You are on page 1of 14

Agricultural Systems 60 (1999) 99±112

www.elsevier.com/locate/agsy

Environmental±economic decision-making in lowland


irrigated agriculture using multi-criteria analysis techniques
D.N. Tiwari a,*, R. Loof b, G.N. Paudyal c
a
PO Box 12609, Kathmandu, Nepal
b
School of Civil Engineering, Asian Institute of Technology, Bangkok, Thailand
c
Danish Hydraulic Institute, Gulshan Road, Dhaka, Bangladesh

Received 30 January 1998; received in revised form 20 November 1998; accepted 26 February 1999

Abstract
The continuing debate on sustainability has raised wide concerns towards integration of environmental and eco-
nomic aspects into the development decision-making process. This paper develops a framework for environmental±
economic decision making that includes the environmental and economic sustainability criteria, and local people's
preferences in the context of a lowland irrigated agriculture system using multi-criteria decision-making techniques.
Several criteria, such as land capability/suitability, energy input/output ratio, water demand and environmental costs,
are considered as environmental sustainability criteria. Economic sustainability is measured from farmers', govern-
ments and societal viewpoints using extended cost-bene®t analysis. The Geographic Information System (GIS) tech-
nique has been used to evaluate spatial sustainability criteria. The involvement of local people at various levels of the
decision- making process is emphasized and their opinions are sought in the decision-making process using a two-stage
®eld survey. The results of the multi-criteria analysis combining both environmental and economic sustainability
criteria are discussed, and economic incentives for sustainable intensi®cation of lowland irrigated agriculture are out-
lined. # 1999 Elsevier Science Ltd. All rights reserved.
Keywords: Economic sustainability; Economic incentives; Environmental±economic decision making; Environmental sustainability;
Multi-criteria analysis

1. Introduction there has been some progress in developing eco-


nomic valuation methods and addressing sustain-
The continuing debate on sustainability raised ability concerns into the CBA (Barbier et al., 1989;
wide concerns towards integration of environ- Bojo et al., 1992; Pearce, 1994; Tiwari, 1996).
mental and economic aspects into the decision- However, reality is complex, and the use of CBA
making process. Decision-making techniques at alone may not be sucient when the decision
the project level have long been dominated by the situation involves consideration of variables which
conventional cost-bene®t analysis (CBA), where cannot be easily quanti®ed into monetary units,
only the quanti®ed direct costs and bene®ts are and the decision-making process is likely to be
incorporated into the analysis. In recent years in¯uenced by multiple-competing criteria. The
complexity involved in the economic valuation,
* Corresponding author. the time dimension, diversity of environmental
0308-521X/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0308-521X(99)00021-9
100 D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112

and natural resource scarcity e€ects, and local countries due to lack of information (El Serafy,
people's concerns, etc., sometimes present pro- 1991). Likewise, the extent of comprehensiveness
blems for monetization of all the criteria, espe- of the method to be adopted also depends upon
cially in the context of developing countries. the nature of local speci®c environmental issues,
Opinions may diverge in the long-term, as more and sustainability concerns to be addressed while
e€orts can be expected towards economic valua- making decisions at the project level.
tion, incorporation of environmental externalities Various studies in the past have used MCDM
into the economic calculus, internalization of these techniques in making environmental±economic
costs and bene®ts, adoption of clean technologies, decisions. For example, Hafkamp and Nijkamp
etc., ultimately may help in arriving at a `win- (1986) used MCDM techniques for integrated
win' situation. In the short-run, however, where economic±environment±energy policy analysis.
environmental±economic decision-making pro- The multi-regional model was developed con-
cesses are constrained by lack of information for structing mutually interactive three parallel lay-
monetization of all the variables and criteria, and ersÐeconomic, employment and environmentÐ
the technological change is also dicult to take using compromising programming (CP) techni-
place, some compromising solution has to be ques. Zekri and Romero (1993) used the MCDM
sought. The decision-making techniques, in such a approach to ®nd a best compromising solution
context, should be redirected towards formulation combining various public and private concerns,
of alternatives which meet various sustainability such as net present value (NVP), employment,
criteria and help to arrive at the best compromis- water consumption and energy use, in agriculture.
ing solution for making a transition towards Likewise, Laxminarayanan et al. (1995) used a
sustainable use of natural resources. In such a multi-objective modeling approach for making
situation, multi-criteria decision-making (MCDM) decisions involving trade-o€ between soil erosion
techniques are particularly helpful in taking into and water quality. Although these various e€orts
account the outcomes of monetary valuation, made in the past have attempted to incorporate
ecological analysis, and local people's perceptions both the economic and environmental dimensions,
as well as decision-makers' viewpoints in the it is equally important that people's perceptions
decision-making process. are also considered while making decisions. While
developing a framework for environmental±eco-
nomic decision making, these various criteria and
2. Theoretical basis for environmental±economic concerns have to be taken into consideration. In
decision making this paper, the use of two widely popular MCDM
techniquesÐCP and analytical hierarchical pro-
Various MCDM techniques have been devel- cess (AHP)Ðis demonstrated by presenting a case
oped and applied in the past for regional planning, study of the Phitsanulok irrigation project (PIP)
agricultural land and water management pur- located in the Northern Plains of Thailand.
poses. Recently, the use of MCDM techniques
has been emphasized for application in environ- 2.1. CP
mental±economic decision making where all the
objectives, or criteria cannot be quanti®ed in CP is a distance-based technique designed to
monetary units (Munasinghe, 1992). The choice of identify non-dominated solutions which are clo-
a particular method for environmental±economic sest to an ideal solution using a quasi distance
decision making, however, is guided by a trade-o€ measure (Zeleny, 1982). Policies for environ-
between comprehensiveness and objectivity (Jans- mental±economic integration are often concerned
sen, 1991). Although a more comprehensive with removing threats to the environment, and
approach is needed for environmental±economic trying to reach a certain desired state of environ-
decision making, a fully comprehensive approach ment (e.g. within the land capability, or suitability
is dicult to apply in the context of developing limits) rather than maximizing the overall results
D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112 101

of the alternatives at a time. In a review of the 2.2. AHP


recent developments in decision-making techni-
ques, Hippel (1992) concluded that application of The AHP is a decision-aided method which de-
MCDM techniques provides a sound basis for composes a complex multi-factor problem into a
solving water resources management problems to hierarchy, and each level is composed of speci®c
satisfy the goals of sustainable development. For elements. It uses hierarchic structures, matrices
application of a MCDM technique such as CP in and linear algebra to formalise the decision pro-
environmental±economic decision making, a set of cesses (Saaty, 1980). The AHP method determines
criteria can be developed and used for arriving at the priorities of each alternative with the assigned
the `best compromising' solution for making the weight for each alternative by analysing the judg-
transition towards sustainable use of resources. mental matrices using the advanced mathematical
In CP, the ideal point or the solution is de®ned theory of eigenvalues and eigenvectors. It inter-
as the best score on each criterion within a given prets the eigenvector associated with the largest
set of criteria which serves as a reference point, eigenvalue as the priorities that indicate the
because, in real practice, it is almost impossible to importance of each alternative in accomplishing
have an ideal solution. In this context, the alter- the objective. AHP combines both subjective and
native that is a `closest' to the ideal point becomes objective judgements in an integrated framework
the highly preferred one. This closeness to the based on ratio scales from simple pairwise com-
ideal point is determined using a family of stan- parisons. Saaty (1980) provided a theoretical
dardized Lp matrices and is mathematically ex- foundation for AHP. Since then the use of AHP
pressed as: techniques has been made in di€erent sectors for
analysing complex environmental and economic
X
N   problems (Golden et al., 1989; Saaty, 1993;
p 1=p
Lp …Aj † ˆ … i …… f i ÿ fij †=… f i ÿ f 
i †† † ; …1†
Alphonce, 1997). The use of AHP stems from its
iˆ1
simplicity, ¯exibility and its ability to incorporate
decision-makers' perceptions as well. The theore-
where Lp (Aj) is the distance matrix, which is a tical framework for application of AHP can be
function of the decision alternative Aj, and the found in Saaty (1980) and Schoner and Wedley
parameter p; i is the standardized form of the cri- (1989).
terion weight wi and:
X
w
3. Study area
i ˆ w i = wi ;
iˆ1
The study area is located in the Northern Plains
of Thailand, about 500 km north of Bangkok. The
f i is the ideal or best value for criterion i; and f 
i irrigation system known as the Phitsanulok irri-
is the minimum or worst value for criterion i. gation project (PIP), consists of a main canal of
The scaling co-ecient in the ®rst equation, p, 176 km. The study area, which is a sub-project of
which is a parameter, makes it possible to include a the PIP, is one of the largest irrigation systems in
relationship between relative size of the e€ect and the country. The irrigation system consists of a
weight into the decision rule (Janssen, 1991). If main canal, 96.0 km from the headwork to the end
there is a relationship such as the balancing of the of this sub-project (here after referred as the pro-
criteria, p must be 1; if there is decreasing marginal ject), and is divided into 39 sub-zones for the pur-
utility, then p must be higher than 1. The value of p, pose of water management. The total area was
in general, depends upon policy objectives. When divided into three zonesÐhead, middle and tail.
only the highest value is important, p must be in- The ®eld survey was carried out in May 1994,
®nitive. For each of these cases, the objective immediately after the rice harvest. The total sample
function equation is reduced to a di€erent form. population surveyed consisted of 209 households
102 D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112

out of a total 12,000 households. About 86% of componentsÐenvironmental, economic and spa-
the population surveyed were entirely dependent tial database management. The environmen-
upon agriculture, the remaining 14% having some tal component is divided into two partsÐland
family members employed in government jobs, capability/land suitability analysis and resource
private companies and daily wage activities. By accounting. The resource accounting sub-compo-
landholding size, the majority of farmers (55%) nent further consists of two partsÐestimation of
had 1.6 and 3.8 ha, 27% above 3.8 ha and the rest energy input/output ratio and the environmental
18% with less than 1.6 ha per family. The level of costs. The irrigation water supply and demand is
income generated from agricultural activities also considered as part of this sub-component.
alone showed the likely impact of the project on The economic component is divided into two
raising the living standard of people. However, partsÐeconomic valuation and CBAÐconsider-
based on the family income, and an average family ing the economic value of irrigation water from
size of 4.6, about 32% of the farmers in the project farmer's, societal and government viewpoints.
area were estimated to be living below the basic Both the ®eld survey and the application of a
needs level. Geographic Information System (GIS) formed
the core part of spatial database management.
Finally, multi-criteria techniques are used for
4. Conceptual framework making decisions among the selection of alter-
natives governed by these various environmental,
Fig 1 shows the overall conceptual framework economic and societal criteria. The next section
for environmental±economic decision making at provides methodological steps developed for
the project level in the context of lowland irrigated environmental±economic decision making using
agriculture. The framework consists of three main multi-criteria framework.

Fig. 1. Conceptual framework for environmental±economic decision making using multi-criteria analysis.
D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112 103

5. Procedural steps in environmental±economic and can be outlined as: (1) maintenance of the
decision making resource base such as soil quality; (2) low depen-
dence on external inputs and lower waste genera-
5.1. De®ning sustainability criteria and setting tions; (3) economic viability; and (4) local farmers'
objectives acceptability. While the conventional approach
such as land capability and suitability analysis is
De®ning sustainability criteria is the ®rst basic considered as a governing criteria for maintaining
step towards developing an analytical framework the resource base over the long-run, energy input/
for environmental±economic decision making. output ratio and NPV with consideration of the
Lowland irrigated areas have always been a major environmental cost from farmers' government and
concern for sustainable intensi®cation of agri- societal perspectives are considered as environ-
culture. Concerns over maintaining food security mental and economic indicators re¯ecting criteria
as well as reducing pressure on the upland fragile (2) and (3). Local people's preferences and their
ecosystem in developing countries is closely rela- choices among various available alternatives are
ted to the resource use patterns and productivity considered as farmers' acceptability criteria.
of the lowland ecosystems. On the other hand,
increasing water resource scarcity and resource use 5.1.1. Land capability/suitability criteria
patterns, and externalities associated with the In the case of lowland irrigated agriculture, this
lowland irrigated agriculture, etc., have raised criterion is related to the maintenance of the soil
serious questions on the capability of these eco- resource base and agricultural productivity. How-
systems to sustain long-term productivity and ever, it is rather dicult to measure the soil nutri-
provide food security to the ever increasing popu- ent balance and its productivity over time, and
lation. Natural resource and environmental scar- provide its value in economic terms. Nevertheless,
city issues, such as water scarcity, increased use of spatial sustainability analysis combining land
agrochemicals for maintaining farm productivity, capability and suitability provides a sound basis as
increasing waterlogging and soil salinity, surface governing criteria for maintaining the soil resource
and ground water pollution, and increased health productivity over the long-run, and integration of
e€ects of pesticides, are drawing wide attention. ecological sustainability criteria into the CBA.
De®ning sustainability criteria in operational This involves selection of cropping patterns by
terms is the ®rst basic step towards integrating assigning and ranking threshold values related to
these wide concerns into the multi-criteria analy- di€erent soil characteristics and topographic fac-
sis. Although there is some disagreement about tors as a set of constraints a€ecting the soil quality
operational de®nition of sustainability, some and agricultural productivity. The land capability
speci®c criteria can be set out as to what it means and suitability analysis based on soil and drainage
at the project level for incorporating various characteristics provide an optimum limit for
environmental and societal concerns. From the change in cropping patterns during the dry season.
standpoint of the environment (or ecological sus- The objective is, thus, to maximize the total crop-
tainability criterion), less reliance on dangerous ping area based on land suitability analysis and
forms of agrochemical (fertilizers and pesticides) the selection criteria are governed by the max-
and diverse crop rotations are preferred than ones imization of the total area.
that are less diverse (OECD, 1995). Moreover,
de®nition of sustainable agriculture is widened 5.1.2. Energy output/input ratio
further as environmentally non-degrading, eco- One main concern for achieving sustainability is
nomically viable and socially acceptable (FAO, to increase the resource use eciency. Various
1991). Developing sustainability criteria at the studies in the past have shown that the modern
project level, thus, varies from de®ning the con- agricultural system is highly inecient. The max-
cept as maintenance of resource productivity over imization of resource use eciency as measured by
time to a socially acceptable agricultural system, the energy output/input ratio for each particular
104 D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112

crop provides a basis for selection of the most costs. Governments and donor agencies are more
ecient area allocation for increasing resource use concerned with the cost-recovery of the project,
eciency. and this approach of consideration of water price
equal to that of the cost-recovery amount is refer-
5.1.3. Water requirements red to in this paper as carried out from the ``gov-
The selection of the mix of crops that utilizes the ernment viewpoint'', which follows the normal
minimum water in the dry season is the main procedure of CBA. Due to growing scarcity of
concern for solving the water scarcity problem. water, the objective of the Royal Irrigation
The objective is, thus, to minimize the total water Department of Thailand has been to minimize the
requirement while trying to satisfy the farmers' rice plantation in the project area in order to save
demand for cultivation of various crops. the water during the dry season. Another objective
has been to raise and collect the water fees for
5.1.4. Environmental costs meeting the project costs. The objective is then, to
Both the on-site and o€-site environmental costs maximize the NPV, in this case government's
of lowland irrigated agriculture, such as crop NPV, computed assuming the economic value of
damages due to waterlogging and ¯oods, potential irrigation water equal to that of the cost-recovery
surface and ground water pollution due to the amount, and with incorporatation of environ-
excessive use of pesticides and chemical fertilizers, mental costs.
and on-farm costs to the farmers, such as crop
damages by insect attacks, etc., have to be mini- 5.1.7. NPV from farmers' viewpoints
mized for a sustainable agriculture system. The From the farmers' viewpoint, they are supposed
objective is, thus, to minimize these environmental to receive an opportunity to cultivate during the
costs. dry season according to their household needs and
preferences, and that they would not prefer to pay
5.1.5. Economic viability with considerations of water fees exceeding their willingness to pay. In
government, farmers' and societal viewpoints this case, the economic value of water is assumed
Economic viability refers to the ability of the equal to that of farmers' willingness to pay, and
lowland irrigated agricultural system to retain an the estimation of NPV, in this case, is considered
acceptable level of pro®tability even when external as from ``farmers' viewpoint''. Likewise, while
costs, such as environmental damage costs and calculating NPV from the farmers' perspective,
resource scarcity prices, are internalized, or incor- only the on-farm crop damage costs, directly
porated into the economic analysis. To achieve accrued to the farmers are considered. The objec-
economic sustainability, the net economic bene®t tive as in the last case is the maximization of
derived should at least be constant over the pro- NPV computed considering these various factors
ject life period, and it should generate surplus for outlined.
investing in maintaining the resource productivity.
The objective is, thus, to maximize the NPV. As 5.1.8. NPV from societal viewpoints
outlined in Section 2, the economic viability cri- From intergenerational viewpoint, however,
teria are de®ned from three di€erent perspectives, scarcity value of irrigation water will have to be
government, farmers' and societal viewpoints, and considered. The NPV in this case is estimated
each of them are brie¯y de®ned in the context of assuming water price equal to the marginal value
this paper. product, or the scarcity rent of water and is con-
sidered as carried out from ``societal viewpoints''.
5.1.6. NPV from government viewpoints In addition, environmental costs are also incorpo-
While carrying out CBA of an irrigation project, rated as in the case of government viewpoints.
usually the economic price of irrigation water is From societal viewpoints, both of these factors,
considered equal to that of the cost-recovery environmental costs and scarcity value of water,
amount including operation and maintenance are highly important. While the objective is to
D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112 105

minimize these costs, the set out objective is, then, to know in determining the economic value of
the maximization of societal NPV. irrigation water, environmental costs, and de-
The computation of NPV from three di€erent signing alternative cropping patterns. Local
perspectives as the governing economic criteria, farmers' preferences, on the other hand, are
however, should not be confused with the normal determined by their household needs. The deci-
CBA procedures which per de®nition is a type sion is carried out by a household to meet these
of analysis that is made from the view of society. needs, and can be elicited using both the direct
For example, if a project is found optimal from (willingness to pay) and indirect (ability to pay)
the view of society, it is not all-optimal for all the economic valuation techniques. Besides, measur-
individuals that are a€ected by the project. In ing and using farmers' willingness to pay for irri-
such a case, paying them compensation for the gation water, local people's preferences are
loss they su€er can provide a solution; how- incorporated while developing alternative crop-
ever, as Morgenstern (1998) argues in his paper, ping practices and in the decision-making process
the compensation mechanism is never truly using a structured questionnaire for pairwise
designed and applied in practice. The CBA pro- comparisons of di€erent sustainability criteria
cess, in this case, only di€ers in the computation and sub-criteria.
of NPV while including the resource price or the
value of water, as farmers' willingness to pay for 5.2. Formulation of alternatives
irrigation water at present may not truly re¯ect
the scarcity value of water as farmers still con- Ten alternatives (Table 1) were formulated for
sider water, as a free gift of nature in develop- helping to solve the water scarcity problem and
ing countries. Second, conventional CBA, which making the transition to sustainable agriculture
considers resource price only, based on the cost- through crop diversi®cation based on land cap-
recovery considerations, may be di€erent when ability/suitability analysis, water availability, local
the scarcity value of such resources is considered. farmers' perceptions and including existing and
These three cases of computation of NPV using worst case scenarios as well for dry season crop-
three di€erent values of water cost-recovery, ping in the project area.
farmers' willingness to pay and scarcity value of
water, termed as government, farmers' and socie- 5.3. Integrating output from spatial analysis using
tal NPV, respectively, still follow the normal GIS, resource accounting and economic valuation
CBA procedures. The computation of NPV in for multi-criteria evaluation
these three cases represent three sub-criteria of
economic viability. Multi-criteria evaluation techniques involve
quanti®cation of several criteria outlined in step
5.1.9. Local farmers' acceptability one. The output generated from spatial analysis
Another major concern of agriculture sustain- using GIS provides a basis for di€erent physical
ability is the local farmers' acceptability. Local environmental criteria such as land capability/
farmers' preferences play a very signi®cant role suitability, energy output/input ratio, and envir-
both in the design and implementation of alter- onmental costs, and design of di€erent alter-
natives aimed at sustainable use of resources. natives on a cropwise basis. The economic value
Alternative methods designed without local of irrigation water obtained from di€erent direct
farmers' acceptance, even though environmen- (willingness to pay) and indirect (marginal value
tally sound, but may not be easy to implement. product, marginal construction costs, scarcity
Implementation of market-based instruments, rent, etc.) methods (Tiwari, 1998), was incorpo-
such as water charging system for ecient use of rated into the extended CBA for estimating the
irrigation water, may also be dicult without NPV from di€erent viewpoints. The output
knowing local people's willingness to pay for obtained for di€erent criteria under di€erent
it. Thus, local people's preferences are important alternatives is shown in Table 1.
106 D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112

A1, continuation of existing cropping patterns; A2, priority for rice cultivation; A3, priority for non-rice crops in highly suitable areas; A4, no cultivation in low-
ÿ665.92
ÿ410.03

resource areas; A5, cropping patterns according to farmers preference; A6, direct reduction in water supply (area with distance <4 km from main canal); A7, direct
reduction in water supply (area with distance >4 km from main canal); A8, creation of ecological bu€er zone along Yom River bank; A9, the best case: rice cultivation
5.4. Decomposition of sustainability criteria on a

493.38
hierarchical basis
A10

0
0
0
0

ÿ280.82
Application of multi-criteria techniques such as
166.19
994.73

AHP for incorporating local people's and decision


34 765

28.68
355
A9

4.8

makers' perceptions involve setting each of the


indicators/criteriaÐeconomic, social and envir-
onmentalÐinto several descending levels of hier-
ÿ107.61

ÿ69.32
432.59
29 795

10.41

archy. Schoner and Wedley (1989) pointed out


213
A8

5.5

that the elements (sub-indicators) under each node


(main indicator group) should be less than nine
ÿ297.92

items, and items should be homogenous as well as


ÿ40.78
ÿ219.7
13 320

6.49

within one order of magnitude. If they are hetero-


108
A7

5.2

geneous or di€erent by more than one order of


magnitude, then they should again be broken
30.571
346.08
20 965

95.36

down into sub-groups of similar nature and mag-


6.39
144
A6

5.6

nitude.
As shown in Fig. 2, the main goal is to achieve
607.14
684.89
1175.9
34 757

10.89

the sustainability of the agricultural system in the


239
A5

5.6

area. The sustainability goal is then broken in-


to two criteriaÐenvironmental and economicÐ
109.19
27 734

11.92
175.4
652.5

which are again divided into seven sub-criteria.


in all areas; A10, the worst case: no cultivation during dry season; NPV, net present value.
213
A4

5.3

The environmental criterion is divided into four


sub-criteria, and economic sustainability criteria
1051.09

into three sub-criteria as de®ned in step one. The


682.57
645.63
33 775

5.98
205

seven alternatives designed for dry season crop-


A3

6.1

ping are included, and each of the alternatives is


Pay-o€ matrix: sustainability criteria versus alternative cropping patternsa

linked to each sub-criterion.


417.71
516.92
1056.4
34 765

12.56
252
A2

5.5

5.5. Development of pay-o€ matrix


Alternatives

ÿ206.28
66.849
27 190

All the elements of the pay-o€ matrix (criteria


21.47

705.3
271
A1

4.8

versus alternatives) are shown in Table 1 The


selection criterion as outlined in step one is either
maximization (of bene®ts) or minimization (of
MCUM
MBaht
MBaht
MBaht
MBaht
Ratio
Units

costs).
Ha

5.6. Assigning weights to each criterion


Maximize
Maximize

Maximize
Maximize
Maximize
Minimize
Minimize
Objective

In any MCDM analysis, the relative weight of


particular criterion plays an important role. The
weight is assigned according to the relative
Energy output/input
Water requirements

importance of the criterion included for analysis.


Environmental cost

Government NPV

Zeleny (1982) describes two kinds of methods for


Land suitability

Farmers' NPV

Societal NPV

assigning weights. The ®rst is the subjective


weight, which is a relatively stable concept of a
Criteria
Table 1

priori criteria importance re¯ecting an individ-


a

ual's cultural, genetic, psychcological, social and


D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112 107

Fig. 2. Hierarchy of sustainability criteria.

environmental background. The second is the assigned giving priority to the economic, soci-
objective weight which is a relatively unstable, etal and environmental criteria at a time. The
context development concept of informational vector of weights assigned in each case is shown in
importance based on a particular set of objectives Table 2.
and a given decision situation. As Zeleny (1982)
further pointed out, these weights are sensitive to 5.7. Development of questionnaires for decision
the changes in both the feasible set of alternatives makers and local people
and criteria, and thus to ¯uctuations in average
information generated by both. This indicates that In the case of the multi-layer analysis, the
weights are to be assigned according to the context quantitative weights are derived from qualitative
of the problem. Although decision makers' pre- statements on the relative importance of the cri-
ference over each criterion compared to another teria obtained from comparisons of criteria by the
would give a better picture of the weight, they can respondents. Saaty (1980) proposed a nine-point
be judged by assigning di€erential weights from scale to express di€erences in their relative impor-
both the government and farmers' viewpoints. tance. The decision maker is asked to compare
In the case of the CP, ®rst, equal weights all pairs of criteria in a nine-point scale, and a
are assigned to all criteria. Then weights are pairwise comparison matrix is then developed
108 D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112

Table 2
Vectors of maximum or best criterion values, minimum or worst values and four sets of criterion weightsa

Criteria f i f 
i Criterion weight sets

w1 w2 w3 w4

Land suitability 34 765 0 0.1250 0.0909 0.2000 0.0769


Energy output/input 6.1 0 0.1250 0.0909 0.2000 0.0769
Water requirements 108 0 0.1250 0.2727 0.2000 0.0769
Environmental cost 5.98 28.68 0.1250 0.0909 0.2000 0.0769
Farmers' NPV 684.88 ÿ665.9 0.1250 0.0909 0.0667 0.2308
Government NPV 1175.92 ÿ410 0.1250 0.0909 0.0667 0.2308
Societal NPV 682.57 ÿ280.8 0.1250 0.2727 0.0667 0.2308
a
w1, Equal weights to all sustainability criteria; w2, high priority to water savings and societal NPV; w3, high priority to environ-
mental criteria; w4, high priority to economic sustainability criteria; NVP, net present value.

according to the results. The ®rst step in deter- 5.9. Computation of consistency index
mining weights is to design a set of questions for
interview with the local farmers' and the project The absolute ranking values for each pair com-
ocials. A set of questionnaires was designed to parison obtained from the survey was fed to the
know the farmers' and irrigation department o- Expert Choice (EC) package for determining the
cials' judgements on pairwise comparisons of sus- local and global priority weights. First, the rank-
tainability criteria and sub-criteria. ings provided by the local people and project o-
cials for each pairwaise comparison were entered
5.8. Administering the questionnaires into the speci®ed format for analysis. Second, the
level of consistency was measured by computing
5.8.1. Interview with farmer leaders the consistency index in each case. The composite
Eighteen farmer leaders were selected for inter- weights can be obtained by taking a geometric
view from di€erent zones from the head, middle mean of the values for each respondent. The value
and tail regions of the system on a random basis. of each alternative under di€erent sub-criteria
They were told about the objectives of the ques- were then entered and the synthesis report was
tionnaires, sustainability criteria and sub-criteria derived in each case. The results are presented in
and on how they can participate in the decision- the next section.
making processes for cropping pattern selection
during the dry season. They were ®rst explained in
group, and interview was taken on individual basis 6. Results and analysis
after they were more or less clear of the purpose
of the interview. Farmers' answers were again 6.1. Decision making using CP
rechecked by asking what they meant by a parti-
cular answer and were found satisfactory for To solve the multi-criterion problem using
inclusion in the analysis. CP, the ®rst task is to calculate the vectors of the
ideal point valuesÐthe best and the worst values.
5.8.2. Interview with project ocials The ®rst two columns in Table 2 show the esti-
Sixteen government ocials at di€erent levels of mated values for each criterion. Table 2 also
hierarchy from the Regional Irrigation Depart- shows the relative weights assigned to each criter-
ment and the Project Oce were asked to express ion assuming that it represents the decision
their concerns on the ranking of di€erent sustain- makers' preference among several criteria at a
ability criteria and sub-criteria. time.
D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112 109

6.1.1. Compromising solution based on equal 2. Savings in energy input (agrochemical and
weights for all criteria farm power) is more important than water
The results obtained in the case of equal weights savings. If they wanted to save water their
assigned to all criteria showed that alternative A3 farm will need more pesticide in order to kill
is the most preferred alternative management plan excess weeds that would grow due to less
among the 10 alternatives for all values of p. availability of water. The cost of pesticide
Conversely, the worst case scenario A10 is found applications in such a case will be more than
to be the least preferred alternative for all values what could be gained by saving the irrigation
of p. water.
3. Farmers' demand for water is also con-
6.1.2. Compromising solution with more priority to strained by the available suitable land, and
water saving and societal NPV pricing of water, if implemented, will play
In this case, the solution is more heterogeneous only a partial role in in¯uencing demand.
and unstable. Alternative A5 was found to be the Farmers' choice for a particular criterion
best preferred alternative when p=1 and p=1. depends upon the type of land they already
Alternative A10 appeared highly favoured at p=6 own and cultivate.
and p=10. 4. For some farmers, land suitability and water
savings are equally important as they think
6.1.3. Compromising solution with more priority to that by scan also be achieved.
all environmental sustainability criteria 5. Compared to large size landholders, small
In this case, A3 and A6 were found to be the size landholders showed more care about
best preferred alternatives at p=1 and 2, and p=6 societal and government NPV. Their view-
and A10, respectively. At p=1, alternative A8 point was that rich people can a€ord to pay
was found to be the best alternative. for environmental cost, but poor people
cannot a€ord such losses and also cannot
6.1.4. Compromising solution with higher priorities pay for it. If the farmers were charged for
to all economic sustainability criteria irrigation water and for excessive use of
In this case, a more stable solution was obtained agrochemicals on a proportional basis, poor
and alternative A3 was the best preferred alter- people will get bene®t from the government.
native except for p=1. For p=1, A5 was found This expression also provides a concept of
the best alternative. compensation from social pricing mechanism
through internalization of external costs.
6.2. Incorporating local peoples perception: AHP 6. Some farmers told that a question regard-
analysis ing saving irrigation water should not be
brought, as they were not receiving irrigation
6.2.1. Local people's observations water.
During the ®eld survey, farmers were also asked
to give their reasoning for the particular choice of 6.2.2. Testing inconsistency index and determ-
pairwise comparisons of di€erent sustainability ination of local and global priority weights
criteria and sub-criteria. From the analysis it was In the case of farmers, the three sets interviewed
found that their preferences were supported by were discarded, because of a high inconsistency
their long farming experiences which are summed ratio. In four cases, although the consistency ratio
up as follows: was found slightly higher than 0.1, they were
found within consistency limits when synthesized
1. Cropping patterns designed and adopted with respect to the goal. The composite weights
according to land suitability conditions are were obtained by taking geometric means of each
more important to them as they will help value obtained from each of the respondents. The
save both water and energy inputs. overall inconsistency index was found higher in
110 D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112

the case of government respondents. In both cases Figs. 5 and 6 show the synthesis of results for
of farmers' and government ocials rankings, each of the alternatives designed from farmers'
highest priority was given to land suitability sub- and project ocials viewpoints. The overall
criteria followed by water savings and NPV esti- inconsistency index was improved in both cases.
mated from farmers' viewpoints. The priority for The results indicated that alternative A5 was the
other criteria varied in both cases. The details are best alternative from the farmers' viewpoint, and
given in Figs. 3 and 4. alternative A2 and A3 were the best one from
project ocials' viewpoints.
6.2.3. Analysis of the results In both cases, environmental sustainability rec-
The results indicated high importance to the eived almost double priority compared to eco-
environmental sustainability criteria compared to nomic sustainability. While in the farmer's case, it
the economic sustainability in both cases of pair- was obtained as 0.673 and 0.327, in the govern-
wise rankings by farmers and ocials. Among ment ocial's case it was obtained as 0.712 and
the environmental sub-criteria, both the local and 0.288. The results indicated that compared to
global priority weights were higher in the case of farmers, government ocials showed more con-
land suitability compared to other sub-criteria. cern for environmental criteria. Likewise, among
Likewise, in the case of economic sub-criteria,
farmers' viewpoints received highest global and
local priority weights compared to governments or
societal sub-criteria.

Fig. 5. Analytical hierarchical process priorities for alternatives


(farmers' viewpoint). A3, priority for upland crops in highly
and moderately suitable areas; A2, priority for rice cultivation
in highly and moderately suitable areas; A5, area allocation
according to farmers' preferences; A7, rice cultivation in all
areas; A4, transfer of low-resource areas for other uses; A1,
Fig. 3. Analytical hierarchical process priorities for sustain- continuation of present cropping practices; A6, reduction in
ability criteria (farmers' viewpoint). water supply based on distance from the main canal.

Fig. 4. Analytical hierarchical process priorities for sustainability criteria (ocials' viewpoint).
D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112 111

best alternative at p=6 and 10. This was mainly


due to reduction in water based on the distance
from the main canal.
In the case of AHP priority setting, while alter-
native A3 was found the most preferred by farm-
ers, alternative A5 was found to be the optimal
when viewed from project ocials' perspectives.
In the latter case, the result was the same as
obtained from the CP. However, the result varied
in other cases. This shows that CP based on expert
Fig. 6. Analytical hierarchical process priorities for alternatives judgement may provide a di€erent outcome com-
(ocials' viewpoint). A5, area allocation according to farmers' pared to the decision analysis with incorporation
preferences; A3, priority for upland crops in highly and mod- of local people's preferences. This further indicates
erately suitable areas; A2, priority for rice cultivation in highly
that incorporation of local people's opinion is
and moderately suitable areas; A7, rice cultivation in all areas;
A4, transfer of low-resource areas for other uses; A1, con-
quite important in the environmental±economic
tinuation of present cropping practices; A6, reduction in water decision-making process and should be considered
supply based on distance from the main canal. at di€erent levels of the decision hierarchy.

the environmental sub-criteria, the land suitability


sub-criterion was given the highest priority by 7. Conclusions: implications for research and policy
government ocials. The government NPV and
societal NPV received lowest priority from the The environmental±economic decision-making
farmers and the government ocials, respectively. situation involves a complex decision-making
process and usually requires a more comprehen-
6.3. Comparison of results obtained from CP and sive framework to arrive at the optimal solution.
AHP This paper provides a comprehensive analytical
framework for incorporating sustainability con-
CP basically derives solutions with the best cerns into the environmental±economic decision-
trade-o€ among the criteria with di€erent weights making process using multi-criteria evaluation
assigned to each criterion. As seen from the above techniques. Use of the GIS technique facilitated
results, a more or less stable solution was found spatial analysis and design of alternatives based on
for the ®rst best alternative A5. This shows that if the land suitability criteria and ecological bu€er
water resource scarcity and environmental aspects zoning concepts. In addition, this paper also
are considered, cultivation of non-rice crops will incorporated farmers' perceptions as well as their
generate a better output than rice cultivation in preferences into the decision-making process using
the dry season. a two-stage ®eld survey. These aspects can be
In other cases, however, the result signi®cantly considered as a major step forward in environ-
varied. The alternative A10 was found to be the mental±economic decision making at the project
best one for p=6 and 10 when more weight was level in the context of developing countries.
assigned to the water saving and societal NPV. Water scarcity has been a national problem in
Both criteria favored water saving and charging Thailand. In this context, e€orts on promoting
irrigation water on scarcity rent basis. In this case, non-rice crops have been considered as the best
it was assumed that farmers might be willing to way towards solving this crisis and achieving the
sell their share of water at a price equal to their objective of sustainable intensi®cation of agricult-
willingness to pay rather than cultivating rice ure in lowland irrigated areas. Both the growing
during the dry season. Likewise, when more water resource scarcity and environmental costs
priority was given to environmental sustainability of excessive use of agrochemicals indicate that
criteria, alternative A6 was found to be the ®rst the project evaluation criteria should be shifted
112 D.N. Tiwari et al. / Agricultural Systems 60 (1999) 99±112

towards evaluating these costs and incorporation Golden, B.L., Wasil, E.A., Levy, D.E., 1989. Applications of
of these concerns into the decision-making pro- the analytical hierarchy process: a categorised annoted bibi-
liography. In: Golden, B.L., Wasil, E.A., Harker, P.T.
cess.
(Eds.), The Analytic Hierarchial Process: Application and
The results obtained using MCDM techniques Studies. Springer, New York, pp. 37±58.
provided useful guidance for selecting optimum Hafkamp, W., Nijkamp, P., 1968. Integrated economic±envir-
cropping patterns taking into consideration the onmental±energy policy and con¯ict analysis. Journal of
economic and environmental sustainability criteria Policy Modelling 8, 551±576.
Hippel, K.H., 1992. Multiple objective decision making in
and local people's perceptions. Unlike traditional
water resources. Water Resources Bulletin 28, 3±10.
emphasis on rice cultivation as a means of enhan- Janssen, R., 1991. Multi-objective decision support for envir-
cing incremental bene®ts from irrigated agri- onmental problems. Free University, Amsterdam.
culture during the dry season, the result of the Laxminarayanan, P.G., Johnson, S.R., Bouzaher, A., 1995. A
study showed that making a shift to non-rice crops multi-objective approach to integrating agriculture and
environmental policies. Journal of Environmental Manage-
will result in more pro®t to farmers and less bur-
ment 45, 365±378.
den to society which can be considered a pre- Munasinghe, M., 1992. Towards Sustainable Development:
requisite for making the transition towards sus- The Role of Environmental Economics and Valuation (The
tainable agriculture. But achieving the goal of World Bank Environment Paper Number 3). The World
sustainable intensi®cation of agriculture requires Bank, Washington, DC.
OECD, 1995. Sustainable Agriculture: Concepts, Issues and
various incentive mechanisms as considered while
Policies in OECD Countries. Organisation of Economic C0-
carrying out multi-criteria evaluation; these operation and Development, Paris.
include: (1) pricing of irrigation water or charging Pearce, D.W., 1994. Valuing the environment: past practice,
it at various levels according to the ®eld condi- future prospects in valuing the environment. In: Serageldin, I.,
tions; (2) minimizing existing and potential envir- Steer, A. (Eds.), Environmentally Sustainable Development
Proceedings Series No. 2. The World Bank, Washington,
onmental costs by taxing agrochemicals; and (3)
DC.
facilitating transfer of water rights to allow farmers Saaty, T.L., 1980. The Analytical Hierarchy Processes.
for making right decisions, especially during the McGraw Hill, New York.
water scarcity period. Design and implementation Saaty, T.L., 1993. Transport planning with multiple criteria:
of these market-based measures will no doubt the analytic hierarchial process applications and progress
review. Journal of Advanced Transportation 29, 81±126.
require institutional re-arrangements and further Schoner, B., Wedley, W.C., 1989. Ambiguous criteria weights
considerations of political economy factors for in AHP: consequences and solutions. Decision Sciences 20,
successful implementation of these mechanisms. 462±475.
Tiwari, D.N., 1998. Determining Economic Value of Irrigation
Water: Comparision of Willingness to Pay and Indirect
References Valuation Approaches as a Measure of Sustainable Resource
Use (CSERGE Working Paper GEC 98-05). University
Alphonce, C.B., 1997. Application of the analytic hierarchy College, London.
process in agriculture in developing countries. Agricultural Tiwari, D.N., 1996. Sustainability criteria and cost-bene®t
Systems 53, 97±112. analysis at the project level. In: Proceedings of International
Barbier, E.B., Pearce, D.W., Markandaya, A., 1989. Envir- Conference on Integrating EIA and Cost-bene®t Analysis.
onmental Sustainability and Cost-bene®t Analysis, Environ- University of Bradford, Bradford, UK.
mental Planning and Management A, Vol. 22 (pp. 1259± Zekri, S., Romero, C., 1993. Public and private compromises in
1266). agricultural water management. Journal of Environmental
Bojo, J., Maler, K., Unemo, L., 1992. Environment and Management 37, 281±290.
Development: an Economic Approach. Kluwer Academic, Zeleny, M., 1982. Multiple Criteria Decision Making. McGraw-
Amsterdam. Hill, New York.
FAO, 1991. FAO denBotsch Conference Report, Food and El Serafy, 1991.
Agriculture Organisation of the United Nations, Rome. Morgenstem, 1998.

You might also like