Monday, October 30, 2017

Getting it wrong

Source


When one is in the business of analysing data and making forecasts, it's easy to make mistakes.  So one should be careful not to mock the forecasting errors others make.

In the investment world one must be intellectually honest.  One must dig out the facts, verify them, check that any models (mental or computer) are logical and internally consistent, and always be ready to admit where one went wrong and why.  If you repeatedly make the same mistake you deserve ridicule.  As the IEA does.

Bjørn Lomborg's thesis has been that there are more important things poor countries could do with their money than try to prevent climate change.  But that led him to advocating the denialist meme after 1998 (an El Niño year, when temperatures reached a peak) that global temperatures--and therefore sea levels--were no longer rising.  He became the darling of the Denialist Right.  One wonders whether this forecast (above) was biased  by his advocacy,  That does seem likely, as he is a trained statistician, and would know all about trend estimation and the statistical problems posed by random variation in an underlying trend.  It's all too easy to be persuaded that what we want to happen will in fact happen.  Alas, life's not like that.

His argument that it might be better to spend  the money earmarked to prevent climate change on other life-enhancing things instead was sort of plausible when renewables were vastly more expensive than fossil fuels (ignoring their impact on global warming).  But that's not true any more.  In fact, it's the opposite: renewables will raise living standards and increase economic growth by cutting the cost of power.  Even if you ignore the benefits for growth and quality of life which would occur if pollution were stopped.  Recently Lomborg appears to have shifted his position.  For example, he now argues that subsidies to fossil fuels should be removed.

I'm sure you could go through this blog and find many examples where my forecasts have been wrong.  It goes with the territory.  But it is important to examine your own biases to see if they are skewing your own forecasts.  I would like to believe I am not, but perhaps I am.  I shall try to be more rigorous in future.

No comments:

Post a Comment