The Intelligence Dilemma

22/08/2013  -  10.02

Professor Peter Waddington

The Intelligence Dilemma

The appalling murder of Daniel Pelka by his mother, Magdelena Luczak, and her partner, Mariusz Krezolek, has yet again been followed by soul-searching and a storm of criticism directed at ‘the authorities’ for their failure to protect Daniel from the child abuse that eventually led to his death. Of course, we must wait for the conclusions of the ‘serious case review’, which doubtless will yet again find failures of communication between agencies, and indecision, prevarication, and hesitation to take action on the part of at least some of their personnel.

Nothing new there, then?

Daniel has now taken his tragic place in a very long line of children abused and eventually killed by their parents or parent surrogates. This line stretches back to 1973 when the murder of Maria Colwell by her stepfather led to an inquiry amid popular indignation. The inquiry found — you guessed it — lack of communication between the agencies who were well aware of her vulnerability. Why does nothing change?

For an answer, perhaps the public should look at itself, for public opinion is nothing if not fickle. In the days and weeks before Daniel’s murder grabbed the headlines, there was another cause célèbre consuming acres of newsprint and hours of mass media programming.

It was the revelation by Edward Snowden (the CIA ‘whistle-blower’) that GCHQ was scanning and retaining the meta-data of all and contents of some electronic communications that left our shores. To say that I wasn’t at all surprised by this is an understatement; I was flabbergasted that anyone was surprised.

What do these two scandals have in common? The answer is that they pose the ‘intelligence dilemma’ from either end of the telescope. The fate of Daniel Pelka prompts us to peer more sceptically beneath the surface appearance of family life, whereas Edward Snowden warns us against being too intrusive.

Intelligence is not merely the gathering of information, but more importantly its assemblage into a coherent picture. As the former head of the domestic intelligence service (MI5) Eliza Manningham–Buller put it in her BBC Reith Lectures, it is a question of ‘joining the dots’.

Often it is the case with abused and murdered children that there was an abundance of dots, it is just that they were not joined together correctly. Time and again we find that different agencies knew snippets of information that were not shared and hence no picture emerged with sufficient clarity to take action.

What would the assembling of these dots require? It would require that agencies with very different purposes, structures, and values should share information with each other. Let us consider perhaps the most sensitive information: medical histories. Some of these children had been treated by medical staff, usually in Accident and Emergency. Should those medics share confidential information about those children with other state agencies, such as the police? If so, should the police be told everything, or only some things?

If only some things, then what? Also, what should the police do with that information because many children pass through A&E on their way to maturity? For the injuries young children sustain to be reported to the police would amount to the collection of vast quantities of medical information about not only those at genuine risk, but also about those at little or no risk.

If we demand, in the wake of some horror, that ‘someone must have known what was happening and should have shared those concerns,’ then we are demanding that the state, through its various agencies, should know almost everything about all of us — a ‘Snowdenian’ nightmare. Such a nightmare is not a fanciful possibility. In another case of child murder — the Soham murders of Holly Wells and Jessica Chapman at the hands of Ian Huntley — the inevitable inquiry, chaired by Sir Michael Bichard, concluded that in order to mitigate risk the police and other agencies needed to pool what they knew about those who came to their attention in a huge database.

In due course this recommendation blossomed into the Independent Safeguarding Authority, which decreed that all those working, in whatever capacity, with children and vulnerable adults must be thoroughly vetted. This created a furore. Leading authors who read stories to young children very publicly refused to be vetted and thereby faced exclusion from such activities.

Volunteers who drove the occasional church outing in a mini-bus found that they too must be vetted. In 2012 the Independent Safeguarding Authority was dissolved and the demands for such stringent vetting ceased and state remained, if not blind, then myopic when it came to the risks being posed to children and vulnerable adults.

This exposes the irony of intelligence: it is best done in retrospect. It’s obvious, isn’t it, that a youth who has sexual relations with girls just under the age of consent and is falsely accused of rape is likely to murder, ten years later, two innocent young girls who he hardly knew? Surveys of teenage girls suggest that around half of them are or have been in sexual relationships, some of which are likely to be with lads a little over the age of consent.

To prevent Huntley murdering Holly and Jessica on these grounds would probably disqualify thousands of men from becoming school caretakers, or come to that doing any of the jobs in which they come into contact with young girls. Retrospection is far easier than prediction and we have very little idea of what biographical indicators predict any form of criminality.

Suppose we did find that there were such tell-tale signs, would resolve the intelligence dilemma? No, it would not, for two reasons. First, such tell-tale signs would be accompanied by as much, if not more, error than contemporary weather forecasts.

Negative errors would result in failing to distinguish the occasional individual who did not display such tell-tale signs and yet posed a danger to children. Positive errors would mean wrongly identifying some people as dangerous when they were not. Both would provoke public scandal and we would be little better off.

The second reason why the discovery of tell-tale signs would not resolve the intelligence dilemma brings us back to Mr Snowden and GCHQ. In order to identify those who exhibit the tell-tale signs it is necessary to scan everyone at some level. The reason that Amazon knows what DVDs I like better than I do, is not only because thousands of people make millions of purchases, from which Amazon’s data–miners are able to extract enduring preferences.

It is also because, armed with that information, they are able to distinguish my online purchasing habits from those of my wife. Unless the veil of privacy is raised at least a little on everyone, then we can never have sufficient information to ‘join the dots’ with any accuracy.

Perhaps we need to recognise reluctantly that there are many things in life that cannot be predicted and that we would not want to be predicted. All we can do is to express our outrage at those who perpetrate such horrors.