A sharp decline in youth crime: reviewing trends in New Zealand’s youth offending rates between 1998 and 2019

ABSTRACT There is a paucity of research investigating New Zealand’s rates of youth offending. This study examines trends in youth offending data between 1998 and 2019, then compares these with the rate at which emerging adults (aged 17 to 24) entered the justice system during this period. Our approach aimed to explore whether changes in youth offending may have had a downstream impact on rates of justice-system contact among emerging adults. Results indicate that the overall rate at which youth were alleged to have offended (per police data) reduced by 58% between 2010 and 2018, which coincided with a significant reduction in the rates at which emerging adults were imprisoned, sentenced or remanded into custody. Additionally, the rate at which young people were formally charged in Youth Court fell by 73% between 2008 and 2018. However, the rate of decline was less for the overrepresented Indigenous Māori and minority ethnic group of Pasifika young people. Serious offending (including by minority groups) reduced at lesser rates than did less serious offending, while reoffending rates remained fairly static. International trends in reductions of youth offending are examined, and recommendations made to better manage the needs of young people at risk of serious offending.


Introduction
Over the last 30 years, crime has declined across most Western societies (Farrell et al., 2011).For many jurisdictions, the decline in the number of young people offending has been so sharp it has driven a reduction in overall crime (Griffiths & Norris, 2019).For example, Payne et al. (2018) examined offending among two birth cohorts born in the Australian state of New South Wales (NSW) in 1984 and 1994.By age 21, there was 49% less contact with the criminal justice system for those born in 1994 than for those born in 1984.There was also a considerable decline in property-related offending among the 1994 cohort.The authors argued that this significant drop in the volume of youth offending was responsible for the overall decline in crime in NSW, as offending rates among adults remained steady.
Such marked declines in youth crime have been widely seen, with the number of individual youth offenders and frequency of offences dropping steadily at different times over this period in North America (Zhang et al., 2016), the United Kingdom (Griffiths & Norris, 2019), Europe (Andersen et al., 2016;Van der Laan et al., 2019) and Australia (Payne et al., 2018).Are youth crime rates also dropping in New Zealand?In the past, the international criminal justice community has held up New Zealand's youth justice system as an innovative example of how they should operate, with widespread use of diversion, decarceration, victims' rights and family involvement (Lambie, 2018a).Research suggests that the efficacy of the country's youth justice system rests largely on these diversionary measures (Maxwell, 2013), which are essential for mitigating the criminogenic risks associated with young people entering the formal justice system (McAra & McVie, 2007).
Yet, New Zealand's young people engage in high rates of offending compared to adults, and New Zealand's high adult prison population is also a cause for concern (Lambie, 2018a).Young people and emerging adults (15-to 24-year-olds) make up 40% of all apprehensions (Lambie, 2018a), but data on patterns of youth offending over recent decades have not been systematically investigated.
This study aims to address this gap in our understanding, in light of international youth crime patterns, such as the decline in youth offending, the social issues associated with offending and government policy impacts.Offending data will be used to determine whether New Zealand's rate of youth offending changed between 1998 and 2019 and, if so, whether changes have differed according to ethnicity.Data on seriousness of offence and reoffending rates (which may signal risks of more persistent offending) will also be analysed.In addition, we will contrast key trends extracted from Ministry of Justice youth offending data with the offending behaviour of emerging adults (from age 17) and adults (aged 30 to 39) to compare cohorts.We hope this process will generate a useful discussion on how to respond to recent trends and create a ripple of positive social, cultural and economic change.

Decline in youth offending
The downwards trend in youth offending globally is likely due to a complex web of social, economic and policy-related factors (Griffiths & Norris, 2019;Payne et al., 2018;Van der Laan et al., 2019;Weatherburn & Rahman, 2021), not one single factor, and diverse explanations have been promulgated by researchers (Farrell et al., 2011;Svensson & Oberwittler, 2021).These include offender pathways (adolescent-limited vs. lifecourse-persistent); social factors (inequity, attitudes and behaviour); and government policy (such as use of criminal sanctions or diversion).

Young offender pathways
Evidence suggests that fewer first-time offenders are now coming into formal contact with the criminal justice system (Sivertsson et al., 2019).So-called 'adolescent-limited' offenders often start offending in their teens, engage in less serious, often peer-led offending and eventually desist from antisocial behaviour upon exiting adolescence (Moffitt, 2017).The underlying 'maturity gap' between their biological and social adulthood and peer associations strongly influence their transient antisocial behaviour (Moffitt, 2017).Due to the justice system's criminogenic influence on young people (McAra & McVie, 2007), effective diversionary measures are usually the most appropriate response to adolescent-limited youth offenders (Clancey et al., 2020).Griffiths and Norris (2019) argue that the reduction in offending among young people fitting an adolescent-limited offending profile is partly due to contemporary youth facing less exposure to offending behaviour.
In contrast, some young people's offending is rooted in early childhood, with early antisocial behaviour and offending persisting well into adulthood, as so-called 'lifecourse-persistent' offenders (Moffitt, 2017).Cohort research, such as Payne et al.'s (2018) in NSW, identified a core group of chronic offenders in the 1994 cohort responsible for more offending than their 1984 counterparts; the 'hardcore' recidivist offenders described by Payne and colleagues tend to fit a 'life-course-persistent' profile (Moffitt, 2017).Those who follow this offending trajectory are responsible for most of the demands and costs associated with criminal justice (Allard et al., 2020), and diversionary measures used to manage adolescent-limited offenders are often seen as insufficient (Zara & Farrington, 2020).They are also responsible for most offences committed by youth, particularly at the severe end of the spectrum (Jolliffe et al., 2017).Research indicates that the proportion of youth crime these life-course-persistent offenders are responsible for might also be growing (Griffiths & Norris, 2019;Payne et al., 2018), and the younger a child is when they first commit a crime (and are caught), the more likely it is that they will keep committing crimes (Weatherburn & Rahman, 2021).Moffitt's (2017) analysis ascribes to these children high rates of violent behaviour, volatile temperaments and entrenched antisocial attitudes, plus high rates of 'significant neurocognitive deficits', such as fetal alcohol spectrum disorder (FASD; Corrado & McCuish, 2015) and severe traumatic brain injuries (TBI; Williams et al., 2018).However, there is a risk of limiting our analysis to only these individual factors, rather than the societal conditions and contexts that are associated with them (such as where TBI is from physical abuse in infancy and FASD is from a lack of alcohol restrictions; Lambie, 2020) and that have influenced their behaviour (Ball et al., 2023).

Social issues
Life-course-persistent offenders can be seen as on the so-called 'prison pipeline' from early trauma and adverse experiences to childhood offending, youth-justice involvement and, finally, adult incarceration (Lambie, 2018b), within social contexts that are inequitable and entrenched.
These young people have almost invariably experienced trauma and grown up in adverse and abusive environments (Craig et al., 2017).Life-course-persistent offenders are also disproportionately 'crossover kids'-victims of significant childhood maltreatment, which leads to their involvement with the care and protection system, including out-of-home care, disrupted attachments and education, early offending with antisocial peers and on to 'crossing over' from the child-welfare system into the youth-justice system (Baidawi & Sheehan, 2019;Reil et al., 2021).
Social harms entrenched for Indigenous people and ethnic minorities within Western nations also affect youth offending patterns (Papalia et al., 2019).Effects of land dispossession, colonisation (and post-colonisation policies), systemic racism, lower rates of employment and school completion and higher rates of negative health and socioeconomic sequelae can leave young people alienated, locked out of life chances and drawn to offending for survival and status (McCarthy, 2021).
Māori are New Zealand's Indigenous people who are overrepresented as victims and offenders at every level of New Zealand's justice system (McIntosh & Workman, 2017).Compared to non-Māori, they also face more offending-related risk factors during childhood (Reil et al., 2021).The impact of colonisation and the resulting marginalisation of Māori perspectives in public life has led to an absence of Māori self-determination within the justice system (Webb, 2017).Many have called for the youth justice system to be based on a more meaningful partnership between the government and iwi (tribal groups) and to correct the tokenism of past attempts at integrating Māori culture into its foundations (Cleland & Quince, 2014).Given this history, it is vital to investigate differences between rangatahi Māori (Māori young people) and non-Māori youth offending rates to observe how these inequities permeate the sphere of youth justice.
Both social and individual risk factors for violence that young people face were the focus of attempts in Scotland to tackle their high rates of youth violence (Klose & Gordon, 2022).Their approach utilised a range of community-based initiatives to address socio-economic inequities such as poverty, educational attainment and social exclusion (Klose & Gordon, 2022).Crucially, they also implemented early intervention and prevention strategies to fully address the needs of children from at-risk environments (Klose & Gordon, 2022).
Researchers have also argued that changes in the attitudes and behaviours of contemporary young people are a necessary part of any explanation of changing youth offending patterns.Owen and Cooper (2013) posited that young people are increasingly bereft of opportunities to engage in 'debut crimes', as their usual antecedents have become rarer-for example, better security measures in homes, cars and smartphones have made opportunistic property offences more difficult (Weatherburn & Rahman, 2021).Baumer et al. (2021) examined self-reported delinquency among a sample of American youth aged between 13 and 16 (N = 204,621).They found reduced offending rates were strongly associated with decreases in 'unstructured socialising' (time spent with peers in an unstructured context with little supervision from adults) and alcohol consumption among the cohort, which have, in turn, been linked to more online socialising and activity (Ball et al., 2023;McAra & McVie, 2018).Similarly, Svensson and Oberwittler (2021) found that reductions in unstructured activities could explain 95% of the between-year variance in delinquency among a sample of Swedish young people (N = 49,000), with fewer evenings spent with friends and consuming alcohol accounting for a large proportion of delinquency-behaviour variance (Svensson & Oberwittler, 2021).
Unstructured socialising has in the past been linked to increased involvement in crime, higher tolerance for antisocial behaviour and broader exposure to antisocial peers among youth (Hoeben & Weerman, 2016), plus more attention from police when disengaged young people congregate in disadvantaged neighbourhoods with little else to do, and get charged (McCarthy, 2021).Online socialising, on the other hand, may include offending behaviour, but that is much harder for authorities to detect and prosecute (McAra & McVie, 2018;McCuddy, 2022).

Government policy
The contribution of government policy is another crucial factor to explore relevant to falling youth crime rates.For example, the United Kingdom's 'net-widening' justice policies of the early-2000s caused much of the antisocial behaviour of UK youth to be criminalised.Most of these incidents would have previously been dealt with informally by police, meaning that the policies disproportionately affected young people, leading to a 22% increase in formal justice system contact for under-18s by 2007 (Bateman, 2014).However, in 2008, the government set a target to reduce the number of first-time entrants into youth justice and began prioritising diversionary responses.This shift in policy correlated with a 73% decline in proven offences committed by children and young people by 2016 (Bateman, 2017).It is unlikely that shifts in the criminal behaviour of young people could account for this striking reversal, over and above the impact of increased diversion.
Diversion may contribute to falling youth offending rates in several ways.Firstly, reducing the number of young people exposed to the formal justice system's criminogenic influence can reduce their risk of engaging in further antisocial behaviour (McAra & McVie, 2007).Secondly, the conditions often associated with a diversionary plan can compel (or at least facilitate) young people to access supports they may otherwise not have.Thirdly, in some jurisdictions, young people diverted from the justice system do not show up in the official crime statistics used to measure offending (Bateman, 2014), although in others, re-offending rates of those diverted are compared with those sent to court.It is important to acknowledge that these diversionary measures are targeted at less-serious youth offenders who have likely experienced little to no justice system contact.New Zealand's main piece of youth justice legislation, The Oranga Tamariki Act (1989), outlined the principles emphasising diversion, decarceration, victims' rights and the empowerment of whānau (families).*(*Note: 'Oranga Tamariki' is New Zealand's ministry for children, including responsibility for care and protection, and for children who offend under age 14; from age 14, they come under the Ministry of Justice youth justice system.)Police were encouraged not to charge most young people they apprehended; instead, they were to be cautioned or referred to specialist Youth Aid officers who plan alternative actions (Cleland & Quince, 2014).If the offending was serious enough or a young person consistently reoffended during or following an alternative action plan, they faced more formal proceedings (Ministry of Justice, 2020b).
In 2010, the government passed an amendment to the Act, which introduced a further principle that any interventions instituted in response to a young person's offending must address its underlying causes.The Act legislates a robust and evidence-based response to youth offending (Jolliffe et al., 2017;Reil et al., 2021).Initiatives such as the Policing Excellence Initiative (2008) and subsequent Prevention First Model (2011) shifted police towards a more proactive and prevention-focused approach to youth offending (both cited in Den Heyer, 2015).
Yet, as noted, young Māori or Pasifika are persistently over-represented in negative offending statistics, and the country continues to have a high adult prison muster (Lambie, 2018a).Although youth offenders are not imprisoned in these facilities, the incarcerated adults are individuals whose offending behaviour started early before becoming chronic and increasingly severe (Lambie, 2018a).As a result, it is important to establish the rate at which young people are reoffending following contact with the Youth Court, as this information has implications regarding the volume of young people offending in a manner that warrants escalation through the system.Similarly, it is crucial to investigate if the proportion of serious youth offending has changed, given that this offending is a marker for individuals at risk of following a life-course-persistent offending trajectory (Moffitt, 2017).

Method
The published data available from the Ministry of Justice for this study included statistics on all young people (aged 14 to 16 years) alleged to have committed an offence between 1998 and 2019.It also included all young people charged with a crime between 1998 and 2019.Three subsets of data were available for 14-to 16-year-olds from 2010 to 2018; namely, 1) the seriousness of offending (where justice-sector scores of seriousness are based on the actual sentences imposed; Ministry of Justice, 2018); 2) reoffending rates (the proportion of 14-to 16-year-old offenders who appeared in the Youth Court, and were proven to have reoffended in any court within the following 12 months); and 3) ethnicity (categorised as either 'Māori', 'Pasifika' or 'European/Other').Any reference to the 'total' youth offending rate refers to changes in the overall prevalence of youth offending, meaning that each young person was counted only once in any given year.
For adults, data were available on all those aged 17 to 39 who were imprisoned, placed in custodial remand, or given a community sentence between 1999 and 2019 (the Department of Corrections adult dataset made available for our research started in 1999; the Ministry of Justice youth justice dataset started in 1998).Note that the data were collected when 17-year-olds were still classified as 'adults' in the New Zealand justice system and were eligible for adult prison.Since 2019, 'adulthood' in the prison system starts at age 18, in line with similar jurisdictions (such as Queensland, Australia where the age shifted in 2018; McCarthy, 2021).
The rationale for the inclusion of adult data was to examine whether earlier changes in youth offending may have seen a flow-on effect in changes in the number of 'emerging adults' entering the justice system.
Ethics approval was not sought for this study as all data were pre-existing and publicly available.

Procedure
The publicly available youth offending data were originally gathered by the Ministry of Justice, Oranga Tamariki and the New Zealand Police (Ministry of Justice, 2020a).The adult data for imprisonment, custodial remand and community sentencing rates were gathered by the Department of Corrections and were also publicly available.Population data from Statistics NZ were merged with the rest of the data into a single dataset.From this, rates per 10,000 members of the population were created to make changes by age group comparable.Since all data used were publicly available, none were unit record data.As a consequence, individuals could not be linked over time; instead, data were aggregated by variables such as age, year and ethnic group.
As a result of this process, we can estimate the potential downstream impact of any historical changes in youth offending on the broader criminal justice system, specifically by comparing the numbers of young offenders within youth justice to numbers of emerging adults and older age groups, to investigate changes in the volume of young people at different stages within the youth justice system by broadly comparing cohorts.

Results
The results show that New Zealand's youth offending rates (for those aged 14 to 16 years) have dropped markedly, but rates for disadvantaged young people and persistent offenders/reoffending have seen smaller reductions, in line with trends in other Western nations.Rates of imprisonment, custodial remand and community sentencing for emerging adults (age 17+) are lower than for older adults (30 years+).
Table 1 shows demographic and offence seriousness information for all youth proceeded against by the New Zealand Police from 2009/10 to 2018/19,* alongside total population percentages.(*Note: The period 2009 to 2018 was chosen as data for all variables of interest were publicly available and this was the period of greatest changes in rates of offending, as shown in Figure 2.) Just under half of all young people proceeded against during the period identified as Māori (48.4%), who comprise 36.7% of the total youth population.European/Other make up the next highest proportion (39.4%), compared to 50% of total youth.Pasifika comprise 8.6% of young offenders and 14.3% of the total population.Most young people in the dataset were male (69.3%), and a significant proportion identified as female (30.7%).Age was distributed relatively evenly, with 16year-olds making up the highest proportion of young people proceeded against by the police (38.2%), followed by 15-year-olds (33.3%) and 14-year-olds (28.4%).Just over half of the offences for which police proceeded against young people were of low (28.3%) or low-medium seriousness (27.1%).Offences with a high rating of seriousness were the least common (8.3%).
The total rate at which youth were alleged to have offended fell by 58% between 2010 and 2018, from 743 young people per 10,000 members of the population to 312 per 10,000.Figure 1 shows this trend broken down by ethnicity.The reduction in youth  offending was largest for the European/Other group, falling 71% during this period.The youth offending rate fell by 64% and 56% for Pasifika and Māori, respectively.
As shown in Figure 2, the overall rate of young people receiving formal charges followed a similar trend to the total youth offending rate.The rate of youth receiving formal charges fell by 73% between 2008 and 2018.Prior to this downwards trend, the rate of young people receiving formal charges had remained fairly steady, increasing by 5% from 1998 to 2008.
While the overall youth offending rate has fallen since 2010, the proportion of youth crime considered to be of a serious nature has risen.As shown in Figure 3, the overall number of young people proceeded against for offences that were deemed to be of medium-high or high-level seriousness reduced by 44% from 2010 to 2018.In 2018, however, 35% of the young people who had offended were proceeded against by the police for offences that were of medium-high or high-level seriousness, compared to 25% in 2010.
The rate at which young people reoffended within 12 months of appearing in Youth Court remained fairly steady from 2009/10 to 2016/17.Reoffending rate is broken down by ethnicity in Figure 4.The reoffending rate for Māori young people was 49% compared to 38% for both Pasifika and European/Other young people.
Figure 5 shows that across the past two decades (1999-2019), rates of imprisonment for those aged 30 to 39 years steadily increased.In contrast, from 2009, rates at which both 17-to 19-year-olds and 20-to 24-year-olds were imprisoned decreased by 83% and 62%, respectively.In the earlier period, from 2001 to 2007, the imprisonment rate for 17-to 19-year-olds increased by 14%, while the rates for both 20-to 24-year-olds and 30-to 39-year-olds rose by 17%.
Figure 7 shows that between 2009 and 2019, the community sentence rate for 30-to 39-year-olds reduced by 10%.This was a smaller reduction than the 57% and 78% decreases for 20-to 24-year-olds and 17-to 19-year-olds, respectively.While the community sentence rate also declined for all three age groups between 2001 and 2007, the decline was not as large for the 17-to 19-year-olds (25%) and 20-to 24-year-olds (23%), compared to the post-2009 decline.In sum, results revealed that between 2010 and 2018, the overall rate at which young people aged 14 to 16 were alleged to have offended by police fell by 58% from a steadily higher rate in the previous decade.Similarly, the rate at which young people were formally charged with an offence also fell, reducing by 73% from 2008 to 2018.Of concern, the proportion of crimes committed by young people considered to be of 'medium-high' or 'high' seriousness rose by 10% from 2010 to 2018.The reoffending rate for youth remained steady during roughly the same period, decreasing by 3% from 2009/10 to 2016/17.There was a significant reduction in the number of emerging adults (aged 17 to 24) beginning prison or community sentences and being placed on custodial remand over the period.However, our results did not demonstrate similar reductions for older adults (aged 30 to 39) across the same measures.

Discussion
This study examined trends in New Zealand's youth offending rates between 1998 and 2019 and compared these findings with Department of Corrections data that included the custodial remand, imprisonment and community sentence rates for adults between 1999 and 2019.Overall, the results showed that a decline in the youth offending rate, particularly from 2010 to 2018, coincided with a sharp decline in the number of emerging adults entering the adult justice system.The overall reduction in volume at different levels of New Zealand's youth justice system is roughly consistent with similar trends in youth offending that researchers have observed in other Western countries over the last 30 years (Andersen et al., 2016;Griffiths & Norris, 2019;Payne et al., 2018;Van der Laan et al., 2019;Zhang et al., 2016).
Our reasoning for focusing specifically on the 2010-2018 period was that (as shown in Figure 2), the rate of alleged offending in youth had remained relatively static and unchanged in the preceding 10 years, and the Oranga Tamariki Act amendments in 2010 called for more intensive diversionary and holistic responses to the drivers of crime, which ideally would have seen rates drop.(There was also the pragmatic issue that the publicly accessible data that covered the most elements-offending, ethnicity, seriousness etc-were for that period.) Although offending rates dropped for Māori, Pasifika and European/Other youth between 2010 and 2018, the decline was smallest for Māori.Young Māori make up 36.7% of the total youth population but 48.4% of youth who are justice-involved (Table 1).The offending rate for Māori young people fell 15% less than the European/ Other group.These results are made more concerning in the face of the historic overrepresentation of young Māori within the criminal justice system (McIntosh & Workman, 2017).Recent data from 2019/20 showed that the Youth Court appearance rate for Māori was 8.3 times higher than it was for European/Other young people (Ministry of Justice, 2020b).Māori youth were also proceeded against and remanded into custody almost twice as often as were European/Other youth (Ministry of Justice, 2020b).The proportion of young people appearing in Youth Court who identify as Māori has increased by 5% since 2011 (Ministry of Justice, 2020b), despite increased cross-agency efforts to reduce their presence in the justice system (Ministry of Justice, 2013).
Additionally, our results showed that the youth reoffending rate has remained relatively static, decreasing by just 3% from the available 2009/10 to 2016/17 figures.As of 2016/17, rangatahi Māori reoffended 11% more than did non-Māori in the 12 months following an appearance in Youth Court.However, data from the following year (2017/18) showed that the proportion of Māori youth reoffending fell to 42%, only marginally higher than the overall youth reoffending rate of 41% (Ministry of Justice, 2020a).These results indicate that the overall drop in the offending rate has not coincided with a decline in the rate at which young people reoffend.Given that these findings are derived from aggregate data, we cannot parse out why the reoffending has remained static in the face of significant changes in the overall rate of youth offending.However, the available dataset defines 'reoffending' as having had an offence proven in any court within 12 months of a Youth Court appearance, rather than a broader definition of general reoffending over a longer timeframe.With typical delays in processing and court times, a 12-month timeframe can be considered very tight, so the reoffending rate may have shifted more than these data imply.The current reoffending dataset also does not allow for a more detailed understanding of the systemic effects on specific ethnic groups as they engage with the justice system (including expressions of racism, bias, barriers to legal and support services etc).Further research in this area is greatly needed.
Given the renewed focus on reducing escalation through the system when dealing with less serious offending (Ministry of Justice, 2013), it may be that fewer young people fitting an 'adolescent-limited' offending profile are now entering the system.Similar to Australia (Payne et al., 2018), this trend may be driving the overall decline in New Zealand's youth offending rate.It would make sense then that those young people still entrenched in the youth justice system may be engaging in more chronic offending behaviour (Kennedy et al., 2019).The shifting proportions in the types of offences committed by young people further indicate that this may be the case.
While our results show that the overall volume of young people entering the youth justice system has markedly reduced, those who do are now more likely to have engaged in offending of a serious nature.More recent results indicate that this proportion has not shifted much from 35% in 2018, with offending of 'high' or 'medium-high' seriousness making up 33% of all youth offending in 2020 (Ministry of Justice, 2020a).This makes the relatively static reoffending rate more comprehensible if a higher proportion of serious, chronic youth offenders are now in the justice system.Further resourcing of evidence-based systemic interventions, particularly for culturally led services, may help to reduce the youth reoffending rate among those young people with more entrenched antisocial behaviour.
Interestingly, our results indicated that, between roughly 2014 and 2016, there was a reversal in the downward trends in imprisonment and custodial remand rates among emerging adult populations.These increases were due to a combination of higher proportions of individuals appearing in court charged with serious offences and judges being more likely to employ custodial responses (Ministry of Justice, 2017).However, our results showed that the increases were much less marked for emerging adults than for 30-to 39-year-olds.The principal reason for this is that there were much more significant reductions in the numbers of those aged 17 to 19 (20%) and 20 to 24 (17%) charged from 2014 to 2017, compared with those aged 30 to 39 (7%).New Zealand's criminal justice policy during this period may have contributed to these trends, with judges acting in response to 'tough on crime' political rhetoric and legislation (Pratt, 2017).This policy may have also spilt over into the youth justice space, as the rate at which the Youth Court charged young people flattened out after decreasing for six years, while the rate of alleged offending continued to decline during the same period.
These results fit within international literature, indicating that the proportion of youth crime that life-course-persistent youth offenders are responsible for might be growing (Griffiths & Norris, 2019;McCarthy, 2021;Payne et al., 2018).This trend may reflect a reduction in adolescent-limited youth offending, as fewer young people face the kind of situational antecedents common to more transient offending behaviour (Griffiths & Norris, 2019).Though our results demonstrated that the reoffending rate has remained steady, and the proportion of youth crime considered to be 'serious' has grown, we did not explicitly measure whether a larger proportion of young people who offend now fit a life-course-persistent profile (Moffitt, 2017), as such data were not available.Local research on those young people who fit that profile remains sparse.
These results are not conclusive evidence that the decline in youth offending is driving a decrease in the offending rates of emerging adults.They do, however, demonstrate a connection between the sharp drop-off in youth offending that began in 2010 and a drop in the rate at which 17-to 24-year-olds are imprisoned, remanded into custody and given community sentences.These same trends were not present in older adult age groups, with imprisonment rates climbing among those between the ages of 30 and 39.So, it may be that we are seeing reductions in youth offending having a downstream impact on the numbers of emerging adults entering into justice services.
If this is the case, this trend would fit our existing knowledge of the criminogenic effect of early justice system contact on young people.With fewer young people coming into formal contact with the justice system, fewer are facing the criminogenic experiences that can then cause their antisocial behaviour to escalate (Capece, 2015;Cowan et al., 2019;Lambie, 2018a).As police divert more young people fitting an adolescentlimited offending profile from the formal justice system, they reduce the opportunities for those young people to shift into more serious offending.Reducing the chance of this shift occurring via early and effective intervention is especially important for those young people at a higher risk of engaging in chronic offending (Craig et al., 2017).
While our results provide clear evidence that significant change has occurred in youth offending over the last 15 years, they do not reveal why this change has occurred.Doing so would require further empirical research into the policy changes instituted by the government in and around the period in which these trends began.Moreover, it would also require further research into the broader societal trends that may have also contributed.However, international hypotheses concerning declining youth crime rates could also be relevant to understanding our results.
For example, the young people of New Zealand are likely no exception to the changes in routine behaviour that international researchers have implicated in reduced youth offending (Baumer et al., 2021;Svensson & Oberwittler, 2021).The proportion of young people in New Zealand who report binge drinking reduced from 36% in 2007 to 22% in 2019 (Fleming et al., 2020).This reduction means that fewer young people are consuming alcohol in the type of patterns most strongly associated with offending behaviour (Baumer et al., 2021).Moreover, New Zealand youth are also spending more time online than ever before (Pacheco & Melhuish, 2019), likely limiting their opportunities to engage in unstructured socialising (Ball et al., 2023).Local research has also indicated that young people with 'maladaptive' personality traits are more likely to engage in aggressive and harmful online behaviour (Kurek et al., 2019).Consequently, the delinquency that some youth previously reserved for their physical environment they may now instead conduct or communicate online (McCuddy, 2022).Future research should investigate changes in the attitudes of New Zealand youth towards offending and delinquency, unstructured socialising and online behaviour to uncover the extent to which international findings generalise to a local context.
Changes in policing practices since 2008 may have also contributed to the reduced offending rate among young people in New Zealand (Den Heyer, 2015).In 2009, the government instituted a new approach to reducing offending and victimisation called 'addressing the drivers of crime' (Ministry of Justice, 2009).The report categorised 'youth' as one of the five main drivers of crime in New Zealand, which may have inadvertently framed young people as a cause of offending and victimisation rather than the risk factors for offending those young people are subject to (Den Heyer, 2015).Nevertheless, attempts by the New Zealand Police to promote diversion and reduce escalation through the system (such as through the 2011 'Prevention First Model'; New Zealand Police, 2011) have likely contributed to the significant downward trends in youth offending, particularly in the number of young people receiving formal charges.
Recommendations: a public health approach to youth offending For many in the media and general public, youth offending is purely an issue of criminal justice.This line of thought sidelines the role trauma and contextual factors, such as social and economic inequities, play in the development of antisocial behaviour.As a result, media and politicians' responses to youth offending (in New Zealand and elsewhere) are defined largely by their 'punitive undertones' (Gordon et al., 2021, p. 40;Pratt, 2017).There may be more utility in taking a public health approach to addressing youth involvement in offending, which aims to intervene before the offending-and victimisation-has occurred (Moore, 1995).That means embedding strategies into a broader population-level approach that addresses the early risk factors at the root of delinquency and violent behaviour, from housing to health (Welsh, 2005).A public health approach also deemphasises the individual motivations and intentions central to a criminal justice understanding of youth offending.Instead, it posits that youth offending arises from a complex causal system underpinned by social structures, access to criminogenic commodities and exposure to situations that lead to offending (Moore, 1995).Youth offending can then be considered an unfortunate symptom tied to the health and well-being of children and their whānau.
Children who 'crossover' from the care and protection system into the youth justice system would benefit from a robust public health approach to youth offending (Baidawi & Sheehan, 2019).Research from He et al. (2021) found that the more persistent and earlier the childhood maltreatment, the stronger its association with youth justice system contact.These findings suggest that intervention efforts should be applied more broadly for care and protection-involved children and whānau.Moreover, the level of response should be subject to the age of first care and protection contact and both the type and number of risk factors present.Robust and culturally appropriate primary prevention strategies can interrupt a child's steady march from care and protection to youth justice.Additionally, when strategies address the risk factors tied to later offending early enough, they can prevent those first steps from happening entirely.
Holistic approaches are critical for those children most at risk of developing into lifecourse-persistent offenders.Individuals whose offending behaviour matches this profile do the most damage to society and should therefore be the target of our most intense prevention and early intervention efforts (Reil et al., 2021;Zara & Farrington, 2020).Our results indicated that rates of youth recidivism and serious offending remain a concern in New Zealand.Indeed, these are hallmarks of life-course-persistent offending behaviour (Moffitt, 2017).An effective approach for this group must be systemic, trauma-informed and should not alienate these young people from their communities (Zara & Farrington, 2020).
As young people access the type of services involved in a public health approach, from housing assistance and education services to family and cultural support, success depends on those services' ability to engage and retain young service users.In the Scottish public health approach, for example, much of the frontline response to youth violence was not from justice-sector staff but from key workers such as mentors, healthcare workers and mental health professionals (Gordon et al., 2021).These individuals are what Gordon et al. (2021) refer to as 'credible messengers'.Those with their own experience of trauma and justice involvement can be best suited to act as trusted communicators and potentially foster the type of participation among justice-involved young adults needed to create positive outcomes.
A public health approach could address the stark inequity evident in Māori overrepresentation in the justice system and providing services that support a Māori way of being (Reil et al., 2021).The partnerships Oranga Tamariki have built with iwi are a good start but have resulted mainly in small-scale initiatives (Oranga Tamariki, 2020).Existing interventions, such as Functional Family Therapy, have also been successfully adapted to better suit Māori clients (Heywood & Fergusson, 2016).However, the government should form a larger-scale partnership with Māori, ideally resulting in a more well-coordinated and resourced kaupapa Māori approach to youth offending prevention.
Partnerships with Pasifika communities, churches and families are also needed in a more holistic public health approach to preventing Pasifika young people from an offending trajectory (Ioane & Lambie, 2016;Paterson et al., 2016).
It is essential that, as a society, we consider both the impact our youth justice policies have on marginalised young people and our responsibility to ensure each response instituted is practical, effective and evidence based.As we continue to evaluate whether our responses live up to this standard, we must incorporate the voices of those frontline workers, children, young people and whānau who are directly impacted.Future research should gather these perspectives, particularly concerning their needs regarding service delivery, equity and accessibility.While we can also draw inspiration from international examples of successful public health approaches to youth offending, there are inherent difficulties associated with policy transfer (Klose & Gordon, 2022)

Limitations
This study's limitations are primarily related to our use of administrative data.The publicly available data used were not unit record data, so individuals could not be linked over time, nor were ethnicity data as nuanced as would be desirable (eg, 'Pasifika' covers a number of cultures and does not distinguish between 'NZ-born ' and 'Island-born').Future research should build on our straightforward analysis by qualitatively investigating the underlying mechanisms behind contemporary trends in youth offending, including the perspectives of young people, families, offenders and victims, diverse communities and frontline staff.
Moreover, these administrative data did not report on risk factors for offending such as young people's substance use, parental imprisonment and school exclusion history (Zara & Farrington, 2020).Undetected youth offending is, by definition, undetected, such as when it occurs online.All forms of sexual offending at all ages are under-reported and under-processed (McCarthy, 2021).Therefore, our results only include that offending behaviour commonly brought to police attention and not necessarily the 'true' youth offending rate.

Conclusion
The current study sought to examine trends in youth offending from 1998 to 2019.In doing so, our goal was to determine whether any changes in the number of young people offending coincided with changes in the number of emerging adults entering the justice system.We found that the drop in volume observed in youth offending coincided with an overall decrease in the volume of emerging adults being placed on remand and starting prison or community sentences.Young people will inevitably make decisions that bemuse adults (Clancey et al., 2020).Their sometimes risky and harmful behaviour is usually a symptom of a not-yet-fully-developed brain trying to make sense of the world around it (Steinberg, 2017).Adolescent risk-taking behaviour will remain something the justice system often manages via diversion and alternative action plans.What is not quite so tolerable is the harmful and chronic offending behaviour of some young people.The lives of these vulnerable and marginalised youth are littered with maltreatment, undiagnosed neurodisabilities and mental health problems, neglect and early system involvement (Craig et al., 2017;Jolliffe et al., 2017).Offending behaviour of this kind requires robust multi-agency responses to support these young people to break out of the prison pipeline.

Figure 2 .
Figure 2. Rate of Alleged Offending and Rate of Formal Charges for 14-to 16-year-olds.

Figure 3 .
Figure 3. Proportion of Youth Offending by Seriousness.

Figure 6 .
Figure 6.Custodial Remand Rate by Age Group.

Figure 7 .
Figure 7. Community Sentence Rate by Age Group.
. Thus, we must consider what works within a local context.Further research efforts should then track samples of system-involved individuals over time and thus shed light on what specific public health interventions are most effective with local populations.

Table 1 .
Frequencies and Percentages of Demographic Information and Offence Seriousness for all Youth Proceeded Against by the New Zealand Police from 2009/10 to 2018/19 and Percentages of Demographic Information for the Overall Youth Population in New Zealand from 2010/11 to 2018/19.Demographic and offence seriousness data prior to 2009/10 was unavailable.Individuals with a prioritised ethnicity of Asian, MELAA (Middle Eastern, Latin American or African) or Other were included in the European/Other group.Unknown group unavailable for overall youth population.Overall population percentages found in brackets.