RESPONSES TO POVERTY AND RISKS IN VIETNAM: HOW EFFECTIVELY DOES THE VIETNAMESE PUBLIC SAFETY NET TARGET VULNERABLE POPULATIONS?

1
RESPONSES TO POVERTY AND RISKS IN VIETNAMHOW
EFFECTIVELY DOES THE VIETNAMESE PUBLIC SAFETY
NET TARGET VULNERABLE POPULATIONS?

• NGUYEN Ngoc Quynh
• Ninth Annual Global Development Conference 2008
• Brisbane, 29-31 January 2008
• Research and Development Department
• Bank for Foreign Trade of Vietnam

2
Outline

• Background
• Motivation
• Research questions and contribution
• Research methodology
• Research findings
• Conclusion
• Policy implications

3
Background Vietnam stylized facts

• Strong average economic growth since the
  mid-1990s, 7.3 per year (World Bank, 2007)
• Considerable variations of poverty rates and the
  speed of poverty reduction across regions
• Increasing disparity between urban and rural
  areas
• Frequent day-to-day fluctuations, uncertainties
  and adverse shocks affecting households income
• Causing households to fall into poverty
• Widening the gap between the rich and poor
  populations

4
Vietnam poverty indexes 1992-2002
5
Background Risk insurance in Vietnam

• Pre-existing risk coping arrangements
• Few, scattered studies on how well households
  income shocks is insured against
• Little information on which risk-coping
  mechanisms are adopted by households
• Few assessments on who are more vulnerable in
  facing with risks
• Government's intervention
• In principle, public safety net programs aim to
  help social beneficiaries, who are in vulnerable
  groups, live above minimum standards
• Poorly funded public transfer programs, for
  instance
• Only 20 of social beneficiaries actually receive
  social allowance (MOLISA 1999)
• Few researches, subjective assessments on the
  performance of existing public transfers in
  insuring the poor in Vietnam
• Less assessments on how much safety net programs
  contribute to poverty reduction

6
Motivation

• Necessary for Vietnam to cope with factors that
  affect welfare to
• Lift out of poverty those who currently poor
• Prevent those who are currently not poor from
  falling into poverty
• Necessary to have research assessing the
  performance of current safety net and providing
  appropriate guidance for effective safety net
  policies
• Desirability of safety net policies depends on
  how well pre-existing risk-arrangements work ?
  essential to have
• Appropriate evaluation of pre-existing
  risk-sharing arrangements
• Identification of vulnerable population and
  coping mechanisms
• Suitable investigation whether governments
  safety net programs reach the most vulnerable
  population

7
Research questions and contribution

• Research questions
• What is the level of risk sharing among
  households in Vietnam?
• Who are vulnerable population?
• What coping mechanisms are adopted by households?
• How well do existing social welfare programs
  target and insure poor people from shocks in
  Vietnam?
• Contribution
• Provide necessary information on households
  exposure to risk in Vietnam
• Provide background for effective policy outcome

8
Research methodology Theoretical model

• Perfect insurance model
• (Townsend 1994, Jalan and Ravallion 1999)

9
Research methodology Empirical approaches
10
Data and Variables

• Data
• Household panel data from the Vietnam Living
  Standard Surveys (VLSS) of 1992-93 and 1997-98.
  4300 households/ 150 communes/ 8 regions
• Idiosyncratic shocks to household
• Unemployment household member became unemployed
• Illness number of days household members were in
  sickness
• Natural disaster household faced with natural
  disaster
• Coping mechanisms
• Borrowing household reported positive net debt
• Private transfer household reported positive net
  transfer
• Precautionary savings household sold asset or
  withdrawn savings
• Returns to human capital household member got
  2nd wage earning job
• Public safety net household received allowance
  from safety net programs

11
Research findings (1)

• Risk sharing among households
• Significant coefficient of change in income on
  change in consumption at regional level
• Good level of idiosyncratic risk sharing taking
  place within smaller communities (i.e. communes)
• No risk-sharing at regional or national level,
  justifying a shortage of effective national risk
  pooling mechanisms
• Uninsured aggregate risks at commune level
• No risk-sharing across communes or across regions
• Most vulnerable population
• female headed households
• ethnic minority households
• households who do not have use-right of
  agriculture land
• households who reside in the Red River Delta
  region

12
Research findings (2)

• Coping mechanisms
• Main coping device self-insurance strategies
  (asset sale, savings withdrawal, private
  transfers)
• Households with unemployment shock depend only on
  self-insurance mechanisms to insure their risks
• Most vulnerable population heavily rely on
  self-insurance mechanisms
• Second coping device borrowing. But still, poor
  households have less accessibility to credit due
  to collateral issue
• Targeting of social safety net programs
• Safety net programs work as a coping mechanism
  only when households face with natural disaster
• While 67 of households face with natural
  calamities, only 3 of them received supports
  from safety net programs
• Most vulnerable households are less likely to
  receive public transfer from safety net programs
  in comparison with other households

13
Conclusion

• Negligible contribution of governments safety
  net program to the outcome of poverty reduction
  in Vietnam
• Shortage of financial resources to cover for
  risks
• Failure of safety net program to target the most
  vulnerable households ? scarce resources have
  been transferred to wrong direction
• Lack of national norms for identifying the poor
  consistently over the country

14
Policy implications

• Introduce consistent mechanisms to identify
  beneficiaries
• Closely monitor administrative procedures on
  local discretion in the implementation of social
  welfare programs
• More government and local authorities
  transparency in the implementation of public
  transfers
• Establishment of an unemployment insurance system
• Better incentives for fiscal redistribution at
  the local and regional levels
• Establishment of better voluntary social
  insurance systems and promotion of individuals
  participation to make safety net accessible to
  everyone

15
Risk-sharing estimation results (1)
Variables OLS estimates OLS estimates OLS estimates IV estimates IV estimates IV estimates IV estimates
Variables With region dummies With commune dummies With region dummies With region dummies With commune dummies
?(income per capita) 0.2 (0.02) 0.15 (0.02) 1.09 (0.16) 0.08 (0.22)
AGE 10.55 (11.05) -3.85 (10.19) -65.29 (23.72) 0.19 (15.79)
AGE2 -0.1 (0.11) 0.01 (0.1) 0.71 (0.24) -0.03 (0.17)
Sex -192.6 (67.76) -75.19 (59.88) -451.65 (121.88) -49.25 (104.72)
?household_size -208.29 (20.65) -174.94 (18.2) -10.5 (48.72) -187.72 (48.13)
?hour_work 0.04 (0.008) 0.002 (0.008) -0.11 (0.03) 0.01 (0.02)
No. of observations 4300 4300 4300 4300
R2 0.2253 0.3697 - 0.3579
16
Risk-sharing estimation results (2)
Variables OLS estimates OLS estimates OLS estimates IV estimates IV estimates IV estimates
Variables Within region Within commune Within region Within commune
?(income per capita) 0.19 (0.02) 0.15 (0.02) 1.17 (0.19) -0.07 (0.27)
?(average income per capita) 0.36 (0.06) 0.37 (0.037) 2.22 (1.51) 1.22 (0.24)
AGE 12.43 (11.01) 4.24 (10.64) -108.88 (34.23) -17.82 (19.83)
AGE2 -0.12 (0.11) -0.057 (0.11) 1.07 (0.32) 0.1 (0.21)
Sex -196.356 (64.95) -152.05 (63.79) -421.55 (140.79) -7.99 (120.95)
?household_size -213 (20.46) -202.25 (19.5) 57.9 (61.85) -181.56 (52.19)
?hour_work 0.04 (0.007) 0.03 (0.007) -0.18 (0.04) -0.008 (0.03)
No. of observations 4300 4300 4300 4300
R2 0.2058 0.2535 - -
17
Identification of vulnerable groups
Variables OLS estimates OLS estimates OLS estimates IV estimates IV estimates IV estimates
Variables Coeff. Robust Std. Err. Coeff. Robust Std. Err.
Wealth groups
Poorest decile (ref. group) 0.28 0.06 0.5 0.3
Poor middle 10-60 -0.15 0.09 0.37 0.6
Rich 60 up -0.29 0.05 -0.38 0.34
HH in urban area 0.05 0.04 0.4 0.26
Farm HH -0.06 0.03 -0.23 0.25
HH with children less than 6 years old 0.03 0.04 -0.03 0.2
HH head is male -0.05 0.04 -0.34 0.17
HH belongs to majority ethnic groups -0.049 0.029 -0.22 0.14
HH that owns agriculture land -0.02 0.03 0.26 0.43
No. of observations 4300 4300
R2 0.3263 0.1119
18
Choice of coping strategies
Have support from safety net programs Have support from safety net programs Have positive net debt Have positive net debt Have positive net private transfer Have positive net private transfer Sold assets or withdrawn savings Sold assets or withdrawn savings HH members get 2nd job HH members get 2nd job
Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err.
HH member was unemployed 0.033 (0.0009) 0.12 0.054 (0.026) 0.064 0.25 (0.06) 0.06 0.12 (0.027) 0.069 -0.005 (-0.001) 0.07
HH member in was sickness 0.00004 (0.0000011) 0.004 0.011 (0.005) 0.002 0.0088 (0.0022) 0.002 -0.005 (-0.001) 0.002 0.002 (0.0004) 0.002
HH faced natural disaster 0.49 (0.0134) 0.11 0.19 (0.091) 0.04 -0.25 (-0.063) 0.046 -0.079 (-0.017) 0.046 0.24 (0.049) 0.04
Log likelihood -10013.346 -10013.346
No. of obs. 4300 4300
19
Accessibility of most vulnerable households to
safety net programs
Variables Interaction between household has support from public safety and household characteristic Interaction between household has support from public safety and household characteristic Interaction between household has support from public safety and household characteristic
Variables Coeff. Std. Err. Marginal Effect
HH head is female
HH member was unemployed -0.14 0.23 -0.07
HH member in was sickness 0.0017 0.008 0.0001
HH faced natural disaster 0.4 0.17 0.022
HH belongs to ethnic minority groups
HH member was unemployed -0.16 0.23 -0.0028
HH member in was sickness -0.015 0.008 -0.0003
HH faced natural disaster - - -
HH do not have the use-right of land
HH member was unemployed 0.11 0.2 0.0016
HH member in was sickness -0.002 0.008 -0.000029
HH faced natural disaster -0.099 0.15 -0.0013
HH belongs to poorest wealth groups
HH member was unemployed 0.036 0.12 0.0019
HH member in was sickness 0.00013 0.004 6.90e-06
HH faced natural disaster 0.49 0.11 0.022
HH resides in Red River Delta area
HH member was unemployed - - -
HH member in was sickness 0.006 0.014 3.14e-06
HH faced natural disaster -0.269 0.31 -0.00017