perceived vulnerability to downside risk felix povel new directions in welfare, oxford, june 29,...

Perceived Vulnerability to Downside Risk

Felix PovelNew Directions in Welfare, Oxford, June 29, 2009

2

Outline

1. Theoretical background

2. Measure of vulnerability to downside risk

3. Empirical application

4. Conclusion

3

Risk

• is a negative future event whose occurrence has a certain probability

• is relevant for current wellbeing, behavior and economic development (e.g.Dercon, 2008, and Elbers and Gunning, 2003)

Vulnerability

• builds on risk and is therefore a forward looking concept measuring something ex-ante (e.g. Cafiero and Vakis, 2006)

• is concerned with „bad“ future outcomes (e.g. Calvo and Dercon, 2005)

Theoretical background

4

Measures of vulnerability using poverty line as benchmark

• Vulnerability as expected poverty (Chaudhuri et al., 2002)• Vulnerability to poverty (Calvo and Dercon, 2005)

Theoretical background

z

a

b

c

d

0

household 1

household 2

household 3

household 4

5

Theoretical background

Critique to poverty line as benchmark of choice

• Consumption change “a” is not captured

• No differentiation between consumption changes “b”, “c”, and “d”

• Consumption change “c” impacts on vulnerability

Solution: Choice of current level of well-being as benchmark

b

a

c

d

0

c

0 0 0

0

z

z z z

z

6

Measure of vulnerability to downside risk

z

a

b

c

d

0

household 1

household 2

household 3

household 4

Measure of vulnerability to downside risk uses current level of well-being as benchmark

with and

7

Measure of vulnerability to downside risk

The number of states of the world household h faces (Nh) is given by

8

Measure of vulnerability to downside risk

Example: Household h faces risks B and A, as well as certainty D

The number of states of the world household h faces (Nh) is given by

9

Measure of vulnerability to downside risk

Example: Household h faces risks B and A, as well as certainty D

The number of states of the world household h faces (Nh) is given by

Consequently, household h faces states of the world.

B A DB DDA D

10

Measure of vulnerability to downside risk

The index of deprivation (dhi) assigned to household h in state of the world i is given by

11

Measure of vulnerability to downside risk

The index of deprivation (dhi) assigned to household h in state of the world i is given by

Example: Household h faces risks B and A, as well as certainty D, i.e. four possible states of the world; severities of impact of adverse events equal

12

Measure of vulnerability to downside risk

The index of deprivation (dhi) assigned to household h in state of the world i is given by

Example: Household h faces risks B and A, as well as certainty D, i.e. four possible states of the world; severities of impact of adverse events equal

Four state of the world-specific indices of deprivation equal

13

Measure of vulnerability to downside risk

The probability (phi) that household h experiences state of the world i is given by

14

Measure of vulnerability to downside risk

The probability (phi) that household h experiences state of the world i is given by

Example: Household h faces risks B and A, as well as certainty D, i.e. four possible states of the world; each risk occurs with a probability of 80%

15

Measure of vulnerability to downside risk

The probability (phi) that household h experiences state of the world i is given by

Example: Household h faces risks B and A, as well as certainty D, i.e. four possible states of the world; each risk occurs with a probability of 80%

B A D

D

A DB D

16

Measure of vulnerability to downside risk

Vulnerability to downside risk of household h (Vh) equals

State of the world 1:

State of the world 2:

State of the world 3:

State of the world 4:

17

Critique to current level of well-being as benchmark of choice

Measure of vulnerability to downside risk

• No differentiation between consumption changes “a”, “b”, and “d”; that is, no identification of households threatened by poverty

However, the measure is still useful for poverty reduction because

1. by concentrating on decreasing degrees of vulnerability in all parts of society one allows all households to redirect resources from risk management strategies to growth enhancing activities.

2. it provides unambiguous policy advice regarding the design of pro-poor policies.

a

0 z

b d

0 0z z

18

Empirical application

„Ingredients“:

• Data from questionnaire‘s section about subjective risk perception

• 11 risks (i.e. households face between 1 and 2048 states of the world): illness, person leaves household, person joins household, flooding, drought, unusually heavy rainfall, storm, crop pests, livestock disease, strong increase in interest rate on loans, strong increase of prices for input

• Perceived severity of impact on income of risks (0, 0.03, 0.06, 0.09)

• Perceived probability of occurrence of risks (0%, 20%, 40%, 60%, 80%, 100%)

• 3875 households (2070 Thai and 1805 Vietnamese)

• α = 2

19

Tobit regression resultsEmpirical application

dependent variable: vulnerability to downside risk (α=2)independent variables Thailand Vietnam

hhsize 0.0013** (-0.0006) 0.0005 (-0.0011)offempl 0.0042 (-0.0036) -0.0206*** (-0.0077)selfempl -0.0027 (-0.0063) -0.0221* (-0.0123)agrempl 0.0253*** (-0.0038) 0.0325*** (-0.0074)

mage -0.0003*** (-0.0001) 0.0001 (-0.0002)age -0.0010** (-0.0005) 0.0009 (-0.0008)

age2 0.0000** (0.0000) 0.0000 (0.0000)female -0.0013 (-0.002) 0.0001 (-0.0046)minor -0.0001 (-0.0041) 0.0084* (-0.005)enroll 0.0014 (-0.001) 0.002 (-0.0012)

enroll2 -0.0001 (-0.0001) -0.0001 (-0.0001)

20

Tobit regression results – continuedEmpirical application

dependent variable: vulnerability to downside risk (α=2)independent variables Thailand Vietnam

llaginc -0.0009 (-0.0008) -0.0035** (-0.0014)ai -0.0014 (-0.001) 0.0006 (-0.0021)

totland 0.0012*** (-0.0004) -0.0034* (-0.0019)totland2 -0.0000** (0.0000) 0.0001 (-0.0001)

shock_exp06 0.0044* (-0.0023) 0.0035 (-0.0027)shock_exp07 0.0093*** (-0.0015) 0.0112*** (-0.0021)shock_exp08 0.0156*** (-0.0024) 0.0187*** (-0.003)

buriram 0.0219*** (-0.0029)ubon 0.0047* (-0.0028)

ha_tinh 0.0038 (-0.0048)hue 0.0134*** (-0.0042)

Constant 0.0345** (-0.0164) 0.0279 (-0.0237)Observations 2008 1701

21

• Choice of current level of well-being as relevant benchmark resolves some of the problems related to the poverty line

• Data on risk perception adds value to vulnerability research

• Vulnerability to downside risk in Vietnam higher than in Thailand

• Determinants of vulnerability to downside risk differ between the two countries

Conclusion

22

Bibliography

Cafiero, C. and R. Vakis (2006), Risk and vulnerability considerations in poverty analysis: Recent advances and future directions. Social Protection Discussion Paper 0610, The World Bank, Washington D.C.

Calvo, C. and S. Dercon (2007), Vulnerability to poverty. CSAE Working Paper Series, 262, Oxford University, Oxford.

Calvo, C. and S. Dercon (2005), Measuring Individual Vulnerability. Department of Economics Discussion Paper Series, 229, Oxford University, Oxford.

Chaudhuri, S., J. Jalan, and A. Suryahadi (2002), Assessing Household Vulnerability to Poverty from Cross-sectional Data: A Methodology and Estimates from Indonesia. Columbia University, Discussion Paper 0102-52.

Dercon S. (2008). Fate and Fear: Risks and its consequences in Africa. Journal of African Economies, 17(2): 97-127.

Elbers, C., and J. Gunning (2003), Growth and Risk: Methodology and Microevidence. Tinbergen Institute Discussion Papers, 03-068/2.

Ligon E., and S. Schechter (2003), Measuring vulnerability, Economic Journal, 113: 95-102.

Townsend, R. (1994), Risk and Insurance in Village India. Econometrica, Vol. 62, No. 3, pp. 539-591.

World Bank (2001), World Development Report 2000/2001 – Attacking Poverty. New York, Oxford University Press.

23

Combination of both reference points

0 z

a

24

Combination of both reference points

0 z

b

25

Combination of both reference points

0 z

b

26

Combination of both reference points

d

0 z

27

Combination of both reference points

Vulnerability to downside

risk

Vulnerability to (more)

poverty