Are Juries Fair?

Constitution and Access to Justice – Analytical Services supports effective policy development and delivery within the Ministry of Justice by providing high-quality social research, statistics and economic analysis to influence decision-making and encourage informed debate. © Crown Copyright 2010. Extracts from this document may be reproduced for non-commercial purposes on condition that the source is acknowledged. First Published 2010 ISBN: 978 1 84099 326 4 Acknowledgements.

Research with juries rightly carries concerns about protecting the secrecy of deliberations, and I am especially grateful to Her Majesty’s Courts Service (HMCS) for facilitating my work with jurors at courts in London, Nottingham and Winchester, and to the jury officers, court managers and judges at these courts for their assistance. Nigel Balmer, UCL Faculty of Laws and Legal Services Research Centre, played a key role in modelling the analysis of CREST data in this report and made an important contribution to the study.

Given the scope of the study and its implications for the criminal justice system, a special Project Steering Group was convened from the Ministry of Justice (MoJ), HMCS, Office for Criminal Justice Reform (OCJR), the judiciary, Attorney General’s Office and Home Office (HO), and I am grateful to all the members for providing invaluable advice on the research. The report also benefited greatly from advice and comments from: Miranda Hill, Lydia Jonson, Gary Hopper, Mike Ainsworth, John Marais, David Perry QC, Dr David Lagnado, Marc Davies, Tim Strouts, John Samuels QC, Tina Golton and three anonymous peer reviewers.

Kevin Dibdin oversaw the provision of CREST data and Rachel Thomas helped collect data in the media reporting study. At MoJ’s Constitution and Access to Justice – Analytical Services (CAJAS), I am especially grateful to Jessica Haskins and Laura Blakeborough for overseeing the project and to Sally Attwood, Eleanor Brown and Michelle Diver for their assistance. Above all, I am indebted to all the jurors who must remain anonymous but so willingly agreed to participate in the studies and generously gave their time to assist with the research.

Author Professor Cheryl Thomas is a member of the Centre for Empirical Legal Studies in the Faculty of Laws at University College London. She is the author of Diversity and Fairness in the Jury System (2007), the precursor to this report. Professor Thomas is a specialist in judicial studies and has conducted research here and in other jurisdictions on juries, the role of diversity in the justice system, legal decisionmaking and the appointment and training of judges.

She has served as a special consultant on judicial affairs to numerous organisations including Her Majesty’s Crown Prosecution Service Inspectorate, Judicial Studies Board, Commission for Judicial Appointments, Council of Europe and the French government. Disclaimer The views expressed are those of the author and are not necessarily shared by the Ministry of Justice (nor do they represent Government policy) Contents List of tables List of figures Summary 1. Context 1. 1 1. 2 1. 3 1. 4 1. 5 1. 6 1. 7 2.

The fairness of jury decision-making Racial discrimination Consistency of jury verdicts Comprehension of legal instructions Jury impropriety Impact of media coverage and the internet Main research questions i 1 1 2 3 3 4 5 6 7 7 7 10 12 14 14 26 35 40 45 51 53 54 Demographics of jurors and local populations in case simulations at Nottingham and Winchester Crown Courts Appendix 2: Case simulation juror decision-making data analysis 57 59 Approach 2. 1 2. 2 2. 3 2. 4 Multi-method approach Case simulation Large-scale verdict analysis (CREST) Post-verdict surveys 3. Results 3.

1 3. 2 3. 3 3. 4 Race and jury decision-making General trends in jury trials 2006-08 Juror comprehension Media reporting and internet use 4. 5. 6. Implications Further research Additional resources References Appendix 1: Appendix 3: All pleas and jury verdicts at all Crown Courts 2006–08 by defendant ethnicity (at charge level) 61 Appendix 4: Relationship between defendant, juror and local population ethnicity for all Crown Courts in England and Wales 62 Appendix 5: Analytical model for analysis of impact of ethnicity on verdict (court effects) 65 List of tables 2. 1: 2.

2: 2. 3: Sample size in case simulation study Sample size for CREST data analysis Sample size for media reporting and internet study 9 10 12 3. 1: Crown Court charges by ethnicity of defendant and offence type: 2006–08 (n=551,669) 21 3. 2: 3. 3: Jury verdicts for White defendants charged with racial offences: 2006–08 All charges, pleas and verdicts in Crown Courts: 2006–08 25 27 A3. 1: All pleas on charges by defendant ethnicity A3. 2: All outcomes where juries deliberated by defendant ethnicity 61 61 A4. 1: BME population, jurors and defendants in the London Region A4.

2: BME population, jurors and defendants in the South Eastern Region A4. 3: BME population, jurors and defendants in the Midlands Region A4. 4: BME population, jurors and defendants in the North East Region A4. 5: BME population, jurors and defendants in the Northern Region A4. 6: BME population, jurors and defendants in the Wales & Chester Region A4. 7: BME population, jurors and defendants in the Western Region 62 62 63 63 63 64 64 A5. 1: Discrepancy between BME defendants in each court and BME representation in the court catchment area population 66 List of figures 3. 1: 3. 2: 3. 3: 3. 4: 3. 5: 3. 6: 3. 7: 3. 8: 3.

9: Jury verdicts in case simulations by defendant ethnicity White juror guilty votes for White, Asian and Black defendants at 3 courts White juror guilty votes by defendant and victim ethnicity Juror first and final guilty votes by gender Defendant not guilty pleas by ethnicity and offence type: 2006–08 Jury conviction rates by defendant ethnicity: 2006–08 (n=68,451) Jury conviction rate by defendant ethnicity and offence type (n=66,889) Outcomes of all not guilty pleas: 2006–08 (n=191,140) Proportion of Crown Court charges and jury verdicts by offence type: 2006–08 15 17 18 19 22 23 24 27 28 29 29 30 32 34 35 36 38 38 39 41 42 42 43.

3. 10: Jury conviction rates by Blackstone’s criminal offence type (n=66,889) 3. 11: Jury conviction rates in homicide-related offences: 2006–08 (n=2,040) 3. 12: Specific offences with highest and lowest jury conviction rates: 2006–08 3. 13: Jury conviction rates for rape by complainant age and gender (n=4,310) 3. 14: Jury conviction rates by severity of offence (n=66,801) 3. 15: Jury conviction rate by number of charges (n=22,907) 3. 16: Jurors’ self-reported understanding of judge’s directions 3. 17: Age groups who fully understood oral instructions on the law (n=195) 3.

18: Juror comprehension with oral and written directions 3. 19: Jurors’ view of need for information about deliberations (n=196) 3. 20: Juror recall of pre-trial and in-trial media coverage (n=157) 3. 21: Where jurors saw or heard media reports (n=265) 3. 22: Emphasis jurors recalled in media coverage (n=163) 3. 23: Juror use of internet during trial (n=643) Summary This research asks: How fair is the jury decision-making process? It explores a number of aspects of jury fairness for the first time in this country, and asks specifically: ? ? ? ? ? ? ? ? Do all-White juries discriminate against BME defendants?

Do jurors racially stereotype defendants? Do juries at certain courts rarely convict? Do juries rarely convict on certain offences? Do jurors understand legal directions? Do jurors know what to do about improper conduct in the jury room? Are jurors aware of media coverage of their cases? How is the internet affecting jury trials? The research used a multi-method approach to examine these issues: ? ? ? case simulation with real juries at Crown Courts (involving 797 jurors on 68 juries); large-scale analysis of all actual jury verdicts in 2006–08 (over 68,000 verdicts); post-verdict survey of jurors (668 jurors in 62 cases).

The study found little evidence that juries are not fair. However, it identifies several areas where the criminal justice system should better assist jurors in performing this vital role. The study also demonstrates that section 8 of the Contempt of Court Act 1981 does not prevent comprehensive research about how juries reach their verdicts and that research from other jurisdictions should not be relied upon to understand juries in this country. All-White juries and BME defendants.

A key question remained to be answered from a recent jury study: Do all-White juries discriminate against Black and minority ethnic (BME) defendants? A large number of all-White juries tried an identical case in which only the race of defendants and victims was varied. This enabled the study to determine if race actually affects jury decision-making. The case simulation was conducted with 41 all-White juries at Winchester and Nottingham Crown Courts (478 jurors). It replicated an earlier study of racially mixed juries at Blackfriars Crown Court in London (27 juries with 319 jurors).

Earlier research found that juries at Winchester and Nottingham will almost always be all-White. The juror catchment area for i Nottingham is predominantly White but includes neighbourhoods with high levels of ethnic diversity; the Winchester juror catchment area is overwhelmingly White throughout. The study examined decision-making at the jury verdict level: ? The key finding was that verdicts of all-White juries did not discriminate against BME defendants. Jury verdicts at both courts showed no tendency for all-White juries to convict a Black or Asian defendant more than a White defendant. ?

All-White juries at Winchester had almost identical verdicts for White and BME defendants, but all-White juries at Nottingham had particular difficulty reaching a verdict involving a BME defendant or BME victim. ? This suggests that local population dynamics may play a role in jury decisionmaking. The study also examined the votes of all individual jurors who sat on these juries: ? White jurors serving on racially mixed juries and on all-White juries had similar patterns of decision-making for White, Black and Asian defendants. But White jurors on racially mixed juries had lower conviction rates overall.

? White jurors in a racially diverse area (Nottingham) appeared sensitive to cases involving inter-racial conflict. These jurors were significantly more likely to convict the White defendant when he was accused of assaulting a BME victim compared to a White victim. No similar trend was found with White jurors in Winchester. ? White jurors serving on all-White juries did not racially stereotype defendants as more or less likely to commit certain offences based on race. The same result was found with both White and BME jurors serving on racially mixed juries.?

The only other personal characteristic that appeared to affect juror decision-making was gender. Female jurors were more open to persuasion to change their vote in deliberations than male jurors. Male jurors rarely changed their mind. Jury verdicts in Crown Courts in England and Wales 2006–08 This study analysed a large dataset of all charges in all Crown Courts in England and Wales (551,669) where outcomes occurred between 1 October 2006 and 31 March 2008. ? ? Such a large dataset enabled a statistically reliable analysis of trends in jury verdicts.

The study examined whether defendant ethnicity, offence type, court, severity of offence or number of charges in a trial had any correlation to jury verdicts. ii Disproportionality for BME defendants in Crown Court trials It is already known that members of BME groups are disproportionately represented among those stopped, searched, arrested, charged and in prison. This study found that: ? BME defendants are consistently more likely than White defendants to plead not guilty to charges in all of the 12 general offence categories used in this study except one (falsification, forgery and counterfeiting).

? ? BME defendants are three and half times more likely to face a jury verdict in the Crown Court relative to their representation in the general population. However, jury verdicts showed only small differences based on defendant ethnicity. White and Asian defendants both had a 63% jury conviction rate; Black defendants had a 67% jury conviction rate. This indicates that one stage in the criminal justice system where BME groups do not face persistent disproportionality is when a jury reaches a verdict. Appearance of jury fairness

While these findings strongly suggest that racially balanced juries are not needed to ensure fair decision-making in jury trials with BME defendants, concerns about the appearance of fairness with all-White juries may still remain. ? The study found that in all Crown Courts, the proportion of BME defendants is greater than the proportion of BME groups in the local population or BME jurors at each court. ? Concerns about the appearance of jury fairness are likely to arise in courts where all-White juries try substantial numbers of BME defendants or try White defendants accused of racial crimes against BME victims.

? To address these concerns, HMCS should ensure that court users understand how jury pools are selected and how representative they are of the locality. Scope and effectiveness of jury trials Most charges brought against defendants in the Crown Court are not decided by a jury: ? ? ? Only 12% of all charges are decided by jury deliberation. 59% of all charges result in a guilty plea by a defendant. Of the remaining charges where a defendant pleads not guilty and therefore gives rise to a potential jury trial, 36% are decided by jury deliberation. iii Juries overall appear efficient and effective: ? ? ? ?

Once a jury is sworn it reaches a verdict by deliberation on 89% of all charges (judges direct jury verdicts on 11% of charges). Once juries deliberate they reach verdicts on virtually all charges (only 0. 6% of all verdicts are hung juries). Juries convict on almost two-thirds (64%) of all charges presented to them. Juries are rarely discharged (less than 1% of sworn juries). Jury conviction rates Offence type had an impact on the probability of a jury reaching a guilty verdict. ? ? Falsification, deception, drugs and theft offences are the general offence types most likely to produce a guilty jury verdict.

Non-fatal offences against the person are least likely to result in a jury conviction, although juries still reach guilty verdicts more often than not here (52% conviction rate). Conviction rates for specific offences within general offence types can vary substantially. ? The category of homicide-related offences has some of the lowest jury conviction rates (threatening to kill 36%, manslaughter 48%, attempted murder 47%) but also some of the highest jury conviction rates (death by dangerous driving 85%, murder 77%).

Differences in jury conviction rates for different specific offences suggest that juries try defendants on the evidence and the law. ? Offences where the strongest direct evidence is likely to exist against a defendant appear to have the highest conviction rates (making indecent photographs of a child 89%, drugs possession with intent to supply 84%, death by dangerous driving 85%). ? Cases where juries must be sure of the state of mind of a defendant or complainant in order to convict appear to have the lowest conviction rates (threatening to kill 36%, attempted murder 47%, GBH 48%).

iv Misconceptions about jury verdicts in rape cases Contrary to popular belief and previous government reports, juries actually convict more often than they acquit in rape cases (55% jury conviction rate). ? ? Other serious offences (attempted murder, manslaughter, GBH) have lower jury conviction rates than rape. A previous Home Office study stating that jury acquittals were more common than convictions was based on a small number of verdicts (181) in a few courts. Current findings cover all jury rape verdicts in all courts in 2006–08 (4,310).

? Jury conviction rates for rape vary according to the gender and age of the complainant, with high conviction rates for some female complainants and low conviction rates for some male complainants. This challenges the view that juries’ failure to convict in rape cases is due to juror bias against female complainants. ? Juries are not primarily responsible for the low conviction rate on rape allegations. Misconceptions about jury verdicts in certain courts There are variations in jury conviction rates between Crown Courts. ?

In courts with over 1,000 jury verdicts in 2006–08, the conviction rate ranged from 69% to 53%. There were no courts with a higher jury acquittal than conviction rate, and this dispels the myth that there are courts where juries rarely convict. ? Variations in court conviction rates could be due to differences in the types of offences presented to juries at different courts; differences in public attitudes to crime and justice in different communities; or variations in police evidence gathering or prosecution or judicial handling of jury trials. ?

It is recommended that the underlying reasons for substantial variations in jury conviction rates between Crown Courts be examined further. Multiple charges ? ? The number of charges against a defendant affected the likelihood of the jury returning at least one guilty verdict. The probability of a guilty jury verdict increased with the number of charges, rising steeply from 40% with one charge to 80% with five charges. v Juror comprehension of judicial directions This study involved 797 jurors at three courts who all saw the same simulated trial and heard exactly the same judicial directions on the law.

? There is not a consistent view among jurors at all courts about their ability to understand judicial directions. Most jurors at Blackfriars (69%) and Winchester (68%) felt they were able to understand the directions, while most jurors at Nottingham (51%) felt the directions were difficult to understand. Jurors’ actual comprehension of the judge’s legal directions was also examined. ? While over half of the jurors perceived the judge’s directions as easy to understand, only a minority (31%) actually understood the directions fully in the legal terms used by the judge. ?

Younger jurors were better able than older jurors to comprehend the legal instructions, with comprehension of directions on the law declining as the age of the juror increased. A written summary of the judge’s directions on the law given to jurors at the time of the judge’s oral instructions improved juror comprehension of the law: ? ? ? ? The proportion of jurors who fully understood the legal questions in the case in the terms used by the judge increased from 31% to 48% with written instructions. The judiciary should reconsider implementing the Auld recommendations for issuing jurors with written aide memoires on the law in all cases.

An assessment should also be made of how many judges already use written instructions, when and how often. Further research should be conducted as a matter of priority to identify the most effective tools for increasing juror comprehension of judicial directions. Jury deliberations and impropriety This study involved 196 jurors at Winchester who had served on a jury and therefore should have been instructed by a judge on improper conduct. ? ? Almost half (48%) of all jurors said they either did not know or were uncertain what to do if something improper occurred in the jury deliberating room.

Most of these jurors (67%) also felt they should be given more information about how to conduct deliberations. vi ? ? An even larger majority of these jurors (82%) felt it was correct that jurors should not be allowed to speak about what happens in the deliberating room. This was only a limited exploration of these issues. The findings indicate that further research should be conducted to determine what jurors understand improper jury behaviour to be; how jurors think they should deal with improper jury conduct; and what type of information jurors want about deliberations.

Media reporting of jury trials and juror use of the internet The study was conducted in three different locations (Nottingham, Winchester and London) and included 62 cases and 668 jurors. The sample included both long, high profile cases and standard cases lasting less than two weeks with little media coverage. ? ? Jurors serving on high profile cases were almost seven times more likely to recall media coverage (70%) than jurors serving on standard cases (11%). Most jurors who recalled media reports of their case saw or heard reports only during the time their trial was going on.

This provides the first empirical evidence in this country of the “fade factor” in jury trials (the further away media reports are from a trial the more likely they are to fade from jurors’ memories). ? ? But a third of jurors (35%) on high profile cases remembered pre-trial coverage. In high profile cases, jurors recalled media reports of their cases from a range of media outlets, with television (66%) and national newspapers (53%) the two main sources. This contrasts with jurors’ recall of media reports in standard cases, where local newspapers accounted for almost all (77%) coverage recalled.

? Most jurors (66%) in high profile cases who recalled media coverage either did not or could not remember it having any particular slant. Where jurors did recall any emphasis, almost all recalled it suggesting the defendant was guilty. ? In high profile cases, 20% of jurors who recalled media reports of their case said they found it difficult to put these reports out of their mind while serving as a juror. The findings show that in high profile cases almost three-quarters of jurors will be aware of media coverage of their case.

It would be helpful to know how these jurors perceive this media coverage, what particular type of pre-trial coverage jurors’ recall and what type of coverage some jurors find difficult to put out of their minds. vii The internet All jurors who looked for information about their case during the trial looked on the internet. ? More jurors said they saw information on the internet than admitted looking for it on the internet. In high profile cases 26% said they saw information on the internet compared to 12% who said they looked. In standard cases 13% said they saw information compared to 5% who said they looked.

? In the study jurors were admitting to doing something they should have been told by the judge not to do. This may explain why more jurors said they saw reports on the internet than said they looked on the internet. ? Among all jurors who said they looked for information on the internet, most (68%) were over 30 years old. Among jurors on high profile cases, an even higher percentage (81%) of those who looked for information on the internet were over 30. The findings raise a number of questions that should be examined further: do jurors realise they are not suppose to use the internet?

How do they use the internet: do they just look for information or do they also discuss the case on social networking sites? What type of judicial instruction would be most effective in preventing jurors from looking for information about their case on the internet? Recommendations The jury system imposes a duty on citizens to participate in the criminal justice system and to decide the most serious criminal cases in this country. It is therefore crucially important that jurors are provided with the most effective tools to carry out that responsibility.

The findings on juror comprehension of the law, impropriety, internet use and jurors’ views about deliberations suggest that jurors want and need new tools to better understand the process. A concerted effort should be made by those responsible for the criminal justice system to identify the most effective means of ensuring the highest levels of juror understanding in criminal jury trials. viii Written juror guidelines To address both jury impropriety in general and juror use of the internet, the judiciary and HMCS should consider issuing every sworn juror with written guidelines clearly outlining the requirements for serving on a trial.

? ? The written guidelines should acknowledge the value of the juror’s role and clearly explain what improper behaviour is, why it is wrong and what to do about it. The judge should review the requirements with jurors as soon as they are sworn. This should include a fuller direction to jurors on why they should not use the internet to look for information or discuss their case. ? ? Jurors should be required to keep the guidelines with them throughout the trial. Piloting should be carried out to determine what form of written guidelines and judicial directions are most comprehensible to jurors and are most likely to be taken seriously.

ix 1. 1. 1 Context The fairness of jury decision-making Even though juries decide less than 1% of all criminal cases in England and Wales, defendants in these cases are charged with the most serious criminal offences and face the greatest possible loss of liberty. The fairness of jury decision-making is therefore of fundamental importance to the criminal justice system. Opinion polls consistently show strong public support for jury trials (Bar Council, 2002; ICM, 2007; Thomas, 2007). To preserve this confidence in the justice system it is crucial that allegations of jury bias are systematically explored.

Yet in this country, little is known about how juries reach verdicts. This lack of knowledge about jury decision-making is usually incorrectly attributed to section 8 of the Contempt of Court Act 1981. This makes it a criminal offence to “obtain, disclose or solicit any particulars of statements made, opinions expressed, arguments advanced, or votes cast by members of a jury in the course of their deliberations”. This does not, in fact, prevent almost all research about and with juries (Thomas, 2008).

But its existence has created confusion about what jury research can and cannot be conducted and has contributed to an information vacuum about juries in this country. In the absence of empirical evidence, inevitably domestic anecdote and research from other jurisdictions has been relied upon to draw conclusions about jury decision-making here. It has long been claimed for instance that juries in certain Crown Courts hardly ever convict (Hansard, 1982). Anecdotal reports also appear from time to time that juries do not understand legal directions or that improper conduct occurs in the jury room (Daly & Pattenden, 2005).

Most jury research has been conducted in the United States. A large amount of this research is concerned with the issue of race 1 and juries, some of which has indicated that White jurors discriminate against non-White defendants (Sommers & Ellsworth, 2003). While this research has developed valuable methodologies, there are important differences between juries in America and in this country that mean caution needs to be exercised in drawing similar conclusions about juries here (Thomas, 2007).

In order to provide clear empirical evidence about juries in this country, this study tackles a number of important issues concerning the fairness of jury decision-making, including: discrimination, consistency, comprehension, improper conduct and media reporting. 1 There is no universally accepted terminology in this field. American jury research uses the term “race”, while the term “ethnicity” is most commonly used in Britain. 1 1. 2 Racial discrimination For several decades there have been claims that the racial composition of juries affects jury fairness in this country.

Where Black and minority ethnic (BME) defendants are tried by an all-White jury, the concerns are two-fold: that all-White juries may actually treat BME defendants unfairly and that all-White juries simply appear unfair. Similar concerns can also arise where all-White juries try White defendants accused of racially motivated crimes involving BME victims. There has been no empirical research in this country to show how often BME defendants or racially motivated crimes are tried by all-White juries, and until recently there has been no research into whether race actually affects jury verdicts.

As a result, American research showing White juror bias against non-White defendants has been relied upon to suggest that juries here are likely to be racially biased. Both the Auld Review of the Criminal Courts (2001) and the Runciman Royal Commission on Criminal Justice (1993) based recommendations for racially mixed juries on this assumption, but both recognised that these recommendations had to be made in the absence of empirical evidence in this country.

In 2007, the Ministry of Justice (MoJ) published the findings of the Jury Diversity Project (Thomas, 2007), which revealed that most defendants in most Crown Courts outside London will be tried by an all-White jury. This does not reflect any failure in juror summoning; it is simply the consequence of population dynamics in court catchment areas and the process of random summoning. The Project also conducted the first empirical study here of how race affects jury decision-making. That study at Blackfriars Crown Court in London 2 found that racially mixed juries did not discriminate against either BME or White defendants.

However, a key question remained to be answered: Do all-White juries outside London discriminate against BME defendants? This study was commissioned to address this question and to determine how often all-White juries try BME defendants and try White defendants in racerelated crimes. As part of the study of racial discrimination, the current research also explores whether jurors in this country racially stereotype defendants. The Macpherson Report (1999) identified racial stereotyping as a sign of institutional racism.

Elsewhere, the type of crime a defendant is accused of committing has been found to affect the likelihood of White juror bias. Sunnafrank and Fontes (1983) found that White jurors in America viewed white-collar crimes (such as embezzlement) as consistent with a stereotype of White criminals, but that more violent 2 Herein referred to as the Blackfriars study. 2 crimes (such as assault) were associated with a Black criminal stereotype. This research looks at whether similar stereotyping occurs here. 1. 3 Consistency of jury verdicts Juries at certain Crown Courts are commonly thought to have very low conviction rates.

Snaresbrook Crown Court, for instance, has long had a reputation for juries that do not convict (Hansard, 1982). It is possible that if jury conviction rates do vary significantly between courts this simply reflects differences in public attitudes to c