C. Northcote Parkinson explained this well over half a century ago:

Indeed, my sense is that as the number of bureaucrats expanded at Sport England — the body set up to allocate billions to create a leaner, fitter Britain — the vitality of grassroots sport declined……..We would later discover that more than 34 per cent of the budget was allocated to administrative overheads.

….

The point of this story, though, is not to reveal failures in sport but to offer a more general warning against those whose “solution” to the present crisis is more spending. Such people are victims of what we might call the input fallacy. They suppose that our commitment to any objective is measured by how much we spend on it. This is why politicians boast about lavishing “ten billion on sport” or “twenty billion on foreign aid” — a pure exercise in virtue-signalling. In the real world we should be concerned not with inputs but with outputs: what did the money achieve?

Quite so, quite so. Another pair of examples:

In a celebrated speech in 2012, Andy Haldane of the Bank of England pointed out that the UK had moved from employing one regulator for every 11,000 people working in the financial sector in 1980 to one for every 300 in 2011.

…..

Meanwhile, the global human resources industry grew from around $343bn in 2012 to $476bn in 2019, and the number of diversity roles has increased by 71% over five years.

It is, of course, possible that diversity, HR, financial regulation, are very important things. But to understand economic numbers we need to grasp that they’re all costs. They have no measurable output but they do have costs associated with them. Therefore they all reduce, in the standard economic numbers, productivity.

A common complaint these days is that British GDP is not growing as we would like it to. Entirely true of course. A corollary to this is that productivity is not growing as we would like. That’s also true. But we do have to grasp how we generate those numbers so that we can understand what might be causing those disappointments.

One thought is simply that we’re being eaten by bureaucracy as above. We’re carrying the costs of the papershuffling and gaining nothing in associated output. Yes, it might be true that there’s value to what is being done but our standard economic numbers don’t include that. For GDP is output at market prices. There is no market price to the output of HR, diversity and regulators - thus there’s no output associated with those inputs. That’s definitely one reason for slower GDP growth.

Please do note this still leaves open - we don’t think so, not at all, but it is logically possible - that this is all worth it. But it won’t appear in the GDP numbers. And given that, it will be measured as a fall in productivity.

Two further examples. We’ve often used the example of WhatsApp. There was a period when there was no fee to use it, no advertising upon it. As Hal Varian has pointed out, GDP doesn't deal well with free. Here, with WhatsApp, output is zero by the conventional GDP accounting. Yet Facebook employed (we asked them) a couple of hundred engineers to run that service. We have costs, those wages, with no associated output - that’s a fall in productivity. For labour productivity is labour hours in compared to market value out.

WhatsApp appears in national accounting as a fall in productivity. And yet 1 billion people gain some to all of their telecoms needs for free from WhatsApp. Which is, we repeat, in the economic statistics as a fall in productivity.

Given that GDP doesn’t measure what we get for free then the wonders of the digital revolution - all those things we get for free - aren’t included. Which would be one damn good reason for disappointing GDP and productivity numbers.

Our second is recycling. Yes, agreed, it is logically possible that we should be doing what we do, recycling things at an economic loss in order to protect Gaia. We do not agree, we insist that argument is tosh but the logical possibility is there. But now note the effect upon the economic numbers as we habitually construct them. Recycling requires considerable subsidy. The value of the shredded plastic output - as an example - is less than the cost of collection and shredding. That’s why plastics recycling requires subsidy, because it is not a profitable endeavour using the normal national accounting methods.

Now, what’s the effect of that on GDP and productivity? We’ve costs - those subsidies, that labour input - into recycling and a loss as the output. No, not because recycling is the very terror, but just because that’s the way GDP works. Recycling that requires subsidy turns up in national accounting as a reduction in productivity.

So, we’re complaining about GDP growth being depressed. We’re worrying about productivity not increasing as we would like. Yes, employing armies of bureaucrats will indeed be some part of it. Both the costs of employment with no associated output and, of course, the productivity reductions from having to obey all the red tape produced. But more than that those standard numbers just don’t deal well with free, those digital goods we now have in such abundance. And finally we’re insisting upon actions - like subsidised recycling - which directly reduce productivity.

As we say, it’s even possible that we should still be doing all of these things. Perhaps they make us richer in some manner not captured by GDP alone. But to be insisting upon all these things and then complaining about the effects upon GDP and productivity is to be insane. Which ain’t, as should be obvious, a good way to run a railroad nor an economy. Madness rarely is.

Oh, and blaming the lacklustre GDP and productivity numbers on neoliberalism, well, that is to laugh, no?