No country for the old, black saturday and ‘stay or go’


Bendigo Fires (Source: Richmeister)

What happens to the elderly when the fire comes?

The first report by the Victorian Police Task Force Phoenix to the Black Saturday Royal Commission provided a basic set of victim data. While the work of Operation Phoenix is ongoing, this data does allow us to do some quick statistical analyses and points towards a significant issue overlooked by the commission in it’s review of the ‘stay or go’ policy.

Taking the Phoenix data (Exhibit 49) and generating a histogram of the percentage of total deaths by age we get Fig. 1.

Fig 1: Percentage of total deaths by age

A quick analysis of the data gives us the following population statistics:

  • average age: 47.994
  • median age: 51, and
  • skew: -0.268 (the left tail is heavier)

OK so what does this all mean?

Well if we compare the frequency distribution of ages to that of the Victorian regional population (ABS 2008) we find that 16% of the Victorian regional population is 65 or older. From Fig. 1 fatalities, those over 60 made up 30% of the total fatalities.

So why would those over 60 be more represented among the dead?

In part this may come back to the noted reluctance of the elderly to leave in the face of a disaster (1). Such reluctance can be attributed to the tendency, cognitive psychologists call it the absurdity bias, to believe that, “I survived the last bushfire, this one will be no different.” (2). Having elected to stay the elderly are more vulnerable to extremes of temperature, smoke and stress.

If we compare the Phoenix data with that published for another major disaster where the age of casualties is available (in this case Hurricane Katrina) we get an even more interesting picture (3). Fig. 2 illustrates how the Katrina data is highly skewed in the positive direction when compared with Black Saturday.

In Katrina public policy was for voluntary evacuation, thus proportionally we would expect a cohort of elderly ‘stay behinders’ to assume a greater proportion of the total fatalities which the data certainly supports. The dark side of this effect is that if voluntary evacuation were made public policy for bush-fires the resultant death toll would become dominated by the elderly.

This of course presents public officials with a difficult problem should public evacuation become a policy. While one can encourage those who are willing, and able, to evacuate what do you do when people are unwilling, or possibly unable, to do so? Further if we are aware that the elderly are prone to such cognitive biases should we discount their decision in an emergency?

Disappointingly the Black Saturday royal commission interim report makes no mention of the effect of cognitive biases upon making a ‘stay or go’ decision, instead assuming that such decisions are made in a completely rational fashion. As Black Saturday and other disasters show this is rarely the case.


1. This is a robust effect noted in a number of disasters including Katrina and Three Mile Island.

2.  Actually there is more than one reason for this reluctance, see Black Saturday – The Not So Clever Country for a broader discussion of how decision-making heuristics can bias a ‘stay or go’ decision.

3.  Note that the Katrina data as published utilised bin sized that terminated at 9, 19, 29 which differ slightly from those used for the Black Saturday data.


Exhibit 49.  Royal Commission Interim Report, Chapter 6, Relocation, Exhibit 49 – Statement of P O’Halloran, Attachment 1 (WIT.3010.001.0007) at 0007, Commission website, 2009, accessed 23 May 2010.

ABS, 2008 statistical analysis of the Australian population, ABS website, 2008, accessed 23 May 2010.