driverless crash testing

I see that the until-recently invisible David Davis has resurfaced to explain away that so-called sector analysis.

When it was first mentioned, I imagine a scene where he had to say something - anything. Perhaps he badly remembered a summary he'd seen? By mentioning it on the record he then had to produce a version of it. Perhaps by stalling for 2-3 weeks it gave him time to ask a roomful of interns to generate the necessary kilos of paper, which one of his team could then speedily redact?

My guess is that one of the 'paid pro-bono' (sic) consultancies produced the basis of the original analysis that Davis remembered.

If I had to plump for one I'd go with KPMG's work, which produced 57 snapshots as re-cuts of 19 industry segments.

The summary of the sector analysis looked something like the one above and was actually an amalgam of ONS and KPMG work. I know there's not 58 sectors in it, but it isn't too difficult to drum up 57 with a bit of table sorting.

Davis admits that he'd not read the main reports in any case, just a summary. He's also saying now that the summaries are qualitative, not quantitative. Although I suppose the use of facts or hard numbers in this situation might mean working more than three days a week?