I am always sceptical about figures used by government to support policies. This is because of my experience on a course at the Civil Service College that I have mentioned in another thread, where we were split into two groups, provided with identical packs containing sets of genuine official statistics and a policy statement, one group being tasked with producing data showing that the statement was true, the other, that it was false. Real-life data is often contradictory, so it was a question of selecting the data appropriate to your task, or, when this was difficult, massaging the data to make it look as if it supported your task at a superficial glance, (for example by producing a graph showing the rate of change rather than the change per se, making it appear that something was reducing with time, when it was actually increasing, but less rapidly). Both groups managed to produce impressive-looking figures, not all of which would have stood up to a detailed analysis by someone who actually understood how to interpret them.
The exercise was intended to illustrate a frequent real-life situation in politics, where policy is made first, often purely on idealogical grounds, and civil servants are subsequently instructed to produce data supporting the policy.