But what if that core assumption is far less robust than advertised?
My answer came in an analysis published last month in Science of Climate Change that used my background as a finance academic the way a financial analyst would interrogate a market hypothesis. In finance, no matter how elegant a theory looks, you still test it against hard data. If the numbers don’t confirm the model, the model gives way, not the other way around..
The starting point is uncontroversial: since the 19th Century both CO2 and global temperatures have risen. The correlation is visually potent. But anyone who has worked with time series knows how deceptive such correlations can be. Ice-cream sales and shark attacks rise together in summer; that doesn’t make one the cause of the other. Two trending variables will often appear tightly linked even when the relationship is entirely coincidental.
Econometricians therefore strip out time trends and examine how annual changes relate to each other. Does each year’s increase in CO2 reliably produce a corresponding nudge in temperature? When we look at the data since reliable measurements began around 1960, the answer is awkward. CO2 has risen sharply, yet the rate of temperature change has not moved in lockstep. In fact, the annual changes diverge. If CO2 were the master control knob, you would expect accelerations in CO2 to march closely with accelerations in temperature. They don’t.
I stress-tested this across multiple temperature and CO2 datasets. The result was the same: the neat, linear linkage between CO2 levels and temperature weakens once you look beneath the headline trend. The famous correlation appears, at least in part, to be a statistical mirage...<<<Read More>>>...
