Joan Smith: So how many really died?

These hyperbolic claims should be read out on every media studies course as a cautionary tale
Click to follow

On Friday evening, after searching the city for bodies, the authorities in New Orleans started damping down estimates of the number of casualties attributable to Hurricane Katrina. A couple of days earlier, the official death toll in Louisiana was still 71, with another 170 in Mississippi, and the suggestion from the mayor that as many as 10,000 people had lost their lives in his city alone had begun to assume the status of holy writ. It was reported that 25,000 body bags were being flown to the Gulf coast, yet it should have been clear that if Katrina followed the pattern of some recent disasters, notably the suicide attacks on the East Coast of the United States four years ago today, such estimates could well prove wildly inaccurate.

As it happens, 10,000 was one of the most widely quoted figures for fatalities in the immediate aftermath of 9/11, and it was used by both the Daily Mirror and The Times. The Guardian showed more caution, suggesting that the death toll "could run to many thousands" but a Scottish newspaper, the Evening Times, speculated that fatalities might reach 20,000. Within a couple of days the former defence secretary Michael Portillo was invoking comparisons with the Vietnam War, in which nearly 60,000 Americans died; the death toll from the destruction of the World Trade Center alone was "in the same league as major acts of war that have scarred nations for generations", he asserted in The Times. In the Mirror, Tony Parsons agreed, claiming that 9/11 "was like the entire trauma and tragedy of the Vietnam War crammed into a few horrific hours".

These hyperbolic claims should, I think, be read out to students on every media studies course as a cautionary tale. Dreadful as the terrorist attacks were, they did not kill 10,000 people, nor even the 6,275 estimated by the New York authorities shortly afterwards. Two years ago, the city's official death toll was reduced to 2,752, a figure that includes 127 passengers and 20 crew on the two hijacked planes but excludes the hijackers. Adding in the most recent figures I can find for the attack on the Pentagon and the fourth plane which crashed in Pennsylvania brings total 9/11 fatalities to just under 3,000.

This is bad enough, but certainly not on the apocalyptic scale feared immediately after the atrocities. Why didn't the authorities and journalists learn from that experience? My suspicion is that casualty-inflation has two causes: one is the human tendency to rely on anecdotes and impressions. Last December's tsunami in South-east Asia was rare among disasters in that initial estimates hugely underestimated the number of casualties; more often, when catastrophes affect parts of the world we know or empathise with - not immediately the case, I'm afraid, with somewhere like Banda Aceh - dramatic TV pictures and newspaper photographs make it impossible to get a realistic sense of scale. Additionally, gossip and hearsay are sometimes presented as fact; I have little sense of whether all the rapes and murders reported in the New Orleans Superdome took place or were just a fantasy about the wholesale breakdown of civilised values in the world's most powerful nation.

This brings me to the other factor at work here, which is the way in which political parties perceive the likely effect of a disaster's impact on their fortunes. After the 9/11 attacks, the metaphor almost everyone reached for was war, and we expect wars to involve large numbers of casualties; indeed the higher the death toll, the more justification the Bush administration could find for launching a couple of real wars, with the tragic consequences we see daily in Iraq.

Four years on, that situation was more or less exactly reversed, with the prospect of thousands of fatalities from Hurricane Katrina having the potential to inflict serious damage on what is widely perceived as an inept and uncaring administration. This factor should be borne in mind as see-sawing fatality figures emerge from Louisiana and Mississippi; it is not just real deaths that are at stake here, but the metaphorical death of a divisive and widely despised president.