Ch 17. Punishment

As Mick Cavadino, one of Britain’s best-known penologists, observes, ‘It remains the case that the main method of ‘perfecting the art’ of sentencing is by practicing on actual offenders.’

Mick Cavadino and Jim Dignan, The Penal System 3rd Edition, Sage, London, 2002, p. 94. See also A. Ashworth, Sentencing & Criminal Justice, 4th Edition, Cambridge University Press, 2005.




The courts are unable to cope. Instead what happens is officially described as ‘a series of pragmatic responses to particular operational challenges’. That means hundreds of thousands of offences are processed through ‘out-of-court disposal’ and that in turn means they will not result in any formal punishment beyond a ticking-off or, at most, a small fixed penalty.1 Police cautions and out-of-court disposals have declined sharply but are not restricted to minor misdemeanours. According to the Magistrates Association, almost half of them involve repeat offenders, many of them violent.2

1In 1998 a range of out-of-court disposals was introduced for adults and youths and in the five years from 2003 such disposals rose from 240,000 to a peak of around 600,000, falling back to 400,000 in 2012. [] These mete out warnings, conditional cautions and penalty notices for disorder called PNDs. This ‘represented a series of pragmatic responses to particular operational challenges’ and was ‘a fundamental shift in how justice is delivered’. (‘Initial findings from a review of the use of out-of-court disposals’, A report by the Office for Criminal Justice Reform, February 2010, p. 2.) []

2 Letter to Justice Secretary Chris Grayling from John Fassenfelt, chairman of the Magistrates’ Association, January 2013. Two months later official figures released under a Labour Party Freedom of Information request disclosed that 10,160 ‘serious’ violent crimes had dealt with informally in 2012, including sexual assaults, malicious woundings, GBH without intent and racially aggravated cases. []




Around 50,000 cases a year lead to an absolute discharge or to a conditional sentence, which in practice is largely a promise of good behaviour.1 That means perhaps over half a million hours are spent by police each year in detecting and convicting crimes after which nothing happens.2

1In theory, with a conditional discharge if the defendant re-offends within the prescribed period he can be sentenced for the original crime as well. But in practice the promise tends to be overlooked on second or subsequent convictions. There are no official data on whether conditional sentencing has any effect on crime.

2 In 2004 there were 15,000 absolute discharges. The Metropolitan Police estimate (to author) that on average each absolute discharge case will have consumed twelve hours of police time, quite apart from court costs and time lost by witnesses. At a conservative estimate, across England and Wales this would amount to 180,000 police hours spent each year in detecting and convicting crimes after which nothing happened. This does not mean the experience might not salutary for some offenders – on the other hand, for others it might entrench a sense that they can get away with bad behaviour. In short absolute discharge is a costly but an entirely evidence-free disposal.




Actually, more than half of all punishments amount to nothing at all in practice. The main reason is that fines make up two-thirds of all disposals – and they routinely go unpaid.

Fines made up 66 per cent of court disposals in the twelve months to June 2012. This figure includes on-the-spot fines which, in spite of their name, are not collected on the spot and often not at all. ‘Criminal justice statistics: quarterly update to June 2012’, Ministry of Justice, 2012. []




Record-keeping is so poor that in 2011 the National Audit Office refused to sign off the HM Court Service’s accounts. Almost £2 billion worth of penalties were then unpaid, with less than half a million likely to be collected.1 Over £100 million of fines were being written off each year.2

1National Audit Office, Press Release, ‘HM Courts Service Trust Statement for the year ended 31 March 2011’, 15 December 2011. []

2 Freedom of Information request FOI-72392, Ministry of Justice, October 2011. [,d.d2k]




Ten years previously the NAO had found that fewer than a third of the financial penalties by magistrates in England and Wales were paid without enforcement and even after enormous costs in bailiffs’ fees a fifth of offenders could not even be traced.

Report by the Comptroller and Auditor General, HC 672 Session 2001–2002: 15 March 2002, The Stationery Office, London. The figure of £385 million includes fines, compensation and prosecution costs.




Complaining of the farce, Lord Justice Leveson, the top presiding judge in England and Wales, cited cases where an offender gave his name and address as a local war memorial and another who had collected eight fines for drunkenness, shoplifting and theft but which were ‘all unpaid with no real prospect of [the offender] ever being able to pay a single one of them’.

The Times, London, 24 December 2007, p. 10. LJ Leveson was quoted as describing unpaid fines as a sometimes like a farce which ‘far from underlining the importance of the law, risks bringing it into disrepute’.




… penalties are watered down because a lot of people can’t afford to pay them

The Criminal Justice Act, 1993, sets out maximum tariffs according to five levels of offence. The maximum fine for the worst (i.e. level 5) offence, e.g. stealing a car, is £5,000, except for a fourteen- to seventeen-year-old (£1,000) or anyone under fourteen (£250) but magistrates must consider the offender’s ability to pay.




In any case research would be difficult because information kept by courts is so ‘unreliable’.1 What we do know, from Scottish figures, is that 40 per cent of people who are fined are convicted of further offences within two years.2

1Ibid., page 9. []

2Reduce, Rehabilitate and Reform, The Stationery Office, Edinburgh, 2004. [] According to Audit Scotland in 2012 reoffending rates have ‘remained relatively static’. [, p. 5]




Ministers worry that voters think it’s simply a soft option and like to rebrand it as ‘Community Payback’ or pledge to make it more punitive. Hence the Justice Secretary pledged in 2012: ‘We’re today putting punishment back into community sentencing. This is about sending a clear message.’1 But essentially the successive governments insist it is effective2 and anti-prison campaigners heartily agree.

1E.g. ‘We’re today putting punishment back into community sentencing’, Justice Secretary Chris Grayling, 23 October 2012. []

2 ‘Community sentences can be an effective means of stopping people reoffending, because alongside punishment comes support to make the offender tackle the underlying factors that make them more likely to offend again.’ Five Year Strategy for Protecting the Public and Reducing Reoffending, Home Office, 2006. []




What we do know is that a third of offenders given supervision orders commit a ‘proven offence’ with twelve months.

Proven Reoffending Statistics, Quarterly Bulletin, January to December 2010, England and Wales, Ministry of Justice, December 2012.[] Proven reoffending is defined as any offence committed in a one-year follow-up period and receiving a court conviction, caution, reprimand or warning in the one year-follow-up. Following this one-year period, a further six-month waiting period is allowed for cases to progress through the courts. []




In Scotland, 42 per cent given community service and 58 per cent on probation are reconvicted within two years.1 In England and Wales, 75 per cent of offenders given tagging and curfews reoffend within a year of completing the scheme, compared to 69 per cent of all young offenders who serve custodial sentences.2

1Criminal Justice Social Work Statistics, 2004–05, Statistical Bulletin Criminal Justice Series CrJ/2006/01: ISBN 0 7559 2927 6, ISSN 0264 1178; and Reduce, Rehabilitate and Reform, Scottish Executive, The Stationary Office, Edinburgh, 2004. []. As you might expect, recidivism tends to increase with severity of sentence because sentences tend to reflect the severity of crime. Thus while 40 per cent of people fined were reconvicted within two years, the figure for those sent to prison was 60 per cent.

2Juvenile Reconvictions: Results from the 2003 Cohort, Home Office, 17 February 2005. This is consistent with other studies such as: Darren Sugg, Louise Moore and Philip Howard, Electronic monitoring and offending behaviour – reconviction results for the second year of trials of curfew orders, Home Office Findings 141, 2001.




But there has never been a head-to-head evaluation of community sentencing and prison. As Mark Henderson of The Times has pointed out, when it comes to sentencing, the legal system is ‘an evidence-free zone’.

Mark Henderson, ‘The blunt end of sentencing’, The Times, 5 November 2008.




Since courts and magistrates are inconsistent anyway, there is nothing to be lost in testing this systematically – and the public deserves to know what works and what does not. But ministers are rarely noted for their scientific literacy or political courage.

Unsurprisingly nor do civil servants feel able to stand up to ministers on this. In 2012 the chief statistician at the Ministry of Justice acknowledged, ‘Arguably, the best way of comparing the impact of two sentence types would be a Randomised Controlled Trial (RCT), randomly assigning offenders to each of the two sentences. This would enable us to control not only for observable and unobservable offender characteristics but also underlying systematic differences in sentencing and offender management. However, randomised assignment of offenders is difficult to implement with regards to sentencing decisions.’2012 Compendium of reoffending statistics and analysis, Ministry of Justice, 12 July 2012, p. 14. [] Presumably the political rationale is that we cannot randomise punishments (and worse, let some people off scot-free) because we dare not tell the public that we do not really know what we are doing. As a result we are like eighteenth-century physicians who refused to stop blood-letting (or those twentieth century oncologists whose routine answer to breast cancer was super radical mastectomy) on the grounds that it was obvious that it worked.




Around a quarter of those given community sentences get reconvicted while the average prisoner is still inside. That amounts to 20,000 extra crimes each year for which the offender is caught and reconvicted.

Figures obtained under the Freedom of Information Act, Daily Mail, 20 August 2012.




Since we know that convictions account for only a small fraction of the crimes actually committed, it is quite possible that hundreds of thousands of crimes are committed each year by offenders who have been left in the community rather than put away in prison. The crime scientist Ken Pease spotted this research bias back in the 1990s and complained about it to the House of Commons Select Committee on Home Affairs.

Home Affairs Committee Third Report Alternatives to Prison Sentences Vol. 2, Minutes of Evidence and Appendices, HMSO, London, 1998, pp. 203–6.




Around 15,000 front-line probation staff supervise over 225,000 offenders, including 13,000 officially regarded as at ‘high or very high risk of committing serious offences’.

Martin Wargent, chief executive of the Probation Boards Association, Today programme, BBC Radio 4, 21 March 2006. Probation officer ‘clients’ include those on parole including life prisoners out on licence, prisoners on home release and 12,000 ‘high-risk’ sex offenders.




Those on release are supposed to be under supervision, but this rarely amounts to more than a half-hour meeting once a week – if that.

In fact it ‘supervision’ is often a misnomer. In 2019 the chief inspector of probation conceded that a great deal of probation was merely a phone call conducted once every six weeks. Research by Manchester Metropolitan University found a ‘dearth of evidence’ that this did any good at all (‘Probation chief inspector: phone-in offender checks don’t work’, The Times, 13 March 2019).

For more insights, see an insider’s view by a former probation officer: ‘Unable to reform persistent criminals, the Probation Service has, over the last thirty years, become their champion’, David Fraser, A Land Fit for Criminals, Book Guild, Sussex, 2006, ISBN 1 85776 964 3, p. 230.




Even when offenders are obliged to live in bail hostels, the staff, however well intentioned, have no resources to tail people in their charge, no powers to search them for drugs or paedophile pornography, and often become cynical at the blatant mismatch between political rhetoric and the practical realities.

BBC’s Panorama 5 November 2006 showed several examples of paedophiles from a bail hostel remorselessly ogling and taking pictures of girls. One, Frank Parker, had killed a child and had been released after thirty-nine years in prison, yet swiftly returned to his obsessions and even managed to smuggle a girl into his room.




Every year spectacular crimes are committed by people on probation or parole, including some seventy-five murders, attempted murders or manslaughters.

Some of the most notorious crimes are by people on probation or parole including the killings of John Monckton (a wealthy banker stabbed along with his wife in his home in west London in 2005), Mary-Ann Leneghan (abducted, raped, tortured and stabbed to death in Berkshire in 2005 along with a friend who was shot but survived), and Sarah Payne (an eight-year-old girl abducted, raped and dumped in a shallow grave in Sussex in 2000). The UK’s Ministry of Justice conceded that 212 homicides were committed by people serving community sentences over a three-year period to 2007, leading to headlines that ‘Criminal freed on probation commit a murder every week’. In some cases the killers wore electronic tags. They were also known to have carried out thirty-eight attempted murders and 155 rapes. Unsurprisingly such inconvenient facts are usually glossed over by advocates of community sentencing.




Because of paperwork, travel, training and court appearances, probation officers can spend only a quarter of their time dealing directly with offenders,1 yet their caseload has risen by more than a third since the turn of the millennium.2 In fact, several probation officers have told me that with fifty (and in one case eighty) clients to look after they have less than an hour a day to meet the people they are supposed to be in charge of. Keeping tabs on people who move requires enormous coordination between probation, police, social services, court officials and sometimes other agencies as well.3

1Form-filling can take two or three hours for every ‘client’ and must be done before sentence and again after sentence, and then updated at least every sixteen weeks. MPs claimed to be ‘staggered’ at how little time probation officers had for direct supervision. BBC News, 17 July 2011. []

2 The annual total probation caseload (court orders and pre and post release supervision) increased by 39 per cent between 2000 and 2008 to 243,434, before falling slightly to 234,528 in 2011. Offender Management Statistics Quarterly Bulletin, April to June 2012, England and Wales, Ministry of Justice, 2012, p. 9. []. Probation officer numbers are from: Probation Service Workforce Information Summary Report: Quarter 2 2012/13, Ministry of Justice, 2012 [].

3The British Home Office set up Multi Agency Protection Arrangements (MAPPA) to try to achieve better coordination but it faced a hugely complex task and was dogged by scandals of released offenders committing spectacularly violent crimes.




There is a time-honoured fashion whereby advocates of new ways of treating offenders are convinced they have found the Holy Grail, and, in the words of David Green of the think tank Civitas, the first report on the experiment reads, ‘like a press release, instead of an independent appraisal’.

David G. Green, The Intensive Supervision and Surveillance Programme, Civitas, London, 2004[]




But by 2005 it became clear that offenders who had been on the scheme were reoffending at a prodigious rate. Over 90 per cent of them were reconvicted, each for an average of seven further crimes over the next two years.

Data supplied to author by Home Office from the ISSP Final Report. Another study, ascribed to Oxford University, found 85 per cent had reoffended within twelve months and that the ISSP scheme (at an extra cost of £12,000 a head) had been no more successful than other less costly forms of punishment.




Prison satisfies a sense of justice and a desire for retribution. It is payback time without the lynchings.

Plainly sentencing has many purposes too. It should assuage victims and restore a sense of equanimity to them, and it should act as a formal conduit for public anger so that people do not take the law into their own hands. For some people that is its only purpose. According to the right-wing polemicist Charles Murray, the function of justice is retribution. The community must exact revenge so as to depersonalise it. Society has a duty to punish culpable offenders to express society’s moral values: ‘That’s it. Nothing about rehabilitation, remorse, or socio-economic disadvantage. Nothing about the bad effects punishment might have on the offender or, for that matter its good effects. The purpose of the sentence is punishment… full stop.’ Charles Murray, Simple Justice, Civitas, 2005, pp. 20–21.






Little wonder that Prison Service targets do not include preventing reoffending.

The Prison Service has many targets but not crime prevention. Targets that are regularly met include security (number of escapes), staffing (days lost through staff sickness, having more ethnic minority staff), efficiency (timely prisoner escorts) and improving facilities for inmates (increasing skills training and resettlement arrangements). Albeit some of these are measured from a very poor baseline. Failed targets usually include overcrowding, suicides, serious assaults on staff, the number of hours spent by prisoners in purposeful activity, and completing sex offender programmes.




If enough crime results in the clang of a cell door then many of the people most inclined to behave badly will be taken off the streets. At very least it will postpone a future offence so our communities are safer for as long as they remain behind bars.

Governments have long maintained that ‘keeping the public safe is the first duty of the criminal justice system’, which is why they say they ensure ‘that those who are dangerous can be kept in prison for as long as necessary’. (A Five Year Strategy for Protecting the Public and Reducing Reoffending, London, The Stationery Office, February 2006.) Of course if prisoners were routinely kept in ‘as long as necessary’ to stop reoffending there would be no reoffending. On the other hand, postponing crime is useful because most offenders eventually grow out of antisocial behaviour.




Americans are five times more likely to be locked up than Britons.

In 2012 the US had on average 735,601 in local jails, 1.4 million in state prisons and 217,247 in federal prisons making a total about 2.35 million. The incarceration rate reached 737 per 100,000 population compared to 615 for Russia and about 140 for England, Scotland, Wales and Northern Ireland combined. The UK imprisonment rate is about midpoint worldwide. Many of the lowest rates are in poorer countries.




Here was a golden opportunity to test the effects of custody on a vast research canvas.

International comparisons can be misleading since imprisonment rates may not be measuring the same thing. Different countries have different crime reporting rates and different conviction rates as well as different incarceration policies. Internal comparisons, like those across US states, are more reliable.




There was enough variation in different years and different states for conservatives and liberals to cherry-pick outcomes that suited their point of view.

According to the Sunday Times believers in prison boasted that recorded crime plunged 40 per cent in Florida between 1993 and 2006 as prison rates rose by 75 per cent; whereas penal reformers crowed that in New York violent crime dropped 60 per cent while its prison population actually fell (Sunday Times, London, 2 March 2008, p. 25).




Some think incarceration might have accounted for as much as a quarter of the cut in violence and a third of the drop in theft.1 British researchers on the whole concurred. One Home Office model suggested that a single 15 per cent rise in the prison population might achieve a 1 per cent drop in crime.2

1 Several economists modelled the effects of incarceration in the 1990s. Alfred Blumstein, Joel Wallman, The Crime Drop in America, Cambridge University Press, 2005, see especially chapters 4 and 5.

2 Frustrated by the lack of evidence about what works in prison, the British government commissioned an inquiry, chaired by a former Whitehall mandarin, John Halliday. The report made depressing reading. It noted ‘the inability of short sentences to make any meaningful intervention in the criminal careers of many of those who receive them’. It reiterated the gloomy prognosis that younger people are more likely to reoffend than older prisoners. And while it showed graphically that longer sentences tend to be followed by fewer reconvictions, this is unsurprising given that long sentences are usually reserved for serious, and thus rare, crimes while short sentences are for common ones: The Halliday Report: Making Punishments Work; a Review of the Sentencing Framework for England & Wales, Home Office, London, 2001, Appendix 6. Other, more controversial, reviews have claimed it is just not possible to draw a hard and fast conclusion or, in one case, that the effect was rather higher than Halliday proposed. Patrick Carter (in a paper attacked by opponents as the ‘Worst Report of Recent Years’) estimated that one-third increase in use of custody in England and Wales between 1997 and 2003 was responsible for a 5 per cent drop in crime (The Carter Report: Managing Offenders, Reducing Crime, Prime Minister’s Strategy Unit, London, 2003. The Carter Report: Managing Offenders, Reducing Crime, Prime Minister’s Strategy Unit, London, 2003).





Eventually, because of its evident unfairness and financial burden ‘three strikes’ lost much of its voter appeal.

In 2012 even California, one of the ‘three strikes’ pioneers in 1994, substantially softened the law with a 69 per cent popular vote. Only serious or violent offences could be counted. Meanwhile Massachusetts bucked the trend, becoming the twenty-seventh state to adopt the policy. []




On the other hand the Dutch have tried a much more carefully focused approach. In effect they introduced a ‘ten strikes’ rule, jailing prolific offenders who proved resistant to everything else … From a strictly crime reduction point of view this is plainly a very promising approach – if only for a tiny cohort of highly recidivist offenders.

Exceptionally long sentences for the most persistent offenders led to a 25 per cent drop in acquisitive crime – precisely the types that the affected offenders committed. There was no impact on violent and sexual crimes which they rarely indulged in – but which were supposedly were the target of the ‘three strikes’ rules. However, ‘these benefits go down rapidly with a more intensive use of the law’.
Ben Vollaard, Preventing crime through selective incapacitation, The Economic Journal, 2012. [doi: 10.1111/j.1468-0297.2012.02522.x]




In a masterly survey, Jeremy Travis, one of America’s leading lights on penal policy, set out the predicament starkly: you can put more people into prison but they almost all come back. Well over half a million prisoners are released every year in the United States, most with no honest job to go to, many with fractured family ties, quite a few with mental illness, and ‘it is hard to imagine another identifiable group of people who exhibit such a high rate of criminal behaviour’.

But They All Come Back by Jeremy Travis, The Urban Institute Press, 2005, ISBN 0-87766-750-0, pp. 87, 95.




In Britain too, the revolving door goes round rapidly and relentlessly.1 In England and Wales over 40 per cent of adults are caught and sentenced for further crimes within a year of getting out of jail, sometimes for multiple offences.2 For young offenders the record is even worse: over half reoffended within a month of discharge and over three-quarters are re-convicted within a year.3

1 Actually, things need not be universally bad. It might be that some prisons work better than others. There are all sorts of penal institutions: open ones which rely on trust, medium security prisons, and high security ones; there are modern blocks with reasonably good facilities, and castellated Victorian fortresses; there are crowded ones and badly overcrowded ones; there are strict regimes and lax ones; there are calm ones and violent ones; well-run ones and prisons which are deplorable. In terms of recidivism, are they all equally good? Has anybody checked? I have looked for published evidence of prison efficacy – something you might have expected proponents of ‘prison works’ to flourish – but I can find no comparative studies of different regimes in terms of recidivism. It would be surprising if they all worked (or failed to work) at the same level.

2 Reoffending of adults: new measures of reoffending 2000–2005, Ministry of Justice, London, 9 May 2008 []

3 Craig Medhurst and Jack Cunliffe, Reoffending of juveniles: results from the 2005 cohort, Ministry of Justice, London, July 2007, London. ‘Those given more severe sentences such as custody are more likely to be at a high risk of reoffending,’ (pp. 9–10) but even so the extraordinary reoffending rate explodes the assumption that custody of ten- to seventeen-year-olds works through rehabilitation. [] See also: Reoffending of adults: new measures of reoffending,2000–2005, Ministry of Justice, London, 9 May 2008. []





Actual reoffending is vastly higher. In confidential self-report questionnaires prisoners confess to huge rates of reoffending without being caught – in one survey inmates admitted to an average of 140 offences in the year before they were convicted.

No one can know how often offenders reoffend without getting caught, but self-report questionnaires suggest that being caught is a relatively low risk. The official Halliday report quotes a survey in 2000 in which prisoners admitted to an average of 140 offences in the year before they were convicted. Nonetheless 85 per cent of people polled believed that prison sentences are very or fairly effective. (The Halliday Report, ‘Making Punishments Work: A Review of the Sentencing Framework for England & Wales’, Home Office, 2001, Appendix 6, p. 130.) []+[]




Four out of five Britons tell pollsters they think sentencing for convicted offenders is too lenient

Online poll of 2,046 adults was conducted between 18 and 20 February 2011: ‘From what you have heard do you think sentencing of people convicted of crimes is too harsh, too lenient or just about right?’ Eighty per cent chose ‘Too lenient’. (In the same poll three quarters thought crime rose or stayed the same between 1997 and 2010 – which was the period which saw the steepest and largest crime drop in human history.) Lord Ashcroft Polls, 3 April 2011.


Life sentence cartoon2


The US has tried all sorts of ‘boot camps’ and ‘scared straight’ regimes, and from the 1980s to the mid-1990s Britain added ‘short, sharp shock’ and military drill to incarceration. But all decent evaluations – including a highly reputable compilation of many studies – showed they simply didn’t work.

Anthony Petrosino, Carolyn Turpin-Petrosino, Meghan E. Hollis-Peel, Julia Lavenberg, ‘Scared Straight and Other Juvenile Awareness Programs for Preventing Juvenile Delinquency: A Systematic Review’, The Campbell Library, 2 May 2013. []

David Wilson, Doris Mackenzie, and Fawn Ngo Mitchell, ‘Effects of correctional boot camps on offending. A Campbell Collaboration systematic review’, 2005, available at: []. These echoed results from a UK-only survey: David Farrington, John Ditchfield, Gareth Hancock, Philip Howard, Darrick Jolliffe, Mark Livingston, and Kate Painter, ‘Evaluation of two intensive regimes for young offenders’, Home Office Research Findings No. 239, London: Home Office, April 2002.




One report catalogued twenty-five violent offences for each 100 inmates and in another survey one in five prisoners said they had been assaulted in the previous month.1 Even the murder rate tends to be higher in prisons.2

1 Kimmett Edgar, Ian O’Donnell and Carol Martin, Prison Violence: the dynamics of conflict, fear and power, Willan, 2003 ISBN 1903240980.

2 On average there are two prison homicides in England and Wales each year. Proportionally this is twice the rate in the general population, and higher than the usual rate for unsupervised adult males. (Figures calculated from national census and crime statistics for 2004/5 compared to Ghazala Sattar, Prisoner-on-prisoner homicide in England and Wales, Home Office Findings 250, 2004, ISSN 1473-8406.)





Bribery, smuggling and drug deals are commonplace. According to official government reports, mandatory blood tests are often dodged and a quarter of prisoners say they can easily get heroin, cannabis, buprenorphine or addictive prescription medicines like Tramadol.

Annual Report 2011–12, HM Chief Inspector of Prisons for England and Wales, House of Commons, October 2012, see p. 36. [] Buprenorphine is often prescribed as a heroin surrogate – like methadone – but has leaked into widespread street use. Under random testing buprenorphine ‘was the most misused drug in eleven prisons and the third most misused drug overall’ and rather than being a substitute for heroin it ‘appears to be an additional problem’. Including it routinely in random testing would increase the detected rate of prisoners on addictive drugs to one in ten.A Survey of Buprenorphine Misuse in Prisons,Ministry of Justice, London, July 2007. []




Time and again prison inspectors have denounced the conditions they report on. There is little training or formal work, a lot of enforced laziness and many prison governors submit ‘seriously misleading’ reports that ‘greatly inflate’ the amount of time prisoners spend out of their cells or in training or any other purposeful activity.

Annual Report of HM Chief Inspector of Prisons for England & Wales 2004–5, Stationery Office London, 2006.




… deterrence works in strange ways. We have less fear of cars, which are dangerous, than flying which is not.

Between 2002 and 2011 the average accidental death rate per billion passenger kilometres was 0 by air, 253 by car and 1,694 by walking. Reported Road Casualties for Great Britain, 2011, ‘Passenger casualty rate by mode 2002–2011’, Dept for Transport, 27 September 2012, p. 234. []




Back in the 1950s Fred Skinner, one of the fathers of behavioural psychology, gave rats and pigeons either food pellets or mild electric shocks and discovered what is known as operant conditioning: reward will encourage actions but punishment will not extinguish an undesirable behaviour, it merely suppresses it.1 Worse still, it has long been shown in humans, not just rodents, that punishment can cause aggression.2

1 The classic reference is B. F. Skinner, Science and Human Behaviour, MacMillan, NY, 1953.

2 Catherine Taylor, Jennifer Manganello,  Shawna Lee, and Janet Rice, ‘Mothers’ Spanking of 3-Year-Old Children and Subsequent Risk of Children’s Aggressive Behavior’, Pediatrics, Vol. 125 No. 5, 1 May 2010, pp. e1057–e1065. [doi: 10.1542/peds.2009-2678]





Experiments have shown adolescents find thrills even more rewarding when they know have cheated danger1 – and their tolerance for novelty and risk is magnified because of inexperience in understanding consequences.2 A Home Office study in England found offenders hugely overestimate of their chances of being caught but even so nearly three-quarters were ‘not worried’ about thepenalties.3

1 So-called ‘called counterfactually mediated emotions’ may be a natural part of cognitive development. Stephanie Burnett, Nadège Bault, Giorgio Coricelli and Sarah-Jayne Blakemore, ‘Adolescents’ heightened risk-seeking in a probabilistic gambling task’, Cognitive Development, 2010, Vol. 25, No. 2, pp. 183–6. [­ /releases/2010/03/100324211144.htm]

2 Agnieszka Tymula, Lior Rosenberg Belmaker, Amy Roy, Lital Ruderman, Kirk Manson, Paul W. Glimcher and Ifat Levy, ‘Adolescents’ Risk-Taking Behavior is Driven by Tolerance to Ambiguity’, Proceedings of the National Academy of Sciences, Vol. 109, No . 42, October 2012.

3  Offending in England and Wales: First results from the 2003 Crime and Justice Survey, Home Office, London, 2004, p. 64. [] This result should not surprise us. Young people suffer high morbidity and mortality because they shrug off the perils of unprotected sex, dangerous driving or experimentation with alcohol and drugs, so it’s little wonder they are least susceptible to warnings of criminal justice retribution.





… what frightens you may for someone else be water off a duck’s back. In fact, each year in England and Wales around 5,000 prisoners turn down the chance of early release. They would rather stay in jail.

According to figures from the Ministry of Justice, London, more than 37,000 inmates opted out of early release schemes between 1999 and 2006. Source: The Times, London, 4 June 2008, p. 9.




We know as an uncontested fact that the penalty for smoking is death and that half of those who smoke will die of it, often in horrible ways and losing on average fourteen years of life.1 Tobacco use kills more people than illegal drugs, alcohol, road accidents, AIDS/HIV, suicides and murders combined.2

1Richard Doll, Richard Peto, Jillian Boreham, Isabelle Sutherland, ‘Mortality in relation to Smoking: 50 Years’ Observation on Male British Doctors’, BMJ, 2004, Vol. 328, pp. 1519–28. []

2  Source: DrugScope []; and the US federal agency CDC (The Centers for Disease Control and Prevention), 2013. []




Did you know life sentences more than doubled between 1994 and 2004?1 Or that you are three times more likely to get a life sentence in one crown court than in another?2 Do villains in Cambridgeshire know that they are two-and-a-half times more likely to be sent to prison than if they carried out the same offence in neighbouring Lincolnshire? Does a crook in Bedfordshire feel any more vulnerable than his counterpart in Warwickshire, where his risk of prison is almost three times lower?3

1The Sentence, The Sentencing Guidelines Newsletter, London, January 2006, Issue 4, []

2 Between 2003 and 2006 the average custodial sentences were compared across all forty-two crown courts in England and Wales. The variation in life and indeterminate sentences ranged from 1.3 per cent to 4.3 per cent. Variance in all other sentencing was also found to be ‘large, particularly for custody rates in magistrates’ courts’. Thomas Mason, Nisha de Silva, Nalini Sharma, David Brown and Gemma Harper, Local Variation in Sentencing in England and Wales,  Ministry of Justice, 2007, pp. 6–7 []

3 In 1994 in Warwickshire 3.5 per cent of offenders were sent to prison; in Lincolnshire 3.8 per cent; in Bedfordshire 9.3 per cent; and for Cambridgeshire the figure was 9.6 per cent. For this and other variations see England & Wales 2004 Sentencing Statistics, RDS NOMS, Home Office, London, 2005, pp. 95 and 123–5.





A great deal of theft is impulsive, most violence is spontaneous and almost half all murders in England and Wales result from angry quarrels that get out of hand.

A study of 763 homicides recorded in England and Wales in 2007/8 reported that 341 took place because of a quarrel or loss of temper, 23 were plainly irrational acts, for example due to mental illness, in 186 cases no rational motive could be identified, and 11 were the result of arson. Only 47 were in furtherance of a crime, and 2 others in resisting arrest. In 153 cases the motive was ascribed to a complex mix of causes. David Povey, Kathryn Coleman, Peter Kaiza and Stephen Roe, Homicides, Firearm Offences and Intimate Violence 2007/8, Home Office, London, 2009, p. 23, Table 1.06. []




In the 1950s there was a classic study by Thorsten Sellin of the University of Pennsylvania, who meticulously crunched the numbers and famously deduced, ‘Executions have no discernable effect on homicide rates.’

Thorsten Sellin, 1959. The Death Penalty, American Law Institute, Philadelphia.




Capital punishment was stopped altogether by the Supreme Court between 1972 and 1976 and after that, even in its peak year of 1999, fewer than 100 people were put to death. That’s about one execution for every 1,500 homicides, always after lengthy and costly appeals. On that basis even the most calculating killer is unlikely to be deterred.

The fact that there are so few executions relative to homicides makes statistical analysis extremely dodgy. Some advocates of capital punishment defend their views by using a technique called multiple regression, but the US execution numbers are so small that a lot of assumptions have to be factored in which makes the results extremely unreliable. More significantly, graphs show homicide rates followed the same trends all across America from the 1960s into the new millennium. During those four decades Texas executed 239 prisoners, California ten and New York none, though all had the same outcomes.




An analysis of FBI Uniform Crime Reports in 2004 showed the murder rate in states with the death penalty was 5.1 per 100,000 population; among those without the death penalty it averaged 2.9.

In the US the FBI keeps crime data for every state []. See also the death Penalty Information Centre [].




And there may be positive effects from abolition: it has long been known that the death penalty makes prosecutions harder.1 (Of course it also risks grievous miscarriages of justice.)2

1Juries have long been reluctant to convict in capital cases when there is any chance of error. One of the first findings of this was in 1793 when William Bradford, the Attorney General of Pennsylvania, published ‘An Enquiry How Far the Punishment of Death is Necessary in Pennsylvania’. (The pamphlet also concluded that the deterrent effect was unreliable but even so Mr Bradford remained in favour of the death penalty.)


2There have been many notorious instances of miscarriage of justice including in the UK several which would have resulted in hangings had capital punishment been continued. These include the Guildford Four, The Birmingham Six and the Bridgwater Four.





For those who do think through the consequences – who premeditate their crimes – surveys consistently show they are far less concerned with penalties than with getting caught.

Anthony Doob and Cheryl Marie Webster, Sentence Severity and Crime: Accepting the Null Hypothesis In: Michael Tonry (ed.) Crime and Justice. A Review of Research, University of Chicago Press, 2003. See also: Paul Robinson, Does Criminal Law Deter? A Behavioral Science Investigation, University of Pennsylvania Law School, 2004, Paper 32.




For convictions after 2005, even long-term prisoners can expect to be released after half their time is done. So in most cases ten years means five. And naturally life does not mean life.

Since 2003 all courts passing a mandatory life sentence have been required to order the minimum term the prisoner must serve before the Parole Board can consider release on licence, unless the seriousness of the offence is so exceptionally high that the early release provisions should not apply in other words, a ‘whole life order’. Normally life will mean fifteen years for adults and twelve years for youths. []




Then there is a discount for time spent on remand. While defendants acquitted of a crime get no compensation, those found guilty do. In one case a fraudster who was jailed for nineteen years was out in eighteen months.

In this case a fraudster Edward Cosgrove conned other businesses out of £2 million and in 2006 was sentenced at Southwark Crown Court to four years for each of seven charges of false accounting and three years for dishonesty, all to run concurrently as though only one offence had been committed. Thus 4 + 4 + 4 + 4 + 3 = 2. In fact when time on remand was taken into account the sentence amounted to eighteen months. Actually Mr Cosgrove was unlucky; like most crooked businessmen he would have got away with it entirely had the police and Serious Fraud Office not taken an unusual interest after his company went into receivership. Business fraud is notoriously difficult to prosecute. But his concurrent sentencing is typical.




Policy-makers have often told me they are afraid that if the sentencing is unambiguously transparent the public will mutiny. But maybe voters can be trusted more than those in authority think. In 2002, along with Fiona Bruce I presented a live experiment for the BBC, perhaps the biggest of its sort ever conducted anywhere in the world.

Cracking Crime: You the Judge, with Nick Ross and Fiona Bruce, BBC One, 18 September 2002. There was a range of views with probation officers rarely favouring prison sentences and junior police more punitive than senior ranks. But when viewers were shown reconstructions of actual crimes and subsequent courtroom exchanges they tended to endorse the actual outcome. The only time the public voted for a much higher tariff than had actually been given, a judge in the studio agreed with the popular consensus. This rather suggests that judges are not out of touch so much as the public, which – outside an experiment like this – has little direct access to such events and so relies on supposition and news reporting in order to form its views on the appropriateness of sentencing. (See below.)




This accords with a great deal of research and measurement including a Royal Commission in Britain and a sweeping international comparative evaluation. It does seem most people are only vindictive when they feel disempowered: ‘The less information people have about any specific case the more likely they are to advocate a punitive response to it.’

Michael Hough and Julian Roberts, Attitudes to Punishment: Findings from the British Crime Survey, Home Office Research Study 179, 1998, p. 12. The paper reviewed large studies of public attitudes to crime from the UK, Australia, Canada, the Netherlands and the US.




In reality UK incarceration rates have been middling by international comparisons, higher than Spain, Hungary or Poland; lower than Germany, France or Norway. In any case such figures are misleading. They relate to population not to recorded crime. An official inquiry found England and Wales sent only twelve people to prison for every 1,000 recorded crimes compared to forty-eight in Spain – but we are prison-mad compared to Sweden, which jailed fewer than five.

A Better Way, Report of the Commission on the Reform of Public Services, Reform, London, April 2003, p. 138.




The old criminal justice model is simply floundering in modern times.

Steep cuts in policing budgets illustrate the disparity between real crime and the criminal justice system. By 2014 England & Wales had 16,000 fewer officers than when staffing was at its peak. Unsurprisingly, many crimes went not just undetected but un-investigated – so much so that a leading politician warned that the authorities were ‘turning a blind eye to the theft, burglary and shoplifting which makes up three quarters of all recorded crime committed in England and Wales’*. Even so, falls in crime continued whether measured by police records or by surveys. (*David Lammy MP, Taking its toll: the regressive impact of property crime in Britain, Policy Exchange, 2015.)