4th Floor, Building 10, National Economics University, Tran Dai Nghia Rd. Hanoi, Vietnam - Tel: 84-4-8693211, Fax: 84-4-8693369        Sunday, Febuary 5, 2012 
 Welcome to MDE
 Overview
 Curriculum
 Facilities
 Organization
 Students
 Applications
 Fees, Loans, Scholarships
 Theses
 Seminars
 Services
 International Cooperation
 MDE News
 Alumni
 
Tiếng Việt | English
Keyword:

Advance search
Goto:
 
Chapter I

Chapter I

INTRODUCTION

1.1. BACKGROUND TO THE PROBLEM

The Millennium Development Goals are a set of goals which all countries in the world are trying to attain. These goals require implementation of a mix of policies, including in poverty degradation and environment protection. Forest protection is one of foremost targets in environmental policies in the world. In recent years, there has been increasing evidence of the inverse impacts of forest depletion on the environment such as global warming, climate change due to the release of carbon, loss of biodiversity, higher intensity and frequency of natural disasters, increased soil erosion, watershed degradation and consequent downstream effects. Therefore, forest protection has been receiving increased concern from the world community. In Vietnam, series of development policies including the Comprehensive Poverty Reduction and Growth Strategy and the Five Million Hectare Reforestation Program have been being carried out.

Despite the increasing demand for protection of natural forests, the intensity and scale of forest use have significantly increased in recent years, mainly in developing countries. According to the Food and Agriculture Organization of the United Nations (FAO, 1993), the loss of forest area of the developing region during period 1980-1990 amounts to 163 million ha, of which 154 million ha is in the tropics. Tropical forests are being lost at the rate of 10 million hectares annually, most of this due to shifting cultivation (Serageldin, 1993). National governments in these countries, especially in

Indonesia, the Philippines, Thailand, and Vietnam, tend to blame shifting cultivators, usually members of ethnic minorities, for rapid loss of forests (Le, 1996).

Viet Nam is not an exception in the common picture of the developing world. Under the Ministry of Agriculture and Rural Development (MARD, 2000), statistics from Vietnam’s Department of Forestry Development show that more than half of the country’s 19.08 million hectares of mapped forested land has now been identified as bare hills and wasteland. Some 60,000 to 70,000 hectares of the nation’s forests are destroyed every year. The World Bank (1995) estimates that the principal causes of deforestation in Vietnam are agricultural encroachment, firewood collection, logging, and fires in which agricultural encroachment, shifting cultivation and firewood are the main contributors in the Northern Mountain Region and the Central Highlands.

Vietnam remains one of the poorest countries in the world. With nearly 80% of its population living in rural areas, Vietnam is dependent on its agricultural and natural resources. The poverty rate is relatively high in upland, remote, and isolated areas and ethnic minority areas. According to

Do (1994), 50 of 54 ethnic groups in Vietnam practice shifting cultivation. He estimates that 7% of Tay, 16% of Nung, 45% of Thai and 100% of almost all the remaining ethnic minority groups are involved in slash-and-burn activities, a main cause leading to deforestation and forest degradation.

how the relationship between deforestation and poverty is?" is established. With an increase in income, the poor may prefer to invest in maintaining the quality of soil on existing cleared land, or increasing yields by expanding agricultural areas. Understanding how the poor make agricultural production and investment decisions will help policy makers find effective solutions not only to improve the living standard of poor people but also to slow the rate of deforestation.

1.2. REVIEWS OF RELATED STUDIES

There are a lot of papers studying the situation of deforestation or poverty in Vietnam, but only at the disaggregated level. A few of them are interested in causes of deforestation, and poverty was implied to be one of the main causes, directly or indirectly through shifting cultivation, increasing the rate of deforestation in Vietnam.

The Center for Natural Resources and Environmental Studies (1998) shows that the socio-economic root causes of biodiversity loss in the mountainous areas of Northern Vietnam consist primarily of high natural population growth rates, traditional agricultural and hunting practices, and the inefficiency of the APCs. This paper also showed that poverty in Vietnam is the root cause of a number of factors. As a rule, poor people are either landless or have been pushed to areas of low agricultural potential with steep slopes and infertile land. Poor farmers extract what they can from the environment to support themselves, and have little time or resources to invest in resource conservation or management.

According to Pham (1999), the violent spontaneous migration in recent years had caused a lot of socio-economic as well as environmental problems for Dak Lak province. The spontaneous migrants of which 30% belongs to more than 30 minority groups, due to the lack of close control, clear good quality forests freely and convert them to arable land. Most of them go deep into the forest to practice shifting cultivation, even into nature reserve areas, and protected and special use forests on the borders with neighbor provinces.

Similarly, Dang et al (2002) showed a bandwagon effect in coffee planting in the Central Highlands. Many farmers even planted coffee on very slopping hillsides or on the top of the hill where soil and water conditions were not suitable for coffee. As a result, the yield was very low and caused high soil erosion.

The papers and studies mentioned above proved that poverty is one of the main causes of deforestation. However, this relationship was only analyzed under the description aspect. Further, due to the lack of generality, a comprehensive picture of Vietnam has still been unclear and indistinct. Therefore, I choose the topic "The link between poverty and deforestation. The case of Vietnam" with the hope of overcoming these defects.

By focusing on the relationship between poverty and agricultural production and investment decisions, this paper also contributes to empirical microeconomic research on the behavior of rural smallholders in Vietnam.

1.3. FOCUS AND SCOPE OF THE THESIS
1.3.1. Focus

This thesis focuses on the key research question: "How is the relationship between poverty and deforestation?” In order to clarify this problem, some following sub-questions will be mentioned.

- What roles does forest have? What are causes of deforestation? How does deforestation affect sustainable development?

- What is poverty? How is poverty measured? What are causes of poverty? How does poverty affect sustainable development?

- Are the relationships between agricultural inputs and income or between cleared land and household size or borrowing constraints positive or negative?

- What should the Government do to decrease poverty but not to increase deforestation?

1.3.2.Scope

The thesis uses the information of poverty and deforestation in the period of 1993-1998 to run the model. Because of insufficient data and information, the thesis only examines the relationship between income (instead of poverty indicator) and cleared land (a main direct cause of deforestation). This study uses a sample of 299 households having forest. 

1.4. SOURCES OF INFORMATION AND METHODOLOGY
1.4.1. Sources of information
The study uses data from "Vietnam Living Standard Surveys" 1992-93 and 1997-98 that are collected by the General Statistics Office (GSO) with financial support from United Nations Development Programme (UNDP) and Swedish International Development Authority (SIDA) and technical assistance from the World Bank. This thesis also uses other information sources from GSO and from newspapers, reports and articles.

1.4.2. Methodology
The thesis applies poverty-deforestation hypothesis (strong hypothesis): increases in household income will be negatively correlated with land clearing and will be positively correlated with the use of inputs that increase or maintain yields. Investigating the relationship between poverty and the land-clearing decision requires taking the household as the unit of analysis.

1.5. LIMITATION OF DATA AND INFORMATION
The limitation of the data and information may prevent the study from better results. Firstly, household income calculated from VLSS 1992-93 reveals inaccuracy. Therefore, the missing or inaccurate data will be corrected from VLSS 1997-98. However, the inconsistent information of income in VLSS 1992-93 compared with VLSS 1997-98 may face some problems and this may affect the results of the study. Secondly, the information for shifting cultivation in VLSS does not exist. Using change in cleared land (including all types of land without forest) representing deforestation will partly limit the estimated results. Moreover, the information for types of land in both VLSS is not completely incompatible, as land could be sold or bought in the period 1993-1998.

Therefore, this data is just convenient for the purpose of identifying the sign between variables. In the future, the author would like to discuss this problem again when there are better data sets for analyzing.

1.6. STRUCTURE OF THE THESIS
The thesis includes 5 chapters: chapter I explains the reasons, objectives, methods and data of the thesis; chapter II goes into analyzing problems related to sustainable development, forest and deforestation, poverty and the link between poverty and deforestation; chapter III presents common deforestation and poverty situations in Vietnam; chapter IV describes data and illustrates the regression results in order to define the relationship between deforestation and income, household size or borrowing constraints, and income-agricultural input connections; and finally, in chapter V, conclusions and suggestions are mentioned.

Chapter II

                ANALYTICAL FRAMEWORK     

2.1 SUSTAINABLE DEVELOPMENT
2.1.1 Concept
According to the definition of World Commission on Environment and Development (WCED, 1987), 'sustainable development is the development that meets the needs of the present without compromising the ability of future generations to meet their own needs'.

2.1.2 Dimensions of sustainable development
The economic view of sustainable development

The ecological view of sustainable development

The socio-cultural view of sustainable development

2.1.3 Challenges to sustainable development in developing countries
Most developing countries are characterized by low per-capita income, high population growth rate and a dependence upon agricultural production. Therefore, their foremost challenge to sustainable development is rural poverty because the poor will continue to destroy their immediate environment to survive.

2.2 FORESTS AND DEFORESTATION
2.2.1 Role of forests
For ecological sustainability

For economic and social development

2.2.2. Causes of deforestation
Property rights

Population

Income

Environmental Policies

2.2.3. Effects of deforestation on sustainable development
a) Environmental impacts

b) Social impacts

c) Economic impacts

2.3 POVERTY
2.3.1. Measuring the poverty indicator
The World Bank's poverty lines

The World Bank (1995) establishes the international standard poverty line based on a minimum requirement of 2,100 calories per day per capita with the assumption that a household devotes 70% its total expenditure on food. The remaining 30% expenditure is assumed to buy other necessary items such as clothes, health, education and transportation.

GSO's poverty line

The GSO (1992) also relied on the method using food expenditure to compute poverty lines. According to GSO, a household is in poverty if its income is insufficient to buy rice for providing 2,100 calories per person per day. This measure makes no allowance for non-food expenses. So it creates a lower threshold than the one used by the World Bank.

MOLISA's poverty line

MOLISA established the rice equivalent poverty line by classifying a household as a hungry one if its monthly per capita rice equivalent income is less than 13 kg of rice. Those with a monthly per capita income that is less than 15 kg, 20 kg, and 25 kg of rice equivalent in mountainous areas, islands, plain, midland rural areas, and urban areas respectively are considered to be poor.

2.3.2 Causes of poverty
Household size and dependants

Gender disadvantage

Education

Isolation disadvantage

Lack of access to productive resources

Employment

Unforeseen shocks

2.3.3 Effects of poverty on sustainable development
Researches and policies have tended to focus on the relationship between poverty and sustainable development, represented by environment factor, in terms of pointing out that the poor are both victims and agents of environmental degradation: victims in that they are more likely to live in ecologically vulnerable areas, agents in that they may have no option but to deplete environmental resources, thus contributing to environmental degradation. However, it is also acknowledged that the poor often have actions that conserve the environment. Great physical and spatial variability in natural resource endowments also seem to complicate the picture (Rachel et la. 1997).

2.4. POVERTY AND DEFORESTATION - WHAT IS THE LINK?
2.4.1 The forest-dependent poor
2.4.2 Poverty - ambiguous impacts on forests

At the macroeconomic level, forest loss is positively related with both economic growth and poverty reduction.

At the microeconomic level, poverty tends to make cheap labor available, which makes it attractive both to cash in low-return forest rents (degradation) and to undertake labor-intensive investment in forest clearing (deforestation).

2.4.3. The poverty-deforestation hypothesis
Simple model

According to Alix Peterson Zwane (2002), the effect of poverty on deforestation is theoretically ambiguous and depends on the relative profitability of intensification (or sustainable agriculture) and clearing. Therefore, in the case of deforestation, environment and development policies must be implemented on the basis of empirical evidence rather than theory.

To find out how poverty may interact with deforestation, we consider the following scenario: Household A lives for 2 periods, P1 and P2, and begins with wealth W. The household cannot borrow, but can save at a certain rate of interest. All wealth is consumed at the end of P2.

The household has two equal-sized plots of forested land (j and k). The household can cultivate on these plots using one of two techniques, A and B. The respective production cost and yield are CA and CB, YA and YB. The cost of clearing land is C and is independent of the production technique used. Assume that YA > YB, and CA < CB, but using technique B allows a plot to be farmed more than one time. If technique A is used, the plot is degraded and must be abandoned.

Figure 2.1 - Household's land-use options

Opt 1

P1

P2

 

Opt 2

P1

P2

 

Opt 3

P1

P2

 

Opt 4

P1

P2

Plot j

A

 

 

Plot j

A

 

 

Plot j

B

A

 

Plot j

B

A

Plot k

F

A

 

Plot k

A

 

 

Plot k

F

A

 

Plot k

B

A

Source: Alix Peterson Zwane (2002)

Figure 2.1 shows the household’s options as well as the ambiguous relationship between land use and poverty. Forest land is denoted by the letter F and abandoned land is imagined by a blank cell. The letters A and B denote cultivated land using one of the possible techniques.

Among options, option 1 is clearly the least costly strategy in period 1. Lack of the ability to borrow, households with the lowest wealth can only pursue this strategy. As wealth increases, whether option 2 or option 3 is preferable depends on each household’s decision; option 4 is the choice that requires the largest investment. When option 4 is chosen, the most output is generated, but if environmental externalities were considered, this might not be the socially preferred outcome.

In this study, the so-called conventional wisdom on the relationship between poverty and deforestation will be used as a strong version of “poverty-deforestation hypothesis” (or written succinctly as “poverty-deforestation hypothesis”). That is: increase in household income will be negatively related with land clearing and will be positively related with the use of input that increase or maintain yields. The weak version of the poverty-deforestation hypothesis might predict that the link between income with both land clearing and input use will be positive, but income elasticity of land clearing will be smaller than the income elasticity of input use.

Now consider a policy that increases initial effective wealth for the poorest households, such as a micro-credit program, but not so much that option 4 becomes feasible. If the poverty- deforestation hypothesis holds, then as W increases, households selecting option 1 will switch to selecting option 3. This means that increasing wealth does not increase environmental degradation in period 1. If households selecting option 1 prefer to select option 2 as wealth increases, there is a short-run trade-off between deforestation and increase in income. Theoretically, there is no way of determining which outcome is most likely. The decision will depend on the relative yields of YA and YB, the relative production costs CA and CB, the cost of clearing land, and the interest rate. Household size and labor availability would also affect this decision if the possibility of an imperfect labor market were introduced.

Extended model

In this part, a more complete model of the household at the forest margin is investigated (Appendix A). The model's results show that the relationship between wealth and land use remains ambiguous. This model provides testable propositions as follows:

Proposition 1: The predicted relationship between income and the land-clearing decision is theoretically ambiguous and must be determined empirically.

Proposition 2: If imperfect labor markets exist, households with more labor per unit of land will be more likely to respond to increases in income by clearing land.

Chapter III

DEFORESTATION AND POVERTY SITUATION

IN VIETNAM

3.1. FORESTS AND DEFORESTATION IN VIETNAM
3.1.1. Forests in Vietnam
3.1.2 Ownership of Forest Land
3.1.3. Deforestation in Vietnam
Deforestation has been widespread for several decades, and remains a serious problem. As indicated by the table 3.1, there was a steady and rather rapid decline in natural forest cover between 1943 and 1990, especially during the war years of 1960-1971. Since 1990, the loss of natural forest has proceeded at a much slower pace - at an annual average of 36,000 hectares per year during 1990-1995. At the same time, the area of planted forest increased at an average annual rate of 61,000 hectares per year. In 1995-1999, areas of both natural and planted forest increased rapidly. However, the increase rate of forests in recent years partly is due to using different definition of forests. Forest quality continues to decline as a result of forest degradation, further impacting water discharge patterns and biological diversity.

Table 3.1 - Changes in the forest cover of Vietnam (1943-1999)

Unit: 1000 hectares

 

1943

1976

1980

1985

1990

1995

1999

Natural forest

14,000

11,077

10,486

9,308

8,430

8,252

9,444

Plantation

0

92

422

584

745

1,050

1,471

Total hectares

14,000

11,169

10,608

9,892

9,175

9,302

10,915

% of total area

43.0

33.8

32.1

30.0

27.2

28.1

33.2

Source: MARD (2000)

3.1.4. Shifting cultivation - a main cause of deforestation in Vietnam
Shifting cultivators in Vietnam

Ethnic minority - shifting cultivation link in Vietnam

3.2. POVERTY SITUATION IN VIETNAM

Firstly, Vietnam is still one of the poorest countries in the world. In 1999, Vietnam’s gross national product per capita was US$370 compared with US$1020 for the Philippines, US$580 for Indonesia, US$780 for China and US$410 for low income countries (figure 3.3)

Secondly, a majority of the poor is concentrated in remote, isolated, and mountainous areas. As many as 64% of the poor live in the Northern mountainous region (Northwest and Northeast), North Central region, Central Highlands, and Central coastal region (figure 3.4). These areas are characterized by difficult living conditions, geographically isolated, very limited access to productive resources and services, underdeveloped infrastructure, harsh natural conditions, and high frequency of natural disasters.

Thirdly, poverty is a widespread phenomenon in rural areas - Rural per capita incomes in some parts of Vietnam are below US$ 100/year. Over 90% of the poor live in rural areas. There is also wide variation across rural regions. The headcount index ranges from 45% in the South East to 79% in the North Central region. Therefore a higher proportion of the population lives below the poverty line in rural areas than in urban areas - 45% compared with 9% (figure 3.5).

Fourthly, the poverty rate is extremely high among ethnic minority groups. While accounting for roughly 14% of the total national population, the representation of ethnic minority groups among the poor is disproportionately high at roughly 29%. The majority  of  ethnic  minority  people  live  in  remote  and  isolated  areas. They are geographically and culturally isolated, and lack favorable conditions for developing infrastructure and basic social services. Ethnic minority people are amongst the poorest in Vietnam. Ethnic minorities make up 14% of the population but account for 29% of poor people in Vietnam. The incidence of poverty among ethnic minorities has come down from 86% in 1993 to 75% in 1998. This compares to the poverty rate for the Kinh majority of 31% down from 54%. Therefore, although ethnic minority poverty is declining, it is falling at a slower rate than for the Kinh population and remains very high (Figure 3.7).

3.3. RELATED POLICIES

3.3.1 Forest land allocation

3.3.2 Programme 327

3.3.3 The campaign for fixed cultivation and sedentarisation

3.3.4 Credit policy for the poor

Chapter IV

does poverty exacerbate deforestation?

empirical evidence from vietnam

4.1. data

This thesis uses two data sets derived from VLSS 1 and VLSS 2 (conducted in two periods of 1992-93 and 1997-98). These data sets are undertaken by General Statistical Office (GSO) of the Government of Vietnam under the LSMS form. They include socio-economic aspects of a large number of households that are highly representative of Vietnamese households. Notably, only 4186 households interviewed in 1993 had been selected in 1997 because some households were sampled again and the other had moved elsewhere. This figure implies that the study on migration might be ignored.

While these surveys contain national data, only 400 households using or controlling forest land in VLSS1 will be considered here. It should be noted that selling or buying land between two period might affect the analysis results and therefore, only 229 households are used in this sample.

4.2 MODEL AND DATA SPECIFICATION

4.3. GENERAL DESCRIPTION

Table 4.1 shows that the average income of households increased in the period 1993-1998 (from 8,326,600 VND to 10,020,600 VND for a year). However, the figure is low compared to the national average income of households (14,010,740 VND for a year). As calculated in Table 4.2, two poorest quintiles in this sample account for 51.96%. Meanwhile, the percentage rate of these quintiles for the whole country is only 37.38%. Thus, the number of poor is concentrated largely in selected regions.


 

Table 4.1 - Summary statistics of variables, 1993 and 1998

Variable

Unit

Mean

Std. Dev

LAND93

m2

10338

8640.1

LAND98

m2

9212.8

8623.8

DEFOR

m2

380.9

5456.5

INCOME93

000 dong

8326.6

10838.9

INCOME98

000 dong

10020.6

8314.9

DURABLE93

000 dong

269.7

526.5

DURABLE98

000 dong

526.8

797.8

AGE93

age

4.16

1.49

AGE98

age

4.33

1.34

ADULT93

persons

5.11

1.74

ADULT98

persons

4.90

1.69

PESEXP93

000 dong

34.93

64.3

PESEXP98

000 dong

139.9

321.8

FEREXP93

000 dong

386.1

383.7

FEREXP98

000 dong

935.5

995.5

LABEXP93

000 dong

78.3

226.8

LABEXP98

000 dong

33.1

227.7

PESUSE93

-

1.28

.449

PESUSE98

-

1.12

.333

FERUSE93

-

1.02

.160

FERUSE98

-

1.00

.093

LABUSE93

-

1.77

.422

LABUSE98

-

1.72

.447

EDU93

-

1.40

.850

EDU98

-

1.60

.855

ETHNIC

-

0.272

0.445

Source: Calculated from 229 observations in the sample based on data from VLSS 92-93 and VLSS 93-98


 

Table 4.2 - Allocating quintiles for the whole country and for the sample

Quintile 1993

For the whole country

For the sample

Frequency

Percentage

Frequency

Percentage

Quintile 1

873

18.19

60

26.20

Quintile 2

921

19.19

59

25.76

Quintile 3

944

19.67

51

22.27

Quintile 4

993

20.69

35

15.28

Quintile 5

1068

22.25

24

10.48

Total

4799

100.00

299

100.00

Source: Author's calculations based on data of VLSS 1992-1993

Non-labor and non-farm income make up a small percentage rate of total income (about 14%) in 1993 but increase to nearly 30% of total income in 1998 (Table 4.3). This suggests that despite of reducing, the vulnerability of household to shocks in agriculture such as pest infestations, bad weather is still large. Such shocks may be mitigated if households can borrow to smooth consumption, but if borrowing potential depends on households' asset positions, the small asset stocks that most households hold suggest that this may be only a limited cushion.

Table 4.3 Non-labor and non-farm income, 1993 and 1998

Year

1993 (000d)

1998 (000d)

Non-labor income

358.8

2668.7  

Non-farm income

811.3

1596.2

% of total income

14%

29%

Source: Author's calculations based on the choosen sample

According to studying results of Do (2001), borrowers are more likely to borrow if they have collateral (in the form of land and/or durable goods). In Vietnam, the combined use of chemical and organic fertilizers has a positive effect on plants, in terms of growth and yield. Some factors, however, reduce the effectiveness of fertilizers. Since the government does not control the importation of fertilizer, there are frequent fluctuations in the price. Such inputs and shocks often need to be financed prior to a harvest, which may be difficult for households with a limited ability to borrow.

Another important issue is also shown in Table 4.1. That is the households in this sample practice high-input agriculture, (especially fertilizer factor – nearly 100% farmers using fertilizer) but hire little outside labor (about 23% households using hired labor). This can be explained by the high level in average available labors (a household has about 5 adults).

Table 4.4 Classifying households clearing land by region in 1993-1998

Region

Mountainous region

Delta region

Coastal region

% households

clearing land

61.36

79.16

88.88

Cleared land area

(m2/per household)

2271.1

1530.7

2250.7

           Source: Author's calculations based on the choosen sample

Also in table 4.1, the rather large positive value (381 m2 per household) of the variable DEFOR (change in cleared land) shows that there happens deforestation in households. Forest can be cleared for cultivation, aquaculture, and other uses. Over time, each household either increase or decrease their holdings of cleared land. Because the sampling does not consider land purchases and salse, increases in cleared land holdings are a result of deforestation.

Table 4.5 Classifying households clearing land by ethnic groups in 1993-1998

Region

Kinh and Chinese

Minority

% households clearing land

88.73

50.50

Cleared land area

(m2/per household)

1792.3

2344.7

          Source: Author's calculations based on the choosen sample

If only concerning about households that really deforest, the cleared forest area will be about 1935 m2 per household in average, ranging from 10 m2 to maximum value of 2,5 hectare, concentrating mainly on mountainous and coastal regions (Table 4.4 in Appendix). 88.7% Kinh or Chinese households attend to clear forest land while this rate in minority households is only 50%, however the more serious rate happens in minority households (2344 m2 compared with 1792 m2) (Table 4.5 in Appendix).

4.4 REGRESSION RESULTS

4.4.1 The land clearing - income relationship

The estimated parameters in the above table indicate that almost all coefficients are statistically significant at 10%. The coefficients of income, durable assets, land, number of adults have positive effects on deforestation.

Table 4.6 - The land-clearing decision

Number of observations: 299

Dependent variable: DEFOR

Variables

Coefficients

t

p-value

LAND93

0.324

11.4

0.000

INCOME93

0.197

1.72

0.086

INCOME932*

0.080

2.12

0.034

DURABLE93

1.375

4.83

0.000

AGE93

-266.4

-2.72

0.007

ADULT93

431.1

3.56

0.000

REGION

177.52

039

0.695

ETHNIC

-1947.9

-2.66

0.000

R square

0.167

Adjusted R square

0.162

F statistic

33.46

Probability > F

0.000

*INCOME932 = INCOME93 ^ 2; the coefficient of INCOME932 is multiplied by 10,000

From the table, we show that there exists a significant relation between cleared land and income. Land-clearing increases with the larger rate when income is high. 1000 dong increase in income causes 0.197m2 forest land to be cleared. Thus, the relation between deforestation and income satisfies the "weak" version of hypothesis.

The number of adults in a household strongly affect on land clearing. Its coefficient is very large and positive. It shows that the larger the number of adults, the higher level of land clearing. This result also expresses that land clearing occurs more in Kinh and Chinese households. This is a strong and expected result.

Also according to the results in Table 4.3, households with larger land holdings or durable assets clear more land. Households with large asset holdings are able to borrow more from different sources than those with fewer assets. The positive sign of the coefficient on asset holdings indicates that households who are eligible to borrow, to buy capital inputs in agriculture for example, choose to clear more land than the others. This will be discussed in more details later in part 4.4.3.

4.4.2. The input expenditure - income relationship

The above results are evidence in favor of the contention that an increase in income is not followed by a decrease in cleared land in Vietnam. Although this is not predicted by the poverty-deforestation hypothesis, this hypothesis cannot be rejected without a further investigate about the relation between the use of inputs that increase yields on previously cleared land and income. If analysis shows that high incomes are correlated with both additional input use and land clearing, this would be consistent with a weak version of the poverty-deforestation hypothesis. On the other hand, evidence that high incomes are uncorrelated with the purchase of yield-increasing or yield-maintaining inputs would be further evidence against the hypothesis.

Table 4.7 presents results from regressions that are similar to those presented in the previous part. In this case, however, the dependent variable is no longer land clearing, but expenditure on purchased inputs, including fertilizer, pesticide, and hired labor respectively, per hectare of cleared land. The results of these regressions suggest that income is positively correlated with the additional use of inputs that increase yields. This is evidence supporting the weak version of the poverty-deforestation hypothesis.

Table 4.7 - The input-use decision

Number of observations: 299

Variables

Dependent variables

 

p-value

FEREXP98

PESEXP98

LABEXP98

 

FEREXP98

PESEXP98

LABEXP98

LAND92

0.014

0.0012

0.0008

 

0.078

0.476

0.434

INCOME92

0.370

0.036

0.0079

 

0.000

0.000

0.067

INCOME922*

-0.100

-0.0072

-0.0019

 

0.000

0.002

0.176

DURABLE92

0.075

0.100

-0.010

 

0.406

0.000

0.334

AGE92

97.00

23.00

12.00

 

0.002

0.000

0.002

ADULT92

-31.00

-4.800

-5.200

 

0.412

0.522

0.255

REGION

-11.00

-89.00

-30.00

 

0.937

0.001

0.068

ETHNIC

-770.0

-89.00

-9.000

 

0.000

0.004

0.628

R square

0.5025

0.3393

0.0347

Adj R square

0.4995

0.3353

0.0289

F statistic

168.34

85.59

6.00

Probability > F

0.0000

0.0000

0.0000

*The coefficients of income922 are multiplied by 10,000

In the case of pesticide expenditures, there is a positive and significant correlation between asset holdings and pesticide expenditures. This correlation may simply reflect the fact that households using pesticide are more likely to grow permanent tree crops such as coffee and tea. These tree crops are an asset that households may add to the estimated value of their total assets, resulting in a positive correlation between asset holdings and pesticide use.


 

4.4.3 The credit market

The direct effect of borrowing constraints is positive, that is, more constrained households clear land. The interaction effect of borrowing constraints and family size is also positive and significant. This positive sign on the interaction term means that, while more constrained households clear land, the effect is intensified for large households

This is further evidence that the land-clearing decisions of large and small households differ. The signs of these coefficients are consistent with the proposition derived from the model that large households are most like to respond to increases in income, or reduced borrowing constraints, by clearing land because these households have surplus home labor that they price below the local wage rate.

Table 4.8 - The effect of borrowing constraints on the land-clearing decision

Number of observations: 299

Dependent variable: defor

Variables

Coefficients

t-value

p-value

BORROW

1409

1.940

0.053

BORROW*ADULT

705

2.650

0.008

ADULT92

500

4.870

0.000

LAND92

-0.368

-14.710

0.000

AGE92

-319

-2.510

0.012

REGION

4973

6.110

0.000

ETHNIC

1409

1.940

0.053

R square

0.1511

Adjusted R square

0.1466

F statistic

33.41

Probability > F

0.000

Turning to the effect of borrowing constraints on the use of inputs, in Table 4.9, results of probit estimates for the fertilizer- and pesticide-use decisions are presented. The results of this estimation show that bigger families are less likely to invest in pesticide but more likely to invest in fertilizer to intensify agricultural production. There is very clear relationship between fertilizer use and borrowing constraints.

Table 4.9 - Effect of borrowing constraints on input use

Number of observations: 299

Variables

Dependent variables

 

p-value

FERUSE98

PESUSE98

 

FERUSE98

PESUSE98

BORROW

0.970

0.876

 

0.000

0.000

ADULT92

0.018

0.027

 

0.000

0.000

LAND92

0.000

0.000

 

0.073

0.000

DURABLE92

-0.023

-0.033

 

0.000

0.000

AGE92

0.002

-0.058

 

0.890

0.008

ADULT92

-0.044

-0.149

 

0.034

0.000

REGION

-0.017

-0.080

 

0.178

0.001

R square

0.9655

0.8574

Adjusted R square

0.9654

0.8567

F statistic

6011.83

1290.07

Probability > F

0.0000

0.0000

 

By using a direct measure of the household’s borrowing constraint, rather than a measure of income, the evidence in this section provides additional support for the claim that the simple version of the poverty-deforestation hypothesis can be accepted for this sample. This part provides evidence that the relationship between poverty and land clearing depends on household size, as predicted by proposition 2. These results also support the contention that larger households, are more likely to invest in better technology.

Chapter V

CONCLUSIONS AND RECOMMENDATIONS

CONCLUSION

This thesis analyzes the interaction between poverty and land-use change in Vietnam. The benchmark is that reductions in poverty will be related with less land clearing, as households are able to make desired investments in intensification on previously cleared plots. In contrast to this so-called conventional wisdom, chapter 2 shows that while households do take expected lifetime borrowing constraints into consideration when making land-clearing decisions, the relationship between poverty and deforestation is theoretically ambiguous.

The findings of this thesis suggest that for households with relatively high incomes, the "weak" version of poverty deforestation hypothesis may indeed be an accurate characterization of households’ preferences.

There are some limitations in the process of studying. First, the information are incomplete, therefore some variables had to be changed by substitutes. Second, the two data set, VLSS 92/93 and VLSS 97/98 are inconsistent, leading to adjustments have been made. Third, the information about forest is less accurate due to problems in statistical methods.

RECOMMENDATIONS

Agro forestry and fixed cultivation policies

Credit supply

Agriculture technology development

Consciousness education

Employment generation

Providing the incentives for farmers to sustainably manage forests

 
<< BACK
   Visitors : 326731  
Copyright© Vietnam-Netherlands Master in Development Economics (MDE) Program.  Contact Us
uggs for cheap uggs for sale cheap uggs cheap uggs uggs for cheap uggs outlet uggs boots sale cheap uggs cheap ugg boots cheap uggs