II Poverty Reduction Strategy

Joachim Harnack, Sérgio Leite, Stefania Fabrizio, Luisa Zanforlin, Girma Begashaw, and Anthony Pellechio
Published Date:
November 2000
  • ShareShare
Show Summary Details

The high-profile political process that launched constitutional democracy in the 1990s and generated Ghana– Vision 2020 placed poverty reduction at the center of economic policy. The main themes of Ghana– Vision 2020 were economic growth, investment in human capital, rural development, and an enabling environment for private entrepreneurship and investment. These themes were carried into the medium-term program for the first five-year period of the strategy, 1996–2000, with human development as the focus for efforts at poverty reduction (Government of Ghana, 1997a). The basic goals in this area were to improve health, life expectancy, and the capabilities of all persons; eliminate extreme deprivation; and ensure an equitable distribution of the benefits of development.

The broad strategy for poverty reduction emphasized rural development, expansion of employment opportunities for the urban poor, and better access of both rural and urban poor to basic public services such as housing, transportation, water supply and sewerage, and family planning services (Technical Committee on Poverty, 1996). The principal elements of the strategy included:

  • promoting equitable economic growth through sound macroeconomic policies that support employment;
  • strengthening the agricultural sector by introducing modern farming methods and marketing practices;
  • increasing investment in human capital to build a more educated, trained, and healthy labor force;
  • providing an environment for the manufacturing and services sectors that enables the creation of new business and employment in the formal sector;
  • promoting development of an indigenous entrepreneurial class to encourage microenterprise and small-scale businesses and strengthen the informal sector; and
  • developing firm targets for poverty reduction and improving data on poverty to ensure more effective targeting of poverty reduction policies.

The medium-term program was the outcome of a decentralized and participatory development planning system that extended into the community level of local administration. It covered all 110 district assemblies, the 10 regional coordinating councils, and various government ministries and departments and addressed the priorities of the private sector, NGOs, and organized labor. 4 During 1996, consultations were conducted through stakeholder workshops and cross-sectoral planning groups convened by the NDPC. The process was completed in July 1997.

The National Economic Forum in September 1997 provided a framework in which poverty reduction could be further discussed and programs developed (Government of Ghana, 1997b). However, given the pressing need to restore a stable macroeconomic environment after the 1996 elections, the main focus was on building a national consensus to reduce inflation and the budget deficit, at the same time shifting expenditure to the social sectors. These developments were viewed as preconditions for pursuing an aggressive poverty reduction program.

Institutional Structure for Reducing Poverty

The Consultative Group, meeting in Paris in 1995, made a concerted push for greater focus and coordination of efforts to reduce poverty, which led to the formation of the Inter-Ministerial Committee on Poverty Reduction (IMCPR) as the main institution responsible for setting and coordinating policies for poverty reduction. Its chairperson is the minister of finance, who is responsible for coordinating donor assistance and making budgetary allocations to poverty reduction programs. 5 The NDPC is the coordinating agency for the poverty programs of the various government agencies, assigning tasks to each agency with the goal of designing implementable poverty reduction programs and projects. The technical arm of the IMCPR is the Technical Committee on Poverty (TCOP), whose chairperson is the director-general of the NDPC. The NDPC has a poverty reduction unit that coordinates all poverty reduction programs and activities of the TCOP. The main functions of the TCOP include:

  • disseminating information on poverty reduction in collaboration with the Ghana Statistical Service;
  • establishing policy guidelines on poverty reduction for district development plans in collaboration with the Ministry of Local Government and Rural Development;
  • commissioning periodic impact studies to monitor the effectiveness of approved programs and projects; and
  • developing training aids and programs to build capacity for designing poverty reduction programs at the subnational level.

The TCOP administers, manages, and coordinates all of the government’s poverty reduction programs, such as the National Poverty Reduction Program (NPRP) and the Social Investment Fund (SIF). The TCOP also fosters linkages among NGOs participating in diverse donor-funded poverty initiatives, connecting them with the government’s development framework for poverty reduction. To enhance the effectiveness of its policymaking and program coordination, the TCOP has established a number of technical working groups to address several aspects of poverty problems. To address poverty in a truly participatory manner under the new development planning system dictated by Ghana– Vision 2000, the TCOP’s working groups are open to all organizations interested in poverty reduction. In March 1999 the Thematic Group on Poverty Reduction (TWG–PR) was established jointly by the NDPC on behalf of the government and the United Nations Development Programme (UNDP) on behalf of donors.

The coordination of poverty activities extends to the district and local levels through the District Planning Coordinating Units of the district assemblies. The district assemblies appoint poverty desk officers, who act as liaisons between these units and the NDPC. At the sectoral level, each ministry, department, and agency has a division responsible for coordinating poverty reduction activities.

Programs to Reduce Poverty

The programs designed and initiated in the 1990s to reduce poverty reflect, to some extent, an effort to build on the lessons learned from PAMSCAD in the late 1980s (Box 2.1). The program strained the implementation capacity of the government and fell short of its objectives, with less than 20 percent of the support pledged by donors actually disbursed on projects. These were widely dispersed by region and sector and deficient in targeting the rural poor.

The 1992 constitution introduced the District Assemblies Common Fund (DACF), which provides financing from the central government, required to be at least 5 percent of domestic tax revenue, for development projects of the district assemblies. Donors also channel funds through the DACF. as does, for example, the World Bank’s Village Infrastructure project. The effort to decentralize the budget and development planning process continued with the undertaking, in the first half of 1999, of a fiscal decentralization project in the Ministry of Finance with assistance from the Canadian International Development Agency (see Box 6.1). A crucial aspect of the design of district budgets is to ensure that they are fully compatible with the medium-term expenditure framework implemented in 1988 (see Box 5.1).

The NPRP is funded by the government and the UNDP and aims at assisting communities in reducing poverty by identifying their needs and participating directly in developing strategies and programs. The program’s main objectives are to build the management capacity of district assemblies, community–based organizations, and NGOs to better coordinate their efforts; to develop the skills of the rural poor to generate employment; to implement the SIF; to develop community and household technologies to improve productivity; and to implement mechanisms to empower women and promote female education. The program covers the period 1997–2000 and is being implemented in five pilot districts (Bongo, Afram Plains, Juabeso-Bia, Dangme West, and Ga Mashie).

The SIF is a community–based, demand-driven, rapid-disbursing fund established with support from the UNDP and the African Development Bank. Funds from the SIF are available to about six districts during the three-year pilot period, 1998–2000. Its main goals are to facilitate access of the poor to basic social and economic infrastructure and services; to provide credit to microenterprises; and to strengthen the delivery of poverty reduction programs by local governments, community-based organizations, and NGOs.

Under these funding sources and programs, resources are channeled directly to poor groups and communities through the district assemblies, NGOs and community–based organizations for implementation of projects that benefit the poor. For example, in addition to expenditures on health, education, and local government from the DACF, each district assembly is expected to devote 20 percent of its DACF allocation to economic activities that will generate employment and income in communities within the district.

Box 2.1.Program of Action to Mitigate the Social Costs of Adjustment (PAMSCAD)

The Ghanaian government recognized early on that some components of the Economic Recovery Program “will exacerbate the economic problems of certain vulnerable groups in the short run, and this may impede the sustainability of the recovery program itself (Government of Ghana, 1987). For example, among the obvious victims of the adjustment program would be laid-off workers in the public sector. Therefore the government developed and began to implement in 1988 a package of measures organized under the umbrella of the Program of Action to Mitigate the Social Costs of Adjustment to protect these groups from the adverse effects of reform. PAMSCAD focused on poor households in rural areas, especially those in the northern regions and those with limited access to income-earning opportunities and social services, as well as the urban poor and unemployed former civil service and state enterprise workers. The program emphasized community initiatives, employment generation through public works and food-for-work programs, training and placement services for those affected by public sector retrenchment, and provision of basic needs such as hand-dug wells, low-cost sanitation, essential drugs, and supplemental nutrition.

At the national level, PAMSCAD was integrated with the government’s structural adjustment program, with joint responsibility vested in the Ministry of Finance and the Ministry of Local Government. A donors conference was convened in February 1988 to discuss a menu of projects that donors might support; this conference generated pledges of support amounting to $84 million for an initial program of 24 projects. PAMSCAD was given attention and support as an innovative means of addressing social concerns while economic adjustment is being promoted. It focused on specific groups and embodied a degree of decentralized policy action.

However, the program got off to a slow start and encountered problems. By the end of July 1990, only $15 million in donor financing had been disbursed. Projects were widely dispersed geographically and across sectors, which strained implementation capacity, and targeting to the rural poor was deficient. Recommendations for improvement included devolution of responsibility for project initiation and implementation to the district level; capacity building at the district level; reduction of the number of donors involved in each project; more attention to recurrent costs; and incorporation of some projects into the regular public investment program.

In 1999 Ghana began implementing a Community–Based Poverty Reduction Project, funded by the World Bank. A poverty measurement and monitoring component will identify core indicators, including district– and community-specific indicators. Other components will address nutrition and food security and the problem of street children.

Poverty Monitoring and Indicators

The main source of detailed information on the social and economic conditions of Ghana’s population, including detailed measures of poverty, is the Ghana Living Standards Survey (GLSS), a multi–topic household survey undertaken by the Ghana Statistical Service. Four surveys, denoted GLSS1 through GLSS4, were conducted in 1987/88, 1988/89, 1991/92, and 1998/99, respectively, each covering a nationally representative sample of households interviewed over a 12-month period. GLSS1 through GLSS3 were used for a consistent analysis of poverty in Ghana over the years of the surveys (Government of Ghana, 1995). GLSS4 was completed in June 1999 and served as the basis for a major government report updating Ghana’s poverty profile and presenting policies for reducing it.

In 1996 the Ghana Statistical Service adopted a new instrument, the Core Welfare Indicators Questionnaire (CWIQ) Survey, for monitoring living standards more quickly than could be done with the GLSS. CWIQ is a nationwide probability sample survey designed to provide timely annual indicators for monitoring poverty and the effects of development policies, programs, and projects on living standards. Rather than undertake the time-consuming and computationally demanding task of measuring expenditure and income, CWIQ employs a set of readily observable indicators to construct poverty quintiles that are closely correlated with the expenditure quintiles based on the GLSS. The first CWIQ survey was conducted from September to November 1997 (Government of Ghana, 1998).

CWIQ represents an important innovation in data collection for poverty measurement and analysis. CWIQ results are available much more rapidly than other data, making them more helpful in tracking developments. The emphasis on measuring easily observable indicators of well-being—the percentage of households with access to safe water, distance to the nearest primary school, and levels of use of and satisfaction with public health services—produces more useful results for monitoring changes in actual well-being, rather than intermediate indicators of inputs or actions taken to improve well-being.

A set of poverty indicators based on the GLSS and CWIQ is under development. Consideration is being given to identifying those indicators that change frequently and whose impact can be easily measured from year to year. Indicators under consideration include:

  • demographic: average household size, total fertility rate, female-headed households, and migration (rural-to-rural, rural–to–urban, and international);
  • economic: incidence of poverty, economic participation rate, sources and levels of income, food share in household expenditure, level of savings, sources of credit, and proportion of adults with access to productive assets;
  • social: literacy rate, primary school enrollment rate by sex, access to health facilities, health personnel consulted, percentage of children undernourished, and access to safety net programs provided by the Department of Social Welfare; and
  • household and community: average number of persons per room, access to potable water, access to electricity for lighting and productive activity, and access to sanitary facilities.

The Ghana Demographic and Health Survey is a third instrument for measuring and monitoring poverty. It has been conducted in 1988, 1993, and 1998 and provides information on fertility, infant and child mortality, child nutrition, utilization of maternal and child health services, and the incidence of HIV/AIDS.

The poverty analysis undertaken with the GLSS4 departed significantly from previous analyses by establishing an absolute poverty line rather than using lines defined by ratios of average household consumption per capita as before.6 Applying this new poverty line raised the incidence of poverty calculated from the GLSS3 but showed a significant decline between the GLSS3 and the GLSS4 (Table 2.1). The incidence of extreme poverty declined from 36 percent of the population in 1991/92 to 29 percent in 1998/99, while poverty as measured by the upper poverty line fell from 51 percent to 43 percent. The increase in poverty in Accra as a result of urban migration between 1988 and 1992 was reversed; extreme poverty in the capital declined from 11½ percent in 1991/92 to 2½ percent in 1998/99, and poverty based on the upper line fell from 22½ percent to less than 5 percent.

Table 2.1.Incidence of Poverty by Region(In percent)
Survey and RegionPoverty Line= ¢900,000Poverty Line= ¢700,000
IncidenceShare of

total poverty
IncidenceShare of

total poverty
GLSS3 (1991/92)
Urban coastal28.34.814.93.6
Urban forest25.85.612.94.0
Urban savannah37.
Rural coastal49.713.830.712.2
Rural forest60.835.445.137.4
Rural savannah72.132.855.936.2
All urban27.518.515.314.2
All rural62.482.045.885.8
All Ghana50.8100.035.7100.0
Urban coastal26.85.317.15.1
Urban forest24.85.915.15.2
Urban savannah42.24.429.74.5
Rural coastal46.316.130.115.1
Rural forest41.431.624.426.9
Rural savannah70.535.658.242.6
All urban22.816.714.515.4
All rural51.683.336.284.6
All Ghana42.6100.029.4100.0
Source: Government of Ghana.
Source: Government of Ghana.

However, the overall decline in poverty was not evenly distributed across regions: modest reductions were seen in both urban and rural areas outside Accra and the rural forest localities, and there was an increase in poverty in the urban savannah. In both 1991/92 and 1998/99, poverty was substantially higher in rural areas than in urban areas, and within both urban and rural areas, poverty was disproportionately concentrated in the savannah. In both 1991/92 and 1998/99, more than half of those in the rural savannah were classified as extremely poor. This area has benefited very little from the overall poverty reduction that has taken place in the country.

The depth of poverty can be gauged by the poverty gap, or the proportion by which the average consumption of poor households falls below the poverty line. For the country as a whole, the shortfall among the extreme poor increased slightly from 35 percent in 1991/92 to 36 percent in 1998/99 (Figure 2.1). The depth of poverty increased in those localities where the incidence of extreme poverty rose, even though the incidence of poverty based on the upper poverty line declined. In other words, this decline was accompanied by an increase in the depth of poverty for those who remained poor.

Figure 2.1.Poverty Gap by Region1

(In percent)

Source: Government of Ghana.

1 The poverty gap is the difference between the average consumption of poor households and the lower poverty line (¢700,000), as a percentage of the latter.

2 Survey year.

Households headed by women accounted for about a third of all households, and that level had been rising slowly over the course of the surveys. The 1987/88 and 1988/89 surveys showed the proportion of these households falling as expenditure rose. In contrast, the pattern in the 1991/92 survey is inconclusive, showing the proportion first increasing and then decreasing as expenditure rose. The CWIQ survey found 35 percent of households to be headed by women in 1997, with significant proportions of these households in the lowest quintile in both rural (53.3 percent) and urban (60.5 percent) areas.

The 1987/88 survey showed that 40 percent of the sampled population age 15 years and over could read. 39 percent could write, and 48 percent could do written calculations. These percentages showed slight improvement in the 1988/89 survey but could not be consistently analyzed in the 1991/92 survey. In both of the early surveys, the proportions of men who could read, write, and calculate were much higher than those of women, with differences of about 25 percentage points. Both literacy and numeracy levels show a positive correlation with expenditure levels for both women and men and are higher in urban than in rural areas. The CWIQ survey reported a literacy rate of 48 percent for all household heads in 1997, with a 62 percent rate for male heads and 36 percent for female heads.

The CWIQ survey found a high negative correlation between education and poverty in both rural and urban areas and for both female- and male-headed households. Poverty declined steadily with education: 54 percent of female heads and 77 percent of male heads of nonpoor rural households were literate, whereas only 20 percent of both female and male heads of households in the lowest poverty quintile were literate.

Overall, women experience greater poverty for a range of reasons. These include lower literacy; less access to education, employment, and farm land; limited access to credit; marital disruption; and social and cultural attitudes that constrain women from making progress in business and other ways that improve their standard of living.

The most important source of income for households in Ghana is self-employment, accounting for at least 70 percent of total income. This predominance is observed in all expenditure quintiles. The CWIQ survey showed 72 percent of household heads to be self-employed in 1997. In the lower quintiles, most self-employment income is generated from agricultural activities, whereas in the higher quintiles it is generated more from nonfarm self-employment. The two groups least affected by poverty are public sector and formal private sector employees, for whom the incidence of poverty is consistently below the national average.

Besides determining its geographic pattern, it is also important to relate poverty and trends in poverty to the economic activities in which households are engaged (Figure 2.2). In 1998/99 in particular, poverty was highest by far among food crop (as opposed to export) farmers. Moreover, their contribution to the national incidence of poverty as indicated in the GLSS4 was far in excess of their population share. Indeed, at the national level almost 58 percent of those identified as poor were from households for which food crop cultivation was the main activity.

Figure 2.2.Incidence of Poverty by Main Economic Activity of Household

(In percent)

Source: Government of Ghana.

1 Survey year.

The GLSS4 showed wage employees in the formal private sector and export farmers to have experienced the largest reductions in poverty. Nonetheless, the incidence of poverty is still high among export farmers, as well as among informal sector wage employees and the nonfarm self–employed. The last group is large: in 1998/99, almost 24 percent of Ghana’s poor were from households engaged primarily in nonfarm self-employment. Poverty also fell among wage employees in the public sector and nonfarm self–employed workers. The number of public sector wage employees fell significantly, with a corresponding increase in the number of nonfarm self–employed workers.

Lessons and the Challenge Ahead

An important lesson from examining the dynamics of poverty in the 1990s is that the incidence of poverty declined, but that decline was interrupted by macroeconomic instability– in particular, by high inflation and slow growth brought about in large part by excessive government spending associated with the 1992 and 1996 elections. Notwithstanding the decline, the incidence of poverty remains high, and much more progress is needed to achieve the poverty reduction goals of Ghana—Vision 2020.

Both Ghana’s institutional structure for participatory development of poverty reduction policies and programs and its survey-based poverty data collection capabilities are well developed not only by African standards but also in comparison with those of developing countries worldwide. Indeed, the successful implementation of the CWIQ survey establishes Ghana as an innovator in data collection. The focus of the CWIQ survey on monitoring well-being more directly and rapidly than was possible before is well suited for developing effective poverty reduction programs.

However, Ghana’s implementation of such programs and of an effective social safety net are less advanced than its institutional structure and data collection capability might indicate. This situation can be rectified by the increased focus on poverty reduction that materialized in the late 1990s in the structural adjustment lending programs of the IMF and the World Bank. Ghana’s institutional structure and record of data collection provide an excellent foundation for its preparation of the poverty reduction strategy paper required by the IMF’s and the Bank’s lending facilities.

The medium-term expenditure framework implemented in 1998 provides an appropriate budget planning mechanism for determining the full cost of poverty reduction programs in a rolling three-year framework (see sections V and VI and Box 5.1). Given its focus on outcomes of government spending this framework can contribute significantly to the development and monitoring of a poverty reduction strategy. It can also be used to improve the efficiency of social spending. Given the setbacks in poverty reduction that resulted from excessive government spending and the consequent macroeconomic instability, the efficiency of such spending on poverty reduction is a fundamental issue in Ghana’s effort to improve its standard of living. To address this issue and capitalize on the medium-term framework’s capability for monitoring a poverty reduction strategy, the quality of national accounts and government budget statistics needs to be substantially improved.

The capacity of district assemblies, local governments, and communities to pursue and complete activities called for in the medium-term expenditure framework, and to meet program deadlines and targets, needs to be factored into project design and implementation. In this context, problems causing delays in the release of project funds, which affect service delivery, need to be solved. Strengthening the capacity of subnational agencies will not only contribute to realistic budgeting under the framework but also broaden its consultation process.

Ghana must use the capabilities developed in the 1990s for achieving public consensus on poverty reduction policies and programs, and for monitoring poverty, to implement a poverty reduction strategy that is efficient and yields tangible results. The system of decentralized and participatory development planning, from central agencies through community-based organizations, needs to focus on the value that it can add to policy formulation and program implementation. Further, district assemblies and local government should receive both resources from the central government and the authority to raise their own resources through taxes and fees, so that they can meet their assigned responsibilities in delivering social services. An important principle in the effective use of resources for poverty reduction is that they should not go too much to project management instead of to the poor. Government policies and actions must be effective and, especially at the local level, focused on service delivery and goals.

The policymaking capabilities developed in Ghana’s democratic environment of the 1990s– a consensus-building institutional structure for policymaking, stronger survey data collection capacity, and the medium-term expenditure framework– put the country in a good position at the beginning of the 21st century to make significant progress in reducing poverty.

    Other Resources Citing This Publication