Italian Seismologists Charged with Manslaughter for Not Predicting Earthquake

Photo: RubyGoes

In our latest Freakonomics Radio podcast, “The Folly of Prediction,” we talk about the incentive structure behind making predictions. Wrong predictions almost always go unpunished, which is why people make so many of them. (You can download/subscribe at iTunes, get the RSS feed, or read the transcript here.)

But recent news out of Italy seems to take the premise of punishing bad predictions a bit too far. From the New York Times:

Seven Italian seismologists and scientists went on trial on manslaughter charges on Tuesday, accused of not adequately warning residents of a central Italian region before an earthquake that killed 309 people in April 2009. Prosecutors say that the seven defendants, members of a national panel that assesses major risks, played down the risk of a major earthquake’s occurring even though there had been significant seismic activity near L’Aquila, the capital of the Abruzzo region, in the months before the quake. The case has drawn the attention of the international scientific community, which argues that it is impossible to predict an earthquake. Prosecutors, in actuality, charge that the panel did not fulfill its mandate and instead conveyed “incomplete, imprecise and contradictory information.”

In an op-ed for NewScientist, Thomas H. Jordan, director of the Southern California Earthquake Center, breaks down the dilemma the seismologists faced in trying to make a judgment call.

Did the scientists do anything wrong? The facts of the case are complex. Seismic activity in the L’Aquila area increased in January 2009, prompting school evacuations and other preparedness measures. Media coverage was inflamed by a series of earthquake predictions issued by Gioacchino Giuliani, a local man who worked as technician in a physics laboratory. These predictions had no official validity but were widely reported. At least two of the predictions were false alarms. No evidence indicates that Giuliani transmitted to the public or any civil authority a valid prediction of the main shock.

The Italian scientists were trapped by a simple yes-or-no question: “Will we be hit by a damaging earthquake?” This was not surprising given Giuliani’s alarms, but it was not one they could answer conclusively. From what they knew a week before the earthquake, a big shock was not very likely: the probability of a false alarm (if an alarm were raised) exceeded the probability of a failure-to-predict (if an alarm were not cast) by a factor of more than 100. Even so, seismic activity had increased the probability of a large earthquake by a significant factor, perhaps as much as 100-fold, above the long-term average.

Distracted by Giuliani’s predictions, the authorities did not emphasise this increase in hazard, nor did they focus on advising the people of L’Aquila about preparatory measures warranted by the seismic crisis. Instead, they made reassuring statements that were widely interpreted to be categorical.

(HT: Katherine Wells)

Leave A Comment

Comments are moderated and generally will be posted if they are on-topic and not abusive.

 

COMMENTS: 3

View All Comments »
  1. Rick says:

    For the people of Italy, I would like to predict that wherever you are, at some point in the next billion years, there will be a major earthquake there. The resulting damage will be catastrophic. Whether this will occur in the next five minutes, or sometime after the human race has wiped itself out, I cannot confirm.
    For your safety, I recommend you take precautionary action (if you choose to move, but fail to realise that this prediction also applies to wherever you move to, that’s your own damned fault).

    Well-loved. Like or Dislike: Thumb up 12 Thumb down 2
  2. Artemis says:

    I’m disappointed that you are actively buying into this as well. The charge is *not* that they failed to predict. They went well beyond failing to predict all the way to advising the populace against standard precautionary measures. It’s one thing to fail to predict an earthquake. It’s quite another to actively tell people to do the wrong thing.

    The scientists are being charged because they knew there was a potential for an earthquake and they intentionally advised to populace to not worry about it. They are being charged because what they did was worse than nothing. If they had done nothing, then the populace would all have done what they always do and went out to an open area. Instead, they stayed indoors, *because the government experts told them to do it*.

    Thumb up 8 Thumb down 8
  3. MW says:

    I’ll try some approximate calculations of the risk.

    Wikipedia says the city had pop. 73000, but about 100,000 during the day. About 300 people died. Lets call this 0.3% chance of death for being in the quake. (Big approximation here – that the actual quake was an ‘average’ size for a damaging quake.)

    Wikipedia lists 6 other significant quakes since 1300, so about one per hundred years.
    http://www.earth-prints.org/bitstream/2122/5570/1/BEE-Crowley_etal_.pdf has a huge variety of seismic hazard maps for Italy, which give a roughly 0.1% chance per year that a randomly chosen L’Aquila building will collapse due to earthquake. These seem vaguely in agreement.

    So, if the hazard were 100 times the background, is it worth sleeping outside? Assuming outside is safe and you sleep 8 hours out of 24, this reduces your risk of death in the quake from 0.3% to 0.2%, or 1 in 1000, if the quake occurs. If the short term rate were 100 times background, and background is once per 100 years, the risk of a quake on a given night is about 1/300. So sleeping outside has approximately 1/300,000 chance of saving your life, per night. If you have 50 years remaining life expectancy, that is about 17500 days, so your expected increase in life expectancy by sleeping outside is about 0.06 days, or 1.4 hours.

    So, under these conditions, is it worthwhile sleeping outside? When I started that calculation, I was hoping it would come out with an answer like change in life expectancy of 2 minutes, or 10 days. Then I could say “I might be out by a factor of 10 because of my approximations, but with a number like that, the answer is still clear.” Unfortunately, I’ve got a middling answer, so getting higher quality data could change the answer one way the other. (I’ve also only considered the chance of death, not of injury.)

    Thumb up 4 Thumb down 0