Get Adobe Flash player
Home Statistics Surveys Poverty in Lao PDR 2008

Poverty in Lao PDR 2008

Contents

Poverty in Lao PDR 1992/3-2007/8

I. Summary

II. Background and Methodology

III. Poverty Trends

IV. Poverty Patterns

V. Sensitivity of Poverty Estimates

VIII. Food Poverty and Insufficiency

VI. Asset Ownership

VII. Trends in Inequality

IX. Access to Public Services and Infrastructure

XI. Conclusions and Summary

XII. References

Preface

The Lao Expenditure and Consumption Survey (LECS) is the most comprehensive household survey conducted in Lao PDR and forms the basis for official poverty estimates. The 2007/8 survey presented in this report is the fourth; the first was conducted in 1992/3 and the additional surveys in 1997/8 and 2002/3.

The findings presented in this report are based on data obtained from sample villages and households with results that are nationally representative for Lao PDR. This survey follows the methodology used for the 2002/3 survey, which is far more comprehensive in coverage than the first two surveys. The survey gathers a wide spectrum of information on the socio-economic status of Lao households, communities and regions. This wealth of information is a key contribution to understanding the development challenges facing Lao PDR.

This report was written within the World Bank supported project on Strategic Statistical Capacity Development. We would like to take this opportunity to express our thanks to all persons engaged on different levels in the survey preparations and operations, and further to the sample communities, village chiefs and households, for their cooperation throughout the time of the survey. We would also like to express our gratitude to the Swedish International Development Agency, Sida, which has contributed funding and technical assistance to the survey.

The work was carried out under the overall direction of Director General Samaychanh Boupha, Head of Division Vixay Santivong, and Deputy Director General Phetsamone Sone. Mr. Somneuk Davading coordinates the World Bank contributions to the Statistical Capacity Building project. Project coordinator Vithanom Keokhounsy, provided support throughout the project and the report benefited from contributions by Kingphet Atsanavong, Santika Chanthalavong, Manyvone Xiengtham and many other DoS staff members. The authors have benefited from technical comments from Hans Pettersson, Statistics Sweden.

I.Summary

At the time of the first LECS survey in 1992/3, 46.0% of the Lao population was unable to afford basic needs. Over the next 17 years economic growth and development helped lift over 500,000 people out of poverty. The drop in poverty over the most recent 5 year period (2002/3-2007/8) has been impressive, from 33.5% to 27.6%. This drop in consumption poverty has been accompanied by a swift expansion in household asset ownership. Assets such as motorbikes and mobile phones link previously remote households to infrastructure and information and are bringing enormous changes to the lives of rural Lao residents. Access to social services and infrastructure also continues to increase. The improvements in access to roads, clean water and sanitation are particularly impressive. Education opportunities have expanded and more than eight in ten children between the ages of six and ten are enrolled in primary school.

However, significant challenges remain. Urban areas and districts along the Thai border have experienced rapid growth and poverty reduction, but other groups continue to lag behind.  The northern part of the country remains poorer than the southern and central regions. Upland areas and priority districts are poorer than lowland and non-priority districts, and poverty reduction in these areas has been relatively slow. Poverty rates vary according to ethnicity, with the Lao-Tai displaying lower poverty incidence. Although government support has been aimed at these groups and regions, they have benefitted less from processes of economic development. Despite improvements, lack of access to infrastructure and markets remain barriers to growth and poverty reduction.

Meeting the needs of these groups for access to services also remains a challenge. Enrollment rates among ethnic minorities and those living in remote rural areas remain far below the national average. Furthermore, improvements in access to healthcare over the past 5 years have been slow. Many households still need to travel long distances on poor roads in order to access healthcare.

Finally, the finding that food-poverty in Lao PDR has increased in both rural and urban areas, reversing the trend observed in earlier surveys, is worrying. Households below the food poverty line are not necessarily unable to meet basic nutritional needs, and evidence from village-level estimates of rice insufficiency and individual-level consumption indicate that food consumption has remained stable or increased slightly. However, given the high level of malnutrition documented in other surveys, and the potential impact of higher food prices, the food security and nutritional status of poor households deserve close attention and further research.

II. Background

II. Background and Methodology

Since the early 1990’s, Lao Peoples Democratic Republic has conducted four national household surveys. This report analyzes poverty and inequality in Lao PDR for 2007/8, and assesses changes during the past fifteen years. The distribution of poverty across geographical regions, provinces and socio-economic groups is presented, together with trends since the first household survey in 1992/3.

Amartya Sen defines poverty as anything that severely decreases a person’s capabilities (Sen, 1999). Poverty is a state of deprivation involving many dimensions, from limited income to vulnerability in the face of shocks and lack of access to social services. This report therefore analyses consumption poverty from a quantitative perspective, but also examines other indicators of deprivation, such as access to healthcare and education. The methodological approach follows several earlier studies using nationally representative LECS household surveys (including World Bank, 2006 and Kakwani et al., 2002).

There are certain limitations to taking a quantitative approach. These include the need to distance the analysis from the experiences of families and individuals and the limits to the understanding of complex causal processes. Qualitative work, such as the two participatory poverty assessments carried out in Lao PDR (ADB, 2002; Chamberlain, 2007) can provide an alternative and complementary perspective to this quantitative analysis.

The quantitative approach gives nationally representative estimates of poverty levels and trends and allows disaggregated analysis across different categories of households and across different regions. In a country with great regional variation, such as Lao PDR, this ability to make disaggregated poverty estimates is important for grasping the development challenges facing policymakers, identifying the factors that are associated with lower or higher poverty incidence.

Making quantitative poverty measures involves making methodological choices that significantly affect the results. Awareness of the methodological issues is essential for interpretation. The following sections give detailed descriptions of the LECS surveys and the methodology employed for measuring poverty in this report.

The LECS Surveys

Lao Expenditure and Consumption Surveys (LECS) have been collected every five years since 1992/3. The surveys are nationally representative surveys of consumption expenditure and a wide range of socio-economic information. Information is collected on individual, household and village levels. Poverty estimates for the country can be derived based on LECS, as well as welfare profiles for different socio-economic population groups. All LECS surveys have been carried out by the Department of Statistics[1] (DOS) with donor support.

The LECS surveys are collected over 12 month periods from March to February to control for seasonal variation in a largely agricultural society.[2] The month of collection is randomly assigned to each village. The sample frame from which respondents are selected covers the entire population of Lao PDR. LECS relies on a stratified and clustered sample. For stratification each province is divided into urban areas, as well as rural villages with and without wet season road access. For LECS IV 518 villages served as primary sampling units or clusters.[3] All of the villages visited in 2007/8 had earlier been surveyed in 2002/3. In each village 16 households were interviewed, half of which had been surveyed in the earlier LECS III. A total of 48,021 individuals in 8,296 households were interviewed.[4]

Table 1: Sample Sizes, LECS1-LECS4

LECS1 1992/3

LECS2 1997/8

LECS3 2002/3

LECS4 2007/8

Villages

147

450

540

518

Households

2937

8882

8092

8296

Individuals

19574

57618

49789

48021

The range of topics covered by the LECS surveys has broadened over time. The initial 1992/3 LECS survey focused on consumption and social indicators. The 1997/8 LECS expanded the coverage of economic issues. The 2002/3 and 2007/8 surveys are broadly similar to each other and collect information not only to measure household income and consumption, but also to investigate a wide range of poverty-related issues. [5] In addition, a village price survey collected information about prices in the local market for most villages covered in the survey[6]. This information is used to construct price indices that capture differences in costs of living within and between urban and rural areas.

Measuring Poverty

This analysis follows common practice in developing countries and earlier LECS analyses and uses a consumption-based welfare measure. While both income and consumption have advantages as measures of welfare, consumption is often preferred for both theoretical and practical reasons (Deaton, 1997).

The consumption based approach is reflected in the LECS questionnaire, where detailed information on household consumption expenditure is recorded. A household diary is the principal instrument for collecting expenditure on a detailed set of food and non-food items. The diary is filled in daily by household members, assisted as necessary by enumerators. All consumption is recorded for a 30-day period, including both cash expenditure and the in-kind value of own produced items consumed. Additional information on purchases of certain high value or durable items is collected for the last 12 months.

Household consumption is defined as the sum of food and non-food consumption. Non-food consumption refers to consumption of all non-food items, except 12 durable goods[7] and rents (cash or imputed). For some high value items the annual purchase value, divided by 12, is included in the consumption aggregate. [8]

Rice, being the most important food staple for most Laotians, is particularly important for deriving accurate food consumption. Some households recorded zero, or very low rice consumption in the household diary. Given the importance of rice in the Lao diet, these figures are not regarded as plausible. Because the household diary reflects only a 30-day recall period, it is probable that these households were consuming rice they had purchased in previous months and stored. Rice consumption for these households was therefore adjusted using information from the separate nutrition module, which records daily rice-intake.

Per capita normalization is used to derive individual consumption needs, whereby consumption requirements of the young and elderly are assumed to equal those of adults. This rules out scale economies in consumption and treats all members as adult equivalents, implying that household needs increase proportionally with household size. This assumption is motivated by the absence of credible adult equivalence parameters for Lao PDR (Kakwani et al., 2002). This practice is consistent with previous LECS surveys and allows for simple and transparent interpretations (Kakwani et al., 2002; World Bank, 2007).

The Poverty Line

Measuring poverty relies on introducing a poverty line. Persons with welfare below this line are defined as poor. The poverty line used in this report was originally defined for 1997/8 (LECS2), based on a calorie requirement of 2100 calories per person per day, plus an allowance for non-food consumption[9]. Consistency between poverty lines used for the LECS surveys is central to enabling comparison of poverty rates between rounds.  However, the cost of living changes over time. Appendix I explains in more detail how the LECS2 line was derived and updated to 2002/3 price levels. This study updates the 2002/3 poverty line used by Richter et al. (2005) based on changes to costs of living over the past five year period.

The DOS constructs a consumer price index (CPI), based on the prices of over 180 commodities. However, the items included and the weights attached to them are not representative of the consumption of the poor. Moreover the CPI is unavailable for rural areas. Therefore updating the poverty line using the inflation rate of the CPI would be inappropriate. A basket of goods that do reflect the consumption of the poorest group, with attached weights, was constructed using the LECS2 diary data.[10] The inflation in the cost of this “poverty basket” was used to update the poverty line from LECS2 to LECS3. For consistency the same poverty basket has been used in this report to inflate the poverty line from LECS3 to LECS4.

Various sources of price information are available. Following the practice of Richter et al (2005), price data from the village price survey are used to inflate the poverty line for urban Vientiane from the LECS3 level. However, this poverty line must be adjusted to account for regional and urban-rural differences in the cost of living. Following practice in the earlier LECS surveys, eight separate poverty lines are used; for urban and rural areas in Vientiane Municipality, North, Center and South respectively. The low coverage of the village and market price surveys outside Vientiane and the fact that, if particular items are not sold their price is missing in the dataset, makes it problematic to use these price data to calculate the cost of living for these regions. Instead, the household diary data, which gives information on quantity consumed as well as total value, is used to derive prices for the goods included in the poverty basket.[11] The cost of the poverty basket is then calculated separately by region and by urban or rural status, and is used to construct price indices comparing the cost of living in these regions to the cost of living in Vientiane. These indices are used to adjust the urban Vientiane poverty line to an appropriate level for each region.

The survey was collected over the course of the year, and the cost of living may have varied by month. To control for seasonal variation, monthly price indices are constructed, using the CPI prices to calculate the monthly price of the poverty basket.

By comparing household consumption with the relevant poverty line, poverty estimates can be made. Aggregate results are adjusted for survey design effects by weighting, based on the results of the 2005 national census. Derived poverty estimates can then be related to other variables to create poverty correlates and profiles.



[1] Formerly the National Statistics Center.

[2] LECS4 was slightly delayed and actual data collection was carried out between April 2007 and March 2008.

[3] The village (baan) is the most basic administrative unit in Lao PDR. This is consistent with the important traditional role for the village as a basic social unit. Koumban, or village cluster, is a recently introduced administrative unit between the village and district level. Districts (muang) are mid-sized administrative unit, are commonly made up of one urban area and a surrounding rural hinterland. The county is divided into 17 provinces (khwaeng). An 18th administrative unit the Xaisomboun Special Region was dissolved in 2006. The provinces are commonly grouped into the four regions of North, Central, South and Vientiane Municipality.

[4] This is an increase from 2937 households surveyed in the first 1992/3 LECS.

[5] Questions on child health and weight were included in the LECS III survey. The questions were not included in LECS IV as similar data on child health had been collected in the Lao PDR Multiple Indicator Cluster Survey, conducted in 2006 by the Department of Statistics of the Ministry of Planning and Investment in close collaboration with the Hygiene and Diseases Prevention Department of Ministry of Public Health and supported by UNICEF (Department of Statistics and UNICEF, 2008).

[6] For more information about the village price survey and its coverage see Appendix I.

[7] Beds, dining and lounge suite, stove with oven, fridge, ax, sewing machines, washing machines, cars & vans, motorcycles, TVs, VCRs, and computers are all excluded from consumption expenditure.

[8] This is the case for tables and chairs, cupboards, desks and sideboards, stools and benches, carpets, lamps, rugs, mats, pictures, stoves (non-electric), irons, electric fans, bicycles, watches, jewellery, airline tickets, expenses abroad, radio or cassettes, camera, other photographic and musical equipment, cellular phone and repairs of such items. The annual purchase value of these items is included in consumption expenditure (divided by 12), so where purchases are recorded in the household diary, these are discarded.

[9] See Kakwani et al (2002) for more information on how the poverty line was constructed.

[10] See Kakwani et al (2002) for more information on how this poverty basket was constructed.

[11] A limited number of commodities in the poverty basket had no matching category in the household diary, or the category in the diary included several types or species. This makes it unclear what the price derived from the diary data represents. In these cases the price from the diary was replaced with the price from the village price survey. These items were: dried chili, fermented fish, canned and frozen fish, fresh fish, sweets, spices and seasoning.

III


III. Poverty Trends

Lao PDR has seen a steady decline in national poverty headcount since the first LECS survey in 1992/3. Figure 1 shows the poverty headcount rate at the national poverty line and the international poverty line of $1.25 a day.[1] In 1992/3 almost half of the Lao population lived in poverty according to the national poverty line. By 2007/8 the poverty headcount rate was just above a quarter. The international poverty line shows a similarly dramatic drop, from 57% to 37%.

Figure 1: Headcount Poverty Incidence, National poverty Line and International Poverty Line, 1992/3 to 2007/8

Figure_1

The poverty headcount measure fails to take into account whether the poor have consumption levels just below or far below the poverty line. It is therefore indifferent to the distribution amongst the poor and insensitive to the degree of poverty. The poverty gap and poverty severity measures address this weakness. The poverty gap equals the average consumption shortfall of the poor relative to the poverty line. The greater the distance, the higher will be the poverty gap. Poverty severity applies an increasing weight to greater distances below the poverty line and is therefore sensitive to the severity of poverty.

Table 2: Poverty Gap and Poverty Severity, national poverty line, 1992/3-2007/8


1992/3

1997/8

2002/3

2007/8

Poverty headcount

46.0

39.1

33.5

27.6

Poverty gap

11.2

10.3

8.0

6.5

Poverty severity

3.9

3.9

2.8

2.3

The poverty gap offers additional support to the picture of successful poverty alleviation during the past five years. The poverty gap fell from 8.0 to 6.5 between 2002/3 and 2007/8. Not only has the share of the population living below the poverty line fallen, as shown by the decline in the headcount, but also the average consumption shortfall of those under the poverty line is smaller than five years earlier.

The severity of poverty declined from 2.8 to 2.3, a substantial reduction just like the headcount and gap measures. Since the poverty severity gives the highest weight to the poorest, this implies that the drop in poverty since 2003/3 benefited the neediest.

Figure 2 and Figure 3 put Lao PDR’s poverty decline in a regional perspective. In the early 1990s Lao PDR was one of the poorest countries in South East Asia but has seen a rapid decrease in poverty, with just over a third of the population living on below $1.25 a day by 2007/8. However, Lao PDR’s neighbors Vietnam and China have been able to reduce poverty even more rapidly, so the country now lags further behind than in the early 1990s. Although poverty reduction has been rapid, the constraints facing Lao PDR have prevented it achieving the rapid decreases in poverty seen in neighboring countries and it remains one of the poorest countries in South East Asia.

Figure_2

Figure 2: Regional Headcount Poverty Incidence, early 1990s and mid 2000

Figure_3

Figure 3: Regional Rates of Reduction in Poverty Headcount



[1] The $1.25 poverty line used here was calculated at private consumption purchasing power parity for 2005, then deflated to 2002/3 levels using national CPI data. It was adjusted to account for differences in the cost of living between areas and between and within survey years using the national poverty lines, which already adjust for this temporal and spatial variation. More information on the methodology used for calculating PPP and updating the line is available in Appendix IV: Estimating Dollar-a-Day Poverty Using the LECS Household Surveys.


IV


IV.Poverty Patterns

Although poverty reduction in the country as a whole has been rapid, certain geographic regions and groups have lagged behind throughout the period, and continue to face significant development challenges, despite efforts by government and donors. Lao PDR’s population is spread unevenly across the country and much of the variation in socioeconomic conditions is related to geography. Ethnicity is another important dimension for understanding variation in livelihoods and welfare. This section examines patterns of poverty within Lao PDR, and how they have changed between 2002/3 and 2007/8.[1]

Urban and Rural Areas

Urbanization and road access have expanded with development. Despite the increase in urban population, the proportion of urban households living in poverty has fallen rapidly, to just over 17% in 2007/8. Households living in rural areas have different livelihood options open to them and progress there has been slower. Poverty in rural areas was 32% in 2007/8, almost twice the rate of the urban areas. Although rural areas made up 71% of the population they accounted for 82% of the poor. However, it is the most remote rural villages that have lagged farthest behind. In villages that are cut off from markets and other infrastructure poverty remains at 43%; a fall of just 5 percentage points since 2002/3.

Figure_4

Figure 4: Poverty Headcount by Area, Region and Priority District, 2002/3 and 2007/8


Figure_5

Figure 5: Share of Population and Share of the Poor by Area, Region and Priority District, 2007/8

Priority Districts

In October 2003, the Government presented a national poverty reduction program, the National Growth and Poverty Eradication Strategy (Government of Lao PDR, 2003). It states the goals of halving poverty by 2006, and eradicating mass poverty by 2010. The strategy is based on poor area development. The implementation focuses on 47 priority districts out of a total of 143, identified based on a set of household, village, and district level indicators. Another 25 second priority districts have also been identified. While they account for about half of all districts, just above a third of the population lives in priority districts. The low population reflects the dominance of sparsely populated highland areas.

In the first five year period with available data, 1992/3 to 1997/8, poverty in first priority districts increased (see Figure 4). This was at a time when national poverty rates declined substantially. Since the initial setback, poverty in priority districts has been declining at a steady pace. Now poverty is just below 44%, a decline from 63% ten years earlier. Despite these improvements, the poverty rate for first priority districts remains more than twice the rate for non-priority districts and the decline in poverty rates has been slower. Although the first and second priority districts together account for only 36% of the population, they account for 53% of the poor.

Regional Poverty Patterns

Lao PDR can be divided into four regions: North, Central, South, and Vientiane Municipality. The North includes mountainous and rugged land, a wealth of natural and ethnic variation, and the old capital, Luang Prabang. The Central region is largely situated along the Mekong Valley with many of the most productive rice growing areas of the country, but also more remote upland areas along the eastern border with Vietnam. The South combines flat croplands with hilly, densely forested areas and remote areas bordering Cambodia and Vietnam. Vientiane, as the country’s capital, has a central role both economically and administratively. The town has benefited from increased integration with regional neighbors and the world economy.

Figure 5 shows substantial regional variation behind the steady decline in poverty. Vientiane Municipality experienced a rapid fall in poverty rates in the early and mid 1990s. An increase in the period leading up to 2002/3 has been reversed in the last five year period, as poverty has fallen to 15.2% in the most recent survey. Throughout the period the North has lagged behind the other regions, and in 2007/8 had a poverty headcount of 32.5%, while the South and Central regions have experienced more rapid poverty reduction, to 22.8% and 29.8% of the population respectively.

Figure_6

Figure 6: Poverty Headcount by Border, Altitude and Ethnicity, 2002/3 and 2007/8

Figure_7

Figure 7: Share of Population and Share of Poor, by Border, Altitude and Ethnicity, 2007/8

Border Districts

Lao PDR is positioned at the center of the Mekong sub-region. In fact, some parts of the northern regions have better developed transport and trade linkages to Southern China and Northern Thailand than the rest of Lao PDR. The western border is shared with Thailand, and coincides for over 1,000 km with the Mekong River. The eastern border with fast-growing Vietnam is the longest, but only a third as many people live here than along the Thai border. The border areas to China and Myanmar in the Northwest, and to Cambodia in the South, are remote and sparsely populated. Figure 7 shows that while the regional population shares have remained constant across the past fifteen years, more than half of the Lao lived in border districts in 2007/8, about 8 percentage points more than in 1992/3.

The districts along the Thai border have seen rapid poverty reduction and have substantially lower poverty incidence than the rest of the country, at 16%[2]. Both the China-Myanmar and Cambodian border regions have seen impressive poverty reduction since the early 90s. The Cambodian border districts have experienced particularly rapid poverty reduction in the past 5 years, and now have lower poverty incidence than the inland districts. This could be evidence to the benefits of having access to foreign markets for trade and employment. Progress in the areas bordering Vietnam has been slower however, with less than a 4 percentage point decrease since 1992/3. More than half of all residents remain in poverty. Here the opportunities of border trade and integration with the neighbors might be hampered by the limited cross-border infrastructure.

Altitude

Perhaps even more important than regional differences are variations by elevation level. More than half of the population lives in lowland areas. These are the plains along the Mekong and other rivers with irrigable and fertile soils. About two thirds of the country is mountainous, with large sparsely populated areas intersected by river valleys, and upland plateaus. The impact of altitude on livelihoods is profound. While many parts of the uplands are deprived of infrastructure, there is an abundance of natural resources such as water, forests and minerals. Despite this abundance of natural resources, upland areas have a poverty rate of almost 43%, and make up 39% of the poor, despite accounting for just 25% of the population. In addition, the decline in poverty between 2002/3 and 2007/8 has been particularly slow in the upland areas.[3]

Ethnicity

Lao PDR is endowed with a wealth of ethnic and cultural variety, which must be taken into account for understanding deprivation and poverty. The largest ethnic group, the Lao-Tai, make up just two thirds of the population according to the 2005 census. Different ethnic groups tend to have distinct livelihoods and live in different locations, which tend to overlap with altitude.[4]

The Lao-Tai traditionally live in lowland areas with agricultural systems based on wet-rice cultivation. The Mon-Khmer today live in many midland areas in Lao PDR. Chine-Tibet and Hmong-Lu Mien groups tend to live in highland areas and engage in uplands agricultural practices, largely in northern Lao.

The large Lao-Tai group have substantially lower poverty incidence than the other ethnic groups. The Mon-Khmer have poverty incidence more than two and a half times the rate of the Lao-Tai. All three non-Lao-Tai groups have poverty rates above 42%. These groups have seen a relatively slow decline in poverty incidence over the past 5 years compared to the Lao-Tai. However, because of the small sample size these results are subject to significant measurement error.[5]

Summary

Economic growth over the past 15 years has benefitted the poorest in Lao PDR, leading to declines in not only the headcount poverty rate but also the poverty gap and the severity of poverty. However, the speed of the gains has varied. The areas with the best links to markets and infrastructure have seen the fastest declines in poverty: urban areas, Vientiane province and the districts along the border with Thailand. Some areas continue to lag behind. Rural villages remain far poorer than the urban areas, and the most remote villages without road access have struggled to reduce poverty. The North, the regions bordering with Vietnam and uplands areas face the largest challenges with poverty. These geographic regions are largely inhabited by ethnic minority groups, who continue to experience high rates of poverty. Despite concentrated efforts to reduce poverty in priority districts, they still account for almost 55% of the poor.

V

V. Sensitivity of Poverty Estimates

It is important to note that these poverty rates are statistical estimates and are therefore subject to various sources of error that affect the interpretation. These issues are discussed in greater detail in Appendix II, but are summarized briefly here.

Firstly, the LECS surveys cover a sample of Lao households, not the whole population. Because of this, it is impossible to be completely certain that the poverty rate reflects the true poverty rate of the population. For example, in the 2007/8 survey it is possible to state with 95% certainty that poverty was more than 25.5 percent but less than 29.7 percent, but not possible to state with certainty that it was exactly 27.6%.

Within smaller groups the “confidence intervals” are larger. For example, it is possible to say with 95% certainty that the poverty rate for the largest Lao-Tai group is between 17.3% and 19.4%, but for the smallest group of “other ethnicity”, it is only possible to be sure that the rate is between 9.9% and 34%. For this reason results for this small group are not reported in subsequent sections. However, even for larger groups the results should be interpreted with care, on the understanding that the small sample size makes the statistics less exact.[6] When differences over time are analyzed, measurement error can be magnified, so estimates of change over time should also be interpreted with care.

Secondly, the fact that consumption is often underestimated in statistical surveys means that poverty is likely to be overestimated. Furthermore, random measurement errors will usually lead to overestimation of poverty. However, these biases are common to all rounds of LECS and to other surveys, so the figures should be comparable across time, and with findings from other countries.

Finally, as discussed in Section II, there are several ways of defining and updating the poverty line. Although the method used has been selected as giving the most appropriate estimates of poverty for Lao PDR, it is important to note that the choice of poverty line makes a large difference to estimated poverty rates. Appendix II demonstrates that increasing the poverty line by 10% would increase the poverty headcount rate by 25%, to 34.5%. Furthermore, the sensitivity of poverty estimates to the choice of poverty line has increased over the 4 rounds of the LECS survey. At the time of LECS3 a 10% increase in the poverty line would have increased poverty by just 23% and at the time of LECS1, by 17%.

Figure_8

Figure 8: Poverty Incidence Curve, Lao PDR, 2002/3

Figure_9

Figure 9: Poverty Incidence Curve, Lao PDR, 2007/8

VIII

VIII.Food Poverty and Insufficiency

While total poverty incidence in Lao PDR has shown a stable decline since the first LECS survey, food poverty trends have varied. It is possible to calculate food poverty rates using the LECS data starting in 1997/8, where food poverty was estimated at 32.5%. By 2002/3, food poverty rates had fallen dramatically to 19.8%. However, food poverty incidence for LECS4 is 24.6%. There has been a significant drop in food poverty over the 10 year period, but the apparent increase between 2002/3 and 2007/8 requires some explanation.

Figure_10

Figure 10: Food poverty headcount, urban and rural 1997/8 to 2007/8

Food poverty rates are derived by relating the monetary value of households’ food consumption with a food poverty line representing the cost of needed to buy food sufficient to provide 2100 calories per day. A household that is food poor spends less on food than is required to buy this “food basket”. A household that is overall poor spends, in total, less than is needed to buy this basket, plus an allocation for non-food consumption. So there may be some households that don't spend enough on food to meet 2100 calories a day and are thus food-poor, but spend a lot on non-food consumption, so are not considered poor overall.[1] In addition, there are some households who spend enough on food to rise above the food poverty line, but have low non-food expenditure, so are below the total poverty line.[2]

While changes in total poverty are driven by changes in total consumption and overall price trends, food poverty trends are driven by the share of consumption devoted to food and food-price inflation. Therefore food consumption shares, displayed in Table 3, are crucial for explaining trends in food poverty. Between 1997/8 and 2002/3 the share of food in total consumption increased from 72.8 to 74.6% of total consumption. This contributed to a rapid decline in food poverty. However, as countries experience income growth and poverty reduction, the share of expenditure on food usually falls. The change between 1997/8 and 2002/3 therefore represents an unusual pattern. The 2007/8 survey saw a decline in food share to 72.3%. Lao households have devoted an increasing share of their consumption to non-food goods. This is true for both urban and rural households, and even in the most remote rural areas. It seems that even the poorest households are spending more on non-food. This is to be expected with economic development, and fits with information reviewed below on asset ownership, but it led to the increase in food poverty relative to 2002/3.

Table 3: Food share of consumption, 1997/8 to 2007/8


1997/8

2002/3

2007/8

Lao PDR

72.8

74.6

72.3

Urban

66.4

69.6

65.4

Rural

74.1

76.1

75.1

Rural with road

70.7

74.6

74.7

Rural w/o road

76.8

78.1

77.5

Furthermore, it is important to note that even if households fall below the food poverty line, this does not necessarily mean that they are suffering from food deprivation. The food poverty line relates to consumption of a basket of goods designed to represent the consumption of the poorest households. If the consumption patterns of the poor have changed, it is possible that many households who fall below the food poverty line nonetheless meet their nutritional needs by using cheaper sources of calories. Furthermore, the price of the basket has been updated using a fixed basket of goods and weights set in 1997/8. If the pattern of consumption has changed, or if the prices used to update the food poverty line are not representative of the prices paid by the poor, it is possible that divergence between the amount needed to afford basic nutrition and the food poverty line has arisen.[3] Further analysis using the LECS data to estimate calorific intake will investigate the seriousness of the food poverty findings.

Rice Insufficiency

In order to better understand the seriousness and implications of rising food poverty rates, this section examines information on food security and on absolute volumes of food consumed and how this has changed over time. Rice is the basic food staple for most Lao. Having access to sufficient rice for the whole year is a basic measure of a household’s food security. In the LECS village surveys, the village chief was asked for how many months in the past year the village had suffered from a lack of rice.[4]

Figure_11

Figure 11: Rice insufficiency, average number of months per year by Lao PDR, urban and rural villages 1997/8 to 2007/8

It is apparent from Figure 11 that there has been a positive trend in rice sufficiency in most areas. In 1997/8 urban areas reported more months of rice insufficiency than rural areas, and within rural areas, the most remote villages suffered fewer months of insecurity than the villages with roads. However, there has been a rapid decline in rice insufficiency in urban areas. In 2007/8 urban villages experienced just 1.8 months of insecurity on average, while rural villages reported 2.6. The decline for areas with road access has been faster than the decline for those without.

Vientiane Municipality has seen an impressive fall in rice insufficiency years and the Central and Southern regions have seen steady progress. The decreasing insufficiency in these three regions is likely to be a reflection of increased rice production in these areas as well as the impact of infrastructure improvements guaranteeing access to outside markets. In the North rice insufficiency appears to have increased over the observed period, but this change is not statistically significant. The ability to purchase rice on outside markets is hampered by higher transportation costs in the northern part of the country, meaning that local insufficiency has not improved as much as in the other regions.

Rice Consumption

Individual-level rice consumption can be estimated using information from the nutrition module, to provide additional insights on the secure availability of this important food staple.[5] Figure 12 shows that there doesn’t seem to be any strong correlation between rice consumption and poverty status. Both poor and non-poor households tend to consume rice in similar quantities to the national average.

Figure_12

Figure 12: Rice consumption, average rice balls of sticky rice plus bowls of steamed rice per day and per capita, 2002/3 and 2007/8

The results don’t indicate any dramatic changes during the five year period between LECS3 and LECS4. National rice consumption has increased slightly. The increase has been driven by an increase in rice consumption among urban households and rural households without road access. The biggest increases have been in the North and South regions. Rural households still consume more rice per person that urban ones. This might reflect a higher reliance among the rural, agricultural population on carbohydrate-rich rice, rather than more expensive sources of calories such as meat and fish.

Summary

Although there has been a dramatic decrease in overall poverty, food poverty rates have risen over the most recent five year period. Given the poor status of nutrition in Lao PDR and the fact that food prices currently remain at a relatively high level, this provides some cause for concern. The reason for this finding is that households are spending more on non-food consumption relative to food consumption. Further analysis is needed to determine the causes of this pattern and whether this means that households are cutting back on essential items, or whether they are exploiting cheaper sources of calories. Analysis of village-level rice insufficiency and individual-level rice consumption suggest that food security and consumption have increased moderately over time and the food poverty findings may not represent an overall decline in access to food.

VI

VI.Asset Ownership

Analysis of asset ownership helps to confirm the findings about consumption poverty, but also adds important insights about how the lives of Lao households have changed in the past five years. Assets can be used for productive purposes. Tractors are the most obvious example found in the LECS data, but other assets, such as motorbikes or phones, could also expand income opportunities. Ownership of motorbikes or bicycles can also allow households to access services such as health or education. Mobile phones can help household access information and communicate with the outside world. Furthermore, as poor households usually lack access to formal financial institutions, assets play an important role in saving and insuring against risk. Finally, ownership of assets can be an important status symbol even for the poorest households. Because many durable assets are excluded from the consumption aggregate, and because households may have made investments that are not recorded in the previous year’s consumption, these transformations are not fully reflected in the poverty analysis.

Figure_13

Figure 13: Asset ownership, all Households, 2002/3 and 2007/8

Figure_14

Figure 14: Asset Ownership, Poor Households, 2002/3 and 2007/8

Figure_15

Figure 15: Asset Ownership, Villages without Road Access, 2002/3 and 2007/8

The rapid increases in asset ownership visible in Figure 13, Figure 14 and Figure 15 reinforce the picture of progress coming from the poverty analysis. Even households who are below the poverty line have made significant investments in durable assets. Ownership of radios, TVs, motorbikes, mobile phones, tractors and mosquito nets have all increased. Increases in the ownership of motorbikes, from 24% to 59% of households and of mobile phones, from just 3% to almost 50%, have been particularly striking. Although richer households are more likely to own assets, a significant number of poor households are also investing. In 2007/8 35% of poor households owned a TV, 35% a motorbike and 21% a mobile phone. All three of these assets give previously isolated communities access to the outside world and to information, and are rapidly transforming the lives of Lao households in unforeseen and unpredictable ways.


VII


VII.Trends in Inequality

In the early 90s consumption inequality, as measured by the Gini index, increased rapidly. In the next five years inequality declined. Between 2002/3 and 2007/8 it has increased once again, to just above the 1997/8 level.

Figure_16

Figure 16: Gini index, Lao PDR, urban and rural 1992/3 to 2007/8

Throughout the fifteen year period, inequality in urban areas has been substantially higher than in rural areas. This is a reflection of the greater variation in livelihoods and thus consumption opportunities in urban areas. Table 4 indicates that the changes in consumption inequality have been largely driven by the share of consumption attributed to the richest quintile. While other groups have seen minor changes in their shares, the richest saw a large increase in the first period, followed by a fall and then another increase. Thus the consumption at the top of the distribution seems to have driven the changes in inequality.

Table 4: Consumption shares by quintile, 1992/3 to 2007/8


1992/3

1997/8

2002/3

2007/8

1

8.8%

7.7%

8.5%

7.9%

2

12.6%

11.6%

12.3%

11.8%

3

16.0%

15.3%

16.0%

15.4%

4

21.5%

20.8%

21.4%

20.9%

5

41.0%

44.5%

41.9%

44.0%

Changes in inequality have an impact on poverty. If economic growth is accompanied by an increase in inequality, the decrease in poverty will be smaller than if the distribution of incomes had become more equal. One way to understand the impact of distribution and growth on poverty is to decompose changes in the poverty headcount into growth and redistribution components (Datt and Ravallion, 1992). The growth component captures the change in poverty due to real consumption growth in the absence of change in inequality. The redistribution component is the change in poverty that would have been observed because of changes in inequality, even if there had been no growth in mean incomes.

Results from these decompositions, along with growth incidence curves, are presented in Appendix V. A common empirical finding internationally is that the growth component accounts for the bulk of changes in poverty. In Lao PDR the average annual fall in national poverty headcount of 4.5% since 1992/3 can be fully explained by the contribution of consumption growth. Without adverse distributional changes the rate of poverty alleviation would have been even higher.

The contributions of growth and distribution vary between survey rounds. Between 1992/3 and 1997/8, inequality rose. In this period, redistribution had an adverse effect on poverty. This was followed by a period when both growth and falling inequality contributed to poverty reduction. During the most recent five year period strong growth pushed down poverty rates, but increasing inequality held back the potential impact of growth on poverty. Growth incidence curves presented in Appendix V illustrate these patterns. Overall, consumption growth among households below the poverty line has been slower than growth for the rest of the population. However, despite increases in inequality, all population percentiles have experienced a substantial improvement in material welfare.

IX


IX.Access to Public Services and Infrastructure

In addition to decreases in income poverty, improvements in access to public services and infrastructure are important for the wellbeing of Lao households and achievement of the MDGs. Access to public services and infrastructure can also help households to break out of poverty by providing new opportunities. Public education can open up new employment opportunities, road infrastructure can give villagers access to outside markets for their products and lower the prices of essential goods bought from the outside, and water and sanitation investments can contribute to improving health outcomes.

The following sections describe and analyze trends in access to some essential public services and infrastructure. This serves to give a broader picture of the progress in alleviating deprivation, as well as providing some hints about the factors driving falling poverty rates in the country.

Access to Education

The importance of education for meeting development goals is evident from the inclusion of the goal of achieving universal primary education as the second Millennium Development Goal (MDG) (UN, 2000). Education can help individuals and households to increase incomes and escape poverty. Figure 17 shows that literacy is systematically related to poverty status. While almost 90% of household heads can write a letter, only 78% of the poor can do so.[1]

Figure_17

Figure 17: Literacy household heads, Lao PDR, 2002/3 and 2007/8

While urban areas have near 100% literacy rates, rural areas without road have literacy rates below 75%. Minority ethnic groups also lag behind the Lao-Thai. The Chine-Tibet group have particularly low literacy rates, below 40%. Progress in improving adult literacy and addressing these disparities has been slow. The relative importance of poverty, remote location, lack of infrastructure and cultural or linguistic barriers in driving these disparities is unclear and merits further attention.

The data from the LECS village survey provide information on the presence of schools in each village, and the distance from schools when no school is present. This provides a direct measure of access to education. In all regions and groups access to schools in the village increased between 2002/3 and 2007/8. 89% of villages have their own primary school, compared to 79% in 2002/3. The increase in access to schools has been most striking in the most remote areas. In 2002/3 only 75% of villages without road access had a primary school. This had increased to 91% by 2007/8. There has also been an impressive increase in the priority districts, an outcome of increased investment in these areas.

Rural villages are now slightly more likely to have a primary school than urban. This reflects the fact that if villages are close to each other, and have good transport links, children can easily attend school in a neighboring village. In more remote areas, by contrast, the school must be located in the village itself if children are to attend.

For secondary schools the picture is reversed. While 36% of urban villages have a lower secondary school, only 4% of the most remote rural villages do. The proportion of villages with secondary schools has also doubled between 2002/3 and 2007/8, but is still relatively low, at 18%. This is confirmed by the evidence on distance to school. Children in rural areas must travel over 8km on average to the nearest secondary school, while in urban areas the average distance is just 3km.[2] Children in villages without road access must travel an average of 17km to secondary school. While children from urban villages can go to the next village to attend secondary school, many rural children will be unable to progress beyond primary education.

Figure_18

Figure 18: Share of Villages with a School, 2002/3 and 2007/8

Figure_19

Figure 19: Average Distance School when No School in Village 2002/3 and 2007/8

Enrollment is one measure of the outcomes of access to schools. Figure 20 indicates that enrolment of 6-10 year olds has increased, in all regions and groups, over the 5 year period.[3] 81% of 6-10 year olds were enrolled in 2007/8, up from 72% in 2002/3. [4]

Although more rural than urban villages have their own primary school, only 78% of rural children were enrolled, against 94% of urban children. This could reflect the fact that, even if a school is present, rural households often struggle to afford school fees and other expenses, or that rural children are needed to work in agriculture. This means that many children enroll late or not at all. The poor and the non-Lao-Thai ethnic groups are less likely to enroll than the average, with children from the Chine-Tibet group particularly likely to be missing school. Less than half of this group was enrolled. In addition to the fact that these ethnic groups are poor and rural, they may also face linguistic barriers. There is a three percentage point gap in enrollment between boys and girls, which has remained remarkably consistent across time. Although children living in priority districts are less likely to be enrolled than their peers, the increase in enrollment over the last 5 year period has been impressive and is evidence that investment in these areas is having some effects.

Figure_20

Figure 20: Enrollment of children aged 6-10, 2002/3 and 2007/8

The results for older age groups mirror those for primary school and are not reported here.[5] For all groups, enrollment of 11-13 year-olds (the official age for lower secondary enrollment) is higher than that of 6-10 year-olds: 87% of Lao 11-13 year-olds are enrolled. Although disadvantaged children enroll later than their peers, most children do attend school at some point: 82% of poor 11-13 year olds are enrolled. Given the lack of secondary schools, particularly in rural areas, it is probable that many 11-13 year-olds are attending primary schools.

However, in the 14-16 age-group (equivalent to upper secondary school age) enrolment drops significantly. Only 63% of children in this group are enrolled, an increase of 5 percentage points since 2002/3. The same pattern of regional, ethnic and economic differences is apparent, but the gaps are larger. Furthermore, the gender gap becomes highly significant for this age group. 71% of boys are enrolled, as opposed to just 55% of girls.

Access to Health Services

Access to health services also plays an important role in increasing household well-being. Table 5 presents village-level indicators of access to health services. There has been a clear increase in the percentage of villages with a medical bag, especially in the most remote areas; 66% of villages had a medical bag or drug kit in 2007/8, compared to 36% in 2002/3. They have also benefitted from an increase in the availability of community health workers, from 56% to 73%. Because these basic services are aimed at rural areas, rural villages are the most likely to have access.

For most of the facilities, however, access has remained roughly similar between the two rounds of the survey, and the gap between urban and rural areas remains large. In 2007/8, 82% of urban villages had a medical practitioner or nurse, and 60% a pharmacy. In the most remote areas only 11% had a doctor or nurse, and 5% a pharmacy.

Table 5: Village access to health facilities, LECS3 and LECS4


LECS3 

LECS4 


Urban 

Rural with road 

Rural no road 

Urban  

Rural with road 

Rural no road 

Villages with medical bag/drug kit (%) 

20%

45%

36%

26%

67%

66%

Villages with traditional birth attendant (%) 

48%

64%

37%

38%

56%

52%

Villages with traditional healer (%) 

38%

49%

37%

33%

40%

36%

Villages with community health worker (%) 

41%

68%

56%

43%

60%

73%

Villages with medical practitioner/nurse (%) 

85%

29%

10%

82%

29%

11%

Villages with dispensary/health centre (%) 

23%

13%

6%

19%

17%

9%

Villages with pharmacy (%) 

59%

26%

10%

60%

17%

5%

Villages where most women give birth in a hospital (%) 

72%

14%

1%

84%

30%

7%

Note: average distances reported for villages with no hospital or dispensary/health centre

Table 6 indicates that rural residents also have long distances to travel to reach medical facilities. In 2007/8 the average village without road was located 27.7 km from the closest hospital. 83% were more than 8kms away. It took on average over 5 hours to reach a hospital and 54% of villages were more than 2 hours away. There has been little improvement in access between the two rounds. There has been a slight decrease in the average distance to dispensaries and health centres, but the average village without road was still located over 13 km, or more than two hours, away from the closest centre in 2007/8.

One encouraging sign is that the proportion of villages where most women give birth in a hospital increased across the board. However, in remote rural areas this is the case for only 7% of villages. Although access to basic health services has improved slightly over time, considerable improvements are still necessary to provide adequate access for the poorest households.

Table 6: Distance and time to health facilities, LECS3 and LECS4


LECS3 

LECS4 


Urban 

Rural with road 

Rural no road 

Urban  

Rural with road 

Rural no road 

Distance to hospital (km) 

4.1

15.5

27.9

5.9

18.7

27.7

Percentage more than 8km from hospital 

20%

69%

86%

31%

75%

83%

Time to hospital (mins) 

13.6

75.6

285.7

16.8

91.4

319.5

Percentage more than 2 hours from hospital 

0%

22%

68%

0%

20%

54%

Distance to dispensary/health centre 

2.3

8.3

17.7

3.1

7.4

13.2

Percentage more than 8km from health centre 

14%

45%

65%

7%

36%

57%

Time to health centre 

10.1

56.7

201.1

12.5

45.5

125.8

Percentage more than 2 hours from health centre 

0%

15%

51%

0%

7%

36%

Note: average distances reported for villages with no hospital or dispensary/health centre

LECS focuses on consumption and economic issues and therefore provides limited information on health outcomes. One measure available, reported in Figure 21, is the treatment rate conditional on self-reported illness. This is defined as the proportion of people who received some form of treatment (either inpatient or outpatient, from any form of provider) if they reported having suffered from any sort of health problem within the last 4 weeks. This proportion has increased from 16% in 2002/3 to 21% in 2007/8. This measure relies on a subjective assessment of what it means to be ill, so bias may arise. For example, the poor may be less likely to report minor ailments as illness. Therefore Figure 22 presents unconditional treatment rates. Unconditional treatment rates have actually decreased, in most groups, over the five year period.[6] However, both measures suggest similar patterns between groups.

The poor are much less likely to have received treatment, both conditional on reported illness and unconditionally, than the non-poor. Furthermore, the proportion treated has not increased over time as rapidly as in the non-poor group. This strongly suggests that the poor lack access to healthcare and that this situation has not greatly improved over time. Women are slightly more likely to receive treatment than men.[7] Non Lao-Thai groups and individuals in remote villages were less likely to receive treatment in 2002/3. The gap had narrowed somewhat by 2007/8 but a large disparity remains. Significant numbers of the poor, those living in remote villages and priority districts report not accessing healthcare when sick. However, the treatment rate in priority districts increased more rapidly than in non-priority districts, so there is some suggestion that investment in these areas has translated into improvements in health outcomes.

Figure_21

Figure 21: Treatment conditional on self-reported illness, Lao PDR, 2002/3 and 2007/8

Figure_22

Figure 22: Treatment unconditional of illness, 2002/3 and 2007/8

Access to Roads

Road access can have dramatic effects on villages. The cost of reaching outside markets to sell products or buy goods is dramatically lowered, increasing farm-gate prices for households with surplus agricultural production and reducing the prices of essential goods, such as fertilizer, fuel, food and medicine. Roads also improve access to public services by reducing travel time and making it easier to recruit skilled individuals such as teachers and health workers. Transport links to the outside world might, however, also carry negative side-effects, for example exposure to social hazards, such as human trafficking and drug trade.

Research has shown that when positive and negative effects of road access are taken together, the overall impact is a lowering of poverty incidence (for an analysis of the effects of roads on poverty in Lao PDR see Warr, 2008). Section IV showed that poverty rates are strongly linked to road access. In urban areas, road access is virtually universal (Figure 23, below) and poverty rates are below 18%. In rural villages that do have access to an all-season road, the poverty headcount is 30%, but in those lacking roads more than 42% of individuals live in poverty. Although the statistics do not imply causality, this is a dramatic difference; rural poverty incidence increases by almost four times as we move from villages with road access to those without.

Figure 23 illustrates the trends in road access at a national level, as well as for urban and rural villages. The village questionnaire asked village chiefs whether it is possible to access the village by road or truck in rainy season and in dry season. Following a slight decline in the early and mid-1990’s rural road access has continued to increase between 2002/3 and 2007/8, while urban road access has been close to 100% throughout the period. The slight variation in road access in urban areas is most likely driven by an expansion of the areas designated as urban to encompass previously rural areas.

Figure_23

Figure 23: Share of villages that can be accessed by car or truck in wet season, 1992/3 to 2007/8

When comparing the regional trends, one important observation can be made; northern Lao used to have the lowest rates of road access but following rapid expansion of access roads in the highland North it has actually surpassed the South. In addition, first priority districts now have better road access than second priority districts.

Access to Electricity

Gaining access to electricity can have a number of positive effects for Lao households. It enables them to discontinue use of fuel lamps and other hazardous light sources with positive health effects, lower fuel costs and a lower risk of fire. In addition it can improve income opportunities. Figure 24 shows a steady increase in access to electricity for Lao PDR.[8] Urban areas had largely gained access to electricity networks at the time of the first LECS survey in the early 1990’s. At that time, only one in ten rural villages had the same privilege. Since then there has been a rapid expansion of rural electrification. In the most recent survey the proportion of rural villages with electricity had actually passed the 50% mark.

Figure_24

Figure 24: Share of villages with access to electricity, 1992/3 to 2007/8

There are large regional differences in electricity access, broadly following the same pattern as roads. Vientiane Municipality has had universal access since the mid 1990’s, with the Central region following and the North and South lagging behind.

Access to Safe Water

One of the 21 MDG targets is to halve, by 2015, the proportion of people without sustainable access to safe drinking water and basic sanitation (UN, 2000). At the time of the third LECS survey, only half of Lao households had a safe water supply in the dry season (Figure 25).[9] This figure has improved rapidly, and a72% of households have improved water according to estimates from the 2007/8 survey. The rapid improvement shows that Lao PDR is on track to meet this important development goal.

Figure_25

Figure 25: Access to safe water in dry season, 2002/3 and 2007/8

Yet, as Figure 25 shows, there are differences in access to safe water between the poor and the non-poor. While 68% of the poor have safe water access, the corresponding figure for the non-poor is 73%. This reverses the situation at the time of the 2002/3 survey, when the poor actually had a higher rate of access to safe water (likely due to a higher reliance on rainwater among poor rural households). The increase in access in priority districts has been particularly rapid.

Access to Improved Sanitation

Improved sanitation is of great importance for a secure health.[10] Without sanitation there is a risk that human waste will enter the water supply and cause disease. In light of this, the finding in the 2002/3 LECS survey that less than half of Lao households had access to improved sanitation was disheartening. The 2007/8 estimates display a clear improvement as improved sanitation can be found in almost 2 in 3 households. This is a truly impressive result given the obvious challenges involved in expanding sanitation to communities spread out over a large country.

Figure_26

Figure 26: Access to improved sanitation, 2002/3 and 2007/8

The difference in access to improved sanitation between the poor and non-poor is even larger than for access to safe water. Only 46% of the poor have improved sanitation, compared with 70% of the non-poor. This highlights the large expense that often has to be covered by the household to improve sanitation.



[1] The direction of causation in this relationship is unclear.

[2] Unexpectedly the average distance to primary schools had increased in urban areas, and is greater than the distance in rural areas. This could be because more and more villages have their own primary school, so the urban villages left lacking a primary school are the least well-off.

[3] Lao schools are closed during the months of June, July and August. Observations during these months were excluded. Because the month of survey is randomly allocated between villages, this should make no difference to the patterns observed.

[4] This group corresponds to the official age for primary school enrollment. However, because of late enrollment and grade repetition, the correspondence will not be exact.

[5] Results available in Appendix III

[6] It is unclear whether this represents improved health among the population or reduced access to healthcare.

[7] This could be because medical care for pregnancy and childbirth are included here.

[8] Access to electricity is measured at the village level. This may differ from household access to electricity- some households in villages with electricity may nonetheless be unable to afford access at home.

[9] Safe water is defined as piped water, protected well or borehole, and rainwater. Unsafe water includes surface water from lakes, rivers or streams and unprotected wells. Households who class their source of water as “other” are classed as having access to safe water, following practice in earlier reporting, as most of these households rely on gravity fed systems, which are usually considered as safe.

[10] Here defined as modern toilet, squat toilet or dry toilet.


XI


XI.Conclusions and Summary

Lao PDR has a successful record of poverty alleviation, as measured in the five national LECS surveys carried out since 1992/3. The fall in the most recent five year period, fuelled by robust economic growth and improvements in infrastructure, has been particularly impressive. The poverty headcount has fallen from 46% at the time of the first survey, to 27.6% in the latest 2007/8 LECS. As well as decreases in consumption poverty, asset ownership has increased dramatically. The rapid increase in motorbike and mobile phone ownership in particular are transforming the lives of the Lao population.

However, large disparities remain. Poverty reduction has been most rapid in the urban areas, where poverty is now at a very low rate. The North remains the poorest region. Upland areas and priority districts remain poorer than lowland areas and non-priority districts despite concentration of poverty reduction efforts on these areas. Poverty rates also vary according to ethnicity. It appears that either further support, or a change in the way the Lao economy grows, will be needed to eliminate poverty in these lagging areas.

One alarming finding is that food poverty in Lao PDR has increased. Both rural and urban areas have seen rapid deterioration in food poverty since 2002/3, although the overall improvement since 1992/3 has been impressive. Evidence from village estimations of rice sufficiency and individual-level nutrition suggest that food security and consumption on the individual level have improved over time. The food poverty findings may be caused by changing household consumption patterns. However, given the poor state of nutrition documented in other studies, this should be a priority area for further analysis.

Access to social services and infrastructure continues to improve. The improvements in access to roads, clean water and sanitation are particularly impressive. Education opportunities have expanded, with almost all villages, even in rural areas, having their own primary school, and distances to secondary schools decreasing. As a result, enrolment of school-aged children has increased. Even in the most disadvantaged groups almost all children attend school, although many enroll late. However, especially in the older age group, significant differences are visible in the probability of enrolment between ethnic groups, regions and by gender. The evidence suggests that distance to school remains a significant problem for rural children. However, an enhanced understanding of all barriers to enrolment will be needed to tackle these problems.

There is some evidence that access to health services has improved over time, although this finding may be subject to substantial bias. It appears, however, that the poor have less access to health services, and that their access has improved more slowly over time than that of richer households. Many households still have long distances to travel to reach healthcare services, and the number of villages where most women give birth in hospitals is woefully low. Significant improvements in access to healthcare are clearly a priority.


XII


XII. References

ADB. 2002. Participatory Poverty Assessment for Lao PDR. Vientiane: Asian Development Bank.

ADB. 2008. Research Study on Poverty-Specific Purchasing Power Parities for Selected Countries in Asia and the Pacific. Manila: Asian Development Bank.

Chamberlain, James R. 2007. Participatory Poverty Assessment II (2006). Vientiane: National Statistics Center & Asian Development Bank.

Datt, G. & Ravallion, M. 1992. “Growth and Redistribution Components of Changes in Poverty Measures: A Decomposition with Application to Brazil and India in the 1980’s.” Journal of Development Economics, Vol. 38, No. 2, pp 275-295.

Deaton, Angus. 1997. The Analysis of Household Surveys: A Microeconometric Approach to Development Policy. Baltimore: The Johns Hopkins University Press.

Department of Statistics and UNICEF. 2008. Lao PDR Multiple Indicator Cluster Survey 2006, Final Report. Vientiane, Lao PDR: Department of Statistics and UNICEF.

Government of Lao PDR. 2003. National Growth and Poverty Eradication Strategy. Vientiane.

GRID. 2005. Lao PDR Gender Profile. Vientiane: Gender Resource Information & Development Center.

Haughton, Jonathan & Khandker, Shahidur R. 2009. Handbook on Poverty and Inequality. Washington DC: World Bank.

Jenkins, S. and Van Kerm, P. 2006. Trends in Income Inequality, Pro-Poor Income Growth, and Income Mobility, Oxford Economic Papers, 58(3), pp 531-548

Kakwani, Nanak. Datt, Gaurav. Sisouphanhthong, Bounthavy. Souksavath, Phonesaly. & Wang, Limin. 2002. Poverty in Lao PDR During the 1990’s. mimeo.