It could be that research and development is not concentrated in the right sectors, says Roger McKinlay, head of the government’s Quantum Technologies programme. “We need an industrial strategy, not just targets,” he says.

No, that’s not it, that’s idiocy. People who can’t choose who should be Prime Minister this week are not to be put in charge of what technologies should be developed over the next decade.

The Office for National Statistics has just admitted it has wildly underestimated private sector R&D spending, which it now reckons was £43 billion in 2020, almost 60 per cent more than thought. If true, this paints a very different picture. No longer a woeful laggard in this area, the UK is roughly average and slightly ahead of France.

But is it true? The ONS thinks its original numbers, derived from surveys, are less accurate than Treasury figures from R&D tax credit claims. Some economists are sceptical, suspecting that the Treasury numbers are too high because companies over-claimed.

We’d just remind of that Hayekian point, that the economy is really complicated, the centre never is going to know what’s happening so detailed planning is impossible.

But there is more here:

Many economists point to the weakness of all business investment, not only R&D. Spending on plant and machinery has been anaemic, despite the cut in corporation tax over the past decade. Yet for Britain’s services-heavy economy it is investment in intangibles that is increasingly key. And here too the figures are dodgy. The Bank of England said the “puzzling” weakness of UK investment data may be due to underestimates of spending on intangibles.

Investment is, by definition, spending that is amortised over more than one accounting period. That’s just what it is. Whatever is a simple cost in this period, right now, is a simple cost and current spending. Whatever is the buying of something that will be used over time is investment.

Which leads to two different observations.

Heard of SaaS? Software as a service? Microsoft Office or Microsoft 365? Same software, same usage, same producer same customer list- people have been migrating. Office is investment, you buy a licence and wait for the next upgrade cycle - a couple of years - to buy another. That’s amortised over the usage period, it’s investment. 365 is a monthly subscription fee. That’s current spending.

There’s more than just this Office thing as well. SaaS seems to be about $10 billion a year in revenue from the UK. $10 billion is an important number when we’re talking of £45 billion in investment. Yes, of course they’re slightly different things, the point is that measuring what is investment is difficult because it’s actually a rather arbitrary definition and tech is moving things across that divide.

This also being true in another sense. Write code these days and it’s considered very odd indeed to capitalise that. The cost of writing the code is - almost always - treated as a current expense, not an investment to be amortised.

That is, the more we invest in such intangibles then the less investment is actually counted as being investment. Again, this is not a perfect mapping, software and intangibles, but the point still stands.

The real answer here is not that the UK invests enough, too little, too much, is inefficient or efficient at doing so. It’s that we’ve no idea of even the number, let alone what it should be. Therefore we should stop faffing around with trying to tweak the number. On the very sensible basis that even all those really clever people who know how to kiss babies have no damn clue.

That’s the solution.