UNITED NATIONS ECONOMIC COMMISSION FOR EUROPE CONFERENCE OF

UNITED NATIONS ECONOMIC COMMISSION FOR EUROPE CONFERENCE OF EUROPEAN STATISTICIANS Seminar on Poverty Measurement Geneva, 5-6 May 2015 CONSTRUCTION AND UPDATE OF A POVERTY MEASURE WHY, WHAT AND HOW João Pedro Azevedo Global Lead Minh Nguyen Economist

WHAT IS POVERTY? Poverty is pronounced deprivation in well-being, and comprises many dimensions Basic concept of monetary poverty Multidimensional aspects Low incomes and the inability to acquire the basic goods and services necessary for survival with dignity Low levels of health and education Poor access to clean water and sanitation Inadequate physical security, lack of voice Insufficient capacity and opportunity to better one’s life

WHAT ARE THE PRINCIPLES OF A GOOD POVERTY MEASURE? It must be understandable and easy to describe It must conform to a common sense notion of poverty It must fit the purpose for which it is being developed It must be technically solid It must be operationally viable It must be easily replicable 3

WHY DO WE WANT TO UPDATE OUR MEASURE POVERTY? To better compare the poverty of different households and regions in the country To better compare poverty over time To better compare poverty between countries at a global level To better compare poverty between countries at a regional level To better monitor poverty To better define a poverty reduction strategy To better evaluate the impact of a poverty reduction strategy The answer to some of these questions will condition the properties and choices of a poverty indicator 4

POVERTY MEASURES Relative Monetary Absolute Objective Basic Needs Approach Poverty Non-monetary Anthropometri c needs Subjective 5 Direct surveys

HOW TO MEASURE MONETARY POVERTY? 1. A welfare measure for individuals, used to derive a distribution of living standards 2. A poverty line, threshold below which individuals are classified as poor 3. A poverty index, summary statistics of poverty in population Non poor Threshold Poor

HOUSEHOLD WELFARE Consumption and income are most widely used Some theoretical advantages of consumption measures Some policy advantages of income measures Both types of household welfare measure need proper adjustments in terms of: Price differences (CPI; Purchasing Power Parity; etc.) Normalize the size of a household (Per capita vs. per adult equivalent scale)

ADJUSTMENTS FOR HOUSEHOLD EXPENDITURES: PRICE ADJUSTMENT Which is more? 20 Somoni in 1990, TJK vs. 30 Somoni in 2010, TJK Adjustment - Use CPI 30 Somoni in 2013, TJK vs. 6 in 2013, USA Adjustment - Use PPP 20 Somoni in 2010, Dushambe vs. 20 Somoni in 2010, GBAO Adjustment - Use Spatial Price Index

ADJUSTMENTS FOR HOUSEHOLD EXPENDITURES: ADULT EQUIVALENCE SCALE VS. PER CAPITA Do children spend less than adults? Many countries replace the number of children with adult equivalent scales (based on daily caloric needs) Maybe food consumption but non-food consumption? The World Bank usually recommends “per capita” for simplicity and transparency

POVERTY LINES International measures 1.25 and 2 per day per capita evaluated at 2005 PPP Regional measures 2.5 and 4 for LAC and 2.5 and 5 for ECA National measures Country specific poverty lines, which are linked to a daily calorie requirement (usually around 2100 kcal per capita per day)

NORMATIVE OR EMPIRICAL BASKETS? Nothing ensures that the items of a normative baskets will be available for consumption There is no guarantee that these items have prices that can be estimated robustly Nothing ensures that the products from a normative food basket can be consumed by a significant fraction of the population (products that are not part of the population food habits). The normative baskets, use multiple criteria (caloric, protein intake, micro nutrients, % protein of animal origin, etc.) end up being ad hoc baskets that depend largely on the expert who built them. The criteria of transparency, acceptance by consensus and reproducibility are hardly met in this case. 11 An empirical basket is preferable

WHY DO WE NEED AN INTERNATIONAL DOLLAR PER DAY POVERTY LINE? National official poverty lines are a function of preferences, norms and level of economic development Note: the ECA Team for Statistical development would like to thank Shaohua Chen and Prem Sangraula for sharing their data 12

REPLICATING A SIMILAR EXERCISE ONLY WITH THE ECA COUNTRIES GIVES US AN AVERAGE REGIONAL POVERTY LINE OF 2.5 USD-PPP 13

FOR ANY GIVEN METHODOLOGY TO ESTIMATE A POVERTY LINE, FINAL RESULTS CAN BE DRIVEN BY Preferences Countries, households or individuals might have different aspirations, as their level of economic development increase. Relative or Weakly-Relative Poverty line argument Circumstances There might be exogenous structural factors that might affect the consumption patterns of different countries at different points in time Demographics, Natural endowments, Climate 14

AND BOTH DEMOGRAPHIC AND CLIMATE CIRCUMSTANCES STAND OUT IN ECA The average dietary energy requirements is positively correlated with the national official absolute poverty lines, and show very distinct regional patterns, reflecting distinct demographic compositions. Source: FAO 15 Many of the ECA countries are among the few with sub-zero average monthly temperatures. Source: World Bank Climate Change Knowledge Portal ( http://climateknowledgeportal.worldbank.org)

POVERTY INDEX Poverty Headcount Rate The most used measure, including WBG Goal of “Ending Extreme Poverty” and MDG 1 A proportion of population below the poverty line A limitation: Insensitive to how far from the poverty line the poor are. Other indices like Poverty Gap Index and Severity of Poverty Index can address this issue

IN A FEW REGIONS THERE ARE IMPORTANT DIFFERENCES BETWEEN THE GLOBAL AND NATIONAL POVERTY NUMBERS and reflects the different levels of economic development 17

BUT ANY MEASURE OF POVERTY IS JUST A MEANS TO AN END. THE MOST IMPORTANT OBJECTIVE IS TO UNDERSTAND HOW POLICIES CAN HELP REDUCE THE POVERTY RATE TURKEY (5 USD-PPP) 2002-2011 2 0 -0.1 -2 -4 -2.3 -3.3 Source: Azevedo and Atamonov (2013) -12.7 -20 -20.0 Source: Azevedo, Atamonov and Rajabov (2013) property income remittances -24 pensions other income agricultural income remittances pensions social assistance wages share of employed -0.7 -1.3 -16 social assistance -11.6 -14 18 -0.4 -12 other income -10 -12 -1.9 wages -7.1 share of employed -7.0 Percentage points -6 -8 -0.9 -8 share of adults -3.3 share of adults percentage points 2003-2009 1.2 0.1 0 -4 TAJIKISTAN (2.5 USD-PPP)

BY HELPING IDENTIFYING THE WINNERS AND LOSERS OF POLICY CHOICES MADE IN OUR COUNTRIES Top 60 Top 60 Those in the bottom 40 percent can differ over time for several reasons: Bottom 40 Period T 19 Demographics Migration Markets Incentives Bottom 40 Period T 5

AND ULTIMATELY UNDERSTAND THE DRIVERS OF INTRA AND INTER GENERATIONAL MOBILITY CLASS TRANSITIONS IN TAJIKISTAN FROM 2003 TO 2009 2003 Poor Vulnerable Middle class Total 2009 Origin (in 2003) Poor Vulnerable 71.8 22.1 6.1 100 48.4 29.2 15.6 43.6 38.8 46.7 54.5 40.8 Middle Class 12.8 24.1 29.9 15.6 100 100 100 100 Source: Azevedo, Atamanov and Rabajov (2013) based on TLSS. Notes: Transition matrix is based on synthetic panel for 2003-2009. Threshold for poor is 162 somoni and for middle class 294 somoni per month. Explanatory variables include year of birth cohort, number of children, education of the head of household, rural/urban and regional dummies, and different interactions between these variables and means are the regional level. Sample: head of households 25-55 years of age. 20

FINAL CONSIDERATIONS Clarity, transparency and simplicity are the most critical aspects of any poverty measure Clear institutional architecture Open methodological guidelines Access to the survey data Clear calendar for the release of the numbers

REFERENCES Azevedo, Joao Pedro & Atamanov, Aziz & Rajabov, Alisher, 2014. " Poverty reduction and shared prosperity in Tajikistan : a diagnostic,"Policy Research Working Paper Series 6923, The World Bank. Azevedo, Joao Pedro & Atamanov, Aziz, 2014. " Pathways to the middle class in Turkey : how have reducing poverty and boosting shared prosperity helped? ,"Policy Research Working Paper Series 6834, The World Bank. Deaton, A., & Zaidi, S. (2002). Guidelines for constructing consumption aggregates for welfare analysis. Living Standard Measurement Survey Working Paper No. 135. Ravallion, M. (1998). “Poverty Lines in Theory and Practice.” LSMS Working Paper No. 133,World Bank, Washington, DC. Ravallion, M. (1992). “Poverty Comparisons: A Guide to Concepts and Methods.” LSMS Working Paper No. 88, World Bank, Washington, DC. Ravallion, M. (2008). “Poverty Lines." The New Palgrave Dictionary of Economics. Second Edition. Eds. Steven N. Durlauf and Lawrence E. Blume. Palgrave Macmillan. The New Palgrave Dictionary of Economics Online. Palgrave Macmillan. 17 September 2013 Ravallion, M. (2012). “Poor or Just Feeling Poor” World Bank Policy Research Working Paper 5968.