Youths going missing continues to be a matter of great concern for the police. This group has been consistently found to comprise most police missing person reports, be more likely to go repeatedly missing, and experience victimization and vulnerability related to these ...
Youths going missing continues to be a matter of great concern for the police. This group has been consistently found to comprise most police missing person reports, be more likely to go repeatedly missing, and experience victimization and vulnerability related to these incidents. This study thus sought to examine single versus repeat versus habitual/chronic missing youth cases to extract differentiated insights to initiate discussions on proactive policing efforts for reducing and preventing missing youth cases. Data employed tracked 2126 young people reported missing for the first time in 2005, for ten years using their police records. Over a third went missing more than once. There was evidence of the power few hypothesis, with the habitual/chronic cases recording over 60% of the missing person reports across the study period, and clear differences emerged between single, repeat and habitual/chronic cases in terms of their demographics, mental health concerns, and justice involvement.
This is a pre-copyedited, author-produced version of an article accepted for publication in Policing: A Journal of Policy and Practice, following peer review. The version of record, Thomas, SDM., & Ferguson, L. (2022). The power few missing persons: A ten-year follow-up study of over 2000 youth missing persons. Policing: A Journal of Policy and Practice, is available online at: https://doi.org/10.1093/police/paac084. When citing, please cite the version of record.
Please direct all correspondence to: Stuart D M Thomas, Criminology and Justice Studies, School of Global, Urban and Social Studies, RMIT University, 124 La Trobe Street, Melbourne, Victoria 3000, Australia. E-mail: [email protected]
Law enforcement agencies worldwide receive millions of reports of missing youths1 annually. The reasons why youths go missing are varied, ranging from parental abduction and foul play to, more routinely, running away from home or a care facility (Babuta and Sidebottom, 2020; Bowden and Lambie, 2015; Huey, Ferguson, and Kowalski, 2020). Due to such variation, Biehal and colleagues (2003) formative conception of the ‘missing continuum,’ from unintentional (wanderers, lost persons, persons unaware they are missing) to intentional (chose to leave, those who drift) missing person cases, assists in providing a classification system organizing such reasons. These incidents can be associated with a panoply of negative sequelae, including the risk of self-harm, predation, abuse by others, substance use issues, exploitation, and violence (Hayden, 2016; Hutchings et al., 2019; Stevenson and Thomas, 2018). To illustrate further, Haynie and colleagues (2009) found a 13% increase in the odds that a child will go missing with every experience of physical abuse by a family member. Most notably, mental health considerations and justice involvement as crime perpetrators (e.g., ‘survival crimes’ like stealing and begging) and victims (e.g., family violence, sexual abuse) are two well-recognized correlations to missing youth cases (Babuta and Sidebottom, 2020; Crosland and Dunlap, 2015; Hayden and Goodship, 2015; Shalev, 2011). A recent Australian estimate of the extent of youth missing person’s justice involvement ranges up to 65-68% for criminal offending and victimization, respectively (Stevenson and Thomas, 2018).
Internationally, youths have consistently been found to comprise the majority of missing person reports (Babuta and Sidebottom, 2020; Biehal et al., 2003; Dedel, 2006; Shalev Greene and Hayden, 2014), ranging between, for instance, 52-59% in Australia (Bricknell, 2017; James, Anderson, and Putt, 2008; Stevenson and Thomas, 2018), 60-65% in the United Kingdom (UK) (Babuta and Sidebottom, 2020; Bureau, 2014), and 55-62% in Canada (Canada’s Missing, 2017; 2018; 2019). Thus, reports involving youths tend to consume the lion’s share of police resources that are dedicated to missing persons.A sizeable proportion of youths who go missing do so multiple times (Huey et al., 2020; Sowerby and Thomas, 2017); these are suggested to be referred to as repeat or habitual/chronic cases (Ferguson and Picknell, 2021). With multiple missing events, there is a greater demand for police resources, due to repeated efforts to find and return the individuals. This is coupled with high urgency to resolve such cases, stemming from the risk of victimization and exposure to risky behaviours escalating with each missing person episode (Stevenson and Thomas, 2018; Huey et al., 2020) and the increased proclivity for the young person experiencing some form of harm. For example, increased justice involvement as a perpetrator and crime victim and mental health-related vulnerability are positively associated with the number of times a youth goes missing (Randone and Thomas, 2022; Tucker et al., 2011; Tyler et al., 2003). As a consequence, police will commonly place a high priority on responding to these reports due to the importance of protecting and safeguarding vulnerable youths (Babuta and Sidebottom, 2020; Hayden and Shalev Greene, 2018, Sidebottom et al., 2020).
The individual's ‘category’ relating to their history of going missing (single, repeat, habitual/chronic) is increasingly considered due to evidence indicating the risk of harm, victimization, and exposure to various adverse experiences increases with each missing event (Huey et al., 2020; Randone and Thomas, 2022). Recent evidence from the UK has raised doubts about the link between being rated as high-risk and the person experiencing harm, with Doyle and Barnes (2020) noting that “89% of the predicted high-risk cases ha[d] no actual harm” (p.160) They reported that harm was more likely to be reported when the person was over 18 years old, defining harm as “death, suicide, self-harm, mental harm, sexual assault and injury” (p.164). However, other evidence indicates that young people who go missing are at risk of harm (more broadly defined), as an antecedent to, and as a consequence of, going missing (e.g., Biehal et al., 2003; Hill et al., 2016).
Ferguson and Picknell (2021) note some important inconsistencies with the terminology that has been used, with ‘repeat’ cases referring to going missing more than once in some studies (Sowerby and Thomas, 2017), compared to others which have noted a specific threshold (e.g., missing more than three times in 90 days) (Harris and Shalev Greene, 2016). Consistent with the contemporary situation, an earlier study reported by Pfeifer (2006) remarked that there is no formal definition for how many times a youth must run away or go missing before being labelled habitual/chronic. For these reasons, scholars have emphasized that additional research is required that disaggregates and compares across single and repeat cases to extract discriminating insights on potentially habitual/chronic missing persons to assist in informing police risk assessment and targeted strategies towards exigent cases and people/groups in highest need of intervention (Ferguson and Picknell, 2021). Some informative recent UK-based research, for example, has highlighted distinct groups of particular concern; that being young people (defined in their research as under 18 years old) in out of home care and those of white ethnicity (Bezeczky and Wilkins, 2022; Sidebottom et al., 2020). Another UK-based study reported that young people who went missing repeatedly were more likely to have a history of abuse and neglect, substance use issues, and be at risk of exploitation (Hutchings et al., 2019).
Sherman’s (2007) concept of the “power few”2 provides a salient context for examining single versus repeat versus habitual/chronic missing person cases. Economists have long discussed the “vital few and trivial many” or the “Pareto curve,” a skewed distribution J-curve constituting the small fraction of population units that account for a sizeable portion of the total volume of any part of that population (Blanchet, Piketty, and Fournier, 2022; Eck et al., 2007; Juran, 1988; Simon, 1955). This form of distribution has been popularised as the “80-20 rule” (Kock, 1999): 20% of some thing or some group leads to 80% of the outcomes (Eck et al., 2007; cf. Sherman et al., 1989; Sherman, 2007). In practice, rarely is it as definitive as an 80-20 split; still, the principle has been held across multiple topics and disciplines to the degree that it has been argued that it is universal (Lipovetsky, 2009).
Scholars in the field of criminology have applied the “power few” concept to a range of matters, such as crime hotspots (Eck et al., 2007; Sherman et al., 1989), victimisation (Farrell et al., 2002), and domestic violence (Bland and Ariel, 2015; Robinson, 2017). One prior published study has applied the “power few” concept to both the location and characteristics of missing person cases. Huey et al. (2020) found that only ten addresses in one city—hospitals, group homes, and homeless shelters—generated over 70% of adult missing person reports and around 69% of youth reports. These were considered the “power few” locations that produced the greatest demand and strain on policing (Huey et al., 2020). In addition to the location-based findings, the authors also reported that youths (aged 16-17) were responsible for half of all repeat missing reports, and that when 14- and 15-year-olds were included, this broader group were responsible for almost 90 percent of repeat missing youth reports. These findings are broadly consistent with other related UK-based research that reported that 20 percent of repeatedly missing youth contributed over 60 percent of missing person cases (Babuta and Sidebottom, 2020; Galiano Lopez et al., 2021).
Against this background, the current study explores the presence and nature of the “power few” of youth missing person cases by identifying which of the types of cases (single, repeat, or habitual/chronic) generate the greatest number of police reports (e.g., is it a few specific youths going missing a considerable number of times, or is it many single episodes?). Understanding this would assist in extrapolating insights on case differences that can offer direction for developing targeted prevention, intervention, and/or reduction efforts towards the cases making up most police missing person reports. In doing so, this study seeks to add a more nuanced understanding, beyond the single versus repeat episode dichotomy that is commonly used to classify missing person cases.
The aims of the study were to explore the presence and extent of justice involvement of young people who were reported as missing to the police and to ascertain whether single episode missing person cases differed from repeat and habitual/chronic cases. First, based on Stevenson and Thomas (2018), it was hypothesized that there would be an association between the number of times the young person was reported missing and their justice involvement as a victim and as a suspect. Second, based on evidence of the relationship between mental health-related vulnerability and going missing (Crosland and Dunlap, 2015; Hutchings et al., 2019; Sowerby and Thomas, 2017), it was hypothesized that there would be an association between the presence of mental health-related vulnerabilities and the number of times the person was reported as missing to the police, and based on more limited and somewhat equivocal evidence, it was further hypothesized that gender would moderate this association. Lastly, following the recommendations by Ferguson and Picknell (2021) about disaggregating and classifying missing person cases, an exploratory hypothesis sought to consider differences between single episode, repeat, and habitual/chronic cases of youth missing persons.
The eligible sample comprised all youths who were reported as missing to Victoria Police for the first time during the calendar year of 2005. A missing person, in the Australian context, is defined as “anyone who is reported missing to police, whose whereabouts are unknown, and there are fears for the safety or concern for the welfare of that person” (NMPCC, n.p., cited in Bricknell, 2017, p.5). Consistent with Australian Bureau of Statistics (ABS, 2011b), and World Health Organisation (WHO, 2022), definitions, a youth was defined as a person under the age of 25 years. For inclusion in this study, a person, therefore, needed to be under 25 years old when they were first reported missing to police in 2005, no other exclusion criteria were applied. The sample consisted of 2126 individuals who generated a total 6150 missing person reports. A longitudinal data set was created, extracting all formal and informal police contacts these 2126 individuals had for the next ten years (up to the end of 2015), as well as extracting all formal and informal police contacts prior to the first missing person report to police.
Data were extracted from the Victoria Police LEAP database which has been in operation since 1993; prior to this, a cardex system was in operation with more limited contact histories as suspects and victims recorded. These paper records are scanned, linked, and attached to the electronic LEAP file. The LEAP database includes all contact individuals have as suspects, offenders, victims, witnesses, and people in need of assistance. Police contact data from other Australian States and Territories were not included in the dataset due to the much more limited data collated on interstate contacts on the LEAP database. The State of Victoria, on the East Coast of Australia, accommodates over a quarter of the Australian population and is the second most populous state in the country.
Contact-based data, in relation to victimisation episodes and criminal charges, were coded using the Australian and New Zealand Standard Offence Classification (ANZSOC; ABS, 2011a). These were then truncated into violent, sexual, and non-violent non-sexual offence categories. Mental health-related vulnerability was coded where warning flags3 entered on LEAP indicated prior knowledge of, or concerns about, mental illness, substance use, spectrum-based disorders, self-harm, and suicidality. Mental health transfers relate to police powers under the Mental Health Act 1986 (Vic) to apprehend and transport a person for the purposes of a mental health assessment where there are grounds that the person may be at serious and imminent risk of harm to themselves or another person. Consistent with prior research (e.g., Hutchings et al., 2019; Shalev Greene and Hayden, 2014; Stevenson and Thomas, 2018), a repeat missing person case was defined as a person who was reported missing to the police up to three times across the eligible timeframe. Habitual/chronic missing person cases, separated from the repeat missing person cases, were defined in line with the existing literature as four or more missing person episodes (Ferguson and Picknell, 2021; Huey et al., 2020).
The research was carried out following contemporaneous Australian guidance on epidemiological research involving databanks (NHMRC, 2007; 2018). The police contact-based data were de-identified at source to protect the privacy of the sample. The study was approved by the Human Research Ethics Committee at the host institution, and by the Police service ethics committee.
First, data were coded, grouped, and plotted to explore distributions of continuous variables and the viability and utility of different categorical groups. Categorical data were cross-tabulated and compared using the non-parametric Chi Squared Test of Association; t-tests were used to compare continuous data between groups, using Levene’s test to indicate where unequal variances needed to be taken into account; this was substituted with ANOVA and the Games Howell post-hoc test where more than two categories were compared. Spearman’s non-parametric correlation was selected to compare continuous variables. All tests were two-tailed; effect sizes were calculated to aid interpretation where appropriate. Multinominal logistic regression was used to compare the characteristics of single case, repeat case, and habitual/chronic missing person cases. The baseline category was selected as single episode missing person cases; repeat missing person cases excluded those who met the classification for being habitual/chronic missing persons. Finally, a logistic regression model was computed comparing habitual/chronic cases with repeat cases. For the purposes of the multivariate analyses, the number of charges, number of victimisation episodes and age at first missing person report were treated as continuous variables, so adjusted odds ratios (AORs) for these variables should be interpreted as a per unit increase.
The sample had a total of 6150 missing person cases over the eligible timeframe, of which were generated by 2126 individuals. Thus, some individuals went repeatedly missing. One thousand three hundred and forty-six (1346, 63.3%) involved cases of individuals going missing just once over the study period. Four hundred and thirty (430, 20.2%) individuals met the criteria for being repeat missing persons, and a further 350 (16.5%) were considered habitual/chronic missing persons (see Table 1 for an overview of the sample). Habitual/chronic cases contributed over 60% (n=3808, 61.9%) of the total number of reports; that is, around 17% of individuals generated about 62% of missing person reports. Of note, individuals in the sample were reported missing up to 115 times across the 10-year follow-up period; the average number of times individuals went missing was 2.89 times.
Nine hundred and eighty-six (986, 46.4%) males contributed 2794 (45.4%) missing person cases and 1140 (53.6%) females contributed 3356 (54.6%) missing person cases. There were no differences in terms of their age when first reported missing (t=0.437, p=0.662). Higher proportions of females were found in the single cases (726/1346, 53.9%), repeat cases (234/430, 54.4%) and habitual/chronic cases (180/350, 51.4%), but no statistically significant association was found between sex and missing person case type (Χ2 =0.841, p=0.657, V=0.020). During the follow-up period, 45 (2.1%) of the sample (34 males and 11 females) became deceased; their average age at death was 22.35 (SD=4.90), the youngest was 14 and the oldest was 33 years old at the time of death.
Table 1: Characteristics of full sample of missing youth by missing type
Age first missing
Any criminal charges
Any warning flags
Deceased during follow-up
Criminal Charge History
A third (735, 34.6%) of the full sample had criminal charge histories recorded prior to their first missing person episode being recorded on the police database; two-thirds (1436, 67.5%) had criminal charges recorded on police files by the end of the follow-up period. Eight hundred and sixty-nine (869, 40.9%) had been charged with violent crimes, 144 (6.8%) were charged with sexual offences, and 1354 (63.7%) had been charged with non-sexual non-violent offences.
Males were significantly younger than females at first criminal charge (M=14.92, SD=3.13 v. M=15.66, SD=3.26, t=4.402 p<0.001) and recorded significantly more criminal charges than the females (M=25.08, SD=41.12 v. M=7.88 SD=18.30, t=12.13 p<0.001) over the duration of the study period. There was a significant positive correlation with how many times a person was recorded as being missing and the number of criminal charges recorded on their police files (r=0.424, p<0.001).
There was a significant association between type of missing person case type and having a history of criminal charges (Χ2 =197.95, p<0.001, V=0.305), the effect size was larger for females than males (V=0.325 v. V=0.287). Single episode cases had the lowest rate of criminal charges (768/1346, 57.1%), followed by repeat cases (344/430, 80.0%), and habitual/chronic cases had the highest rate (324/350, 92.6%). Further analyses revealed significant differences in the number of violent charges, with the post-hoc Games Howell test indicating that habitual/chronic cases had significantly more violent charges than repeat cases (p<0.001) and single cases (p<0.001). The same pattern was found with charges for sexual offences (p=0.003 and p<0.001 respectively) and for non-sexual non-violent offences (p<0.001 and p<0.001 respectively) (see Table 2).
Table 2: Counts of criminal charge type by missing person case type
Habitual/chronic case (n=350)
M=1.39, SD=4.04, Mdn=0
M=3.02, SD=5.71, Mdn=1.00
M=6.69, SD=9.19, Mdn=3.00
M=0.16, SD=1.22, Mdn=0
M=0.44, SD=1.18, Mdn=0
M=1.06, SD=1.65, Mdn=1.00
Non-sexual non-violent charge
M=7.00, SD=20.38, Mdn=1.00
M=14.33, SD=26.80, Mdn=4.00
M=33.15, SD=38.86, Mdn=16.50
Six hundred and fifty-eight (658, 31.0%) of the sample were recorded as having been a crime victim prior to their first missing person episode; just under two thirds (1372, 64.5%) had victimisation episodes recorded on police files by the end of the follow-up period. Nine hundred and twenty-seven (43.6%) were recorded as having been the victim of a violent offence; 449 (21.1%) had been the victim of a sexual offence; and 865 (40.7%) had been the victim of non-violent non-sexual offences.
While there were no significant differences between males and females regarding their age at first victimisation (males M=15.58, SD=5.30 v. females M=15.77 SD=5.10, t=0.664 p=0.507), females had significantly more victimisation episodes recorded than males (M=2.85, SD=3.98 v. M=1.80 SD=2.73, t=7.154 p<0.001) over the duration of the study period.
Similar to the criminal charge data, there was a significant association between the type of missing person case and having a victimisation history (Χ2 =120.78, p<0.001, V =0.238). The effect size was larger for females than males (V=0.248 v. V=0.232). Single episode cases had the lowest rate of criminal charges (758/1346, 56.3%), followed by repeat cases (315/430, 73.3%), and habitual/chronic cases had the highest rate (399/350, 85.4%). Further analyses revealed significant differences in the number of violent victimisations, with the post-hoc Games Howell test indicating that habitual/chronic cases had significantly more violent charges than repeat cases (p<0.001) and single cases (p<0.001). The same pattern was found with charges for sexual offences (p<0.001 and p<0.001 respectively); while the habitual/chronic cases had significantly more non-sexual non-violent victimisations recorded than the single cases (p<0.001) there was no significant difference in the number of non-sexual non-violent victimisations between the habitual/chronic and repeat cases (p=0.108). See Table 3.
Table 3: Counts of victimisation type by missing person case type
Habitual/chronic case (n=350)
M=0.74, SD=1.50, Mdn=0
M=1.18, SD=1.68, Mdn=1
M=1.89, SD=2.62, Mdn=1.00
M=0.28, SD=1.01, Mdn=0
M=0.45, SD=1.09, Mdn=0
Non-sexual non-violent victimisation
M=0.69, SD=1.28, Mdn=0
M=0.95, SD=1.47, Mdn=0
Mental Health-Related Factors
One in four of the full sample (545, 25.6%) had mental health-related warning flags recorded on the police database across the study period. There was a significant association between having a mental health warning flag and sex, with a higher proportion of males than females having one or more mental health flags on their police files (277/986 v. 268/1140, Χ2 =5.83, p=0.016, ɸ=0.052).
There was a significant association between having a mental health-related warning flag and the type of missing person case (Χ2 =129.37, p<0.001, V=0.247). The lowest proportion was found among single cases (240/1346, 17.8%); this rose to a third of the repeat cases (147/430, 34.2%) and just under half of the habitual/chronic cases (158/350, 45.2%). The effect size was larger for males than females (V=0.272 v. V=0.245).
One hundred and eighty-nine (8.9%) of the sample were transported by police for a mental health-related assessment; almost half (90, 47.6%) of this group were subject to a mental health transfer more than once across the study period. There was a positive correlation between the number of times the person was reported missing and the number of times they were subject to a mental health transfer by the police (r=0.243, p<0.001). Being reported as missing to the police preceded records of police having completed a mental health transfer in all but two instances. Mental health transfers were more common for males than females (101/986 v. 88/1140, Χ2 =4.16, p=0.041, ɸ=0.044). There was a significant association between missing person case type and being subject to police transfer for mental health-related assessment (Χ2 =112.67, p<0.001, V=0.230); the effect sizes were similar for males (V=0.246) and females (V=0.245). Single cases had the lowest rate of mental health transfer (56/1346, 4.2%), followed by repeat cases (60/430, 14.0%), and habitual/chronic cases had the highest rate of transfer (73/350, 20.9%).
The “Power Few”: Comparing Single, Repeat and Habitual/Chronic Missing Persons
The multinominal regression revealed similar patterns for both the repeat and habitual/chronic missing persons when compared with single episode missing persons, with increased numbers of criminal charges, increased numbers of victimisation episodes, younger age first reported missing, and the presence of mental health-related warning flags being significantly more common for the repeat and habitual/chronic youths. The one difference between the repeat and habitual/chronic cases related to sex; sex did not differentiate between repeat and single episode missing persons when other factors were taken into account. Refer to Table 4 for further details.
As a final step, an exploratory analysis comparing repeat and habitual/chronic cases revealed that habitual/chronic cases were more likely to be younger at the first missing episode and also to accumulate more criminal charges and victimisations as compared to repeat cases over the ten-year follow-up period (see Table 5).
Table 4: Multinominal logistic regression comparing repeat and habitual/chronic missing persons with single episode missing persons
Repeat missing persons
Age first reported missing
Sex - female
Any mental health-related concerns
Number of criminal charges
Number of victimisations
0.940 – 0.996
0.675 – 1.090
1.633 – 2.760
1.008 – 1.018
1.025 – 1.106
Habitual/chronic missing persons
Age first reported missing
Sex - female
Any mental health-related concerns
Number of criminal charges
Number of victimisations
0.852 – 0.962
0.535 – 0.962
2.123 – 3.866
1.020 – 1.030
1.092 – 1.179
R2 0.177 (Cox and Snell), Nagerkerke 0.211, Model Χ2 = 414.06, p<0.001.
Table 5: Logistic regression comparing repeat and habitual/chronic missing persons
Habitual/chronic missing persons
Age first reported missing
Sex – female
Any mental health-related concerns
Number of criminal charges
Number of victimisations
0.864 – 0.951
0.911 – 1.801
0.971 – 1.886
1.011 – 1.021
1.033 – 1.130
R2 0.130 (Cox and Snell), Nagerkerke 0.174, Hosmer and Lemeshow test Χ2 = 22.83, p=0.004.
This study sought to explore differences in the types and extent of justice involvement in an Australian cohort of young people who had been reported missing to the police for the first time in 2005. The focus of the study was contrasting single episode missing persons from repeat missing person and habitual/chronic missing persons to examine the potential application of the concept of the “power few.” The full life history of police contacts provided ample time to explore people’s missing person and criminal justice trajectories through youthhood and into adulthood. This afforded the researchers ample time for individuals to ‘reach criteria’ for being a habitual/chronic case and explore common factors associated with this. Some interesting demographic differences emerged among the sample in relation to their justice involvement and related risk and vulnerabilities, with support found for our first hypothesis regarding levels of justice involvement and the number of times the young people in the sample were reported missing. Consistent with Vo (2015), as cited in Bricknell (2017), the males were much more criminally involved, attracting substantially more criminal charges than the females in the sample. By contrast, the females had significantly more victimisation episodes recorded on police files across the ten-year period, but recorded rates were much lower than the level of criminal involvement overall.
Regarding our second hypothesis, mental health-related concerns were more commonly reported for male youths in this sample, as compared to females, both with respect to having a range of indicative warning flags noted on the police records system and through police transfers, under powers set out in the Mental Health Act (Vic) 1986, for specialist health assessment. The overall rate reported here, one in four of the sample, is somewhat on the lower side of other reported estimates, which have ranged from as low as 4% to as high as 80% according to police data across Australian jurisdictions (Bricknell, 2017). A more accurate estimate would require access to linked health records, or mental health screens (or similar) with young people upon their return. These types of methodology are uncommon in this literature; however, by way of example using linked data, Sowerby and Thomas (2017) reported that over half of their sample of youth and adult missing persons had established histories of contact with public mental health services. The finding that forty-five (2.1%) of the sample were reported as deceased across the ten-year follow-up is a cause for concern. Prior research suggests that only a small percentage of people die while they are missing (Fyfe et al., 2015). This finding may just be inherently linked to the young people’s risky lifestyles and having to survive in a range of high-risk environments (Svensson and Pauwels, 2010) both prior to, or as a consequence of, going missing. We did not have access to further detail to consider the timing, causes and associated circumstances around their deaths, although prior research suggests a strong association with mental health-related concerns (Perkins, Roberts, and Feeney, 2013). This finding needs to be explored in more detail, linking these police data with health and Coronial databases, as well as value-adding through other collateral data sources.
This study revealed a number of common risk factors for being a repeat missing person or a habitual/chronic case; these included being at an increased likelihood of having criminal charges and being a crime victim, being younger when first reported missing to the police, and having mental health-related warning flags. Addressing our third hypothesis, when exploring differences between a repeat case and habitual/chronic case, many similarities were found, with the habitual/chronic cases having an increased likelihood of more criminal charges and victimisation experiences, and a higher likelihood of experiencing mental health-related concerns. Considered as a whole, a small but significant number of persons who go habitually/chronically missing are likely to experience more significant involvement with health and justice agencies and therefore, arguably, be more likely to experience some form of harm due to repeat contact with the justice system (Esposito et al., 2017).
The results are congruent with the “power few” argument (Huey et al., 2020; Sherman, 2007), with approximately one in six of this sample (350/2126) being categorised as habitual/chronic missing persons, contributing 62% of the missing person reports. This finding resonates closely with recent UK-based publications (e.g., Babuta and Sidebottom, 2020; Galiano Lopez et al., 2021), despite the extended age range used in this study. Collectively, this evidence adds further weight to our understanding of how police services can potentially target scarce or otherwise limited resources to the greatest potential effect by focusing on the group of young people who go on to be habitual/chronic cases. Future research should seek to explore and detail the temporal relationship between crime involvement and being reported missing and consider the proximal nature of crime perpetration and victimisation to being reported missing (Randone and Thomas, 2022) as well as the person returning or being found. Our results revealed that a third of the sample criminal charges and the same proportion had victimisation episodes recorded prior to them first being reported missing to the police. The inter-relationships between health, justice, and social welfare risk factors remain difficult to disentangle but must remain important foci of future research efforts to better understand the “power few” of missing persons, and the actuating or contributing mechanisms to these repeated incidents. It will be these patterns that should underpin the development of police responses and case handling, with particular focus on advancing prevention and intervention strategies. To consider if there is anything different or unique about this missing person cohort, and perhaps the need for a different health/justice/welfare response, future research could also consider how and where the health and justice involvement of youths who go missing may differ by comparing these trajectories to a sample of youth who are not reported missing to police.
When interpreting these findings, there are a couple of limitations that need to be considered. Firstly, by virtue of relying on a police contacts-based database with data that were not collected for express purposes of research, we were somewhat limited with the scope of the data available. For example, some extant literature notes the relevance of considering the role of out-of-home care and differences according to ethnicity (Bezeczky and Wilkins, 2022). The police database did not routinely capture this information, so we were not able to consider these as partial or alternative explanations in our analyses. As such, police contact data cannot be fully relied upon to give a complete picture of the young person’s risks, vulnerabilities and justice involvement. Secondly, while we were able to create a proxy variable for mental health-related vulnerability, this was not verified through health data, so likely represents a lower prevalence estimate..Thirdly, record keeping of missing person cases is reliant on several processes (Boivin and Cordeau, 2011); not least of which includes the young person being reported as missing and then the police officer deciding to record the person as missing on the police database. There is known to be hesitancy about both reporting missing persons to police, and frustration and complacency from police officers, especially when the young person is a repeat missing person or otherwise known to police and considered to be a burden on police time and resources (Newiss, 1999). Concomitantly, a lack of trust in police, especially when young people have had prior negative experiences, will impact the likelihood that they will come forward (e.g., to report victimisation experiences) or disclose personal details to the police (Bezeczky and Wilkins, 2022; Thomas, 2014). Indeed, Birch and Sicard (2021, p.394) describe the relationship between young people and the police as, at best, turbulent. As such, there needs to be a continued emphasis on efforts to build trust and rapport between police and young people. Finally, the Australian definition of a youth ranges up to age 25, whereas other jurisdictions define a youth as up to 18 years old; some caution should therefore we applied when comparing results between studies.
Consistent with Ferguson and Picknell (2021), we conclude with a clarion call for consistent definitions to be used when reporting and researching missing person cases; this will serve to both further academic debates, and more accurately inform policing policy and practice reform. Our findings indicate the utility of differentiating between single episode, repeat and habitual/chronic cases, which can meaningfully contribute to important policy and practice-based discussions for how we theorize about and develop prevention and intervention strategies for at-risk youths, in particular focusing on the “power few” who go on to become habitual/chronic missing person cases. Quite apart from looking to redress some of the economic arguments made by police around the burden associated with responding to missing person reports (Hayden and Shalev-Greene, 2018), these findings also point to how police need support from, and to work alongside, other services, in a ‘joined-up’ way (van Dijk et al., 2019), with proactive, sustained partnerships that extend outside of their current professional silos (Crofts and Thomas, 2017). The call for this kind of cooperation and partnership between health, justice and social welfare agencies is not a novel or new recommendation (e.g., Newiss, 1999), and it is not by any means straightforward (Bezeczky and Wilkins, 2022). However, these study findings reinforce the continued need to target/support young people prior to them becoming habitual/chronic missing individuals to reduce the chances of further occurrences and change the trajectory of their health and justice involvement.
Thanks to Emily Stevenson and Jasmine Randone for coding the original data set and for their excellent contributions to prior publications on this topic.
No funding to report.
The author reports no conflicts of interest.
Data availability statement
Due to the limited nature of the ethical approval received, and sensitive data involved, the data are not available to be shared.
Australian Bureau of Statistics. (2011a). Australian and New Zealand Standard Offence Classification (ANZSOC). Available at: http://www.abs.gov.au/ausstats/[email protected]/mf/1234.0.
Australian Bureau of Statistics. (2011b). In focus: Crime and Justice Statistics. Available at: http://www.abs.gov.au/ausstats/[email protected]/Lookup/4524.0Chapter300September%202011.
Babuta, A. and Sidebottom, A. (2020). ‘Missing children: on the extent, patterns, and correlates of repeat disappearances by young people.’ Policing: A Journal of Policy and Practice, 14(3): 698–711.
Bezeczky, Z. and Wilkins, D. (2022). ‘Repeat missing child reports: Prevalence, timing, and risk factors. ‘Children and Youth Services Review, 136, 106454. https://doi.org/10.1016/j.childyouth.2022.106454
Biehal, N., Mitchell, F. and Wade, J. (2003). Lost from view: Missing persons in the UK. Policy Press.
Birch, P. and Sicard, L.A. (2021). ‘Young people, the police and policing’. In P Birch, M Kennedy & E Kruger [Eds.]. Australian Policing. Critical Issues in 21st Century Police Practice. Routledge. p. 383-396.
Blanchet, T. and Piketty, T. (2022). ‘Generalized pareto curves: Theory and applications.’ Review of Income and Wealth, 68(1): 263-288.
Boivin, R. and Cordeau, G. (2011). ‘Measuring the impact of police discretion on official crime statistics: A research note.’ Police Quarterly, 14, 186-203.
Bowden, F. and Lambie, I. (2015). ‘What makes youth run or stay? A review of the literature on absconding.’ Aggression and Violent Behavior, 25, 266–279.
Bricknell, K. (2017). Missing persons” Who’s at risk? AIC Research Report 08. Canberra: Australian Institute of Criminology.
Canada’s Missing. (2017). Background – 2017 Fast Fact Sheet. Available at: https://www.canadasmissing.ca/pubs/2017/index-eng.htm.
Canada’s Missing. (2018). Background – 2018 Fast Fact Sheet. Available at: https://www.canadasmissing.ca/pubs/2018/index-eng.htm.
Canada’s Missing. (2019). Background – 2019 Fast Fact Sheet. Available at: https://www.canadasmissing.ca/pubs/2019/index-eng.htm.
Colvin, E., McFarlane, K., Gerard, A., and McGrath, A. (2018). ‘We don’t do measure and quotes: How agency responses criminalise and endanger the safety of children missing in care in New South Wales, Australia’. The Howard Journal of Crime and Justice 57(2): 231-249.
Courtney, M. and Zinn, A. (2009). ‘Predictors of running away from out-of-home care.’ Children and Youth Services Review, 31(12): 1298–1306.
Crofts, N. and Thomas, S. (2018). ‘Law enforcement and public health: Finding common ground and global solutions to disparities in health and access to criminal justice.’ Journal of Community Safety & Well-being, 2(3): https://doi.org/10.35502/jcswb.56
Crosland, K. and Dunlap, G. (2015). ‘Running away from foster care: What do we know and what do we do?’ Journal of Child and Family Studies, 24(6): 1697-1706.
Dedel, K. (2006). Juvenile runaways. Problem-oriented guides for police problem-specific guides series no. 37. Washington DC: Office of Community Oriented Policing Services, US Department of Justice. Available at: http://www.cops.usdoj.gov.
Doyle, R., and Barnes, G.C. (2020). ‘Targeting missing persons most likely to come to harm among 92,681 cases reported to Devon and Cornwall Police.’ Cambridge Journal of Evidence-Based Policing, 4, 160-177.
Dudfield, G., Angel, C., Sherman, L. W. and Torrence, S. (2017). ‘The “power curve” of victim harm: targeting the distribution of crime harm index values across all victims and repeat victims over 1 year.’ Cambridge Journal of Evidence-Based Policing, 1, 38–58.
Eck, J., Clarke, R. and Guerette, R. (2007). ‘Risky facilities: Crime concentrations in homogeneous sets of establishments and facilities.’ Crime Prevention Studies, 21, 255–264.
Esposito, M.H., Lee, H., Hicken, M.T., Porter, L.C., and Herting, J.R. (2017). ‘The consequences of contact with the criminal justice system for health in the transition to adulthood.’ Longitudinal Life Course Studies, 8(1): 57-74.
Farrell, G., Sousa, W.H. and Weisel, D.L. (2002). ‘The time-window effect in the measurement of repeat victimization: a methodology for its examination, and an empirical study.’ Crime prevention studies, 13, 15–28.
Ferguson, L. (2021). ‘Profiling persons reported missing from hospitals versus mental health facilities.’ International Journal of Police Science & Management, 23(4): 372-384.
Ferguson, L. and Koziarski, J. (2021). ‘How far do they go?: A spatial examination of missing persons from hospitals.’ Policing: An International Journal, 45(2): 235-251.
Ferguson, L. and Picknell, W. (2021). ‘Repeat or chronic?: Examining police data accuracy across the ‘history’ classifications of missing person cases.’ Policing and Society, DOI: 10.1080/10439463.2021.1981899.
Flores, R. J. (2002). NISMART questions and answers. NISMART: National Incidence Studies of Missing, Abducted, Runaway, and Thrownaway Children, 4, 1-2. Available at: http://www.ncjrs.org/pdffiles1/ojjdp/fs196760.pdf.
Forst, B., Lucianovic, J. and Cox, S. (1978). What happens after arrest? A court perspective of police operations in the District of Columbia. Washington, DC: INSLAW.
Fyfe, N. R., Stevenson, O. and Woolnough, P. (2015). ‘Missing persons: The processes and challenges of police investigation.’ Policing and Society, 25, 409-425.
Galiano Lopez, C., Hunter, J., Davies, T., and Sidebottom, A. (2021). ‘Further evidence on the extent and time course of repeat missing incidents involving children: A research note.’ The Police Journal: Theory, Practice and Principles, https://doi.org/10.1177/0032258X211052900
Harris, M. F. and Shalev Greene, K. (2016). ‘Police attitudes in England to return interviews, in repeat missing person cases.’ Journal of Investigative Psychology and Offender Profiling, 13, 253-266.
Hill, L., Taylor, J., Richards, F., and Reddington, S. (2016). ‘No-one runs away for no reason: Understanding safeguarding issues when children and young people go missing from home.’ Child Abuse Review, 25(3): 192-204.
Pfeifer, J. (2006). Missing persons in Saskatchewan. Police policy and practice. Available at: http://publications.gov.sk.ca/documents/9/86185-Jeff%20Pfeifer%20Report%20on%20Missing%20Persons%20-%20Police%20Policy%20and%20Practice.pdf.
Safyer, A. W., Thompson, S. J., Maccio, E. M., Zittel‐Palamara, K. M. and Forehand, G. (2004). ‘Adolescents' and parents' perceptions of runaway behavior: Problems and solutions.’ Child and Adolescent Social Work Journal, 21, 495–512.
Shalev Greene, K. and Hayden, C. (2014). Repeat reports to the police of missing people: locations and characteristics. Available at: https://researchportal.port.ac.uk/files/1548257/Repeat_reports_to_the_police_of_missing_people.pdf.
Smith, R. and Greene, K. S. (2014). High risk?: attitudes to the risk assessment process in missing person investigations. Available at: https://pure.port.ac.uk/ws/files/1548080/Attitude_to_the_risk_assessment_process_final.pdf
Sowerby, A. and Thomas, S.D.M. (2017). ‘A Mixed Methods Study of the Mental Health and Criminal Justice Histories of Missing Persons.’ Police Practice and Research: An International Journal. doi:10. 1080/15614263.2016.1230062.
Svensson, R. and Pauwels, L. (2010). ‘Is a risky lifestyle always “risky”? The interaction between individual propensity and lifestyle risk in adolescent offending: A test in two urban samples.’ Crime & Delinquency, 56, 608-626.
Tyler, K. A., Whitbeck, L. B., Hoyt, D. R. and Johnson, K. D. (2003). ‘Self‐mutilation and homeless youth: The role of family abuse, street experiences, and mental disorders.’ Journal of Research on Adolescence, 13, 457-474.
Tucker, J. S., Edelen, M. O., Ellickson, P. L. and Klein, D. J. (2011). ‘Running away from home: A longitudinal study of adolescent risk factors and young adult outcomes.’ Journal of Youth and Adolescence, 40, 507-518.
van Dijk, A. J., Herrington, V., Crofts, N., Breunig, R., Burris, S., Sullivan, H., Middleton, J., Sherman, S. and Thomson, N. (2019). ‘Law enforcement and public health: Recognition and enhancement of joined-up solutions.’ The Lancet, 393, 287–294. https://doi.org/10.1016/S0140-6736(18)32839-3