Criminological Highlights Vol. 12, No. 4 – March 2012

View this issue as a PDF: PDF iconCrimHighlightsV12N4.pdf

This issue of Criminological Highlights addresses the following questions:

  1. Do stiff fines stop people from drinking and driving?
  2. Are youths who behave badly in custodial institutions likely to reoffend when released?
  3. Would the public tolerate giving judges discretion in the sentencing of murderers?
  4. How has a reduction in US imprisonment rates been accomplished?
  5. Are people who live in punitive jurisdictions happier about sentences than people living in locations with less punitive criminal punishment practices?
  6. What can be done to increase the accuracy of eyewitness identifications in criminal cases?
  7. Are online child pornography offenders likely to commit offences involving sexual contact with children?
  8. Are job training programs for people leaving prison useful?

Increasing the size of fines handed down for drinking-driving offences will not reduce re-offending.

Fines are a relatively common sanction in criminal courts.  In Canada, fines are imposed almost as often as prison sentences. For criminal code driving offences (the impaired driving offences, dangerous driving offences, etc.) fines are imposed in Canada about five times as frequently as imprisonment.  For impaired driving offences, there are almost 10 times as many fines imposed as prison sentences.  More generally – for less serious offences and in other countries – fines are a very common penalty. 

Previous research has suggested that the imposition of mandatory minimum fines has not had a measurable general deterrent impact. In other words, mandatory minimum fines are no more likely to keep people from committing drinking-driving offences than penalties set by judges in which the judge has discretion on the size of the fine.  But there is less research on the effect of fines of different amounts on the likelihood that those who receive the fine will reoffend.   However, other research would suggest that the size of the penalty an offender receives has no deterrent effect on the likelihood that the offender will reoffend (see Criminological Highlights, 11(4)#2, 11(1)#1, 11(1)#2).  

This study examined the subsequent drink-driving offending of all of those charged with driving with blood-alcohol concentrations above the legal limit in New South Wales, Australia in 2003 and 2004.   The study takes advantage of the fact that there is substantial variability in the fines handed down by different magistrates.  Various controls were introduced related to the offender (age, sex, prior record of a drinking-driving offence) and the offence (urban or non-urban setting, blood alcohol content, plea, whether the offender was represented by counsel).  

Looking at the likelihood of a subsequent drinking-driving offence within three years, the results show that males, those with more serious original drinking-driving offences, those who faced their original charge without lawyers, and those with previous convictions for drinking driving offences, had a higher likelihood of reconviction.   However, there was no indication of an impact of the size of the fine that was handed down on the likelihood of reoffending within three years. 

Overall, almost 10% of the 12,658 offenders reoffended.  There was a good deal of variation in the fines handed down when they were convicted.  The lowest 25% of the fines were $400 or less.  The top quarter of the fines exceeded $800.   Thus the conditions for an adequate test of the specific deterrent impact of the fine were met.  Hence, had there been even a small deterrent impact of the size of the fine, an effect would have shown up. 

Conclusion:  Since the size of the fine appears to have no impact on the likelihood that a drinking driver will re-offend, it is reasonable to ask why this might be the case.  One possibility, of course, is that the perceived likelihood of apprehension may be too low.  Australian governments, aware of this problem, spend a good bit of effort on random breath testing and advertising campaigns designed to emphasize the risks in drinking and driving. “The perceived risk of apprehension, however, may be more dependent on the number of times a driver has been stopped by police while intoxicated or after drinking than on the publicity surrounding random breath testing, or the total number of times he or she has been stopped by the police or the number of times police have been seen performing random breath tests on other people” (p. 799).   What is clear, however, from this study and others is that raising the penalty size is not going to reduce this type of reoffending. 

Reference: Moffatt, Steve and Don Weatherburn (2011).  The Specific Deterrent Effect of Higher Fines on Drink-Driving Offenders.  British Journal of Criminology, 51, 789-803.

Back to top

The behaviour of young offenders in custodial institutions is not a good predictor of their offending after they are released.

It is understandable to assume that institutional misconduct by incarcerated young offenders predicts reoffending when these youths are released.  The theory is a simple one: if a youth can’t behave well in a custodial institution, then the youth must be out of control and is likely to offend when released.  Institutional behaviour, then, is often used in various decisions about the youth’s future such as whether it is safe to release the youth into the community. 

This study followed 1,804 violent male delinquents released from a large US juvenile correctional institution between 1987 and 2004.  These were very serious offenders: almost a quarter were incarcerated for a homicide or attempted murder and the rest were in custody for serious violent offences.  In this jurisdiction, these youths were eligible for ‘blended’ sentences, initially served (until age 18) in a youth facility.  At age 18, a hearing was held at which time the youth could be released (perhaps on parole), held until age 21, or transferred to an adult facility.  The study followed those who were not transferred to the adult facility for an average of about 5.7 years, and looked at re-arrests during this period.  

Six different measures of institutional conduct (expressed as rates per year incarcerated) were used to predict re-arrest: assaults on staff or on other youths, possession of a weapon, gang activity, staff reports that the youth was a danger to others, and total misconducts (which was largely made up of less serious types of misconduct).  Various other forms of controls were used:  race, age at release, time served, previous delinquencies, commitment offence, previous substance abuse, and whether the youth was supervised on parole on release. 

Looking first at those youths (n=1433) who were followed for at least three years, none of the measures of institutional conduct predicted either of the two outcome measures: re-arrest for any offence and re-arrest for a felony.  When ‘rate of re-arrest’ (for any offence) was examined, there was a very small, but statistically significant, effect of ‘total misconduct.’  When the size of this effect was compared to the predictive effects of the other known factors (e.g., race, whether there was a history of substance abuse, previous criminal record, and whether he had a history of involvement with gangs) the added impact of total institutional conduct was minor. 

Conclusion:  In this study “a variety of indicators of misconduct offered little explanation of post-release reoffending for this specific group of serious and violent male delinquents” (p. 722).   It may be that if one does not control for other factors (e.g. criminal record and personal characteristics of the youth), institutional behaviour might have a larger predictive value. But the findings clearly demonstrate that the best predictors of future offending for these youths (who had committed serious violent offences) were characteristics that they entered the institution with. But even these predictors are far from perfect.  This study simply demonstrates that ‘institutional behaviour’ adds little to our ability to predict future offending.   The fact that institutional behaviour is more recent – and more salient and intuitive to those making decisions – does not make it predictive of behaviour outside the institution. 

Reference: Trulson, Chad R., Matt DeLisi, and James W. Marquart (2011).  Institutional Misconduct, Delinquent Background, and Rearrest Frequency Among Serious and Violent Delinquent Offenders.  Crime & Delinquency, 57(5), 709-731.

Back to top

Residents of England and Wales want judges to have discretion in the sentencing of murderers.

In England and Wales (referred to hereafter as England), as in Canada, judges typically have a good deal of discretion in the sentencing of offenders.  The exception in both England and Canada is murder for which the mandatory sentence of life in prison is the only legal sentence.  The argument has been made that anything less than a life sentence would undermine public confidence.  

Simplistic public opinion polls have suggested that the public would prefer life without parole or the death penalty for murder, even though there are strong arguments against its use (see Criminological Highlights 9(1)#7).  When asked, in 2007, which penalty they preferred for someone who killed a stranger 34% preferred the death penalty, 44% preferred life without parole, and only 19% preferred ‘a long prison sentence with a chance of parole.’  As has been shown previously in a study of life without parole for youths, those favouring life without parole may be doing so because they don’t think about possible alternative sentences (see Criminological Highlights 12(2)#3).

This paper examines public knowledge of sentencing for murder and explores whether, for 1027 respondents to a 2010 survey, life sentences are truly the preferred option. Most respondents were unaware of the fact that homicide rates in England had decreased somewhat in the previous decade.  64% of respondents thought, incorrectly, that the rate had increased.  48% of respondents were reasonably accurate on how long murderers actually spend in prison, but 42% underestimated the actual time whereas only 10% overestimated the actual time spent in prison.  In focus groups, it became clear that people knew that a ‘life sentence’ didn’t mean life without the possibility of release, but they didn’t have a clear idea what a ‘life sentence’ meant. When asked to estimate the percentage of  life prisoners who had been released to the community  and were subsequently returned to prison for an offence (the actual percent is about 2%), 74% of the respondents thought that the correct answer was 10% or more.  Only 9% were unsure and didn’t venture a guess and 17% guessed  5% or less.   

Respondents were given 10 specific scenarios describing actual cases in which people were convicted of murder.  They were then asked which of 5 alternatives they would prefer as the sentence: up to 9 years in prison, 10-19 years in prison,  20-29 years in prison, 30 years or more in prison with release at some stage, or imprisonment for the offender’s natural life.  There was huge variation in  preferences across the descriptions of the different murders.   For example, for the most severe alternative (imprisonment for life without release), the percent favouring this alternative ranged from 4% to 52% with an average, across scenarios, of 26%.   In other words, for all but one scenario, most respondents favoured a sentence with a definite ending.   But equally importantly, their preferences for the length of murder sentences varied considerably with the facts of the case.  

Over 80% of respondents thought that sentences (or sentences in murder cases) were too lenient, with at least half of these indicating that sentencing was much too lenient.   Those who were less well informed about the facts related to homicide in England were more critical of sentencing in general, and were more punitive in their preferred sentences in the murder scenarios. 

Conclusion:  When asked general questions, people say that they favour life without parole (or death) sentences for murder, but when asked about actual cases, they opt for less severe sentences and, more importantly, their preferred sentences reflect the specific facts of the case.  Mandatory ‘life sentences’ are preferred in some cases, but not in others.   A sentencing scheme for murder that reflected the values of a public informed about the cases and the alternatives that are possible would, then, be much more flexible than the current structure that exists in England and in Canada.

Reference: Mitchell, Barry and Julian V. Roberts (2012).  Sentencing for Murder: Exploring Public Knowledge and Public Opinion in England and Wales. British Journal of Criminology, 52, 141-158. 

Back to top

The number of people in American state and federal prisons declined in 2010 for the first time since 1972. The reduction occurred as a result of a variety of changes in state policies.

At a time when Canada has been implementing policies designed to increase the prison population, the US – the unrivalled world leader in imprisonment rates – has been decreasing its imprisonment rate.  Part of the motivation to decrease imprisonment appears to be fiscal concerns.  In other instances, courts, not legislatures, made the decision that reduction of imprisonment is necessary because of overcrowding.  California, for example, was ordered by the US Supreme Court to reduce its inmate population by about 46,000 inmates, a reduction that is about 30% larger than the total Canadian prison population.  (California’s resident population is 9% larger than Canada’s.) The mechanisms for change, however, appear to be quite varied across states. 

The most obvious way for states to reduce the use of imprisonment is through changes at the ‘front door” – in sentencing policies.   What is most notable about the sentence reduction changes that have been made is that they appear to have two major targets.  In the first place, imprisonment of drug offenders has been targeted.  A number of states have either eliminated or further limited the requirement of imprisonment for certain drug offences (e.g., in Arkansas, Connecticut, Delaware, Kentucky, North Carolina, Ohio).  Some have reduced the use of imprisonment for drug offences by increasing the availability of drug courts (e.g., in Florida, Idaho).   

Alternatively, states have reduced imprisonment less visibly by changes at the ‘back door’ – by allowing prisoners to be released earlier or by making parole more available or easier to obtain. This has occurred in Arkansas, Colorado, Connecticut, Kentucky, Louisiana, Maryland, Montana, Nebraska, North Carolina, North Dakota, Oklahoma, Ohio, Oregon, Rhode Island, Vermont, and West Virginia. In Louisiana, for example, prisoners over 60 and who have served 10 years in prison can be released early, if they are low risk. Similarly, severely ill prisoners in Rhode Island can be given medical parole.

A third way of reducing imprisonment is by limiting the ‘re-entry’ of those serving sentences in the community who violate the conditions of their community sanction (e.g., probation) or violate conditions of release (parole).  Changes allowing non-prison responses to these ‘technical violations’ have occurred in Alabama, Louisiana, Maryland, and North Carolina. In South Dakota the problem of parole violations was addressed by allowing early discharge from parole for certain prisoners.  In Texas, the problem of the imprisonment of those committing technical violations while being supervised was addressed by allowing for a reduction in the period of community supervision and by giving financial incentives to counties if they reduce the number of probationers who are returned to prison. 

There have also been a number of attempts to reduce the collateral consequences of imprisonment such as allowing records of certain offences to be expunged (Arkansas, Idaho, Indiana, Oregon) and eliminating certain statutory exclusions of offenders from rights such as voting,  accessibility of welfare programs, or employment (Delaware, North Carolina, Ohio, Texas, Utah). 

Conclusion: What is important about the changes that took place in 2011 to reduce imprisonment in the US is that they are varied and widespread.  They are not concentrated in any one region or in states with a particular political orientation.  They are broad, but largely they target specific types of offenders.  Furthermore, they seem to be more likely to target release from prison and to limit re-incarceration for technical violations than they are to address the sentencing policies that landed the offender in prison in the first place.

Reference: Porter, Nicole (2012).  The State of Sentencing, 2011.  The Sentencing Project. Available online: www.sentencingproject.org 

Back to top

People who live in relatively punitive jurisdictions are not more likely to be confident that sentences are harsh enough.

In many countries (e.g., England and Canada) a substantial majority of the public thinks that sentences are too lenient.  In Australia, 12 separate surveys carried out between 1984 and 2007 showed that between 70% and 85% of Australians believe that harsher sentences should be handed down in criminal courts.  The suggestion, from political leaders who favour high imprisonment, is that people would have confidence in the courts and in sentences if sentences were more punitive.  

The problem with this assertion is that it assumes the view people hold about the appropriateness of sentences imposed in their jurisdictions is the result of a careful assessment of sentences actually handed down. This is almost always not the case, in part because the data for such an assessment is typically not available to ordinary citizens. No ordinary citizen examines court sentences systematically and then makes a reasoned decision that they are too harsh, too lenient, or just right. 

This study examines a fairly simple question: are people in those locations in Australia that are most punitive (measured by imprisonment rates) likely to express confidence with sentencing?  Australia has 8 separate states and territories (each with its own criminal law).  There is considerable variation in imprisonment rates: the most punitive four jurisdictions have imprisonment rates that are, on average, more than twice that of the average of the four least punitive jurisdictions. 

In a survey that took place in all 8 jurisdictions, people were asked 7 questions designed to measure each individual’s confidence in sentencing.  Seven other questions measured the respondent’s desire for more punitive sentencing.  Not surprisingly, those who had the lowest confidence in sentencing were most likely to desire harsher sentences.

Although there was, as noted above, a fair amount of variation in actual punitiveness (as measured by the rate of imprisonment), there was little variation in the average confidence in sentencing or in people’s desire for harsher sentences. However, some of the states/territories did differ from one another on these attitude measures. 

Most important is the finding that there is a “lack of concordance between the public attitude measures and imprisonment rates across jurisdictions” (p. 379).  Said differently people were not more confident in the sentences that were handed down in their jurisdictions when they lived in a punitive state or territory.  Similarly, people were not less likely to desire harsher sentences than those currently imposed when they lived in a state in which relatively harsh sentences were already handed down. 

Conclusion:  “Wide differences in sentencing practice [across states and territories in Australia] and political activity on sentencing are not linked to any major differences in public attitudes between jurisdictions” (p. 381).   People living in those states with the lowest imprisonment rates are no more likely than people living in states with high imprisonment rates to hold punitive attitudes or lack confidence in the sentencing system.  In other words, beliefs about sentencing and beliefs about the courts appear to be independent of what is actually happening in the courts. 

Reference: Roberts, Lynne D., Caroline Spiranovic, and David Indermaur (2011).  A Country Not Divided: A Comparison of Public Punitiveness and Confidence in Sentencing Across Australia.  Australian & New Zealand Journal of Criminology, 44(3), 370-386. 

Back to top

Seventy-two studies demonstrate that the accuracy of the identification of suspects can be improved if police present eyewitnesses with lineups sequentially rather than simultaneously. 

For many decades, it has been known that a frequent cause of wrongful convictions is mistaken eyewitness identification of the offender by a well-meaning witness.  (Criminological Highlights 12(1)#3). The ‘simultaneous lineup’ in which the suspect (or a picture of the suspect) is shown to an eyewitness along with a number of other people or pictures is the standard approach most police services use in conducting a lineup.

Since the number of wrongful convictions is not trivial (Criminological Highlights 9(4)#5, 4(1)#6, 7(5)#5), the challenge is to find an approach for identifying offenders that has fewer errors.   The difficulty with the standard simultaneous lineup in which a person sees the suspect along with a number of other people is that there is a tendency for eyewitnesses to identify the person who most closely resembles the offender, rather than to make a judgement that the suspect is the offender.

The solution to this problem in the identification process has been known for some time: the witness should be shown pictures (or live people) sequentially – one at a time – rather than simultaneously (all pictures or people at once).  The eyewitness is shown a single picture (or a person) and decides if he or she is the culprit.  Then the witness is shown the next person or picture and indicates whether  that person is the culprit. Typically the witness does not know how many pictures (or people) are to be presented.  Ideally – in any identification procedure – the person administering the procedure should not know who the suspect is so that unconscious cues cannot be communicated to the eyewitness. 

This study brings together the research that compares the accuracy rate of simultaneous lineups and sequential lineups.  Seventy-two separate experiments were located.  Obviously these studies varied in a number of different ways.   Twenty-seven of the studies used a ‘full design’ in which simultaneous and sequential lineups were compared, in which, for some witnesses,  the culprit was present and for some the culprit was not.  That way the accuracy can be estimated more  precisely since the researchers can see whether, in the ‘culprit absent’ lineups, the witness correctly indicated that nobody in the lineup was the culprit. 

Overall, the results are straightforward: sequential lineups are more accurate in identifying the true culprit.  However, the detailed results are a bit more complex.  ‘Choosing rates’ are higher in the simultaneous lineups than in the sequential, perhaps because guessing is implicitly encouraged.  The result of this higher tendency to pick someone, results in the culprit being more likely to be chosen in the simultaneous lineup.   But the advantage of the sequential lineup can be seen very clearly in the cases in which the culprit was absent:  eyewitnesses subjected to this procedure were much more likely to identify nobody (the correct decision) than in the simultaneous lineup.

Conclusion: The standard in most police departments seems to be the simultaneous lineups.  These data suggest very strongly that, in comparison with a sequential presentation of the lineup, the simultaneous lineup is much more likely to result in a mistaken identification of an innocent person.   The reluctance of police departments to change may be the result of the other finding: an estimated 8% drop in culprit identifications in the sequential lineup.  There are, however, a number of police departments in the US that use sequential lineups apparently without difficulty.  These departments appear to understand that ultimately their goal should be to use the most accurate test that is available. 

Reference: Steblay, Nancy K, Jennifer E. Dysart, and Gary L. Wells (2011). Seventy-two Tests of the Sequential Lineup Superiority Effect: A Meta-Analysis and Policy Discussion. Psychology, Public Policy, and Law, 17(1), 99-139.  

Back to top

It should not be assumed that users of child pornography either have committed ‘contact’ sexual offences in the past or are likely to do so in the future.

There is understandable concern that ‘online’ child pornography offenders (those whose offence involves possession or viewing of child pornography on their computers) have committed or will commit sexual offences with children involving actual physical contact with their victims. In addition, of course, there are other concerns about child pornography:  the demand for child pornography encourages the exploitation of children to meet this demand, and its mere existence offends community standards and values.

This paper looks at the users of child pornography and asks two questions: (a) Are ‘online offenders’ likely to have committed offences in the past? (b) After being apprehended for an online child pornography offence, are ‘online’ child pornography offenders likely to commit further offences?   These are important questions because the answers to them are likely to be helpful in shaping policies related to society’s responses to online child pornography offenders.

The study identified 24 studies that examined the criminal histories of online offenders. One important finding from these studies is that there is substantial variation in the rates of previous offending.  Not surprisingly, when one looks at a sample of those who are arrested, the rates are lower than they are for those who are identified because they are in a correctional institution.  And both of these are lower than rates for those referred for clinical treatment.  These differences are “probably the result of contact offence history having an effect on whether someone is incarcerated [or]… being referred for assessment or treatment” (p. 125).   In addition, those who are incarcerated and in treatment programs are often encouraged to admit to offences [even if they did not take place exactly as stated], since admitting to prior offences is sometimes seen by clinicians as a sign of progress in treatment.

Looking at the percent who had previously been charged or convicted of any contact sexual offence, the range was from 0% in one study to 43%.   When official records are examined (for 4,464 online offenders), 12.2% had prior contact sex offences. Looking at all records (including self reports), 17.3% had reports of prior contact sex offences.  For the (largely) clinical sample, using self reports, about 55% reported prior sexual contact with children.  

The recidivism data are a bit more complicated, in part because the followup periods varied (from 1.5 years to 6 years, with most periods under 4 years).  Looking at the recidivism of 1,247 online offenders, 2% reoffended with a contact sexual offence and 3.4% recidivated with another child pornography offence.  Information about (non-sexual) violence recidivism was available for 983 online offenders.  The rate was 4.2%.  The relatively low rate of recidivism is consistent with other studies of sex offence recidivism (Criminological Highlights 6(6)#8, 5(1)#4, 3(3)#3, 6(3)#3, 9(2)#5). 

Conclusion: It is almost certain that the arrest or charge data under-estimate the involvement in contact sex offences.  The self report measures suggest that up to about half of online sex offenders may have committed contact sex offences in the past.  But whatever measure one looks at, it appears that “there is a distinct group of online offenders whose only sexual crimes involve illegal (most often child) pornography or, less frequently, illegal solicitations of minors using the Internet” (p. 136).  But it is also true that “online offenders rarely go on to commit detected sexual offences” (p. 136) and “pedophilic interests do not necessarily result in contact sexual offences against children” (p. 140).  Initial research evidence “suggests that the same risk factors matter for online or offline [contact] sexual offending” (p. 137).  Policies, therefore, should reflect the fact that online offenders do not constitute a homogeneous group of offenders. 

Reference:  Seto, Michael, R. Karl Hanson, and Kelly M. Babchishin (2011).  Contact Sexual Offending by Men with Online Sexual Offences. Sexual Abuse: A Journal of Research and Treatment 23(1), 124-145. 

Back to top

Although transitional job training programs for people leaving prisons have not, overall, been very successful in reducing reoffending, successfully completing these job programs may serve as a signal that the offender has chosen to stop offending and become part of the ordinary workforce. 

A number of meta-analyses of noncustodial employment programs for people with a history of offending have resulted in rather disappointing results.  The safest overall conclusion would be that they do not, as a rule, have an impact on long term employment or reoffending.  

There are a number of possible reasons for this, among them the tendency to treat ‘offenders’ as a homogeneous group, rather than as a quite varied group in terms of their needs and skills. Hence it may be that more fine grained analyses (see, for example, Criminological Highlights, 4(3)#6) would demonstrate positive effects for certain types of offenders (in certain kinds of programs).   In addition, of course, there is no reason to believe that all ‘programs’ that label themselves as providing transitional jobs and/or job training are going to be equally effective.  It may be that there are only certain components of the programs that are successful, and that only some programs have components that reduce offending (see Criminological Highlights 6(4)#5). 

One of the more intriguing findings, however, is “the well-known fact that graduates of employment training programs recidivate at lower rates than non-graduates, and they perform better in the labour market” (p. 22).  In evaluating overall program effects, this form of self-selection is a problem in that those who drop out have chosen not to receive the full ‘dose’ of the program.   Typically, then, the ‘program’ as a whole is evaluated by its ability to affect all of those initially assigned to it, not only those who choose to complete it.  

Evaluators of programs see dropouts as a problem. But the difference between dropouts and those who complete the programs may itself be interesting.  Completion of an employment program may serve as “a valid and powerful demonstration of the potential function of program completion as a desistance signal during reentry” (p. 22).  Said differently, those offenders who choose to do the hard work of successfully completing an employment program may be ‘signalling’ the fact that they have decided to stop offending.  Thus the program may provide an opportunity, in a timely and intuitive fashion, to identify those who have themselves chosen to make efforts to stop offending.  This is similar to a suggestion made some years ago about the mechanism for positive impacts of restorative justice programs: they  provide offenders with “an opportunity to facilitate a desire, or consolidate a decision, to desist” (Criminological Highlights 9(5)#1). 

Conclusion:  The overall effects of employment programs for people leaving prison are not encouraging except in one important way. Doing the hard work to successfully complete these programs provides a valid signal, to anyone watching, that the offender is much more likely than those who did not complete the program (or than similar offenders who did not have an opportunity to enrol in the program) to stop offending.  For a potential employer, then, successful completion of the program is a valid signal that the former prisoner is ready to go straight.

Reference: Bushway, Shawn D. and Robert Apel (2012).  A Signaling Perspective on Employment-Based Reentry Programming: Training Completion as a Desistance Signal. Criminology & Public Policy, 11(1), 17-50. 

Back to top

Publication Type

Volume Number

12

Issue Number

4