The position of GDP growth as the be all and end all of economic policy may be coming to an end.
Mathematically, GDP is the sum of consumption, investment and government spending—plus exports, minus imports. Since World War II, it has been used as a measure for the growth of an economy. And it was generally assumed that economic welfare correlated with GDP growth.
But it has unravelled on both counts. People are seriously questioning whether GDP captures all of what matters in the modern economy. Then there’s the longstanding critique of GDP as a meaningful measure of progress: it says nothing about sustainability or wellbeing.
And what the state does not see is invisible in policymaking. Statistically invisible phenomena do not feature in political debate until they become impossible to ignore—at which point it may be too late.
In October 2017, Diane Coyle and Benjamin Mitra-Khan were joint winners of the Indigo Prize for their essay, “Making the future count”, in which they reimagined economic statistics. Here, we published an edited transcript of a conversation with Professor Coyle, an economist at the University of Manchester.
What does GDP actually measure and where did it originate?
It measures the total amount of economic activity in a country, or a geographical region, and it can be defined as the sum of all income, all spending or all output—they should be the same, although in practice they never are. It originated during the Second World War as a tool for the American and British governments to know what was going to be available for the war effort and how much consumption needed to be curtailed to allow all those resources to be reallocated. So it had a very specific origin, and then it was shaped over the postwar years for the kind of economy we had then: mass production, much more focused on manufacturing.
Why did it stick?
Partly because it was a technical standard. Everybody agreed to it through a United Nations committee process because they wanted a standard measure that would allow comparisons across countries. And once you agree on a standard, it does become sticky. So part of it is just the political economy of how these things come about. And the other part is that it seemed a very suitable measure for what policymakers wanted to know at the time. It had a very clear link to the tax base, and it told them about the kind of growth they were interested in.
Why is it no longer so relevant?
Some of the criticisms are longstanding; the environmental one goes back to the 1970s. And throughout there have been criticisms of the fact that economic activities in the home – which do not have a market price – are not included in GDP, even though other, non-price sensitive government spending is.
But I think the reason it has become such a focus now is probably two-fold. One is that the environmental cost we paid for growth after so many decades is now looking quite serious. Even if you do something that is harmful, like polluting, there is no reduction in GDP to account for that externality. The other reason is that the structure of the economy has changed so much. It’s already 80% services in the UK, and there are similarly high proportions in other developed countries, and the concept of productivity and a product doesn’t really make sense in the services sector in the same way as it does in manufacturing.
The digital changes have also been key: changes in business models and behaviour that simply aren’t being captured by GDP anymore. For example, there’s the fact that we get all these free goods on our smartphones instead of having to buy the watch, the diary, the maps and so on that we used to. A lot of the vocal criticism of GDP right now is coming from [people in] the tech sector, who really doubt that productivity is actually flatlining because they know that they are delivering so many product and service innovations.
So that’s the problem. In your essay, you set out a two-tier strategy for dealing with it. The first part involves immediate modifications to GDP figures.
We suggested three different things.
One is to do a better job of measuring intangible investment, because the great majority of what companies invest in now is not the physical, the machines and buildings that are reflected in GDP, but the intangible. It might be creating software, or creating reputation, brand values, designs and so on. All of those need capturing.
The second is paying attention to income distribution by, instead of looking at average per capita GDP, looking at the median per capita GDP. And, of course, the median is much lower than the average when you have some people with very high income.
And then the third thing is taking out a change made last time the standard was amended that put financial speculation into GDP. We think that was a big mistake. A lot of the boom before 2007/2008 was due to this artefact of measuring the risk-taking by the financial sector, and that needs to come out.
And these are all changes that could be done pretty quickly.
The second, long-term part of your strategy is to replace GDP entirely with a balance sheet of six new measures.
We recommended a balance sheet of six forms of capital. Physical and financial capital — pretty obvious. Intangible capital, which I just talked about. Then natural capital, social capital, and human capital.
We have to look at it like a balance sheet, because if you care about sustainability, then that’s the only way you’re going to give future generations a voice in policy decisions now. It will also get round problems like fact that a natural disaster that wipes out lots of assets is great for GDP because of all the rebuilding afterwards.
It’s clear that those measures would be useful, but how would you quantify them?
This is the difficult part. There are two aspects to the problem: we don’t have the raw data that we would need, and even then we don’t know how we would put a value on it. I’ve been looking at natural capital with the [UK’s] Office of National Statistics and it’s partly just a question of raw data, like collecting air quality data or looking at the amount of forest cover in different geographies, and we don’t have all of that. Then you have to value it, and that’s tricky with some of these forms of capital, not least because you have non-linearities and tipping points and we don’t yet have the techniques to deal with that. That’s why we suggested an interim step.
To your eyes, what are the valid defences of GDP?
A while ago I would’ve said that it was a good enough indicator of what we need to know about the economy. But I’m just less and less confident that it tells us anything very useful. I suppose it’s a known entity. And if you want to create a single index of the economy, then you want to be careful about how you weigh everything together, and with GDP we know what weights we are using—they’re just market prices. Lots of alternative indices don’t even think about weights, or just mush things together without thinking about the implications of what their implicit weighting schemes are. In essence, we understand what GDP is telling us. That’s its great advantage.
Other countries and organisations have tried to get beyond GDP in the past. What do you think of things like Bhutan’s Gross National Happiness, or the Social Progress Index?
I don’t think directly measuring wellbeing is very informative. It’s not a good way of holding policymakers to account, because it doesn’t change very much over time. And it’s surprisingly similar in different countries at different levels of income because of what psychologists call a hedonic fixed point: people just get used to how things are. So looking at a measure of happiness can find, oh, people are six out of ten on the scale, therefore we don’t need to do very much – even though they might be living in very different countries.
The Social Progress Index isn’t an economic measure. And the trouble with alternative indices – the OECD Better Life Index is another – is that there’s such a proliferation of them, and if you don’t have something that’s rooted in economic theory then you get lots of different dashboards with lots of indicators. Then it’s very hard to get a message out of those. It’s very important to get a consensus and we opted for a relatively small dashboard of measures that has a clear message. After all, one of the purposes of statistics in a democracy is to hold the government to account.
(Picture credit: Flickr/Phrawr)