitable distribution of income or greater democracy. As far as possible, the limited social security institutions to be found in developing and middle-income countries must be assisted to improve their management and administration, their benefit levels and, above all, their coverage of the population.

Another possibility considered in this report is the development of self-supporting, self-administered schemes among the population of the informal sector. Such schemes — essentially micro-insurance schemes — can offer basic social protection in return for small contributions, especially important in the area of health care services. The schemes are based on the principles of solidarity and risk-pooling. The members themselves define their own set of principal needs and participate in the management of the scheme. Such micro-insurance structures have the advantages of cohesion, direct participation and low administrative costs. But there are also many problems associated with them which have yet to

be tackled. These include the difficulties of providing benefits to beneficiaries whose contributions are voluntary rather than mandatory. There are technical problems associated with the administration and management of such institutions. And the institutions must rely on a collective and moral policing of contributors and beneficiaries. Nevertheless, and in spite of these issues, the human dependency on the results is too great for an effort not to be made. Ultimately, what is sought is a fusion between formal and informal social security mechanisms, which will enable social protection to be spread widely and generously across the population.

Part I of the report examines the extent of income insecurity and the changing context to which income security policies are having to respond. Demographic, family and labour market structures (Chapter 2) all have decisive effects on income security and are undergoing enormous changes, many of which are interrelated, such as increasing participation of women in formal employment, declining birth rates, reduced capacity of the extended family to provide the care necessary for the young and the elderly, and so on. Not only do changes in these areas have critical implications for social needs, but employment and demographic trends will greatly affect the adequacy of society’s response. Social security expenditure and the economy (Chapter 3) constitute a crucial part of the changing context. The trends in expenditure as a percentage of GDP are described and projections are presented. An analysis is made of the effects which social protection is often alleged to have on economic performance.

Part III looks to future needs and prospects. Of greatest concern is the need to extend the coverage of social protection to the majority of the world’s population who still have none (Chapter 10). Social protection systems face many other challenges which call for restructuring and innovation (Chapter 11). These include: the evolution of gender roles and the demand for gender equality; the changing nature of the labour market and the need to ensure an appropriate combination of flexibility and security; and the public demand for democratically managed schemes upon which they can rely for courteous and efficient service. Finally, without in any way producing a blueprint, the report proposes some broad policy conclusions (Chapter 12) about the way in which social protection systems should develop in the twenty-first century.

The changing context

Despite the increasing general levels of prosperity observed in most countries of the world, many changes are taking place which are affecting income security in adverse ways. These include changes in demographic and family structures and in labour markets.

The extended family, the traditional mainstay of income security for the majority of people in the developing world, is becoming smaller and more dispersed. In major developing countries about 20 per cent or more of families are headed by a woman with no spouse present. Such developments have major implications for income security and call for policy responses in the areas of social security, social services and employment. These responses have in many cases been inadequate, partly because of disapproval of divorce and single parenthood based on traditional or religious values. However, it is increasingly understood that the welfare of the children matters most and that a judgemental stance is highly prejudicial as far as they are concerned. Even taking into account the social security benefits they receive, households headed by a lone mother are three times more likely to be in poverty than two-parent households in major industrialized countries. The Nordic countries have been much more successful than others in avoiding this problem, essentially through the provision of good childcare facilities and employment opportunities. Recent years have seen the emergence of high child poverty rates in two or three major advanced economies, related no doubt to economic and social policies as well as to changes in family structure (figure 1).

In most parts of the world the population is ageing rapidly, partly because of longer life expectancy, but above all because of falling fertility rates. Attention is invariably focused on the increases which this will bring in the old-age dependency ratio. What is less commonly realized is that total dependency ratios will fall in the coming decades in the developing world, and fall quite steeply in the least developed countries. This is an invaluable opportunity to establish universal primary education for the much smaller cohorts of children reaching school age and to improve greatly the quality of the education provided — which is the best hope of achieving income security in these countries.

The rise in the old-age dependency ratio naturally has longterm implications for pension schemes. However, the financial position of these schemes depends not just on demographic factors but on other variables, notably labour force participation rates and coverage rates, which directly affect the number of active contributors. Labour force participation of women is projected to increase and there is enormous scope for increased participation rates among older workers, following a long period during which many millions have been pushed into premature retirement in their fifties. As disability-free life expectancy rises, people will have the capacity to remain active in the labour force longer, particularly in the non-manual jobs which occupy a larger and larger proportion of the labour force. Older people have much to contribute to society, provided that they are given the opportunity. The real problem is not demographic but economic in nature: how to ensure employment for all who are willing and able to work.

One-third of the world’s labour force is either unemployed or underemployed. Since the mid-1970s unemployment rates in the advanced economies have more than doubled and high rates have been registered in many of the developing countries for which meaningful figures are available. The increase in unemployment rates understates the problem almost everywhere, because it takes no

account of the many workers who ultimately give up looking for work and who leave the labour force either temporarily or permanently. In certain countries, especially among the economies in transition, low unemployment rates mask the reality of a situation where millions of workers remain attached to enterprises, but do little or no work and receive no pay. In most countries youth unemployment rates are two to three times higher than adult rates. Long-term unemployment — which involves loss of skills, motivation and productive potential — now accounts for a much larger proportion of total unemployment than it did in the 1970s. This has happened in spite of the fact that, both for the young and for the long-term unemployed, benefits are either low or non-existent.

The structure of employment has undergone changes which are of significance for income security, but which are by no means uniform across countries. The past two decades have seen a continuation, in numerous countries, of the secular decline in the proportion of the labour force which is self-employed. But in others the decline has stalled or been reversed, most spectacularly in the economies in transition.

Some of the new self-employed are wage earners under another guise. This kind of self-employment — and indeed much informal sector employment — is conceived either primarily or in part as a means to avoid deductions at source (taxes and social security contributions) which usually apply to wages and salaries. Workers have little choice but to accept work on a self-employed or even clandestine basis, so long as there is an oversupply of the labour in question. They are often not covered by social security; they are not subject to employment protection legislation and usually have no job security whatsoever; they are not covered by occupational health and safety legislation where this applies only to employees; they do not qualify for coverage under any employer or enterprise pension or health scheme; and their contractual status usually means that they cannot be represented by a trade union but must deal individually with the enterprise for which they work.

Certain other forms of employment are also on the increase, such as part-time employment, temporary employment, casual employment, home work and teleworking. Most of these tend, in one way or another, to provide less income security than permanent, full-time employment on the premises of the employer. However, national social security systems differ significantly in the way they deal with these various phenomena, in terms both of their legislation and of the effectiveness with which that legislation is implemented.

Social protection began to play a much greater part in the economies of most countries in the last quarter of the twentieth century. Expenditure levels, expressed as a proportion of GDP, have risen in most nations, although exceptions are to be found, notably in Africa and Latin America. These percentages continue to vary widely, being many times higher in the industrialized countries, where most of the population is covered, than in the developing countries, where few are covered and benefits are much less comprehensive. Projections for the next 50 years for countries with the highest expenditure levels show that, if existing systems and policies are maintained, increased social security expenditure in the OECD countries will be equivalent to about 20 per cent of the increase in real earnings (figure 2). Projections for selected Central and Eastern European countries indicate that, by the year 2050, social security expenditure will be equal to just over 25 per cent of GDP. This is the same level as projected for the OECD countries, whereas in the 1990s the figure for the Central and Eastern European countries was one or two percentage points higher than for the OECD countries.

Most social security expenditure — with the exception of health care and social security administration — takes the form of transfers; that is, it does not use up physical and human resources, unlike defence or education or road construction. Be that as it may, the possible economic impact of social security is nevertheless a subject which arouses widespread concern. Attention is usually focused on the cost side and thus on the potentially adverse effects on the economy such as duration of unemployment. However, social protection also has positive effects on economic performance, notably through the benefits it provides. Economic performance has different dimensions: the level of employment, the level of unemployment, labour force participation (the combination of the first two), and

the level and growth of productivity, particularly labour productivity. No systematic relationships are to be found between the level of a country’s social protection expenditure and those economic indicators, when comparing countries of a roughly similar level of development.

A well developed social protection system is a necessary component of a well organized market economy. It has a positive economic role to play as a facilitator of structural change. This is the case of unemployment insurance, which China and other economies in transition, for example, have introduced in recent years. Some forms of social protection, notably employer or enterprise benefits, on the other hand, may place obstacles in the way of labour mobility, as has been observed particularly in the states formerly part of the USSR.

Globalization has significant implications for social protection. Countries with the most open economies are most exposed to the vicissitudes of global markets, and research shows that they are the countries which have the highest social security expenditure. However, some correlation has been found between the increasing share of trade in GDP in recent years and a reduction in social security expenditure — which suggests that globalization may be making it more difficult for countries to finance social protection. The increased mobility of capital has certainly made it much more difficult to tax capital, so that governments are increasingly resorting to taxes on labour and on consumption which may meet with greater resistance from voters.

High social security contributions or payroll taxes are a subject of intense debate. Employers take the view that these push up labour costs and thereby undermine their competitiveness with producers in other countries. Most economists tend to think that the burden of contributions, including those paid by employers, is absorbed by workers in the form of lower wages, at least in the long term. International comparisons tend to support the latter interpretation; in Denmark, for example, where payroll taxes are extremely low, total labour costs have not been any lower than in France, where payroll taxes are rather high. In the short term, however, increases in employers’ social security contributions do raise labour costs. Similarly, in the short term their reduction would no doubt reduce labour costs. Recent research suggests that social security contributions have no longrun impact on unemployment (Nickell, 1997).

The most spectacular effect of globalization has been on international capital flows; the deregulation of markets and the introduction of information technology has facilitated a huge expansion of short-term investment in emerging markets, notably by foreign pension and mutual funds. The volatility of these flows is judged by many to be a major factor in recent financial crises. The agenda for pension reform being canvassed in financial circles would imply an enormous expansion of these funds at the global level, even if only one of the major industrialized countries were to adopt the proposal to convert to an advance funded system.

Existing mechanisms of social protection

Health care systems affect income security in two ways. The availability of adequate preventive and curative care is vital to ensure that workers are fit to earn a living for themselves and their dependants. And it is the function of the health care financing system to ensure that the large and unpredictable costs of health care do not fall directly on individual households, with the catastrophic effects which that would have on their budgets. Health care is equally important for everyone, regardless of labour force participation or employment status. However, many workers, particularly those who are outside regular wage employment, do not have any satisfactory health care coverage, at any rate in many developing countries. Their situation has become particularly difficult since numerous governments have been obliged to implement structural adjustment programmes which have led to sharply reduced expenditure on public health services. Higher user charges have significantly raised the barriers for many people with low health status and low incomes. The countries worse affected are generally those where social security schemes also have the least extensive coverage.

Micro-insurance schemes for health care are being developed in various parts of the world to bring the advantages of risk-pooling to people who are otherwise unprotected. However, the extent to which these schemes succeed in pooling risks depends on the size of the groups covered and the extent to which they are able to join up with each other in networks. The development of these schemes is highly dependent at the present time on support from governments, NGOs and international organizations such as the ILO. They contribute, through popular participation in their setting up and in their management, to the social and political inclusion of the excluded.

Social protection during incapacity is relevant to all who work for their living, not only wage earners but also self-employed persons, many of whom have few financial resources to fall back on if sickness or disability prevents them from working. Employees in large companies and in public organizations typically receive their full salary during periods of sickness, often by virtue of agreements negotiated by trade unions and employers. However, growing labour market flexibility is leading, in certain industrialized countries, to more and more workers not being covered even by social security sickness benefits (which may be just 50 per cent, or less, of their normal earnings). And the majority of the labour force in many developing countries remains without any social security coverage for either short-term or long-term incapacity.

Income replacement benefits are vital, but they are far from being sufficient. The ultimate objective is to restore the worker’s health and earning capacity. This requires good occupational health services and rehabilitation systems. These are still seriously underdeveloped in most countries, even in the industrialized world. The growing numbers of workers who are retiring prematurely on the grounds of disability in some countries strongly suggest that there is an urgent need to place much greater emphasis on these services and systems — and of course on the

worker’s reintegration into employment. However, the impact of unemployment on the numbers receiving disability benefits has been amply documented and points to the problems that arise when workers are poorly protected against the first contingency and well protected against the second. This is yet another illustration of the way in which social security systems are undermined by unemployment.

Old-age and survivors’ pensionsare almost invariably the costliest element in a country’s social protection system. The only exceptions tend to be developing countries where pension systems have been established only very recently and where health care is still a larger item. The low rate of social security coverage in many developing countries means that most of their elderly population receive no pensions. A small number of these countries now provide a minimum income for their elderly. The situation in the industrialized countries was transformed in the course of the twentieth century by the success of social security schemes in almost eliminating old age as a cause of poverty. However, insecurity in old age is still a problem in some of these countries for certain groups of the population, particularly women. Divorce, which has rapidly increased in recent years in many countries, has left many women with little provision for their old age if they have been either not active in the labour force or else engaged in some form of employment not providing them with any pension entitlements.

Formal pension systems in the developing world are sometimes financed out of general government revenue, but in these cases the benefits are flatrate and very low. Such pensions are typically subject to a means or income test, though there are one or two examples of universal schemes. Given the weakness of government finances in most countries, pensions are financed from contributions payable by the insured themselves and, where relevant, by their employers. These pensions tend of course to be much higher and are almost always related to the individual’s previous earnings and contributions. In the absence of wide pension coverage most people in developing countries have been dependent in their old age on their grown-up children. However, due to economic, demographic and social changes, these traditional arrangements are no longer as adequate as they once were. This is particularly the case for those who have no (surviving) children.

Contributory schemes take various forms. The most common is social insurance, which permits the pooling of various types of risks, as well as a degree of vertical income redistribution. Mandatory retirement savings schemes, on the other hand, like provident funds, leave most of the risks for the individual pensioner to bear and they ensure that inequality in old age is (at least) as great as during working life. Mandatory retirement savings are attractive to those who wish to reduce or contain pension costs without having to spell out what this will mean in future in terms of lower pensions. The unpredictability of pensions under such schemes, however, does not recommend them to workers and pensioners. The costs of transition from a pay-as-you-go or partially funded social insurance scheme to a fully funded mandatory retirement savings system make it a

prohibitively expensive proposition for most governments. The latter, in any case, may be unwilling to assume the risk of underwriting a system which could leave many pensioners with inadequate benefits if real rates of return on retirement savings are lower than expected.

Social insurance is an appropriate technique for financing pension schemes in developing countries, as contribution rates can be set initially at low levels and increased only gradually over an extended period, in order to avoid any hardship for low-income earners and to spare employers any big increase in labour costs. Social insurance also allows pensions to start being paid within a relatively short period, under transitional provisions for workers over a certain age at the time of the scheme’s introduction. It is essential to clarify the long-term financial implications of these measures through rigorous actuarial projections.

Governments have a responsibility not only to ensure that workers are covered by a solid, compulsory pension system, but also to provide an enabling environment for the development of additional voluntary retirement provision. Collective bargaining has proved to be an effective way of establishing dependable private schemes and joint management of these schemes by workers and employers can provide better safeguards than even the most complex government regulations.

Unemployment benefitsexist primarily in the industrialized countries and in a certain number of middle-income developing countries, although they are also to be found in China and Mongolia. It is estimated that, of the 150 million unemployed in the world, not more than onequarter receive some kind of unemployment benefit. This reflects not only the absence of such schemes in many countries but also the fact that, where they do exist, many workers are not covered. Furthermore, many others have not contributed long enough to be entitled to benefit or else have already exhausted their benefit entitlement. Unemployment benefit, unlike many other social security benefits, is not payable throughout the contingency, but only for a limited duration which tends to be determined largely by financial considerations.

The proportion of unemployed workers receiving benefits from unemployment benefit schemes in Latin America and the Caribbean is very low, because of

the restrictive rules governing coverage and entitlement. There is now more emphasis on developing employment services in a number of these countries. The Asian financial crisis has accelerated the development of the Republic of Korea’s unemployment insurance scheme and prompted certain other governments in the region to give more active consideration to the advantages of having such a scheme. In the less developed countries, notably in South Asia and in Africa, employment-intensive programmes have served to provide a limited form of unemployment protection. These programmes provide some income security to workers who choose to participate, while avoiding the costly and cumbersome process of means-testing associated with social assistance.

Social benefits for parents and children were introduced in order to help families in general to cope with the greater costs which they face as a result of having children. They also have an important role to play in promoting gender equality. Over the years, some countries have modified their systems by subjecting some family benefits to a means test, usually in order to economize on expenditure. In addition, new systems, such as the earned income tax credit in the United States, have been developed in recent years to subsidize low wages paid to workers with family responsibilities. Family benefit systems have tended to be relatively neglected in recent decades and to see their share of social security expenditure unduly eroded. Developing countries have in many cases been reluctant to introduce family benefits, in the belief that these would aggravate the problem of high fertility (a belief for which there is little real supporting evidence). However, there are indications that benefits for children can greatly contribute to income security and can help to eliminate the problem of child labour, particularly if they are tied to school attendance. They are normally paid to the mother as the parent most directly concerned with the care of the children, which improves the intra-family income distribution and helps to promote gender equality. Greater provision of childcare services has also contributed to this objective by giving mothers greater autonomy and opportunity to enter the labour market.

Maternity protection is an important aspect of income security for working women and one which the ILO is currently in the process of strengthening by the revision of the international labour Convention on the subject. The provision of maternity benefits through social security ensures that costs are shared between women and men workers and their respective employers. Making maternity benefits an employer’s liability, as in certain developing countries, fails to ensure this solidarity and can lead to discrimination in recruitment against women of child-bearing age. Parental benefits have been introduced by a number of countries, in order to allow either parent to take time off work to care for a child in the early months of its life. The guiding principle is that parents should have the right to choose who should receive the benefit. For those who wish to share, an option is usually available which allows each parent to receive part of the benefit. The introduction of this option has helped to foster greater gender equality.

Social assistance provides benefits to people in need who receive either no benefits at all or else inadequate benefits from other forms of social protection. It is often lack of social insurance that generates the demand for social assistance. However, relatively few of the developing countries give much priority to social assistance schemes. If they do have such schemes, then usually they are restricted to limited categories of the population and provide very low benefits. There are some exceptions, such as the relatively high assistance pensions in Argentina, Brazil and Uruguay.

Social assistance benefits are normally subject to an income test, but as it is difficult, especially in developing countries, to obtain information on the income of a claimant, other methods of targeting may be employed, using other indicators (such as ownership of certain items) or a process of self-selection (as when work is involved or when benefit is provided in a form that will be unattractive to the non-poor). The extent to which people are dependent on social assistance is (in industrialized countries) in some degree an indicator of inadequacies elsewhere in a country’s social protection system. Social assistance will always be necessary for crisis situations and in order to plug gaps, but good social policy aims to prevent the gaps from opening up in the first place.

Future needs and prospects

Extending personal coverage is probably the greatest challenge facing social protection systems. In many developing countries the proportion of the labour force in formal sector employment is small and, recently, has even fallen, partly as a result of structural adjustment programmes. If they opt for social security coverage, the self-employed and others who have no easily identifiable employer have to pay the full contribution themselves and, generally speaking, feel unable to afford this. To include such people in compulsory social security coverage may therefore be feasible only through the creation of special schemes providing a more restricted package of benefits in return for a lower contribution.

In most developing countries the existing social insurance schemes apply to less than half of all employees, because coverage has not yet been extended to those in small enterprises, which account for a relatively large percentage of total employment. Valid administrative reasons lay behind the original decision to cover only enterprises employing more than 10 or 20 workers. However, in many cases inertia and vested interests have also come into play. The principle is that every employee should be covered “as soon as the collection of contributions ... can be organized and the necessary arrangements can be made for the administration of benefit” (Income Security Recommendation, 1944 (No. 67), Paragraph 20). Some countries in the developing world have shown that this is feasible under certain circumstances, given the political commitment and the administrative resources. Whenever possible, legislation should lay down a phased plan for extending coverage, giving due notice to all concerned and permitting the social security administration to take all the necessary preparatory measures.

Self-employed and informal sector workers who are not covered by social security schemes are organizing themselves in some cases in micro-insurance schemes, to help themselves and their families cope with increasingly burdensome user charges for health care. In this way, they aim to avoid the problem of indebtedness, which can otherwise be brought about by high medical bills. Through such schemes they can also succeed, as a group, in obtaining better value for money from health care providers than they ever could as individual consumers. Such schemes cover only very small numbers of the poor and are not available in most areas. These schemes need to be analysed and evaluated to assess their development potential. By and large, this implies a fresh approach in the field of technical cooperation. Support can be provided by governments and by social security agencies to create the necessary legal and administrative structure. These schemes should not be viewed as a substitute for social security schemes but, within the national social protection system, as a complement to extend coverage to the excluded. They are not a way for the State to escape its responsibility in the field of security but a means to establish a new partnership between governments, public institutions and civil society to enhance income security for all.

In the light of the problems discussed in this report, the restructuring of social protection systems is on the agenda in many countries. First and most obvious is the need to improve the scope and level of social protection. Efforts in this direction must take account of changing patterns of employment and of changing gender roles and family structures. Special schemes — or, in some cases, adaptations to existing schemes — will be necessary to facilitate the extension of personal coverage. Numerous changes are required in order to eliminate the barriers to gender equality within social protection systems, including occupational schemes.

Certain trends may be observed in the way that resources are being allocated within national social protection systems. In various countries an increasing reliance is being placed on social assistance. Another trend, visible in many countries, is the greater emphasis that is being put on employer-based or private provision. This bifurcated restructuring of social protection is tending to cement divisions in society — between the poor and the non-poor — which decades of solidaristic social security had helped to break down.

Social security systems themselves have come under close scrutiny and measures are being taken to make them work better. Governments are trying to improve strategic planning of social security and indeed of social protection as a whole. Better institutional arrangements are being made in order to increase democratic accountability and to prevent failures of governance such as have occurred in a number of developing countries. Finally, improvements in administration at the operational level are being made in order to improve compliance and enforcement and to avoid excessive administrative costs.

Main policy conclusions

The main policy conclusions identified by the report are concerned with:

extending the coverage of social protection. This will involve the extension of existing schemes to cover currently excluded employees, with whatever adaptations may be necessary for specific groups such as domestic employees. It will also involve the development of special schemes for the self-employed and for those working in the informal sector without any identifiable employer. Wherever possible these schemes should be compulsory. Where this is not feasible, as in low-income developing countries, support should be given to micro-insurance and other grass-roots initiatives which can provide some form of risk-pooling. Appropriate social assistance measures should be developed for the most vulnerable groups outside the labour force;

the need for good governance. Not only is good system design vital, but close attention has to be paid to the establishment of suitable institutional arrangements and to the efficient administration of social protection schemes. At the design stage, enough time has to be allowed for research and planning. In determining the institutional structure, it is essential to bear in mind that schemes almost invariably work better if the workers whom they serve participate in running them. Finally there is no substitute for well- trained and motivated staff equipped with the systems and powers necessary to collect and record contributions and to calculate and pay benefits accurately and on time;

the link between social protection and gender. Improved income security for women presupposes greater equality between women and men both in the home and in the labour market, with improved access of women to paid work. Practical measures are required to help men and women combine paid employment and caring responsibilities within the family. Compulsory social security coverage needs to be extended to the special categories of employment in which women are heavily represented, notably part-time and home work. Conversely, schemes which directly or indirectly discriminate against women should be eliminated or at any rate accorded a reduced role. When gender equality leads to cutbacks in benefits for women, it is indispensable to have a careful and gradual transition process;

affordability and the positive economic effects of social protection. The level of social security expenditure deemed to be affordable differs widely, even between countries with similar income levels. Affordability is a subjective matter, influenced by political culture and traditions but also by the distribution of income and the design of the social protection system. Improving the income distribution, which is particularly skewed in many developing countries (figure 3), is something which can be achieved in the long term, for example, through better education and training. Thus, these measures not