布里斯托作業(yè)：淺析災(zāi)難對孟加拉國沿海地區(qū)收入的影響
08-22, 2014
由于氣候變化,孟加拉國成為世界上最脆弱的國家之一。這種現(xiàn)象體現(xiàn)在各種類型的自然災(zāi)害上。孟加拉國的臺風(fēng)和潮汐是最頻繁發(fā)生的災(zāi)害類型,經(jīng)常出現(xiàn)在沿海地區(qū)。幾乎每年的沿海地區(qū)都會受到颶風(fēng)的嚴(yán)重影響，最終給孟加拉國人民造成了極大的損失。在沿海地帶，這樣的氣候影響整個社會經(jīng)濟結(jié)構(gòu)。沿海地帶受影響最大的是居民收入的經(jīng)濟指標(biāo)。沿海地區(qū)這些災(zāi)害的頻繁發(fā)生造成了收入不平等問題。大多數(shù)生活在沿海地區(qū)的人民，生活收入降低,這些人通常是非常容易受到氣候災(zāi)害的沖擊的。因為災(zāi)害嚴(yán)重影響窮人的收入來源,研究分析，災(zāi)難會給家庭收入帶來顯著變化。本研究主要探討災(zāi)難對收入不平等的影響。本文使用電子表格計算收入的不平等,而且比較了沿海災(zāi)難過后收入的前后變化。
Disaster And Income Inequality: Coastal Areas Of Bangladesh
Bangladesh becomes one of the most vulnerable countries in the world due to climate change. This phenomenon manifests itself through various types of natural disasters. For Bangladesh cyclone and tidal surge are the most frequent disasters, especially to the coastal regions. Almost every year the coastal regions are severely affected by cyclones causing a lot of damages. Such climatic shocks affect the entire socioeconomic structure in the coastal belt. The most affected economic indicator in coastal belt is income of the people. Frequent occurrence of these disasters creates income inequality problem in coastal regions. Most of the people living in coastal areas possess lower standard of living and these people are usually very much vulnerable to any climatic shock that comes about in the form of disasters. Since disasters affect the poor people badly, the study analyzes the change in income of the household due to disaster. This study investigates the influence of disaster on income inequality. A spreadsheet computation of income inequality is used in this paper to measure the income inequality, and a comparison between past and present income levels of people is figured out once a disaster hits the coastal belt.
Key Words: Disaster, Income Inequality, Gini Index, Hoover Index, Labor Income.
Introduction
Since independence in 1971, Bangladesh has started her war against numerous socio-economic, political and socio-political problems with a view to escalating her status from bottom level LDC to top level LDC. Among a long list of existing problems like rapid population growth, poverty, illiteracy, gender disparity, low level of economic growth etc. income inequality takes the crucial role that actually undermines the sustainable development process of Bangladesh (Khan and Hossain, 1989). A decade ago the major causes behind income inequality were mostly socioeconomic ones, particularly in the coastal belt of Bangladesh. However, at present the most powerful determinant creating income inequality in coastal belt does not belong to socioeconomic aspects, rather to climatic factors in the form of natural disasters (Carter and Maluccio, 2003).
Natural disasters are most frequent events in Bangladesh. Bangladesh often makes top news all over the world due to natural disaster, which usually means a huge death toll and massive destruction for one of the poorest nations in the world (Brammer, 1996). Recently, on November 15, 2007, thousands have died and millions were rendered homeless by a super-cyclone named SIDR. It devastated southern and central Bangladesh. According to official sources, some 3,313 people were killed and 28,128 injured by the deadliest cyclone in the history of post-independence Bangladesh. Due to SIDR the estimated damage in terms of US dollar was 2300 million (EM-DAT, 2010). However, the unofficial sources said that the total death will be between 5,000 and 10,000. Cyclone SIDR swept across 16 coastal districts with wind speeds of 240 km/h (150 mph). The winds whipped up waves three to five meters high. The cyclone flattened homes, uprooted trees, electricity, and telephone poles, washed away roads and bridges, destroyed crops, and killed livestock. The death toll would have been far worse if limited emergency plans had not been in place (Bilham and England, 2001). An early warning system enabled the evacuation of an estimated 3.2 million people to higher ground and safety. In contrast, during 1991, Cyclone Gorky claimed at least 138,000 lives and in 1970, Cyclone Bhola killed about half a million people in Bangladesh (Nizamuddin, 2001).
Both the frequency and the scale of natural disasters have increased in recent decades, especially the frequency of disasters does have adverse effect on developing countries like Bangladesh (Barkun, 1977). Despite increasing frequency, natural disasters have often been ignored by economists because of the difficulty regarding making generalized conclusions about their impacts. Disaster is mainly of two types and they are the natural disaster and the manmade disaster (Benson and Clay, 2004). It is quite difficult to calculate and make a conclusion on all disaster. In this paper, only a cyclone has been taken into account as natural disaster and it is the recent devastating super cyclone SIDR occurred on 15th November 2007.
As we have already mentioned that income inequality is one of the major problems in Bangladesh, which reveals the degree of disparity between high and low incomes. The degree of income inequality is often regarded as an important indication of the fairness of the society we live in. A high level of income inequality may be harmful to the level of social integration among the members. Inequality in income and other welfare indicators reflect underlying inequality in asset ownership pattern (Ferreira et al., 2005). The rise in income inequality in Bangladesh has attributed to the widening disparity in landholdings and education levels along with the rising income inequality between the urban and rural sectors (Deininger and Squire, 1996, Sen and Ali, 2005). Now a days the natural disasters are highly responsible for creating income inequality in Bangladesh. Therefore, this paper focuses on the impact of disaster on income change in the study areas between pre and post SIDR period.
1. Objectives of the Study
The objectives of this study are to – assess whether disaster affect the income or not; and what extent disaster influences the factors that are directly related with income distribution.
2. Literature Review
Burton (1978) attempts to measure the impact of disaster and finds that in a disaster situation, those living near the poverty line can easily slip below it. The landless and near landless may be forced to sell their limited assets for survival needs. Those who depend upon wage labor for subsistence are forced to compete with those entering the labor market. The labor market, in turn, becomes further depressed by the loss of harvests and alternative sources of employment.
According to Lindell and Prater (2003), empirical work at a macro-economic level concludes that natural disasters have no long term adverse economic effects, even at a community level. However, household surveys suggest that households living at or near the subsistence level are unable to recuperate from natural disasters, and that a bad disaster can have intergenerational effects on household health and productivity.
Again, Hubner (2008) attempts to identify communities respond to disasters by the following strategies: reducing consumption, drawing from savings, selling productive assets, migration, and borrowing money. Lower income populations do not have savings to utilize or the resources to migrate. Their remaining options lead to reductions in both current and future consumption. Therefore, there is a disconnection between large-scale surveys reporting aggregate recovery and micro-level research implying long-term reductions in consumption. If an economy recovers evenly across all income levels, then we should see no long term changes in income distribution or consumption. However, if assets are flowing from lower to higher income deciles, then we should see increased income inequality and lower consumption marked by increased volatility at lower income deciles.
Tol and Leek (1998) observe and argue that damages from disasters are hard to estimate accurately for several reasons. Especially in more developed nations, total estimated costs of disasters may not include preventative costs aimed at mitigation. In developing nations, damages to poor households will not be accounted for in formal book-keeping, insurance receipts or other market indicators.
Bibbee et. al. (2000) assess that the incomes and businesses of lower income deciles are more vulnerable to disruptions in their supply and demand chains, and have a harder time receiving formal credit to begin renovating their businesses.
Britton (1986) mentions that disaster is also viewed as a mental construct imposed upon experience. This is because to understand disaster knowing the number of deaths, the value of property destroyed or the decrease in per capita income is not sufficient. The symbolic component requires knowledge of the sense of vulnerability, the adequacy of available explanation and the society’s imagery of death and destruction.
Knowles (2001) points out that there is a negative correlation between inequality and economic growth when consistently measured expenditure data, which take redistribution of income into account, are used. This paper also showed that an unequal pre-tax distribution of income leads to pressure for distorted transfers from the rich to the poor, which will in turn reduce the rate of growth.
Moreover, Fielding and Torres (2006) show how to estimate simultaneously the relationship between income inequality and other indicators of social and development by using cross country data. The overall picture which is shown by this paper is that there is a correlation between reductions in inequality and improvements in development indicators such as per capita income, literacy and life expectancy.
Different views have come from the above discussion and most of the researchers have tried to focus on the disaster management and preventive measure. Again, in case of income inequality, the researcher only mentions its importance as a tool to measure economic development. However, they haven’t focused the relationship between the disaster and income inequality. As income inequality is a serious problem and disaster is also an annual phenomenon, it is very much essential to find out the interdependency between these two and we have tried to do it through this paper.
3. Materials and Methods
The study is an out and out primary data-based work. The data were collected through questionnaire survey and observation method. Collection of data was conducted in different units from the study area and after compilation of data; various outcomes of household factors are shown accordingly.
3.1 Impact Measuring Variables: There are some parameters to conduct the study to know the existing setup and the total scenario of the study area. For this study, the following parameters have selected to know the present situation of the study area.
3.2 Study area: To investigate the impact of disaster on income distribution, the study area was chosen on the basis of intensity and frequency of disaster in that particular area. One of the significant coastal cyclone prone areas of Bangladesh is Kolapara Upzilla [1] and it has been selected for this study. Moreover, Kolapara is one of the most affected areas because of SIDR. The study was carried out in three Unions [2] namely- Kolapara Pourosova [3] , Nilgonj and Mithagonj; and these Unions are under Kolapara Upazila in Potuakhali District, a southwestern coastal region of Bangladesh. These study areas are located on the Southeastern part of Kolapara Upazila. To select the location of the study area, emphasis was given mainly on the mostly damaged part of Kolapara Upzilla, which comprises of the three different locations as stated above.
3.3 Sampling Unit and Sample Size Determination: For this study we adopted firstly, stratified random sampling targeting income differential among households, and then secondly, proportional random sampling was used to get the target group from different strata. In this study, each Household (HH) is considered as a unit. Side by side, sample size determination was also an important factor of this study. To avoid error it is essential to consider a very good proportion as sample size and hence 324 households were taken as samples.
4. Socio Economic Conditions analysis
4.1 Population: Population table below shows demographic feature of the study area. In the study area total number of population is 7836, among them 49.72% is male and 50.28% is female. The total number of household is 1518. The average member of the household is around 5.
4.2 Age-sex structure: The major age group of the population is 0-9 and 18-34 years (1795 and 1713 respectively). Age Sex distribution of the people of the study is shown below.
4.3 Educational Status: The literacy rate of the survey area is not satisfactory enough. More than 38.27% people are illiterate in the surveyed area and only 8.64% people are graduate. A significant portion of people has completed their secondary education and it is 27.16%.
4.4 Toilet Facility: In the study area, condition of toilet facility is not hygienic as because most of the peoples rely in unhygienic practices as the super cyclone SIDR has destroyed all the sanitation facility in the surveyed area. Only 34.56% of total households have the sanitary toilet facility. Around 61.72% people have no access of sanitary latrine.
4.5 Water Supply: There is no government water supply system in the study area. The major source of safe drinking water of study area is deep tube well which is provided by the different local NGOs as well as Union Parishad but it is not enough in respect to their need. They usually use pond for drinking as well as domestic purposes but due to SIDR the salinity of pond water is increase. The pure source of drinking water is far away from the locality and they have to rely on the pond water, which is unhealthy.
4.6 Condition of the House: The people in the surveyed area are mainly poor and they are using kucca [4] residence for their living. Around 73% household are using kucca house. Only 19.20% household are using semi-pucca [5] house and 7.81% household are using pucca [6] house.
4.7 Income Expenditure Pattern: Before SIDR average income of the low income group people are 5347 tk per month and for middle and higher income deciles, it is 11572 tk and 22311 tk respectively. A decrease of average income is found after SIDR and for lower income deciles it is now 4035 and it is also true for middle income deciles but average income for higher income deciles is increasing and it is now 23656 higher than before.
The expenditure pattern of the surveyed has changed due to SIDR. Normally the people spent more on food compared to others. Before SIDR, 64% of the total expenditure belongs to food. Again, only 10% of the total expenditure goes to educational purposes. The people have very little scope of education and they spend only 3% of their total expenditure on recreation. But after SIDR, expenditure on food decline from 64% to 58%. The expenditure on health increases from 3% to 10% because due to SIDR people become more affected on diseases than before. Moreover, expenditure on education decreases by 1% after SIDR.
4.8 Residential Ownership and House Condition Cross-tabulation: The cross tabulation analysis between residential ownership and housing condition shows that 78.4% self-owned residence are kucca and only 2.7% are pucca. In the survey, only 8 persons have rented house and it is kucca. An important findings in this analysis is that the household who got their house inheritably, most of their houses are pucca and it is around 67% (66.7% exactly) and 33.3% are semi-pucca.
4.9 Education and Occupation Pattern Cross Tabulation: The cross tabulation analysis between occupation pattern and education shows that most of the illiterate people have engaged them in fishing and agriculture. About 54.8% and 32.3% people are engaged in fishing and agriculture respectively who are illiterate. Again, 83.3% people who have completed their S.S.C are engaged in agriculture. Most of the people who are educated are doing service. Around 85.7% people are doing service who are graduate. Almost 50% people are doing businesses that have completed their H.S.C education.
5. Data Analysis and Discussion
5.1 Loss during SIDR: The super cyclone SIDR affects badly to the people of the surveyed area and it destroyed all. The loss was too severe to explain. It destroyed houses, trees and washed away animals, crops, net etc. The total loss during SIDR is divided into income loss and asset loss. The cyclone hit the poor people badly.
5.1.1 Income loss during SIDR: Among the income loss, there are five categories occupation, rent, livestock, crops and business. From the field survey, it is found that among the income loss, 76% belongs to damage of crops, the second largest damage goes to the loss in the business, and it is 17% of the total income loss.
Asset loss during SIDR: Due to SIDR, a huge damage has occurred in case of asset. Among the total loss, 32% loss goes to house and then to gher and it is 20%. Normally the people of the surveyed area are engaged in fishing and they loss their net due to SIDR and the loss of net is around 13%
5.2 Statistical Analysis:
5.2.1 Correlation between Income before and after SIDR: Correlation has done between incomes before and after SIDR and the result is 0.850. It shows that there exists a high degree of positive correlation between income before and after SIDR.
5.2.2 Partial Correlation between Income before and after SIDR: Partial correlation has done between the income before and after SIDR and in this case the family size has considered as control variable and the result have shown in the following 6.2 table.
The result shows that if we consider the size of the family as a control variable then there exist a high degree of positive correlation between income before and after SIDR and it is .846 and the result is significant at 5% level.
5.2.3 Correlation between family size and income before SIDR: The correlation between family size and income before SIDR has done and the result (0.277) shows that there is very low degree of correlation between family size and income before SIDR.
5.2.4 Correlation between family size and income after SIDR: The correlation between family size and income after SIDR has done and the result is .191, which shows very low degree of positive correlation. The result is not significant at 5% level.
5.2.5 Partial Correlation between Income before and after SIDR: Partial correlation has done between income before and after SIDR where year of education is considered as a control variable. The result is .851, which shows high degree of positive correlation, and it is significant at 5% level.
5.2.6 Pre and post SIDR income Regression Model Analysis
In the pre SIDR income regression model, the relationship of income before SIDR is explained with year of education and size of the family. The R2 for this model is 0.083 which means that 8.3% of the model is explained.
To explain the model we can write as…………………… (i)
Where,
= Income in pre SIDR period
= Constant = Error term
= Coefficient of Year of education = Year of education
= Coefficient of family Size = Family size
The following coefficient table (Table 6.6) shows the value of constant and coefficient of different independent variable. This table shows that all the independent variables are positively correlated with the dependent variable. Besides, all independent variables under this model are significant at 5% level of significance.
Again in the post SIDR regression model, the relationship of income after SIDR is also explained with year of education and size of the family. The R2 for this model is 0.057 which means that 5.7% of the model is explained.
To explain the model we can write as…………………… (ii)
Where,
= Income in post SIDR period
= Constant = Error term
= Coefficient of year of education = Year of education
= Coefficient of family Size = Family size
The above coefficient table (Table 6.6) shows the value of constant and coefficient of different independent variable. This table shows that all the independent variables are positively correlated with the dependent variable but in this case year of education, one of the independent variable is not significant since the p value is more than .05 while the other independent variable family size is significant.
5.2.7 Pre and post SIDR Labor Income Regression Model Analysis
In the pre SIDR labor income regression model, the relationship of labor income before SIDR is explained with income from service, agriculture, fishing, business and day labor. The R2 for this model is 0.912 which means that 91.2% of the model is explained.
To explain the model we can write as-…………………… (iii)
Where,
= Labor income in pre SIDR period
= Constant = Error term
= Coefficient of service = Labor income from service
= Coefficient of agriculture = Labor income from agriculture
= Coefficient of fishing = Labor income from fishing
= Coefficient of business = labor income from business
= Coefficient of day labor = labor income from day labor
The following coefficient table (Table 6.7) shows the value of constant and coefficient of different independent variable. This table shows that all the independent variables are positively correlated with the dependent variable. Besides, all independent variables under this model are significant at 5% level.
Similarly, in the post SIDR labor income regression model, the relationship of labor income after SIDR is explained with income from service, agriculture, fishing, business and day labor. The R2 for this model is 0.966 which means that 96.6% of the model is explained.
To explain the model we can write as=…………………… (iv)
Where,
= Labor income in post SIDR period
= Constant = Error term
= Coefficient of service = Labor income from service
= Coefficient of agriculture = Labor income from agriculture
= Coefficient of fishing = Labor income from fishing
= Coefficient of business = labor income from business
= Coefficient of day labor = labor income from day labor
The above coefficient table (Table 6.7) shows the value of constant and coefficient of different independent variable. This table shows that all the independent variables are positively correlated with the dependent variable. Besides, all independent variables under this model are significant at 5% level.
5.3 Income Inequality Measurement
The concept of inequality is distinct from that of poverty and fairness. Income inequality metrics or income distribution metrics are used here to measure the income inequality of the Kolapara Upzilla.
The most common metrics used to measure inequality are the Gini Index (also known as Gini Coefficient), the Theil Index, and the Hoover Index. In this research, to measure the income inequality of the surveyed area Gini Index and Hoover Index is used.
5.3.1 Gini Index
The Gini Index is the most frequently used inequality index. The reason for its popularity is that it is easy to understand how to compute the Gini Index as a ratio of two areas in Lorenz curve diagrams. To measure the income inequality, at first Gini Coefficient have used. In the following table the spreadsheet computation of Gini Coefficient and Hoover Index is presented. And from this table, the income inequality of the study area is measured.
According to the spreadsheet, the income of the people in the surveyed area has divided into three categories. They are
Group one (0 to less than 8000 Tk. per month)
Group two (8001 to less than 15000 Tk. per month)
Group Three (15001 Tk. and above per month)
Each of the range or group has the same number of member and in this research, it is 108 and the total sample is 324
Calculation of income Inequality through Gini Index
(a) Gini Coefficient before SIDR
Member per Group, Income per Group,
Members of the Group 1, A1 = 108 Income of Group 1, E1 = 577533
Members of the Group 2, A2 = 108 Income of Group 2, E2 = 1249800
Members of the Group 3, A3 = 108 Income of Group 3, E3 = 4236933
Income per Individual, Accumulated Income,
ē1 = E1/A1 = 5347.52 K1 = E1 = 577533
ē2 = E2/A2 = 11572.22 K2 = E2 + K1 = 1827333
ē3 = E3/A3 = 22311.11 K3 = E3 + K2 = 4236933
Gini per Group,
Gini for Group 1, G1 = (2 * K1 - E1) * A1 = 62373564
Gini for Group 2, G2 = (2 * K2 - E2) * A2 = 259725528
Gini for Group 3, G3 = (2 * K3 - E3) * A3 = 654940728
So, ΣA = 324 ; ΣE = 4236933; ΣG = 977039820
Income Inequality of the study area before SIDR
Gini = 1 – [{ΣG/ΣA}/ΣE]
= 1- [{977039820/324}/4236933]
= 1- 0.71
= 0.29
(b) Calculation of Gini Coefficient after SIDR
Member per Group, Income per Group,
Members of the Group 1, A1 = 194 Income of Group 1, E1 = 782820
Members of the Group 2, A2 = 52 Income of Group 2, E2 = 547900
Members of the Group 3, A3 = 78 Income of Group 3, E3 = 1845200
Income per Individual, Accumulated Income,
ē1 = E1/A1 = 4035.15 K1 = E1 = 782820
ē2 = E2/A2 = 10536.54 K2 = E2 + K1 = 1330720
ē3 = E3/A3 = 23656.41 K3 = E3 + K2 = 3175920
Gini per Group,
Gini for Group 1, G1 = (2 * K1 - E1) * A1 = 151867080
Gini for Group 2, G2 = (2 * K2 - E2) * A2 = 109904080
Gini for Group 3, G3 = (2 * K3 - E3) * A3 = 351517920
So, ΣA = 324; ΣE = 3175920; ΣG = 613289080
Income Inequality of the study area before SIDR
Gini = 1 – [{ΣG/ΣA}/ΣE]
= 1- [{613289080/324}/3175920]
= 1- 0.60
= 0.40
Findings of Gini Index Measure
The result of the Gini Index shows that income inequality prevails in the study area and pre SIDR period, the value of Gini Coefficient was 0.29. The range of the Gini Index is between 0 and 1 (0% and 100%), where 0 indicates perfect equality and 1 (100%) indicates maximum inequality. So, the value shows that there exists income inequality in pre SIDR period and it increases more due to SIDR and it is now 0.40. So, it can be said that disaster is increasing the income inequality.
5.3.2 Calculation of income Inequality through Hoover Index
(a) Hoover coefficient before SIDR
Member per Group, Income per Group,
Members of the Group 1, A1 = 108 Income of Group 1, E1 = 577533
Members of the Group 2, A2 = 108 Income of Group 2, E2 = 1249800
Members of the Group 3, A3 = 108 Income of Group 3, E3 = 4236933
Relative Deviation per Group, Hoover per Group,
D1 = E1/ΣE - A1/ΣA = -0.20 H1 = abs (D1) = 0.20
D2 = E2/ΣE - A2/ΣA = -0.04 H2 = abs (D2) = 0.04
D3 = E3/ΣE - A3/ΣA = 0.24 H3 = abs (D3) = 0.24
So, ΣH = H1 + H2 + H3 = 0.48
Income Inequality of the study area before SIDR
Hoover = ΣH/2 = 0.48/2 = 0.24
(b) Hoover coefficient after SIDR
Member per Group, Income per Group,
Members of the Group 1, A1 = 194 Income of Group 1, E1 = 782820
Members of the Group 2, A2 = 52 Income of Group 2, E2 = 547900
Members of the Group 3, A3 = 78 Income of Group 3, E3 = 1845200
Relative Deviation per Group, Hoover per Group,
D1 = E1/ΣE - A1/ΣA = -0.35 H1 = abs (D1) = 0.35
D2 = E2/ΣE - A2/ΣA = -0.01 H2 = abs (D2) = 0.01
D3 = E3/ΣE - A3/ΣA = 0.34 H3 = abs (D3) = 0.34
So,
ΣH = H1 + H2 + H3 = 0.70
Income Inequality of the study area before SIDR
Hoover = ΣH/2 = 0.70/2 = 0.35
Findings of Hoover Index Measure
The result of the Hoover Index shows that income inequality prevails in the study area and pre SIDR value of Hoover Coefficient was 0.24. The range of the Hoover Index is between 0 and 1 (0% and 100%), where 0 indicates perfect equality and 1 (100%) indicates maximum inequality. So, the value shows that there exists income inequality in pre SIDR period and it increases up to 0.35 after SIDR. So, it can be said that disaster is increasing the income inequality.
6. Conclusion
No one can get rid of the destructive impact of disaster. The damage of cyclone is enormous as well as hard to estimate. The higher class people have huge amount capital and assets in the form of cultivable land, livestock, gher etc. and as they have huge asset the loss of disaster is severe. Again, the lower income deciles people have negligible portion of assets from which they have to survive. Due to disaster, these portions of people lose all their assets and they are now living below the subsistence level (Burton et al., 1978). Income inequality is one of the major reasons for the backwardness of the country and it also prevails in the study area. The analysis of the study shows that before SIDR the value of Gini Coefficient was .30 and after SIDR it is now .37. The result is alarming for the country because poor people are the main victim of these consequences. Another point which is needed to taken into consideration that after SIDR, the lower income deciles people are more vulnerable and they are exploited by the higher class people. This is because the rich people are now giving loan which is named as “Dadon” to the poor people at higher interest rate. So, the fact is that the earning of the rich people is increasing more and the poor people are becoming poorer.
Actually it is very difficult to model the effects of disaster. Despite these difficulties, modeling the effects of disasters, even with region specific models, remains an important task. More information is needed to come to a conclusion on the effect of disaster. While economists may still be a long way away from one general model of disaster recovery, building models at a regional level, or specific to types of disasters can help foster a core understanding of if and how the poor recover, and what mitigation and alleviation strategies are successful.
(Knowles, 2001, Anon, 1998, Lindell and Prater, 2003, Fielding and Torres, 2006)
Disaster And Income Inequality: Coastal Areas Of Bangladesh
Bangladesh becomes one of the most vulnerable countries in the world due to climate change. This phenomenon manifests itself through various types of natural disasters. For Bangladesh cyclone and tidal surge are the most frequent disasters, especially to the coastal regions. Almost every year the coastal regions are severely affected by cyclones causing a lot of damages. Such climatic shocks affect the entire socioeconomic structure in the coastal belt. The most affected economic indicator in coastal belt is income of the people. Frequent occurrence of these disasters creates income inequality problem in coastal regions. Most of the people living in coastal areas possess lower standard of living and these people are usually very much vulnerable to any climatic shock that comes about in the form of disasters. Since disasters affect the poor people badly, the study analyzes the change in income of the household due to disaster. This study investigates the influence of disaster on income inequality. A spreadsheet computation of income inequality is used in this paper to measure the income inequality, and a comparison between past and present income levels of people is figured out once a disaster hits the coastal belt.
Key Words: Disaster, Income Inequality, Gini Index, Hoover Index, Labor Income.
Introduction
Since independence in 1971, Bangladesh has started her war against numerous socio-economic, political and socio-political problems with a view to escalating her status from bottom level LDC to top level LDC. Among a long list of existing problems like rapid population growth, poverty, illiteracy, gender disparity, low level of economic growth etc. income inequality takes the crucial role that actually undermines the sustainable development process of Bangladesh (Khan and Hossain, 1989). A decade ago the major causes behind income inequality were mostly socioeconomic ones, particularly in the coastal belt of Bangladesh. However, at present the most powerful determinant creating income inequality in coastal belt does not belong to socioeconomic aspects, rather to climatic factors in the form of natural disasters (Carter and Maluccio, 2003).
Natural disasters are most frequent events in Bangladesh. Bangladesh often makes top news all over the world due to natural disaster, which usually means a huge death toll and massive destruction for one of the poorest nations in the world (Brammer, 1996). Recently, on November 15, 2007, thousands have died and millions were rendered homeless by a super-cyclone named SIDR. It devastated southern and central Bangladesh. According to official sources, some 3,313 people were killed and 28,128 injured by the deadliest cyclone in the history of post-independence Bangladesh. Due to SIDR the estimated damage in terms of US dollar was 2300 million (EM-DAT, 2010). However, the unofficial sources said that the total death will be between 5,000 and 10,000. Cyclone SIDR swept across 16 coastal districts with wind speeds of 240 km/h (150 mph). The winds whipped up waves three to five meters high. The cyclone flattened homes, uprooted trees, electricity, and telephone poles, washed away roads and bridges, destroyed crops, and killed livestock. The death toll would have been far worse if limited emergency plans had not been in place (Bilham and England, 2001). An early warning system enabled the evacuation of an estimated 3.2 million people to higher ground and safety. In contrast, during 1991, Cyclone Gorky claimed at least 138,000 lives and in 1970, Cyclone Bhola killed about half a million people in Bangladesh (Nizamuddin, 2001).
Both the frequency and the scale of natural disasters have increased in recent decades, especially the frequency of disasters does have adverse effect on developing countries like Bangladesh (Barkun, 1977). Despite increasing frequency, natural disasters have often been ignored by economists because of the difficulty regarding making generalized conclusions about their impacts. Disaster is mainly of two types and they are the natural disaster and the manmade disaster (Benson and Clay, 2004). It is quite difficult to calculate and make a conclusion on all disaster. In this paper, only a cyclone has been taken into account as natural disaster and it is the recent devastating super cyclone SIDR occurred on 15th November 2007.
As we have already mentioned that income inequality is one of the major problems in Bangladesh, which reveals the degree of disparity between high and low incomes. The degree of income inequality is often regarded as an important indication of the fairness of the society we live in. A high level of income inequality may be harmful to the level of social integration among the members. Inequality in income and other welfare indicators reflect underlying inequality in asset ownership pattern (Ferreira et al., 2005). The rise in income inequality in Bangladesh has attributed to the widening disparity in landholdings and education levels along with the rising income inequality between the urban and rural sectors (Deininger and Squire, 1996, Sen and Ali, 2005). Now a days the natural disasters are highly responsible for creating income inequality in Bangladesh. Therefore, this paper focuses on the impact of disaster on income change in the study areas between pre and post SIDR period.
1. Objectives of the Study
The objectives of this study are to – assess whether disaster affect the income or not; and what extent disaster influences the factors that are directly related with income distribution.
2. Literature Review
Burton (1978) attempts to measure the impact of disaster and finds that in a disaster situation, those living near the poverty line can easily slip below it. The landless and near landless may be forced to sell their limited assets for survival needs. Those who depend upon wage labor for subsistence are forced to compete with those entering the labor market. The labor market, in turn, becomes further depressed by the loss of harvests and alternative sources of employment.
According to Lindell and Prater (2003), empirical work at a macro-economic level concludes that natural disasters have no long term adverse economic effects, even at a community level. However, household surveys suggest that households living at or near the subsistence level are unable to recuperate from natural disasters, and that a bad disaster can have intergenerational effects on household health and productivity.
Again, Hubner (2008) attempts to identify communities respond to disasters by the following strategies: reducing consumption, drawing from savings, selling productive assets, migration, and borrowing money. Lower income populations do not have savings to utilize or the resources to migrate. Their remaining options lead to reductions in both current and future consumption. Therefore, there is a disconnection between large-scale surveys reporting aggregate recovery and micro-level research implying long-term reductions in consumption. If an economy recovers evenly across all income levels, then we should see no long term changes in income distribution or consumption. However, if assets are flowing from lower to higher income deciles, then we should see increased income inequality and lower consumption marked by increased volatility at lower income deciles.
Tol and Leek (1998) observe and argue that damages from disasters are hard to estimate accurately for several reasons. Especially in more developed nations, total estimated costs of disasters may not include preventative costs aimed at mitigation. In developing nations, damages to poor households will not be accounted for in formal book-keeping, insurance receipts or other market indicators.
Bibbee et. al. (2000) assess that the incomes and businesses of lower income deciles are more vulnerable to disruptions in their supply and demand chains, and have a harder time receiving formal credit to begin renovating their businesses.
Britton (1986) mentions that disaster is also viewed as a mental construct imposed upon experience. This is because to understand disaster knowing the number of deaths, the value of property destroyed or the decrease in per capita income is not sufficient. The symbolic component requires knowledge of the sense of vulnerability, the adequacy of available explanation and the society’s imagery of death and destruction.
Knowles (2001) points out that there is a negative correlation between inequality and economic growth when consistently measured expenditure data, which take redistribution of income into account, are used. This paper also showed that an unequal pre-tax distribution of income leads to pressure for distorted transfers from the rich to the poor, which will in turn reduce the rate of growth.
Moreover, Fielding and Torres (2006) show how to estimate simultaneously the relationship between income inequality and other indicators of social and development by using cross country data. The overall picture which is shown by this paper is that there is a correlation between reductions in inequality and improvements in development indicators such as per capita income, literacy and life expectancy.
Different views have come from the above discussion and most of the researchers have tried to focus on the disaster management and preventive measure. Again, in case of income inequality, the researcher only mentions its importance as a tool to measure economic development. However, they haven’t focused the relationship between the disaster and income inequality. As income inequality is a serious problem and disaster is also an annual phenomenon, it is very much essential to find out the interdependency between these two and we have tried to do it through this paper.
3. Materials and Methods
The study is an out and out primary data-based work. The data were collected through questionnaire survey and observation method. Collection of data was conducted in different units from the study area and after compilation of data; various outcomes of household factors are shown accordingly.
3.1 Impact Measuring Variables: There are some parameters to conduct the study to know the existing setup and the total scenario of the study area. For this study, the following parameters have selected to know the present situation of the study area.
3.2 Study area: To investigate the impact of disaster on income distribution, the study area was chosen on the basis of intensity and frequency of disaster in that particular area. One of the significant coastal cyclone prone areas of Bangladesh is Kolapara Upzilla [1] and it has been selected for this study. Moreover, Kolapara is one of the most affected areas because of SIDR. The study was carried out in three Unions [2] namely- Kolapara Pourosova [3] , Nilgonj and Mithagonj; and these Unions are under Kolapara Upazila in Potuakhali District, a southwestern coastal region of Bangladesh. These study areas are located on the Southeastern part of Kolapara Upazila. To select the location of the study area, emphasis was given mainly on the mostly damaged part of Kolapara Upzilla, which comprises of the three different locations as stated above.
3.3 Sampling Unit and Sample Size Determination: For this study we adopted firstly, stratified random sampling targeting income differential among households, and then secondly, proportional random sampling was used to get the target group from different strata. In this study, each Household (HH) is considered as a unit. Side by side, sample size determination was also an important factor of this study. To avoid error it is essential to consider a very good proportion as sample size and hence 324 households were taken as samples.
4. Socio Economic Conditions analysis
4.1 Population: Population table below shows demographic feature of the study area. In the study area total number of population is 7836, among them 49.72% is male and 50.28% is female. The total number of household is 1518. The average member of the household is around 5.
4.2 Age-sex structure: The major age group of the population is 0-9 and 18-34 years (1795 and 1713 respectively). Age Sex distribution of the people of the study is shown below.
4.3 Educational Status: The literacy rate of the survey area is not satisfactory enough. More than 38.27% people are illiterate in the surveyed area and only 8.64% people are graduate. A significant portion of people has completed their secondary education and it is 27.16%.
4.4 Toilet Facility: In the study area, condition of toilet facility is not hygienic as because most of the peoples rely in unhygienic practices as the super cyclone SIDR has destroyed all the sanitation facility in the surveyed area. Only 34.56% of total households have the sanitary toilet facility. Around 61.72% people have no access of sanitary latrine.
4.5 Water Supply: There is no government water supply system in the study area. The major source of safe drinking water of study area is deep tube well which is provided by the different local NGOs as well as Union Parishad but it is not enough in respect to their need. They usually use pond for drinking as well as domestic purposes but due to SIDR the salinity of pond water is increase. The pure source of drinking water is far away from the locality and they have to rely on the pond water, which is unhealthy.
4.6 Condition of the House: The people in the surveyed area are mainly poor and they are using kucca [4] residence for their living. Around 73% household are using kucca house. Only 19.20% household are using semi-pucca [5] house and 7.81% household are using pucca [6] house.
4.7 Income Expenditure Pattern: Before SIDR average income of the low income group people are 5347 tk per month and for middle and higher income deciles, it is 11572 tk and 22311 tk respectively. A decrease of average income is found after SIDR and for lower income deciles it is now 4035 and it is also true for middle income deciles but average income for higher income deciles is increasing and it is now 23656 higher than before.
The expenditure pattern of the surveyed has changed due to SIDR. Normally the people spent more on food compared to others. Before SIDR, 64% of the total expenditure belongs to food. Again, only 10% of the total expenditure goes to educational purposes. The people have very little scope of education and they spend only 3% of their total expenditure on recreation. But after SIDR, expenditure on food decline from 64% to 58%. The expenditure on health increases from 3% to 10% because due to SIDR people become more affected on diseases than before. Moreover, expenditure on education decreases by 1% after SIDR.
4.8 Residential Ownership and House Condition Cross-tabulation: The cross tabulation analysis between residential ownership and housing condition shows that 78.4% self-owned residence are kucca and only 2.7% are pucca. In the survey, only 8 persons have rented house and it is kucca. An important findings in this analysis is that the household who got their house inheritably, most of their houses are pucca and it is around 67% (66.7% exactly) and 33.3% are semi-pucca.
4.9 Education and Occupation Pattern Cross Tabulation: The cross tabulation analysis between occupation pattern and education shows that most of the illiterate people have engaged them in fishing and agriculture. About 54.8% and 32.3% people are engaged in fishing and agriculture respectively who are illiterate. Again, 83.3% people who have completed their S.S.C are engaged in agriculture. Most of the people who are educated are doing service. Around 85.7% people are doing service who are graduate. Almost 50% people are doing businesses that have completed their H.S.C education.
5. Data Analysis and Discussion
5.1 Loss during SIDR: The super cyclone SIDR affects badly to the people of the surveyed area and it destroyed all. The loss was too severe to explain. It destroyed houses, trees and washed away animals, crops, net etc. The total loss during SIDR is divided into income loss and asset loss. The cyclone hit the poor people badly.
5.1.1 Income loss during SIDR: Among the income loss, there are five categories occupation, rent, livestock, crops and business. From the field survey, it is found that among the income loss, 76% belongs to damage of crops, the second largest damage goes to the loss in the business, and it is 17% of the total income loss.
Asset loss during SIDR: Due to SIDR, a huge damage has occurred in case of asset. Among the total loss, 32% loss goes to house and then to gher and it is 20%. Normally the people of the surveyed area are engaged in fishing and they loss their net due to SIDR and the loss of net is around 13%
5.2 Statistical Analysis:
5.2.1 Correlation between Income before and after SIDR: Correlation has done between incomes before and after SIDR and the result is 0.850. It shows that there exists a high degree of positive correlation between income before and after SIDR.
5.2.2 Partial Correlation between Income before and after SIDR: Partial correlation has done between the income before and after SIDR and in this case the family size has considered as control variable and the result have shown in the following 6.2 table.
The result shows that if we consider the size of the family as a control variable then there exist a high degree of positive correlation between income before and after SIDR and it is .846 and the result is significant at 5% level.
5.2.3 Correlation between family size and income before SIDR: The correlation between family size and income before SIDR has done and the result (0.277) shows that there is very low degree of correlation between family size and income before SIDR.
5.2.4 Correlation between family size and income after SIDR: The correlation between family size and income after SIDR has done and the result is .191, which shows very low degree of positive correlation. The result is not significant at 5% level.
5.2.5 Partial Correlation between Income before and after SIDR: Partial correlation has done between income before and after SIDR where year of education is considered as a control variable. The result is .851, which shows high degree of positive correlation, and it is significant at 5% level.
5.2.6 Pre and post SIDR income Regression Model Analysis
In the pre SIDR income regression model, the relationship of income before SIDR is explained with year of education and size of the family. The R2 for this model is 0.083 which means that 8.3% of the model is explained.
To explain the model we can write as…………………… (i)
Where,
= Income in pre SIDR period
= Constant = Error term
= Coefficient of Year of education = Year of education
= Coefficient of family Size = Family size
The following coefficient table (Table 6.6) shows the value of constant and coefficient of different independent variable. This table shows that all the independent variables are positively correlated with the dependent variable. Besides, all independent variables under this model are significant at 5% level of significance.
Again in the post SIDR regression model, the relationship of income after SIDR is also explained with year of education and size of the family. The R2 for this model is 0.057 which means that 5.7% of the model is explained.
To explain the model we can write as…………………… (ii)
Where,
= Income in post SIDR period
= Constant = Error term
= Coefficient of year of education = Year of education
= Coefficient of family Size = Family size
The above coefficient table (Table 6.6) shows the value of constant and coefficient of different independent variable. This table shows that all the independent variables are positively correlated with the dependent variable but in this case year of education, one of the independent variable is not significant since the p value is more than .05 while the other independent variable family size is significant.
5.2.7 Pre and post SIDR Labor Income Regression Model Analysis
In the pre SIDR labor income regression model, the relationship of labor income before SIDR is explained with income from service, agriculture, fishing, business and day labor. The R2 for this model is 0.912 which means that 91.2% of the model is explained.
To explain the model we can write as-…………………… (iii)
Where,
= Labor income in pre SIDR period
= Constant = Error term
= Coefficient of service = Labor income from service
= Coefficient of agriculture = Labor income from agriculture
= Coefficient of fishing = Labor income from fishing
= Coefficient of business = labor income from business
= Coefficient of day labor = labor income from day labor
The following coefficient table (Table 6.7) shows the value of constant and coefficient of different independent variable. This table shows that all the independent variables are positively correlated with the dependent variable. Besides, all independent variables under this model are significant at 5% level.
Similarly, in the post SIDR labor income regression model, the relationship of labor income after SIDR is explained with income from service, agriculture, fishing, business and day labor. The R2 for this model is 0.966 which means that 96.6% of the model is explained.
To explain the model we can write as=…………………… (iv)
Where,
= Labor income in post SIDR period
= Constant = Error term
= Coefficient of service = Labor income from service
= Coefficient of agriculture = Labor income from agriculture
= Coefficient of fishing = Labor income from fishing
= Coefficient of business = labor income from business
= Coefficient of day labor = labor income from day labor
The above coefficient table (Table 6.7) shows the value of constant and coefficient of different independent variable. This table shows that all the independent variables are positively correlated with the dependent variable. Besides, all independent variables under this model are significant at 5% level.
5.3 Income Inequality Measurement
The concept of inequality is distinct from that of poverty and fairness. Income inequality metrics or income distribution metrics are used here to measure the income inequality of the Kolapara Upzilla.
The most common metrics used to measure inequality are the Gini Index (also known as Gini Coefficient), the Theil Index, and the Hoover Index. In this research, to measure the income inequality of the surveyed area Gini Index and Hoover Index is used.
5.3.1 Gini Index
The Gini Index is the most frequently used inequality index. The reason for its popularity is that it is easy to understand how to compute the Gini Index as a ratio of two areas in Lorenz curve diagrams. To measure the income inequality, at first Gini Coefficient have used. In the following table the spreadsheet computation of Gini Coefficient and Hoover Index is presented. And from this table, the income inequality of the study area is measured.
According to the spreadsheet, the income of the people in the surveyed area has divided into three categories. They are
Group one (0 to less than 8000 Tk. per month)
Group two (8001 to less than 15000 Tk. per month)
Group Three (15001 Tk. and above per month)
Each of the range or group has the same number of member and in this research, it is 108 and the total sample is 324
Calculation of income Inequality through Gini Index
(a) Gini Coefficient before SIDR
Member per Group, Income per Group,
Members of the Group 1, A1 = 108 Income of Group 1, E1 = 577533
Members of the Group 2, A2 = 108 Income of Group 2, E2 = 1249800
Members of the Group 3, A3 = 108 Income of Group 3, E3 = 4236933
Income per Individual, Accumulated Income,
ē1 = E1/A1 = 5347.52 K1 = E1 = 577533
ē2 = E2/A2 = 11572.22 K2 = E2 + K1 = 1827333
ē3 = E3/A3 = 22311.11 K3 = E3 + K2 = 4236933
Gini per Group,
Gini for Group 1, G1 = (2 * K1 - E1) * A1 = 62373564
Gini for Group 2, G2 = (2 * K2 - E2) * A2 = 259725528
Gini for Group 3, G3 = (2 * K3 - E3) * A3 = 654940728
So, ΣA = 324 ; ΣE = 4236933; ΣG = 977039820
Income Inequality of the study area before SIDR
Gini = 1 – [{ΣG/ΣA}/ΣE]
= 1- [{977039820/324}/4236933]
= 1- 0.71
= 0.29
(b) Calculation of Gini Coefficient after SIDR
Member per Group, Income per Group,
Members of the Group 1, A1 = 194 Income of Group 1, E1 = 782820
Members of the Group 2, A2 = 52 Income of Group 2, E2 = 547900
Members of the Group 3, A3 = 78 Income of Group 3, E3 = 1845200
Income per Individual, Accumulated Income,
ē1 = E1/A1 = 4035.15 K1 = E1 = 782820
ē2 = E2/A2 = 10536.54 K2 = E2 + K1 = 1330720
ē3 = E3/A3 = 23656.41 K3 = E3 + K2 = 3175920
Gini per Group,
Gini for Group 1, G1 = (2 * K1 - E1) * A1 = 151867080
Gini for Group 2, G2 = (2 * K2 - E2) * A2 = 109904080
Gini for Group 3, G3 = (2 * K3 - E3) * A3 = 351517920
So, ΣA = 324; ΣE = 3175920; ΣG = 613289080
Income Inequality of the study area before SIDR
Gini = 1 – [{ΣG/ΣA}/ΣE]
= 1- [{613289080/324}/3175920]
= 1- 0.60
= 0.40
Findings of Gini Index Measure
The result of the Gini Index shows that income inequality prevails in the study area and pre SIDR period, the value of Gini Coefficient was 0.29. The range of the Gini Index is between 0 and 1 (0% and 100%), where 0 indicates perfect equality and 1 (100%) indicates maximum inequality. So, the value shows that there exists income inequality in pre SIDR period and it increases more due to SIDR and it is now 0.40. So, it can be said that disaster is increasing the income inequality.
5.3.2 Calculation of income Inequality through Hoover Index
(a) Hoover coefficient before SIDR
Member per Group, Income per Group,
Members of the Group 1, A1 = 108 Income of Group 1, E1 = 577533
Members of the Group 2, A2 = 108 Income of Group 2, E2 = 1249800
Members of the Group 3, A3 = 108 Income of Group 3, E3 = 4236933
Relative Deviation per Group, Hoover per Group,
D1 = E1/ΣE - A1/ΣA = -0.20 H1 = abs (D1) = 0.20
D2 = E2/ΣE - A2/ΣA = -0.04 H2 = abs (D2) = 0.04
D3 = E3/ΣE - A3/ΣA = 0.24 H3 = abs (D3) = 0.24
So, ΣH = H1 + H2 + H3 = 0.48
Income Inequality of the study area before SIDR
Hoover = ΣH/2 = 0.48/2 = 0.24
(b) Hoover coefficient after SIDR
Member per Group, Income per Group,
Members of the Group 1, A1 = 194 Income of Group 1, E1 = 782820
Members of the Group 2, A2 = 52 Income of Group 2, E2 = 547900
Members of the Group 3, A3 = 78 Income of Group 3, E3 = 1845200
Relative Deviation per Group, Hoover per Group,
D1 = E1/ΣE - A1/ΣA = -0.35 H1 = abs (D1) = 0.35
D2 = E2/ΣE - A2/ΣA = -0.01 H2 = abs (D2) = 0.01
D3 = E3/ΣE - A3/ΣA = 0.34 H3 = abs (D3) = 0.34
So,
ΣH = H1 + H2 + H3 = 0.70
Income Inequality of the study area before SIDR
Hoover = ΣH/2 = 0.70/2 = 0.35
Findings of Hoover Index Measure
The result of the Hoover Index shows that income inequality prevails in the study area and pre SIDR value of Hoover Coefficient was 0.24. The range of the Hoover Index is between 0 and 1 (0% and 100%), where 0 indicates perfect equality and 1 (100%) indicates maximum inequality. So, the value shows that there exists income inequality in pre SIDR period and it increases up to 0.35 after SIDR. So, it can be said that disaster is increasing the income inequality.
6. Conclusion
No one can get rid of the destructive impact of disaster. The damage of cyclone is enormous as well as hard to estimate. The higher class people have huge amount capital and assets in the form of cultivable land, livestock, gher etc. and as they have huge asset the loss of disaster is severe. Again, the lower income deciles people have negligible portion of assets from which they have to survive. Due to disaster, these portions of people lose all their assets and they are now living below the subsistence level (Burton et al., 1978). Income inequality is one of the major reasons for the backwardness of the country and it also prevails in the study area. The analysis of the study shows that before SIDR the value of Gini Coefficient was .30 and after SIDR it is now .37. The result is alarming for the country because poor people are the main victim of these consequences. Another point which is needed to taken into consideration that after SIDR, the lower income deciles people are more vulnerable and they are exploited by the higher class people. This is because the rich people are now giving loan which is named as “Dadon” to the poor people at higher interest rate. So, the fact is that the earning of the rich people is increasing more and the poor people are becoming poorer.
Actually it is very difficult to model the effects of disaster. Despite these difficulties, modeling the effects of disasters, even with region specific models, remains an important task. More information is needed to come to a conclusion on the effect of disaster. While economists may still be a long way away from one general model of disaster recovery, building models at a regional level, or specific to types of disasters can help foster a core understanding of if and how the poor recover, and what mitigation and alleviation strategies are successful.
(Knowles, 2001, Anon, 1998, Lindell and Prater, 2003, Fielding and Torres, 2006)
如果您有論文代寫需求，可以通過下面的方式聯(lián)系我們
點擊聯(lián)系客服