The concept of environmental sustainability or wise stewardship of natural resources is not a new idea. Aldo Leopold (1949) proposed the institution of what he called the "land ethic". By land, Leopold was really considering limited resources such as soil, water, and biota. Even in the middle 20th century, scientists, educators, and government leaders were thinking about future environmental preservation. Other authors and leaders like John Muir Theodore Roosevelt before Leopold and Rachel Carson in the 1960s were beginning to persuade us to adopt a more scientific, objective approach to environmental stewardship. Conceptions of environmental sustainability have evolved considerably and now we have the tools and technologies to greatly improve to management of our natural resources; we only have to have the desire to do so.
Defining and Measuring Environmental Sustainability
The Environmental Sustainability Index (ESI) by Esty et al. (2005) defines sustainability as “systems that maintain themselves over time”. The concept is abstract, multifaceted, and includes a wide range of issues including: natural resource consumption, economic progress, politics and governance, poverty, human health, water quality, pollution, biodiversity and more. The ESI as outlined by Esty et al. (2005) establishes benchmarks using a quantitative, systematic approach and serves to identify issues in policy making. The goal is to set objective measures of efficacy of governance that policy makers can use to improve environmental stewardship at local, national, and international levels over the time scale of several decades. This index also strives to set goals for improved data collection and standardization. The index has undergone considerable review and modification to improve its utility, robustness, and sensitivity to variation of indicators.
The ESI is a collaborative effort focusing on evaluating important environmental decision making outcomes using by using multivariate statistical techniques (Esty et al., 2005).
The ESI gathered a wide array of data from 146 countries using 76 data sets. Ultimately establishes 21 indicators that fall into five broad categories or components. Environmental Systems Reducing Environmental Stress Reducing Human Vulnerability Social and Institutional Capacity Global Stewardship. In order for counties to be included in the ESI they must have data coverage of at least 60% of the variables. Nations were asked to fill any data “gaps”, update, and verify there statistics. Small countries with populations <100,000 and land areas <5,000 km2 were also excluded from the ESI (Esty et al., 2005).
To standardize variables for comparison among countries the ESI uses a series of common "denominators" such as socio-economic indices, demographic information, and species richness data (Esty et al., 2005).
The indicators are equally weighted and averaged for purposes of statistical analysis. Data were then transformed and extreme data points were removed to eliminate "skewedness" and/or kurtosis. Major deficiencies of the ESI are a paucity of certain data and standardization in measurement among countries. Where these data are lacking or metrics are unavailable, proxies become an important tool. The 21 indicators of the ESI are at the heart of its continued development and refinement. Because of lack of data, several important indicators have been omitted; e.g. wetlands protection, waste management, and ecosystem function (Esty et al., 2005). Other data concerning biodiversity, forest and agricultural management, and sustainable fisheries are seriously limiting. Hopefully, in future iterations of the model these and other variables will be included with the development of better data collection standards and new metrics.
Statistical Methods and Sensitivity Analyses
The ESI uses Principle Components Analysis (PCA), stepwise regression, and cluster plots to compare countries and their environmental indicators. PCA is used to identify a number dimensions in the index and to show the influence of indicators. This allows investigators to determine whether or not to weight each variable or allow them to all be equal in the model. Further details of the PCA procedure can be found in Appendix A- Methodologies (Esty et al., 2005).
Indicators can be manipulated or removed to detect the ESI sensitivity. By doing this we can see the magnitude of an individual variable's effect on a given county's rank teasing out which factors have the greatest impact. For example, by using regression analysis to plot ESI verses Per capita Gross Domestic Product (GDP) we can see that this factor explains about 23% of ESI variability and that the Growth Competitiveness Index (GCI) accounts another 19% (Esty et al., 2005). Researchers manipulating the model are using the model to test the index's robustness and sensitivity to uncertainties including: 1) imputed data, 2) weighting of variables, 3) level of aggregation- indicator vs. components, and linear vs. non-linear aggregation schemes (i.e. those countries with some extreme values may dramatically shift in rank if a certain indicator were added or subtracted. It appears that the index is fairly robust and imputing data to fill gaps creates uncertainties in rank of an average magnitude of 10 (Esty et al., 2005).
Cluster Analysis: An Example using Economic Growth and the ESI
Cluster analysis reveals some interesting and potentially significant relationships and trends. The example above; the relationship of environmental sustainability and economic performance is one of the most interesting. Some traditional economic theory has held that commitment to environmental preservation and economic growth may be at odds. Porter and van der Linde (1995) suggest this may not be the case. Porter and van der Linde argue that there has been a paradigm shift away from the old idea that there is an environmental tradeoff for economic progress and that now environmental sustainability and innovation are improving competitiveness of industries. The ESI model has the ability to test these hypotheses. The model allows one to compare peer counties with similar indicators and see where departure of rank occurs and look for contributing factors.
Countries tend to fall out in seven distinct clusters with certain shared characteristics; see Table 14, p.30 in (Esty et al., 2005). Clusters can generally be characterized as follows: 1) high population density, industrialized nations with social and institutional capacity, 2) undeveloped countries, 3) large countries with low population densities, 4) 'eastern block countries', 5) central and south American countries with relatively intact ecosystems, 6) Russia and former Soviet states, and 7) high population density countries with stressed ecosystems like Bangladesh, India, and Mexico. Cluster six contains Australia, Canada, Finland, Iceland, New Zealand, Norway, Sweden, and the United States while the United Kingdom, France, and Germany reside in cluster one.
Correlation Analysis
A comparison among statistically significant environmental indicators and the GCI and GDP/cap reveals some interesting trends that can play an important role in governance and policy making. Of the 21 ESI indicators, most were highly significant (> 0.01 level). Four of the 21 had little or no significance; biodiversity, water quantity, reduced ecosystem stress, and eco-efficiency (Esty et al., 2005). The highest indicators with respect to governance were: civil and political liberties, environmental governance, governmental effectiveness, democratic institutions, rule of law, and participation in international environmental agreements. All of these six variables had R2 values between 0.50-0.60 suggesting that strong, effective governance are more likely to achieve environmental stewardship goals (Esty et al., 2005).
The ESI does seem to have been a catalyst by influencing some countries to begin to seek environmental solutions and more proactive policy decisions. The 2005 ESI reports about 100,000 downloads of 2002 report on Columbia University alone. The ESI rankings have sparked review and assessment of Best Management Practices (BMPs) among leaders. Mexico, South Korea, the United Arab Emirates, and Belgium. Esty et al. (2005) goes on to point out that the ESI has been instigative in improved water monitoring programs and scholarly research and in education.
Limitations
Some of the most significant limitations are:
- Data gaps and inconsistency in collecting and reporting information are two of the greatest problems with the ESI.
- The index does not seem to consider some "common" areas like oceans, seas, rivers or polar regions.
- The index provides a "snapshot view" as acknowledged by the authors.
- Comparison of countries with large differences in: land area, population densities, and wealth are difficult to compare.
Other Sustainability Indices
Appendix F of the ESI by Esty et al. (2005) compares other sustainability indices using correlation analysis; results are: Ecological Footprint Index R2= 0.15, Environmental Vulnerability Index R2=0.03 and Millennium Development Goal 7 Index with R2=0.29. The Ecological Footprint Index is negatively correlated with the ESI; rich countries with a larger footprint tend to more capable of investing in environmental preservation. The Environmental Vulnerability considers resource depletion and natural disasters shows a week relationship to the ESI. Finally, the United Nations Millennium Development Goal (MDG) 7 Index is concerned with forest preservation, greenhouse emissions, water quality, and sanitation. Although the MDG has a good correlation with the ESI, only 56 countries could be included in the comparison because of incomplete data sets.
The Pilot 2006 Environmental Performance Index (EPI) is under continuing development by collaboration among: the Yale Center of Environmental Law and Policy, Center for International Earth Science Information Network at Columbia University, the World Economic Forum in Geneva, Switzerland, and the Joint Research Centre of the European Commission, Ispra, Italy. The EPI has two objectives: 1) reduce environmental stresses on human health and 2) protect ecosystem function. These objectives are gauged using 16 indicators within six policy categories: Environmental Health, Air Quality, Water Resources, Biodiversity and Habitat, Productive (Esty et al., 2006). Analytical techniques are similar to the ESI.
Conclusions and Future Directions for ESI
The ESI has improved significantly since its inception and seems to be providing benchmarks for performance. Developers and policy makers are moving in the right direction by making the whole decision making process more analytically based (Esty et al., 2005). Future directions will include more rigorous data collection and validation, more widespread use of Geographic Information Systems (GIS) and remote sensing methods for collection data, and continued feedback from scientists and policy makers.
The Demotechnic Index Relating Human Population and Resource Consumption
The Demotechnic Index (DI), sometimes also referred to as the "D-index), is summarized by Mata et al., (2006). The DI, as described by Mata et al., relates human population and rates of natural resource consumption in such a way as to compare developed vs. developing countries using a common measure. This is another way of understanding and predicting trends in environmental sustainability that focuses on the relationship between demographics and consumption. Like the ESI discussed above, this index can be used by policy makers to guide their decision making process; especially on a global scale.
Since the Earth has finite non-renewable resources that determine human population carrying capacity, the DI considers demographic data as a valid indicator of sustainability (Mata et al., 2006). Based on work by Vallentyne (1982), the DI strives to allow for the direct comparison of per capita resource consumption among countries with very different socio-economic patterns. Other studies have called for an energy consumption "multiplier" to account for differences in consumption among countries at different stages of development (Goodland et al., 1994; ICPD, 1994).
How the DI Works
Since population density is not a good indicator of sustainability by itself the DI adjusts population by consumption to produce an index of energy consumption rates that accounts for technology, thus allowing a common unit for comparison among countries (Mata et al., 2006). The exact calculation of the index is outlined in Appendix-1 in (Mata et al, 2006). Essentially the DI units are calculated as a ratio of technological energy consumption to physiological consumption expressed in kilocalories.
Strengths and Weaknesses of the D-index
The DI takes into consideration differences in consumption rates of countries across a range of developmental stages. This allows us to relate natural resource consumption rates directly. Problems include: 1) evaluating types of energy consumed since various energy sources have different environmental impacts (Mata et al., 2006). For example, fossil fuel, hydroelectric power, wind and solar generated energy all have different consequences in both temporal and spatial scales (i.e. greenhouse gas emissions, thermal pollution, etc.). Another problem is that the index does not account for variation in waste production among various consumptive activities, e.g. solid waste, nuclear contaminants, nutrification, etc.
Population and Consumption Trends
D-index numbers for 1990 rank countries from 1-144 (see Table 1., Mata et al., 2006). The top 10 countries from highest to lowest consumption rates include: Qatar, the United Arab Emirates, Bahrain, Canada, Norway, the United States, Iceland, Sweden, Kuwait, and Finland. Note that had relatively high ESI values are considered less sustainable under the DI. Some countries that ranked relatively low with the ESI appear to have greater sustainability under the DI. For instance, India is ranked 104 under the DI and Bangladesh is near the bottom of the list at 142.
In 1990, the DI listed the following countries in order of population density from higher to lower and their corresponding adjusted consumption rates:
Country | Population (millions) | Consumption Adjusted Population (millions) |
China | 1,139 | 9,329 |
India | 853 | 3,907 |
USSR | 289 | 16,828 |
US | 249 | 22,993 |
Canada | 27 | 3,159 |
Mata et al., 2006
The top 29 countries listed in Table 3, Mata et al., (2006) shows that these 29 countries accounted for more than 85% relative contribution to global consumption among the 144 countries rated. Several of these nations ranked high and mid-range with the ESI. This seems problematic since the two indices do not correspond well.
Conclusions
In general sustainability indices all have limitations and weaknesses owing to gaps in data, standardization of methods, model complexity, statistical rigor, etc., but they are still useful tools in gaining new perspectives and setting benchmarks for natural resource management and policy making at all scales. As models become more robust this utility will only improve. I believe the biggest hurdle for governments now is to initiate policies and strategies that will achieve sustainability in the face of rapid human population growth. The challenges are great and the outlook often bleak, but individuals, governments, and international organizations must not continue to ignore the problems of natural resource depletion and environmental destruction or deny that they exist.
Literature Cited
Esty, Daniel C., Mare Levy, Tanja Srebotnjak, and Alexander de Sherbinin. 2005. Environmental sustainability Index: Benchmarking national environmental stewardship. New Haven: Yale Center for Environmental Law and Policy. www.yale.edu/esi/
Esty, Daniel C., Marc A. Levy, Tanja Srebotnjak, Alexander de Sherbinin, Christine H. Kim, and Bridget Anderson. 2006. Pilot 2006 Environmental Performance Index. New Haven: Yale Center for Environmental Law & Policy. www.yale.edu/epi/
International Conference on Population and Development (ICPD). 1994. Draft program of action, International Conference on Population and Development. http://www.un.org/popin/icpd2.htm
Goodland, R., H. Daly, and J. Kellenburg. 1994. Burden sharing in the transition to environmental sustainability. Futures 26(2):146-155.
Leopold, Aldo. 1949. A sand county almanac. Oxford University Press. 226p.
Mata, Francisco J., Larry J. Onisto, and J.R. Vallentyne. 2006. International Conference on Population and Development. www.ecouncil.ac.cr/about/speech/secrar/consump.htm.
Porter, Michael E. and Class van der Linde. 1995. Toward a new conception of environment-competitiveness relationship. Journal of Economic Perspectives, 9(4):97-118.
Vallentyne, J.R. 1982. A new approach to membership dues schedules for use by international organizations, Biology International, 5:10-12.