Spotting tomorrow's criminals in today's pushchairs
We can predict outcomes and change them, says Blair crime review
Analysis The Government's efforts to spot and intercept potential criminals during childhood, which first broke surface last summer, have come into sharper focus with the publication of this week's policy review, Building on Progress: Security, Crime and Justice,. "Universal checks throughout a child's development," says the review, will "help service providers to identify those most at risk of offending."
These checks, effectively a series of 'spot the teen crim' tripwires, "should piggyback on existing contact points such as the transition to secondary school." This however should be seen in the context of the Government drive for increased data exchange among agencies dealing with children (via, for example, the Information Sharing Index), and on ever-increasing statutory requirements on professionals dealing with children to gather and record information. Overall, therefore, the system being constructed will increasingly 'crim-check' children whenever they come into contact with authority, agencies and services, not simply when they leave one school for another.
The policy review would have us believe that sound science underpins its proposals. "Early intervention can be highly effective in preventing future crime", it tells us, and "individuals can move in and out of risk." So far, so uncontroversial. "However, by using intelligence on risk factors (such as conduct disorder or living in very low income families), high risk individuals can be identified early and specific, tailored interventions used. The availability of this kind of intelligence is increasing, meaning that a more systematic approach can be taken both in identifying which interventions work best and in applying them."
If you thought you could spot a transition from stating the bleeding obvious to voodoo social science there, you would not be entirely wrong. "There is strong evidence that," the review adds, "when targeted effectively, early intervention and prevention can have a significant impact and be cost-effective... Agencies are increasingly able to identify those with significant problems and at risk of future offending early in their life."
This increased ability on the part of Government agencies is at the very least debatable - what is it that they are finding out now that they didn't know already? And some of the evidence cited by the review is perhaps a little less than earth-shattering. It points to US studies and initiatives, including Nurse-Family Partnership and Perry Preschool, and to an earlier Strategy Unit document*, Predicting adult life outcomes from earlier signals, by Leon Feinstein and Ricardo Sabates.
The general data from the first two is however not particularly controversial. Children who start disadvantaged through poverty, single and/or inadequate parenthood, in families with a history of crime, drug abuse, tend to do worse in later life, and tend to be more likely themselves to continue the cycle of crime and disadvantage. By giving such families more and better targeted support early on, however, it is possible to reduce the incidence of criminality and disadvantage in the child's later life. That's essentially what the Nurse-Family and Perry Preschool data tells us, and frankly one really does not need to go all the way to the United States in order to obtain this kind of information. The party currently in Government believes as an article of faith (and in the shape of a well known slogan, "tough on the causes of crime") that deprivation, crime and disadvantage breed deprivation, crime and disadvantage, and if the state itself did not at least to some extent agree, then we would not feel the need for social workers, check?
So maybe there's less to all this than meets the eye. Social services, probation officers, health services all focus on these areas already, and all are at least intended to make an effective contribution to breaking the cycle.
But there are a couple of problems here. The review presents predictable and unremarkable study data as a new and increasing body of evidence, and rather than proposing better support for the core systems and organisations dealing with the problems, it favours more in the way of "targeted crime prevention programmes such as Youth Inclusion and Support Panels (YISPs) and Youth Inclusion Programmes (YIPs)," adding that "there is scope to go even further by intervening earlier in a child's life." Meaning another series of organisations and programmes to add to the throng of agencies and professionals already crowding around the hapless child? Probably.
Feinstein and Sabates highlight a second problem: "In our view," it says, "it would be irresponsible and socially and economically inefficient to ignore this very high level of capacity to identify early on those at risk of high cost, high harm outcomes." And, while it notes that there are issues "about the ethics and legality of access to data and about data linking and about the use of the data in the targeting of interventions... information is regularly gathered in schools, doctors' surgeries and elsewhere that in fact might be considerably more predictive for adult outcomes than that collected in the datasets investigated in this research study. Moreover, teachers, social workers and other practitioners routinely form assessments and perceptions of children that can be remarkably accurate about their level of risk."
And it's Feinstein and Sabates report that favours monitoring from birth: "It seems likely, therefore, that the most useful framework for developmental measurement and assessment would start from birth with indicators of childhood health and development, together with measurement of family income, education, parenting skill and social ties to the neighbourhood or in terms of wider social and familial networks."
There are a few trade-offs to be considered here. Professionals who're dealing with a child on a regular basis should quite clearly be in a position to form accurate assessments of them, and should be able to some extent to make a positive intervention where necessary. There may be some benefit to be gained by making such assessments more widely available among other professionals involved with the child, or who could become involved with the child, but on the other hand, by spreading the net wider one might find that the assessments became more cautious, less informed, and their quality degraded. And scoring systems, although they can be useful tools, can also contribute to this degradation if in deployment they effectively deskill assessments through automation.
Nor is it inevitably the case that the more organisations and systems involved, the better the outcome. Organisations peripheral to the case of a particular child will nevertheless, when presented with a statutory box to check, check it, in part for reasons of arse-covering and in part because that's what organisations do. Clouds of doubtful and marginal data static will build up around the individual, possibly resulting in more, but less effective, intervention, with outcomes less favourable (or at the very least, less cost-effective) than the ones you started out with.
Feinstein and Sabates favour "a system of risk monitoring at which certain levels of risk would trigger greater monitoring and assessment, and ultimately, if judged appropriate, intervention. This is the same process as is followed in relation to medical practice." Well, up to a point... They also note that "it may not be appropriate to respond to these signals of risk... if there were no appropriate interventions." In fairness, we should be clear that they are alive to the importance of the quality of the assessment and the validity of the intervention. But will the system be? One might doubt whether an agency, presented with an amber or red signal on a child's file, would be prepared to do nothing on the basis that there was nothing worthwhile that could be done, and risk accusations of negligence a couple of years down the line. Once the data exists, it will multiply, and it will produce pressure for more and more intervention, whether or not it is appropriate.
Other effects of the drive for data collection and matching have also been noted. Individuals and families who might be viewed as being at risk have begun to withhold information from care workers, on the basis that they know that such information will be passed on, resulting in multiple interventions from multiple agencies. Mention a slight problem and - depending on who you are - you get at least one social worker. And related to this effect we have a tendency for intervention to cluster around easier and more easily identifiable cases and 'issues', while those most in need are less likely to get support, partially because they're difficult to identify, and partially because they know who they are, fear (with some justification) the consequences, and therefore take steps to avoid identification. As the Government pushes for more data and more sharing in this and other areas, we're likely to see this effect magnify. ®
* Leon Feinstein, co-author with Ricardo Sabates, is director of the Centre for Research on the Wider Benefits of Learning, and keeps a fascinating CV there. His PhD is in economics, and he lists an impressive series of research funding wins, largely in the field of education. The DfES contributed in excess of £1.3 million between 2001 and 2004, while funding has also been forthcoming from the Treasury, the Cabinet Office and the Smith Institute, which is one of those think tanks the media satirically refers to as "left-leaning". We're jolly impressed, doubly so considering that we're the sort of shifty characters who could never bring ourselves to be so commendably open about money.
Sponsored: DevOps and continuous delivery