Understanding community attitudes toward miscarriages of justice: the role of social characteristics on perceptions of wrongfully convicted exonerees

ABSTRACT

Objective: Not all exonerees are perceived equally. Attitudes toward exonerees are negatively impacted by social characteristics, such as race and criminal history; providing significant challenges for exonerees attempting to reintegrate into the community. The present study examined whether exoneree race, gender and criminal history influenced ratings of desired social distance. Method: A community sample of 1043 females and males completed an online survey. Participants were asked to read a vignette of a mock news article that described a wrongful conviction, then complete a desired social distance scale. Exoneree gender, race and criminal history were manipulated in the news article, with a minimum of 120 participants randomised to eight vignette conditions. Results: Exoneree race and the presence of a prior criminal history had significant main effects on participant attitudes regarding desired social distance. There were no interactions between exoneree gender, race and priors, but differences did emerge suggesting differential views of desired social distance for females and older participants when the exoneree had a prior offence. Conclusion: Social characteristics, demographics and individual differences may be particularly important in forming community attitudes toward wrongful conviction. This holds important implications for exoneree reintegration and community attitudes toward the Australian criminal justice system.

KEYWORDS:

• Wrongful convictions
• gender
• race
• criminal history
• attitudes
• social distance

Since the first DNA exoneration in 1989, over 3,100 wrongful convictions have been identified in the United States (US) alone (National Registry of Exonerations, 2022). In response, an ‘innocence movement’ emerged as a driving force of criminal justice reform by uncovering error and overturning the convictions of innocent people imprisoned for crimes they did not commit (Norris, 2017). By exposing flaws in the US, this movement has increased the global awareness of wrongful convictions and the need for legal, social, and political reform required to achieve justice (Dioso-Villa, 2015; Naughton, 2014; Roach, 2015).

To date, the study of wrongful conviction has been dominated by attempts to understand the immediate causes of error within the criminal justice system (Leo, 2017). Subsequent research has developed a list of contributing factors including eyewitness testimony, unreliable forensic science evidence, false confessions, incentivised witnesses, government misconduct, and inadequate defence counsel (Garrett, 2020). In combination with the advocacy efforts of the innocence movement, this research has contributed to reforms in areas such as police interviewing and forensic science with the aim of reducing wrongful convictions (Bang et al., 2018; National Research Council, 2009). Despite the success of these reforms, the rate of wrongful conviction has not slowed down. Critiquing this work, some authors have pointed to the importance of additionally considering social context when understanding how and why wrongful convictions occur (Dioso-Villa et al., 2016; Norris & Bonventre, 2015). A number of notable earlier studies that did consider social context found that populations that are more likely to be dispossessed in society are more inclined to believe that wrongful convictions occur frequently (Bell et al., 2008; Bell & Clow, 2007; Ramsey & Frank, 2007; Ricciardelli & Clow, 2012; Smith et al., 2011; Zalman et al., 2012). While much of this earlier work was exploratory in nature, their examination of issues like exoneree stigma and criminality offered a potentially valuable additional means of understanding societal expectations and pressures faced by exonerees upon their return to the community.

Literature review

Where advocates and activists have long acknowledged the relationship between wrongful conviction, institutional racism, sexism, and marginalisation, scholarship has often left this relationship under-explored (Free & Ruesink, 2016; Macfarlane & Stratton, 2016). The emerging research in these areas highlights how a broad range of social, cultural, and political factors contribute to wrongful conviction.

The gendered nature of wrongful convictions is defined by the nature of crimes and the evidence forming the basis of their conviction; wrongfully convicted women are more likely to be convicted of harming family members and children. False confession and false or misleading forensic evidence have been found to be common contributing factors (or errors) for women who have been found to be wrongfully convicted (Lang, 2021; National Registry of Exonerations, 2022). Other research has highlighted that women are over-represented in wrongful convictions where no crime has occurred; these often being linked to cultural expectations and gender norms particularly surrounding women in caregiving roles, women’s sexuality, and use of violence (Free & Ruesink, 2016; Jackson & Gross, 2014; Lewis & Sommervold, 2015; Ruesink & Free, 2005). Exploring this argument further, Parkes and Cunliffe (2015) suggest that women, particularly First Nations women, are more likely to plead guilty to crimes they did not commit. The authors posited that this was due to the pressures exerted by various legal and social mechanisms such as sentence discounts, denial of bail, and a sense of responsibility for the harm despite their innocence. These findings support a range of earlier and more recent studies (Hussemann & Siegel, 2019; Jones, 2011; Stubbs & Tolmie, 2008), and reflect data from the United States (US) that indicate that women constitute less than 10% percent of all known exonerees and nearly 20% of no-crime exonerees (Henry, 2018, p. 693). The impact of these errors is significant, with women experiencing both personal trauma and social stigma (Konvisser, 2015; Westervelt & Cook, 2010). Based upon the distinctive characteristics of women’s wrongful convictions identified in the broader literature, coupled with women’s self-reported experiences of stigma, and the unique history and profile of wrongful convictions in Australia, gender may have an important role to play in public perceptions of exonerees.

Acknowledging the innocence movement’s late recognition of gendered factors pertaining to wrongful conviction, some have rightly noted the lack of other intersectional approaches to the harms of wrongful conviction (Webster & Miller, 2015). For example, significant racial differences have been reported in the types of offences where wrongful convictions occur, with Free and Ruesink (2016) highlighting that African-American women are more likely to be wrongfully convicted of drug offences (40% of total exonerees) as compared to white women. While similar effects have been seen in Australia, the identification of ‘factual’ innocence remains limited; a result of unidentified vulnerabilities as an outcome of a colonial history often unable to recognise the privileges of power and the inherent discriminatory power of the criminal justice system toward racial minorities (Stratton & Sigamoney, 2020). The absence of ‘factual’ innocence, though, does not disguise particularly concerning broader justice outcomes for First Nations women in Australia who have been found to represent the fastest-growing incarcerated population in the country (Tubex & Cox, 2020). Parkes and Cunliffe (2015) posit that the absence of these types of intersectional conversations about wrongful convictions represents an avoidance in addressing the racial bias that exacerbates vulnerability to wrongful conviction for Indigenous people.

Broader social patterns of racial differences and discrimination are also reflected in the outcomes of wrongful conviction that see racial minorities experience more obstacles and lengthier pathways to exoneration (Gross, 2008; Olney & Bonn, 2015). These findings reflect the role of racism, racial discrimination and stereotyping on the occurrence of wrongful conviction that authors have long argued have historically been overlooked by the criminal justice system (Free, 2017; Smith & Hattery, 2011). The history of racial over-representation is aptly highlighted in the first repository of Australian wrongful conviction, recording 71 cases between 1922 and 2015, in which Aboriginal and Torres Strait people represented 17% of these cases (compared to 3.3% of the Australian population) (Australian Bureau of Statistics, 2016; Dioso-Villa, 2015). Other cognate research supports this, reporting that Indigenous status is a precursor to differential treatment in the criminal justice system (Roach, 2015; Sangha et al., 2010).

An individual’s criminal history can also expose them to error within the criminal justice system. While the ‘tunnel vision’ of investigators and legal professionals has often been highlighted as a factor behind error, the presence of criminal history and perceptions of criminality have also been found to contribute to these biases (Gould et al., 2014; Lowrey-Kinberg et al., 2019). These findings are aligned with global trends in police training and legal systems which encourage investigators to identify suspects based on prior criminal activity, as well as for prosecutors to present tendency or other related character evidence in order to demonstrate the potential for criminal behaviour based on the person’s prior behaviour (Redmayne, 2015).

Perceptions of exoneree

Exonerees’ reintegration challenges are well-documented in the international literature (e.g. Kauzlarich et al., 2001; Wildeman et al., 2011). An important finding from this research has been exonerees’ consistent self-reported experiences of stigmatisation by the community upon release, including from former friends and family (Grounds, 2004; Westervelt & Cook, 2008). Broader community reintegration issues and problems around securing employment and housing have also been recorded (Clow, 2017; Goldberg et al., 2020; Kukucka et al., 2020; Shlosberg & West, 2011). Researchers have investigated the public’s attitudes toward and responses to exonerees and have found that exonerees are stigmatised similarly to convicted offenders (e.g. Clow & Leach, 2015a, 2015b; Thompson et al., 2012). This has been conceptualised as an extension of stigma-by-association processes (Goffman, 1963; Pryor et al., 2012), such that exonerees are negatively stigmatised as a result of their interactions with convicted offenders and the prison system (Tudor-Owen et al., 2019), in spite of their wrongful conviction outcome.

While broader research has demonstrated a greater desire for social distance from innocent individuals who were associated with crimes (e.g. Shechory & Idisis, 2006), the wrongful conviction literature has yielded inconsistent results to date. Thompson et al. (2012) and Tudor-Owen et al. (2019) found no differences between exonerees and controls (e.g. ‘non-offenders’), but Clow and Leach (2015a) found a greater desire for social distance from wrongly convicted individual than non-offender controls, but less desire for social distance from convicted offenders. The lack of consensus may be explained by the sampling methods used, with all but the Tudor-Owen et al. (2019) study using solely student samples which (arguably) lack generalisability to broader public sentiment. In addition, Tudor-Owen et al. (2019) suggested that their findings differed from other international research due to potential differences surrounding perceptions of wrongful convictions in the Australian context. Of note, they point to a potentially raised public awareness and understanding of wrongful convictions due to extended media coverage of high-profile wrongful convictions, such as the cases of John Button and Andew Mallard, and argue that this may contribute to raising the awareness of, and reducing the stigma associated with, wrongful convictions (Bell et al., 2008; Stratton, 2013). Understanding these nuances in method and the broader social-cultural context, both of the exoneree and the community member (i.e. participant), are vital to further our understanding of, and ability to respond to, public perceptions (Boppre & Boyer, 2022; Faison & Smalarz, 2020; Webster & Miller, 2015) in order to improve experiences and outcomes for exonerees.

The current study

The current study sought to extend previous research on community attitudes toward wrongful conviction within the Australian context. It extends the extant literature on wrongful conviction by examining perceptions of exonerees through the cultural constructs of race, gender and criminal history and their intersections.

In lieu of evidence pertaining to wrongful conviction, but based on gendered patterns and stereotypes evident in the Australian criminal justice system (Jeffries & Bond, 2010; Strub & McKimmie, 2016), it was hypothesised that there would be differences in desired social distance depending on whether the exoneree was male or female. Second, while the race and wrongful conviction research remains equivocal (Howard, 2019; Scherr et al., 2018), based on the over-representation of First Nations people in prison and wrongful conviction populations (Dioso-Villa, 2015; Roach, 2015), it was hypothesised that there would be differences in desired social distance depending on the race of the exoneree. Thirdly, based on Clow and Leach (2015a), as well as recent recommendations from Kukucka et al. (2020), it was hypothesised that there would be differences in desired social distance according to whether exonerees had prior criminal histories or not. Fourth, based on the importance of intersectional arguments (Boppre & Boyer, 2022; Webster & Miller, 2015), it was hypothesised that there would be an interaction between exoneree gender, race and priors and ratings of desired social distance. Finally, an additional exploratory hypothesis sought to consider the potentially confounding effects of participant age, sex and Aboriginal or Torres Strait Islander status on perceptions of desired social distance from exonerees.

Method

Design

This study employed a national community survey, with participant recruitment targeted to be at proportional rates to the populations of the eight Australian States and Territories. The design incorporated a 2 (race: indigenous or non-indigenous) x2 (gender: male or female) x2 (criminal history: prior conviction or no prior conviction) embedded 3-way factorial design. The study received ethical approval from the host University (ref: 22933); all participants indicated their consent prior to commencing the survey.

Participants

The eligible sample comprised adult males and females who were residents of Australia and proficient in the English language. A survey company, Qualtrics, was used to recruit the sample through a series of online survey panel providers that interested individuals can sign up to; those who complete surveys receive incentives such as redeemable points or vouchers. Based on an effect size of 0.2, with alpha set at .05 and power of .95 the minimum sample required was determined to be 560. A total of 1043 participants completed the survey, with a minimum of 120 participants being randomised to each of the eight vignette conditions.

Materials

Mock news article

The eight vignettes were designed to imitate a newspaper article; the content remained the same across all eight vignettes, apart from the one variable that was manipulated in each vignette: race (Indigenous or no mention of race), gender (male or female), and prior criminal history (criminal history or no criminal history) (See Appendix A for example vignette). The gender of the exoneree was manipulated by maintaining a gender-neutral name (Jamie Brooks) across all vignettes, but changing their pronouns (his/her, she/he, man/woman). For vignettes depicting an Indigenous exoneree, race was noted in the heading and main text of the mock news article, as well as reference being made to their use of Aboriginal services. To manipulate priors, the exoneree was depicted as either having a prior criminal history (unspecified nature and extent) or not. The crime of murder was used to describe the type of offence for which the exoneree was wrongly convicted for across all vignettes, as violent crimes comprise more than half of known wrongful convictions in Australia (Dioso-Villa, 2015).

Desired Social Distance Scale (DSDS)

This scale was originally adapted from the Bogardus Social Distance Scale (1993) used in previous exoneree stigma research (Thompson et al., 2012; Tudor-Owen et al., 2019). The DSDS operationalises desired social distance in relation to three domains: reintegration support, social distance, and future criminality. An example statement included, ‘I would be willing to spend time alone with Jamie Brooks.’ Responses were measured on a Likert-type scale from (1) strongly disagree to (7) strongly agree, with higher scores indicating more favourable judgements of exonerees. Cronbach’s alpha was .68, indicating acceptable reliability.

Procedurally, participants provided brief demographic details, then were directed to read a mock vignette after which they were asked to complete the Desired Social Distance Scale based on the details provided about the exoneree in the vignette. This formed part of a broader survey on community attitudes towards exonerees, the results of which are considered elsewhere.

Approach to analysis

The dependent variable of interest was total desired social distance, with lower total scores indicating a greater desire for social distance. Initial univariate analyses explored factors associated with total desired social distance, using independent t-tests and Spearman’s correlations, computing Cohen’s d where appropriate and stratifying by gender, age group, and Aboriginal and Torres Strait Islander background to aid interpretation and consider differences between participants. The vignette conditions of gender, race and prior criminal history were recoded into dummy variables. Using GLM, a three-way (2 × 2 × 2) between subjects ANOVA was computed to consider main effects as well as possible two- and three-way interactions. Estimated marginal means were reviewed, and Levene’s test of homogeneity of variance was computed to assess the overall model.

Results

Demographics

The participant sample included 1043 participants; 671 (64.3%) were female and 363 (34.8%) were male, ranging in age between 18 and 86 years old (M = 45.0 years. SD = 18.0). The majority were born in Australia (738, 70.8%), and 45 (4.3%) identified as being of Aboriginal or Torres Strait Islander origin. Over half (567, 54.3%) were in part or full-time work, 188 (18%) were retired, and 186 (17.8%) were not currently employed or receiving benefits. Over a third had completed a degree program (378, 36.3%).

Participant correlates of desired social distance

The average desired social distance score was 117.49 (SD = 26.52) and varied from 35 up to 161. Broadly speaking, female participants scored significantly higher on the desired social distance scale than males (M = 119.18, SD = 26.87 v. M = 114.39, SD = 25.54, t = 2.790, p = 0.005, d = 0.18), meaning they reported a lower desire for social distance from exonerees regardless of vignette condition. There was a significant positive correlation between desired social distance and age (r = 0.120, p < 0.001), such that older people recorded a decreased desire for social distance from exonerees regardless of vignette condition. Desired social distance did not differ between those participants born in Australia as compared to overseas (t = 1.226, p = 0.221), or by education level as assessed by being university educated or not (t = 0.498, p = 0.619). Reported levels of desired social distance from exonerees, regardless of vignette condition, were lower for participants of Indigenous or Torres Strait Islander origin, but not significantly so (M = 115.93, SD = 27.79 v. M = 117.57, SD = 26.47, t = 0.404, p = 0.687).

Vignette condition and desired social distance

Gender of exoneree

An initial review of vignette condition and desired social distance revealed that there were no significant differences in ratings of desired social distance based on the manipulation of gender (male or female exoneree) in the vignette (t = 0.269, p = 0.79). A further exploratory analysis, stratifying by participant gender, then Aboriginal and Torres Strait Islander status, and then age, revealed that there were no differences when considering male and female participants separately (p = 0.93 and 0.92 respectively), Aboriginal or Torres Strait Islander and non-Aboriginal and Torres Strait Islander status participant views separately (p = 0.75 and p = 0.73 respectively), or older and younger (under 40 years old) participants separately (p = 0.44 and p = 0.50 respectively).

Race of exoneree

Participants randomised to the Indigenous race vignette scored higher on the desired social distance measure than those randomised to the white race vignette (M = 119.07, SD = 26.18 v. M = 115.91, SD = 26.78, t = 1.925, p = 0.05, d = 0.12), meaning that participants reported less desire for social distance when the exoneree depicted in the vignette was Indigenous. Further exploration revealed that there was some suggestion that this was more the case for younger (p = 0.07) and male (p = 0.09) participants, as compared to older and female participants, but these differences did not reach statistical significance. Participants who were not of Aboriginal and Torres Strait background reported a stronger desire for social distance, as compared to those from Aboriginal and Torres Strait Islander backgrounds, but not statistically so (p = 0.06 and p = 0.73).

Prior criminal history of exoneree

Participants randomised to the vignette where the person did not have a prior offence history scored higher on the desired social distance scale than those who were randomised to the vignette where the exoneree described had a prior offence history noted (M = 120.53, SD = 26.48 v. M = 114.52, SD = 26.24, t = 3.683, p < 0.01, d = 0.23), meaning that participants reported the need for social distance more strongly when the exoneree in the vignette had a prior offence history. Exploring further, this association between the exoneree having priors and the need for social distance was statistically significant for female participants (M = 115.05, SD = 26.25 v. M = 122.92, SD = 26.90, t = 3.829, p < 0.01, d = 0.29), but not for male participants (p = 0.51). Further, older participants reported a significantly higher desired social distance from exonerees with priors (M = 116.44, SD = 25.91 v. M = 123.79, SD = 25.84, t = 3.412, p = 0.01, d = 0.28), but these differences were less pronounced for younger participants (p = 0.07). Participants from Aboriginal and Torres Strait Islander backgrounds did not differ with respect to ratings of desired social distance according to whether the exoneree had priors or not (p = 0.54), but participants who were not of Aboriginal and Torres Strait Islander background indicated a significantly increased desire for social distance when the exoneree had priors (M = 114.57, SD = 26.11 v. M = 120.63, SD = 26.51, t = 3.640, p < 0.01, d = 0.23).

Intersectionality of gender, race and priors

A three-way between subjects ANOVA supported the finding of a main effect of having priors and of being Indigenous on total desired social distance. Levene’s test was non-significant (p = 0.80). There were no statistically significant interactions between gender, race and having priors (Tables 1 and 2).

Table 1. Generalised Linear Model examining effects of gender, race and priors on desired social distance.

Variable	Type III Sum of Squares	df	Mean Square	F	Sig	Partial Eta Squared
Intercept	14394251.89	1	14394251.89	20709.31	<0.001	0.952
Male gender	58.73	1	58.73	0.084	0.771	0.000
Indigenous race	2555.13	1	2555.13	3.676	0.055	0.004
Has priors	9303.99	1	9303.99	13.386	<0.001	0.013
Male*Indigenous	38.10	1	38.10	0.055	0.815	0.000
Male*priors	976.59	1	976.59	1.405	0.236	0.001
Male*Indigenous*priors	169.55	1	169.55	0.244	0.621	0.000

Dependent variable: Total Desired Social Distance.

Table 2. Estimated Marginal Means for gender, race and prior criminal history relating to desired social distance.

Gender vignette	Race vignette	Priors vignette	Mean	Std Error	95% CI
Female	White	No priors	117.55	2.37	112.90–122.19
		Priors	114.42	2.30	109.91–118.92
	Indigenous	No priors	121.97	2.32	117.41–126.52
		Priors	117.02	2.30	112.52–121.53
Male	White	No priors	119.49	2.30	114.99–123.99
		Priors	112.29	2.30	107.79–116.79
	Indigenous	No priors	122.95	2.30	118.44–127.45
		Priors	114.33	2.31	109.79–118.87

Dependent variable: Total Desired Social Distance. Note marginal means are different to total desired social distance mean scores reported in text.

Discussion

This study sought to explore community attitudes towards wrongful convictions and explore the effect of gender, race and having a criminal history on people’s ratings for desired social distance. A vignette was used, keeping all content identical but changing one independent variable at a time, with participants randomised to one of eight vignette conditions. A point of difference from the extant literature was the use of a broad community sample and the exploration of possible intersections between exoneree gender, race and criminal history.

Reflecting similar findings to Tudor-Owen et al. (2019), our study suggests that, overall, exonerees tend to be perceived relatively positively by the general public and that they are not stigmatised in the same way as convicted offenders. This finding stands in contradiction to similar prior international research (Clow & Leach, 2015a, 2015b; Thompson et al., 2012), potentially reflecting cultural and social differences between Australian and North American experiences. These differences may also be indicative of the cultural moments and different timeframes when these studies were conducted, that may reflect changes in media attention and shifting public awareness of miscarriages of justice over time (Bell et al., 2008; Tudor-Owen et al., 2019). In the absence of studies examining perceptions of wrongful conviction in Australia, some have suggested that there is a community trust in the protections afforded within the criminal justice system in Australia that allows for greater acceptance on the rare occasion wrongful convictions are formally acknowledged (Finlay, 2003, p. 14). While public confidence and trust in criminal justice institutions have varied and declined (Indermaur & Roberts, 2009; Neave, 2021), confidence in the Australian courts seems to have remained relatively constant, despite indications of it declining elsewhere (Australian Law Reform Commission, 2020). Subsequently, Australians’ level of confidence in their justice system might, in part, ameliorate the effects of stigma by association processes witnessed elsewhere.

Given the gendered experience of wrongful convictions in the US (Free & Ruesink, 2016) and research on gender-crime congruency effect between socio-cultural stereotypes and assessments of criminal defendants (Strub & McKimmie, 2016), we expected to find differences according to whether the exoneree was male or female. However, no differences were found. The null finding here was not expected and is incongruent with earlier work by Konvisser (2012) who reported that female exonerees experience stigma and bias upon release from prison. The type of crime described in the mock news article may partially explain this finding. Although common among men’s wrongful convictions, murder is not a typical crime for women (Ruesink & Free, 2005). A stereotypical crime, or a ‘no-crime’ conviction, may have differentially impacted participant’s assessments of female exonerees. The Australian context and potential role of the media may also contribute to these findings, with high-profile cases, like that of Lindy Chamberlain, resonating in the public consciousness. These narratives are often the only point of access surrounding wrongful conviction for community members, which might imply that the role of the media is more important in shaping individual exoneree’s experiences of stigma.

In contrast to some international research that has found little effect of race on perceptions of exonerees (Karaffa et al., 2017; Thompson et al., 2012), this study observed a negative attitudinal effect of exoneree race biased towards white exonerees. Where other extant research has suggested a perceived risk of offending and aggressiveness towards exonerees from racial minorities (Howard, 2019), our findings offer a contrary finding suggesting greater acceptance by general community members of Indigenous exonerees than white exonerees. Conceptually, this may be supported by the results on criminal history, race, and attenuating biases in other exoneree perception work (Scherr et al., 2018), and may reflect earlier international research demonstrating white mock jurors attenuate their biases to appear unprejudiced when race is a salient factor in criminal trials (Sommers, 2006; Sommers & Ellsworth, 2000, 2001). Australian research published around the same time also suggested that Aboriginal offenders received more sympathy from mock jurors and subsequently more lenient judgements (Feather & Souter, 2002). Additionally, some research across the US, Canada, and Australia has indicated a similar ‘positive discrimination bias’ toward Indigenous defendants in sentencing outcomes (e.g. Jeffries & Bond, 2012).

While differences in attitudinal outcomes of interest between studies may account for some of these inconsistencies in study findings, a specific race manipulation check was not included in this study. Although the mock news article in the Indigenous vignettes included multiple mentions of the exoneree’s race, as well as the use of the Victorian Aboriginal Legal Services to account for this, ‘white’ was not actively manipulated in this study. Rather, all mentions of race were removed from the mock news article to act as a control for Indigenous race manipulation. Consequently, it cannot be guaranteed with certainty that participants noticed the race of the exoneree described in the article. Further, while the survey was anonymous, and therefore the potential for social desirability bias may be reduced, other research has suggested that the anonymity provided by online surveys does not always reduce socially desirable responding (e.g. Gnambs & Kaspar, 2017). While some other authors have used alternative methods, including visual imagery (e.g. a picture) to indicate race or included a specific race manipulation check, these methods fail to take intersectionality into account and are not without their own challenges and limitations (e.g. Ma-Kellams, 2021).

There was support for the hypothesis that exonerees with criminal histories would be perceived differently from those with no reported criminal histories; in our study, participants reported a greater desire for social distance from exonerees with priors, indicating less favourability to exonerees with criminal histories. This is broadly consistent with Karaffa et al.’s (2017) findings with regard to perceptions of exoneree’s deservingness of financial compensation, and Evans et al. (2019) regarding landlords’ willingness to rent according to the persons disclosure of a criminal history. Our findings add to this, but indicate that these perceptions differ for different community members, in that the influence of prior criminal history deleteriously impacts the views of females and older community members, but less so for males and younger people. While there is no extant literature regarding exoneree populations to help frame or otherwise understand these exploratory findings, they are certainly consistent with the wealth of international evidence around broader issues relating to community members perceptions of issues around community safety and fear of crime (e.g. Dodd, 2018; Gault & Sabini, 2000). It is possible that general community members perceptions of crime and community safety extend to exoneree populations, either by virtue of stigma by association, and/or other factors not overtly measured, manipulated, or accounted for here, and that these factors that influence perceptions, for at least some community members, regardless of factual guilt or innocence. Some studies have noted differences in perceptions and/or treatment according to crime type (Chataway & Hart, 2019); therefore future research should explore desired social distance (and related outcomes), to extend and test differences for a broader range of crimes, as well as considering additional differences by race and gender (Evans et al., 2019; Zannella et al., 2020).

Limitations and future directions

The study used a mock news article vignette, despite some previous research highlighting the potential for limited ecological validity of the vignette-style method (Thompson et al., 2012). Within our study, participants’ only exposure to questions about exonerees was the vignette they were randomised to. Prior research has demonstrated that differences in attitudes toward wrongful convictions can be dependent upon how information is delivered (e.g. reporting numbers/statistics, written/spoken narrative, use of imagery, etc), the salience of the manipulations being explored using different methods (Zannella et al., 2020), and the psychometric properties of the outcomes of interest (Norris & Mullinix, 2020). Subsequently, the mock news article itself, and/or the measure of desired social distance may not have been sensitive enough to elicit differences or sufficiently strong attitudinal responses, or perhaps sway other pre-existing views participants may have held about exonerees and wrongful convictions. This may explain, in part, why there was a lack of intersectionality found with our findings. Clow and Leach (2015a) have highlighted the need for wrongful conviction research to employ alternative methodological approaches beyond attitudinal response measures to gain more nuanced understandings of the extent and associated impacts of stigma faced by exonerees. Indeed, the social and systemic complexity of wrongful conviction may imply a limitation of the vignette method in understanding the issue, particularly in contexts where they occur and are reported less regularly.

A second limitation relates to the sample, with community members from Australian and Torres Strait Islander backgrounds and non-binary gender identifications being underrepresented in our sample. This is potentially problematic because the experiences of the justice system are known to differ between members of these communities and racial and gender-conforming majorities, and these factors may result in substantially different sentiment toward wrongful convictions (Cops & Pleysier, 2011). Future research should seek to explore these issues within these specific populations.

Additional research is also necessary in order to better understand the practical consequences that social characteristics hold for exonerees and the role of the media in mediating these effects, as well as the wider implications for stigma, bias and discrimination associated with people’s involvement within the criminal justice system. The social characteristics of ex-offenders and exonerees have been shown to hold important implications for understanding public perceptions of criminality and how they are constructed (Strub & McKimmie, 2016). It is possible that exonerations and wrongful conviction are potentially seen as less of a ‘crime’ or ‘legal’ issue and more of a ‘justice’ issue. This remains to be tested, but if this is the case, this may suggest the need for different ways and means of capturing and measuring public responses to these errors in the justice system.

Conclusion

The results of this study support previous research indicating that not all exonerees are perceived equally. Our findings suggest that biases, prejudice, stereotypes, and stigma directed towards exonerees may be different in the Australian wrongful conviction context. As such, while wrongful conviction may be a global issue, regional and other systemic differences may influence the public’s perception of error, as well as their perspectives towards justice and fairness more broadly. Future research is required to investigate the drivers and mechanisms for successful community reintegration for exonerees and how they can be accepted and supported by the broader community. This has significant implications not only for exonerees, but also for others who have been justice involved.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Appendix: Mock news article vignette (example provided woman, Indigenous, priors)

.