If Karl Marx were alive today he would have a field day updating his theories on the crisis of capitalism to encompass the shifting morass that forms the economics of software.

With the current recession there has been a lot of debate about the use of economic stimulus with Keynesian theory: how government can effectively jump start economies by spending and helping money to circulate.

I can see how it works well for doing things like building roads and railways. There is a nice linear relationship. The technology is well understood. You spend 250 million on roads and you get real tangible economic value out the other end. I would appreciate it if some of that effort could be made around my neighbourhood. We know how to build roads. Building roads employs lots of people including many without sophisticated skills. It generates a lot of good stimulus for an economy because low skilled, low income people have no choice but to spend all their income to live. Which means all the government spending recirculates quickly into the general economy.

Software on the other hand has many horrible properties that, in my opinion, make it dreadfully unsuited to Keynesian economics.

Firstly, people who are good at it tend to be in high demand already. So putting economic stimulus into software generally does not benefit people who need it most. Secondly, if you put money into the pocket of a high income earner they won’t spend it all – so the economic stimulation is lessened. The likely result is little or no tangible economic value out the other end.

Oh dear.