The idea of quantifying aspects of society in order to better understand our current situation is not new. Since ancient times governments have collected data on the population and other resources under their command, for a variety of reasons and purposes.
One of the earliest known collections is that of Sir William Petty who, in 1665, presented estimates of population, income, expenditure, stock of land, other physical assets and human capital for England and Wales. At the time, the second Anglo-Dutch war was underway, and Petty's estimates were intended to provide a quantitative framework for the implementation of fiscal policy and mobilisation of resources (Maddison 2004). From Petty's foundational work 'the art of reasoning by figures on things relating to government' was developed, otherwise known as Political Arithmetick (Schumpeter 1954).
Skipping forward to the 1930s, it was the absence of information of pressing need that led to the creation of the national accounts system, on which GDP is based. National income accounts rose to prominence at that time as policymakers, trying to recover from the Great Depression, recognised that they knew too little about how the economy was actually performing. Policymakers, even in advanced countries, were attempting to steer their economies out of the Great Depression armed with only partial information on some facets of the economy (BEA 2000).
Given the recurrent criticism of GDP, it is easy to forget that the system of national accounts represents a milestone in our measurement systems (Hall et al 2010). Simon Kuznets won the Nobel Prize in Economic Sciences in 1971 - in only the third year in which it was awarded - for his empirically founded interpretation of economic growth.4 And, at the close of the 20th Century, the U.S. Department of Commerce concluded that the invention of the national economic accounts was its greatest achievement of the century (BEA 2000).
2.1 Learning lessons from history (GDP is not a dirty word)
As mentioned, the development of national accounts arose from the need to understand the state of the economy during the Great Depression.
Today, issues and pressures around the environment, particularly climate change, the use of natural resources and questions about patterns of development have raised concerns that continued use of the Earth's limited resources at current (or increasing) rates may endanger the economic possibilities available to our descendants. These concerns are driving demand for additional measures to better understand and inform policy to promote sustainable development - as highlighted, for example, by Jose Manuel Barroso: 'we cannot face the challenges of the future with the tools of the past' (Beyond GDP 2007). As with the national accounts in the 1930s, what we require are indicators better suited to helping measure and address the challenges of today.
An increased focus on progress has led to an abundance of measures. The demand for better indicators of wellbeing and the sustainability of wellbeing (discussed below) are arguably precursors to the next phase in the development of our modern statistical systems. Over the past two decades there has been an explosion in the number of alternative indicators, and a surge of initiatives tasked with developing new indicators of performance - economic, social, environmental, political and others. This explosion is not surprising given the multi-dimensional nature of progress, but the problem from a policy making perspective - and from the perspective of broader community acceptance - is that the abundance of measures can lead to confusion. Not surprisingly, different perspectives and different measures can provide conflicting signals. Sometimes, it is hard to keep sight of the big picture.
The Australian Bureau of Statistics (ABS) has identified over 70 indicator projects worldwide and about 50 for Australia alone - and they are still counting. Before developing competing sets of indicators, there are lessons to be learnt from history.
It is useful to draw on the factors that have made GDP, and more generally, the system of national accounts, such a success, separate from it meeting a clear need. Three in particular are worth highlighting: GDP is a simple and straightforward measure; national accounts are based on a conceptual framework, grounded on a clear understanding of the problem; and the national accounts are largely based on a set of international standards. These factors have assisted greatly in obtaining community acceptance. It is also the case that GDP provides important and useful information. For example, as a measure of market production, GDP is a useful tool in managing the macroeconomy because it is a key measure of the level of economic activity, and hence employment, in the market sector. Also, while far from perfect (see below), GDP per capita is still a useful measure of material living standards.
A problem is that, over time, GDP has been interpreted as a broader measure of progress by some policymakers, commentators and the public. This is despite a clear understanding by the inventors of the national accounts that 'the welfare of a nation can, therefore, scarcely be inferred from a measurement of national income' (Kuznets 1934, p.7). While it is true that GDP provides important and useful information, focusing on a single measure, for purposes for which it was never intended, and using it without acknowledging its limitations can lead us astray.
Statistics are admirable attempts at summarising complex activities taking place in society, and will invariably involve compromises. This is not a concern as such, as there are nearly always good reasons for these compromises (Stiglitz et al 2009). However, it is of concern if the users of these measures are unaware of their limitations.
This suggests a cautious approach to adopting new measures. Just as one number cannot tell us all we need to know about the economy, it will not be possible to develop a new single measure that can give us a complete summary of society's progress. After all, you cannot capture happiness on a spreadsheet any more than you can bottle it (Cameron 2010).
With this in mind, it is important to confine the use of measures to the purposes for which they are fit.
4 The first Nobel Prize in Economic Sciences was awarded jointly to Ragnar Frisch and Jan Tinbergen in 1969, followed by Paul Samuelson in 1970 (http://nobelprize.org/).