滋賀大学学術情報リポジトリ

(1)

Role of Environmental Aid in Economic Growth and Poverty Reduction:

Evidence from Bangladesh! ７３

Role of Environmental Aid in Economic

Growth and Poverty Reduction:

Evidence from Bangladesh

!

Anupam Saha

5. Does Aid Contribute to Economic Growth? - Methodological Approach

As with the Burnside and Dollar（2000）paper a simple production function model is used, which is,&_*#!%_*$, with#"!and !!$$". The planner maximizes subject to a resources constraint given by "_*"$_*$!%*$"# ('")('% &

Where,$_*is the time t investment, F is foreign aid which has been seg-mented into ea as environmental aid and nea as non-environmental aid. Capital evolves according to

%_*""# "!#% &%*"$*,Where!!#!"

Aid is important for a developing country because it generates capital ac-cumulation. Towards the convergence of steady state the savings rate first rises then falls with the income level, can be explained by diminish-ing returns to capital. This is because the marginal propensity to save in the developing country is high, although the average propensity to save is small and also these countries do not generate marginal income from which it can accumulate capital rapidly. One time donation adding to de-veloping countries GDP would cut a significant span of life to steady state. In this setting it is not difficult to see that foreign assistance in the form of aid is an effective instrument for raising capital accumulation and

(2)

７４ The Hikone Ronso No. 369 November 2007

growth rate of country. 5.1 Data Sources

Bangladesh has only thirty-two years history of foreign aid. Due to unavailability of data during the period 1970 s and because of political unrest, validity or correctness of the model may be blurred to some ex-tent. The foreign aid, as referred to regular concept has used the measure that lumps together grants and concessional loans（converted to constant 1995 dollars）. In my study the amount of disbursed aid is measured by adding the grant component of concessional loans to outright grants di-vided by countries real GDP（in PPP$）. This measure includes all types of aid component including technical assistance. Two types of aid are considered in this study: environmental aid and non-environmental aid. Environmental aid comprises aid disbursed for environmental protection, social security and infrastructural development, in other way, for keeping the environment clean. The data are in constant in US dollar and are in 1995-dollar constant rate. Finally I divided the aid figure by real GDP in constant 1995 prices. The data cover the years from 1979 to 2004 and economic policy variable are also assigned to each year for the period 1979-2004. The dependent variable in my study is real GDP growth rate and explanatory variables are initial income, aid, and population, which are taken from the database of Ministry of Finance, Bangladesh. There are many other national and international organizations based in Dhaka like Bangladesh Bureau of Statistics, Board of Revenues, Investment Board, World Bank and Asian Development Bank are also the source of data for this study.

(3)

Evidence from Bangladesh! ７５

5. 2 Structural Hypothesis

The hypothesis１２is designed to find relationship between aid pol-icy and growth and it is somewhat different from that introduced by Burnside and Dollar（2000）. Model is outlined as follows:

$'#!!"!))'!""!""'"!"&#"'&#'"!$#$#'"!&!&'!""' ……（1） Let$_'be the growth rate of GDP of the country during period t. Let )'!" be the level of GDP in country at the beginning of the period t,"'be the level of aid not as a fraction of GDP received by the country in period t, &#_'is a policy index of time t,$#'is a government expenditure in the year t and &!be the population growth rate of the country in a time period t and"is the error term.

Aid and policy are endogenous and depend on the independent variables in the system and on each other. Thus we consider aid may be a function of policy or policy may be a function of aid.

"'#!"%")'!")"&#'!"&#"(_'$!"(""'" ………（2） And

&##!' &#%")'!&#)""'!&#""(_'$!&#(""' &

………（3）

Here x captures other explanatory variables. The system may contain si-multaneous equation bias. To check simultaneous biasness we need some instrumental variables. Bone（1996）has shown that there are good instruments for aid that do not belong to the aid regression, notably popu-lation, infant mortality rate and proxies of donor’s strategic interests. Be-cause of unavailability of data we skipped simultaneous equation bias test although it purges the correlation of aid with the error term in the growth regression. We rather testified whether autocorrelation exists among er-rors or accept!!: &#!. If &#!,errors may follow the first order au-１２）In this system of equations simultaneous equation approach and its biasness is not considered.

(4)

７６ The Hikone Ronso No. 369 November 2007

toregressive model and the errors will not be autocorrelated. For the null hypothesis of no autocorrelation, we can use !!: )#!, where )is the parameter of error term in an autoregressive model. Durbin-Watson test has been done for each of the scenario to verify the existence of autocor-relation.

We presume that aid and policies are not simultaneously deter-mined. This means that we reject either of the situation: when the country is upon good policy environment, it receives aid or without aid-commitment, country does to go for the improvement of policy. Therefore, it turns out that aid appears to have little or no impact on the policy index and vice-versa. For the simplicity of the model we will ignore this reciprocal inter-relationship. We can now draw down a semi log functional form as be-low:

%*#!("!*+*!""!$#"(%$#$ %"!* '$#&(%'$#$ %"!* $#)#$#_*)#_* "!'$#)#'$#*)#*"!%$%$*"!)!)_*!""………（4） Where$#*and'$#*are environmental aid and non-environmental aid at time t respectively. %*is GDP growth rate at time t, a dependent variable which is to be estimated. In the growth literature the popular and fre-quently taken independent variables are per-capita GDP, population growth rate（see Levine and Renelt, 1992）. I took these variables consid-ering the fact that many other related articles have used the GDP growth rate as dependent variable and variables mentioned above were used as independent variables. Moreover, these independent variables in the above growth equation are very particular and relevant in the context of Bangladesh economy. Population growth rate is taken on the basis of the assumption that a vast number of population live below poverty line（in Calorie intake method it is 2122 kk）in Bangladesh and growth rate in population is higher among the poor group rather than in affordable

(5)

Evidence from Bangladesh! ７７

group. So growth rate of population will eventually represent the popula-tion growth rate in poor community. Government expenditure is one of the important variables for growth also involves strategic development works that indirectly help the poor to raise their income. Policy index is constructed using the following formula. This formula is depicted from a noble literature of Burnside and Dollar（2000）who construct this index using a panel data of 57 countries. Using the following formula, index for Bangladesh has been constructed.

!#(

! "is Policy Index＝1.3 + 5.4 × Budget Surplus − 1.4 × Inflation + 2.1 × Trade Intensity１３.

The interaction term"#&#(&#(allows us to capture effects of aid and pol-icy, suggested in the neoclassical model.

5. 3 Growth Regression

As my intension is to rationalize the effectiveness of environ-mental aid on growth in a policy environment of a particular country, I choose Bangladesh as one of the leading aid recipient country in the world. A method of simple OLS regression analysis（Table 1 ）is used and found that the interaction term ‘Environmental aid × policy’ is posi-tive and significant when we run the full model. Four scenarios depict a clear picture of effectiveness of environmental aid on growth. The Durbin-Watson d statistics is used to determine if there is first-order se-rial correlation in the error term of an equation by examining the residu-als of estimation of the equation 4 and subsequent scenarios. We did not find any existence of autocorrelation in scenarios 1, 3 and 4. The equa-tion for the Durbin-Watson d statistic for t observaequa-tions is used here as: $"# " " %(!%(!! # $"_!# ! "

(6)

７８ The Hikone Ronso No. 369 November 2007

１３）Taken from the policy index equation of Burnside and Dollar.

１４）Null Hypothesis implies positive serial correlation. If the d value is such that#!#_"we can say that there is no evidence of positive serial correlation, if#!!#!#"then the test is inclusive and

if#!!#then there is evidence of positive serial correlation.

Independent Variables Structural Model （Ist Scenario） In Absence of Non-Environmental component （２nd Scenario） Absence of Environmental Component （３rd Scenario） Absence of Policy Component （４th Scenario） Intercept 39.49605 （3.510915） 36.52187 （3.846236） 38.32881 （2.804569） 4.097596 （0.532177） GDP Per capita −12.8824** （−3.53463） −12.9778** （−3.56959） −7.67725 （−1.93526） −3.15752 （−0.80905） Environmental Aid −3.90139 （−1.08721） −2.91014 （−1.25005） −− 4.940717 （2.240249） Non-environmental aid −0.50287 （−0.09588） −− −10.2366* （−2.65274） −1.17243 （−0.45612） Environmental Aid × Policy 0.034195* （2.484528） 0.026542** （4.268446） −− −− Non-environmental Aid × Policy −0.0096 （−0.47758） −− 0.038165** （3.441239） −− Population Growth rate 222.3273** （4.602189） 219.4196** （4.556253） 149.7615** （2.823747） 106.0993 （1.899485） Government Expenditure −0.13933** （−5.50842） −0.12788** （−5.54164） −0.13492** （−4.3822） −0.07043* （−2.70022） Adjusted R Square 0.795377 0.791078 0.693915 0.613644 Durbin-Watson Test （d value） 2.66 （We can not

reject Null-hypothesis）１４

1.11 （Inclusive）

1.87 （We can not

reject Null-hypothesis）

1.89 （We can not

reject Null-hypothesis）

Source: Own calculation. Note: ** mark denotes significance at 1% level and * denotes 5% level of significance. Figures in parenthesis are t-statistics.

(7)

Evidence from Bangladesh! ７９

Scenario 1: When aid is merged with policy it is seen that aid has very positive and significant relationship with the growth. In various studies it is reported that aid has little or no relationship with the GDP growth rate but appropriate allocation of aid to appropriate places where such aid is deemed necessary should instigate economic growth. In our result, not any form of aid either environmental or non-environmental, alone could lead to economic growth. The mere aid-growth relationship is found insignificant and negative.

Scenario 2: In this scenario we exclude ‘non-environmental aid’ term. In absence of non-environmental aid, growth is also affected by the environmental aid provided it is to be treated under policy environment. The interaction term ‘environmental aid × policy’ which is highly signifi-cant and positive imply that under good policy environment aid entails growth. On the other hand environmental aid alone can not contribute to any economic growth and coefficient of this variable is also insignificant. In this scenario we have also seen that even environmental aid can not work without policy environment.

Scenario 3: Let us see what happens when we run our regression without any environmental component. We find the same result as visu-alized in scenario 2. If we want aid to work it is to be treated under pol-icy environment. Non-environmental aid alone is negatively significant but positively significant when policy is incorporated into non-environmental aid. It flourishes a very interesting story. Non environmental aid negatively affects growth. This might be because of mismanagement of this form of aid in implementation process. It also proves in other way the fact that Bangladesh needs more environmental aid than non-environmental aid.

Scenario 4: Importance of policy is acknowledged in this sce-nario. Here aid is not interacted with policy. Neither environmental nor

(8)

８０ The Hikone Ronso No. 369 November 2007

non-environmental aid could demonstrate any contribution to economic growth as coefficients of both types of aid carry insignificant value. This scenario describes that without policy environment aid does not work．

In a sense that this paper is different from Burnside and Dollar （2000）, they consider 56 countries whereas in this paper the single coun-try, Bangladesh is considered. Although both the paper proceeds with the same production function model. Moreover, generalized impression may not be applicable to a country specific situation. In our paper aid has been segmented into environmental-aid and non environmental aid and non-environmental aid even interacting with polity was found neither efficient nor desirable to the country, Bangladesh, which is just opposite to what Burnside and Dollar has established. As mentioned, this is because of typical characteristics of a specific country. A bulk of aid is targeted to environmental projects which comprises disaster management, environ-mental preservation, aforestation, modernization of agri-environenviron-mental organism and so on directly benefit disaster victims and poor relying on environmental resources. It is to be noted that almost every year natural calamities cost Bangladesh a significant amount of economic loss which is about 8% of country’s GDP and Poor are the vain victims of these dis-asters. So environmental aid helps and protect the country from the shocking condition and thus works through environmental preservation and short term employment generation for the poor.

There are reasons to be skeptical about the OLS results presented above. TSLS might be necessary to check the robustness of the result. The reasons why we ignored TSLS are:（1）The data set used in this study is for country specific, we did not take into account of other income countries like low or middle income country what Burnside and Dollar’s paper has considered,（2）Unlike their paper we ignored or did not

(9)

con-Role of Environmental Aid in Economic Growth and Poverty Reduction:

Evidence from Bangladesh! ８１

sider any outliers, which is likely to distort the result（3）Moreover in our system of equations we considered, all variables including aid and policy are exogenous as per the assumption of the model, as a result there is hardly any chance for the variables to be simultaneously biased. We also run our regression with heterokedasticity consistency test but did not find any significant difference from our simple OLS. It is also to be noted that the degree of freedom of our sample is only 19, which is not sufficient, but we had to go with this number firstly because Bangladesh has only 35 years of history from its emergence since 1971 and secondly unavailability of data of other important variables concerned.

Reader may also be skeptical with the negative sign of initial GDP per capita. The coefficient for initial GDP per capita is negative. This part of the study result has actually kept us silent and unexplained. But this negative sign of GDP per capita is consistent with the findings of other researchers including Burnside and Dollar. Still we can attempt to explain it in a way that the nation with greater economic freedom tented to worsen over time（Hekelman and Knack, 2005）. Another explanation of the negative coefficients associated with initial GDP per capita is that conditional convergence in national income is realized. Bangladesh is low income country. Low income country face institutional and socioeco-nomic obstacles and weakness in government planning（please see At-sushi Iimi, Yasuhisa Ojima, JBIC, 2005）．

Huge burden of poverty on the economy may hit the economy negatively unless country practices appropriate allocation of aid and pol-icy. In a good policy environment population growth is also contributing to economic growth. The relationship between GDP growth rate and population growth rate is all along positive but government expenditure is all along negative. Lack of transparency, efficiency and corruption

(10)

ex-８２ The Hikone Ronso No. 369 November 2007

ists in government spending and absence of government commitment and inefficient spending result in slow down of economic growth.

There are two aspects of the derivative of growth with respect to aid with which we are concerned. First, the slope in the policy dimension significantly positive means that aid is not effective in worse policy envi-ronment１５. As regards Bangladesh the mean policy index is 1.31 which is below than mean of global policy standard. More significant and positive relationship might be established if policy index goes close to global standard. The conditional aid for various projects is often redirected due to its failure to meet the requirement of the donor. As a result govern-ment has to take huge amount of aid burden and falls in the trap of debt. In this way if the county maintains good policy or environment where aid can work, foreign assistance in terms of aid contributes to growth. Sec-ond, derivative is negative when examined without any policy dimension. When Aid/GDP alone is introduced in the growth regression, the coeffi-cient of aid/GDP came out with negative sign（Table 1）. These results imply that the impact of aid on growth is a function of both level of pol-icy and the type of aid.

For a country with mean policy and mean aid and an additional 1 percent of GDP in aid would add. 46 percentage point to the growth rate and this impact increases with the level of policy and decrease with the level of aid（See cross country analysis of Collier and Dollar, 2002）. Consistent with the other authors it can be proclaimed that aid has little or no effect on growth, but depends on country’s economic, social and envi-ronmental conditions. For a country like Bangladesh, aid seems essential for the alleviation of poverty and also for the environmental conserva-１５）Good policy is considered at the level 2.5, which is one standard deviation above the mean

(11)

Evidence from Bangladesh! ８３

tion. However a robust finding is that aid has a positive impact on growth in a good policy environment. For Bangladesh, we found low and insufficient score for policy index. It signifies the reason why in spite of huge amount of aid-inflow Bangladesh has seen merely an insig-nificant growth.

Thus policy plays a vital role as catalyst in any development proc-ess. Government of Bangladesh has recently taken ambitious project in cooperation with leading international organizations to alleviate poverty in a sustainable manner. A policy draft has been formulated with a view to attaining the goal of the project. A careful attention has been given in the process of formulating the draft. Failure in attaining the goals is eventually due to the failure to make good policy.

We will reiterate why aid why not other forms of foreign assis-tance are considered in the alleviation of poverty. Before we come into this point we are supposed to discuss what comes first: economic growth or poverty alleviation. Two hypotheses exist in this field of research. One is top-down approach and another one is bottom-up approach. Some research warns sustainable development is to be ensured through poverty alleviation. Because poverty not only delays the development process but also pose harm to the environment. To stop deterioration of environ-ment and to stop the poor rely heavily on environenviron-ment for their lively hood, poverty alleviation is a must. With regard to top-down approach, achieving economic development by the blessings of economic growth and income generation help the country to take off. It approaches that trickle down effect automatically pay off the poor and thus help the poor to get out of poverty. This paper did not argue which approach would better fit to Bangladesh economy rather comparatively reasonable ap-proach has been chosen where aid is assumed to be necessary. In the

(12)

８４ The Hikone Ronso No. 369 November 2007

bottom-up approach, role of foreign assistance in the form of aid is inevi-table when the most of the country’s population is poor. This is because, this big task is only be attainable by government intervention or capital injection through foreign aid. Generally no private entities will be will-ing to finance for poverty alleviation. So aid is one of prime options that fills up the gap of country’s savings thus help the poor to come out of poverty line. On the contrary, it is also true that a bulk of aid has been targeted to poverty alleviation program in Bangladesh, there is success but insignificant. This may be because of lack of good policy or its prac-tices. It also seems reasonable to think that only aid is not sufficient enough to attribute to any growth, it is the policy and other economic fac-tors which are to be interacted with aid to persuade growth.

6. Economic Growth and Poverty

The following part of this paper is to draw a link between eco-nomic growth and poverty. In this part no statistical model is used rather application of data interpretation in a logical way have given validation to the claim that economic growth has positive impact on poverty eradica-tion, although there exits so many arguments in this view. Before we go insight let us see how poverty is commonly derived. Poverty is measured in two common ways. The head count ratio of poverty is generally pre-sented（a）on the basis of cost of basic needs and（b）on the basis of calo-rie intake. Here both measures have been used to avoid marginal change in poverty measurement.

6. 1 Income Poverty Trends

It will be pertinent to present a disaggregation of poverty trends by rural and urban location. Locational pattern of economic growth may

(13)

Evidence from Bangladesh! ８５

then be linked with the poverty trends. Even if a strictly locational disag-gregation of economic growth data is not available, the sectoral pattern of growth may provide a relevant basis for analysis. Data on head count ra-tio（HCR）１６of poverty incidence based on a ‘poverty line income’ has been presented in Table 2. The poverty line is estimated on the ‘cost of basic needs’（CBN）１７method and Daily Calorie Intake１８method. The ta-ble provides the following features of poverty trends:

Poverty incidence in Bangladesh has been continuously declining during 1988-89 to 2000 and decline has been continuous since 1985-86 if DCI１９based HCR are considered. According to CBN method poverty in Bangladesh has increased during 1986 to 1992. In the rural areas, pov-erty incidence increased at a higher rate. There has been an 8.1 percent-age point increase during this period. Among the urban households, pov-erty increased by 2.0 percentage points during these seven years. During the following four years, that is, during the first half of the nineties, both the urban and rural poverty situation improved, but the improvement was less pronounced for the rural areas, compared to the urban areas: the rate of poverty reduction has been 1.1 and 2.5 percentage points per year for the rural and urban households respectively. A contrasting trend has been observed during the later part of the nineties. Rural poverty de-clined by 3.7 percentage points while the urban poverty situation in-creased by 1.6 percentage point during this four- year period.

１６）HDR means ‘Head Count Ratio Index’ of Poverty. １７）CBN means Cost of Basic Needs.

１８）Note: Method uses the calorie intake level of 2122 Kilo calorie as the cut off point for moderate

poverty and 1805 kilo calories for extreme poverty. In the CBN method, poverty lines represent the level of per capita expenditure at which the members of the household can meet their basic needs（defined by a standard bundle）

(14)

８６ The Hikone Ronso No. 369 November 2007

So far I did not cover the earlier period and examined only the pe-riod of 1985-86 to 1999-2000. There are methodological problems in making direct long-term comparison of poverty incidence for the earlier period. Nonetheless, the data for 1973-74 and 1981-82 are comparable. The extents of poverty reduction during the two sub-periods are shown in Table３．

Even if there are methodological differences, the difference in the rate of poverty reduction is so large between the two periods that one may accept the observation that the incidence of poverty declined at a faster rate in the first period than in the second period. The same view has been expressed by other researchers２０.

Year HCR based on CBN（％） HCR based on DCI（％）

Rural Urban National Rural Urban National

1985/86 53.1 42.9 51.7 54.7 62.6 55.7

1988/89 59.2 43.9 59.1 47.8 47.6 47.8

1991/1992 61.2 44.9 58.8 47.6 46.7 47.5

1995/1996 56.7 35.0 53.1 47.1 49.7 47.5

2000 53.0 36.6 49.8 42.3 52.5 44.3

Source: Bangladesh Bureau of Statistics（various years）, WB（1998, 2002）, Note:

Head Count Ratio is based on Upper Poverty Line

Area Percentage point change per year

1973-74 to 1981-82 1983-84 to 1995-96

Urban 2.97 2.93

Rural 1.10 0.40

National 1.29 .92

Source: Bangladesh Bureau of Statistics, 2001, 2002 and Khan（2000）

Table 2: Rural and urban poverty trend

(15)

Evidence from Bangladesh! ８７

6. 2 Growth in Reducing Poverty

A number of studies（Mujeri 2000, Bhattacharya 2001）high-lighted the high annual growth rates of GDP during the recent years. An-nual growth rates are influenced by the growth in the previous years. The trend growth rates of GDP should therefore, be examined to assess the performance of the economy during longer periods.

The trend growth rate of GDP for the three sub-periods have been presented in Table 4. The average growth rate was 2.46 per cent per an-num during 1986 to 1991. The trend growth rates rose to 4.50 per cent and 5.29 per cent per annum during the first and second half of the nine-ties respectively. Even if there has been an acceleration of the rate of growth, it achieved during the latest period was much less than the growth rate targeted for the Sixth Five-Year Plan（1998-2003）.

The acceleration of GDP growth during the latest years has been modest because of the devastating floods of 1998; GDP growth would

２０）A.R. Khan in his recent public lecture on Globalization, delivered on September 12,2001 at BIDS. Year

Percentage per annum Percentage point per year GDP growth rate Change of GDP

growth rate

CBN method based on HCR of Poverty Moderate Poverty Extreme poverty

1986-1991 2.46 1.04 1.18↑ 1.48↑

1991-1996 4.50 0.79 1.43↓ 2.08↓

1996-2001 5.29 -- 0.23↓ 0.98↓

Source: Statistical Year Book, 2001, 2002, Bangladesh Bureau of Statistics. Note: upward directed arrow indicates ‘increase in poverty’ and downward directed arrow indicates ‘de-crease in poverty’

(16)

８８ The Hikone Ronso No. 369 November 2007

have been taken a better shape otherwise. The trend rate of growth could be much higher if the flood damage of 1998 did not occur. Such hypo-thetical case should not be overstretched because natural calamities are facts of life in this country and are observed to occur frequently. Such calamities caused a downturn in economic growth at least once during each of the five-year periods: 1986 to 1991, 1991 to 1996, and 1996 to 2000.

The GDP growth rate shows the expected relationship with pov-erty reduction. During the first sub-period, growth rate has been low; and HCR of poverty has increased. The GDP growth rate accelerated during the second sub-period, i.e., 1991 to 1996. During this period, head count ratio of moderate poverty declined at a rate of 1.4 percentage point per year. The decline of extreme poverty was faster 2.1 percentage point per annum.

During the second half of the nineties, GDP growth has increased. HCR further declined during this period, though the decline per year was smaller than 1991-96. The decline in HCR of moderate and extreme poverty per year was 0.23 and 0.98 percentage points respectively during this period. The percentage point of acceleration of GDP growth has been 0.79 points per year during the second half of the nineties, which was smaller than the acceleration of GDP growth rate during the previous sub-period. The crucial importance of GDP growth for poverty reduction is reflected in association with a smaller acceleration of GDP growth rate with a smaller rate of decline of poverty during this period. Nonetheless, growth appears to have been less pro poor during the second half of the nineties.

Desegregation of growth in sectoral form will give more promi-nent picture in analyzing linkage between poverty and economic growth.

(17)

Evidence from Bangladesh! ８９

From the various source it can be revealed that two main sectors that con-tribute to the Bangladesh economy are agricultural sector and manufac-turing/industrial sector. Growth in agricultural sector helps rural poor to fight against poverty. It is to note that rural poverty is directly connected with agriculture. Poverty is severely concentrated in rural area and bears major poverty concern of the government of Bangladesh. On the other hand urban poverty reduction is assumed to be the blessings of manufac-turing sector. This section will examine the relationship between the ag-ricultural growth and poverty reduction followed by a comparative analy-sis of industrial growth and urban poverty.

The following tables（5 and 6）do not show a direct relationship between poverty and growth, however, can give a fair picture when we compare total GDP growth with the sectoral GDP growth. In the previ-ous section it is demonstrated that poverty reduction and economic growth go with the same direction, if the sectoral analysis produces the same relationship, this methodological approach can be claimed to have justified that economic growth does contribute to poverty reduction.

Fluctuating sectoral growth rate during the period of 1986-2001

Year GDP growth rate in Agriculture Reduction of Moderate Poverty in Rural Area Reduction of Extreme Poverty in Rural Area Percent per annum Percentage point per annum

1986-1991 2.19 1.35↑ 1.67↑

1991-1996 1.50 1.13↓ 1.88↓

1996-2001 5.07 0.82↓ 0.98↓

Source: Bangladesh Bureau of Statistics, 2002, Rahman and Islam, 2003. Note: upward

directed arrow indicates ‘increase in poverty’ and downward directed arrow indicates ‘decrease in poverty’

(18)

９０ The Hikone Ronso No. 369 November 2007

depicts ambiguous link between growth and decline in poverty. During 1986-1991, GDP growth rate in agriculture and industry are 2.19（Table 5）and 5.06（Table 6）respectively which led to an increase in poverty both in rural and urban areas. However, growth during 1991-1996 in ag-ricultural sector is lower than the previous period but could address a de-cline in poverty but the growth in the same period in the industrial sector is higher than its previous decade and as expected, played a good role in the reduction of poverty. This can be explained in a way that natural ca-lamites in 1988 and 1998 affected both agricultural and industrial sectors. After the devastating flood in 1988 and also in 1998 a vigorous social safety net program had been undertaken with the huge assistance from in-ternational community which was assumed to have strong long term ef-fect on poverty reduction in successive years. As a result even with low economic growth poverty has also decreased in the agricultural sector.

There are other factors involved that results in inelastic relation-ship between sectoral growth and poverty. Inequality due to economic imperfection takes away major payoff from the industrial sector and thus it fails to benefit to urban poor. Government intervention during the time

Year GDP growth rate in Industry Moderate Poverty in Urban Area Extreme Poverty in Urban Area

Percent per annum Percentage point per annum

1986-1991 5.06 0.33↑ 0.57↑

1991-1996 7.81 2.53↓ 2.40↓

1996-2001 6.42 0.36↑ 0.71↓

Source: Bangladesh Bureau of Statistics, 2001, Rahman and Islam, 2003. Note: upward

directed arrow indicates ‘increase in poverty’ and downward directed arrow indicates ‘decrease in poverty’

(19)

Evidence from Bangladesh! ９１

of any disaster plays a role in determining this relationship. For example, when agriculture faces setback due to adverse condition government tries to ease credit and other services to accelerate growth of industry and other urban activities. Such activities, therefore, encourages movement labor force to urban area. So lower agricultural growth may not necessar-ily lead to a worsening of rural poverty. Similarly, when industry lags, the government makes more investment in agriculture and infrastructure building. Thus intervention may lead inverse pattern of growth in agri-culture and non-agriagri-culture sector and contribute to an indirect relation-ship. However, growth pattern in agricultural sector shows plausible re-lationship with respect to poverty decline. In this sector it is observable that high growth could result in poverty declination. For last three dec-ades this has been true specially in agricultural sector where most of the poor people concentrate in.

Economic growth is inevitable for countries development, it is in-evitable also for poverty reduction and environmental conservation―this is true even for Bangladesh. Sectoral contribution to economic growth is direct and evident but capital accumulation necessary for production or environmental friendly production, through foreign aid is doubtful and obscure. But in this paper we proved of were foreign assistance（aid）to generate capital accumulation were treated under good policy environ-ment, it would have worked more efficiently. Out of many other factors, aid plays a significant role in instigating growth and reducing poverty in Bangladesh because, benefit of aid directly goes to the poor. If aid is pro-poor and allocated in such a way that the poor can get maximum benefit from it, it leads to the economic growth（Collier and Dollar, 1999, 2002）. In that sense, Bangladesh being one of the poorest countries in the world, is eligible to receive aid and it is expected that aid would work

(20)

９２ The Hikone Ronso No. 369 November 2007

well. But our findings say that aid, particularly the environmental aid, works efficiently if it is treated under good policy environment. Manage-ment of aid and proper allocation of it within the country is important for the aid to work. Environmental aid component in our analysis include those aid which goes to the investment in clean technology in agriculture and environment along with other environmental preservation projects. So the sectoral contribution to economic growth, in other way, is the con-tribution of the effectiveness of environmental aid. Two-way effect of environmental aid on growth: one that directly effect growth through the accumulation of capital and the another one that effect through sectoral contribution, lift the poor out of the poverty.

7. Conclusion

We conclude that reduction of poverty is possible through eco-nomic growth and also through the preservation of environment, if Bang-ladesh is given more environmental aid than the generalized aid. Mere aid alone does not sufficiently provoke growth in Bangladesh. In our case study of Bangladesh, we found that environmental aid along with non-environmental, do not affect the growth unless it is well treated and well managed in light of good policy of the government, as we see the in-teraction term of ‘environmental aid × policy’ is highly significant and positive but coefficient of aid without policy is insignificant. The objec-tive of good policy, apart from other macroeconomic policy, is supposed to be that aid has to be targeted to environmental conservation and environmental-friendly infrastructure where poor can directly be bene-fited from. As we see that environmental aid, macroeconomic stability, competitive markets, and public investment in physical and social infra-structure are widely recognized important requirements for achieving

(21)

Evidence from Bangladesh! ９３

sustained economic growth, reduction in rural poverty and attainment Millennium Development Goal. Moreover allocation of aid must be ap-propriate and be ensured for the vulnerable poor and the targeted group. A country with good policy environment, economic growth is automati-cally entrusted with proper allocation of aid. Thus policy has to be given pivotal importance when sustained poverty reduction is demanded. Lastly, economic growth is one of the pertinent factors, not the sufficient factor responsible for poverty reduction. Foreign assistance is helpful but not the ultimate solution for any development. However, it plays a good role in alleviating poverty through reducing pressure on natural resources specially at the time immediately before economy takes off. Though this channel is not all but its influential impact stimulates the poverty reduc-tion process considerably. In addireduc-tion, the rural poor constitutes quarter of the country’s population, no development can be accomplished unless pro-poor government policies like access to land and credit, education and health care, support services, nutrition, etc. are ensured through well-designed public works programs and other transfer mechanisms.

――――――――― References

- Aghion, P. and Howitt, P.,（1992）, “A model of growth through creative destruc-tion”, Econometrica 60, Page 323-351.

- Andreoni, James and Arik Levinson,（2001）“The Simple Analytics of the Envi-ronmental Kuznets Curve”, Journal of Public Economics, Vol. 80, Page.269-286. - Arrow, K., and Kurz, M.,（1970）, “Public Investment, the Rate of Return and

Opti-mal Fiscal Policy”. The John Hopkins Press.

(22)

９４ The Hikone Ronso No. 369 November 2007

Economic Studies, Vol.29, Page 155-173.

- ‘Assessing Aid’,（1998）, World Bank Policy Research Report.

- Balke, N.S. and Slottje, D.J.,（1993），“Poverty and Changes in Macroeconomy: A Dynamic Macroeconometric Model”, The Review of Economics and Statistics, Vol 75, No.1, Page 117-222.

- “Bangladesh Economic Review”,（2002）, Ministry of Finance, Govt. of Bangla-desh.

- Barro, R.J.,（1990）. “Government spending in a simple model of endogenous growth”, Journal of Political Economy” Vol.98, Page 103-125.

- Bhattacharya, D.,（2001）, “Bangladesh Economy in FY 2000: Macro Economic Outlook in Changes and Challenges”, A Review of Bangladesh Development, CPD and UPL, Page 18-32.

- Bone, P.,（1996）, “Politics and Effectives of Froeign Aid”, European Economic Review, Vol.40, Page. 289-329.

- Burnside, C. and Dollar, D.,（2000）, “Aid, Policies and Growth”, American

Eco-nomic Review, Vol 90, Page 847-868.

- Collier, P., and Dollar, D.,（2002）, “Aid Allocation and Poverty Reduction”

Euro-pean Economic Review, Vol.46 Issue 8, Page 1475-2000.

- Collier, P. and Dollar, D.,（1999）, “Aid Allocation and Poverty Reduction”, Policy Research Working Paper 2041, Development Research Group, World Bank. - Dasgupta, P.,（1998）, “The Economies of Poverty in Poor countries”,

Scandina-vian Journal of Economics, Vol.100, Issue 1, Page 41-68.

- Flow of External Resources to Bangladesh,（2001）, Ministry of Finance, Govt. of Bangladesh.

- Grandmount, Jacques,（1986）, “Money and values” Cambridge University Press. - Greuber, A., Semmler, W. and Gong, G.,（2005）, “The forces of Economics

Growth: A Time Series Perspective”, Princeton University press, Princeton. - Greuber, A. and Semmler, W.,（1999）, “An Endogrowth model with public capital

and government borrowing”, Analysis of Operation Research.

(23)

Evidence from Bangladesh! ９５

Economy”, MIT press, Cambridge.

- Grossman and Krueger,（1995）“Economic Growth and Environment”, The Quar-terly Journal of Economics, Vol.110, Issue 2, Page. 353-377.

- Hansen, H. and Tarp, F.,（2000）, “Aid and Growth Regression.” Journal of

Devel-opment Economics, Volume 24.2, Page 375-398.

- Heckelman, J. and Knack, S.,（2005）, “Political Institutions and Market-Liberalizing Policy Reform”, World Bank.

- IIMI Atsushi and OJIMA Yasuhisa,（2005）, “Natural Resources, Economic Growth and Good Governance: An Empirical Note”, JBIC Institute.

- James, A. and Levinson, A.,（2001）, “The Simple Analytics of the Environmental Kuznets Curve”, Journal of Public Economics, Vol.80, Page 269-286.

- Klasen, S.,（2003）, “In Search of the Holy Grail: How to achieve Pro-Poor Growth”, Discussion paper no.96, Ibero-America Institute for Economic Research.

- Levine, R., and Renelt, D.,（1992）, “A Sensitivity Analysis of Cross-Country Growth Regrassions”, American Economic Review, Vol.82, Issue.4, Page 942-963. - Lucuas, R.E.,（1988）, “The Mechanism of Economic Development”, Journal of

Monetary Economics, Vol.22, Page 3-42.

- Mbaku, J.M.,（1993）, “Foreign Aid and Economic Growth in Cameroon” Applied

Economic. Vol.1, No.4, Page. 1309-1314.

- Mujeri, M.K.,（2000）, “Poverty trends and Agricultural Growth Linkage”, FMRSP working paper No.26, Government of Bangladesh, IFPRI.

- Mustafa K. Mujeri,（2002）, “Bangladesh: Bringing Poverty Focus in Rural Infra-structure Development”, Discussion Paper, ILO Geneva.

- Nurul Islam,（1992）, “Foreign Assistance and Economic Development: A Case of Pakistan” The Economic Journal, Page 502-530.

- OECD home page: http://www.oecd.org/dataoecd/56/21/1867292.gif

- Panayotou, T.,（1997）, “Demystifying the Environmental Kuznets Curve: Turning a Black Box into a Policy Tool”, Environment and Development Economics, Page 465-484.

(24)

Bangla-９６ The Hikone Ronso No. 369 November 2007

desh”, Discussion paper No.10, International Labor Organization, Geneva. - Ravallion, M. and Chen S.,（2003）“Measuring Pro-Poor Growth”, Economic

Let-ters, No.78（1）, Page. 93-99, 2003.

- Romar, M. and M.K., Gugerty,（1997）“Does Economic Growth Reduce Poverty?” CAER II Discussion paper no.5, Harvard Institute for International Development. - Selden, T. and D. Song.,（1994）, “Environmental Quality and Development: Is

there a Kuznets Curve for Air Pollution Emissions?” Journal of Environmental

Economics and Management. Vol.27, Page 147-162.

- Semmler, et. al.,（2005）, “The Forces of Economic Growth: A Time Serious Per-spective” Princeton University Press.

- Semmler, W.,（1999）, “An Endogeneous Growth Model with Public Capitl and Government Borrowing”, Analysis of Operation Research, Vol.88, Page 65-79. - Uzawa, H.,（1965）, “Optimum technical Change in an aggregative model of