What’s the point in doing studies if you know better than the studies?
The Government did something clever: it started a few small-scale pilot trials of a new policy. It was a policy with clear goals, so outcomes could be easily measured. If the trials improved those outcomes, they could be reasonably sure that it was a good policy, and it would make sense to roll it out nationwide. That is an excellent way to govern, because that is how you work out what works and what doesn’t.
The Government, or more specifically Chris Grayling, the Justice Secretary, then did something startlingly dumb, according to The Economist: it rolled out the policy anyway, before the results of the pilot trials came in. Because, it seems, Chris Grayling can see the future.
To be honest the specific policy doesn’t really matter. But for the record, it was payment by results for new private- and voluntary-sector parole services. Essentially, the new groups would be paid more if the people who passed through them didn’t reoffend as much.
It sounds very sensible and market-solutiony, which is good. But it might end up with perverse incentives, leading to groups cherrypicking the simplest cases. Which is why running the pilot schemes was a good idea. In an ideal world, they’d have run randomised controlled trials on it, but in general, testing your policies – having the humility to realise that you aren’t omniscient, and that not everything that seems like a good idea really is one – is a good idea.
So it’s doubly ridiculous to set up a trial (as Mr Grayling’s predecessor Ken Clarke did) and then, when the trial is running, decide you know the results of that trial in advance (as Mr Grayling did) and launch two-footed into the policy.
The Economist has the rest of the story; go read it. I hope payment-by-results works. But whether it does or does not, Mr Grayling has significantly reduced his odds of making the right decision. So bravo for that, Chris.