First, a caveat: strictly speaking, we should not refer to ‘sentencing’ girls. Juveniles are not ‘sentenced’ after ‘conviction’; instead, since the implementation of the Children and Young Persons Act 1933, the youth court ‘makes an order upon a finding of guilt’ (s 59) in relation to those minors who have been successfully prosecuted.

Second, we simply do not know enough about this decision-making process with minors. In 1986, Heidensohn reviewed research about girls and young women as well as adult women. That research – stemming mainly from the 1970s and early 1980s when a more welfare-focused juvenile justice system operated – concluded that ‘delinquent girls’, like adult women, were ‘subject to a particular regula­tory censure’ stemming from ideas about what was normally expected of ado­lescent girls.[744] In particular, the response to male offending signified by ‘boys will be boys’ contrasts with the response to the ‘waywardness’ of delinquent girls.[745] So, for Heidensohn, research evidence that females of all ages risk the penalties of double jeopardy emphasises the fact that, ‘while this is a man’s world it will be his conception of justice which will prevail’,[746] in relation to minors as well as adults.

It is tempting to use twenty-year-old research to prove discrimination in the present. For example, Annie Hudson, in her chapter on ‘Troublesome girls’ in a compendium on youth justice issues published in 2002, perhaps does rely too much on conclusions drawn in the 1980s.[747] A NACRO briefing in 2001 noted that data was not available ‘at a sufficient level of detail’ to know whether the factors identified in the 1980s currently influence sentencing outcomes for offending girls. Further, even in 1987, Harris and Webb had concluded: ‘With the sole (and major) exception of the control of girls’ sexual behaviour, the differences between the treatment of boys and girls are of degree and not kind.’[748]

The reliance on possibly out-of-date research on the experiences of girl offenders is also problematic because of changes in penal policy and theory since the 1980s, which may be influencing the treatment of women offenders and reducing the likelihood that girls are still facing the discriminatory treatment that research in the 1980s suggested.[749] First, the ‘just deserts’ sentencing framework set up by the Criminal Justice Act 1991 mandated an approach primarily in terms of pro­portionality to offence seriousness and, whilst the Criminal Justice Act 2003 has amended this framework in relation to persistence of offending, this retributivist starting point is still important.[750] The current focus on the young person’s responsibility and accountability also increases the focus on the offence. In theory at least, these developments should have reduced the sentencing discretion that permitted different outcomes for girls and women.

However, in 1994, Hudson argued that the focus on proportionate punishment could work against the interests of female offenders: ‘Young women’s needs are marginalised as the criminal justice system becomes ever more offence-focused, and the sort of justice they are offered is of equal access to male provision, rather than gender-appropriate provision.’[751] A potentially gender-levelling focus on proportionality may also have been outweighed by the focus on risk. At the sen­tencing stage, minors and adults who have committed sexual and violent offences and are deemed to be dangerous must be dealt with ‘disproportionately’.[752] New provisions in the Criminal Justice Act 2003 relating to persistence (previous convictions) may also undermine proportionality. The punishment stage is also infused with risk management and the use of ever more intensive community programmes – for adults and minors.

According to Hudson,[753] ‘the factors which had been seen as bringing about harsher and more interventive sentencing of socially disadvantaged offenders. . . have reappeared as “risk of offending” factors’. The focus on risk factors can mean that life history factors increase an offender’s risk score whilst the choice of pre­ventative programmes or sentencing outcome ignores that context. For girls and women, then, the assessment of risk may itself be discriminatory.

Let us focus on a girl or boy from an abusive home who has committed a criminal offence. ASSET is the young offender assessment profiling tool used by youth offending teams to assess children and young people in relation to various interventions. There is some evidence from a national evaluation of Final Warning Projects that ASSET is not always properly used and needs are consequently not addressed.[754] One component of the form (section 2) on ‘family and personal relationships’ includes not only factors about the criminality or the health problems of the young offender’s family but also has tick boxes for ‘experience of abuse’ and ‘witnessing other violence’ in the family context. At the bottom of the page, the professional conducting the assessment has to rate on a scale of one to four the extent to which the family relationships ‘are associated with the likelihood of further offending’. This is one of 12 ratings which have to be totalled at the end with a maximum high-risk score of 48. A high score may lead to intensive supervi­sion, which ignores the factors which led to the risk score. For a girl, this might mean a mixed-sex community project or pressure to stay with her family, when in neither context does she feel safe and may not be safe.

As noted above, Home Office research on siblings also points to gender implica­tions of the prediction of risk of offending.[755] The finding that ‘socio-economic factors such as low social class, low family income, poor housing and large family size predicted offending more strongly for sisters than brothers’ might also confirm that girls and young women in custody are more likely to have experienced ‘struc­tural’ disadvantage in their lives, which needs more targeted programmes as part of punishment. There is also an increasing awareness that not only are the young disproportionately at risk of violence, but that females experience different forms of violence, notably violence within the home.[756]

A telephone survey of criminal justice practitioners by Beatrix Campbell sug­gested that professionals in the front line are aware of gender issues.[757] The cases summarised by practitioner respondents reveal boys ‘watching their mothers beaten to hell by their fathers’ and mothers who felt disempowered and defeated. Campbell noted that government guidelines on the use of parenting orders ‘refer only to “inadequate” or “harsh and erratic” parenting as risk factors, and cite no references to the significant volume of research on domestic violence and its impact on children’.[758]

Campbell’s comments were based on draft guidance, the final version being issued in March 2004. This states that both parents should be seen if they both participate in the child’s upbringing, ‘unless a parent is estranged, for instance because of domestic violence’.[759] It further states that ‘information that emerges during the intervention and assessment process about domestic violence or abuse will need to be passed on to police and social services for action’ and practitioners should establish with other agencies whether they have such information.[760] How­ever, as research on practitioners in the family justice system has shown[761] and as one of Campbell’s respondents phrased it, domestic violence ‘won’t come out unless you create a climate’.[762]

The violence and harm that may be done to girls and young women – as offenders, as the mothers of offenders or as young offenders who are themselves mothers – does not get highlighted. Instead, as the Howard League pointed out, the focus is on the violence done by girls, and that results in an increasing tendency to give more severe sentences to girls than boys where violence is involved.[763]