Climate study has flown the coop

This is a follow-up post to “Maths, Economics, and pseudo-science” posted last week, to which Stickman has given a thorough response.  He no-doubt will also give us his thoughts on the second half of the post in due course, which we look forward to.

The point of this post is to further clarify our position on the scientific method and the search for truth in the field of economics.  We are using the example of climate science because a) it is illustrative of a broader problem of pseudo-science in intellectual endeavour, and b) because the proponents of anthropogenic global warming (AGW) propose a massive centralisation of global economic policy which will affect all of us economically.

Mathematics is highly valuable and has helped and will help us better understand the world around us and how to thrive in that world.  But maths is servant to the mathematician, not the other way around.  In economics, maths can be used to represent a logically derived theory in a very elegant and simple way.  But maths cannot be used as the starting point for a theory.  Even game theory which is supposedly maths-derived is really first built on a set of basic behavioural assumptions.  Mathematics in economics has been abused and caused the economic fraternity to arrive at false conclusions about the way the world works under the guise of ’science’.

The problem is that we can apply a mathematical method to many social and natural phenomena and produce results that have a quantitative veneer.  But the real question is: what are the legitimate fields of study in which to do this, and how much store can we place in the results.

Let me give you a very clear example: An economist, using a bona fide and scientifically accepted mathematical/statistical method, can produce an erroneous theory and yet state its validity with “95%” confidence.  The reason this can happen is because he might be using data that is simply well correlated but not at all causal.  The strong correlation may give rise to a reasonably good prediction model, and as a result, even though the causality might be non-existent or weak, the repeated predictive accuracy may give rise to an erroneous theory built on scientific method.

This theory, ’scientifically’ proven and peer-reviewed by fellow scientific economists, gains traction as hard economic theory, which then drives policy prescriptions or various other actions.  Acting upon this information runs the risk of grave error and ultimately economic folly and degeneration.  Specifically, we may end up trying to control one variable in the belief it has a causal impact on another, when in fact it is only correlated.  When it does not work we revisit the statistical models and conclude with 95% confidence that the scientific method must be correct.  So we conclude that we must simply persist with our policy action, or even increase its magnitude, until the expected result arrives.

But it doesn’t arrive.  And so we repeat our folly.  This is the textbook case of what is now happening with deficit spending.  Erroneous neo-Keynesian models say that deficit spending and money pumping will create growth by increasing aggregate demand to a level that soaks up the supply of goods, thereby creating employment.  When it does not work, as we are seeing in the developed world now, we hear loud calls from proponents of these methods that the deficit spending is too small and money printing too light.  Nobel Laureate Paul Krugman is one such voice.  Human Action wonders if history will correctly record him as the highly decorated vocal economist who helped drag the modern world into serfdom?

For Krugman and his ilk, the economy is like one big sausage machine.  Feed in your ingredients (people, capital, labour, land, ideas, technology), tweak the buttons and knobs and dials (interest rates, money printing, deficits, taxes, regulations) to the “just right” setting, monitor progress, tweak again half way through as required, and voila, out spits jobs, economic growth, prosperity and everyone’s happy.

Economics doesn’t work like that.  The system is complex and organic.  So complex in fact that tweaking certain policy levers at once sets in motion a set of events which not even the  greatest mind or model could anticipate.  Some of these consequences may be good for many, while others may be bad for many.  The systemic complexity makes modelling and policy prescription not only practically futile but downright dangerous.

And this is where we come to climate.  The inherent feature of meteorology is its systemic complexity.  No-one can control this complexity, isolate it at the macro level, study it in controlled ways and make truly hard scientific conclusions about it.  And even if we think we can approximate such conditions, we are forced to admit that our confidence levels are very low, simply because we cannot control the system.

So we can understand quite well some of the properties of the components of the earth’s atmosphere, but as a total system we cannot legitimately reach hard scientific conclusions.

But this is where the hot-for-climate-science people slip up.  They say that ’scientific’ studies DO show that we make hard conclusions.  They cite 95% confidence and statistical models and mathematical formulas.  But, exactly as in our economic example, we have seen that someone can arrive at a seemingly scientific conclusion, and even say with 95% confidence that this conclusion is correct, and still be in grave error!

This either means that the very essence of our methods for testing scientific truth is flawed, or that we are applying sound scientific methods to fundamentally unscientific realms.  Given that the scientific methods serve us so well in some sciences and so poorly in others, surely it is sensible to conclude that there are no problems with the methods per se, but with applying them to the wrong things.

S0, this leaves us with the conclusion that scientific methods are appropriate for certain realms of study, only partly appropriate for some, and wholly inappropriate for others.  Thus when applying confidence levels to results we have to take into account that scientific methods CAN produce erroneously high confidence levels, especially when applied in the wrong realm.

So what is it then that defines to which areas of study we can legitimately apply scientific, quantitative, statistical methods?  We think the issue comes down to one of systemic complexity.  Now let’s be clear here, that does not mean that just because something is complex we cannot try to understand it with the scientific method.  On the contrary, the scientific method was adopted in order to grapple with and understand the complex phenomena around us.  HOWEVER, it is only as we are able to understand and control systems better that we can place any meaningful store in the scientific results.

That means there are at least two main areas of study: those that are still being understood, and those that we basically understand.  Let’s take gravity as an extreme example.  We can go to the edge of a building and quite easily drop a ball off the edge 1000 times and record what we observe.  We can then repeat this experiment another 1000 times with balls of different sizes and densities.  From this we can begin applying statistical methods and start making very accurate predictions about how objects will fall to the ground.

Now let’s take climate study.  We cannot repeat experiments as observed phenomena take place over centuries or even millennia.  We cannot isolate the system.  We cannot run comparison studies on another planet.  We have to rely on data from history that is a) uncertain and b) selective.  We cannot experimentally alter one variable and observe its impact on other variables.

As a result, climate study is just that, a study, an intellectual pursuit.  It sits in the realm of things we are still trying to understand, like string theory, the origins of the universe, and a myriad of other complex phenomena.

The AGW climate study fraternity however claims that their field is one that is well understood.  But it’s not, and no amount of insistence by the IPCC to the contrary can change that.

Climate study has flown the coop and is fraudulently posing as hard science when in fact it is still in the discovery pen.  It’s time to rein it in to its proper place before we let these ’scientists’ and their government supporters tell us what to do with our lives, liberty and property in the name of saving the planet.

Oh, one last point: If these folk are so gung-ho about their theories and are so good at predicting climate, why try and leverage off state funding?  Why not set up private forecasting companies and actually make money off it (and I’m not talking about carbon trading which is not a free market sector but is propped up by laws governing carbon emissions and carbon credits).  If you’re so sure of global warming, make some REAL money off it.

Ok, very last point: THE ‘SCIENCE’ IS NOT SETTLED!!!! (1MB file)