1. Introduction In 1987, the need to strive towards development aimed at making it possible to satisfy the present needs without compromising the needs of future generations is generally acknowledged. The Rio Earth Summit in 1992 propagated Agenda 211, which is composed of principles to reconcile environmental protection, economic development and social equity. More then 150 countries committed to undertaking concrete actions and measures. The Johannesburg World Summit on Sustainable Development in 2002 strove to address the obstacle and further priorities for implementation of Agenda 21: from an often vague definition of the objectives to an irregular control of activities and to the lack of final evaluation, several reasons explain why Agenda 21 has partly failed to achieve the expected goals so far. To improve these processes as well as to ensure their continuation, local authorities must be able to achieve actions of high quality in terms of sustainability, and develop the capacity to assess and evaluate efficiently the different projects. This currently remains difficult. How does one measure progress toward sustainable development? The general aim of this dissertation is to identify the key aspects in the sustainability field in order to set up an integrated method, which will help local actors to plan, monitor, assess and communicate the results of their actions in an easy and efficient way. The developed sustainability assessment procedure will meet the requirements of a comparative approach such as Life Cycle Assessment. This thesis aims at exploring ways to study consistently the three sustainability fields simultaneously, with a focus on the Swiss and regional context of the Vaud county. It has therefore a strong interdisciplinary character and an exploratory nature. Special care is put on the coherence between the used methodologies, the details of each methodology being not at the center of this work. The more detailed objectives of this thesis are to determine the essential steps for action assessment and specifically: To develop an assessment framework, which will help to consider all the sustainability aspects relevant to action evaluation (chapter 2). To define, on the basis of the assessment framework, a step-by-step procedure including a visualization and communication method for presenting the results (chapter 3). To set up a toolbox with compatible quantification and/or qualification methodology for the different sustainability fields (chapter 3 and 4). To study the appropriateness of the Input-Output approach for Swiss local assessment of Agenda 21 project, quantifying environmental, economic and social impacts in a consistent way (chapter 4). To apply the methodology described above to real case studies meeting the needs of local authorities (chapter 5). To achieve this, an interdisciplinary project called "Priority 21" was launched in 2000 involving two EPFL Laboratories: the GECOS (Ecosystem Management Lab) and the LASUR (Urban Sociology Lab). It was financially supported by the ARE (Federal Office for Spatial Development), and the cities and cantons concerned. The project objectives were to develop a project assessment method to help improve local projects and to identify the key sustainable development issues in the area through a representative survey of the population. The following four Swiss communes, chosen because of their diversity in size and function, took part to this project: Lausanne (VD), urban locality, on the shores of Lake Léman, with 114'304 inhabitants of which 35 % are foreigners, Vevey (VD), urban locality, on the shores of Lake Léman, with 15'364 inhabitants of which 39 % are foreigners. Essertines/Yverdon (VD), rural locality with 666 inhabitants of which 6 % are foreigners and Mannens-Grandsivaz (FR), rural locality of 505 inhabitants of which 5 % are foreigners. At the same time, a smaller project was launched with the city of Neuchâtel (NE, urban locality, on the shores of the Lake of Neuchâtel with 31'571 inhabitants of which 31 % are foreigners,), aiming at defining a set of transparent and communicative indicators allowing the follow up of the Neuchâtel local Agenda 21. Both projects have been carried out in parallel. The main focus of "Priority 21" was the methodology development with tests. Neuchâtel project focussed on the systematical application of the developed tools. 2. Assessment Framework The developed analysis framework (Figure 1) is composed of four interdependent fields: the societal characteristics or locality characteristics, the individual or collective behaviors, the sustainability impacts and the inhabitants' perception of these three fields. The link between society and behavior can be studied through sociological analyses (link 1). The link between behavior and environmental impact can be established through Life Cycle analysis and Input/Output analyses (link 2). At each level, sustainable development indicators can be defined. For the present assessment of project, indicators are mainly retained at two levels: a) indicators of behavior and practices to ensure the regular follow-up of the project consequences on practices b) end objective indicators to measure the environmental and socio-economic impacts of the considered project. As an illustration of the framework application, one performance indicator has been selected: the primary non-renewable energy. The citizen energy consumption has been studied and expressed in energy slaves (= 3'150 MJ/Year). A Swiss citizen needs altogether 69 slaves, 20 for his private consumption of goods and services, 19 for housing, 11 for transport, 11 for food and 8 for his consumption of public goods. If 1/5 of the inhabitants of Lausanne, wanting to live in a rural environment, actually moved out of town, the region would need 10'436 more slaves for 3'500 persons living in country areas, thus the interest to promote urban quality of life. The use of one single indicator limits the interpretation of the results. In the presented cases, the aspects of quality of life, money and time are very important in a decision-making case. 3. A step by step procedure with its toolbox A short and concise six-step assessment procedure (Figure 2) facilitating the collection and presentation of project repercussions has been developed. The aim is to assess a) the efficiency, the degree to which the project has been productive in relationship to its resources, b) the effectiveness, the degree to which the goals have been reached and c) the impacts, the degree to which the project resulted in changes. The first step corresponds to the in depth project study with its detailed objectives, its goal and scope, the means chosen to reach them and the induced behavior change. The second step consists in establishing an inventory of the required resources, available data and existing analogous projects. The third step is the choice of indicators and tools adapted to the project. The fourth step is the quantification of practices or behavior. The fifth step is the impact evaluation. The last step is the interpretation of the results, including e.g. communicative dashboards for the results presentation. A toolbox comprising different quantitative and qualitative methods is proposed to support steps four and five. : Questionnaires and surveys can be advantageously used. These primarily qualitative methods make it possible to create direct links between the societal characteristics and the behaviors, as well as between the perceptions and the behaviors. They provide indicators of behavior and practices. To quantify action-induced impacts, life cycle energy balances in parallel with life cycle costing are used for first screening. They provide end objective indicators. Next, a complementary model, the extended Input-Output model also called Leontief model3, is studied and adapted. The fundamental purpose of the Input-Output framework is to analyze the interdependence of industries in an economy. Extended Input-Output models go further and deal with energy consumption, environmental pollution and employment associated with industrial production. Input-Output models can potentially provide both types of indicators. Figure 3 illustrates the coverage of the different methods. 4. Extended Swiss IO model for sustainability assessment: Development and Suitability The fundamental purpose of the Input-Output (IO) framework is to analyze the interdependency of industries in a regional or national economy. Extended IO models go further and deal with energy consumption, environmental pollution and employment associated with industrial production. The development of an extended interregional IO Swiss models for local sustainability assessment is the core of this chapter, testing its ability to assess environmental, economic and social performances in a consistent way. Currently the main weakness for performing this type of analysis in Switzerland is certainly the current and long-term data availability. The Swiss basic inter-industry transaction table dates from 1995 and has 37 sectors. Sectoral environmental emission factors are not collected in Switzerland at all. Compared to the US, with its 500 sectors IO model extended to more than 1'000 pollutants implemented on different software, like LCNetBaseTM, MIET, Switzerland has really poor statistics. We therefore developed a Swiss extended IO model with the available statistics. To summarize, the issues encountered were the following: he available Swiss economic data is in an old Swiss classification (NGAE classification). The number of sectors, 37, is very limited. The recent regional and national employment data is in the new Swiss classification (NOGA classification). A key to transform the old classification into the new one is needed. Environmental data from European countries is available in the European classification (NACE classification) in NAMEAS4. To build a coherent model, the data has first to be studied. Then coherent extrapolation for Swiss factors must be found and an original classification-transformation key has been created (NACE-NOGA). A way to make the different elements compatible while safeguarding as far as possible the data quality and the most detailed information possible i.e. the number of sectors, the number of pollutants, has been developed. As an application the environmental impacts and the total output generated from the households' final demand on the country and on the neighbors have been assessed. This first illustration shows the primary energy necessary to meet the citizens' final demand calculated with the developed extended Swiss IO model. We can see that primary energy comes from abroad for the sectors A (agriculture, hunting, forestry), B (fishing and fish farming), C (mining and quarrying), D (manufacturing), E (electricity, gas and water supply) and I (transport, storage and communication). This is not a surprise except for E. For E, about 1'753 mioCHF/year is imported goods. The fact that primary energy is so high is most probably linked to the energy sources. For CO2, we can also see that the sectors with a high-energy consumption are the same that emit important quantities of CO2; they are D (manufacturing), E (electricity, gas and water supply) and I (transport, storage and communication). For NOx, the highest shares correspond to manufacture by foreign companies (D) and domestic transport (I). Again, a strong link to energy is to be seen. This applies to the dust and SO2 case as well. For electricity, this means that the assumption made when extrapolating coefficients for Switzerland corresponds to the UCTE mix. This could be a valid assumption for Switzerland as changes in Swiss electricity consumption only marginally affect Swiss production and rather correspond to a change in imports from the European network. A sensitivity study would be interesting to further investigate these points. Extended IO models potentially offer a wide range of possibilities. For Switzerland, data had however to be gathered or improved in relation to the study. The final Swiss IO, with its 34 sectors, is not an all-purpose, all-case bulldozer model. If the sectors are further disaggregated, it can give thorough, comprehensive and exhaustive results. Special care must be taken when analyzing a project. It is necessary to check exactly what the IO is considering and what is to be left out or needs to be corrected, such as the direct household emissions that are not included. 5. Case study: Sustainable Mobility in Neuchâtel In 1999 the city of Neuchâtel undertook action to elaborate an Agenda 21, which consists of performing twenty-one actions, then evaluating and following them over time. The objective of this part of the thesis was to analyze the mobility actions of Neuchâtel with the developed methodology in this thesis. This case study has been chosen because this is a topic, which is very often present in Agenda 21 Action Plans. Moreover, this mobility case study provides thorough analysis and conveys the research results quite well. For this action, the link between societal characteristics and practices is evaluated on the basis of a sociological questionnaire. The link between practice and performance is defined by using a life cycle assessment LCA (assessment of non-renewable primary energy, of CO2 and NOx). The economic aspects are dealt with using a cost analysis. The extended IO model presented in chapter 4 is not used for this case study, mainly because no regionalization at the Neuchâtel city level is available. The sociological mobility analysis brings interesting elements to the surface. The automobile is far from being omnipresent in Neuchâtel: one-third of those surveyed use public transport or walk in their daily mobility. The proportion of "two-wheelers" is very low. The same can be said of motorcycles and scooters. As for the system analysis, the results are the following. For vehicles as a whole, the phases of use and supply of energy are dominant, for non-renewable primary energy as much as for CO2 and NOx emissions. The phases of production, maintenance and elimination of the vehicles, as well as the infrastructure can nevertheless be significant, particularly for electric vehicles. The electric Swiss Mix (CH) (composed of 59 % hydraulic and 41% nuclear) is clearly a better option than the Mix-UCTE (14 % hydraulic). It is important to ensure that electric vehicles are recharged by the Mix-CH, or in an autonomous way (solar energy), so that using them does not have a negative impact. To improve environmental and economic performance of automobiles and public transportation, it is necessary to improve the occupation rate. The evaluation method is adapted to this type of projects. The combination of behavioral studies with the environmental and economic aspects is possible and proves to be very interesting. Understanding practices and household transportation equipment makes it possible to better target the visualized actions in the future. Presenting the results in the form of a dashboard opens up interesting perspectives to the measure in which the latter make it possible to completely visualize the situation "at a single glance". The method and toolbox have been tested on several case studies. They can be found in Annexe 3.1 and in the final report of the Priority 21 project5. A toolbox is available on-line for project assessment6. 6. Conclusion The analysis framework developed with its four interdependent fields and the complementary sustainable development indicators is consistent and helps to evaluate actions objectively. The developed procedure has been tested on more than 20 different actions; the behavior quantification linked to the impact assessment works to give interesting and trustworthy results, though the procedure is not adapted to be used directly by the action leaders. The choice and especially the use of the proposed tools (sociological and system analysis) is a specialist issue. Further developments to have a user-friendly interface are necessary. The creation of a database with already analyzed actions could be of great help for non-scientist users. Comparing actions that are very different in nature and size is difficult. The best way to solve this issue is to know about the state of the region (total expenses, total emissions, etc) and to compare the projects impacts with the state. This then gives a clue as to the direction in which the action is going. In the present thesis, the institutional aspects have intentionally not been considered in detail in the developed framework that has been adopted. This institutional side needs further development. Parallel to that, the analysis of participative processes, either included in the process of definition of a local Agenda 21 plan of action or as projects, should also be further investigated. The analysis of participative processes is an objective in itself and should be studied thoroughly. The final decision concerning a project remains political. The assessment provides information for an expert decision. The results sets the basis for example used in open discussion with the stakeholders. When using a toolbox, the decision situation stays open, to enable diverging ways to combine the different types of information. Openness is a part of effective communication. The audience that is interested in sustainable development performance includes the general public and experts, individual households and international organizations, and public and private sector decision makers. Thus an accent in research in communication and participation is required. ------------------------------ 1 Earth Summit Agenda 21, The United Nations Programme of Action from Rio, downloadable: 2 A European System of Environmental Pressure Indices, First Volume of the Environmental Pressure Indices Handbook: The Indicators Part I: Introduction to the political and theoretical background (available at 3 Leontief W., Input-Output Analysis, october 1951, Scientific American. 4 NAMEA for National Accounting Matrix with Environmental Accounts, it consists of a conventional national accounting matrix extended to include environmental accounts in physical units. 5 [Corbière-Nicollier & al., 2003] Corbière-Nicollier T, Jolliet O., Ferrari Y., Jemelin C., Priorité 21, rapport final, Lausanne, août 2003. 6