Most countries lack a comprehensive and workable
monitoring system to gauge progress against poverty.
Countries need a comprehensive but workable poverty monitoring system to gauge their progress towards the targets for eradicating extreme income poverty and reducing overall income poverty. The monitoring system should also track progress in the fight against human poverty. The main weaknesses of traditional surveys: they do not illuminate the causes of poverty or generate policy-relevant information. To say that the poor need more income and therefore the economy should grow does not take us very far anymore. Large income and expenditure surveys will have to share the bill with rapid monitoring surveys focused on human poverty and with participatory assessments.
Among the countries whose poverty programmes were assessed for this report, several have set up a poverty monitoring system. (As chapter 1 noted, fewer than a third of developing countries have set targets, but more than three-quarters have made estimates of income poverty for the 1990s.) Uganda monitors income poverty annually and uses the human poverty index to monitor human poverty. It has also started using participatory poverty assessments. China conducts annual household income and expenditure surveys from which it derives poverty estimates. The Philippines recently switched from triennial Family Income and Expenditure Surveys to an Annual Poverty Indicators Survey and began implementing an innovative Minimum Basic Needs Information System at the community level. Thailand has a new monitoring and evaluation system that includes biennial Surveys of Basic Human Needs from which it can derive estimates of human poverty. Benin has one of the most comprehensive systems of poverty monitoring.
Some countries still have relatively undeveloped monitoring systems. Uzbekistan lacks an official definition of income poverty, has abandoned its Family Budget Survey and has not replaced it with a survey that can produce reliable poverty estimates. South Africa still has no viable monitoring system. Mauritania, while in other respects quite advanced on poverty issues, still has a weak monitoring system. Morocco's most recent estimate of income poverty dates from the early 1990s. Many countries still do not have estimates after 1995, the year of the Social Summit.
Producing Poverty Data More Frequently
Many countries are trying to conduct surveys more frequently or provide data more representative of sub-regions. And many have conducted a recent flurry of surveys - often prompted by donor funding. But many others do not plan to conduct income and expenditure surveys more than once every five years - and some even less frequently.
While conducting such surveys no more than once every few years might make practical sense, because of the expenses involved, it will not provide recent - or policy-relevant - information on changes in poverty. Even in the Philippines, where the last triennial survey was in 1997, there is no up-to-date information on the effect of the financial crisis on the poor. And even when surveys are completed, the results are unlikely to be available until a couple of years later.
Some countries have recently started conducting a poverty survey every year (Kyrgyzstan) or every two years (Mali). If these are income and expenditure surveys, the repetition can be expensive. Yemen is attempting to carry out a National Poverty Alleviation Survey of 60,000 households to generate reliable subregional poverty estimates. The results could be used for better geographical targeting of poverty funds. But the expense implies that the survey cannot be repeated very often.
Lighter surveys - using shorter questionnaires (focusing on data for human poverty indicators or proxy indicators for income poverty) and canvassing a smaller sample of households - would have to be used in between large income and expenditure surveys to monitor poverty more regularly - the objective of UNDP's new Rapid Poverty Monitoring Survey. Benin uses such lighter surveys for its Social Change Observatory. And Armenia has recently used one to gauge the conditions of refugees.
The emphasis should be on producing policy-relevant information. For this, participatory poverty assessments are particularly useful - and should be added as a third monitoring tool in any monitoring system. Such assessments in a sample of communities can add valuable insights on how poor people assess their situation and what they think should be done.
Moving to Evaluation
There is growing consensus that poverty monitoring should be linked more tightly to anti-poverty policy-making. How fast information is produced is one issue. Whether the information reveals the causes of poverty is another. But the most troublesome is determining whether and to what extent policies are having an impact on poverty - a matter of evaluation, not just monitoring.
The existing systems of monitoring and evaluation are not up to this task. Most surveys have been used to target resources geographically. Morocco is a good example of a country that has used a sophisticated system of indicators to determine first the poorest provinces and then the most deprived communities within those provinces (see the country profile). This task can be done with a set of surveys at one point in time. To monitor poverty over time requires a series of surveys producing comparable results. Only a few countries have such surveys. Fewer still have fielded a system to both monitor and evaluate the impacts of their poverty programmes. So there is little systematic verification of what policies work - and what policies do not - to help countries move closer to their poverty eradication targets.
Monitoring Income Poverty
Most poverty monitoring systems continue to rely on income poverty measures, as do most poverty profiles. To help provide data for such purposes, UNDP continues, for example, to support the MECOVI (Improvement of Surveys and the Measurement of Living Conditions) programme in Latin America and the Caribbean. Working closely with the Economic Commission for Latin America and the Caribbean, the Inter-American Development Bank and the World Bank, UNDP has helped Bolivia, Guatemala and Nicaragua improve surveys and data analysis and expects to do the same in Ecuador, Haiti and Honduras.
Some economies, such as Moldova and the occupied Palestine territory, have only recently carried out, with UNDP assistance, their first comprehensive income and expenditure survey (box 10.1).
In some cases the results have been unexpected. Palestine Poverty Report 1998 (Palestine National Poverty Eradication Commission 1998), the first publication to explore poverty in the occupied Palestine territory, reveals that 23% of households are poor. It also highlights marked regional disparities and the economic vulnerability of women. Its policy recommendations, such as targeting female-headed households and the poorer rural population, will inform the design of a poverty eradication plan for the territory.
The Fiji Poverty Report, produced by the government and UNDP (1997), debunks the myth that Fiji is a society of self-sufficient farmers living comfortably off the bounty of nature and protected from poverty by a strong social support system in local communities. Instead, the report found that the country suffers from marked inequalities and rising poverty. A new government has made poverty a national priority, and the Ministry of National Planning is developing an anti-poverty strategy.
Other publications, such as the Pacific Human Development Report (UNDP 1999b), report similar conditions on other islands. With UNDP support, household income and expenditure surveys have been conducted in the Cook Islands, Niue, Samoa and Tokelau. The findings for Samoa raised new concerns about meeting people's basic food needs and led the government to target its projects for job creation and microcredit to those most in need. Some of the island states have focused more on vulnerability than on income poverty to underscore their unique problems. Once vulnerability is taken into account, their level of development generally turns out to be lower and more precarious. Maldives has developed a composite human vulnerability index to underscore the diversity of deprivations across its islands (box 10.2).
With the limits of income and expenditure surveys better understood, there is increasing recognition of the need to supplement them with other survey instruments, such as rapid monitoring surveys and participatory poverty assessments. The issue is how to do this in a way that takes advantage of the strengths, and avoids the weaknesses, of each tool. Benin's Social Change Observatory attempts to implement an integrated system, and other countries are beginning to experiment with similar systems (see the country profile).
Monitoring Human Poverty
Although countries did not explicitly agree to human poverty targets at the Social Summit, they adopted many individual targets, such as reducing child malnutrition and adult illiteracy. These commitments should now be framed more explicitly as reductions in human poverty and used as objectives to guide monitoring.
Adopting New Tools to Monitor Human Poverty
Light, rapid monitoring surveys, such as the one used recently in Armenia, can be used to collect data on human poverty. Armenia's survey, supported by UNDP, covered a sample of 2,000 households, focusing on refugees. Its brief questionnaire drew on UNDP's Rapid Poverty Monitoring Survey as well as the World Bank's Living Standards Measurement Survey and the Armenian Ministry of Statistics' Social Indicators of Poverty project. Using its own version of the human poverty index, the analysis of the Armenian data shows that human poverty affects 36% of the refugee population but only 22% of the local population. A rapid survey to monitor food security and poverty has been used on a regular basis for years in Bangladesh (box 10.3).
Following the lead of UNDP's global Human Development Report, some countries have started using the human poverty index to monitor progress (table 10.1). Uganda is one example. And Mali and Thailand are modifying their survey systems to monitor dimensions of human poverty. Analysing the varying levels of deprivation in the components of the human poverty index yields pertinent policy information (table 10.2).
Many recent national human development reports have produced estimates of the human poverty index. But some have removed variables, such as access to health services, and added others, such as access to electricity. And some have included mortality rates, such as for infants or children, but not standardized them to 100%. The variety in the use of the human poverty index suggests a need for greater understanding of how to define and measure human poverty. UNDP could offer valuable advice on recommended indicators for the human poverty index and technical definitions to guide data collection.
Tracking Human Poverty and Income Poverty in South Asia
The South Asia Poverty Monitor (Bangladesh Institute of Development Studies 1999), a report of the South Asia Poverty Alleviation Programme, uses its own version of the human poverty index quite extensively for analysis, usefully contrasting results for human poverty with those for income poverty and drawing out the policy implications of its findings.
The monitor targets faulty governance as a major reason that poverty programmes have not been more successful in South Asia. Too often, such programmes have been poorly managed, with beneficiaries never receiving the funds intended for them or being allowed to participate fully. The key to success for the South Asia Poverty Alleviation Programme is the poor organizing to take more control of their lives.
The Slow Pace of Poverty Reduction The monitor notes that using a common income poverty line for the countries of South Asia is infeasible. The purchasing power parity method - which is intended to generate comparable estimates across countries - still produces some questionable results. And the national poverty lines cannot be compared because they use different methods.
Even so, some general conclusions about regional income poverty trends are possible. One is that while the proportion of poor has been declining over the past 15 years, the pace of decline has been painfully slow and the trend erratic, with poverty increasing in some periods even when the economy is growing. And except in Sri Lanka, the number of poor people has been on the rise (table 10.3).
The Scale of Human Poverty Although income poverty remains a severe problem in South Asia, human poverty affects a larger share of the population. In reaching this conclusion, the monitor uses its own version of the human poverty index. For deprivation in knowledge it adds an indicator for the percentage of children aged 6 -14 not attending school to the indicator for adult illiteracy. In place of lack of access to health services, it uses the percentage of children not fully immunized combined with the percentage of births not in institutions. For deprivation in public provisioning it adds the percentage of the population not living in electrified houses.
Whether human poverty is higher than income poverty in a country or region depends on how both are measured.In income poverty Pakistan appears to do better than India, but in human poverty India does better. In general, South Asia has made faster progress in reducing human poverty - because of favourable public policies or interventions by civil society organizations - but big shortfalls remain.
South Asia has made very little progress in eliminating illiteracy. India has barely achieved a 50% literacy rate; Pakistan's is still below 40%. Reproductive health in the region is in a dire state, because of neglect of women's health and nutrition during pregnancy and lack of facilities for safe delivery. The incidence of malnutrition among children is still terribly high. These statistics reveal what household-level income poverty data perhaps cannot show: women and children bear a disproportionate burden of deprivation.
Investments and Redistributive Reforms These social outcomes are directly linked to the quality of public policies. If governments paid more attention to how public assets and services are distributed and encouraged the active participation of communities, the deplorable features of human poverty would be eradicated more quickly.
Progress against human poverty can, in turn, accelerate progress against income poverty. Much appears to depend on initial conditions in countries and regions, but initial conditions are invariably due, at least in part, to past public policies. Past investments in human capabilities (education, health and nutrition) and in physical infrastructure (roads, irrigation and electricity) can make a big difference. Successful poverty reduction, as in West Bengal, suggests that reforms in institutional arrangements (tenancy) and in governance (decentralization) can speed progress.
In many cases economic reforms to make markets more efficient do not reduce poverty unless they are combined with redistributive reforms, such as reallocating land or providing more equitable access to education and credit. One reason is that some of the reforms enhancing equity - such as educating the poor, improving their food security or providing them with microcredit - improve the productivity of the poor and encourage them to make productive investments. In these cases equity also enhances efficiency.
Faulty Governance Maintains Poverty The policy implication: dramatic reductions in poverty follow not from market reforms alone, or even necessarily from higher growth rates, but from structural reforms that attack poverty at its roots. What the poor need to bring to markets - even if the markets have been "liberalized" - is more economic power.
Why do the poor lack market power, despite the plethora of poverty programmes over the years? Mainly because of the lack of good governance, particularly of pro-poor governance. The South Asia Poverty Monitor concludes that a significant share of the funds channelled through these programmes has been squandered through corruption or directed to the non-poor. The criticism applies not just to governments but also to many civil society organizations.
Including the Poor in Poverty Assessments
The poor are often excluded from poverty assessments - as they are from poverty programmes. Whether expensive or cheap, rapid or slow, surveys that fail to incorporate the views of the poor are likely to miss the essence of the problem. After all, they are the people most directly affected - keenly aware of the problems they face and probably the most knowledgeable about solutions.
Supporting Participatory Assessments
UNDP has been supporting efforts in many countries to carry out more participatory assessments. In Nigeria it backed the Vision of Development, a survey to determine how people throughout the country viewed poverty, how they suggested reducing it and what their vision of devel-opment was. The objective was to have the government incorporate these perceptions into its development policies. UNDP provided additional technical and financial assistance to enable states to translate people's visions into action plans and budgets for poverty reduction.
In São Tomé and Principe UNDP supported a series of consultations with the poor that revealed how rural and urban poverty differ, which groups need assistance and how best to address their problems. The rural and urban poor had different perceptions of poverty, and groups had different priorities. For example, the most vulnerable group, female heads of household, was most concerned about the lack of productive employment. The government's Strategic Framework to Fight Poverty is incor-porating the results of the consultations to better focus its development efforts on reducing poverty.
Uganda's participatory assessment is very revealing - particularly on governance issues.
Incorporating the Poor's Perceptions in Uganda
In 1999 the Uganda government, with support from donors including UNDP and the World Bank, conducted a participatory poverty assessment in 36 rural and urban communities. The goals: to consult with poor people on the impact of government programmes, ensure that their voices could influence policy-making and build national capacity for such assessments.
The results are revealing - and helpful for crafting useful anti-poverty policies. People saw poverty as multi-dimensional - as powerlessness as well as lack of means to satisfy basic material and social needs. It is also clear that poverty differs by population and locality, so that some interventions need to be carefully targeted and location-specific while others might have a wide application.
The results show that poverty is a dynamic, complex problem that requires multifaceted programmes. It can be seasonal, particularly in rural areas. And it can trap people in vicious cycles in which some forms of deprivation (ill health) reinforce others (lack of income). Because of the specific and changing nature of much poverty, local governments need some flexibility to adjust the general priorities set by the central government to combat it.
Different Groups of Poor People's responses to interviews indicated that the poor could be grouped in three broad categories: The destitute, with no assets and little hope for change. The very poor, with few productive assets and usually socially disadvantaged. The majority of the poor, with some resources and actively seeking to escape poverty.
Interventions would have to be very direct to help the destitute, relatively targeted to reach the very poor, but broader for the majority and relying more on self-help.
Food security is critical - but achievable by providing access to some basic productive assets, such as land, improved seeds and fishing nets, and by supplying credit and access to markets. Declining agricultural productivity and fragmentation of landholdings are prime causes of food insecurity. Facing such conditions, people often have to resort to casual employment to obtain food or the money to buy it.
According to the assessment, people's access to social services is limited and the quality of such services tends to be low. Primary education may be broadly available, but the poor cannot afford the costs of uniforms, materials and required contributions to school funds. Government policies of "cost sharing", it was discovered, tend to exclude the poor from services.
People identified lack of access to markets as one of their major problems. There are few all-weather roads, so poor people cannot get products to market or reach such services as schools and health clinics.
Concern for Good Governance People identified lack of good governance as an overriding problem. They tended to favour decentralization but thought that it was being implemented too quickly and that it put poorer districts at a disadvantage. Disgusted and frustrated with the extent of corruption, people called for greater accountability and transparency in the use of funds. While local council leaders were appreciated, people had a low opinion of higher leaders because they failed to visit and to consult people after elections.
A big complaint: lack of information - on laws, policies, development plans, marketing opportunities and credit availability. People suspected that leaders were deliberately withholding information from them, and they questioned the usefulness and equity of the government's graduated tax system. Taxes were often collected during periods of hardship, not after harvests, and people saw little benefit in return.
Communities in which people cooperate were thought to develop faster and to withstand crises better. Outside interventions - by government or donors - were not always appreciated, especially if they undermined community cohesion and traditions of self-reliance.
