The end of GDP?
During my first four years of high school, I delivered the Afrikaans daily Die Burger to about 50 homes in my neighbourhood every morning. For a fee, subscribers would have access to national and global world news over breakfast. Two decades later, nothing much has changed; a printed version of Die Burger is still delivered to subscribers. At the same time, everything has changed: there are far fewer subscribers to printed newspapers – the circulation of the Die Burger has halved since my newspaper delivery years – but far more consumers of digital news, including myself. I don’t buy a printed newspaper anymore, I read the news online.
The conundrum for national statisticians who are interested in measuring the living standards of South Africans over time is that the decline of newspaper sales will register as a decline in the Gross Domestic Product, the widely-used indicator for the general prosperity of society. And if people do not pay for their online news service – but instead get the news from free services like News24.com or the Huffington Post – then there will be no concomitant increase in GDP. What would appear like lower levels of living standards will in reality be significantly improved welfare; no need to get out of bed for a sometimes-late newspaper soaked in an unexpected rain shower when you can pick up your phone and read up-to-the-minute news in the comfort of your bed, for free.
This, in essence, is the problem with using Gross Domestic Product as an indicator of living standards. To be sure, GDP raises many problems that economists have known about for a long time: when Simon Kuznets (who is most famous for his work on inequality) first designed the measure in the 1940s, he was unsure whether to include home work in the measure. If you appoint a gardener, for example, it will be counted as part of GDP. If you instead choose to do the gardening yourself, it will not be included. (The obvious conclusion: If you marry your gardener, GDP will fall!)
But the rise of ICT – or what Michael Jordaan, former CEO of FNB and now CEO of Montegray Capital, has termed the ‘free economy’ – has made matters much worse: our ability to connect with friends (Facebook), or make international calls (WhatsApp) or study any topic on earth (Wikipedia) free of charge has without question boosted our living standards but won’t be reflected in GDP. It may well have resulted in a decline: we buy fewer albums, address books and encyclopedias. We may not be consuming more, but we are certainly consuming better, faster, and a greater variety with more comfort.
The failure of GDP to capture these advances is beginning to worry economists. The April 30th edition of The Economist calls for a ‘fresh approach’. Several new books by prominent economists – of which Diane Coyle’s ‘GDP: A Brief but Affectionate History’ is my favourite – have noted the need for change.
Measuring the improvement in living standards accurately is important, for at least two reasons. First, it has an influence on the political process. Weak GDP growth and opposition parties have more ammunition. But a focus on incomes only may miss the large improvements in living standards derived from both free public services and technology. Even the way we measure poverty may need adjustment to this new world of freeconomics.
Second, and often neglected, mismeasuring the improvement in living standards also implies a mismeasurement of the causes of that improvement. Poor growth suggests that technology has had little impact on our welfare. This has become known as the productivity paradox: Nobel-prize winning economist Robert Solow famously quipped: ‘You can see the computer age everywhere except in the productivity statistics’. The remark was aimed at the inability of computers to have the cataclysmic effect of the dynamo or the steam engine of an earlier age. But perhaps it did (and still does), but we are simply not measuring it accurately. I suspect the computer age has added far more ‘quality’ to our lives than what has been measured in GDP estimates.
Of course there have been many attempts at constructing a better measure of living standards. A few years ago, ‘happiness’ was touted as a better measure of general welfare. The problem is that, beyond a certain level of development, happiness changes very little. So too for light density, a measure of economic performance that rely on images from satellites. (Light density does have its uses, though: A new paper by Maxim Pinkovskiy and Xavier Sala-i-Martin in the Quarterly Journal of Economics use light density to show that living standards in developing countries have risen much faster, and that the world income distribution has become more equal, than previously thought. This is because the quality of surveys in these countries is often poor.)
A new NBER working paper by Charles Jones and Peter Klenow – Beyond GDP? Welfare across countries and time – is another attempt to measure the general improvement in living standards. They construct a composite indicator that consists of consumption, leisure, mortality, and inequality – the four main things, they argue, that determines welfare. Although welfare and GDP are correlated, there are large deviations – including South Africa. We rank much lower on welfare than we do on GDP because of our high level of inequality and low life expectancy.
And yet, even this composite measure fails to account for the immense gains from technology in the Digital Age, like free news, communication and information. GDP will remain a very useful measure, but it is not invincible. Politicians and their electorate should take note.
*An edited version of this first appeared in Finweek magazine of 19 May.