‘The Ministry of Justice continues to mislead, referring to ‘reoffending’ when it means ‘reconviction’’. So says Richard Garside from the King’s College Centre for Crime and Justice Studies (via twitter). He’s right. Reconviction rates show us when people get caught and convicted of committing a crime after being convicted previously: reoffending rates (which are usually much, much higher) show when people commit a crime after having committed a previous one.
The problems with the language of reoffending rates don’t stop there, however – which leads me to want to share some points with you that are pretty important when it comes to understanding crime.
Currently, in most developed countries, governments measure reconviction rates (even if they do wrongly call them reoffending rates). There’s no fixed international reporting standard, however, so reconviction rates are still measured in a number of ways. Now having just trawled through the latest report on “reoffending” in England and Wales, I can tell you that these different ways of measuring “reoffending” are certainly not transparent. Still, the key variables when measuring reoffending rates are:
- The time period for reconviction. In the UK, this is 1 year (although there is also a “six month waiting period” to pick up crimes committed within that year but not logged until later). In the U.S. data is gathered rarely (see details of available Federal data here) but one of the most comprehensive exercises measured reconviction within three years.
- When the measurement period starts. In the UK, the reconviction period starts when a prisoner is released from custody or when an offender starts a sentence in the community. In much U.S. data, reconviction data is only collected on those who have served prison sentences.
- Whether the measure is a ‘yes/no’ measure or a ‘frequency’ measure. A yes/ no measure just tells us whether offenders have been reconvicted of any other crime during the follow up period. A ‘frequency’ measure tells us how many crimes are convicted in the follow up period. In the UK, we measure both but the media tend to refer to the simpler ‘yes/no’ statistics, as it’s easier to understand. Still, the measures tend to correlate quite well anyway. If yes/no reconviction rates are falling, so generally will frequency of reconviction fall.
- What counts as a ‘crime’/ ‘serious crime’. Different countries count different acts as crimes and some countries (including England and Wales) have additional measures for assessing ‘serious reoffending’ – again meaning reconviction for a serious offence.
To state the obvious, different choices on each of these measures make it almost possible to assess reconviction rates (or reoffending rates) fairly across countries. If a country is measuring reconviction rates over shorter periods, starting the measurement period from the date of conviction and only including reasonably serious offences, its reconviction rates are likely to be lower (relative to actual reoffending rates) than countries with more inclusive measures. So let’s just look at the UK reconviction rates – and see how we’re doing. Source: MoJ 2010
OK, I see you thinking. Great. A pretty steady fall in reconviction rates since 2002. How wonderful that we’re getting better at rehabilitating offenders! Erm… maybe. But (as ever) it’s a bit more complicated than that. Reconviction rates will vary depending on:
1. Levels of crime. Generally, lower crime equals lower conviction and (therefore lower reconviction) rates, all things being equal. Falling crime rates should mean (all things equal) falling reconviction rates. This is of course not true if falls in crime are entirely due to new people coming into the system who don’t commit crime – but most analysis suggests that this is very rarely (never?) the sole reason for crime falling. Crime in the UK has fallen significantly in recent years.
Verdict: Falling UK reconviction rates are partly due to falling crime
2. Sentence lengths. In general, people commit fewer crimes as they get older. Why is much debated but it seems most people learn eventually that committing crime is not a very good way of having a happy life – and that middle aged social habits aren’t as conducive to criminal activity, and particularly the kinds of criminal activity that get noticed. If sentence lengths are rising, people age more in prison (where they can’t be reconvicted – unless they commit a really serious offence) so they are marginally less likely to offend if they receive longer sentences. In the UK, there has been a general trend towards more punitive sentencing.
Verdict: Falling UK reconviction rates might be due (in small measure) to increasing sentence severity, and are almost certainly influence by changes in police practices.
3. The use of formal sanctions & the general efficiency of the criminal justice system. Police and the wider criminal justice system have discretion over how they classify misdemeanours and crimes. A shift in favour of more formal procedures will lead to higher reconviction rates. Targets can have a big impact on how cases are processed and the strong focus on targets in the UK in the 2000s almost certainly pushed up the likelihood that an offence, if encountered by authorities, would be dealt with formally. The ‘offences brought to justice’ (OBTJ) target also had the effect of encouraging the police to focus on minor low level crimes as all crime were given equal weight. From 2008 onwards however, this target was modified to give more weight to serious offences and the focus on OBTJ has since dropped somewhat. Effective police, fast court procedures and a well funded prosecution team will all increase the proportion of crimes that end up with a conviction – and therefore the number of people (re)convicted. The best estimates would be that the criminal justice system is getting more efficient and effective: certainly it was spending a lot more money each year until 2010.
Verdict: Up until at least 2008, police activism and improved CJS effectiveness almost certainly inflated reconviction rates. It is possible that from 2008 onwards reduced focus on minor crimes contributed to falling reconviction rates.
4. The mix of crime. Some types of crime have lower reconviction rates. For example, reconviction rates for murderers are very low (around the 2% mark in the U.S.). This is partly due to the fact that such crimes earn a severe sentence (so there is more time for offenders to ‘grow out’ of crime) and partly because some murders are ‘crimes of passion’ territory – you only usually have one wife you can kill.
Verdict: It’s difficult to say if changes in crime mix have played a role in falling reconviction rates
5. The concentration of crime in specific cohorts/ groups. Research from a number of studies shows that just 4 to 6% of people commit upwards of fifty percent of all crimes, according to self-report records. These studies are from different countries, which suggests that these proportions are relatively consistent, but we still don’t know how much the concentration of crime actually varies internationally or how the concentration of crime varies over time.
Verdict: It’s unlikely that there have been changes in the concentration of crime that have had a knock on effect on reconviction rates
6. The extent to which police target ‘known offenders’. Across the world, police to a greater or lesser extent focus their attentions on offenders known to them. There are also policies directly targeted at monitoring ex-offenders, for example Multi-Agency Public Protection Arrangements (MAPPA) for ‘dangerous’ criminals in the UK. The impact of increased monitoring could either reduce reconviction rates (either by deterring criminals or supporting their rehabilitation) or could increase them (by ensuring higher detection rates for their subsequent crimes).
Verdict: It’s not clear whether targeting of ex-offenders has had an impact on reconviction rates
7. The ability of offenders to resist the temptations of crime. OK, so only now I get to this but we do know that some rehabilitation programmes can make a difference in reoffending rates. Typically decent long-running programmes targeted at offenders’ needs reduce reoffending rates by about 10% – though some programmes claim an even greater impact. There has been somewhat increased investment in rehabilitation in the 2000s (particularly drug treatment for ex-offenders) even if the provision of good quality services remains the exception rather than the rule, particularly for short-sentence prisoners.
Verdict: Offender rehabilitation programmes are likely to be improving slightly – which may be contributing a small part to the falling reconviction trend shown
So, what’s the point of all this. Well, basically, it’s that looking at reconviction rates in simple terms might not tell you as much as you think it does. The most important points that are missed by casual observers are points 1, 2, and 3. Lower crime societies will usually have lower reconviction rates even if they’re doing nothing particularly special on rehabilitation. And more punitive regimes focused on imposing formal sanctions for bad behaviour will almost always have higher rates of reoffending, even if they are doing a pretty good job on providing rehabilitation programmes to ex-offenders.