Gauging public opinion on sentencing: can asking jurors help?

Foreword | Using jurors to ascertain public opinion on sentencing has a number of advantages over other methods. Jurors' opinions are based on informed judgements rather than uninformed, intuitive responses and they have detailed knowledge of the offence and a sense of the offender as a real person. Using a sample of jurors from Tasmanian courts, this study examined the utility of using jurors to gauge public opinion on sentencing and as a means of informing the public about crime and sentencing issues. Results indicate that the opinion of jurors towards sentences is not as punitive as public opinion polls would suggest and it would appear that specific knowledge of a case may moderate harsher sentencing attitudes. While there is evidence that jury participation increases confidence in the criminal justice system, the study found that pre-existing perceptions about lenient sentencing may be difficult to change. This was particularly the case when it came to sentencing for sex offences, where jurors were least satisfied with sentencing severity. The study determined that despite there being potential issues around obtaining truly representative population samples, surveying jurors as a means of ascertaining informed public opinion about sentencing seems a good option. A useful view of the general public's opinion on appropriate penalty levels for particular crimes can be elicited, along with their knowledge of crime and sentencing matters.

Judy Putt
General Manager, Research

Why do public attitudes to sentencing matter?

There are at least three reasons why public attitudes to sentencing matter. First, because of the contribution attitudes to sentencing make to public confidence in the criminal justice system. Second, it is generally accepted that sentencing policy and practice should be responsive to public opinion (Roberts 2008). Third, perceptions of public opinion can force changes to the law. Recognising that this is so has led to a burgeoning interest not only in measuring public opinion but in finding the best method of accessing views that are based on an informed judgement rather than an uninformed and intuitive response. For example, the Victorian Sentencing Advisory Council is developing 'a suite of methodologies' to gauge public opinion through representative surveys using vignettes, focus groups and deliberative polls (Gelb 2006). This paper suggests that using members of the public who have been involved in a jury trial has considerable potential as a new methodology. While juries have been the focus of a range of academic studies, there appears to be only one reported study in which jurors were asked anything about sentencing. That study, undertaken in the United Kingdom, explored a wide range of issues, including jurors' understanding of evidence and procedures and their experience of the jury system. However, out of a series of 80 questions, it included just one question about the sentence imposed (Zander & Henderson 1993).

The advantage of using jurors in a sentencing survey and focusing on the particular case they heard is that they will have both a 'strong sense of the offender as a real person' and detailed knowledge about the offence (Lovegrove 2007: 770). The inability of participants to gain any sense of the offender as a person is a constant deficiency in the conventional means of ascertaining public opinion, even in surveys that use vignettes or case studies based on real trials. Moreover, jurors have had their interest in the case aroused by being given an important civic duty in relation to it. They have had direct decision-making responsibility over the question of innocence or guilt. Consequently, they are likely to be in a state of mind to absorb relevant sentencing information and give a considered view. Another advantage is that jurors are randomly selected members of the public who are seen as representing the views, attitudes and beliefs of the general community. These considerations no doubt prompted the former Chief Justice of the High Court of Australia to suggest that asking jurors for their reactions to the sentence in the case they have heard would provide useful information to courts and governments (Gleeson 2005). The study discussed in this paper was inspired by this suggestion.


The study, which began in September 2007 to recruit jurors from trials over a two-year period, has two central aims. First, it explores the possibility of using jurors as a means of ascertaining informed public opinion on sentencing by surveying jury members about sentencing issues in general and about the case that they have heard. Second, it seeks to investigate the usefulness of using the jury as a means of better informing the public about crime and sentencing issues.

This paper will focus on the first aim by examining the response rates from the first 51 trials. It seeks to answer the preliminary research question posed by the study relating to the willingness and feasibility of using jurors as a source of public opinion about sentencing issues. Some preliminary results relating to jurors' reactions to the sentence in the case they heard will also be discussed.

Approach and methodology

In Tasmania there are only two levels of courts: the Magistrates Court and the Supreme Court. All jury trials are heard in the Supreme Court, which sits in three locations: Hobart, Launceston and Burnie. The study recruits jurors from all trials delivering guilty verdicts with the aim of yielding between 100 and 150 trials over the two-year period of the study. All trials with guilty verdicts are included in the study whether the sentence is imposed immediately (which is rare) or on a later occasion, and whether sentencing submissions are made immediately after the verdict (the usual practice) or subsequently. The project tracks the attitudes of participants at three stages. Stage one asks jurors for an initial opinion based on their knowledge of the facts of the case before the sentence is imposed. Stage two occurs after the judge has imposed the sentence and the jurors have read a package prepared by researchers containing the judge's reasons for sentence and further information about the process of sentencing, crime patterns and other contextual matters. Stage three involves an interview, which allows deeper exploration of the jury members' reactions to the case, the reasons for their opinions about the particular sentence and sentencing matters in general.

Questionnaire responses are entered into an SPSS database and the responses and demographic details of each juror are matched to case details (e.g. type of offence, sentence, age of offender etc.). At this stage, the responses have been analysed using univariate and bivariate analysis to reveal only primary descriptive information. The interviews are recorded, transcribed and analysed.

Are jurors willing to be used as a source of public opinion about sentencing?

The study is demanding of jurors, but current responses show that a significant number are willing to participate. After discharging their responsibility for handing down the verdict, jurors are invited by the judge to stay and listen to the sentencing submissions. They are then asked to complete a consent form and short questionnaire which also invites participation in stage two. Stage two requires participants to read the sentencing comments and a booklet about crime and sentencing patterns before completing a second questionnaire. If sentencing submissions have been adjourned, they also have to read the transcript of the submissions. At the end of questionnaire two, respondents are given the option of agreeing to a face-to-face interview in their own time. No incentives are offered for participation. Jurors are already considerably inconvenienced by jury service and remuneration is not always adequate to cover loss of earnings. In some cases, trials are long and the jury may take many hours to reach a verdict. Jurors generally find the responsibility onerous and difficult.

Previous jury studies suggest that jurors are prepared to participate in research projects on topics related to jury service and this study confirms that trend. In a jury study conducted for the New Zealand Law Commission in 1998, an average of 54 percent of jurors in a total of 48 trials participated in interviews of more than an hour's duration about their understanding of the law, the judge's directions and their perceptions of the trial process (Young, Cameron & Tinsley 1999). In a New South Wales study, the response rate for completing questionnaires in sexual assault trials was 92 percent but this dropped to between six to eight jurors per trial if they were allowed to take away the questionnaire rather than complete it in the jury deliberation room (Cashmore & Trimboli 2006). An earlier New South Wales study on the influence of publicity on trials achieved a juror response rate of 36 percent (Chesterman, Chan & Hampton 2001). A study which examined facets of the quality and scope of the jury experience in New South Wales, Victoria and South Australia achieved a response rate from empanelled jurors of 75 percent (O'Brien et al. 2008). Zander and Henderson's English Crown Court jury survey, which asked one sentencing question, was completed by 85 percent of jurors from trials for which at least one juror responded (Zander & Henderson 1993). It is not clear from the Zander and Henderson study whether procedures for obtaining informed consent were used.

This study requires more of jurors than most other jury studies because they are required to remain in court to listen to sentencing submissions after they have delivered their verdict. Sentencing submissions are rarely prolonged; they normally last no more than about half an hour. However, if there is a factual dispute and evidence is called they can last much longer. From 51 trials, 257 jurors participated in stage one of the study by completing questionnaire one - an overall response rate of 42 percent. The median response per trial is five jurors (of a possible 12) but the rate per jury has varied from a nil response (in three trials) to 11 out of 12 jurors (in two trials). Relevant factors appear to be the time of the verdict (trials with a verdict late in the day attract a lower response rate), jury dynamics, degree of juror engagement in the case and factors related to attitudes towards the offender. If the verdict is late in the day, jurors who are willing to participate have been allowed to take home questionnaire one to return or post. As found in a previous study, this adversely affects the response rate, as not all questionnaires are returned. (Cashmore & Trimboli 2006)

Table 1 Juror response rate (as at 31 July 2008) (number)
Trials Jurors asked to participate Questionnaire one responses Agreeing to stage two Questionnaire two returned Jurors agreeing to interview
51 612 257

The three cases in which no responses were received are informative. The last trial in Hobart before Christmas 2007 was a case of causing death by dangerous driving. The verdict was reached late in the afternoon after many hours of deliberation. Sentencing submissions were adjourned and the jurors were allowed to take the questionnaire home. Some jurors did so but none were returned. In another trial, a case of a minor wounding, the judge's associate indicated that the jury appeared uninterested in the case and did not seem to consider the matter to be a serious one. In the third case, the judge's associate suggested that the intimidating nature of the offenders and their families may have explained the nil response. Factors affecting the response rate will be further explored in the interviews.

Of the 257 jurors who participated in stage one, most were willing to participate in stage two. As explained above, questionnaire two, the sentencing remarks and the information booklet, are sent out to the jurors. The current return rate for questionnaire two is 62 percent of stage one participants or 23.5 percent of all jurors. To date, 63 jurors have agreed to participate, so there will be no difficulty in filling the planned quota of 45 interviews (see Table 1 for summary). While the response rate for stage one is not as good as the response rate achieved in some other jury studies, it compares favourably with the response rate in other surveys of public attitudes to sentencing such as the Australian Survey of Social Attitudes which achieved a response rate of 42 percent in the 2007 survey and the New South Wales Sentencing Council's 2008 survey which had a 'nominal response rate' of 11.1 percent (Jones, Weatherburn & McFarlane 2008; Roberts & Indermaur forthcoming).

How representative are jurors of the general population?

A possible flaw in using jurors to measure public opinion is that they may not be representative of the general adult population, a bias that may be exacerbated by the self-selecting nature of jurors who are willing to participate in a jury sentencing survey. The jury is promoted as being 'representative' of community members. However, the extent to which the modern jury is truly representative of the public in the sense of being a cross-section of the community has been questioned. The wide range of exemptions from jury service and the ease with which jurors are excused from service are mentioned as reasons why a jury may not be truly representative (Victorian Law Reform Commission 1997). Citizenship and English proficiency requirements mean that jurors do not reflect the ethnic and cultural diversity of the community (Australian Law Reform Commission 1992). Peremptory challenges further interfere with the ability of jurors to be truly representative (French 2007; Horan & Tait 2007).

In Tasmania, the Juries Act 2003 (Tas), which commenced on 1 January 2006, has drastically reduced the number of occupations that render a person ineligible for jury service and tightened the grounds for application for excuse. However, the jury pool is unlikely to be representative of the general community in terms of ethnic background because eligibility for jury service depends both on electoral enrolment (Juries Act 2003, s 6(1)), which in turn depends on citizenship (Electoral Act 2004 (Tas) s 31; Commonwealth Electoral Act 1918 (Cth) s 93) and an adequate ability to communicate in or understand English (Juries Act 2003, s 68(3) Schedule 2, item 10). Successful applications for excuse and deferral could also reduce the representativeness of the jury pool.

In theory, the representativeness of juries can be tested empirically. Early Australian studies have shown significant age and gender discrepancies between jurors and the general population (e.g. Wilson & Brown 1973). However, a recent study of civil juries in Victoria found that jurors were a fair cross-section of the community in terms of gender and age. Jurors from non-English speaking backgrounds were marginally under-represented and university educated citizens were over-represented (Horan & Tait 2007).

There are no studies of the representativeness of Tasmanian juries and the court does not collect data on this aspect of jury service. However, comparing demographic information collected from questionnaire one for the first 51 trials with Australian Bureau of Statistics 2006 census data for Tasmania shows that respondents are roughly similar in age, gender and country of birth distribution to the general Tasmanian population. Women are slightly over-represented; so too are persons in the 45 to 64 year age group. Those aged 65 and over are under-represented, but this is to be expected because persons over the age of 70 may be excused from service (Juries Act 2003 s 11(3)). As expected, because of citizenship and language requirements, Australian born respondents are also somewhat over-represented in the survey (90% of respondents were Australian born compared with 83% of the Tasmanian population). The major area of difference between Tasmanian jurors and the general population is in educational level. The respondents are more likely to hold post Year 12 qualifications (such as a diploma or bachelor degree) than Tasmanians aged 15 years and over. While the higher level of education among respondents in this study may have affected outcomes, the groups are comparable.

Some preliminary findings of jurors’ sentencing views

Stage one

Questionnaire one asked jurors to nominate the sentence they considered appropriate for the given case. This occurred before the judge imposed the sentence. Juror responses were compared with the judge's sentence. This revealed that when all crime types were aggregated, respondents were more likely to suggest a sentence that was less severe than the judge's sentence. The figures show that 50 percent are less severe, 46 percent are more severe, and only about five percent suggest a sentence of the same severity. This result accords with other research which has found that when the full facts of a case are explained, suggested sentences tend to approximate those imposed by the courts (Lovegrove 2007; Roberts et al. 2003). However, as Figure 1 shows, when the responses are cross-tabulated by type of crime, a more complex pattern emerges. In cases involving sex offences, the juror's proposed sentence was more severe than the judge's in 53 percent of cases. However, the juror's proposed sentence was less severe in 66 percent of property offence cases. Suggested sentences for drug offences were evenly divided between more and less severe sentences than the judge's sentence, but for violent offences, jurors' sentences were marginally less severe than judges' sentences.

Figure 1: Severity of jurors' proposed sentence compared to judge's sentencea (percent)

Figure 1 Severity of jurors' proposed sentence compared to judge's sentence (percent)

Table 2: Juror's opinion on sentencing: questionnaire one and two (percent)
  Type of offencea
  Sex Violence Drugs Property
a: Chi square test indicates significant differences between juror's opinions in questionnaires 1 and 2 p
Q1: Too lenient 79.6 75.5 59.4 55.4
Q2: Too lenient 66.8 70.0 51.8 43.1
Q1: About right 17.6 22.5 33.7 40.2
Q2: About right 31.2 29.3 38.0 51.1
Q1: Too tough 2.8 2.0 6.9 4.5
Q2: Too tough 2.9 0.7 10.2 5.8

Questionnaire two also asks about the aggravating and mitigating factors in relation to the crime tried. This, when analysed, will clarify factors influencing judgements of offence severity. In both questionnaire one and questionnaire two, jurors were asked their views on sentencing in the abstract using a five point Likert scale. Juror opinion on sentencing became less negative in questionnaire two, but respondents still tended to view sentencing practice as too lenient. In Table 2, information from the two questionnaires has been aggregated to three levels. As shown, juror opinion about current sentencing varied between types of offence. In questionnaire one, nearly 80 percent of respondents stated that sentences for sex offences were too lenient, but only 55 percent considered sentences for property offences were too lenient. Sentences for violent and drug offences were found to be too lenient in 76 percent and 59 percent of responses respectively. This is consistent with previous research. Over the last decade, the Australian Survey of Social Attitudes has consistently found between 70 and 80 percent of respondents agree that people who break the law should be given harsher sentences (Indermaur & Roberts 2005; Roberts & Indermaur forthcoming). A recent New South Wales study found 66 per cent of respondents indicated that sentences are either a little too lenient or much too lenient (Jones, Weatherburn & McFarlane 2008).

Asked the same questions in questionnaire two, the proportion of jurors who thought sentences were too lenient was lower. This difference was statistically different for each type of offence. For example, as shown in Table 2, just over two-thirds of questionnaire two respondents thought that sentences for sex offences were too lenient compared to nearly four-fifths in questionnaire one. But despite the overall reduction, the basic pattern remained the same. Jurors were most likely to consider that sentences for sex offences were too lenient and least likely to consider that sentences for property offences were too lenient.

For each crime type, the proportion of respondents who stated that current sentencing practices are 'about right' increased. However, a clear majority of respondents still stated that sentences for sex offenders and violent offenders were too lenient, with the most common response being that sentences for sex offenders were much too lenient. When more data is available, jurors' general views will be cross-tabulated with their views about the particular case they heard. It appears that while most jurors thought that the sentence in their particular case was appropriate, a majority thought that in general, sentences for sex offenders and violent offenders were too lenient. For violent offences in particular, there appears to be a dichotomy between jurors' views in the abstract and their views about the sentence in a given case. This is relevant to the resilience of general perceptions about sentencing severity and public confidence in the criminal justice system. While there is evidence that jury participation increases confidence in the criminal justice system, this study suggests that pre-existing perceptions about issues such as lenient sentencing may be difficult to change (O'Brien et al. 2008). This has implications for efforts to correct public misperceptions about crime and justice and the role of sentencing councils in attempting to do so.

Preliminary results from the study therefore suggest:

  • Given knowledge of a case, the opinions of jurors toward sentences is not as punitive as public opinion polls suggest.
  • There is a dichotomy between jurors' views about sentencing in the abstract and jurors' views about the sentencing in particular cases in which they have been involved.
  • Jurors appear least satisfied with the severity of sentences for sex offenders.


Despite pervasive problems in attaining truly representative public samples, the current study provides some evidence that surveying jurors is a promising method of ascertaining informed public opinion about sentencing. Jurors appear willing to participate and a reasonable response rate is achievable. This could perhaps be improved if funding were provided to allow jurors to be compensated for their involvement. While the time taken to gather a reasonable sample size in a small jurisdiction like Tasmania is considerable, this would not be a drawback in most places. There are cost advantages to this method. The fact that jurors view a real trial means that the costs associated with preparing and presenting scenarios for focus groups or deliberative polls can be avoided. Costs can be lowered further by using court staff to assist with the survey by transcribing sentencing submissions, handing out and collecting self-completed questionnaires and mailing them to the researchers.

Preliminary findings suggest that surveying jurors can elicit interesting and useful information about sentencing. By asking about the particular offence tried, informed responses can be obtained about public views of appropriate penalty levels for particular crimes. Comparison with general views about sentencing, analysis of the demographic juror data and changes in views will provide interesting insights into the nature of public opinion and its relationship with knowledge of crimes and sentencing matters.

Perhaps the best way to understand and view jury surveys is as a complementary approach to measuring public opinion. Insofar as results are similar to those found in more formal surveys, a degree of convergent validity is present and any differences observed are worthy of further investigation. Thus, jury surveys provide a useful addition to the current suite of methodologies.


Australian Bureau of Statistics 2006. Census of population and housing, Tasmania (State). ABS cat. no. 2068.0

Australian Law Reform Commission 1992. Multiculturalism and the law. Report no. 57

Cashmore J & Trimboli L 2006. Child sexual assault trials: a survey of juror perceptions 2006 Crime and justice bulletin no. 102

Chesterman M, Chan J & Hampton S 2001. Managing prejudicial publicity: an empirical study of criminal jury trials in New South Wales. NSW: Law and Justice Foundation of New South Wales

French V 2007. Juries—a central pillar or an obstacle to a fair and timely criminal justice system? Reform 90: 40–42

Gelb K 2006. Myths and misconceptions: public opinion versus public judgment about sentencing. Melbourne: Sentencing Advisory Council

Gleeson AM 2005. Out of touch or out of reach? The judicial review 7: 241

Horan J & Tait D 2007. Do juries adequately represent the community? A case study of civil juries in Victoria. Journal of judicial administration 16: 179–199

Indermaur D & Roberts L 2005. Perceptions of crime and justice, in Wilson S, Meagher G, Gibson R, Denemark D & Western M (eds), Australian social attitudes: the first report. Sydney: UNSW Press: 141–160

Jones C, Weatherburn D & McFarlane K 2008. Public confidence in the New South Wales criminal justice system. Crime and justice bulletin no. 118

Lovegrove A 2007. Public opinion, sentencing and lenience: an empirical study involving judges consulting the community. Criminal law review 2007: 769

O’Brien K, Goodman-Delahunty J, Clough J & Pratley J 2008. Factors affecting juror satisfaction and confidence in New South Wales, Victoria and South Australia. Trends & issues in criminal justice no. 354. series/tandi/341-360/tandi354.html

Roberts JV 2008. The evolving role of public opinion, in Freiberg A and Gelb K (eds), Penal populism, sentencing councils and sentencing policy. Sydney: Hawkins Press: 15–30

Roberts JV, Stalans L, Indermaur D & Hough M 2003. Penal populism and public opinion: lessons from five countries. Oxford: Oxford University Press

Roberts L & Indermaur D (forthcoming). What Australians think about crime and justice: results from the 2007 Survey of Social Attitudes. Research and public policy series. Canberra: Australian Institute of Criminology Victorian Law Reform Commission 1997. Jury service in Victoria, final report. Vol 3

Wilson P & Brown JW 1973. Crime and the community. Queensland: University of Queensland Press

Young W, Cameron N & Tinsley Y 1999. Juries in criminal trials, part two: a summary of research findings. Wellington: Law Commission of New Zealand

Zander M & Henderson P 1993. The royal commission on criminal justice: crown court

This paper is taken from the report of research commissioned by the Criminology Research Council.