Background

The Inter-Regional Inequality Facility ran from 2004-2006.

In 2005, the international community focused its attention on progress towards the Millennium Development Goals (MDGs). These are the set of eight goals for freeing the world of extreme poverty and deprivation, achieving gender equality and empowering women, and establishing a new international partnership for global development. Recent progress toward the various targets associated with these goals has been disappointingly slow (UN Millennium Project 2005: 13-27).

In debates about achieving the MDGs, the issue of inequality has featured prominently. On the one hand, several studies have argued that some of the targets associated with the MDGs will not be met unless attempts are made to reduce inequality. To give one example, the rates of economic growth required to halve $1-a-day poverty by 2015 in Sub-Saharan Africa, assuming inequality does not change, are generally considered to be too high to be achievable (e.g. Hanmer and Naschold 2000; Clemens et al. 2004). However, if inequality can be reduced, the poverty target can be met at much lower, and more achievable, rates of growth.

On the other hand, studies have also shown how progress towards the MDGs at the national level often masks highly unequal rates of progress within countries (e.g. UNDP 2005). This is not necessarily against the ‘letter' of the MDGs, which overwhelmingly include national-level targets. Nevertheless, it is arguably against the ‘spirit' of the MDGs, according to which – in the words of the 2005 Human Development Report – “progress … should be for all people, regardless of their household income, their gender or their location” (UNDP 2005:5).

Key issues and questions

Inequality is therefore a key issue in debates about achieving the MDGs. Nevertheless, there are still important issues which need to be resolved before one can design sensible and effective policies for addressing inequality. These issues are highlighted by three main questions:

• which inequalities matter?
• what drives inequalities, and how can they be affected by policy?
• which specific policy instruments work best where?

Which inequalities matter?

Like poverty, inequality can exist in a range of dimensions (e.g. income, life expectancy, mortality, education). One needs to first ask which inequalities matter, and of those that matter, which matter most.

One argument is that income inequality matters since it is often possible, at least in theory, to reduce absolute poverty by reducing income inequality. This need not always be the case, not least because in a context of low average income, a reduction of inequality could conceivably place more people below the poverty line. Nevertheless, evidence suggests that reductions in inequality – or more accurately, ‘pro-poor distributional shifts' – do often make large contributions to the amount of poverty reduction (Ravallion 2005).

Inequalities in income and assets can have a whole range of other important effects. There has been a long debate about whether a country's level of income or asset inequality affects its rate of economic growth. As yet, the empirical evidence on this particular matter is inconclusive (World Bank 2005:103). Nevertheless, there is a generally robust positive relationship between measured income inequality and several variables widely agreed to be bad for growth, including corruption, insecure property rights, low school enrolment, high fertility, low life expectancy, and macroeconomic instability (e.g. Knack and Keefer 2002, Odedokun and Round 2004, Wilkinson 2005).

Inequalities also matter for intrinsic reasons, although in this case opinions differ according to one's philosophical or moral viewpoint. A common argument is that equity or social justice requires equality of opportunities but not equality of outcomes. However, other perspectives, such as the Rawlsian theory of social justice, do contain an intrinsic aversion to inequality in welfare outcomes. Less controversial is the view, associated most notably with a human rights approach to development, that all citizens should be entitled to an equal set of civil, political and social rights, including the means to exercise these rights effectively.

What drives inequalities?

Inequalities in income and other welfare indicators reflect underlying inequalities in the ownership of assets, which are often highly persistent and change only slowly over time. This is typically attributed to the combination of two things: political forces, which make it possible for groups in power to protect their wealth, and capital market imperfections, which make it difficult for those on low incomes with low savings to accumulate capital (Li et al. 1998).

Inequalities in income and other welfare indicators also reflect the effects of external shocks and domestic policy reforms. These can have significant distributional implications, through their effects on the prices of goods and services and factors of production. Some argue that the implementation of ‘Washington Consensus' policies has contributed to a widespread trend towards higher income inequality within countries over the past two decades (e.g. Cornia 2004).

Discrimination and exclusion are increasingly being recognised as key determinants of inequalities. Research has shown that so-called ‘horizontal' inequalities (inequalities between groups defined according to gender, ethnicity, region, and so on), which typically reflect underlying discrimination, often make up a significant proportion of overall inequality (e.g. Stewart 2004; World Bank 2005: 40-43). Discrimination and exclusion are in turn often deeply embedded in cultural norms and attitudes, which are difficult to change.

This suggests that there are three main things policy-makers need to do to address inequality.

1. Lower barriers to the accumulation of assets by those on low incomes.
2. Avoid any disequalising effects associated with external shocks and domestic policy reforms.
3. Tackle discrimination and exclusion.

The balance between these three strategies will vary, depending on the relative contributions of capital market imperfections, external shocks, policy reforms and discrimination in generating and maintaining inequalities in any one country context.

Which policy instruments work best where?

Associated with each of the three broad strategies for addressing inequality are a range of more specific policy instruments. These include:

• conditional cash or in-kind transfers, subsidised education and training, support to small-scale enterprises, land reform (reducing barriers to asset accumulation);
• price subsidies for basic food items, public works programmes, unconditional cash or in-kind transfers (avoiding disequalising effects of shocks and reforms);
• equal opportunities legislation, affirmative action and public awareness campaigns (tackling discrimination).

The third and final key issue is therefore which of these more specific instruments should be used in any one country context?

Here there are two main considerations. First, most policy instruments require additional government spending. To be consistent with basic principles of public expenditure, it must be the case that the benefits of a chosen instrument exceed those which could have been gained by allocating the expenditure elsewhere (e.g. on roads or infrastructure). This does not mean that nothing can be done about inequality, but it does mean that careful attention must be paid to issues of targeting and impact.

Second, policy makers also need to take into account political economy considerations. In particular, the most ‘efficient' policy instrument(s) for addressing inequality might not be those most likely to command widespread political support. For example, universal programmes are sometimes argued to be more politically sustainable than means-tested programmes, and for this reason to provide greater benefits to poorer groups in the long-run even if they are less efficiently targeted (e.g. Besley and Kanbur 1993).

These considerations – the need to choose efficient instruments, but also those which are politically acceptable – mean that decisions as to which precise instruments should be used to address inequality tend to be highly country-specific, and there are few general prescriptions. Nevertheless, there is still a lot that policy-makers can learn from the experience of policies applied elsewhere.

Findings from Regional Synthesis Papers

During 2005 the IRIF commissioned a set of three Regional Synthesis Papers. The aim was to review the most recent evidence on inequality in Africa, Asia and Latin America, placing a particular focus on research done by institutions within each region. Here we summarise some of the key findings of these studies.

Research on inequality in Latin America is well-established and has a relatively long history. The most recent examples are the World Bank ‘Breaking with History' study (De Ferranti et al. 2004), and the ECLAC-UNDP-IPEA (2002) report on meeting the MDG poverty target. These studies emphasise the ‘extensive, pervasive and resilient' nature of inequality in the region. They also emphasise the adverse consequences of inequality on growth and on poverty reduction. There is a considerable amount of consensus that poverty in the region cannot be reduced or eradicated unless the high and persistent levels of inequality are explicitly addressed.

Research on inequality in Asia is also relatively well-established. Early studies emphasised how low and sometimes declining levels of inequality contributed to the region's undoubted success in delivering high and sustained levels of economic growth combined with poverty reduction. More recent research suggests a more diverse picture however. Significant increases in income inequalities have been witnessed, including most notably China, but also Vietnam, Pakistan and Bangladesh. Nevertheless, even in these cases, the effect has mainly been to slow already quite rapid reductions of poverty, rather than to prevent poverty reduction altogether. This has made addressing inequality seem less of a policy priority, as compared with Latin America.

Research on inequality in Africa , by contrast, is a relatively new phenomenon. In the past, the general impression has been that levels of inequality in the region were low, and that poverty reduction would come mainly from higher economic growth. This changed towards the end of the 1990s, with the increased availability of detailed household surveys in the region. These showed that levels of inequality in the region are, at least in terms of income or expenditure, as high as they are in Latin America (Table 2). They also showed that economic growth was having only a limited impact on absolute poverty in many countries in the region. Despite this, there is much less consensus, compared to Latin America, that addressing inequality is central to the policy agenda in the region.

An increasing amount of research has been carried out in each region on the different dimensions of inequality, and also in decomposing measures of inequality into their different components. In Latin America, studies have documented the large contribution of educational inequality to overall income inequality, which has important implications for policy. Studies have also demonstrated the significant differences in income, education and access to services between ethnic groups. Evidence of significant ‘horizontal' inequalities between groups is also available in Asia (e.g. India, Malaysia and Vietnam).

Findings from Policy Briefs

During 2005 the Inter-Regional Inequality Facility commissioned 15 Policy Briefs, covering 5 topics. The aim here is to provide accessible information about which policy instruments for addressing inequality work best in which country contexts. We summarise some of the main findings emerging from these studies.

Targeted cash transfer programmes were studied in South Africa, Colombia, Nicaragua and Jamaica. These aim to allow households to make costly investments in the health, education and nutrition of their members, while at the same time reducing poverty directly. Evaluations have shown generally high levels of targeting efficiency, and significant beneficial impacts on school enrolment, child health and household expenditure on food and children's clothing. Challenges remain however, in terms of increasing levels of take-up among the poorest households still further, and also in securing fiscal sustainability, since in three of the programmes (Colombia, Nicaragua and Jamaica) at least half of programme funds have come from external sources.

Social safety net programmes were studied in Indonesia and India. In Indonesia, the objective was to help those on low incomes cope with the effects of the East Asian crisis, through a targeted rice subsidy, an employment creation programme, education scholarships, and free health services. There were certain problems with the implementation of these schemes, partly because of the speed at which they were introduced (prior to the crisis, safety nets were largely informal). Nevertheless, studies have shown that households participating in the schemes were significantly less likely to experience a large reduction in income following the crisis, compared with other households. In the Indian state of Maharashtra, the Employment Guarantee Scheme aims to help people on low incomes deal with the seasonal nature of employment and other income-generating opportunities. Recent studies have shown that this scheme provides significant transfer and income stabilisation benefits to participating households, the majority of which are poor or extremely poor.

Programmes to increase access of low-income households to education or health were analysed in Uganda, Sri Lanka and India. The Ugandan programme has involved eliminating all tuition fees for primary education, while the Sri Lankan programme involves providing free school uniforms and textbooks to all children in primary and secondary education. Both have raised school enrolment and attendance among groups with the lowest initial levels of enrolment and attendance, thereby contributing to declining inequality. Challenges remain however: in Uganda for example, significant differences in resources per pupil persist between rural and urban areas. The Indian programme aims to increase use of hospitalisation services by poor households on low incomes, through the provision of health insurance. This has been successful in attracting a large number of participants, particularly when insurance is provided via ‘nodal agencies' (e.g. micro-finance institutions, state co-operative societies) which overcome the informational disadvantages and high transaction costs involved in providing insurance to low-income groups.

Youth and adult training programmes were studied in Chile, Argentina, Tunisia and Ghana. These aim to provide skills and training to people who have dropped out of formal education at an early age, and are, as a result, highly vulnerable to unemployment and poverty. In three cases (Chile, Argentina and Tunisia), the emphasis is on providing labour market skills for the young (ages 15-24), through a combination of basic occupational training and job placements. In Ghana, the emphasis is on providing basic literacy skills to groups of adults with particularly low literacy rates (e.g. rural communities, women), although classes are combined with occupational skills and livelihood improvement activities. All have succeeded in enrolling large numbers of adults, and demonstrate generally positive impacts on the skills, productivity and labour market outcomes of participants.

Affirmative action programmes were studied in Malaysia, India and Nigeria. In each case the aim is to address inequalities between groups - defined by ethnicity in Nigeria and Malaysia, and mainly by caste in India - largely through the use of quotas. In Malaysia, quotas were established in 1970 for ethnic Malays and other indigenous groups (together, the ‘Bumiputera') in the ownership of share capital, employment in modern-sector firms, and enrolment in higher education. In India, there are quotas for Scheduled Castes and Scheduled Tribes in appointments and promotion in government services, admission to public education institutions, and seats in the legislature. In Nigeria, the country's ‘Federal Character Principle' requires that there is no predominance of persons from a few states or from a few ethnic groups in the government or any of its agencies. Substantial progress has been made towards meeting the quotas in Malaysia and India – although as part of an explicit goal to eradicate poverty irrespective of race or ethnicity, and a certain amount of flexibility in their enforcement. Progress has been slower in Nigeria however, where there have been calls for the monitoring and enforcement of quotas to be strengthened.

Conclusions and next steps

That inequalities matter for development has been long recognised in development discourse. This concern has re-emerged more recently in the context of debates around countries' progress towards the various targets set out at the UN Millennium Summit in September 2000. Nevertheless, policies designed to address inequality often fail to feature highly in the policy agenda. This reflects a number of different constraints, including limited or inaccessible evidence about the magnitude of inequalities, and their effects; concerns about the effect of redistributive policies on incentives, and economic growth; and entrenched power interests resistant to policy change.

The findings in this pack suggest that inequalities matter and can be addressed effectively and sensibly by policy. Inequality is not, as was once thought, a purely middle-income country issue; the balance of evidence suggests that reducing inequality is more likely to increase, rather than reduce economic growth; and reductions in inequality are possible and have made a significant contribution to poverty reduction in recent decades. There is also a wide range of specific policy instruments by which inequalities can be, and are being, tackled, and a lot of scope for countries to choose those most suited to their own conditions. With donor support, there are no countries for which the tool-box is empty.

In terms of next steps, there is scope for more case studies of other types of policy instruments, such as land reform, equal opportunities legislation and regional funds. There is also scope to expand understanding of the different dimensions of inequality, and the relative contributions of capital market imperfections, external shocks, policy reforms, discrimination and politics in explaining inequality, in different country contexts. There is also scope to increase the impact of existing and future research on policy, though national-level workshops, inter-regional visits and exchanges, media work and other advocacy activities.

References

Besley, T. and Kanbur, R. (1993). Principles of targeting. In M.Lipton and J. van der Gaag (eds), Including the Poor, Washington D.C.: World Bank.

Clemens, M., Kenny, C. and Moss, T. (2004). The trouble with the MDGs: confronting expectations of aid and development success. Working Paper No.40, Center for Global Development, Washington D.C.

Cornia, G. (2004). Inequality, Growth and Poverty in an Era of Liberalization and Globalization. Oxford: Oxford University Press.

De Ferranti, D., Perry, G., Ferreira, F. and Walton, M. (2004). Inequality in Latin America: Breaking with History? World Bank, Washington D.C.

Dollar, D. and Kraay, A. (2002). Growth is good for the poor. Journal of Economic Growth, 7: 195-225.

ECLAC-UNDP-IPEA (2002). Meeting the Millennium Poverty Reduction Targets in Latin America and the Caribbean. Santiago, Chile.

Hanmer, L. and Naschold, F. (2000). Attaining the International Development Targets: will growth be enough? Development Policy Review 18(1): 11-36.

Keefer, P. and Knack, S. (2002). Polarization, politics and property rights: links between inequality and growth. Public Choice 111: 127-154.

Li, H., Squire, L. and Zou, H. (1998) Explaining international and intertemporal variations in income inequality. Economic Journal 108: 26-43.

Odedokun, M. and Round, J. (2004) Determinants of income inequality and its effects on economic growth: evidence from African countries. African Development Review, 16(2): 269-284.

Ravallion, M. (2005). Inequality is bad for the poor. Mimeo, Development Research Group, World Bank.

Stewart, F. (2004). Horizontal inequalities: a neglected dimension of development. Working Paper No.1, Centre for Research on Inequality, Human Security and Ethnicity, University of Oxford.

UN Millennium Project (2005). Investing in Development: A Practical Plan to Achieve the Millennium Development Goals. London: Earthscan.

United Nations Development Programme (UNDP) (2005). Human Development Report 2005. New York: UNDP.

Wilkinson, R. (2005). The Impact of Inequality: How to Make Sick Societies Healthier. London: Routledge.

World Bank (2005). World Development Report 2006. New York: Oxford University Press.

This overview was written by Edward Anderson of the Overseas Development Institute, London. For more details or comments, please contact: e.anderson@odi.org.uk.