Criminological Highlights Vol. 17, No. 5 – November 2018

View this issue as a PDF: PDF iconCrimHighlightsV17N5.pdf

This issue of Criminological Highlights addresses the following questions:

  1. What kinds of police activities suppress voter turnout?
  2. How are people affected by police shootings of unarmed civilians?
  3. Are politicians right when they suggest that higher rates of pretrial detention would reduce crime?
  4. Who benefits from high concentrations of immigrants in a neighbourhood?
  5. When punishments are decreased in a jurisdiction and crime goes up, is it possible to determine whether one caused the other?
  6. How good are people at evaluating forensic science evidence in court?
  7. Should restorative justice conferences be used with youths charged with crimes?
  8. Does it matter where accused people sit in court during their trials?

High rates of aggressive policing result in lower rates of voter turnout in English cities.

Previous research has demonstrated that imprisonment can have harmful effects both on those imprisoned and on their families (see The Effects of Imprisonment: Specific Deterrence and Collateral Effects on our website). Recent research (e.g., Criminological Highlights 14(4)#1) has demonstrated that contact with the justice system, and imprisonment in particular, reduces people’s level of political engagement. This paper suggests that high levels of aggressive policing activities in a neighbourhood can reduce voter turnout in that neighbourhood.

This study examines two types of aggressive policing.  First, it examines the use of anti-social behaviour (ASBs) policing in which the police are encouraged to target any behaviour “that caused or was likely to cause harassment, alarm or distress to one or more persons not of the same household [as the perpetrator]” (p. 1235).  The police then can use legal means to put orders on citizens to stop this behaviour, even if the behaviour itself would not normally be considered to be criminal.  Second, it examines “stop and search” actions by the police. It was suggested that these two forms of aggressive policing activities could cause political resentment or alienation in a neighbourhood because both actions are fundamentally adversarial in nature, they are targeted at people in particular neighbourhoods, and they are largely perceived to be illegitimate and/or unfair.

Data were obtained on ASB policing and voter turnout in three years (2012, 2014, and 2016) in over 500 wards in Metropolitan London.  Data on stop-and-search activities in over 300 wards (in 2015 and  2016) in each of Greater Manchester and the West Midlands were also gathered.  Various characteristics of the wards (variables presumed to be related to voting) were included as controls. These included housing (percent homeowners, average housing price, percent social housing), percent single-parent households, unemployment, ethnic background and education.   The goal was to examine, above and beyond these control variables, the impact of ASB activities and stop and search by police on ward-level voter turnout

The effects on voter turnout were consistent across types of policing and years: large numbers of ASB stops per capita and large numbers of stop-and-searches per capita in a city ward were associated with lower rates of voter turnout in that ward. Specifically, with respect to ASB policing, it was estimated that if the ward with the lowest rate of ASB policing were, instead, to have received the highest rate, this would have resulted in a 3-4 percent reduction in voter turnout in that ward. The effects of stop-and-search policing were similar, though somewhat smaller (between a 1.2% and 2.6% reduction in voter turnout).

Surprisingly, “the results suggest that as the Black, Asian and Muslim share of the population [in a ward] increases, the demoralizing effect of community-level [ASB and stop-and-search] policing diminishes” (p. 1246). The exception was that, for one year, the effect of stop-and-search policing on reducing voter turnout was greater as the percent Black in that ward increased.   More generally, of course, the focus of the study was on ward-level policing activities and ward-level voter turnout rather than the impact of being a recipient of a particular kind of policing and individual voter turnout. Hence the experiences of different groups on voting could not be assessed.

Conclusion: The effects of aggressive policing (the policing of anti-social behaviour and police stop-and searches) on civic engagement (e.g., voting) have previously only been explored in the US and the effects have often been explained within that country’s history of high imprisonment rates and voter suppression/ disenfranchisement.  These results demonstrate that each of these forms of aggressive policing can reduce political engagement in other countries as well – in this case, England.  More importantly, the results of this paper “suggest that the demobilizing relationship between criminal justice contact and political participation is more fundamental than previously characterized….” (p. 1250).

Reference: Laniyonu, Ayobami (2018). Police, Politics and Participation: The Effect of Police Exposure on Political Participation in the United Kingdom.  British Journal of Criminology, 58, 1232-1253.

Back to top

The killing of unarmed Black Americans by the police has broad negative impacts on the mental health of Black Americans who live in the state in which the killing took place.

It is estimated that in about 29% of police killings of Black Americans, the victim was unarmed.  The comparable figure for White Americans is 19%.  This study examines the short-term impact of police killings of unarmed Black Americans on the mental health of Black Americans living in the state in which the person was killed.

“A large literature has shown associations between racism and health outcomes” (p. 302), but it has been difficult to test whether this relationship is causal.  In this study, by taking advantage of the essentially random nature of the timing of police shootings of ordinary black citizens and comparing the impact on Black and White Americans of shootings of armed and unarmed Black and White citizens, causal inferences can be made with more confidence.

In an ongoing telephone survey (the US Behavioural Risk Factor Surveillance System) between 2013-2016, approximately a million American adults (about 10% of whom were Black) were asked, among other things, how many days in the previous 30 days was their mental health ‘not good’ (p. 304). It was demonstrated that the timing of police killings is unrelated to other factors that might influence mental health.  Nevertheless, controls were included for year, month, day of the week when the interview took place, age, sex, and education.

The hypothesis was that police shootings of apparently unarmed Black people in a respondent’s community (defined conservatively as within the same state as the respondent) that occurred recently (within the previous 3 months) would lead to increased reports of ‘not good’ mental health.

The results are simple to describe: When Black Americans are exposed to one or more police killings of other unarmed Black Americans in their own state in the previous 90 days they are more likely to report higher numbers of days in which their mental health was ‘not good.’ In line with the idea that the harmful effects of these shootings are likely to be due to “heightened perceptions of threat and vulnerability, lack of fairness, lower social status, [and/or] lower beliefs about one’s own worth” (p. 308), it was found that there were no significant effects of police killings of armed Black Americans.  Similarly, police killings of White (armed or unarmed) Americans had no impact on the mental health of  Black Americans.   Furthermore, there were no significant effects of the killing of Black or White, armed or unarmed, residents on the reported mental health of White Americans.

Further support of the hypothesis that police killings of unarmed Black Americans caused a decline in mental health came from the fact that the decline in mental health of Black Americans only occurred after the police shooting.  Those Black Americans who were interviewed prior to a police killing of an unarmed Black American showed no impact.

Conclusion: Police killings of unarmed Black people were associated with worse mental health of Black Americans in the three months following the shooting, compared to ratings of mental health by similarly placed Black Americans who did not live in a location where such an event occurred in the previous 3 months.  The effect that was measured is probably less strong than the total impact, since the ‘comparison’ areas – those without a police killing of an unarmed Black person – were often exposed to the same media information. “The results provide rare causal evidence about the impact of events widely perceived to reflect structural racism on the mental health of Black Americans” (p. 308).

Reference: Bor, Jacob, Atheendar S. Venkataramani, David R. Williams, and Alexander C. Tsai (2018). Police Killings and Their Spillover Effects on the Mental Health of Black Americans: A Population-based, Quasi-experimental Study.  The Lancet, 392, 302-310.

Back to top

Pretrial detention has harmful effects for both the defendant and society more generally: it increases the likelihood that a person will plead guilty to a felony and, in the long run, increases reoffending after the case is completed.

In the United States, one out of every 550 adults is currently in pretrial detention and, in recent years, the rate is growing (in contrast to sentenced imprisonment).   This study follows 245,060 felony cases in New York City to examine the impact of pretrial detention on case outcome and reoffending.

Numerous studies have examined the relationship between being detained pre-trial and being convicted. The problem is that there may be unobserved or unmeasured differences between cases in which the defendant is detained and those in which they are released.  Although it is hard to argue, normally, that society benefits from the large variation across judges in decisions which result in accused people being detained or released, one advantage of this variation is that the assignment of cases to judges is typically unsystematic.  Thus whether the case is being handled by a ‘tough’ or ‘lenient’ judge is unlikely to be related to case characteristics.  If, then, cases assigned to judges who are likely to detain have different case outcomes from cases assigned to judges who are more likely to release accused people, it is likely to be the result of whether, overall, the accused was detained or released.

This study looks at cases heard by 212 judges, each of whom heard at least 500 arraignment (pre-trial detention) cases between 2009 and 2013.  If a judge decides that a person can be released, cash bail is usually set, though in many cases the accused is unable to come up with the bail amount.  A case was defined as ‘detained’ if the accused was formally detained or not released because they were not able to come up with the required cash.   Case characteristics, for felony cases, were clearly unrelated to the average severity of the judge who dealt with the case.  For misdemeanour cases, there was some suggestion that certain cases were somewhat more likely to go to ‘tough’ judges.  Hence inferences about the impact of detention for misdemeanors are somewhat more problematic than inferences concerning felonies. Nevertheless, the results for the 728,750 misdemeanor cases are largely similar to those described below for the felony cases.

The results for felonies demonstrate that “being detained increases the probability of conviction by 13 percentage points and the probability of pleading guilty by 10 percentage points… [suggesting] that detention primarily affects conviction by inducing some individuals who would not have pled guilty if released to plead guilty after they are detained” (p. 543).  This effect was more pronounced in felony cases in which the accused has no criminal record. Compared to those who receive pretrial release, those who are detained are also less likely to receive a reduction in the class of offence that they plead to.

Obviously, those who are detained for the period before their cases are disposed of are less likely to be re-arrested before the disposition of the original case. However, these ‘benefits’ are lost when one considers the fact that those detained are, within two years of the final decision on the case, more likely to be re-arrested even though, for those who were initially detained, some of the study participants spent time in sentenced custody.   Hence, the short-term incapacitation effect of pretrial detention is largely lost by what happens within two years of the end of the case.

Conclusion:  The results are clear.  Those who have the misfortune to be arraigned before a ‘tough’ judicial officer are not only more likely to be detained.  They are more likely to plead guilty, they are less likely to be offered attractive plea deals, and they are more likely, after the case is disposed of, to commit future offences.  “Tough” decisions at the pretrial detention stage then are “effective” both in getting accused to plead guilty and in increasing the likelihood of future offending.

Reference: Leslie, Emily and Nolan G. Pope (2017).  The Unintended Impact of Pretrial Detention on Case Outcomes: Evidence from New York City Arraignment.  Journal of Law and Economics, 60, 529-557.

Back to top

Using victimization data (rather than measures of crimes reported to the police) demonstrates that high concentrations of immigrants in a neighbourhood provide protection from crime. The data show that both economically advantaged and disadvantaged people enjoy the crime-reduction benefit of high concentrations of immigrants in their neighbourhoods.

There is a substantial research literature (see Criminological Highlights 17(4)#1) demonstrating that when the concentration of immigrants (including ‘undocumented’ immigrants) in a neighbourhood increases crime tends to decrease.  In general, however, this research uses either homicide data or police-based measures of crime (i.e., reports of crime that are recorded by the police).

This paper examines victimization data – from US victimization surveys carried out in 2008-2012 – of residents’ reports of their own victimizations.  It examines whether the ‘protective’ impact of immigrants is concentrated on particular groups of people. Hence the paper examines the relationship between the concentration of immigrants and the non-lethal violent victimization (sexual assault, robbery, and assault) experiences of Whites, Blacks, and Latinos. The key variable of interest is the proportion of the neighbourhood population that was foreign born.  Various other predictors of victimization were also examined, including county labour market data, population density, age, family structure, and various neighbourhood measures of economic disadvantage.  Part of the reason for examining the race of those who are possible victims is that in the US, Whites tend to live in neighbourhoods with lower proportions of foreign born (9%) compared to Blacks (13%) and Latinos (26%).  Hence it is possible that Whites do not benefit from increased concentrations of foreign born.

Overall, a high concentration of immigrants in a neighbourhood was associated with lower likelihood of violent crime (reported or not reported to the police). This held true even when 9 neighbourhood variables, 10 county labour and housing market variables, region, and 12 variables related to the individual respondents were controlled for.

The results appear to show, however, that this protective effect is primarily experienced by those who live in neighbourhoods with high concentrations of immigrants; there is little difference in the likelihood of victimization for those living in low vs. moderate concentrations of immigrants.  Latinos who tend to live in neighbourhoods with high concentrations of immigrants experience the largest change in the protective impact of immigrants as one moves from low to moderate concentrations of immigrants.

It appears that the protective impact of high concentration of immigrants was larger for Latinos than for Blacks and Whites. However, perhaps more important than these details is the finding that there is a “significant protective role of immigrant concentration on [violent] victimization for Whites, Blacks, and Latinos that is not dampened by a person’s low education level, low income, and unemployment or by measures of labour market competitions between citizens and noncitizens or between Latinos and non-Latinos” (p. 324).

Conclusion:  Consistent with a large body of research, these findings, using reports of victimizations (reported or not reported to the police) rather than police-recorded data, show that high concentrations of foreign-born people in a neighbourhood are associated with lower rates of victimization.  Given that the findings hold for all groups that were examined (Whites, Blacks, and Latinos), the findings might be “useful to encourage local communities to evaluate their attitudes toward immigration and create a more immigrant-friendly environment….” (p. 324).

Reference: Xie, Min and Eric P. Baumer (2018). Reassessing the Breadth of the Protective Benefit of Immigrant Neighbourhoods: A Multilevel Analysis of Violent Risk by Race, Ethnicity, and Labour Market Stratification.  Criminology, 56(2), 302-332.

Back to top

The small increase in crime in California that immediately followed an attempt to reduce state imprisonment was not caused by the change in imprisonment policy. Instead, it seems most likely to be ordinary ‘random’ variation in crime rates.

In November 2014, a citizen-initiated proposition (Prop 47) in California required that certain property and drug offences be charged as misdemeanors rather than felonies. The clear purpose was to reduce imprisonment.  However, in 2015, there were increases in violent crime (8.4%) and property crime (6.6%) in the state, leading some to believe that the increase in crime was caused by the reduction in the use of imprisonment.

This paper examines whether the increases in crime in California can reasonably be attributed to the change in law. It separately examined the changes that took place in homicide, rape, aggravated assault, robbery, burglary, larceny and motor vehicle theft rates.   At the time the study was done, post-Prop 47 data were only available for one year.   However, from a public policy perspective it is often important to examine the effects of single events quickly in order to determine whether additional legislative change is needed or warranted.

The challenge, of course, is to identify a ‘counterfactual’ comparison that allows one to estimate what would have happened if the change in law had not occurred.  The problem is that simply looking at what the situation was before the policy change isn’t adequate since crime rates aren’t stable. Crime rates do not always move in a simple linear fashion. Often there are year-to-year irregularities that are best explained as random events (see Criminological Highlights 17(3)#4).  And examining some nearby jurisdiction(s) may not be adequate because the pattern of crime rates may not have been similar to that in the jurisdiction of interest.  In this study, a “synthetic control group” for California was created by weighting data from other states in a way that “fits California’s crime trends from 2010 to 2014 [the year of Prop 47]” (p. 700).

For homicide, rape, aggravated assault, robbery and burglary, there was no evidence that Prop 47 had an impact.  When compared to the “synthetic control’ which had crime trends between 1970 and 2014 that were very similar to those of California (and in many cases, like California crime rate data, changed direction frequently during this period), California and the ‘untreated control’ looked similar.

For larceny and motor vehicle theft, however, initial comparisons suggested that California, in the year after Prop 47, had higher rates than would be expected.   However, even for these two offences, the data are problematic: the change in rate for motor vehicle thefts was well within the range that one would expect on the basis of random variation across states. For larceny, it would appear that the marginally significant effect was sensitive to the particular states chosen for comparison.  Hence even if one were to accept the idea that this one offence (out of 7) showed an increase after the implementation of Prop 7, the finding is both weak and needs to be interpreted with caution.

Conclusion: Overall, there is no evidence that the reduction of the use of imprisonment that came into force in late 2014 had any effect on 4 violent offences.  The increases in crime that occurred in 2 of the three property offences “are both sensitive to alternative specifications [of the control group] and too small to rule out spuriousness” (p. 708).  These conclusions are based on only one year of data after the reduction in the use of imprisonment.  At a minimum, however, they suggest caution is necessary in assuming that the small increases in crime that followed the change in law had anything to do with the change in law. The paper also illustrates the importance of, and difficulty in, examining changes in crime that occur after a policy change.  One safe conclusion is that caution and care are both needed in assessing the impact of changes in punishment policy on crime.

Reference: Bartos, Bradley J. and Charis E. Kubrin (2018).  Can We Downsize Prisons and Jails Without Compromising Public Safety?  Findings from California’s Prop 47.  Criminology & Public Policy, 17(3), 693-715.

Back to top

The amount of experience a forensic examiner has (e.g., in assessing fingerprints or bitemarks for criminal trials) is more important to jurors in determining how much weight to give forensic evidence than whether the assessments are based on reliable scientific evidence.

In 2009, the US National Academy of Sciences (NAS) found “many forensic sciences to be sorely lacking in scientific support and practices.”  Soon thereafter work began “testing the foundational assumptions of some forensic disciplines” (p. 401).  This paper examines the way ordinary people – potential jurors – assess forensic evidence.

There appear to be at least three factors that might be relevant to the judgements that ordinary people make about whether to believe forensic science evidence.  First, has the method been scientifically tested?  The NAS found that “much forensic science evidence… is introduced in criminal trials without any meaningful scientific validation, determination of error rates, or reliability testing….” (p. 402).  Second, is the forensic expert experienced?  There is some evidence suggesting that when forensic evidence is linguistically complex, jurors rely heavily on the credentials of the expert. Third, does the technique appear to be sophisticated?  It is possible that jurors assume that technological sophistication implies validity. The last two factors (experience and apparent technical sophistication) are obviously problematic in that they are not logically related to validity.

In the first of two experiments 441 volunteers on the web were randomly assigned to read one of two 700-word cases: one involving sexual assault and bitemark evidence; the other involving murder and fingerprint evidence. The forensic evidence in all cases supported the prosecution. Cases described the technique used as having been subjected to a great deal of scientific testing or no testing.   The forensic examiner was described as either having extensive experience or having worked on only a handful of cases.  Finally, the technique that was used was described as being ordinary photographic techniques or as using high-tech ultraviolet photographic equipment. Participants read one of the 16 different descriptions of the case and were asked about the strength of the evidence and the guilt of the accused.

For both cases (the sexual assault/bitemark case and the murder/fingerprint case) the evidence was seen as being stronger when it was described as coming from a validated source and an experienced examiner.  Sophisticated technology increased the perceived strength of the case in the bitemark case.  However, neither the scientific validity of the technique that was used nor the technical sophistication of the technique affected the perception of guilt of the accused. On the bitemark case, however, those who heard that the examiner had lots of experience were more likely to see the accused as guilty (the result consistent with the examiner’s conclusion).

In the second experiment, only one case was used (involving fingerprints). Ordinary citizens, in groups of 6-8, saw a video of a ‘trial’.  They gave pre-deliberation views of the validity of the evidence and the guilt of the accused.  They then deliberated and gave (individual) verdicts again. The consistent finding was that jurors, in their evaluation of the validity of the technique, were affected by the apparent experience of the examiner not the evidence that the technique that was used was scientifically valid or sophisticated.

Conclusion: Contrary to scientific norms, the demonstrated validity of the forensic technique appeared to be unimportant to jurors in these jury simulation experiments.  It would appear that they relied, instead, on the amount of experience the examiner had.  In other words, jurors use the background and experience of the examiner as a proxy for the value of the evidence.  The problem with this is simple: an examiner’s impressive background is more important for jurors when evaluating forensic evidence than the demonstrated validity of the evidence itself.

Reference: Koehler, J.J., N.J. Schweitzer, M.J. Saks, D.E. Mc Quiston (2016) Science, Technology or the Expert Witness: Jurors Judgements about Forensic Science Technology. Psychology, Public Policy and Law, 22, 401-413.

Back to top

Restorative justice conferencing may not be appropriate for young offenders.

There is a growing amount of research suggesting that formal contact with the youth justice system can have negative effects on youths (Criminological Highlights [CH] 16(4)#5, 14(6)#1, 11(4)#3, 17(4)#8).  These negative effects are, perhaps, most obvious in the case of incarceration of youths (CH 17(3)#3, 11(4)#3, 12(5)#7, 12(1)#8, 10(6)#1).

The negative effects have led some people to advocate for approaches involving restorative justice conferences (RJC) for youths notwithstanding the research suggesting the importance of careful matching of youths to programs (CH 17(3)#6) if one is interested in reducing reoffending or improving the lives of youths.

Restorative justice programs, however, have their own issues (CH 6(6)#2, 5(3)#1), among them being that they are often used with youths who have committed relatively minor offences (CH 7(3)#8) and who may not benefit from them. Though some scholars have noted that these programs can be justified on broader dimensions than just their effects on recidivism (CH 9(5)#1), the effect of RJCs on participants are somewhat inconsistent (CH 6(6)#2,  10(1)#6), 16(1)#7).

This paper examines RJCs for youths in the context of research on the developmental and cognitive capacities of youth and suggests that restorative processes may not be ideally suited for youths.  The problem is straightforward: young people may not have the cognitive capacity to be appropriate participants in RJCs. The use of RJCs with youths may be more acceptable to the general public than RJCs with adults but this, alone, does not justify their use with youths.

More important is the suggestion that “RJC is seen as encouraging youth offenders to develop empathy and moral reasoning, make amends to victims, and [as a result] successfully reintegrate into their communities” (p. 452-3). This raises an important question: are youths mature enough to adequately experience two emotions – empathy and shame – that are often seen as important components of RJC?  Not surprisingly, some accounts of RJC suggest that young offenders often drift “from apologetic discourse to mitigating accounts and back again” (p. 457).  The research on the not-fully-developed cognitive processes of youths is consistent with the findings of a systematic review of high quality research on RJC suggested that, if anything, RJCs may be more effective with adults than youth and, perhaps, more effective with respect to serious offences and with those who had already gone through the court process rather than participating in RJCs instead of court (CH 15(4)#4).

Conclusion:  Restorative justice conferences (RJCs) are often seen as an effective alternative to formal youth justice processing as if these were the only two possible approaches to responding to offending by youths.  This paper suggests that just because RJCs may be more acceptable to the public as alternatives to normal criminal justice processing for youths than adults does not mean that they are effective or appropriate  for youths.  In addition, the manner in which RJCs are implemented, and the selection of participants vary so much across programs that it is risky to conclude that a given program is going to be effective unless its selection and operational approaches follow well-researched effective programs. At a minimum, however, given the specific concerns about the appropriateness of RJCs with youths it is important “to identify what specific skills or abilities youth offenders lack, such as verbal ability, and address these issues…” (p. 461).

Reference:  Suzuki, Masahiro and William R. Wood (2018).  Is Restorative Justice Conferencing Appropriate for Youth Offenders?  Criminology and Criminal Justice 18(4), 450-467.

Back to top

Accused people who appear in court in enclosed “prisoners’ docks” are more likely to be convicted (in realistic jury simulation experiments) than if they were sitting next to their lawyers in the body of the court.

Secure prisoners’ docks, where the accused is fully enclosed in a secure cage, are becoming more common in many jurisdictions.  The alternative – which is common in most Australian states in situations in which an accused is not in custody – is that those on trial are seated with their lawyers or are seated immediately behind their lawyers.

The problem with prisoners’ docks, as with accused people showing up in court in prison clothing, is that this “could be seen as a ‘brand of incarceration’ [that] could undermine the presumption of innocence” (p. 323) or could lead to the inference that the accused person is dangerous.   This study used a trial court in Sydney Australia in which a trial, loosely based on a terrorist conspiracy case, was re-enacted before 404 research participants recruited from the jury-eligible list for the city.   This trial was re-enacted 9 times by actors under three conditions.  For approximately one-third of the participants, the accused was sitting at the same table as his lawyer.  For another third of the participants, the accused was at an open dock, and for the third group of participants, the accused was in a glass enclosed dock.  Each condition was presented 3 times.  Ordinary instructions from the judge on ‘reasonable doubt’ were given as were instructions concerning the elements of the offence that needed to be proved.

The participants were assigned unsystematically to one of the three groups (accused sitting with his lawyer, in an open dock, or in a glass-enclosed prisoners’ dock). To address pre-existing differences in participants’ punitiveness, their ‘punitiveness’ was assessed two weeks before participating in the study and the scale value for ‘punitiveness’ was included as a factor in the analyses.  This also allowed the researchers to see if the location of the accused in the court had an effect both for highly punitive and less punitive jurors.

Overall, 36% of the jurors who watched the (simulated) trial with the accused sitting with the lawyer thought he was guilty.   47% of those who saw the accused sitting in an open or glass-enclosed dock thought he was guilty. This relationship held both for those who were high and those who were low on the ‘punitiveness’ scale, though in all conditions those high in punitiveness were more likely than those low on this scale to find the accused guilty.

The research participants were also asked to indicate how strong they found the evidence. Not surprisingly, across all conditions, those who found the evidence weak almost never thought that the accused was guilty.  Similarly, those who saw the evidence as strong – across all conditions – were likely to find the accused guilty.  It was for those in the middle – who saw the evidence as moderately strong – that the placement of the accused mattered.  For this group, only 28% of those who saw the accused sitting with his lawyer would have found him guilty.  This increased to 50% for those who saw the accused in either an open or glass-enclosed dock.

Conclusion: “Being in the [open or glass-enclosed] dock increases the chances that a defendant will be found guilty [in comparison to the situation where the accused is sitting with his lawyer]” (p. 340).  In particular, “jurors whose assessment of the evidence strength is neither very strong nor very weak may turn to other cues (such as whether the accused is in a dock) to assist their decision” (p. 340).

Reference: Rossner, Meredith, David Tait, Blake McKimmie, and Rick Sarre (2017). The Dock on Trial: Courtroom Design and the Presumption of Innocence.   Journal of Law and Society, 44, 317-344.

Back to top

Publication Type

Volume Number


Issue Number