Getting to grips with 'sustainability'
For ‘social responsible investments’ in government bonds, it is often required to rank countries with regard to ‘sustainability’, or to establish which countries ‘perform above average’ and which perform below average. Here we will focus on 26 OECD countries (Slovakia, Iceland and the Czech Republic were excluded due to lack of certain information). Clearly, one needs a robust approach and a plausible story: there are 10,400,600 ways to divide 26 countries into two equally sized groups, and there are 400m trillion ways to rank them. Our suggestion will be just one of them…
Now ‘sustainability’ is one of those dynamic, live concepts, which are hard to define operationally. It is essentially an ethical, responsibility concept: sustainability means to manage the affairs of the present generation without diminishing the options for the future generations. The difficulty to define it operationally is shared by many concepts, for example ‘profits’.
In spite of the disagreement about the precise meaning of sustainability however there is a broad consensus that the concept integrates economic, social and environmental objectives. Countries are expected to do well on all scores, not just on a few of them. There is a clear analogy with the ranking of athletes in a decathlon contest, where the winner is expected to do well in every discipline. We decided to take this analogy seriously and let the countries quasi ‘compete’ with each other on the various dimensions of sustainability, as determined by well-designed indices. We chose six indices, two for each dimension:
ESI: The Environmental Sustainability Index, as developed by the WEF(Yale&Columbia,). This index comprises 67 indicators on the state of the environmental systems, the stresses on those systems, the human vulnerability to environmental change, the social and institutional capacity to cope with environmental challenges and finally, the ability to respond to the demands of global stewardship.
REF: ‘The Relative Ecological Footprint’ as defined by us as the ratio of the Ecological Footprint to the Biological Capacity, the latter two were developed by the WWF International (Mathis Wackernagel). Basically, one adds the various, mutually exclusive regions that are occupied to produce the resources consumed and to absorb the waste generated by the population. The result is expressed in area units per person and divided by the available area per person that is biologically productive. Countries with a lower score are ranked higher.
CCI: The Current Competitivenes Index as developed by the Harvard Business School (Michael Porter), which purports to measure an economy’s effective utilisation of its current stock of resources. It uses indicators for institutions, market structures, and economic policies supportive of high current levels of prosperity.
GCI: The Growth Competitiveness Index as developed by the Harvard University (Jeffrey Sachs). Building on theoretical and empirical macroeconomics the GCI represents a best estimate of the economic’ underlying prospects for growth in the medium term. It focuses on the level of technology, the quality of public institutions, and the macroeconomic conditions for growth.
SWI: The Social Watch Index ranks countries with respect to their progress towards achievement on measurable targets as set in 1995 at the World Summit for Social Development in Copenhagen. This relates primarily to health and education, development aid and military spending.
HDI: The Human Development Index as designed by the UN, based on life expectancy, educational achievement and adjusted real income.
Each of these indices defines a ‘competition’ and yields a ranking. A snapshot is given in table 1 for three countries (recall that there are 26 countries to compare, ‘1’ refers to the top-country, ‘26’ to the worst country):
As expected, it is not the case that one country ranks first on all indices, nor is there a country performing worst on all scores. We have to somehow balance the various rankings in a way that honours the requirement to find the best all-round ‘athletes’. Our approach is as follows:
We pit each country against any of its competitors and assign points in a super-additive way. As an example, suppose we pit country A against country B and find that A is worse then B on all indices. Naturally it will receive 0 points. If it outperforms B on one index it will get, say, 5 points. But if A is better than B on two indices it will not get 10 points but, say, 15 points. So the outperformance on one additional index is awarded more than additive. Similarly, if A is better on yet another index it will not get 20 points but, say, 30 points. And so on (if A is superior to B on all accounts it will get 100 points). We perform this calculation for every pair of countries. The final ranking is determined by the total net count of the scores.
It is important to note that the sequence of numbers referred to (0, 5, 15, 30,…,100) is just an example. There is nothing special about it; there is no explicit justification for these numbers other then that they are super-additive. In fact we generated thousands of sequences of super-additive numbers and determined in each case the corresponding ranking of the countries.
Table 2 summarises the main results. The first column indicates the range of the ranks the countries may obtain depending on the particular sequence of points. Clearly, it does make a difference which sequence one chooses. The second column yields the percentage of times the countries were deemed to be ‘above average’. As one can see, there is a strong consensus as to which country belongs to this special class. So the exact rankings may be different but not the partitioning. The third column finally gives a consensus ranking based on the median of the ranks observed.
A client who disagrees with the partitioning in above versus below average performance can essentially only question the indices chosen, but not the particular points assigned.
We presented an approach for the ranking and partitioning of countries with regard to sustainability, which does not require a precise operational definition of sustainability, but only the rankings of countries on indices reflecting various facets of the concept. The approach is relatively robust; it does not depend excessively on arbitrary assumptions. We will repeat the analysis on a regular basis, taking new developments with regard to indices into account.
Professor Dr Theo K Dijkstra is with SNS-REAAL Group/SNS Asset Management, and the University of Groningen/ Econometrics & OR in the Netherlands