Most, if not all, members of the SPE will spend much of their working time with statistics of one sort or another.  During my career, the sheer quantity of data available and the ease of accessing it has increased overwhelmingly.  Yet we probably don’t spend long enough considering how it was all produced, by whom, and whether it really answers the question we are interested in.  Nor do we consider the value judgements inherent in its construction.  We should understand these issues better – and so we should ALL read this important book.  It is full of interesting facts and certainly opened up this reader to new ways of looking what really matters.

It asks a very big question: how should we measure economic progress in a way that relates to the improvement in people’s lived experiences and well-being; arguing that this matters profoundly as better measurement would drive better decisions.  Do we really know, just by looking at real GDP per person, whether we are getting richer or poorer?

Coyle’s key argument is that a great deal has changed since the present System of National Accounts (SNA) was established in the 1940s, and the changes to the SNA since then have not kept up.   Of course, there is a well-known list of issues with GDP (it omits unpaid work, doesn’t cover nature, etc) and there are various efforts, in the UK and elsewhere, to improve GDP to remedy these deficiencies – often by the use of satellite accounts. We may worry about the ‘price’ of some additions, but she makes the interesting observation that UK GDP is already more than 60% non-market or imputed (mostly general government, rents and FISIM).  More and more of GDP is becoming ‘hard to measure’.

The bulk of the book looks at the various big issues which increasingly bedevil the statistical interpretation of the world.  This starts with ‘dematerialisation’: manufactured goods which come bundled with services, companies which outsource all their manufacturing and other trends making it hard to pin down where value lies, and indeed where it is produced.  Cloud computing, for example, presents special problems.  If industrial policy aims to boost manufacturing – do we really know where the manufacturing boundary lies?

Other topics include the substitution of consumers’ time for producers’ (not just supermarket checkouts though that is the most obvious example), and most dramatically the rise of ‘free’ digital goods.  As we haven’t worked out how to value these, they don’t appear in the conceptual farmwork for GDP – well-described as ‘trying to see in the dark with normal spectacles rather than night vision ones’.  If we do attempt to include free goods, it increases the proportion of GDP which is imputed.

The digital economy also makes it harder to assess what crosses borders, in particular data.  And a construction new on me: the ‘digital stack’ or digital public infrastructure which India and Estonia are already making use of.

My own involvement with the ONS on consumer price statistics made the discussion of prices especially interesting.  We are all familiar with the problems around measuring quality, or the arrival of new products.  Perhaps less familiar is the problem with assuming households can simply substitute one good (or service) for another – which tends to be less true for low-income families.  There is a fascinating table of how many hours of work are needed to buy different products at different dates – items which have risen in ‘price’ on this basis may be higher quality – but with the cheaper item no longer available, this isn’t much comfort to those on low incomes.  Real-terms data does not fully capture changes in welfare.

Coyle argues that there is a major failure in focusing on GDP as it ignores what is happening to wealth, and we ought to pay much more heed to the loss of wealth happening at the moment due to the depletion of our natural resources.  Increases in wealth, fully measured, bring rising welfare.  The ‘Six Capitals’ – physical capital, natural, human, social/institutional and knowledge should comprise our wealth, but at present most are not measured well. Or they are used too simply: in particular, surveys of people’s well-being have been hastily used with the ‘same technocratic confidence as economists have long had using conventional statistics.’

The conclusion is that we need other statistics to capture ‘economic progress for society as a whole’.  Having dismissed the suggestions of well-being, a broader GDP measure, and dashboards, she recommends the comprehensive wealth measure.  Importantly, this builds in sustainability and also moves towards capturing ‘capabilities’.  An alternative, or complementary, measure would be to consider time-accounting – looking at how effectively (or enjoyably) we use our time.

These are complex proposals – but are responding to a complex and pressing problem, outlined effectively.  While many of GDP’s shortcomings are well-known, Coyle sets out elegantly and compellingly why these issues have now become so numerous, and so serious, that we should rethink radically how we measure our progress.