You know better than that, Nigel. You are the one insisting we look at long term trends, and then you want us to extrapolate on two adjacent measurements?

But if we really want to split hairs, there is not really enough information, because the figures are rounded to one decimal place and they each have a margin of error, so it is theoretically possible, albeit unlikely, that the growth is actually the other way round, if we had the underlying actual figures once everything is known.