Poverty

National Poverty

Poverty has many aspects of deprivation. People are poor if they do not have adequate resources to buy all the commodities that they need. Similarly, people who lack the ability to live and function properly in society or are more vulnerable to shocks and disasters have less well-being and are likely to be poor. Poverty is also associated with insufficient outcomes with respect to many human development indicators, such as health, nutrition, education and literacy. Poverty is also related to social exclusions and deficient social relations, vulnerability and insecurity, and to low voices, power and self-confidence. Poverty is also a lack of opportunity, and an inability to make use of existing opportunities (NIS, 2006)[1].

Cambodia Socio-Economic Survey 2004 was conducted by National Institute of Statistics (NIS), covering 15,000 sample households across the entire country. The total poverty line was updated by adjusting inflation. At current prices, the total poverty line in 2004 was estimated at 2351 Riel (US$ 0.59) in Phnom Penh, 1952 Riel (US$ 0.49) in other urban areas and 1753 Riel (US$ 0.44) in rural areas. The World Bank and Ministry of Planning use different method from FAO to calculate the food poverty line by translating the food intake into Riel to achieve a level of 2,100 Kcal/person/day; using this method the food poverty line was estimated at 1782 Riel (US$ 0.45) in Phnom Penh, 1568 Riel (US$ 0.39) in other urban areas and 1389 Riel (US$ 0.35) in rural areas (NIS, 2006).

According to CSES 2004, 35 percent of Cambodian population was estimated to live below the national poverty line, and 20 percent below the lower food poverty line. The incidence of poverty was highest in rural areas (39%) and considerately lower in urban areas (5% in Phnom Penh and 25% in other urban areas)[2]. The incidence of poverty was estimated by five geographical zones[3] (Phnom Penh, Plains, Tonle Sap, Coastal, and Plateau/Mountains). The higher incidence of poverty is observed in Plateau/Mountain zone, while the lower incidence of poverty is observed in Phnom Penh (Map 3.1).

For the geographical comparable sample, poverty incidence declined from 39 percent in 1993/1994 to 28 percent in 2004. The poverty incidence declined in all three regions, but reduction rate was not uniform. The rural areas still experience much higher poverty rate (NIS, 2006).

Map 3.1: Poverty Rates

Compared to other developing countries in the region, the percentage of population below the national poverty line in Cambodia is higher than that of other developing countries, such as Phillipines, Indonesia, Vietnam, etc (Figure 3.7).

Figure 3.7: Comparison of Percentage of Population below National Poverty Line

Source: CSES 2004, UN MDG Data

The World Bank Cambodia Poverty Assessment 2006 found that the fall in consumption poverty has been accompanied by an improvement in non-income indicators such as ownership of consumer durables, housing quality and schooling. However, some indicators appear to be deteriorating. Cambodia was  ranked 130th out of 177 countries in the 2005 UNDP Human Development Index and it ranked 57th out of 87 medium human development countries in the Human Poverty Index, which reflects educational, health, and nutritional outcomes as well as access to safe water and health care services.

Despite progress, poverty remains widespread and multidimensional. According to the 2006 World Bank poverty assessment poverty rates are highest in remote rural areas. The rural population, which represents 85 percent of the total, faces a number of problems, including lack of secure land tenure, remoteness from markets and services, lack of productive assets, low levels of education, and high dependency ratios.

Inequality is another concept related to poverty, which focuses on the distribution of an attribute, such as income or consumption across the population. Inequality has risen considerately from 0.35 in 1993/1994 to 0.40 in 2004 for geographical comparable sample. The figures for inequality in 2004 are actually more serious when the full sample is used. In 2004, the Gini coefficient stood at 0.42 (Gini of 0 represents perfect equality, and Gini of 1 represents perfect inequality).

The lessons to be drawn from international experience are open to debate, as findings are highly dependent upon the choice of countries, reference periods and analytical technique. Until perhaps the last five years the consensus in international development theory was that: (i) inequality inevitably rose with economic development and the transition from “traditional” or socialist society to a market-based society; (ii) however, changes in inequality, in either direction, occurred only slowly; (iii) growing inequality was of little practical significance so long as absolute deprivation (i.e. poverty) was falling; and (iv) there was little that public policy could do to control growing inequality, certainly not without impeding growth rates and poverty reduction (World Bank, 2006).

Figure 3.8: Inequality in Cambodia

Source: CSES 2004

Small Area Estimation Technique

Small areas estimation technique was developed by Elbers et al. (2001) in order to make estimates at a micro-level from census-based secondary data that are normally available on a whole country basis. The core of the method is to identify the predictive indicators from the census survey information that could explain the information obtained from household surveys. A household survey usually collects very detailed information on household characteristics, including consumption level, but the sample coverage is generally limited and only representative over a relatively large geographical area. On the other hand, a population census has a complete coverage of all households, but usually collects very limited information on household characteristics. Hence, this method tries to combine the advantages of detailed information on household characteristics obtained from household surveys with the complete coverage of a population census.

In 2002, the United Nations World Food Programme (WFP) in close collaboration with Ministry of Planning conducted a joint study of poverty mapping at commune level using the small area estimating technique.

Existing data from the Cambodian Socio-economic Surveys (CSES) allow for the estimation of poverty indicators, but this is only accurate over large geographical areas. Thus, pockets of poverty or poor areas surrounded by non-poor areas cannot be identified from the socio-economic surveys alone. The surveys do not provide for a reliable estimation of poverty at the provincial, district or commune levels.

It is often the case that what policy makers really need is information at geographically disaggregated levels. Therefore, poverty maps have been created by combining data from the 1997 socio-economic surveys, the 1998 Cambodian Population Census, and other geographic data compiled with a GIS. Using a recently developed statistical technique “Small Area Estimation”, the strengths of these various datasets were combined to generate poverty estimates at geographically disaggregated levels thus providing more useful information.

The mapping exercise uses consumption (the monetary value of the goods and services consumed by a person) as the measure of poverty. Data is taken on household consumption from the socio-economic survey, which is combined with data on household and village-level characteristics from the population census and GIS data. Statistical analysis is applied to these data to estimate a relationship between consumption and household and village-level characteristics. This relationship is known as the consumption model and is used to generate estimates of consumption for large numbers of households across the country.

Next, a given level of consumption is chosen as a poverty line. Households whose consumption is below this level are considered poor, and those with consumption above it are considered non-poor. From here, standard poverty indicators, such as poverty incidence, poverty severity, and poverty gap can be calculated. The calculations also produce measures of error for the estimates, which can be used to assess their validity. Finally, maps are generated to show the geographical distribution of these indicators across communes, districts and provinces (Map 3.2).

Map 3.2: Poverty Rates at Commune Level 1998

Rural - Urban Bias

Of the total number of the poor, more than 90 percent live in the rural areas. This implies that Cambodia’s poverty is rooted in its large agricultural sector, which has low productivity and low growth, but provides livelihood to the vast majority of the country’s population. As well as living in rural areas, the poor tend to have low levels of education and limited access to land and other productive assets. This results in households depending on low paying manual employment (e.g. daily labor).

In both urban and rural areas, the poor have less access to modern amenities and services. They reside in houses of inferior quality with no or limited access to basic services. The poor are more likely to reside in households with larger membership sizes, have more children, and have a household head that is less educated. They also have much less access to public services.

Rural poverty probably has declined at a much slower rate than poverty in Phnom Penh or other urban areas (UNDP, 2007). The significant fall in poverty in the bigger cities has been due to strong urban bias in growth and concentration of public investment. Cambodia’s economic success has been a largely urban phenomenon with the primary drivers – garments, tourism and construction – having few linkages with the majority of the population, who depend on agriculture as the main source of livelihood. Figure 3.9 and 3.10 compares poverty in Phnom Penh and in rural areas of Cambodia.  Though poverty in Phnom Penh has decreased more than 50 percent in the period 1994 - 2004, rural poverty rates as seen no similar improvement.

Figure 3.9: Gini Index

Source: World Bank Cambodia Poverty Assessment and East Asia Update 2006

Figure 3.10 : Population below the poverty line

Source: World Bank Cambodia Poverty Assessment and East Asia Update 2006

Consumer Price Index

The national consumer price index (CPI) indicates an upward trend of food prices from 2000 to 2005. On average, food prices increased faster than the overall inflation rate since 2003. This can be explained as a result of production shortfalls in several provinces caused by drought in 2004 and 2005, as well as the inflationary effect of the distorted trade regime. As a result, market prices for rice have almost doubled, implying that informal and official imports have only a limited stabilizing effect on food prices.

Figure 3.8. Cambodia, Annual Average Inflation Rates (2000-2005, in %)

Source: National Institute of Statistics

In a normal year, rice prices would generally increase from 600 to 900 Riel/kg during the dry season (April-May) and the lean season (August-November) and decrease from December to March (harvest and post-harvest seasons). However, in years of significant production shortfall, the rice price increase can fluctuate between 1,000 and 1,500 Riel/kg, especially during the lean season.

According to NIS data covering Bat Dambang, Kandal, Kampong Cham, Siem Reab, Sihanoukville and Phnom Penh, the food price increase is reflected by rice price increase (in real terms) at provincial level over the last five years (Figure 3.9). The real price level of rice is low in Bat Dambang compared to other provinces because of its surplus production status. However, the increase of the real price (of both the second quality rice and the first quality rice) is one of the highest in Bat Dambang. As can be seen below, the price increase was higher in the drought years of 2004 and 2005. In 2006, the real price increase was maintained despite a relatively good rice harvest compared to the previous two years.

Key informant interviews held by the IPC mission in Kampong Chhnang and Kampong Speu indicates that production shortfalls may have led to increased dependency of small farmers (net consumers with less than half a hectare) on markets. The poorest households are likely trapped in a vicious cycle of debt as they borrow rice from traders with a 100 percent interest rate pay back these loans, in cash or in-kind, after the next harvest season. Following consecutive years of drought and production shortfalls, such a practice increases vulnerability and reduces resilience to shocks.

Various reports indicate that rice production was negatively affected by drought in 2004 and 2005. A compilation of MAFF data suggests Kampong Spueu, Rotanak Kiri, Mondol Kiri and Kracheh provinces were the most adversely affected, with an average of 10 percent of rice cultivating areas destroyed by drought from 2004 to 2006. A drought assessment conducted by OXFAM in 2005[4] concluded that rice production was also severely affected by drought in Kampong Spueu, Svay Rieng, Prey Veng and Takeo. Reportedly, farmers with small plots of land only produced enough rice for 2 to 3 months in 2005, down from an average of 7 to 9 months in a normal rainy year.

Figure 3.9. Real Retail Price of Rice (yearly average 2002-2006, in constant terms of 2000)

Source: National Institute of Statistics (NIS) price data

Looking at prices of rice in Cambodia and comparing between spatial price differences and transport costs suggests inter-provincial trade is constrained by high transfer costs (Table 3.2). Aside from agricultural productivity, the main costs explaining the difference between surplus production areas and deficit production areas, i.e. consumption markets, are transport costs, unloading costs, processing costs, interests to be paid on loans, margins and losses[5]. Transport costs are the main component of transfer costs. In the absence of other cost components, it is considered as a good proxy indicator of transfer costs. Considering Phnom Penh as a major distribution center to provincial markets, the spatial price difference (between Phnom Penh and the province) is compared with WFP transport costs of rice from Phnom Penh to other provinces. The table below indicates high price correlation, but higher transport costs are more significant than the spatial price differences between Phnom Penh and other provinces, especially for the lower quality rice consumed by poor households. This implies that the inter-provincial rice trade is constrained by high transfer costs. However, annual average price differences hide seasonal variations. Transfers between markets can be cost effective when the price gaps exceed transport costs, during some seasons of the year. However, addressing this seasonality issue would require an analysis of monthly transport costs. Furthermore, market integration analysis requires greater provincial coverage of price data and transport costs. Limitations in available data, both by season and by province, do not permit such analysis at this time.

Table 3.2. Spatial Price Differences Compared to Transport Cost (Riel/Kg)

Source: National Institute of Statistics price data and WFP transport costs, Author’s estimates.

Household Consumption

According to CSES 2004, the mean per capita per day household consumption in Cambodia was 3,606 Riel in 2004 at average 2004 Phnom Penh prices (NIS, 2006). Average levels of per capita consumption vary greatly between regions and between urban and rural areas within regions. Figure 3.10 presents estimates of per capita household consumption by agro-ecological region, disaggregated into the urban and rural areas within each region. The estimates of mean per capita consumption are expressed in 2004 Phnom Penh prices. The urban Phnom Penh population clearly enjoyed the highest real per capita level of consumption, having an estimated level well over twice the national average. The rural areas of Tonle Sap and Plateau/Mountain regions had the lowest levels of real per capita consumption. These two poorest rural areas account for over one third of the country’s population, i.e. 26 percent in Tonle Sap and about 9 percent in Plateau/Mountain (World Bank, 2006).

Figure 3.10 : Average Level of Per Capita Consumption by Regions

Source : CSES 2004

Expenditures

Based on an OXFAM 2007 survey, the below were the common expenditures typically incurred by rural households.

Food: Among the better off and middle households, expenditure on food does not change much between a typical and a bad year. The better off usually purchase fish, pork,eggs, vegetables, oil, sugar, fruits while the poor household’s expenditure is mostly used to purchase rice, vegetables and fish.

Farm inputs: The better off and middle households purchase pesticides for rice production in addition to ammonium phosphate and urea fertilizers.

Utilities:  This expenditure mainly consists of purchasing firewood, kerosene, batteries and battery charging.

Transport and telephones:  The better off and middle households spend money on fuel/gasoline for their motorbikes. The expenditure for poor households is on using public transport facilities.

Health and hygiene: For all groups the money is spent while paying for medicines especially from the localvillage stores.

General Economic Characteristics

The Cambodian economy has experienced strong growth in recent years. Cambodia became the second LDC to accede to the WTO, in October 2004. Overall economic growth averaged nearly 8 percent between 2002 and 2006, while exports increased from US$1.8 billion to US$3.0 billion over the same period[6]. Similarly, foreign direct investment (FDI) expanded from US$139 million to US$394 million[7]. While this economic performance is certainly encouraging, it would be misleading to assume that such trends indicate an unqualified parallel with the export-led development miracles in neighboring countries. Cambodia currently has a trade deficit of approximately US$1.3 billion, and total outstanding debts in excess of US$2.2 billion or 34 percent of GDP[8]. However, with a short-term debt load of US$220 million and foreign exchange reserves in excess of US$1.0 billion, it is not currently vulnerable to the type of currency-driven crisis that affected East Asian economies in the late 1990s. According to the IMF, general inflation is projected to remain stable at about 3.5 percent over the next three years (2007-2009)[9], but volatility in global oil prices may exert upward pressure on the inflation rate. Exports from the garment sector have been rising rapidly in recent years, increasing by over 10 percent to US$2.2 billion in 2005[10]. Most of these exports are destined for the US and EU markets and are highly dependent on trade quotas placed by these markets on Chinese manufactures. Tourist arrivals increased by 35 percent to reach a total of over 1 million visitors in 2004.

As the World Bank notes, most of the economic activity related to tourism and the garment industry is concentrated in Siem Reab and Phnom Penh. Therefore, the major impacts on the rural economy are in the form of remittances. Recent studies indicate that only 13 percent of rural households received such remittances.[11] Economic concessions have recently been given to explore for petroleum, natural gas, gold, and bauxite in Cambodian territory. The economic opportunities represented by the possible discovery and exploitation of these resources must, however, be balanced with realistic expectations related to corruption and the ‘resource curse’ experienced by other developing countries.[12] Official employment data, which shows the unemployment rates at 1 percent or lower, cannot be relied upon or, due to the nature of definitions for employment used in Cambodia, used for comparison with other countries. However, data from the 2004 Inter-Censal Population Survey suggests that 7 percent of the population is unemployed. This figure counts those involved in agriculture and other seasonal occupations as employed, however, suggesting very high rates of under-employment. The same data indicates that 74 percent of the working population is employed in the primary sector, including agriculture, forestry and fisheries; 7 percent employed in the secondary sector, which comprises mining, manufacturing, and construction; and 18 percent in the tertiary sector, which includes services and the public sector[13]. Seasonal migration is common among agricultural households, with many migrants crossing into neighboring countries to find wage labor opportunities.