Hossain m 20150709_1730_upmc_jussieu_-_room_101_(building_14-24)

Our Common Future Under Climate

Change Conference

UNESCO & UPMC, Paris

7-10 July 2015

Response to Climatic Hazard: Does Vulnerability Measurement of the

Coastal People Matter?

Associate Professor M Hossain and Professor E Selvanathan Griffith Business School, Griffith University

Brisbane Australia

[email protected]

7-10 July 2015

Paris, France

Outline

• Introduction

• Vulnerability Measurement from Climatic Hazard

Matters: A New Approach for Measuring Vulnerability

• Climatic Hazard and The Study Region

• Results

• Conclusions

Introduction

The major aim of this study is to measure the vulnerability due to climatic hazard of a frontline region in South Asia, for making adaptation policies. More specifically, the study focuses on the following objective:

‘to explore how to make the vulnerability measure proposed in

the present study robust out of a primary survey in the Bay of Bengal delta. Measurement of vulnerability is essential for making adaptation policies of the coastal regions, particularly, developing and small island nations’.

In other words, a micro-analysis with a macro implications.

Response to Hazards in the Bay of Bengal Delta

Major Cyclonic Storms: 1960-2014

Table 1: Frequency of Cyclone over last half a century

Table 1: Major cyclonic storms some with hurricane intensity in the Bay of Bengal Delta, 1960–2010

2011 1 Myanmar (Nargis) 150-200 6-8 2014 1 Orissa (Paleen) 200-220 4-5

Total 38 cyclones hit over the last 54 years --------------------------------------------------------------------------------------------- Source: BBS (2008 & 2010), DS (2014)

Study Region

Vulnerability Measurement: The Approach

Vulnerability as Expected Poverty (VEP)

We use field data to estimate household’s vulnerability for the two villages by calculating the chance (or probability) that each household consumes below a certain amount after hit by a hazard. To quantify a household’s vulnerability we first need to calculate its expected consumption and the variance of its consumption. To predict the mean consumption of household ‘h’ at a particular time period and its variance of consumption following Jha, Dang and Tashrifov (2009), we estimate the following heteroscedaticity regressions:

ln ch = Xh +h (1)

with

var (h) = = Xh + eh (2)

where ch is household food consumption expenditure, Xh represents a bundle of observed household characteristics, is a vector of parameters and h is a zero mean disturbance term that captures idiosyncratic shocks that contribute to different consumption levels for households, which are otherwise observationally equivalent.

We would estimate and using a three-step feasible generalized least squares (FGLS) procedure suggested by Amemiya (1977), which is described in detail in Chaudhuri (2000).

Using the estimates of and, for each household h, we find the expected log consumption.

Vulnerability Measurement (cont.)

E[ln ch | Xh] = Xh (3)

and the variance of log consumption

V[ln ch | Xh] = = Xh (4)

By assuming that consumption is log-normally distributed (that is ln ch is normally distributed) we then estimate the measure of Vulnerability as Expected Poverty for household h (VEPh) by (see Chaudhuri et al (2002),

(5)

where is the cumulative log-normal distribution function.

Chaudhuri, Jalan and Suryahadi (2002) argue that a household should be classified as being vulnerable if its vulnerability computed as above (equation 5) exceeds a certain threshold. In this study, we follow their recommendations and use a vulnerability threshold of 0.5. That is, a household i would be included among the vulnerable if VEPh > 0.5.

VEPh = (ln ch < ln z) =

Methodology: The Village Census

Table 2: Basic Information of Two Villages under Study

Lorenz Curves of Two Villages, 2011

Lorenz Curves, 2012

Methodology: Vulnerability as Expected Poverty

More Specifically, what exactly we are after from this model of

vulnerability measurement?

• To apply the model in two villages hit by a climatic hazard;

• To find pathways for speedy recovery in post emergency period of a

hazard;

• To recommend socio-economic-legal protection measures for

sustainable recovery and resistance from climatic hazards in the

future.

Methodology (cont.)

Results

Table 4: Percentage of households (HH) vulnerable, 2011–12 Note: BD is Bangladesh and WB is West Bengal (India). Source: Present Study

• Source: Present Study

Results (cont.)

Table 5: Vulnerability in two villages with two segments of households (6%)

Source: Present Study

Summary 1. The VEP results show that the households of the

Indian village recovered within two years of the cyclone (2009–11), whereas, the Bangladesh village took three years for recovery (2009–12).

2. After the recovery, among the landowners of the Indian village, the poverty increased again due to economic cycle, when landless households’ condition remained unchanged.

3. In the Bangladesh village, the households took one extra year to recover, the landowners come out of poverty fully and the landless improved their conditions remarkably in recovery period.

4. In the span of one year the poverty declined from 51% of the landless people to only 6% in BD village. Certainly, the improvement has been exemplary. One of the main reasons for this achievement was the increased employment opportunity created in the village through re-establishment of the shrimp farming in the region. The cyclone Aila destroyed this industry in 2009.

Conclusions

Now, how to integrate this investigation into the adaptation policy and sustainable development at macro level?

1. Policy makers both in developing and advanced nations are looking for micro level investigation on vulnerability to make adaptation policy for region (s) hit by climatic hazards.

1. Sustainable development goals (post-MDGs) cannot be achieved by leaving out millions of coastal population from the equation with constant threat from climatic hazards with high intensity and frequency due to global warming?

1. Our study focuses both these issues with micro level investigation of victims of climatic hazards for relevant and appropriate adaptation policies for effective response by the macro players.

References * Amemiya, T. (1977) ‘The maximum likelihood estimator and the non-linear three stage least squares estimator in the general nonlinear three stage least squares estimator in the general nonlinear simultaneous equation mode’, Econometrica 45: 955–68.

Barnett, J., Lambert, S. & Fry, I. (2008) ‘The hazards of indicators: Insights from the environmental vulnerability index (EVI)’, Annals of the Association of American Geographers 98: 1, 102–19.

BBS (2010) Annual Statistical Dictionary, Bangladesh Bureau of Statistics, Dhaka

BBS (2008) Annual Statistical Dictionary, Bangladesh Bureau of Statistics, Dhaka

Briguglio, L. (1995) ‘Small island states and their vulnerable economies’, World Development 23: 1615–32.

Chaudhuri, S. (2000) ‘Empirical methods for assessing household vulnerability to poverty’, mimeo, New York: Department of Economics, Columbia University.

Chaudhuri, S. (2003) ‘Assessing vulnerability to poverty: concepts, empirical methods and illustrative examples’, mimeo, New York: Columbia University.

* Chaudhuri, S., Jalan, J. & Suryahadi, A. (2002) ‘Assessing household vulnerability to poverty from cross-sectional data: A methodology and estimates for Indonesia’, Economics Department Discussion Papers, New York: Columbia University,

Appendix: Regression Results

Regression results for VEP measure, 2011–12

Dependant 2011 2012 variable Shukhdara (BD) Jharkhali (WB) Shukhdara (BD) Jharkhali (WB)

Age (Head) 0.0047 –0.0050** (0.361) (0.083) Family size 0.0384 0.0579* (0.397) (0.010) Income 0.0001* 0.0007** (0.067) (0.000) Child <5 –0.0021 –0.0477 (0.980) (0.207)

HH No. 82 73 F 1.99 (0.00) 37.05 R Square 0.3062 0.6792

Consumption – – 0.2547 0.9524* 2011 (t-1) (2.28) (0.04)

HH No. 82 73 F 1.61 62.29 R Square 0.0198 0.4604

Notes: BD is Bangladesh and WB is West Bengal (India) * and ** indicate the coefficient is significant at the 10% and 5% level, respectively. p-

values are in parentheses.