Monitoring Poverty More Regularly: Can labor force surveys and market research data help? Joao Pedro Azevedo LCSPP 1 More Frequent, More Timely & More Comparable Data for Better Results 20 April 2011
The need Food crisis Environmental shocks Financial crisis Policy makers need to understand the consequences of such crisis to design responses More Frequent, More Timely & More Comparable Data for Better Results 220 April 2011
How can we handle this? Econometrics + Assumptions in Famous Poverty-Growth Elasticity Imputation methods Collect new data and/or collect old data in different ways Continuous household surveys Listening 2 LAC Use existing information Labor force surveys Market research data Administrative records (i.e. vital statistics) 20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 3
Two examples of using existing data in a different way Labor Income Poverty Index (LIPI) – Poverty and Labor Brief Prices and Quantities information from Nielsen on food and beverages 20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 4
LABOR INCOME POVERTY INDEX 20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 5
Labor Income Poverty Index (LIPI) How we do it: – Labor Income (70% of total income) – 12 months moving average or 3 months moving average – Poverty line (national + usd/ppp) – Price data (monthly or quarterly) – Index Q (base=1) Mexico March,2010 (CONEVAL) 20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 6
Labor Income Poverty Index (LIPI) (U$ 2.5 a day) Jan Jul April 2011 More Frequent, More Timely & More Comparable Data for Better Results 7
MEXICO - Labor Income Poverty Index (LIPI) (U$ 2.5 a day) Q – Q April 2011 More Frequent, More Timely & More Comparable Data for Better Results 8
Testing the validity Colombia and Peru Continuous household survey Design – 2 random samples Sample 1 : Poverty incidence (full income) Sample 2 : LIPI (labor income) 20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 9
Sensitivity to the welfare aggregate in Colombia 20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 10 Poverty3 ma12 ma Extreme Moderate
Sensitivity to the welfare aggregate in Peru 20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 11
Sensitivity to poverty and inequality measures CORRELATIONS Moderate poverty LevelChange fgt (0) fgt (1) fgt (2) Extreme poverty LevelChange fgt (0) fgt (1) fgt (2) Inequality LevelChange gini SPEARMAN Moderate poverty LevelChange fgt (0) fgt (1) fgt (2) Extreme poverty LevelChange fgt (0) fgt (1) fgt (2) Inequality LevelChange gini April 2011 More Frequent, More Timely & More Comparable Data for Better Results 12
MARKET RESEARCH DATA 20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 13
More Frequent, More Timely & More Comparable Data for Better Results RMS – Nielsen Areas 95% Population coverage Zona no cubierta 5% población Atlántico: Atlántico, Bolívar, Cesar, Córdoba, Guajira, Magdalena, Sucre Atlántico: Atlántico, Bolívar, Cesar, Córdoba, Guajira, Magdalena, Sucre Antioquia: Antioquia Oriente: Boyacá, Meta, N. Santander, Santander Cundinamarca: Cundinamarca Centro: Caldas, Huila, Quindío, Risaralda, Tolima Pacífico: Cauca, Nariño, Valle 20 April
More Frequent, More Timely & More Comparable Data for Better Results Categories included: 20 April
More Frequent, More Timely & More Comparable Data for Better Results Countries covered in the LAC Region Argentina Brasil Chile Mexico Venezuela CAM – Panama – Nicaragua – Costa Rica – Honduras – Guatemala – El Salvador Information included: – Volume sales variation per country – Consumption Price variation per Country – By Item, Outlet and Geographic location 20 April
More Frequent, More Timely & More Comparable Data for Better Results 20 April
Volume and Price series 20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 18
20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 19
20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 20
20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 21
20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 22
Applications Tracking of wellbeing, proxy by consumption Reweighting of the index to better reflect the consumption patters of the poorer Better estimation of elasticities and cross- elasticities of substituion Distributional analysis 20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 23
Final considerations Labor income explains most of the poverty variation High correlation between income, expenditure and labor income poverty measures Highlights the importance to improve income measure in the region Market research data can help 20 April 2011 More Frequent, More Timely & More Comparable Data for Better Results 24