Multicriteria Analysis
– multicriteria analysis as a method for project evaluation – Multicriteria analysis takes also non monetary effects into account – Illustration with an example
1. Limitations in cost-benefit analysis • CBA is based on the knowledge of all the costs and benefits that are linked to the project, and expressed in monetary values • Some costs and benefits are difficult to express in money, because they don’t have a price (intangible costs and benefits), for example (-) air pollution by cars, (+) attractiveness of the agricultural landscape, …
• multicriteria analysis takes into account both tangible and intangible costs and benefits: balance of effects expressed in different units is possible • The decision is based on different criteria (for example: environment, economics, …) each with their own weight (one criteria can be more important than another one)
2. Basic principles of MCA • Multicriteria analysis: – Method for the comparison of different alternatives that are evaluated for different criteria – Is based on pare wise comparison of alternatives – an alternative is preferred to another alternative if it scores enough better than the other alternative in the majority of the (important) criteria
3. Method • Step 1 – per criteria, express the relative difference between the alternatives by preferential functions (6 types); in this way, criteria can be compared
• Step 2 – Express the importance of criteria by attributing weights – the following formula can be used: n
1 gj =∑ i =k i k = the rang order of the criteria j (with k=1 for the most important and k=n for the least important criterium)
• Step 3 – aggregation by calculation of the preferential indicators 1 – Preferential indicators P ( a, b ) = ∑ g j [ ê j ( a, b ) ] n
met
êj(a,b) = f(ej(a) - ej(b)) if a >j b =0 if a ≤j b ej(a) = score of alternative a for criterium j ej(b) = score of alternative b for criterium j gj = relative weight of criterium j n = total number of criteria
• Step 4 – comparison preferential indicators: P of >>> if P(a,b)>>>P(b,a) – strong preference p of > if P(a,b) > P(b,a) – weak preference I if P(a,b) = P(b,a) and both indicators have low values - indifference R if P(a,b)=P(b,a) but both indicators have high value – incomparable
Determination via PIR-filter (threshold value)
• Step 5 – Determination of the order of the alternatives on basis of: • or PIR relation (if order is desired) • Or determination of a distance index (if further analysis is desired) - distance index determines the position of the alternative in min-max interval (pareto efficiency alternatives)
• Step 6: – Conclusions on basis of outcome and sensitivity analysis: • • • •
Uncertainty with respect to alternatives Criterium study Preference uncertainty Uncertainty of estimation
4. Example • Train connection between 2 places – Split the distance between the places into homogeneous zones (14 zones) – definition of evaluation criteria – evaluation of the zones on basis of quantitative en qualitative scale – identification of all possible connections between the zones – evaluation of each route – Comparison of possible routes
Possible roads Alternative A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 A13 A14 A15 A16 A17 A18
Sequence of zones 1-2-6-9-13-14 1-2-6-9-11-12-14 1-2-5-9-13-14 1-2-5-9-11-12-14 1-2-5-8-10-12-14 1-2-5-8-11-12-14 1-2-5-7-10-12-14 1-2-5-7-11-12-14 1-3-5-9-13-14 1-3-5-9-11-12-14 1-3-5-8-10-12-14 1-3-5-8-11-12-14 1-3-5-7-10-12-14 1-3-5-7-11-12-14 1-3-4-8-10-12-14 1-3-4-8-11-12-14 1-3-4-7-10-12-14 1-3-4-7-11-12-14
5. Conclusions • Multicriteria analysis is a useful instrument for support of decision with regard to projects en investments • advantages: – Different criteria can be taken into account – Non monetary values can be taken into account
• disadvantages: quiet complex when different alternatives are possible