Can we trust police and hospitals figures?

We have to accept that many statistics are estimates with a known margin of error — but it's a different matter for MPs to be told that police statistics have been deliberately rigged; and for the National Audit Office to announce that hospital waiting time figures don't stand up to scrutiny either.

The figures in question differ from surveys such as the Crime Survey of England and Wales. Hospital and police crime figures are drawn from administrative records: they start out as numbers recorded by busy policemen and women, doctors, nurses and hospital staff, go up through their management, and only then reach the independent statisticians.

There's scope for inconsistency while recording these figures, and MPs have questioned whether the collection of this data has been affected by performance-related targets.

The Crime Survey is regarded as a reliable indicator of crime trends. It's run from beginning to end by independent official statisticians under the supervision of the UK Statistics Authority. Everyone can see what they ask, who they ask, and how precise the figures can be.

In an interview on Radio 5 Live yesterday, Full Fact's Director Will Moy said:

"The UK Statistics Authority has responsibility for statistics that start off with people outside its sphere of accountability - police or nurses — but wasn't given the ability to say to the regulators of the police force and hospitals, 'You must be checking that this data is coming out well.' They weren't given the powers they need to really take responsibility for how trustworthy those kinds of statistics are. We need better backing for that authority."

Voters use these statistics to decide whether governments have made things better or worse. Regulators, hospitals and police use these statistics to keep us safe. We need a watchdog that has the means to uphold standards throughout the process: from where statistics begin at the front line right through to publication.