DFG Research Unit 756 Collecting Data for Vulnerability Measurements: some initial findings from Thailand and Vietnam Hermann Waibel - PowerPoint PPT Presentation

Loading...

PPT – DFG Research Unit 756 Collecting Data for Vulnerability Measurements: some initial findings from Thailand and Vietnam Hermann Waibel PowerPoint presentation | free to download - id: 66437c-MmUyZ



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

DFG Research Unit 756 Collecting Data for Vulnerability Measurements: some initial findings from Thailand and Vietnam Hermann Waibel

Description:

DFG Research Unit 756 Collecting Data for Vulnerability Measurements: some initial findings from Thailand and Vietnam Hermann Waibel Seminar at Thailand Development ... – PowerPoint PPT presentation

Number of Views:8
Avg rating:3.0/5.0
Date added: 5 December 2019
Slides: 49
Provided by: Michael2231
Learn more at: http://www.vulnerability-asia.uni-hannover.de
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: DFG Research Unit 756 Collecting Data for Vulnerability Measurements: some initial findings from Thailand and Vietnam Hermann Waibel


1
DFG Research Unit 756 Collecting Data for
Vulnerability Measurements some initial
findings from Thailand and Vietnam Hermann
Waibel
Seminar at Thailand Development and Research
Institute, September 25, 2009
2
Outline
  • Project Objectives
  • Why Vulnerability Research?
  • Data Collection to Measure Vulnerability
  • Results of 1st Phase Some Examples
  • Plan for 2nd Phase (2010-2012)

3
Overall objective of the Project
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
4
Expected Results
  1. A comprehensive conceptual and empirical basis
    for the assessment of vulnerability to poverty in
    Thailand and Vietnam
  2. The relative importance of the sectoral and
    regional determinants of vulnerability
  3. Strategies for more timely, better targeted and
    more cost effective social risk management

5
Project Components
6
Study Areas
7
Motivation for the project
  • I. Development Policy
  • Pockets" of poverty in emerging market economies
  • Poverty is a dynamic issue
  • Shocks make people vulnerable
  • How efficient do HH cope with shocks?
  • What policy interventions are appropriate?

8
Motivation for the project
  • II. Research
  • Collection of data specifically for measuring
    vulnerability
  • Explore the black box (risks and shocks)
  • Comparative study of vulnerability concepts
  • Understand Household decision processes (e.g.
    dynamics, multi-location)
  • Risk matters From fate to fear!

9
Why vulnerability Research?
  • (1) Reducing vulnerability has an intrinsic value
  • Well-being having good prospects now to have
    enough in the future
  • (2) Reducing vulnerability has an instrumental
    value
  • Poverty trap ex ante risk mitigating prevents
    adoption of high average return portfolios

10
Poverty Typology and Measurements
  • Criterion Description
  • Perspective ex post, ex ante
  • Time Horizon static, dynamic
  • Welfare indicator consumption, income,
  • assets, non-monetary
  • Poverty Type chronically-poor,
    transient-poor
  • - structural
  • - stochastic

11
Asset based Vulnerability
Income (I)
Asset poverty line VTP0.5
Structural poverty
Poverty line
Probability to be poor
Assets
Transient poverty
0ltVTPlt1
12
Four prominent definitions of vulnerability
  1. Uninsured exposure to risk and shocks (e.g. Jalan
    and Ravallion, 1999)
  2. Expected Poverty (e.g. Pritchett et al., 2000)
  3. Low level of expected utility (Ligon and
    Schechter, 2003)
  4. Expected deprivation or individual vulnerability
    to poverty (Calvo and Dercon 2005)

13
Vulnerability as Expected Poverty (VEP)
Vh vulnerability ch consumption of HH z
poverty line p probability EP expected
poverty
14
Measurement of vulnerability (Calvo and Dercon
2007)
Vulnerability as expected deprivation -
accounting for the probabilities of negative
future events and their severity
  • V V(z,y,p)
  • z benchmark, i.e. poverty line
  • y vector of outcomes across n states of the
    world
  • p a vector of corresponding probabilities.

15
Data Collection Procedure
  • Limit to Thailand and Vietnam
  • Large Provincial sample gt 400 - 1000
  • Panel data, (at least four waves planned)
  • Multi-location households, (include the migrants)
  • Combine different fields of economics
  • Welfare Economics and Poverty Dynamics
  • Agricultural Economics
  • Financial Institutions
  • Economic Geography

16
Sampling procedure
Study Areas and sampling procedure
purposive sample (location, per capita income)
province
stratified random sample (e.g. population
density, share agricultural households)
districts
villages
random sample
random sample (based on available lists)
households
2 surveys à 4,400 respondents
17
Modules of questionnaire and variables to measure
vulnerability
Module / Content Module / Content Main Variables Contribution to vulnerability research
1 General survey information Location Location factors, Poverty and vulnerability mapping
2 Actual Household Characteristics Size, composition and dynamics, education, health Testing and advancing the concept
3 Shocks experienced during past five years and perceived risks for the next five years Type, timing, duration, scope, severity, financial consequence, ex post coping measures, covariance, subjective assessment of risk and well-being type, frequency, severity, consequence of expected risks and ex ante mitigation measures Vulnerability concept, causes of vulnerability, coping and mitigation strategies
4 Land, Agriculture and Natural Resources Land size and ownership status, land value, crop and livestock technology and production, self-consumption, productivity, costs, returns, timing and extent of natural resource extraction Source of income, causes of vulnerability
18
Modules of questionnaire and variables to measure
vulnerability
Module / Content Module / Content Main Variables Contribution to vulnerability research
5 Off-farm employment including wage labour Type, contractual arrangements, location, travel costs, job acquisition costs, duration of work, wage and fringe benefits Source of income, causes of vulnerability
6 Non-farm self- employment including cottage industries Type, investment, costs and returns Source of income, causes of vulnerability
7 Borrowing, Lending, Public and other transfers and insurance Type, sources, contractual arrangements, conditions amounts, payment frequencies Source of income, causes of vulnerability
8 Household Consumption Food and non food items, other expenditures Consumption
9 Assets including house Purchase value depreciation and service life Coping capacity
19
Shock module
20
Risk module
21
Selected Results
  • Descriptive Statistics
  • Income Composition
  • Demographic Characteristics
  • Pattern of shocks
  • Data Robustness Test and Aggregate Vulnerability
  • Risk Perception and Risk Attitude
  • Key messages of 1st project phase

22
Indicators of Data Quality
  • Total sample in 2007 4,381
  • Missing cases 2007 ( 2 )
  • Total sample in 2008 4,284
  • Attrition 2
  • Replacement rates 20
  • Standard Errors are acceptable
  • Initial assumptions confirmed
  • New issues emerged

23
Age structure in Buriram versus Thailand
Sources DFG FOR756 Household survey 2007,
Thailand Institute for Population and Social
Research (2003)
24
Type of shocks by province for Thailand and
Vietnam
Source DFG FOR756 Household survey 2007.
25
Educational Attainment of Household members
Source DFG FOR756 Household survey 2007.
26
Results Income composition
Share of income component in , crop year 2006/07
27
Results Income composition
Share of income component in , crop year 2007/08
28
Income composition
29
Income composition
30
Change in crop prices - Vietnam
31
Change in crop prices - Thailand
32
Robustness Tests Stochastic Dominance Relations
  • Are conclusions on vulnerability driven by the
    choice of the measure ?
  • Compare cumulative distributions of income and
    consumption at provincial level and search for
    stochastic dominance relations between these
    distributions
  • Search for stochastic dominance relations (FSD,
    SSD,TSD) below the threshold income
  • Comparisons should allow for robust conclusions
    on welfare with non censured data
  • Provide a benchmark for various vulnerability
    measures

33
Distribution of per capita income in Thai
provinces
Source DFG FOR756 Household survey 2007.
34
Dominance relations for provincial distributions
and critical values Per capita consumption
Source DFG FOR756 Household surveys 2007.
35
Consumption distributions cross-country
comparison
Source DFG FOR756 Household surveys 2007, 2008.
Source DFG FOR756 Household survey 2007.
36
Per capita income distributions 2007 - 2008
Source DFG FOR756 Household surveys 2007, 2008.
37
Per capita income distributions 2007 2008
(lower part)
Source DFG FOR756 Household surveys 2007, 2008.
38
Summary Observations
  • No meaningful FSD relations -gt all vulnerability
    comparisons depend on the measure used and the
    poverty line
  • For consumption data SDD occurs -gt consistent
    with FGT vulnerability indices and the
    Calvo/Dercon measure
  • Consumption Vulnerability ordering by province in
    Thailand NP lt BR lt UR
  • In Vietnam Ha Tinh lt Hue lt Dak Lak
  • Income and consumption vulnerability not
    consistent
  • No difference between per capita per adult
    equivalent

39
Key Messages of 1st phase
  • Vulnerability data base
  • reasonably valid data can be collected
  • "but it takes time and it is expensive"
  • spatial dimension of vulnerability is feasible
  • Vulnerability concept
  • determinants of vulnerability vary
  • downside risk matters
  • Alternative benchmarks seem to be useful
  • Household heterogeneity including gender matters

40
Key Messages of 1st project phase
  • Agriculture
  • Part-time farming dominates (TH)
  • Shocks and risk perception affects
    diversification within agriculture and outside
    (THVN)
  • Majority of shocks are in agriculture (THVN)
  • Shocks can contribute to natural resource
    degradation (VN)
  • Typical patterns of coping observed for shocks in
    farm households (TH)

41
Key Messages of 1st project phase
  • Financial Markets
  • Consumption insurance of rural HH only partially
    possible
  • Village level microfinance institutions improve
    access to credit but are not an effective "shock
    absorber" (TH)
  • Economic Geography
  • non-agricultural wage employment has potential to
    reduce vulnerability (TH)
  • income diversification has positive effect on HH
    welfare (VN)
  • Employment in rural-based industries more stable
    in the large firms (TH)
  • Higher participation rates in non-agricultural
    wage labour among peri-urban HH

42
Project plan for next phase
  • Add another two panel waves in 2010 and 2011
  • Refine risk module and risk experiment
  • Add investment module
  • Add village business survey
  • Interview domestic migrants once (in 2010)

43
Research Topics 2nd Project Phase
Risk preferences and perceptions
Agriculture
Thailand
- Economic crisis and de-velopment of agriculture
in peripheral areas - Economic crisis and
natural environment - Risk perceptions and
decision making
Vulnerability and HH dynamics
Management of Vulnerability Database
Vietnam
Financial institutions and employment
Non-agricultural income
44
Thank you very much for your attention!
45
Risk attitudes
  • Risk is a major variable in vulnerability
    assessments
  • Combining risk experiments with risk questions in
    surveys
  • Simple self-assessment is validated by survey
    evidence
  • Risk experiment may need to be adjusted to
    cultural context and level of education
  • Gender does not seem to matter
  • More testing is needed !

46
Results Risk attitudes, survey based
47
Results Risk attitudes, experiment
48
Results Coefficient of partial risk aversion
(experiment)
About PowerShow.com