Thursday 19 August 2010

The Sustainable Society Index

This is a comment from Geurt van de Kerk in response to Jon Hall's last blog post titled From homo economicus to homo edoctus: evolution in the information age. We thought the comment was so interesting we would publish it as a new post.

You have written an appealing and challenging blog, Jon. It provokes me to write a reaction. You have posed a question about measuring and I happen to know the answer to that question. But first of all, what do you want to measure? Success? Progress? Progress to what? Happiness? Or ‘just’ to replace GDP by a better measure?

Of course we can measure everything that is important or at least interesting. But should our main concern not be about progress to sustainability? Without living within the ecological limits imposed by our planet and the social limits set by the community of mankind, life cannot be continued. Happiness is less important than sustainability. If we are talking about progress, it should be about progress on the way towards sustainability.

We can use the famous Brundtland definition for expressing what we mean by sustainability and add a third sentence to this definition to make explicitly clear that Environmental wellbeing and Human wellbeing are both included. The extended Brundtland definition may run as follows:
A sustainable society is a society

  • that meets the needs of the present generation,
  • that does not compromise the ability of future generations to meet their own needs,
  • in which each human being has the opportunity to develop itself in freedom, within a well-balanced society and in harmony with its surroundings.

There is overwhelming proof of the risks we run by the ever increasing GHG emissions, by the social inequalities worldwide, by the loss of biodiversity, by damaging our environment. Let’s not quarrel about the question which indicator or set of indicators is best. Any set of indicators that covers the main aspects of sustainability and is built upon reliable data is OK.


Let’s not fool ourselves by measuring things like happiness. Of course, it is awfully nice if everybody is happy. Apart from the question how to measure happiness, what does it tell us? What does it mean in the long run?


Allow me a sideline. You say that you’re “pretty sure that enforcing a massive cut in CO2 emissions, would make most Westerners decidedly less happy at the moment”. I am not pretty sure, but absolutely sure that this is not necessary. Will you be less happy if your electricity is produced by solar or wind energy than by fossil fuels? Will you be less happy if you buy local or regional food, having travelled less kilometres before it ends up in your stomach? Will you be less happy if you live in an well insulated house, requiring only some quart (or less) of the present energy consumption? I am sure your answers will be negative. And if we turn it round, how can you be happy if temperature rise will pass the tipping point and will cause yet unknown environmental and social damage? For the sake of happiness, we’ll have to cut the GHG emissions.


I admire Bhutan for having replaced the GDP by their own index, the Gross National Happiness. It is an audacious step. It would be even better to replace GDP by a set of indicators that measures the conditions for the possibility of being happy, not only today, but also on the long run. That is a set that measures the main aspects of sustainability in its broad sense, as expressed by the extended Brundtland definition. And for those who prefer a single figure, the scores of all indicators may be aggregated into one overall figure.


Oh sorry, I nearly forgot to answer your question about measuring. One of the possible solutions may be the Sustainable Society Index (SSI), a set of 24 indicators, covering all main aspects of Human, Environmental and Economic Wellbeing. And showing one overall score for sustainability as well as the scores of each indicator separately. So it’s up to you which one you prefer.


Let’s act. It requires all our efforts to prevent mankind to evolve from homo sapiens to homo stupidus.


Geurt van de Kerk

Tuesday 10 August 2010

Enriching our poverty measures


One of the most widely recognised Millennium Development Goals (MDG) is halving the proportion of people living on less than $1 per day. And yet, how many of us have actually taken time to think about what that could mean, to live on less than $1 a day? Indeed, for most people in the developed world, this represents such a miniscule sum as to be almost meaningless. How can we better understand what life is actually like for the world’s poor?

A new international poverty measure – the Multidimensional Poverty Index (MPI) – devised by the Oxford Poverty and Human Development Initiative (OPHI) and the UNDP Human Development Report, goes some way to addressing this problem. A person may not be considered poor according to the traditional monetary threshold and yet still have woefully inadequate access to schooling, food, healthcare, electricity, water or decent housing.

The MPI takes these different dimensions into account, and finds significant differences between the income-poor population and the population facing wider deprivations. For example, in Ethiopia, only 39% of the population is counted as being income-poor using the $1.25-a-day measure (the $1 figure was revised to account for inflation), yet 90% of the population are “multi-dimensionally” deprived according to the MPI.

As the MPI directly measures the hardships experienced by individual households, it can give a much more nuanced picture of patterns within countries. In Kenya, the overall poverty rates for the Kikuyu and Embu ethnic groups are similar; however the experience of poverty for the two groups is quite different. The Kikuyu suffer from higher rates of child mortality and malnutrition, whereas the Embu are more likely to face insufficient access to electricity, cooking fuel and adequate sanitation.

From a policy perspective, this kind of detail is crucial to be able to identify the challenges faced by different groups and to design effective solutions accordingly.

There are obvious links between lacking income and being deprived in a wider sense, and monetary measures of poverty remain very useful. However, poverty is a complex and multifaceted phenomenon – a reality that is sometimes obscured by income statistics. The availability of international comparable data of the breadth and depth of detail provided by the MPI can hopefully result in more informed policy and tangible change in the lives of poor people.
Of course, poverty is not just a problem for the developing world and the need for more comprehensive measures of household living conditions is increasingly being recognised in OECD countries too. The Stiglitz-Sen-Fitoussi Commission made recommendations in this vein last year, and surveys such as the EU Survey on Income and Living Conditions are starting to give a richer picture of poverty and social exclusion. The OECD publication Growing Unequal provides an analysis of non-income poverty indicators in OECD countries.

These developments are important because how we define and measure a problem dictates how we respond to it. By using a narrow definition of poverty, we risk ignoring large groups of disadvantaged people and implementing ineffective policy. Better data can provide a powerful (and much-needed) advantage in the fight against poverty.

This post was contributed by Kate Scrivens of the OECD’s Statistics Directorate and was released on the OECD insights blog. Kate is working on a project researching indicators of household vulnerability and resilience in OECD countries.

Thursday 5 August 2010

From homo economicus to homo edoctus: evolution in the information age

Let me start with the proposition which is at the heart of wikiprogress. We get what we measure, and so the indicators we chose to reflect success become the things we strive for. It follows that if we measure the wrong things we may get the wrong outcomes. At the moment GDP growth is the dominant goal of policy makers. But economic growth is not a good measure of success. At best it is partial. At worst it is downright misleading.

So, what’s the problem? All we need do is replace, or at least complement, GDP with another measure or measures that will focus attention on genuine wellbeing.

But of course it isn’t quite as easy as that. Even armed with the world’s most lucid arguments, backed by some of the most brilliant minds, we cannot click our fingers and ‘replace’ GDP. Thousands of policy makers, media commentators and others have jobs built around analysing and using GDP, and much of the architecture of government relies on the figures. So there is a systemic inertia to change …. It reminds me of the old joke about “how many psychiatrists does it take to change a lightbulb?” (one, but the bulb needs to want to change).

It would be a very interesting piece of economic history to chart GDP’s rise to power. Why, despite the warnings of Kuznets and others, did GDP so quickly become the pre-eminent measure of progress? There are many reasons, but the fact that it is just one number is surely among them: relying on a single measure to judge progress is much simpler than trying to analyse several indicators that will often move in different directions. We are busy, we are lazy and we prefer simplicity to complexity. And in this information age the surplus of information creates a deficit of attention. The ideas of using sets of indicators for decision making isn’t new, but one of the reasons the social indicators movement of the 1980s collapsed is probably the lack of any single indicator.

It follows, therefore, that if we want to dethrone GDP, we might have more chance of success if the pretender to the crown is one number rather than a set of them. Of course relying on any single number runs the risk of repeating the problems of relying on GDP. A single measure of progress must be a simplification of that complicated system we call life. Naively charting a course from just one number might lead to distorted behaviour and skewed policies: hitting the target and missing the point. In a perfect world we would rely on a great deal more information – at the very least on a set of indicators. But the world isn’t perfect and if we want to change the paradigm we need to be pragmatic and a single number is the pragmatic choice. I think we could find a better single number than GDP. Perhaps not perfect. But better.

Now, if we want a single measure to replace it then we have two choices. We can create a composite index like the Human Development Index, or we can select one ‘simpler’ measure of something else (eg life expectancy, literacy etc) to summarise progress.

There are lots of good arguments in favour of composite indices and they serve a useful purpose. But ultimately they rely on taking a whole bunch of stuff that is important (education, health, income, happiness, etc) and adding it together. Before you can add different things together you must express them in a common unit of measurement. And that just about always requires some subjective decision about the importance of the components. You can claim everything is equally important and take a simple average (which is the same as saying you have no idea what is most important), or you can weight them together according to their importance based on some underlying theory or econometric model. But there is no single model or theory on which everyone agrees and so often the debate around the indicator will end up revolving, not around the number itself, but around the selection of the weights. Composite indicators can generate more heat than light.

If we don’t want a composite indicator, we must find a single non-composite measure of something that somehow summarises the progress of a society. And increasingly I think the only possibility is a measure of how we feel about our lives (what the experts would define as our subjective wellbeing, and what the media would call happiness). Call it what you will, but an overall measure of life satisfaction must depend – at least in part – on many of the other bits of life that are important and we can measure objectively: our health, our income, our relationships, the quality of our environment. So it is a broad summary measure. It is also readily understandable and resonates with everyone. Parisiennes aside, we all want to be happy.

There are those, like Richard Layard, who argue that just about all government policy can be designed to maximise subjective wellbeing. I can’t imagine this happening, not least because sustainability questions don’t seem to feature very prominently in determining people’s current levels of wellbeing. I’m pretty sure that enforcing a massive cut in CO2 emissions would make most Westerners decidedly less happy at the moment, even if our grandchildren will be a lot happier as a result.

I can imagine a measure of wellbeing being used in two ways: to challenge the economic output is progress paradigm, and to provoke a more informed and intelligent facts-based debate about what matters for policy making. If commentators spent as much time analysing the latest subjective wellbeing figures as they did discussing economic growth then perhaps GDP will be dethroned. Imagine if each and every time a politician discussed the pros and cons of a new policy in terms of its effect on economic output, the press and public answered “So what? What will it do for our wellbeing?”.


There are those who will doubtless criticise the way in which wellbeing is measured and point to all the problems with the numbers and how they are collected. They aren't perfect. But, neither, when you take a closer look at what goes into and what stays out, is GDP. In any case, I’m not suggesting that the measure of wellbeing is picked up and used blindly to set policy. Rather it is a way to change the focus of the debate and ultimately deepen it.


Analysis of the happiness measure would surely lead to a debate about the underlying data. Let’s say that this month’s figures show the subjective wellbeing of middle aged women is suddenly much lower than that of men. The next step would be to investigate differences between men and women across a broad range of socio economic data to see what has happened. Imagine the rich new policy conversations that might start as a result. Then, and only then, might we claim to have moved into the information age. Long live homo edoctus!


Jon