Already published in the series:
Promises to the Poor: the Record of European Development Agencies
Aidan Cox and John Healey
Approaches to Sustainable Livelihoods for the Rural Poor
Diana Carney
The Meaning and Measurement of Poverty
Simon Maxwell
The Poor and their Money
Ana Marr
Making Adjustment Work for the Poor Tony Killick
Trade Liberalisation and Poverty
Henri-Bernard Solignac Lecomte and Maurizio Bussolo
Towards National Public Expenditure Strategies for Poverty Reduction
John Healey, Mick Foster, Andy Norton, and David Booth
Forthcoming:
Poverty, Violent Conflict and Aid
Margie Buchanan-Smith
The authors are Research Fellows at ODI. To contact them email: j.healey@odi.org.uk
ODI wishes to acknowledge grants from the Swedish International Development Cooperation Agency, Sida,
and the UK Department for International Development towards this series. However, the opinions expressed are those of the authors.
| The picture in brief |
Projections of likely future scenarios suggest that:
- In developing countries as a whole, extreme poverty can be more than halved by 2015 provided that currently forecast GNP growth rates (higher than past growth rates) are realised.
- The Latin America and Caribbean regions will only halve poverty if they generate conditions for a more pro-poor pattern of growth. Sub-Saharan Africa will not halve poverty in any likely future scenario.
- High levels of income-inequality limit the poverty reducing effects of growth. Higher growth and pro-poor policies will improve poverty reduction prospects in both high and low-inequality countries but high-inequality countries will need to grow twice as fast as low-inequality countries to halve poverty by 2015. This is not feasible.
| Some challenges for external funders |
- To encourage pro-poor growth through strategies which:
– encourage national capital stock to grow at least as fast as the labour force;
– ensure the most efficient use of new investment;
– help to achieve higher agricultural yields and linkages between farm and non-farm activities in South Asia and sub-Saharan Africa. - To give attention to the impact of policies on the distribution of income and assets particularly in those high-inequality countries (mainly in sub Saharan Africa) where low growth prospects will otherwise severely limit poverty reduction.This inter alia requires more attention to:
– the scope for redistribution of land ownership or use and especially greater access of the poor to these assets;
– shifting the distribution of public subsidies and services more towards primary education and primary health especially for girls, than in the past;
– the likely impact on the poor of the functioning of markets at the micro-level, especially removing bureaucratic obstacles in the informal sector. - To ‘mainstream’ gender equality within government and to make more efforts to challenge normal gender roles.
I. The international development targets
In 1996, the OECD/DAC adopted a set of international development targets to guide the efforts of a global development partnership between development agencies and their developing country partners. These targets focused on raising the standard of living, human development and well-being in developing countries. A major target was to halve the proportion of people living in extreme poverty by 2015. Extreme poverty refers to the percentage of people living on less than $1 dollar a day, using 1985 purchasing power parity exchange rates to convert the dollar into local currency.
Will extreme poverty be halved? This paper presents the results of recent ODI research commissioned by the Department for International Development on prospects for all developing countries. The results point to some major challenges.
II. Determinants of poverty
The conditions for growth and the degree of inequality are two key factors which appear to determine the extent of poverty reduction, given the rate of per capita economic growth.
Poverty reduction depends on the economic conditions that generate pro-poor growth. There are many conceptions of what economic conditions are required to generate ‘pro-poor’ growth. This econometric study supplemented by country case studies, tested some possible conditions against past experience and found that four of them were associated with a greater reduction in poverty. First, there is more reduction in the proportion of poor people if the national stock of capital grows faster than the labour force. It is necessary to generate enough extra capital to gainfully employ the extra (often poor) hands and raise their productivity. Second, poverty was reduced more in those countries where capital was used more efficiently (i.e. where a given amount of investment produced a larger amount of additional goods and services). Although this only gives us a rough guide, it does indicate that efficiency (in the use of capital at least) is not the enemy of poverty reduction but an aid to it. Third, in South Asia and sub-Saharan Africa (though not elsewhere) the performance of the agricultural sector appears critical because poverty was lower where agricultural productivity per worker was high relative to the modern sector. Fourth, historically those countries with more ‘open’ economies have lower poverty ratios. Taking into account levels of national income per capita, a higher proportion of those countries which had less restrictions on international trade, had less poverty. The criteria for ‘openness’, included the number of controls on exports and imports and the extent to which governments controlled the prices of exports. However, this last finding must be interpreted with some caution as some other recent studies have not confirmed a favourable relationship (See ODI Poverty Briefing 6: December 1999).
The importance of income equality for poverty reduction. With a given rate of economic growth, poverty falls faster in those countries where inequality of income is lower. The so called ‘Gini Coefficient’ measures statistically the degree of inequality in the distribution of household income. The higher the coefficient, the more extreme is the observed inequality. Analysis of developing countries in the 15 years between 1985 and 1990, showed that they fell into two groups. In the low income-inequality countries (average Gini coefficient = 0.34), 10% economic growth was associated with a fall in the proportion of people below the poverty line by 9 percentage points. In the high income-inequality countries (average Gini coefficient = 0.55) 10% growth was associated with only a 3 percentage point reduction.
III. Two future scenarios
Two basic future scenarios were projected.
Table 1 gives the past and assumed future (usually higher) per capita growth rates for each scenario and shows the projected poverty reduction outcomes under these. Bold numbers indicate where the target might be attained.
| Table 1. Poverty in 1990 and future projections of poverty in 2015 | ||||
| Poverty (% under $1 a day at 1985 purchasing parity prices) | ||||
| 1990 | 2015 No change in conditions | 2015 Pro-poor, higher growth rate | ||
| A | B | |||
| Sub-Saharan Africa | 44 | 42 | 36 | 25 |
| Middle East & North Africa | 3 | 2 | 1.6 | 1 |
| East Asia and Pacific | 31 | 12 | 12 | 9 |
| South Asia | 47 | 30 | 24 | 16 |
| Latin America & Caribbean | 28 | 19 | 17 | 12 |
| Eastern Europe & Central Asia | 9 | 5 | 4 | 3 |
| Developing Countries | 36 | 22 | 18 | 13 |
| Note: The past and future assumed growth rates per capita (in % p.a.) are respectively: sub-Saharan Africa, 0.0/1.4; Middle East and North Africa, 0.1/1.4; East Asia and Pacific, 5.4/5.6; South Asia, 3.7/2.3; Latin America and Caribbean, 3.0/1.3; all developing countries 3.0/4.0. Bold numbers indicate the target is attained. | ||||
IV. The influence of inequality
The influence of inequality on poverty reduction is illustrated by the results of two exercises. First, countries are assumed to stay in the same ‘high’ or ‘low’ income- inequality category until 2015. Projections (see Table 2) show that countries which fall into the low income- inequality group, attain the poverty reduction target with no change in the pattern of growth and lower growth rates. By contrast, countries that fall into the high-inequality group will halve extreme poverty by 2015 only if growth is pro-poor and higher than it has been in the past (average 4% per capita per annum). High-inequality countries need over twice as much growth as low-inequality countries (7.1% growth per capita per annum), to reach the poverty target. This is hardly feasible.
A second exercise for all developing countries also indicates the importance of inequality for poverty. A growth rate of 4% per annum per capita is projected using a World Bank data-set (which differs from that used in Table 1). This exercise suggests that if all countries in the developing world were to belong to the high-inequality group then although poverty falls, the target of halving poverty by 2015 would not be attained. If on the other hand, all these countries belonged to the low-inequality group then the target is easily met and poverty is halved as soon as 2005 (see Figure 1).
Figure 1 Global Poverty in 2015 - the effect of income inequality
(with growth at 4% p.a.)
(High-inequality is the blue line)
| Table 2 Projected poverty in 2015 for high and low-inequality countries | ||||
| Poverty incidence 2015 as % of 1990 level | Annual per capita growth needed to halve poverty by 2015 | |||
| No change, past growth | Pro-poor, With no change | With pro-poor conditions | higher growth | |
| High-inequality countries | 68 | 49 | 7.1 | 3.7 |
| Low-inequality countries | 47 | 33 | 3.7 | 1.5 |
V Pro-poor growth: major areas of policy challenge
Progress in reducing poverty has been limited since 1987. For developing countries in aggregate, poverty fell only 4 percentage points from 28.3% to 24%, between 1987 and 1998 and in many individual countries, poverty is on the increase. Economic stagnation or low growth rates only partly explain this dismal performance. In some countries, little growth has trickled-down to reach the poor. For example, poverty reduction in both China and India was slower than expected, even though growth was higher than anticipated. Certain conditions and policies are therefore needed to launch countries onto a pro-poor growth path.
Challenge 1: Adequate growth and efficient use of the capital stock
Our findings point to rather ‘old fashioned’ economic messages: the vital importance of sufficiently high investment levels and the efficient use of that new capital. After all, poverty was reduced in East Asian societies in the 70s and 80s through rapid capital accumulation in manufacturing – which built on the successful transformation of agricultural production. The scope for creating livelihoods for unskilled/semi-skilled poor workers through substitution of labour for capital has been limited. This reinforces the need for a policy environment which favours domestic and foreign ‘productive’ investment and for continued and increased aid flows for public investment in economic and social infrastructure.
The importance of efficient use of capital for poverty reduction points to the need for very careful investment appraisal and choice in the public sector especially in relation to the spatial location of public infrastructure investment particularly in rural areas. The poorest people often live in marginal areas, frequently isolated from markets and services. It is sometimes difficult for them to migrate to other areas. Development of the poorer regions and districts requires substantial investment in rural feeder roads, irrigation, conservation, extension services, etc. to improve access and productivity. This may pose difficult trade-offs and choices because investment in the more backward areas may yield lower or at least more delayed rates of return on the new capital invested.
Challenge 2: Importance of agricultural productivity
Improved agricultural yields were associated with poverty reduction in South Asia and sub-Saharan Africa. This reinforces the case for a rural-based development strategy. The promotion of agriculture can stimulate linkages between farm and non-farm activities which are important for poverty reduction. Recent comparative evidence from Indian states (Ravallion and Datt, 1999) supports this because the number of people below the poverty line was found to be more responsive to the growth of non-farm output in those states where farm yields and rural spending levels were higher. Policies to improve incentives for both farm and non-farm enterprises in rural areas have been recognised for some time (e.g. deregulation of pricing and marketing in agriculture, better marketing opportunities, information and extension services for smallholders crops and small producers of non farm products). Credit networks and similar measures to encourage livelihood creation in non-farm rural businesses are needed but so far few strategies for ensuring access to credit for remote rural areas and the poor seem to have worked well. An enlarged role for formal financial institutions is now seen by some as more important given the limited success in reaching the poor through micro-credit institutions and targeting. Flexible credit repayment periods and provision of crop insurance would assist the poor together with development of ‘Green revolution’ type higher yields for African crops like maize, sorghum and cassava.
Challenge 3: More attention to the distributive impact of policies
In those high-inequality countries where growth prospects will not be enough to achieve the poverty reduction targets by 2015, reducing the degree of inequality would speed up the process of poverty reduction dramatically. These undoubtedly include some sub-Saharan countries where inequality is high. Against this must be set the historical evidence that in most countries, Gini inequality coefficients have changed only slowly over time. Nevertheless some options for the reorientation of government actions on asset and income distribution present themselves. First, policies to reduce high degrees of income inequality need to focus on the access to and distribution of, assets which determine income flows for the poor - especially land. Where inequality is high, agricultural growth does not seem to reduce poverty so well. The potential for redistributive land reform is clearly greater in countries with a high degrees of inequality in land holdings and where there is political support for land redistribution (e.g. South Africa). Since actual land reform programmes being currently implemented in sub-Saharan Africa typically exacerbate inequality, their design must be improved. Policies for liberalisation of land markets may also need to be reconsidered (White et al., 2000). Second, greater equality of income is served by improving the productivity of the poor, more especially through improving basic education, health and especially the skills of poorer groups (e.g. landless labourers and marginal farmers), so they can better take up rural non-farm options or urban-oriented livelihood opportunities. Action by the state to redistribute state subsidies from tertiary to primary and sometimes secondary levels of education and health services and to improve the coverage and quality of these services, has been recognised as necessary for sometime. Despite this, there has been a limited shift of expenditures in that direction in the 1990s and we rarely know how far the poor actually benefit from these services. Third, policy makers need to know much more about how the poor participate in the economy and to establish the distributive consequences of changes in taxes and charges, privatisation under different regulatory frameworks, trade liberalisation, and the removal of government subsidies and price controls. To achieve more pro-poor outcomes requires greater state capacity to regulate market activity and privatised firms in what are often non-competitive market conditions. We know for sure that reduction of bureaucratic interference with small-scale activities in the informal sector would benefit the incomes of the poor.
Challenge 4: Mainstream gender equality
Gender inequality is a constraint on growth and poverty reduction. An increase in the number of girls in school and female literacy is likely to reduce poverty and means that over the longer term, fertility rates will fall (i.e. the number of children borne to a woman in her lifetime). Also child survival will improve. These factors themselves contribute to higher productivity and per capita growth. The low levels of technology available to women, their time constraints and discrimination in their access to credit and other markets, constrain the supply response to any new incentives for small-scale producers.
Our country case studies showed the unfavourable effects on poverty and infant mortality (Pakistan) and girls’ attendance at school (Uganda) which arise from women’s low level of political and economic empowerment. Yet government policies can help create an environment where traditional conceptions of appropriate or ‘normal’ roles for men and women are challenged. Demonstration effects are important and so affirmative action programmes that ensure that women are represented in key professional positions – the judiciary, ministers in government and positions in local government for example – can help to stimulate change. Governments can take action to support women and in some cases with legal representation to ensure that statutory rights are adhered to. Only with such changes in culture will households decide to send children to school rather than have them work and allocate time and money to use health care facilities for their women and children rather than alternative uses.
References
White, Howard, et al. (2000) Africa Poverty Status Report 1999. World Bank for the SPA.
Ravallion, M. and G. Datt (1999) ‘When is growth pro-poor?’, December (www.worldbank.org/research/growth/wpauthor.htm).
© Overseas Development Institute 2000
ISSN 1465-2617
Hanmer, Lucia & Felix Naschold (2000) ‘Attaining the International Development Targets: Will Growth be Enough?’, Development Policy Review, March,11–36.
Readers are encouraged to quote or reproduce material from ODI Poverty Briefings for their own publications, but as copyright holder, ODI requests due acknowledgement and a copy of the publication.The views expressed in this paper are those of the authors and do not necessarily reflect any official position of ODI.
Email: publications@odi.org.uk
Overseas Development Institute Home Page