One of the big problems in this information age is the use of heuristics, or short cuts (a point made very eloquently by Daniel Kahneman in Thinking Fast and Slow) which can lead us to make decisions on what is very often low information data dressed up as the opposite. Perhaps the most obvious from a markets’ perspective is the biggest number of all, GDP. This is not so much a number as a Totem, the higher the GDP number the ‘better’ off we all are apparently, so not surprisingly those rewarded for high GDP numbers – politicians usually – will do whatever it takes to flatter those numbers – and certainly make sure they at least look as good as possible. However, amidst all the busy hands working to manipulate the Covid data, they missed one trick, which is that the UK measure GDP in a slightly different way to everyone else. Normally this doesn’t matter – and is actually a sensible and more accurate way of doing things – but in Covid, it has made a meaningful difference and it thus appears that the UK is doing a lot worse than other countries.

This is all explained by the Government Statisticians in this note back in May. In brief, what it points out is that all countries struggle with measuring the value of output for something that is essentially free – like education or healthcare – and as a result most countries simply equate value with cost of production. If we think about this for a moment we can see the issues this raises with services and with the public sector in particular. To use an old example of public job creation – hiring two teams of people, one to dig a hole in the road and another to fill it in can create GDP ‘growth’. Even better (or worse depending on your viewpoint) if you give them all a pay rise, the cost – and by extension the measured output – also rises. Now obviously this simplistic example doesn’t really happen in the real world, although hiring teams of bureaucrats to collect data and monitor it and others to publish reports on it comes close, but it is a huge issue in areas like public health and education. Hiring more teachers or nurses does not, of itself, increase the value added in health and education. Nor does a pay rise. As a result the UK government statisticians have been trying to create other measures of output for the last decade or more – a fact which may well explain why countries with large and growing public sectors (eg France) appear to be growing faster than the UK in recent years.

All this has meant that during Covid, while the measured cost of these public services has not gone down, in anything approaching the real world we should concede that the value of the output most certainly has. Fortunately, to our mind at least, the good men and women of the Office for National Statistics have tried to allow for this and, as the linked article points out, in education at least, the output has fallen, even if you include remote learning by up to around 30%, something which has been seized upon by anti government campaigners at home and abroad as a demonstration of how badly the UK has dealt with the crisis. While not excusing them in the slightest from culpability in the economic downturn, this is one area that they can not be blamed for.

Does this all matter though? Well aside from the fact that the ‘surprise’ recovery in q4 (assuming a return to school) will likely be dismissed as ‘fake news’, it should lead us to reassess our dependence on single figure press releases that very few people understand and even fewer bother to do the research on. Arguments that x or y are good policies because they will increase GDP need much more detailed scrutiny. GDP per capita is more important for example, as is the balance between state and private sector and the ability to measure utility or value added needs to be seen as more art than (fake) science. The old expression about ‘teaching to the test’ will always mean that Politicians and civil servants will skew policy to that which is being measured and by which they are being judged/rewarded. To the extent that part of that measurement/reward is the positive response of markets to macro data, it is incumbent on markets to also make sure they are looking at the right data.

‍