Getty Images/EyeEm
If we're measuring GDP wrong in a digital world, what does that say about the economy?
Conventional measures of economic success such as GDP fail to account for the nature of free digital services. How different would the UK economy appear if that changed?
Our modern economy is stuck in the slow lane. In fact, it’s truly in the pits. Growth is slow to the point of being invisible, wages are for many lower than they were a decade ago, productivity numbers are dire and Lord knows we’ve a problem with the worst inequality since Dickens were a lad.
That’s a reasonable summation of the reports we see in the newspapers, the common refrain across the spectrum from the OECD, IMF, Jeremy Corbyn and even the prime minister.
It’s also something that really shouldn’t be happening. For we – and it’s us, in tech, driving it – are in the middle of a technological revolution. The two things simply aren’t consistent with each other, and one of them must be wrong.
We cannot be having lovely new ways of doing things, and also have low productivity growth and low real wage growth. It’s like saying we’re going down uply, something that only happens in Escher drawings – and economics is weird but not that weird.
So, something is wrong, as I’ve mentioned here before. One answer is that we’ve just not got that tech revolution going on, and so productivity isn’t improving. This might be true of NHS IT but it’s most unlikely to be true of all IT out there.
The other explanation is that we’re counting wrong – as we obviously are in part when we measure WhatsApp as having no output, no value, and in fact as a decline in productivity. At which point there’s a new paper out trying to measure these things.
Digital value
According to research by the National Bureau of Economic Research, search engines are worth $17,530 a year per person, email $8,414 and digital maps $3,648. Even Facebook is worth $322. Those numbers are, of course, a joke, and a little titter at their spurious accuracy is appropriate. But given the best techniques we have they’re about right to the number of digits and the size of the first one.
The researchers – good ones too – have used the same techniques we use to try to work out the costs of climate change, the value of parks, even the statistical value of a human life. That last one being important as it’s what we use to decide whether a specific piece of accident prevention is worth it or not.
We can’t observe what people pay for these things because, obviously, they’re not doing so. But we can ask, well, how much would you have to be paid to not have them? Ask many people, many times, with different price ranges, over a period of time and that’s the best we can do. These numbers are not right, not in any detail, but they are correct.
It’s important to realise that GDP is only a proxy for what we want to know. It’s the measure of all value added at market prices. What we want to know is what value can people consume, and our GDP definition clearly leaves out things that don’t have market prices.
Feminists have made hay with this distinction for decades as it leaves out all household labour – and they’re right to complain. But it also, quite obviously, leaves out what we get to do or use for free.
Read more about the digital economy
- How will the UK supercharge its digital economy?
- Experts from across UK technology sector will help government implement digital economy policies
- The UK has been ranked among the top 10 nations in terms of digital capabilities
Well, not quite. Google is in GDP at the value of advertising it sells per user – that’s a market transaction we have a price for. That value is maybe $25 a person a year, something we’ll note is different from $17,000 or so. Even if we pay for our email it’s not more than $250 a year which is again less than the $8,000 plus change value, isn’t it?
Yes, there are always things which work this way, things of benefit not included in GDP. But as a general rule of thumb we say they’re of about the same value as GDP. So, total value consumed is 200% of GDP as recorded when we include that consumer surplus – the value we get without having to pay for it.
If that value we get is very much greater then that solves our conundrum of where all the growth is. In fact, if that consumer surplus from digital tech were 10 times, not double, what we’re recording in GDP then everything else falls neatly into place.
No longer is value creation growth low to non-existent. It’s only that we’re not counting it in GDP. Nor are real wages stagnant – for the only useful definition of real wages is what we get to consume. As you can see, the best estimation we’ve got is that digital tech is worth very much more than that threshold we’re seeking.
Key metrics
Our other key metrics are derived from value created. Productivity is value divided by the hours of labour used to create it – if value is greater then so is productivity.
Inequality measures how that value is distributed. And we all have equal access to those free things meaning that this extra value identified is equally distributed across us all – in theory – which brings the societal measures on actual inequality down considerably.
A similar exercise by the TUC some years ago measured the effect of free government services, for example schools, the NHS and so on. Real inequality fell from 12:1 between the top and bottom 10% of households to 4:1. Maybe that’s too much, or maybe it’s too little inequality, but that’s certainly different and an effect from including items free at the point of use.
There is no doubt within economics that if we are under-measuring the value of digital goods then all of those other numbers are wrong - incomes, productivity and inequality. The only question is, are we and if so by how much?
The major question in political economy for our times is what are we going to do with stagnant incomes, tepid productivity growth and soaring inequality? The answer might well be just start measuring them all properly. Instead of, as most plans currently on offer seem to assume, completely rebooting civilisation to deal with a problem that may well not exist.
