Vulnerability to poverty in the Mekong region countries

Views:
 
     
 

Presentation Description

Presented by Prof. Hermann Waibel, University of Hannover, Germany.

Comments

Presentation Transcript

Vulnerability to Poverty in Mekong Region countries – data collection and methodological issues Hermann Waibel Leibniz University Hannover http://www.vulnerability-asia.uni-hannover.de/: 

Vulnerability to Poverty in Mekong Region countries – data collection and methodological issues Hermann Waibel Leibniz University Hannover http://www.vulnerability-asia.uni-hannover.de/ Food for thought seminar series at the World Fish Centre, Penang Malaysia, February 20, 2012

Outline of the Presentation: 

Outline of the Presentation Background and Motivation of the Project The Vulnerability to Poverty Concept Data collection procedures Some Results (examples) Non sampling error Risk experiments vs survey- based risk questions Future Plans

Motivation….... : 

Motivation…....

Motivation : Poverty Experiments „ Pedicab Driver in the Philippines“ : 

Motivation : Poverty Experiments „ Pedicab Driver in the Philippines“

Motivation: Poverty Experiments „Street Vendor in Manila“: 

Motivation: Poverty Experiments „Street Vendor in Manila“

The DFG Research Project for Vulnerability to Poverty in Asia: 

The DFG Research Project for Vulnerability to Poverty in Asia Objective: “To advance the concept and the measurement of vulnerability to poverty and assess the sources of vulnerability through interdisciplinary economic research in Thailand and Vietnam”

PowerPoint Presentation: 

Project Concept Agricultural Economics Financial Institutions Economic Geography Vulnerability + HH - dynamics Vulnerability database Agricultural Economics Financial Institutions Economic Geography Vulnerability + HH - dynamics Vulnerability database

PowerPoint Presentation: 

Project Areas

PowerPoint Presentation: 

V h : vulnerability c h : consumption of HH z: poverty line p: probability EP: expected poverty Vulnerability to Poverty (working concept)

PowerPoint Presentation: 

Assets Income (I) Poverty line Transient poverty 0< VTP<1 No poverty VTP=0 Poverty line VTP=0.5 Probability to be poor Chronic poverty VTP=1 Structural poverty Stochastic poverty Source: Chiwaula, Witt & Waibel, JDS, 2011 Theoretical considerations

PowerPoint Presentation: 

Vulnerability Survey Survey locations and sample size: 6 provinces in Thailand and Vietnam Purposive selection of provinces („typical“) 220 rural subdistricts/communes Sampling: PPS 440 villages 2 villages per subdistrict/commune Fixed sample per village: 10 HH 4400 households in rural and peri-urban households Rural Household Survey in 2007, 2008, 2010, (2011) Migrant tracking survey in 2010

Organization of Vulnerability Survey: 

Organization of Vulnerability Survey Provincial Supervisor Country Coordinator Hannover Supervision Team

Supervision of Enumerator Team in Vietnam: 

Supervision of Enumerator Team in Vietnam

Questionnaire: 

Questionnaire

PowerPoint Presentation: 

Data Collection: Shock Experience

PowerPoint Presentation: 

Data Collection: Risk Expectations

Panel Data size and Attrition: 

Panel Data size and Attrition No. of questionnaires (rural) 2007 2008 2010 2011 Migrant survey (2010) T 2186 2136 2071 917 653 (Bangkok) V 2195 2148 2100 675 310 (Ho Chi Minh City) (1 province per country) Attrition rates 2007 2008 2010 2011 T 0.6% 2.3% 3.0% 2.13% V 0.2% 2.1% 2.2% 2.18% (1 province per country)

Selected Results: Example 1: 

Selected Results: Example 1 Non-sampling Error and Data quality: What can we learn from Surveys to collect Data for Vulnerability Measurements? Tung Duc Phung, Bernd Hardeweg, Suwanna Praneetvatakul and Hermann Waibel (for „ World Development“ )

Types and Sources of Non-sampling Errors: 

Types and Sources of Non-sampling Errors Coverage error Non- response error Measurement errors Coverage error: occurs when the list from which the sample is selected does not cover the right targeted population Non-response error: results from failure to obtain observations from some sample units or some items Measurement errors: that occurs when the data obtained are not correct or different from actual values of the sample units

PowerPoint Presentation: 

All sections Sensitive Questions Log of interview duration (minute) 15.040*** -0.034 Interview in morning (dummy) -8.331*** -2.948*** Interviewed in the afternoon (dummy) -6.567*** -2.431*** Harvested time (dummy) 3.569*** -0.040 Sex of interviewer (dummy) -4.560*** -1.863*** Age of interviewer (year) -2.052*** -0.765*** Square age of interviewer 0.033*** 0.009*** Local interviewers (dummy) -3.446*** -0.661** Sex of respondent (dummy) 2.037* 0.752** Ethnicity of respondent (dummy) -3.755*** -2.317*** Model 1: Explaining Missing Value Note: selected variables only, (*); (**) ; (***) significant at 10%, 5% and 1% respectively

PowerPoint Presentation: 

All sections Sensitive Questions Age of respondent (year) 0.809*** 0.040 Square age of respondent -0.008*** -0.001 Respondent education 1.417*** -0.023 Square respondent education -0.078*** -0.005 Respondent is household head -3.704*** -0.514* Interaction of gender -1.057 -1.030*** Number of observation 8223 8223 Adjusted R2 0.637 0.219 Model 1 (continuation) Note: selected variables only, (*); (**) ; (***) significant at 10%, 5% and 1% respectively

Selected Results - Example 2: Experimentally-validated survey evidence on individual risk attitudes in rural Thailand: 

Selected Results - Example 2: Experimentally-validated survey evidence on individual risk attitudes in rural Thailand Bernd Hardeweg, Lukas Menkhoff and Hermann Waibel (to be published in „ Economic Development and Cultural Change“ )

Data items: 

[ 23 ] Data items 1. Self-assessment of general risk attitude: “Are you generally a person who is fully prepared to take risks or do you try to avoid taking risks?” Response scale 0 – 10 0 meaning “not at all willing to take risks” 10 meaning “fully prepared to take risks” 2. Hypothetical investment “Imagine you had just won 100,000 Baht in a lottery and you can invest this money in a business. You are not sure if the business goes well or not. If it goes well you can double the amount invested after one year. If it does not go well you will lose half the amount you invested. What fraction of the 100,000 Baht would you invest in the business?” domain-specific, more concrete, large pay-off 3. Experimental elicitation of risk attitude 4. Behavioural indicators: Purchase of lottery tickets Self-employment

2. Data - Experimental measure of risk attitudes: 

2. Data - Experimental measure of risk attitudes 3. Experiment conducted after completion of the survey form x

Results: Risk attitudes, survey based: 

Results: Risk attitudes, survey based

PowerPoint Presentation: 

Results: Risk attitudes, experiment

Summary Results Example 2: 

[ 27 ] Summary Results Example 2 Large heterogeneity in risk preferences usefulness of individual indicators Individual characteristics and risk attitude female risk aversion not significantly higher unexpected signs for household size, being married and being civil servant Self-assessment of general risk attitude validated by experimental results single significant predictor consistency with observed behaviour Closer fit for better educated sub-group indicates need to adjust survey items/experiment sequential approach might support less educated respondents

Plans for an interdisciplinary project..: 

Plans for an interdisciplinary project..

Sketch of future Research Project: 

Sketch of future Research Project Climate change, vulnerability to poverty and inclusive growth in the GMS region (preliminary Title) Use concepts developed in vulnerability project and expand the existing household and migrant data base from Thailand and Vietnam to include Yunnan (China) Stung Treng (Cambodia), Savannakhet (Lao PDR) Interdisciplinary approach to assess climate change and natural resource degradation on inclusive economic growth, poverty and inequality, food security and structural change in agriculture, labour markets and migration, regional trade and other issues of regional cooperation. Bio-Economic and econometric modelling (Climate, water resources, pathogens, production potential, factor and product markets, trade)

Possible Research Partners: 

Possible Research Partners Research Groups in Economics and Natural Science of Hannover University Research Group in Economics of Goettingen University Research Institutes and regional Universities in Vietnam Research Institutes and regional Universities in Thailand Centre for Chinese Agricultural Policy (CCAP) of the Chinese Academy of Sciences Research Institutes and regional Universities in Laos Research Institutes and regional Universities in Cambodia Asian Development Bank IWMI Other CGIAR Centres

Thank you for your attention ! (….“my plan for 2013“…..): 

Thank you for your attention ! (….“my plan for 2013“…..)

Potential and Actual Research Output: 

Potential and Actual Research Output PalgraveMacmillan Book (2012): Vulnerability to Poverty: Concepts, Measurement, and Data Issues (Klasen/Waibel eds) Part I: Introduction Part II: Conceptual and Measurement Issues of Vulnerability Part III: Empirical Applications (Food Price Schock, Diversification, Schock and Coping Interactions and others) Special Issue in World Development (2013): Vulnerability as downside risk (Povel/Klasen) Vulnerability and risk attitude (Glöde/Menkhoff/Waibel) Non sampling error in vulnerability surveys (Tung/Waibel et al) Vulnerability, Shocks and Recovery Time (Junge/Diez/Schätzl) Does Migration reduce vulnerability (Grote/Nguyen/Raabe) Nutrition-based Wages and Employment Vulnerability (Strulik).

PowerPoint Presentation: 

Typical Age structure in villages Sources: DFG FOR756 Household survey 2008, Thailand: Institute for Population and Social Research (2003)

PowerPoint Presentation: 

Consumption distributions – cross-country comparison Source: DFG FOR756 Household surveys 2007, 2008. Source: DFG FOR756 Household survey 2007.

Vulnerability levels in Thailand and Vietnam: 

Vulnerability levels in Thailand and Vietnam Source: FOR 756 household survey panel , own calculations