“It is no coincidence that the last time the number of prisoners fell, Mr Clarke was responsible for penal policy – and crime was higher than it has ever been before or since.”, Editorial, Daily Telegraph, 22 October 2010.

The Background

Justice Secretary Ken Clarke’s promotion of alternative punishments to short sentences has proved one of the more controversial Coalition policies, particularly with sections of the Conservative party.

So as it was announced that this week's Spending Review could mean a cut of 3,000 prison places for short term sentences, the row was reignited.

The move has rankled with disciples of Michael Howard’s ‘Prison Works’ doctrine, who point to a general trend over the last fifteen years of rising prison populations and falling crime.

It is such trends that are alluded to in today’s Daily Telegraph, but have they got their numbers right?

Analysis

Ken Clarke was Home Secretary from the 1992 election until he was succeeded by Michael Howard in May 1993. So as a rough approximation of Mr Clarke’s time in charge of penal policy, we will look at what was going on in 1992.

In the sentences preceding the quotation above, the fall in recorded crime is mentioned, so again this will be used as the yardstick

This snapshot gives backing to the Telegraph’s claim – to an extent.

Due to a change in methodology, recorded crime figures are not comparable up to the present day, but at least until 1999 when changes were made to Home Office counting methods, the highest year for recorded crime was indeed 1992.  The year also saw a reduction in the number of people incarcerated.

But to say that year was the “last time the number of prisoners fell” is not accurate.

The Ministry of Justice numbers show that between 1999 and 2001 the prison population also fell.  During this period crime rose from 5,109,089 reported incidents in 1999 to 5,301,187, but then fell to 5,170,843 in 2001.

So while there are questions over the wording of the Telegraph’s claim, there are also questions about the choice of statistics.

Given that the claim compares crime rates over a period that transcends changes in recording methodology, it seems that a fairer way to measure crime over time is the British Crime Survey. Indeed the BCS has been backed by head of the UK Statistics Authority, Sir Michael Scholar, as the best data source for doing this.

If we use the BCS to gauge crime rates, a slightly different picture emerges, as the graph below shows.





Against the same trends in prison population which began rising after 1993, save for the hiatus mentioned above, BCS crime peaks in 1995, two years after Ken Clarke’s stint in charge of prisons policy

Of course the possibility of a time lag in the data between Government policy, prison population and crime rates, should not be ruled out - in which case, neither should it be ruled out for Mr Clarke’s solitary year at the Home Office.

Likewise, simply looking at the figures like this rules out any other factors, such as police numbers and strategy, and economic conditions on the rate of crime.

Pinning the crime rate for one year on the policies and the man who was Home Secretary for that one year only therefore perhaps seems a bit harsh.

Conclusion

Looking at the figures in context, it appears the Telegraph's assessment of Ken Clarke’s stint as Home Secretary is selective.

It is true that he oversaw prison policy in the last year before the number of inmates began a steady rise, and in this year recorded crime hit a high point.

But given that crime was rising in the years prior to 1992, while the prison population had been coming down, the situation in that year seems more attributable to previous incumbents at the Home Office.

Likewise a look at the BCS shows that, by a different measure 1992 was not even the highpoint of the early 1990s crime wave. Crime continued upwards to 1995 while prison numbers increased.

This is not to suggest one way or another whether ‘prison works’, but more to show that as the debate around prisons policy hots up that there are pitfalls in comparing two factors over time.

Consider also Full Fact’s comparison of police numbers, spending and crime rates since the 1980s.

Yes crime has fallen since its early- to mid-nineties highpoint, while more criminals were locked up, but linking this trend to one policy, or even one person, stretches the point.

It may not be complete coincidence, but nor is there direct cause and effect.