The countries of former socialist system were severely affected by the collapse of the system of central planning. As a consequence of domestic and external factors, the countries have experienced many negative impacts. These include slow down of socio-economic development, destruction of the social safety system, drastic increase in unemployment, and worsening of living standards. These negative trends imposed sharp changes in starting from macro structure to lives of individuals. Health and education sectors fail to stand solidly on feet due to current state budget shortages.

Poverty has become an acute social problem in newly in transition countries. Prior 1990, in those countries poverty was not reported since basic needs were met and access to a full range of social services guaranteed.

The extent of poverty and the choice of policies to reduce poverty have become of interest to governments in transitional economies. Poverty measurement therefore became an urgent task. The objectives of this paper are to outline our practical steps we are taking to address this problem and to stimulate the exchange of ideas amongst participants of Working Group of Statistical Experts, eleventh session.

In spite of the growing literature on poverty, no specific consensus on a concept and definition of poverty exists. However, the dominant concepts of poverty are the absolute, relative and subjective poverty. Absolute poverty identifies the poor regardless of others and it refers to the state of "inability to attain a minimal standard of living. It therefore means the lack of what is regarded as basic human needs, the attainment of which is a necessary component for a "decent living". In most of the transition countries the absolute approach is adopted for statistical measurement. 

The Household surveys, national accounts and consumer price indices give basic information for assessing and determining poverty level in the transition countries.

For the first time in 1991 NSO (National Statistical Office) determined the poverty line or population minimum subsistence level. This was based on the value of a basket of essential foodstuffs and other consumer goods and service. The basic approach was application of the primary consumption norms and lowest prices as determined by the State Planning Committee in the 1970's-1980's. Considering variations in geographical location, active businesses and consumption patterns, the consumer basket was formed differently for urban and rural areas. In addition, the foodstuff calorie intake by the poor households was estimated to be equal to two-thirds of the household average calorie intake from the Household Surveys conducted by the NSO for approximately thirty years. The population minimum subsistence level was fully complied with cabinet requirements. 

The first living standards measurement survey was conducted under World Bank assistance in 1995 and the consumer basket approach was utilized, but the government refused to grant permission for most of the results of the survey to be published, and vigorously attacked the Bank's methodology. In 1997 the Government of Mongolia revised the Law on Statistics. According to the Law the statistical system of Mongolia has being introduced more efficient structural changes to produce impartial and reliable statistics. In addition in 1998 the Parliament of Mongolia was approved the Law on "Population poverty line and its Determination by the National Statistical Office". These laws help to create a legal framework indispensable for permitting to continue the estimation of poverty line and release without any interference coming from the politicians. 

The Government of Mongolia recognizing the value of the survey as a tool for monitoring poverty, decided to repeat the Living Standards Measurement Survey in 1998. 

In June 1998 with the support of UNDP the NSO conducted the survey for the second time with a view of seeing what changes were made in poverty of the population since 1995 as well as poverty incidence, depth, profile and causes. In addition it aimed at collecting detailed information on housing, education, health and employment of the population. Under the survey adequate data and information were collected on the population's living standard. The estimation of the poor and other concerned indicators was based on the poverty line determined by the same methodology in the 1995 and 1998 surveys. Therefore, the results of the surveys can be compared.

The consumer basket consists of food and non-food consumption and the commodities have been selected on the basis of the household income and expenditure survey. The food consumption value was determined as 40 per cent of households with lowest consumption per capita as the Household Survey shows. In other words, consumption composition was selected on the basis of real consumption (Table 1).

This consumption pattern basket was then scaled up and adjusted to represent a daily average caloric intake of 2100 calories per capita. This time the food basked was compiled of the similar commodities without any urban and rural variations.

Food poverty line: To estimate the food poverty line, the daily minimum-needs of food basket of an adult equivalent is multiplied by market prices in local areas (in each of the aimags). Consumer prices were collected by rural and urban variations. For the purpose of making data more detailed food poverty line was determined by urban and rural areas. As well in order to make consumption data comparable, the conversion factor was used thus enabling to reflect a person equivalent to the average male adult.

The Food Research Institute under the Ministry of Health and Social Welfare designated the daily required calorie intake of a particular age group and gender

The food share on the log of the ratio of total consumer expenditures was regressed to the food poverty line and other variables related to household composition and geographic location:

• S_i is the share of spending on food for the _ith household
• EXP is total household expenditures for the _ith household
• FP is the food poverty line specific to the location of a household

The intercept of the regression (a) represents the average share of spending on food for households whose total spending just equals the food poverty line. A single regression was run across all regions, and region specific (urban/rural) total poverty lines were obtained using regression results applied to the average values of independent variables for each region. The regions are determined based on petrol price criteria.

Periodic revision of the poverty line in Mongolia provides an opportunity to carry out timely analysis of population wealth and living standards and, to produce data to formulate targeted programs. However, the periodic revision of the poverty line limits direct comparisons of the number of people below poverty line over time because whenever the line is revised income groups shift may have occurred.

The most commonly used poverty indicators are:

1. The head count Index, % (H): This is simply the proportion of the population with a standard of living below the poverty line. This indicator is only good in telling how many are poor.
2. The Income-Gap Index (I): This is the percentage that falls short of the average income of the poor as indicated by the poverty line.
3. Poverty sensitivity index (Foster, Greer and Thorbecke Index) Px: This incorporates sensitivity with regard to distribution within the poor themselves.
4. Gini coefficient or unequal distribution of income/expenditure: The Gini coefficient measures the inequality in income and expenditures.

It is important to remember that all poverty indicators are relative. Statistics show as a relative judgement, that around 40 per cent or 20 percent or 50 percent of Mongolians are poor. They do not provide any information about characteristics and profile of the poverty. The living standard varies over time and by geographical location. The poverty profile depends on what is regarded as a basic need. For instance, the basic need of a person in a developed country is impossible to be compared with that of a person in a developing nation. Likewise safe water from central delivery is a basic need in densely populated urban areas whereas that might be not the case for persons living in a vast rural territory with very low density.

In the 1998 LSMS, the poverty of Mongolia was studied in absolute and relative terms. Absolute poverty defines the poverty line, while relative poverty describes the distribution of the poor in and around the poverty line. Actually, Mongolian poverty is absolute, but it was investigated in terms of relative measures.

During the centrally planned economy the government of Mongolia, as other former socialist countries, was addressing social and economic problems including education and health areas with much reliance on assistance and cooperation from former USSR and COMECON. In the previous economic system, the government implemented various programs whereby all people with working capabilities were provided with employment, disabled people were covered by basic social safety nets, the population was given basic education and total population had a free access to health services. During the transition period the social safety net underwent fundamental changes and the opportunities for implementing these programs by the government has become limited. At the same time the lives of people have deteriorated. Therefore, the government began determining a poverty line since 1991 with a view of address this issue and take appropriate actions. A program targeted at those people who fell into poverty as a consequence of economic changes is being undertaken.

1998 LSMS results show that 35.6 per cent of the population of Mongolia are poor, of which 34.1 per cent belong to Ulaanbaatar (capital) city and 32.6 per cent to rural areas. This is a decrease of 0.5-1.0 point compared to the poverty incidence in 1995.

During the last three years, rural poor have increased by 5.7 per cent and in urban areas by 4.8 per cent.

Poverty is strongly correlated with employment and unemployment. By the end of 1998, the labor force participation rate was 68.4 per cent, employment rate - 64.4 per cent and unemployment rate - 5.8 per cent, of which for female these ratios are 64.6, 60.5 and 6.4 per cent respectively. More than half of all unemployed is female.

According to the 1998 LSMS poverty depth (P1) ranges from 9.8 per cent to 13.9 per cent and poverty severity (P2) increased by 0.8-2.9. These imply that there has been an increase in the gap between average consumption expenditure of the poor and the poverty line and demonstrate that purchasing power of the poor in urban and rural areas to buy basic needs is falling. The average consumption expenditure of the poor is 10 % below the poverty line. This expresses that the number of poor is increasing, and self-sustainability, to obtain basic essentials, is worsening and the results of government measures against poverty alleviation are not clear. As well as far the country goes into a market economy structural reforms in the economy and society are directly affecting the wealth and income distribution of the population and household. However, these results need to be interpreted with some caution.

The Gini coefficient, showing the unequal income distribution reached 0.35 per cent in 1998 from 0.31 per cent in 1995.

The pattern of poverty in Mongolia is unusual. In most countries, poverty is concentrated in rural areas, and poor people in urban areas live in better living conditions in many ways than the poor in rural areas. In contrast the living in rural areas in Mongolia is better and as a result some people move to rural areas. It might be the results of the many measures on structural adjustments in the economy and society, privatization, promotion of private businesses and small to medium enterprises. Livestock privatization has boosted household business and a majority of herders- households stay in rural areas. Most industries, services and markets are concentrated in the capital city while large scaled industries and services in aimag centers have been replaced by small services and trade units, which barely sustain the needs of households. This has contributed to increase in unemployment and expansion in poverty in aimag centers. Therefore, poverty alleviation program and measures should be targeted at urban areas especially to aimag centers.

The majority of Mongolian poor households are headed by females. This can be partially attributed to the implementation of pronatalist population policy promoting high fertility rates and provision of free childcare and prenatal care at maternity homes. Due to the transition, these incentives were reduced or eliminated, which impacted heavily on urban and rural female headed households.

Regardless of household location poor households are big sized.

There is a tendency for increase of poverty incidence among the employed population. This is directly related to labor cost.

In order to study the nation's living standards, the National Statistical Agencies of the transition countries have conducted a family budget survey for about 30 years. This survey was organized by the sectorial selection method.

The system of continuous observation becomes extremely burdensome for the families surveyed. This circumstance obviously has an adverse effect on the quality of the entries.

In the transition to a market economy, the sectoral selection method does not guarantee representativeness in the case of family budget surveys. At present the statistical agencies conduct household surveys by the snapshot method.

The World Bank, Asian Development Bank, UNDP and TACIS give methodological and technical support for conducting the sample survey of households in transition countries.

As a result, most of the transition countries are producing and releasing the basic indicators of poverty according to internationally recommended concepts and definitions.

However, there are more work needs to be done to strengthen the analytical capacity of National Statistical agencies in order to utilize the survey data.