I'm 45. About 10 years ago I was pretty confused about economics, sociology, etc and did some pretty intensive research. My background at the time was in business for myself and mainly libertarian and right wing philosophies.

Once I started to realise what a complex science it really was I started taking notes and indexing it all. I ended up with about 700 pages of indexed notes and about 6 cubic feet of photocopied stuff read but not indexed.

As part of all that I delved pretty deep into the mathematics behind it all as that's one of my hobbies. Eventually I came to the conclusion that when you looked deeply enough into both sides there was no single theory to explain everything, but that like physics there was two viewpoints.

Each viewpoint tends to be self referential, which means that according to the facts that fit with it they seem to imply the other side doesn't exist. Randomness and probability describe the world pretty well and much of the last 200 years was built on it, with the discoveries of Pascal and Gauss. No one knew what chaos was so no one knew how to run an economy to reduce catastrophes like recessions, runaway inflation, high unemployment, etc.

You can do an equally good job explaining the world in terms of chaos and assuming that everything depends on something else, which is like the idea that people cannot be regarded as separate from the field like society they inhabit. Communism was just an expension of socialistic ideas of the 19th century. These came from the quite obvious defects that capitalism had at the time, and a trial and error attempt to fix them. Communism was never meant to be a system by itself, but to establish a symbiotic association with capitalism.

It accomplished this quite well in Europe, in Russia it became extreme and ultimately failed. It did work reasonably well though, certainly enough for a relatively poor country to threaten the world. Systems based on chaos work because they eliminate most chances of something bad happening. By reducing so much of this though they often eliminate risk taking which causes them to lose in other ways.

When you approach the problem from each of these two viewpoints mathematically it explains a lot by assuming the other side doesn't exist. For example probability is based mathematically on the assumption the objects analysed are independant of each other. Often this is not the case but making this assumption early on makes it much easier to get answers. The problem is in economics those answers are only approximately correct because some things do depend on each other.

For example if you go into the car park of a shopping center then whether you want to park and shop has little do with anyone else so probability explains the ebbs and flows of shoppers quite well.

If someone sets off a fire alarm probability fails and chaos takes over because people become like a mob in which their actions are controlled by what other people do. While chaos describes this situation quite well it isn't good for describing the shopping center when nothing bad is happening. You can design it with extra exits and fire extinguishers by calculating the effects of catastrophe on a situation, e.g. people running around in circles or following others who don't know where the exits are.

So even in this small case it is quite difficult to get the balance right between allowing for chaotic movements of people and random ones. No one today would ignore the potential for either chaos or randomness in town planning though. The melding of the two has become an established science throughout the economy.

The stock market is the same. Investors are assumed to be independant in working out the values of shares. If someone publishes bad news that can make a panic in which other people see big sales of shares and dump their shares as well. This is again similar to the shopping center fire.

Probably the best example of this was Black Tuesday (IIRC) where computers all decided to sell at the same time and nearly bankrupted the country. Because they all assessed the stock market with probability none of them could understand how this led to a chaotic situation. If the program had been able to recognize this then it could have been stopped earlier. So in this case purely random probabilistic situations created chaos and then couldn't understand it. Human investors realised it was a chaotic baseless panic and pulled all the plugs.

So this situation happens spontaneously in random systems where chaotic elements take charge suddenly, and like in a shopping center the system is more stable if it is designed with this in mind.

In the same way welfare can be wasteful but in a sudden unforeseen downturn the job loss turns chaotic as it has the last few years where companies panic and get rid of too many workers which leads to more firings elsewhere.

Such events can be crippling to an economy and if they are caused by chaotic panic then many people end up unemployed and if there is no safety net then they lose their houses which depresses the housing market, and it can take them a long time to get back on their feet. Capitalism assumes nothing should be done to help so this chaos creates much more damage than it otherwise would. The mathematics of this is completely different from market forces.

So this chaotic cascade can do enormous damage to an economy, That's why Keynesian economics was formed so that in recessions extra government spending stops some of this cascade by creating jobs temporarily and then running a surplus later.

So a safety net is not meant to be a mattress to lay on but to catch and arrest catastrophic declines. Insurance is the same which is why people insure their homes against fire. If insurance didn't exist the economy would probably be bankrupt from catastrophe in 10 years.

The problem with both of these systems is both extremes want to set it up to give themselves a free ride. Welfare queens don't want to work and rich people find they can bribe the government to get extra subsidies or loosen the laws to do something dishonest. Usually it's the middle class that suffers on both sides. At the moment Bush is paying off his base, eventually the Democrats will pay off their base. Bush deregulates so his base can get away with more dishonesty. The democrats pander to their base by reducing prison terms for the lower classes.

So the whole thing is like a giant incomprehensible machine running on complex mathematical formulae not completely understood with much of the data not even possible to analyse in principle. To run all this science needs to do its job with peer review, journals, etc. It's far beyond what laymen can comprehend and advise on, unless you want to spend a couple of years reading and understanding it.

If you do that then you can understand what Krugman and De Long are talking about and why virtually all the economists in the world are saying Bush is doing things wrongly. If you don't read up on it thoroughly in mainstream books then you have about as much chance of understanding it as knowing how computer programs work without learning any languages.

The problem is demagogues like Limbaugh and others on the left don't understand it and try to convey the impression it can be understood in a dumbed down way. It can't be.

Imagine for example if a new passenger plane was to be designed by all the demagogues with all the rhetoric. We don't do that because we leave it all up to scientists though we can't understand as laymen much of it. In the same way economics is like that, it's a science that is more complex probably than modern physics as it has to run a whole world. It is imperfect because it can't model people accurately, but it is very good indeed.