Relative Efficiency Evaluation on Water Resource Utilization

Sep.2011Joumal of Northeast Agricultural University (English Edition)Vol.18 No.3 60-64Relative Efficiency Evaluation on Water Resource UilizationMA YingSchool of Agriculture and Forestry Economics and Management, Lanzhou University of Finance and Economics, Lanzhou 730020, ChinaAbstract: W ater resource allocation was defined as an input-output question in this paper, and a preliminary input-output indexsystem was set up. Then GEM (group eigenvalue method)-MAUE (muliattribute utility theory) model was applied to evaluaterelative efficiency of water resource allocation plans. This model determined weights of indicators by GEM, and assessed theallocation schemes by MAUE. Compared with DEA (Data Envelopment Analysis) or ANN (Artificial Neural Networks), the modelwas more applicable in some cases where decision-makers had preference for certain indicators.Key words: water resource allocation, group eigenvalue method, mult-attribute utility theory, evaluationCLC number: X322Document code: A Article ID: 1006-8104(201 1)-03-0060-05and output indicators. Indexes should be chosen in con-Introductionformity to two rules: reasonableness and practicability.The number of indexes should be concise; otherwise,As the development of economy, water resourcesthe model would be too complicated to lose the valuebecome more and more scarce. Therefore, the ques-of practicability.tion that how to optimize the allocation of finitewater resources attracts extensive attention, but theInput indicatorspast related studies often just highlighted the benefitsThere are few researches on input indicators ofand production effects on allocation plans withoutwater resources distribution. At present, area waterpaying enough attention to allocation costs. In thisresource allocation behaviors can be divided intopaper, water resource allocation was defined asfour categories according to its inner mechanism:an input-output question, and a preliminary input-configuration by verge costs price, distribution byoutput index system was set up. Then GEM (groupadministrative management, configuration by watereigenvalue method)-MAUE (multi-attribute utilitymarket and allocation by users' demand, so theretheory) model was applied to assess water resourceare main four kinds of water resource configurationallocation plans. This model determined weights ofresolutions: market configuration, administrativeindicators by GEM, and assessed the allocationconfiguration, users participating configuration andschemes by MAUE.synthesis configuration". No matter which allocationmodel it is, each configuration resolution has its ownEvaluation Indexesconfiguration cost. How to reckon the allocation costsreasonably is very important to the allocation plans'Assessment indexes are composed of input indicatorsassessment, so further studies on it should be done.中国煤化工Received 6 April 2010MA Ying (1981-). male, lecturer, Master, engaged in the research of population and envirMYHc N M H G2046@yahoo com.cnE-mail: xuebaenglish@neau.edu.cn.MA Ying. Relative fficiency Evalution on Water Resource Uilization.61.Here, distribution costs are divided into three sub-assessment model sententious.indicators: EC, EEC, and UC.Environmental and ecological costs (EEC)Economic costs (EC)EEC is the environmental and ecological sacrificeW ater resource allocation economic costs are the coststhat corresponds to configuration resolutions. Therewhich can be measured and expressed by currencyare many causes which can induce EEC, and it candirectly during water resource configuration process.appear in many different forms, which also include theThe constitution of EC varies according to distributionenvironmental and ecological losses of areas losingmethods. Except for the costs of water resourcewater discussed above. In the early 1990, Colby BGtransfer, transport, store, etc, there are two kinds ofwho studied the western water market of Americacosts which are easy to be ignored:yet talked about the environmental and ecological(1) The costs which are produced during the processexternal diseconomy produced by water right transferof introducing water rights market, such as informationfrom agriculture to non-agricultural usages', andcollecting fee, institution design fee, correspondingthis external diseconomy was a form of EEC. Therelaws, regulations' design and approval fee, water rightshad been researches on the quantification of EEC');initial allocation fee, litigation fee caused by con-here EEC was monetized, and was presented by thetradictions during the allocation process, the main-currency equivalent.tenance and supervision fee of water rights market.Uncertain costs (UC)Past assessment studies of the efficiency of waterBecause water resource allocation processe is sorights market configuration often just emphasized thatcomplicated and uncertain, there will be many randomwater market can make the verge benefit of water userscosts which can't be expected beforehand in processescome to the same level so as to maximize the socialof real configuration assessments; therefore, UCtotal welfare, but ignored the relative institutionalindicator is used to describe them.transfer costs. Laura McCann had pointed out thattransaction costs of introducing water right market onOutput indicatorsthe base of administrative allocation was decided byAccording to the principles of reasonability andthe local water resources' own characteristics, relatedpracticability talked above, roundly considering theinstitutions, more comprehensive social institutionallarge scale system of the human-ecology and environ-environment and interactions among them' .ment-economy involved in water distribution pro-(2) In the past, evaluations of water resourcescesses, an output indicator system like this (Fig. 1)transfer from one area to another area, researcherswas set up. The indicator c5 (the degree of satisfactionoften underlined the social, economic, environmentalfor water resource allocation) was used to expressand ecological benefits obtained by the area acceptingthe fairness and social agreement of water resourcewater, but didn't pay enough attention to the sideconfigurations.effects for the area losing water. So when the measureof water transfer is taken, direct and indirect economicGEM-MAUElosses of the area losing water resources should becredited to EC; its environmental and ecologicalThese two models, DEA (data envelopment analysis)sacrifice should be credited to the next indicator EEC.and ANN (artificial neural networks), needn't fixMeanwhile, there are some chance costs, time costsindicators' weights, and just constitute the productionand so on caused by judicial litigations and democraticfrontier by processing indicator values then to evaluate中国煤化工participations during the configuration process whichthe relative eompared withalso should be credited to EC in order to make the them, GEM-NYH_CNMHGerwasmorehttp: /publish.neau.edu.cn.62.Journal of Northeast Agricultural University (English Edition)Vol.18 No.3 2011applicable in some cases where decision- makers possible to develop area economy; but for developedhave special preference for indicators. For example,areas or some areas where the tourism of naturalin the inchoate period of economy development,landscapes is regarded as economic polar, they willon the premise of assuring that the ecological andpay more attention to the environmental and ecologicalenvironmental status wouldn't degenerate seriously,qualities. This kind of preference can be expressed bythe decision-makers often use as much water asweights of indicators.Outputindicatorsystem _EconomicEnvironmentalienefitand ccologicalSocial benefitindicatorsRatio of the waterDomestics Area GDPamount used byRate ofrivers'Degree ofContributionproduction grosswater qualitysatisfaction foratio of一下 increase rateenvironment andper ton of 5多reachingwater resourcescientifc(c2)ecology to the totalwater(el)standards (c4)allocation (c5)progress (c6)Fig. 1 Output indexes systemideal expert's scoring vector can make S a, becomeGroup eigenvalue method (GEM)minimum. After fixing the indexes weights vector,Group eigenvalue method abbreviated as GEM is MAUT model can be used to assess the relativeproposed by Professor Qiul0!. This method fixes theefficiency of plans in the following.indicators' weights by constituting experts' judgmentmatrix, which makes the decision process more simple Multi-attribute utility theory (MAUT)and convenient. It includes two steps:Multi-indicator decision making, also known as multi-(I) Constructing experts' scoring matrix.atribute decision making, is an important part ofLet each one of the expert groups score all the in-multi-objective decision to build the theory, and isdicators directly so as to form an mxn scoring matrixmainly used to rank decision making schemes with(integer m is the number of experts; and integer n ismultiple attribute-indicators. According to MAUT,the number of indicators):the result value of every attribute has some utilitiesX=(xy)mwnfor decision makers. So the result value of everyHere, x。is the ith expert's scoring value for the jthattribute can be converted into dimensionless utilityindicator.value through some utility function relationships. As(2) Calculating the indicators' weights.utility values are dimensionless, different attributes'Make the matrix F=x"X, and then figure out theutilities can be synthesized into an integrated utilityeigenvector X according to the largest eigenvalue ofvalue, which makes the result value of muliattributematrix F. At last, normalize the eigenvector to formquantified completely; then the best plan can bethe indicators' weights vector which is also calledchosen in accordance with the quantification value."the ideal expert's scoring vector". Marking theFisistosetupaThe ic*中国媒化工i8intersection angle between a n-dimensional vector and"super targTYHCNMHG: U(p,P>the ith expert's scoring vector as a(i=l, ., m), the Pn) is the心，刚P2,一，P, are differentE-mail: xuebaoenglish@neau.edu.cnMA Ying. Relative Eficiency Evaluation on Water Resource Utilization.63.attributes). The principle of the method is just asMark the standardized decision-making matrix asfollows!-81B=(by)mxn, and the weighted standardized decision-Mark the set of multi attribute decision-makingmaking matrix as C=(c)mxm here ci=wyxby.plans as X={x, x2, ., xm}, the set of indicators asSetcj=max{cji=, 2, . m}=max {wb,i=l,2, ... m}P={Pl, P2, ", p}, and the index right weight vector=max{b,li=l,2, .. m}w=b*w, j=1,2, .,nis figured out by GEM as W={w, W2, . w}, whichHere, b'; is the ideal value in the jth column ofhas been normalized. aq represents the atribute valuematrix B, and then the ideal solution C={ct,cz, ...of plan x for indicator P, and A=(ay)mxn is the deci-c} is acquired. The evaluation target value of a plansionmaking matrix. Normally, there are differentis defined as the distance between the plan and thetypes of indicators, such as cost-based pattern, bene-ideal point, which is expressed by the sum of error'sfitbased pattern, fixed pattern, interval pattern andsquare:so on. Because there are contradictions among diffe-d(=(gjcjB(wb,-w,b)rent types of indicators and they can't be measuredtogether, the decision-making matrix needs standar-=2 w?(b; b")}(i=l,2, . m)dizing.The indicators proposed in this paper just involvedObviously, the less the value of d, is, the closerin cost-based patterns and benefit-based patterns.the plan is to the ideal point, which means the planBenefit- based indicators are processed as below:further close to the ideal plan. Plans should be rankeday-a,"inaccording to the value of d; thus, the relative effi-ag,*-a,ciency of plans is obtained.=l,2, .,. m;j=l,2,.., nHere, a,"ax and a" im are the maximum and minimumExamplein the jth column, respectively.As for cost-based indicators, they are processed asA calculation example is established to demonstratebelow:the using of GEM-MAUE model. Supposing there areag,x*-ayfive schemes for an area's decision makers to choose,those plans have different index values which are=l,2, .. m;j=1,2, ...nshown in Table 1.Table 1 Index values of decision -making unitseec .Plan(Ten housnd Yuan) (Ten holusand Yuan) (Ten thousandYuan) c1(%) c2(%) c3(%) 4(%) c5(%) 6(%)1008020308(40212070605(57(6(0018:50__7S50_The above model is applied to assess the plans.0.155, 0.114, 0.115)" by using GEM model and thenThere are five experts (S1, S2, . S5) who score thestandardize Table中国煤化工，above nine indicators, and the results are shown inAt last, figureCNMHG318, d,=Table 2.0.0544, d;=0.0416,:DYHGet x=(0.093, 0.121, 0.104, 0.100, 0.103, 0.095,plans is as below: Plan3>Plan5>Plan1>Plan4>Plan2.http: /publish.eau.edu.cn.64.Jourmal of Northeast Agricultural University (English Edition)Vol.18 No.3 2011Table 2 Experts scoring matrixItemSS2S.s4Ss54ec2:3:5Table 3 Standardized index valuesPlanee:2c3c4c5c0.3330.5710.2500.4000.1250.6671.0000.0000.5000.6000.3750.857.0000.2000.2860.7500.6250.1673 Colby B G. Transactions costs and efficiency in western waterConclusionsallocation [J]. American Jourmal of Agricultural Economics, 1990.72(S): 1184-1192.GEM-MAUE model which is concise and practical4 Loomis J. The economic value of insteam flow: methodology andsuited some situations where people have specialbenefit estimates [J]. Environ Manage, 1987, 24: 169-179.preferences for asssment indexes. Evaluation factors5 Schluter M, Savitsky A G, McKinney D C, et al. Optimizingcould be chosen, according to the concrete backgroundlong-term water allocation in the Amudarya River delta: a waterof real cases. And the configuration plans should bemanagement model for ecological impact asessment [小}. Environ-measured in more comprehensive input-output views.mental Modelling and Software, 2005. 20(5): 529-545.6 Qiu W H. An eigenvalue method on group decision [小ApplicedReferencesMathematics and Mechanics, 1997, 18011): 1027-1031.1 Wang H, Wang J H, Qin D Y. Research advances and diretion on7 Wu W Q, Dong Y M. The ideal solution of multiple decisions [].the theory and practice of reasonable water resources allocation [J]Joumal of Kunming Metallurgy College, 1999, 15(4): 62-63.Advances in Water Science, 2004, 15(1): 123-128.8 WuZ N, Cui M, Cao Q. Application of BP artificial neural network2 McCann L, Easter K W. A framework for estimating the transac-model in evaluation of water resources utilization [小South-to-tion costs of alternative mechanisms for water exchange and alloca-north Water Transfers and Water Science & Technology, 2004.tion [U]. Water Resources Research, 2004, 40(9): 1-6.2(3): 25-28.中国煤化工MHCNMHG .E- mail: xuebaoengish@ neau.edu.cn