On August 12, the UK ONS published data showing UK GDP falling by 20%. This makes the UK the one of the worst hit of any comparable countries. Is this believable?
The GDP fall is huge and it is partly a result of the terrible handling of the crisis by the Johnson government. The failure to control covid has had a disastrous economic impact.
The results have badly hit transport, schools, shops, offices, restaurants etc. But almost everywhere else in Europe has been hit too. So how can the UK be reporting twice the decline in output compared to the likes of Germany on higher than France, Spain and Italy?
The contrast is partly one of timing. The UK locked down nationally and late so the covid effect has hit the UK later. There is a greater impact in Q2. Spain and Italy had problems spread over Q1 and Q2.
But it is also due to how differing countries measure their economies and specifically how they value the output of things like health and education. In theory all advanced countries follow the same rules – especially those in Eurostat land.
Everywhere things that are not bought or sold are difficult to value in GDP. To get around this most countries simply add up the inputs (wages of teachers, costs of buildings, books etc) and say that is the value of the output.
So if a teacher’s wage is say £35,000 then this equals £35,000 of output.
This is a simple solution to valuing a non market output but it is not very satisfactory. (1) So there have been various attempts to improve on this.
The ONS has developed specific measures of output for education and health. In education, for example, it looks at numbers of pupils, teaching hours worked, exam results to directly measure the output of the education sector. The UK seems a statistical outlier.
All European economies that closed their schools kept paying teachers salaries and other running costs so those who did not adjust recorded little to no decrease in education.
The UK on the other hand measured how many children were still actually attending school and the actual hours worked by teachers and estimated that the output of education had fell by 34%.
As education is roughly 6% of the economy and health 7.5% these falls contribute hugely to the decrease in UK GDP. Together they represent 4 % points of the 20% GDP fall.
But some countries will have recorded no decrease in health and education. Some may even have recorded an increase in health output from extra spending. The question is which is which?
What does this mean?
If we look across both Q1 and Q2 and control for the way UK measures health and education then the fall in UK GDP is much more in line with the likes of France and Italy.
This is still terrible but nowhere near as terrible as it seems in current international comparisons. It does not mean that the UK approach is wrong either. That is another argument. It simply means it may not be comparable.
Why is nobody noticing this?
I have always had an interest in data and have discussed this issue with several people. The question is why does it take an outsider to ask questions?
The answer is that too many economists pride themselves on their maths and running tests on big data sets but not on actually looking at the data and its reliability or comparability.
It is worrying that the economic profession including those who once sat on the BoE MPC and decided interest rates do not seem to understand this.
Bluntly most economists don’t have much of a clue about the data they use.
If you are a true nerd
The GDP deflator is an alternative measure of inflation in the economy. It is the result of dividing nominal by real GDP. The current GDP deflator is hugely impacted by the health and education measures as the ONS acknowledges. This is no indication of inflationary pressure in the economy.
(1) You can find a discussion of this and other measurement issues in chapter 3 of my Productivity, Agenda, 2020.
This is excellent and needs to be read by a great many people (particularly those who are confused by the implied deflator). The subtlety is that a lot of it won’t show up in current price (nominal) GDP as ONS share a similar approach to others using the sum of costs. Volume GDP is where the divergence happens and hence the strange implied deflator.
[…] this is not so for G. As was explained in a previous post the rules of accounting for government output allow for countries to apply differing methodologies. […]
[…] reflect the falls in the output of the health and education sectors (a point explained well here by Mike Haynes). Most countries in Europe base their estimates for these sectors on the fact that […]