Poverty in Yemen is deeper and more severe than in other MENA countries, report [Archives:2007/1113/Reportage]
A recent report on poverty assessment in Yemen said poverty rate dropped from 40 percent in 1998 to 35 percent in 2006. Titled 'Yemen Poverty Assessment', the report was prepared by the government of Yemen, World Bank, and UN Development Program (UNDP). This report was based on a household budget survey conducted from April 2005 to March 2006.
Below is an executive summary for the report.
From what was historically known as “Arabia Felix” – a land of prosperity and happiness – Yemen has become the most impoverished among the Arab countries.
The government of the united Yemen, formed in 1990, has launched so far three five-year economic reform plans with the goal of restoring Yemen's prosperity. Have these efforts succeeded? What policies are needed to further reduce poverty? The Poverty Assessment report aims to answer these questions. This report measures poverty in Yemen in 2005/06, and evaluates the change in poverty compared to 1998, the two years for which comparable household budget surveys are available. The period between the two survey years (1998 and 2005/06), more or less overlaps the first two five-year economic plans and captures the effect of the economic reform programs launched since 1995.
In addition to measuring poverty, this report has three objectives: evaluating the role of growth and past reforms on poverty, identifying better ways to target the vulnerable poor through public action, and an assessment of the poverty monitoring system. By examining the effect of the key policies on poverty, such as the petroleum price reform and the government's social protection mechanisms between 1998 and 2005/06, the study aims to equip policy makers and development partners with the knowledge needed to improve the effectiveness of their efforts to reduce poverty in Yemen.
The main conclusion of the study is that poverty in Yemen declined.
Real GDP per capita grew, on average, by 2.1 percent per year between the two survey periods, while poverty (measured as the headcount ratio) declined by almost 2 percent on an annual basis. The percentage of poor declined from 40.1 percent in 1998 to 34.8 percent in 2005/06. The percentage of the poor declined in both urban and rural areas. However, due to continued high population growth, the absolute number of poor remained at around seven million, the same number as seven years ago. The pace of poverty reduction is modest compared to the MDG goal adopted by the government. The goal of reaching the first MDG (halving the percentage of poor) will require the quadrupling of the current one percent per year rate of growth in per capita consumption achieved over the last seven years.
The urban areas of Yemen witnessed a remarkable decline in poverty.
The urban areas benefited remarkably from the predominantly oil-led growth, which resulted in the rapid decline in the percentage of urban poor from 32.2 percent to 20.7 percent, despite an increase in urbanization. Although poverty declined in four of the seven urban regions, poverty did not change in the remaining three.
The decline in national poverty for the rural areas is not robust to changes in poverty lines. In rural Yemen, the percentage of poor declined from 42.4 percent in 1998 to 40.1 percent in 2005/06. However, unlike in the urban areas, this decline is not robust to alternative definitions of poverty line. If slightly higher poverty lines were used than the already chosen ones (higher by about 12 percent), the percentage of poor in the rural areas would have increased between 1998 and 2005/06. The decline in poverty as measured by the poverty gap and severity measures is also reversed at slightly higher poverty lines. However, examined at regional levels, the change in rural poverty becomes robust. In three of the seven rural regions where nearly 40 percent of the country's poor live (Central North, Central South and Eastern1), poverty unambiguously worsened by 10 to 15 percentage points.
The distributionally sensitive measures of poverty show a greater decline in poverty. During 1998 to 2005/06, the poverty gap index declined at a greater rate than the headcount index, and the severity of poverty index fell at an even greater rate. This indicates that the ultra poor benefited even more than the average poor.
Despite the recent reduction, poverty in Yemen is deeper and more severe than in other MENA countries. The poverty gap index is 8.9 percent, implying a monthly poverty deficit per capita of about YR 497. On average, a poor person should receive YR 1,431 per month to be lifted out of poverty2 – about one-third of the mean consumption of the poor. Perfect targeting of the poor would require only about YR 124.4 billion per year (about 4 percent of GDP) to fill the gap between the actual spending of poor households and the poverty line, thus lifting everyone out of poverty. The severity of poverty index (which attaches greater weight to the poverty gaps of poorer families) at 3.3 percent is relatively high by MNA country standards (Figure E.2). The food poverty gap averages about YR 2,100 for the food-poor, some 75 percent of the average consumption of the food-poor. Meanwhile the government's cash transfer program which targets the food-poor (the Social Welfare Fund) is inadequate as it currently has a benefit ceiling of YR 2,000 per family.
Poverty differs considerably across regional dimensions. For the first time for Yemen, utilizing the 2004 census data and the 2005/06 HBS, poverty at the district levels was projected with statistical precision. There are large intra-governorate differences in the incidence of poverty (Figure E.3). Poverty varied between 5.4 percent and 71 percent in 2005/06 among governorates. Poverty is highest in the rural part of the Amran governorate, where 71 percent of the population is poor. Amran is followed by Shabwah and Al-Baida (60 percent). The incidence of poverty is the lowest in the Al-Maharah and Sana'a City governorates. The ranking of governorates remains unchanged for other measures of poverty. Because of the faster reduction in urban poverty, regional differences in the rates of poverty have accentuated between 1998 and 2005/06.
Educational attainment of the poor has improved, and the poverty incidence has
marginally declined among illiterate families. The share of poor households headed by a person without any formal education declined from 86.8 percent in 1998 to 68.3 percent in 2005/06. The illiterate dominate among those without formal education. The incidence of poverty among families headed by an illiterate person declined from 47.3 percent in 1998 to 44 percent in 2005/06. The lowest poverty rate was found among households headed by persons with university degrees and above, although puzzlingly the difference between the poverty rates of university graduates in urban and rural areas is very large: 5 percent versus 29 percent. This difference could arise from inefficiencies of the labor market such as costs of job-search. Poor workers receive 10 percent less return on education than the non-poor as they probably cannot secure better paying jobs which require better social connections.
However, at a time of rapid surges in enrollment, poor children are falling behind. The enrollment rates for children in the age group 6-14 in the poorest decile declined for both urban and rural areas by 5 percentage points, while the richest deciles increased enrollment by 1 percentage point in urban areas and by 21 percentage points in rural areas. Unavailability of schools and difficulty in commuting are the obstacles that keep children out of school. Other reasons given for non-enrollment include a lack of interest by the family, and/or having to work.
Poor families spend less on health expenditure as a share of their total expenditure compared to the non-poor. Inability to pay and difficulty of access may be discouraging them from utilizing medical care. Out-of-pocket payments for health services as a share of total expenditure has increased between 1998 and 2005/063. Households in the poorest quintile spend about one-half what the richest quintile spends as a share of their total consumption expenditure. This distinct pattern suggests that the poor in Yemen may be discouraged from utilizing medical care since they cannot afford it. The inability to pay may contain demand for medical care among the poor.
Children from poor families suffer more from malnutrition. Child malnutrition remains a concern in Yemen, as nearly one-third of children between two and five years old are severely stunted4. Data on severe stunting shows a greater disparity between urban and rural children than other types of malnutrition (Table E.1). Poverty is clearly associated with the prevalence of severe stunting and underweight among Yemeni children5. Since the WHO guidelines for the international reference population changed, it was not possible to compare the prevalence of child malnutrition with the results of earlier health surveys. On the other
hand, available data suggests that the prevalence of child malnutrition did not decline over the last few years.
More people are seeking medical treatment when sick, but the gap persists between the rich and poor. In the 2005/06 HBS, the percentage of individuals who were ill during the preceding month of the survey and sought treatment ranged from 56.6 percent in the poorest quintile6 to 79.7 percent in the richest quintile. Compared to the 1998 HBS, the 2005/06 HBS figures show that proportionately more of the poor population seeks medical care when ill; however, there still exists a persistent gap between the poor and the rich.
High health cost is the single most important reason why the poor do not seek care when they are sick. For the poor, inability to pay for medical care is the most significant reason for not seeking treatment. Unavailability of needed medical service and difficulty in physical access were the second and third reasons for not seeking the medical care by the poor7. Inability to pay for medical care and unavailability of care become less significant barriers as the expenditure level goes up. In the richest quintile, the major reason for not seeking care is that the illness was considered too minor to receive the medical attention.
Using self-reported status, the poverty rate of female-headed households in Yemen is not significantly different from male-headed households. Only 5 percent of the population in Yemen lives in households headed by a female. These households represent 8 percent of all households. While the mean poverty incidence for self-declared female-headed households is 32 percent versus 35 percent for male-headed, this difference is not statistically significant, as the sample of female-headed families produces a high standard error.
Among female-headed households, there is heterogeneity in risk of poverty. Poverty incidence is highest among those female-headed households which have a spouse residing in the same household. In most cases these husbands are physically or mentally disabled. Next in incidence are the households where the husbands are away. Widows and divorced women headed households rank next in poverty risk. The lowest risk (7%) is for households headed by single women who have never married.
Female-headed households allocate resources better than male-headed households. Controlling for other factors in a typical demand system, the HBS data established that female headed households (self-reported and de-facto) allocate resources better for current welfare and for the future. Households that are headed by women have lower shares of consumption on adult-consumption goods such as tobacco and qat, and spend more on education8. This applies both to households headed by single women, and households where the male head of the family is mostly absent. Having a female as head of household increases the share of expenditure on food by 1.2 percent, while the demand for non-essential goods such as tobacco and beverages declines by 10 percent.
THE IMPACT OF GROWTH AND REFORMS ON POVERTY AND UNEMPLOYMENT
The decline in poverty rates observed in Yemen between 1998 – 2005/06 was largely a result of oil-driven economic growth. However the oil sector employs few Yemenis, and fewer if any from the poor population9. The spending impulse from government expenditure expands mostly the service sector based in urban areas.
Oil-based growth does not benefit the poor, especially the rural poor.
Empirical evidence has shown that oil rents do not trickle down to most of the economy, in particular the poor. When income level is controlled, countries which have high levels of natural resource endowment tend to have higher rates of poverty and Gini coefficients than their non resource rich peers. The oil sector does not contribute to employment among the rural poor population, being a highly industrialized sector demanding skilled labor, which is mostly imported.
The direct impact of tariff cuts has been minimal. Trade liberalization is generally considered to be beneficial for economic development due to long-term efficiency gains, but when applied to import-competing sectors, it can have immediate adverse effects on the poor. During the structural reform periods of the 1990s, Yemen's unweighted average tariff decreased from 20 percent in 1996 to 12.6 percent in 200311. However, as Yemen is without a strong initial industrial base, the country is unlikely to have been adversely affected by import tariff cuts.
The devaluation has probably contributed the most to transitional adverse affects on the poor. Between 1998 and 2005/06, the Yemeni Rial depreciated by 41 percent.International evidence points to an average pass-through coefficient of 0.712, implying that up to one-third of the observed price increase of 89 percent over the period could be explained by depreciation. But devaluation improves the incentives in the export-oriented sectors, and the expansion of these sectors could benefit the poor. Partly in response to the depreciation, non-oil exports in Yemen expanded more than two-fold in nominal US$ terms between 1998 and 2005.
The partial removal of petroleum price subsidies between 1998 and 2005 may have resulted in a cost of living increase of around 21 percent. Petroleum prices have been highly subsidized in Yemen, the highest in the MNA region, representing up to 9 percent of GDP. Faced with the prospect of dwindling oil reserves, and thus declining fiscal revenues, the authorities decided to reduce petroleum sector subsidies. Increased disbursements from the Social Welfare Fund were meant to mitigate any adverse effects.
The increase in the labor force out-paced the growth in jobs during 1999 – 2004.
The labor force increased at a rate of 4.3 percent per year, while the number of jobs increased by 3.7% per year. The labor force participation rate increased slightly from 38.5% to 39.2% between the two reference years. Therefore the rate of unemployment increased slightly by 2.6 percentage points. Youth unemployment (ages 15 – 24) is considerably higher than the average general population, at 28.3%.
Unemployment increased between 1999 and 2004. Unemployment in Yemen
(ages 15+) increased from 13.7 percent in 1999 to 16.3 percent in 2004, even though there was a significant decline in poverty (from 40 percent to 35 percent approximately) .. According to international evidence, there is no clear relation between poverty and unemployment. While micro-economic analysis may show a positive correlation between unemployment and poverty, from a macro-economic standpoint, poverty is a household phenomenon and unemployment is related to the individual. An unemployed individual may not be poor if there are other income earners in the household. Moreover, poor households cannot afford to be unemployed, but likely earn very low wages that perpetuates the cycle of poverty. Similarly, changes in policy can cause divergent outcomes for poverty and unemployment. For instance, a significant reduction in real wages may create more jobs, but lower incomes would create higher poverty.
Unemployment rate among women has increased during 1999 – 2004.
This is in part because female labor force participation increased from 7.0 to 9.6 during 1999 – 2004. The female unemployment rate also increased from 25.4 to 39.5 percent in 2004. The high level of female unemployment is slightly more evident in the urban areas, where nearly half the female labor force is unemployed. In comparison, in rural areas, female unemployment is 33 percent. There is also a large difference between male and female unemployment rates. The ratio of male to female rate of unemployment is approximately 3:1, similar to that in Egypt
Social spending in Yemen has decreased in the recent past to 7 percent of GDP.
Social spending, defined to exclude petroleum subsidies, has declined from 8.6 percent of GDP in 2003 to 7 percent in 2006. Health spending is only about 2 percent of GDP. In fact, petroleum subsidies alone equal or exceed total social spending.
POVERTY MONITORING
A coherent and functional poverty monitoring system does not currently exist in Yemen, despite multiple donor efforts.
Most monitoring and evaluation is done in response to varying donor agendas. Moreover, the information is not available publicly. Competing demands for information exist under the PRSP, TFYP, and MDG frameworks, leading to poor formulation of objectives and indicators. Past inefficiencies exist in the system, such as the lack of coordination and duplication of work at various levels of MOPIC, which have hampered growth and poverty alleviation programs. Even after three years of initiating M&E, monitoring systems are still not completely functional. Discussions regarding the list of most important indicators are still ongoing, and progress is very slow. In addition, there are also concerns regarding implementation capacity.
But, improvements to the monitoring system are planned.
A successful strategy for M&E must focus on the essential aspects of poverty monitoring and capture primarily the level of outcomes, as well as scrutinize the input side (the budget, its execution, and medium-term financial planning). It should include surveys that can assess the effectiveness of policy measures on the ground. It should be built upon the existing administrative information systems in the sectors, but it should also critically examine and validate their results. The new Third Five Year Plan (TFYP) recognizes these deficiencies and prioritizes poverty monitoring as a major objective under the TFYP. Proposed restructuring of MOPIC and reassessing the role of all stakeholders, including line ministries and donors, should lead to improvements in the monitoring and evaluation processes.
Greater civil society involvement is the key to success in monitoring.
The involvement of the parliaments and civil society in the discussion about the achievements and successes in implementation of the PRSPs is so poor, that in effect, neither accountability nor dialogue is possible amongst the governments, the parliaments and the citizens. Since the objectives of the poverty monitoring process include a qualitative change in the political culture and a more intense dialogue between governments and civil society, considerable efforts are necessary in order to maintain the dynamics of participation that emerged during the formulation of the PRSP.
——
[archive-e:1113-v:15-y:2007-d:2007-12-20-p:report]
