Astrophysics/Climate Science/Systems Biology is a pseudo science.

It does not follow the basic principle that knowledge is proved and disproved by experiments. You can find all the sophisticated mathematical methods you want in geophysics, and none of them would stand a test of reality. Most of so called 'models' come with disclaimers against various features, and pretty soon you discover that our whole life is one big exogenous phenomenon.

Your criticism applies to the study of almost any complex multiscale system. The fine scales are poorly understood, and most models come with various disclaimers. Many parameters are unknown, and reasonable quantitative predictions are impossible. Plenty of examples show models to fail, but they are still used since they sometimes work.

Life is messy, and studying a mess leaves you with a messy field. It's a pseudo-science when you stop admitting that, and pretend you have all the answers. I don't know of any economists who have reached that stage.

Mainstream economics refuses to admit it deals with a complex nonlinear system. And econometrics claims it has all the answers through gaussian statistics.

Most econometricians would admit it, it's just that the linear models, and Gaussian models are easier to handle since they require much less data to calibrate, and the results have more or less confirmed economic theory. There is no doubt that economic theory does not have near the predictive power that physical sciences have attained, but that is due in part to how recent new developments have been, and how sparse the data available to economists is. Indeed, estimating all the parameters necessary for non-linear models is significantly trickier. Anybody that developed robust nonlinear economic models could safely bet on receiving a good academic appointment at a top-20 school, and very probably an economic prize.

I can't think of a single econometrician who claims to have all the answers through Gaussian statistics.

I don't think you are aware of much modern economics. In general, the Chicago school exactly takes the viewpoint that one is dealing with a complex, nonlinear and nondeterministic system. Neoclassical economics takes a similar viewpoint.

See also [complexity economics](http://en.wikipedia.org/wiki/Complexity_economics).

As for econometrics, they do exactly what you said economics doesn't do: try to fit the models to real life, and reveal when the models fail.

Thanks for questioning my knowledge. Let me return the compliment. I don't think you are aware of the fact that a complex dynamical system (the economy) is deterministic. The randomness implied by stochastic models does not exist in reality. Complex dynamics looks similar but is a different animal altogether. Attempts to build global predictive models are futile in principle and can be compared to attempts to build a model that predicts the weather a month from now.

Economics needs a philosophical paradigm shift, and of course it is happening (slowly, and Santa Fe Institute is the place to watch), but is it not what you find in Samuelson books and in the mainstream academia. That is the point I am making.

I know a fair amount about dynamical systems. As for whether the economy is deterministic at the most fundamental level, I suppose that really depends on your interpretation of quantum mechanics.

Concerning weather prediction, that's my point. Economics is as much of a science as meteorology is. Economics and meteorology will never achieve the level of simplicity and accuracy that fundamental physics does. Doesn't make them unscientific.