Criminal Law and the Science of Prediction

It was a devastating earthquake on many levels: loss of life as over three hundred people died, loss of property as buildings crumbled, and loss of history as an ancient medieval fortress town came tumbling down. All it took was seconds as the earth shook on April 6,2009 in the tiny hilltop village of L’Aquila, Italy. Yet, behind the disaster were years of earthquake readiness and scientific predictions. Behind the devastation was months of tremors; warning signs that something was not quite right. Yet, disasters happen and this one certainly did.

But was there someone to blame? The government thought so when they charged six scientists and one government official with manslaughter. Finally, after a lengthy trial, on October 22, 2012, Judge Marco Billi found them guilty of manslaughter and sentenced them to six years imprisonment. Those convicted included the head of the Serious Risks Commission, the Director of the National Earthquake Centre, and a physicist. The prosecutor built a case of mismanagement, inaccuracy, and the withholding of crucial information, which could have saved lives. The defence emphasized the unpredictability of earthquake prediction and the “chill effect” such a verdict would bring. Indeed after the verdict, a number of Italian scientists quit government posts as scientists across the globe warned of the harmful effects the verdict would have on future research efforts.

There are clearly two sides to the issue but, when looked at more closely, the two sides are not in opposition. The prosecutor is correct in characterizing erroneous or even untimely information as an act or omission, which can and should ground a criminal charge. In Canada, manslaughter requires an underlying unlawful act, which can be viewed as objectively dangerous in nature, which then causes the death of a human being.  Here, the allegation the scientists knowingly mislead the people of L’Aquila, may be a basis for such a manslaughter charge.

However, a mere failure to accurately predict a major disaster is not, in itself, a basis for a manslaughter charge. If it were, the defence’s concern that such prosecution would curtail scientific innovation would be correct. Such a prosecution would be the antithesis of the scientific method. A scientific hypothesis must be first tested before accepted. If the experiment does not produce the results expected, then the hypothesis is modified: such trial and error is needed to produce a final result. Without the ability to make mistakes with impunity, many medical treatments would not be developed.

There is also the difficulty with predictions, whether under the rubric of science or not as seen by the successes and failures of polls and the pollsters who interpret them. As Nate Silver, predictor extraordinaire, explains in his new book "The Signal and the Noise," although some outcomes are predictable if we crunch enough numbers and gather enough data, some, like earthquake prediction, is not possible at this time. How then can we expect a standard of behaviour in an area that is impossibly non-standard?

There is also, a “half-way” opinion between the two: that certain failures or breaches of rules should stay in the regulatory arena and sanctioning should be through the controlling regulatory scheme as opposed to the criminal law.

All of the above is, of course based, on the facts and the level of liability would be commensurate with those facts. In the Italian case, if the findings of the trial Judge were as submitted by the prosecutor, then not unlike the Walkerton, Ontario tainted water incident, the criminal law is properly engaged. However this case is viewed the decision has caused much debate. To sample the various viewpoints, read this, this, this, and this. To read more about public disasters that attract the intervention of criminal law, please read my posting from January 15, 2012 on Public Disasters and the Criminal Law and my posting from February 25, 2012 entitled Safety First: Laboratory Safety and the Criminal Code.