Beware neo-optimists (and pessimists) using "TED-Stats" to sell you a story. It starts with complicated but pretty graphics that combine 4 different things via multiple confusing axis scales mixed in with colour and different sized spots. Then notice that the graphics are only available as part of a youtube video of a powerpoint presentation with no citations or links to the underlying data. Dig deeper and you'll find the following common statistical strategies.   

1) Excessive accuracy. For instance total global population to an accuracy of 1000 in 7,000,000,000. Or total CO2 production against global GDP when this is hugely dependent on figures from China and India. One example is the de-coupling carbonisation story that requires a pause in CO2 production during a continued rise in GDP. Except that both figures just got restated for the last 10 years by 10 times the small change you're looking for. And with China the figures are coming out of a hierarchy with a huge incentive to lie about them. And as Eris knows, truth in communication is very hard in strongly hierarchical systems where self-serving lies are the norm.

2) Mistaking percentages for absolute figures. Total global population has been following linear growth for 4 decades with each successive billion taking about 12-13 years to add. If anything the time to add a billion is shortening from 15 to 12 years. But if you measure the growth as a percentage of the total, of course the percentage growth is dropping. You can sell this story as "Population growth has been dropping for 4 decades". Then extrapolate this drop in percentage out as a straight line and you can argue that it will drop to zero and so population will peak. Except that the linear growth is still happening and we're still adding a billion every 12 years. Even the most reliable source of figures, the UN, can simultaneously predict a drop in population growth due to fertility rates and also adjust the date we get to 9b closer by 6 years. So which is it; is growth slowing down or is it speeding up?

There's a similar kind of statistical lying in the story about CO2 production flat-lining. So the total yearly emissions in GtC/y didn't increase so the year on year rate changed by 0% or went down by 0.1%. Great. Except that yearly emissions were already at a record high and will be pretty much the same next year.

3) Focussing on detail and sub-sections instead of the holistic big picture. The classic example here is to argue that because the USA and developed world reduced their onshore fossil fuel use or pollution or CO2 output then things are getting better. If everybody could do this, then we'd solve the climate change problem. But this ignores the global picture that the developed world reduced it's emissions by outsourcing all its manufacturing to the developing world. And we outsourced all our pollution as well. Not only that but the shipping involved is a major polluter as well.

Now add all this together and attempt to argue that things are getting better because the numbers of people in extreme poverty are reducing rapidly due to capitalist globalisation. See, the percentage of people in extreme poverty is coming down year on year! If we keep this up we can wipe out extreme poverty by 2030. Add in a lot of hand-waving arguments about average incomes, education, fertility rates, urbanisation to justify that. But, wait. At the same time as the percentage was dropping we were adding 1b every 12 years. How come there were 1000m in extreme poverty in 1820 and still 1000m in 2010? And it's essentially the same people in subsistence farming in the poorest countries that are classified as in extreme poverty. Didn't we achieve anything in the last 200 years of progress? Clearly mankind has achieved an astonishing amount in the last 200 years and if nothing else we've added 5.5b people who are NOT IN EXTREME POVERTY. That's a cause for optimism as long as you don't think too hard about the implications of continuing to add more non-poor people to the mix indefinitely. But what we haven't done is to reduce the absolute number who are in poverty. Except maybe we have. Look closely and it's possible that the absolute numbers have been dropping a little this century. But be extremely wary of the accuracy of that reporting.

It makes you wonder. Does anyone know WTF is going on and can describe it accurately without bias and an agenda?

Bonus links

[from: Google+ Posts]

[ << In preparation for The Crisis, read this ] [ Why is David Cameron's "Greenest Government Ever" turning away from investment in renewable infrastructure... >> ]
[ 23-Jan-16 10:54am ]