This essay starts from the assumption that most democratic governments with a relatively free economy want to and should play at least some role in addressing poverty amongst its population. But what should this role be? The first issue to be resolved in addressing this question is whether the government should intervene directly in alleviating poverty or whether the emphasis should be on creating an environment in which economic growth can alleviate poverty. This question is at the core of the continuing growth versus redistribution debate (See Section 2 below).

This essay also assumes that there should be at least some direct government action to address poverty. (Section 1) Various ideological approaches have attempted to identify the ‘root-causes’ of poverty. Such projects varied from socialist experiments (that identified bourgeois control of the means of production as cause) to the ultra-liberal schemes of Reagan and Thatcher in the 70s and 80s (that identified interference in the free market as cause). This strategy was based on an image of society as a seamless web of closely interwoven social, political and economic aspects. In terms thereof fundamental change in one sphere leads to fundamental change in another (Higgins 1977:101). Once governments lost confidence in identifying and eradicating such ‘root-causes’, they started targeting more ‘mixed approaches’ that tried to address the structural causes of poverty, while at the same time taking more immediate short-term action to address the plight of the poor.

At the same time there has been a departure from thinking of single causes of poverty to a more empirical investigation of the multitude of causes of poverty. This has also meant that poverty reduction strategies target a number of different issues and not any single determinant cause of poverty. This can even be seen in recent World Bank and International Monetary Fund (IMF) initiatives such as the Poverty Reduction Strategy papers (PRSPs) and the Comprehensive Development Frameworks (CDFs) that conceptualize poverty-reduction in terms of several broad poverty outcomes. The realization of this complexity in the causes of poverty has also made the goal of poverty reduction a far more complex target to pursue.

The complexity that has emerged in development theory shows in the fact that not even the definition of poverty is a simple issue any more. No longer can poverty be defined simply by referring low levels of income. Poverty is now defined as both low levels of income and consumption and low levels of achievement in education, health and nutrition status (non-income indicators) (World Bank 2000: S6 and Ghai 1997:1).

The result of these developments is that a poverty reduction strategy has to target more than one goal at a time. On the basis of the above definition alone any government wishing to reduce poverty would have to target the improvement of both income and non-income indicators. While the choices in reducing income poverty are relatively simple, the reduction of non-income poverty introduces new complexity. Questions about the relative impact of a unit spent on health or on education on non-income poverty are very difficult to answer. At the same time it is possible to address some of these issues. Within the health for example there is unambiguous evidence that spending on primary health care is more efficient in improving the health-status of the population than spending on tertiary and secondary health care. These issues are discussed in section 3.

While it is important to think about the ideal role that a government should play in poverty reduction, the potential role of most governments, especially those in developing countries, is limited by various capacity constraints. Given their reality, impact and permanence in developing countries one cannot treat such impediments as unwelcome guests. Any poverty reduction strategy should integrate these limitations into the strategy. The relevant government should plan and allow for such limitations. These issues are addressed in section 4.

This paper forms part of a preliminary effort to design a framework for evaluating the current South African government's poverty reduction strategy. The conclusion looks at a preliminary draft of such a framework (section 5).

1. The Role of the State in the Economy

Economic theory suggests that government (democratic, free market) should intervene in the economy to correct market failure, distributional inequality, to create an enabling environment for the private sector and to remove some of the structural barriers to accessing markets:

• When a market economy fails to allocate resources efficiently (i.e. some important goods are not provided), market failure occurs. The justification of government provision of pure "public" goods, for example, is clear. The non-rival and non-excludable (and the consequent inability to charge for) characteristics of these goods imply that the private sector will lack the incentive to supply them. National defense and to a certain extent, roads, have these features. The market often also fails to allocate certain private goods to the poor such as health care and schooling.
• In many cases, equity reasons lead governments to provide private goods--those that are disproportionately consumed by the poor--as a means of transferring resources to the needy in a targeted manner. A market-based economy may distribute income in socially unacceptable ways and it is often the case that governments intervene to protect the vulnerable. Food and housing services are some of the antipoverty programs offered by many governments. Equity concerns are however largely addressed by cash transfers through social security schemes.

• Government intervention is also justified to provide an enabling environment for the private sector. Much of the role of government can be viewed as establishing infrastructure in its broadest sense--educational, technological, financial, physical, environmental, and social. Since markets cannot operate in a vacuum, such infrastructure is necessary for markets to play their central role in increasing wealth. Because constructing the broad infrastructure is beyond the capacity or interest of any single firm, it must be primarily the responsibility of government.
• Government intervention in the economy is also justified to remove some of the structural barriers to accessing markets. The historical concentration of capital in white hands and government efforts to facilitate Black Economic empowerment in South Africa is a good example of such intervention.

All four the rationales for government intervention in the economy stated above have an impact on the quality of life of the poor:

• Market failure affects the provision of private (for example life and health insurance) and public goods (for example the provision of rural roads) to the poor.
• The unequal distribution of resources has a very direct impact on the income of the poor.
• The absence of an environment conducive to private investment affects the rate of economic growth that ultimately, if indirectly, has some impact on the lives of the poor.
• Lack of access to markets disempowers the poor

Many of these possible government actions are in conflict with each other. The most obvious conflict is between creating an environment conducive to economic growth and the provision of public and private goods to the poor. Past a certain point high government spending drives up taxes and inflation and ‘crowds out’ private investment. The apparently conflicting nature of these interventions implies that governments need to choose between them or opt for a mixture of them. The next section of this essay looks at the choice, and interaction, between growth and redistribution.

2. Growth and/or redistribution

Most of the recent debate on poverty in South Africa has centered on this debate. The aim of the government’s initial development plan (the Reconstruction and Development Program) was to foster economic growth by investing government resources in people and communities that had been impoverished by apartheid (Barberton and Woolard 1998:36). The basic logic was that economic growth would be stimulated through redistribution to deprived communities. The demand for goods created in this manner would be the impulse for economic growth.

The subsequent program for Growth, Employment and Redistribution (GEAR) outlines a macro-economic policy framework that makes other policy objectives subject to creating macro-economic conditions that will persuade foreign and domestic investors to invest in the South African Economy. GEAR rests on the tenet that long-term poverty reduction is impossible without sustained economic growth and that the key to economic growth is private investment (Barberton and Woolard 1998:36).

GEAR’s further argues that government spending can only play a limited role in redistribution and that this should rather happen through the creation of private sector employment. GEAR thus not only proposes growth but the kind of growth that creates large numbers of jobs. Ironically Barberton and Woolard (1998:37) base their criticism of GEAR on exactly this point when they argue that the benefits of investment-led growth normally benefits urban, educated and experienced people.

The first lesson of development economists in the 1960s was that improvements in the economic system do not necessarily lead to improvement in the social system. Economic growth (growth in national income) does not automatically filter through to improvements in the social system. This points to a need for planning in terms of social systems as they actually operate (Higgins 1977:99). Higgins (1977:104) contends that at low levels of development in countries with sharply dualistic economies, growth works to the disadvantage of lower-income groups. As dualism weakens in the course of development, the middle-income group is the main beneficiary. But the poorest section benefits only when the government plays an important role and widespread efforts are made to improve human resources.

In a draft of the 2000/01 World Development Report (WDR) the authors argue that there is a strong association between national economic growth and national level income and non-income poverty indicators, but that there is a wide variation across countries in the extent to which economic growth reduces poverty. They find that this variation depends on the pre-existing degree of inequality in both income and assets (e.g. land and education). Addressing these structural inequalities should thus be of key importance in any poverty reduction strategy.

The authors of the WDR also find that inequality impacts negatively on growth and that redistribution can thus support and encourage growth. This means that governments are no longer faced with a simple choice between growth and redistribution. The GEAR document reflects an understanding of this dynamic when it emphasizes that government expenditure that survives the deficit reduction strategy should be re-prioritised towards the poor.

The trade-off between growth and redistribution thus becomes doubly complex. To take spending on old-age pensions in South Africa as an example: Over the period in which the government deficit has been reduced, old-age pensions have not shown any real growth. In South Africa these pensions are important poverty reduction tools because in many areas they support entire households, and not only the elderly. The choice made by the government to reduce these pensions in real terms for the sake of reducing the deficit could thus be counter-productive even in terms of GEAR. It can therefore be concluded that poverty reduction strategies should carefully consider the interaction between growth-enhancing measures and redistribution.

Debates on addressing poverty in South Africa have paid relatively little attention to where government should spend the money that does have available for the poor. This debate has paid virtually exclusive attention to the preceding debate playing growth off against redistribution. The next section opens the debate on which goods government should provide in order to have maximum impact on poverty reduction.

Squire and Kanbur (1999) describe in a recent paper how the definition of poverty has broadened from a simple consideration of income to include longevity, literacy and health (in the 1980s) and vulnerability, powerlessness and lack of voice (in the 1990s).

Considerations of powerlessness and lack of voice deal largely with institutional reform and with the impact of public spending directly. We will therefore deal with it in the next section of this paper that deals with government capacity. Public spending that impacts on the reduction of poverty can otherwise be discussed under two headings, namely spending that enhances the income security and non-income security (opportunity) of the poor.

3.1 Security

The draft World Development Report for 2000/01 argues that the security of the poor can be improved by protecting them against negative shocks on the one hand and by managing national-level economic shocks on the other.

1. Protecting the poor against negative shocks

The WDR first of all argues for a modular approach that has different schemes targeting different groups with different risks. The risk of a retired person (major illness) would for example be significantly different than that of a young employed person (unemployment). Such an approach depends heavily on information about the effects of each form social protection on the population.

The WDR also argues for a need to shift from policies for coping with risk (such as various social security programs) to policies for reducing or mitigating risk such as social welfare programs, health insurance, labor market risk management access to credit and financial assets.

Even more important to developing countries is the need to manage the impact of national-level economic shocks on the poor. The WDR indicates that there is a strong link between macro-economic downturns and rising poverty in terms of income and non-income poverty measures. Macro-economic crises further tend to ratchet up inequality since the poor are more vulnerable to such crises. Responses to such shocks should therefore protect the poor from drastic cuts in living standards and deterioration in their human and physical assets. Adequate pro-poor management of such crises should at least: help the poor maintain adequate consumption levels and access to basic social services; avoid permanent reversals in the accumulation of human and physical capital. Programs that are particularly important to the poor should be protected from budget cuts such as basic education, preventative health care, water and sanitation and rural infrastructure.

Such national level shocks are often the intentional result of macro-economic policies. In such cases plans for fiscal austerity measures should include targeted poverty alleviation measures. When economic pressure is on, targeting limited resources (social security, education and health) to reach the poorest groups becomes very important (Ghai 1997:23). If such pressure persists inequalities in access to health, education and social security develop, especially with regard to quality (Ghai 1997:41). In the case of economy-wide crises, cash transfers, welfare programs, social funds and school scholarships can be used as safety nets to cope with the short-term costs of economic reforms.

Income security is only one dimension of poverty alleviation. This section will deal with the impact that government spending should have on non-income poverty (the need to improve access to physical and human capital and increase the rates of return to such assets).

The poor’s access to physical capital can be addressed via land reform programs and investment in rural infrastructure in the case of the rural poor and investment in physical infrastructure in the case of the urban poor.

Since labour is the most important asset of the poor investment in education and health has an immediate impact on the human capital that the poor have access to. This means an emphasis on adult literacy, universal basic education and primary health care (pre and post-natal maternal and childcare, prevention of insect-borne and infectious diseases, improvement of health education) and (Ghai 1997:ii).

It is also not sufficient to only give the poor access to human and physical capital. Their rate of return on such capital should also be improved by making the market work for them. This can be done by reducing state intervention in the economy at micro level by limiting the intervention of local bureaucracies in small-scale trading activities.

In this essay we presume that most governments are at least nominally interested in poverty reduction. Hunter (in Higgins 1977:114) states that there are at least two requirements for economic and social development. Firstly the presence of the political will to create a development situation. This ‘political will’ is often tested where decisions have to be made that will encounter strong opposition from the class from which the leaders come. Secondly the provision for the expression of local needs. We presume that these pre-conditions are present in the countries to which this framework will be applied. Even so, governments in many countries--more in developing than in industrialized--have limited resources and low capability for effective intervention in the economy. Countries differ in their ability to play an active role in addressing market failures. For example, the ability to establish and enforce laws and regulate financial markets varies considerably. Similarly, the capacity of developing countries to address social equity also varies. At the lower end of the spectrum, countries struggle to provide the very basics--food and shelter--to the poor and needy. The delivery is made mostly in a non-targeted and costly way. Expecting countries with low capability and few resources to undertake more complicated redistributive functions such as providing social safety nets (health care and retirement) could be a tall order to fill. In all the cases studied by Ghai (1997:16) the administrative capacity of the state to reach down to major segments of the population was crucial to widespread social provisioning.

Beyond the general lack of funds, the capacity of governments to alleviate poverty can be described in terms of allocative and operational efficiency (Schick 1998). Allocative efficiency refers to the ability of governments to spend money in a way that reflects their policy objectives whereas operational efficiency refers to its ability to provide the chosen services as cost-efficiently as possible. In a formula allocative efficiency refers to ‘doing the right things‘ while operational efficiency refers to ‘doing things right’.

Admittedly there are technical limitations to what a government can do. Such limitations can only be addressed by some form of training or external input into the system. In the meantime policy should take account of what is possible in the given context.^* Beyond such limitations the design and implementation of policy should reflect the overall goal of poverty reduction. The WDR refers to the need of the poor to influence policy to their needs and police the implementation of such policies as a need for ‘empowerment’. The report argues that poverty is the result of economic processes as well as social and political interactions. The WDR argues that these two issues can be addressed by involving the poor in formal democratic process (thus enhancing allocative efficiency) and giving the poor access to processes of accountability (improving operational efficiency).

Involving the poor in the budget process would improve allocative efficiency in a government that has poverty reduction as its overall goal. The WDR argues that the role of the poor in such policy processes is greatly influenced by the quality and orientation of organizations in civil and political society to the extent that they mediate the voice of the poor to the state. Access to such processes is thus only the first step in involving the poor in policy processes. The configuration of political parties, the system of voting and the political culture of a country all play an important role determining the quality and quantity of participation of the poor in such processes.

Once policy has been formulated and money has been budgeted for it on each departmental budget, things can still go wrong in implementation. Consequently one of the IMFs requirements for a PRSP is the design and implementation of monitoring systems (IMF1999). Besides the concrete mechanisms to facilitate such monitoring of implementation, the availability of accurate information is also of key importance to the exercize. For the purposed of monitoring the public availability of information on long-term goals for key poverty reduction as well as annual targets for intermediate and proxy indicators for the monitoring of progress is of vital importance.

Conclusion

This essay is a first attempt at constructing a framework for evaluating the poverty reduction strategy of governments. Such a framework would have to answer questions at several levels:

• Macro-economic strategy: To what extent does the government’s macro-economic strategy take account of the interaction between growth and inequality? Are poverty and income distribution targets integrated into the model (IMF 1999)? When macro-economic and structural measures are likely to affect the poor, does the macro-economic model explain how the poor will be protected in the short-term e.g. Social safety nets (IMF 1999)?
  Income security: How well targeted is the government’s risk reducing strategy? To what extent does government spending mitigate risk, rather than just enabling the poor to cope with it after the fact? To what extent has the government’s responses to shocks protected the poor from drastic cuts in living standards and deterioration in their human and physical assets?
• Opportunity: Has access to physical capital improved for the rural poor (roads, land)? Has access to physical capital improved for the rural poor (housing, public transport)? To what extent has value been added to human capital (labour) via education and health (adult literacy, universal basic education and primary health care)?
• Empowerment: What mechanisms are in place for the participation of the poor in the policy and budget process? Which factors impede such participation? Which mechanisms exist for enabling the monitoring of the implementation of budgets? Has the government formulated explicit long-term and interim poverty reduction targets and are these public information?

The advantage of the above framework is that it provides a systematic framework for asking questions of different aspects of a government’s poverty reduction strategy. What it does not do is to consider the interaction and feedback between some of these aspects, for example: healthier children perform better at school, better-educated mothers have healthier families etc. These are to some extent context-specific and would have to be reconsidered for each country that is evaluated in terms of this framework (Higgins 1977:122). An understanding of such interaction would enable one to engage with the trade-off and sequencing questions that governments have to ask themselves.

Bibliography

Barberton C. and Woolard I. 1998 ‘The extent of poverty and inequality’ in Barberton C. et.al. (eds.) Creating Action Space: The challenge of poverty and democracy in South Africa IDASA: Cape Town

Ghai D. 1997 Social Development and Public Policy: Some Lessons from Successful Experiences Discussion Paper 89, United Nations Research Institute for Social Development, Geneva

Higgins, Benjamin. 1975. ‘Economic development: Seamless web or patchwork quilt?’ In: Manning Nash (ed.). 1977. Essays on economic development and cultural change in honor of Bert F. Hoselitz. Supplement to Economic Development and Cultural Change 25

International Monetary Fund 1999 Poverty Reduction Strategy Papers: Operational Issues Discussion Draft

Kanbur R. and Squire L. 1999 ‘The evolution of thinking about poverty: Exploring interactions’ World Bank, Mimeo

Shick 1998 A Contemporary Approach to Public Expenditure Management World Bank Institute

World Bank 2000 World Development Report 2000/1: Attacking poverty Discussion Draft