« A Good Year for Environmentalists | Main | McIntyre's Bulldog or The Hockey Stick Illusion by A W Montford - A Review »

Lord Stern's 1.3% Error or Typo?

Roger Pielke Jr. points out that the online version of The Stern Report has quietly been changed and an erroneous 1.3% of GDP costs that hurricanes may cost the USA changed to a more believable 0.13%.
He says " now the Stern Review math does not add up". So it wasn't a mere typo that was corrected, but a figure that was relied on in calculations.

Eli Rabbett predictably says it was just a typo and the correct figure was used everywhere else in the paper.

Tim Worstall, no great friend of Lord Stern, channels Professor Tol, a knowledgeable critic of The Stern Report to agree it is just a typo.

It would seem it should be easy to work out the truth, Stern puts up some numbers, and then summarises the total effect. I'm not familiar enough with the report to easily work my way through it but I've tried. As far as I can see it is a bloody mish-mash of figures and claims all constrained by what-ifs. A key paragraph is this:

Costs of extreme weather alone could reach 0.5 - 1% of world GDP by the middle of the century, and will keep rising as the world warms.
• Damage from hurricanes and typhoons will increase substantially from even small increases in storm severity, because they scale as the cube of windspeed or more. A 5 – 10% increase in hurricane windspeed is predicted to approximately double annual damages, resulting in total losses of 0.13% of GDP each year on average in the USA alone.
• The costs of flooding in Europe are likely to increase, unless flood management is strengthened in line with the rising risk. In the UK, annual flood losses could increase from around 0.1% of GDP today to 0.2 – 0.4% of GDP once global temperature increases reach 3 to 4°C.

So world GDP damage goes up to 0.5 to 1%, on those figures Europe's damage will be well below average at 0.2 to 0.4% So somewhere else in the world must have a much higher than average damage to adjust the average up, considering the percentage of world GDP that Europe contributes. But it can't be the USA if the USA only weighs in at 0.13% for hurricanes (other natural disasters may also happen but are small beer in this account).

So where then is the section of the world economy that will be devastated to get his average up?
The EU and the USA make up over half of the world GDP so it would need to be large.
Stern seems to be silent on that point.
Of course if he had used the 1.3% for the USA then his maths easily adds up.

(The other frightening possibility is he made the same mistake that I nearly did when I was looking at it half asleep, adding up the 0.4 from the EU plus 0.13 from the USA to make 0.53....)

So my feeling is that the wrong figure was used and now corrected the report doesn't add up, but if you can prove it either way please tell us.


I think your basic logic is fine although he is talking about all extreme weather events for the global figure whereas the other figures are for specific events. Europe (the 0.4 is for the UK, 0.02% is the EU figure) would have to include other events such as heatwaves. The 2003 heatwave is said to have cost farming, livestock and forestry $15 billion, about 0.1% of EU GDP according to Box 5.4. It is still difficult to see how he could come up with 0.5 to 1% without the US having a figure greater than 1.0%. However, the real problem is that he does not (which in my opinion is very shoddy) show his calculations. I'd have expected to see a table listing every event from every developed country, multiplied by the countries GDP, all totted up and then given as a % of total global GDP. Without these workings, it looks like the 0.5 to 1% was a thumb suck. Would it be possible to get Stern's workings under FOI?

Another typo would appear to be that in Table 5.2, the US cost of hurricanes was listed as 0.06% for 2005 (it is stated that "Numbers in brackets show the costs in 2005") whereas in Box 5.4, Hurricane Katrina is said to have cost 1.2% of US GDP.

As you link to, I asked someone who actually knows. It's a typo, nothing more.

There are all sorts of things wrong with Stern but this particular one isn't one of them. The sums use the correct number, the table really is just a typo.

For example, increased hurricane damage isn't because hurricanes are becoming stronger or more frequent. It's because a richer society builds more things in the path of hurricanes (this is the explanation for the rise in hurricane damage over past decades).....houses on outlying islands etc.

So, Tim, where does the 0.5 to 1% figure come from? Where can we see the sums?

Post a comment