Poverty Measurement in Tajikistan Statistics Agency under the

Poverty Measurement in Tajikistan Statistics Agency under the President of Tajikistan

Poverty Measurement in Tajikistan: Background • There was some documentation on the methodology which was prepared together with the World Bank during the Living Standards Survey in 1999, 2003, 2007 and 2009 and was published in reports. • Food poverty line was defined as a cost to purchase food for intake of 2, 250 calories per day per capita • Changes in the poverty headcount at $2. 15 Purchasing Power Parity (PPP) – absolute poverty line in 1999, 2003, 2007 and 2009 (%)

Changes in the absolute poverty line based on LSS in 1999, 2003, 2007 and 2009 82 90 80 64 70 53. 5 46. 7 60 50 40 30 20 10 0 1999 2003 2007 2009

BACKGROUND • • • Household Budget Surveys (HBSs) are conducted in Tajikistan from 1952. Sample size was 1, 250 households. Starting from 1992, due to complicated political, economic and financial situation in the country, the HBS sample size was halved (600 households). Starting from 2000, as the political and economic situation in the country improved and there was a stronger demand for HBS data, the sample size was increased up to 925 households. Government Resolution No. 497 dd. 1 October 2008 increased the number of surveyed households up to 3, 000 (1, 150 households in urban areas and 1, 850 households in rural areas). Starting from 2009, HBSs are conducted in all 5 regions of the country. Surveyed households by regions: • Dushanbe – 400 h/h • Sughd – 860 h/h • Khatlon – 900 h/h • Gorno-Badakhshan region (GBR) – 240 h/h • Regions of republican subordination (RRS) - 600 h/h

Purpose of HBS • • • Household budget statistics is one of areas of economic statistics addressing living standards of the population. Budget surveys contain data on the role of specific sources in income, dependence of consumption on income and help to review changes in the consumption demand over time. In addition to characterizing specific population subgroups, HBS data are widely used in various social and economic estimations: • for estimating gross domestic product and its distribution; • for calculating real population income; • for preparing a balance between agricultural production and consumption; • for calculating cost-of-living index (index of consumption prices for goods and services); • in the national accounts system where households indicators play an important role, etc.

Activities of the Statistics Agency together with the World Bank in 2012 -2014 1. Trainings and study tours for Statistics Agency’s and Mo. ED’s specialists on the collection and processing of household survey data (India, Switzerland, Kazakhstan and Kyrgyzstan arranged by the World Bank); 2. New data entry program based on СSPro and trainings delivered for HBS specialists on the data entry and processing for all 5 regions; 3. Two trainings on poverty measurement based on SPSS programme for the Statistics Agency and Mo. ED (April and June 2014) 4. Advice on estimating poverty levels and middle class 5. Methodological Guidelines on Poverty Measurement in Tajikistan 6. Trainings for HBS interviewers in all regions of the country and familiarization with a new questionnaire for an integrated household budget and labour force survey

National methodology • The development of the national methodology for poverty measurement was set to be a priority in the National Development Strategy up to 2015. • In 1999 -2009, Tajikistan used Living Standards Survey (LSS). • Starting from 2009, for poverty estimates Tajikistan has been using economic growth and poverty elasticity for poverty forecasts. The limitations of this approach imply that over time poverty estimates generated by using this method would not be reliable. • Starting from the second half of 2012, HBSs are used for measuring poverty in the country and the resultant figures are not comparable with the figures generated prior to 2009.

New methodology • Household Budget Survey (HBS) will be used for poverty measurements under the new methodology • Prepared under the leadership of the Statistics Agency and in cooperation with the Ministry of Economic Development and the World Bank • Internationally recognized methodology

Purpose and actions • Short-term solution (2012 -2016) – New approach – immediate solution • Long-term solution (post 2016) – HBS Enhancement Project to support the development of integrated HB and LS survey • Supplement to a questionnaire • Changes in the sampling design methodology • Changes in the field survey procedures • 2015 -2016 will be a period of transition from the short-term to the longterm solution during which some changes will be introduced stepwise

Poverty Council: expectations • Attendance of technical meetings • Securing confidence as an active member of the Poverty Council • Sharing practice on similar financial and technical assistance received earlier, both domestically and abroad

Old vs. new method • Advantages – HBS is a national property – A sample size of 3, 000 households provides more opportunities for measuring poverty (2009 LSS had a sample size of 1, 500 households) – HBS is a continuous survey (households are visited 4 times a year), less subject to bias and enables taking into account seasonality

Old vs. new method • Limitations – Does not fully match LSS methodology – Outdated data entry system in HBS – HBS sample from the 2000 census – Questionnaire does not reflect all necessary aspects and needs of the new policies – Data collection and analysis processes can be improved – Data access and dissemination can be improved

Welfare aggregate Total household expenditures per capita Adjusted food basket quantity Food basket quantity Spatial deflator Calorie conversion factors Food basket cost prices Food basket unit prices Selection (a priori) of a reference group Calorie intake standard Deflated expenditure percentile (all households) Poverty scope Total poverty line Food basket cost Engel’s coefficient

Welfare Aggregate: Composition •

Welfare Aggregate: Composition • Welfare aggregate can be expressed as a sum of two welfare subaggregates THC = FHC + NFHC • where: – THC = total household consumption – FHC = food household consumption – NFHC = non-food household consumption

Poverty line – required calories • 2, 250 k. Cal/day is set to be the required average calorie intake per capita in Tajikistan • In Tajikistan households were selected in 3 -36 welfare deciles • Median value of 1 calorie for the selected group of households is estimated to be 0. 0017 somoni/kcal • Food poverty line = 114. 28 somoni/month

Poverty line (non-food) • Non-food component is calculated as follows: – First, identify those households whose total expenditures are close to the food poverty line – Second, calculate the ratio of food to total consumption (Engel’s coefficient) – Third, divide the food poverty line by Engel’s coefficient • Mean Engel’s coefficient for the selected households was 77. 86%. • The sum of food and non-food component give a poverty line of 146. 77 somoni/month

Poverty Line

Welfare Aggregate: results Somoni/month % food % non-food 233. 89 60. 4% 39. 6% 243. 73 62. 1% 37. 9% 181. 46 74. 2% 25. 8% 176. 88 73. 4% 26. 6% 153. 65 73. 0% 27. 0% Urban 223. 77 63. 5% 36. 5% Rural 187. 95 71. 3% 28. 7% Q 1 (the poorest) 90. 24 78. 7% 21. 3% Q 2 134. 64 76. 4% 23. 6% Q 3 174. 13 74. 1% 25. 9% Q 4 224. 69 69. 4% 30. 6% Q 5 (the richest) 375. 44 59. 8% 40. 2% 199. 83 68. 4% 31. 6% Groups Dushanbe Sughd Region Khatlon RRS GBR Location Quintile TOTAL

Poverty indicators, by regions 60% 50. 9% 50% 45. 9% 41. 3% 40% 36. 4% 30% 20% 40. 0% 29. 3% 23. 5% 21. 5% 10% 0% Tajikistan Dushanbe Sogd Khatlon RRP GBAO Urban Rural

Poverty indicators: urban vs. rural urban 26. 7% rural 73. 3%

Poverty indicators, by regions RRS 29. 6% GBR 11. 7% Dushanbe 6. 5% Khatlon 35. 7% Sughd 16. 6%

Long-Term Solution Action Plan • • • Q 1 2015 – Publication and dissemination of the national methodology Q 1 2015 – Amendments to the poverty measurement methodology Q 3 2015 – Data entry Q 3 2015 – Sampling Q 3 2015 – Questionnaire

Methodology Review • Questionnaire • Methods • Sample • Field activities • Data management • Publications of results Multidimensional To reduce the period for keeping a diary To introduce recall method Frequent visits Based on the 2009 population census Inspections Performance indicators New data entry system New automated output tables

Thank you!