I have written around this topic many times before but have never been very happy with the results. Unlike gsgs I am not particularly bothered about experts not putting probabilities or numbers like deaths, economic losses etc. against a potential pandemic but I am very unhappy about what I see as a potentially catastrophic failure to understand the relationship between severity and its effects on the economy.
<o
></o
>
There are several factors that need to be considered and I will try and talk a little about each and their influences. I have been goaded into to writing now by two of today’s posts.
Lloyd's Report - Pandemic - Potential Insurance Impacts http://www.flutrackers.com/forum/showthread.php?t=83249
&
Flu Pandemic May Cost World Economy Up to $3 Trillion http://www.flutrackers.com/forum/showthread.php?t=83211
In 2005 the World Bank made an $800 billion estimate and now they pick $3 trillion. Lloyds go for a 1 to 10% drop in Global <st1:stockticker>GDP</st1:stockticker> given a 1918 style pandemic (Global <st1:stockticker>GDP</st1:stockticker> ~ $50trillion so ~ half a trillion per percent). I would argue this is like estimating a figure for medical costs given only that the patient has been shot with no indication of where the patient was hit or by what.
<o
></o
>
The first point is estimating the range of severities of a pandemic. Again I have posted on this several times before but the main points are
* We don’t know what will cause the next pandemic. With over 6 billion people spreading into new areas the scope for zoonotic emergence has never been higher and HIV, SARS and the ongoing situation in South Africa have given us fair warning.
* H5N1 is still our most immediate threat and we only have reasonable data on 3 past flu pandemics which is in not a large enough data set to draw any useful conclusions about the range of symptoms flu could present. Worst case scenarios based on 1918 could be little more than wishful thinking.
* H5N1 - unlike all previous human flus (as far as we know) - is Highly Pathogenic (HP) and can therefore cause systemic – rather than localised – infections which further reduces the usefulness of comparisons with previous seasonal or pandemic flus.
<o
></o
>
The previous section concerns the futility of trying to say anything too dogmatic about the range of events which planners may need to accommodate. Now we need to concern ourselves with the consequences of different types of pandemic.
It is my opinion that the relationship between severity and public reaction (or more accurately public fear) and so to economic damage looks linear initially and then turns exponential. To try and explain lets look at three scenarios.
1] If a pandemic starts and it seems like an ordinary flu season, or a little more severe (1957/1968), then economic damage would be in the form of days lost, like any other flu season, based on the public going to work as usual. Changed behaviour would be limited to some social distancing hitting concerts, sporting events etc.
2] If the pandemic was like 1918 then many, if not most, individuals would personally know someone who had died or been close to it. The fear of lost income would become secondary to the fear of becoming infected, or worse, bring the infection home to your family. A sizeable portion of the population in the richer industrialised countries would switch modes to one in which the primary driver of all actions would be to avoid the infection of themselves and those they love. If available to them they will shelter in place (SIP), run up debt on essential items, ignore non essential bills and wait it out. Newly industrialised nations dependent on the production of manufactured goods will have no orders, subsistence farmers will – paradoxically - be least effected economically and the 60% of the population of Somalia, and all those others dependent on food aid, will starve and die in their millions.
3] If the pandemic is 1918+ the effects will be as above but cut deeper.
<o
></o
>
This quote from Jason Gale’s article
<o
></o
>
></o
>
illustrates the problem nicely. <st1:stockticker>IMO</st1:stockticker> it correctly identifies that the indirect consequences of the publics attempts to avoid infection will far outweigh the direct ‘days lost due to illness’ but then vastly underestimates the damage done to tourism and air travel “In the worst-case”. Even if 1918 were the worst case I just don’t believe 80% of people would not change their ‘tourism’ plans and airlines will be lucky if they keep 20% of their customers. <o
></o
>
<o
></o
>
<o
></o
>
<o
></o
>
Next I would like to look at the effects of the behavioural phase change predicated above.
The ongoing global economic instability conveniently illustrates the effects of ‘efficiency’ in the global market.
In the computer age we have become able to accurately calculate the probabilities of a wide range of events for which we have statistically significant quantities of data. This has had the effect of tightening margins in the insurance industries as all the competing companies know exactly what the probability of my being killed in a car crash given my age, job, address, insurance history etc. The only remaining variable is how much more they can get away with charging than their actuarial tables say they will pay out. Scale of operation, along with the reinsurance of risk, further reduces the dangers to the insurer of aberrations from the statistical norm. The system in extremely efficient but at a cost to its resilience. If claims exceed expectations then reinsurance spreads the load to an ever wider circle and the hit is absorbed the corollary is that if the hit is too large, and exceeds the planners’ models, then the reinsurers fall like dominos and the problems is uncontained and uncontainable. The same basic rules apply to the efficiencies of the Just-in-Time (JIT) economy. All the savings made in not holding dead stock are predicated on the global conveyer belt not breaking down and, as it never has before, it is assigned a low probability of failure in the future. On the other hand it has never had to deal with mass global absenteeism before; there is no other system capable of delivering food to the vastly increased, and newly urbanised, global population. The question is how much slack is there in this system and – like the sub-prime market – at what point does it cease to be self-righting.
<o
></o
>
As I am neither a biologist nor economist please treat the above as a personal opinion offered as a basis for discussion.<o
></o
>
<o
There are several factors that need to be considered and I will try and talk a little about each and their influences. I have been goaded into to writing now by two of today’s posts.
Lloyd's Report - Pandemic - Potential Insurance Impacts http://www.flutrackers.com/forum/showthread.php?t=83249
&
Flu Pandemic May Cost World Economy Up to $3 Trillion http://www.flutrackers.com/forum/showthread.php?t=83211
In 2005 the World Bank made an $800 billion estimate and now they pick $3 trillion. Lloyds go for a 1 to 10% drop in Global <st1:stockticker>GDP</st1:stockticker> given a 1918 style pandemic (Global <st1:stockticker>GDP</st1:stockticker> ~ $50trillion so ~ half a trillion per percent). I would argue this is like estimating a figure for medical costs given only that the patient has been shot with no indication of where the patient was hit or by what.
<o
The first point is estimating the range of severities of a pandemic. Again I have posted on this several times before but the main points are
* We don’t know what will cause the next pandemic. With over 6 billion people spreading into new areas the scope for zoonotic emergence has never been higher and HIV, SARS and the ongoing situation in South Africa have given us fair warning.
* H5N1 is still our most immediate threat and we only have reasonable data on 3 past flu pandemics which is in not a large enough data set to draw any useful conclusions about the range of symptoms flu could present. Worst case scenarios based on 1918 could be little more than wishful thinking.
* H5N1 - unlike all previous human flus (as far as we know) - is Highly Pathogenic (HP) and can therefore cause systemic – rather than localised – infections which further reduces the usefulness of comparisons with previous seasonal or pandemic flus.
<o
The previous section concerns the futility of trying to say anything too dogmatic about the range of events which planners may need to accommodate. Now we need to concern ourselves with the consequences of different types of pandemic.
It is my opinion that the relationship between severity and public reaction (or more accurately public fear) and so to economic damage looks linear initially and then turns exponential. To try and explain lets look at three scenarios.
1] If a pandemic starts and it seems like an ordinary flu season, or a little more severe (1957/1968), then economic damage would be in the form of days lost, like any other flu season, based on the public going to work as usual. Changed behaviour would be limited to some social distancing hitting concerts, sporting events etc.
2] If the pandemic was like 1918 then many, if not most, individuals would personally know someone who had died or been close to it. The fear of lost income would become secondary to the fear of becoming infected, or worse, bring the infection home to your family. A sizeable portion of the population in the richer industrialised countries would switch modes to one in which the primary driver of all actions would be to avoid the infection of themselves and those they love. If available to them they will shelter in place (SIP), run up debt on essential items, ignore non essential bills and wait it out. Newly industrialised nations dependent on the production of manufactured goods will have no orders, subsistence farmers will – paradoxically - be least effected economically and the 60% of the population of Somalia, and all those others dependent on food aid, will starve and die in their millions.
3] If the pandemic is 1918+ the effects will be as above but cut deeper.
<o
This quote from Jason Gale’s article
<o
<o``People's efforts to avoid infection are five times more important than mortality and more than twice as important as illness'' in terms of economic impact, the authors said. In the worst-case, they assumed that air travel would slump by 20 percent for the whole year, and that tourism, restaurant meals, and use of mass transportation would decline by the same amount.
illustrates the problem nicely. <st1:stockticker>IMO</st1:stockticker> it correctly identifies that the indirect consequences of the publics attempts to avoid infection will far outweigh the direct ‘days lost due to illness’ but then vastly underestimates the damage done to tourism and air travel “In the worst-case”. Even if 1918 were the worst case I just don’t believe 80% of people would not change their ‘tourism’ plans and airlines will be lucky if they keep 20% of their customers. <o
<o
<o
<o
Next I would like to look at the effects of the behavioural phase change predicated above.
The ongoing global economic instability conveniently illustrates the effects of ‘efficiency’ in the global market.
In the computer age we have become able to accurately calculate the probabilities of a wide range of events for which we have statistically significant quantities of data. This has had the effect of tightening margins in the insurance industries as all the competing companies know exactly what the probability of my being killed in a car crash given my age, job, address, insurance history etc. The only remaining variable is how much more they can get away with charging than their actuarial tables say they will pay out. Scale of operation, along with the reinsurance of risk, further reduces the dangers to the insurer of aberrations from the statistical norm. The system in extremely efficient but at a cost to its resilience. If claims exceed expectations then reinsurance spreads the load to an ever wider circle and the hit is absorbed the corollary is that if the hit is too large, and exceeds the planners’ models, then the reinsurers fall like dominos and the problems is uncontained and uncontainable. The same basic rules apply to the efficiencies of the Just-in-Time (JIT) economy. All the savings made in not holding dead stock are predicated on the global conveyer belt not breaking down and, as it never has before, it is assigned a low probability of failure in the future. On the other hand it has never had to deal with mass global absenteeism before; there is no other system capable of delivering food to the vastly increased, and newly urbanised, global population. The question is how much slack is there in this system and – like the sub-prime market – at what point does it cease to be self-righting.
<o
As I am neither a biologist nor economist please treat the above as a personal opinion offered as a basis for discussion.<o
Last edited by a moderator: