This is true to some extent. The biggest problem is, if I dig ditches for a living (oh, if I could only find such a job!), why should I be taxed to pay for somebody who wants to play video games, monitor equestrian breeding lines or transfer doughnut futures to Iraq? I want ditch-digging software. The argument can be made that the equestrians will build more fields and require more ditch-diggers, but that is far removed from me. I probably won't reap enough benefits to cover my additional taxes. This is sort of the Robin Hood approach. If you are a Keynsian economist, it makes sense to charge more for the software because it increases the velocity of money. That, in turn, stimulates the economy, produces more goods, and increases everyone's standard of living. All that being said, the government's job is to protect us from ourselves. They are charged with playing Robin Hood in many instances because us poor dumb ditch diggers (no offense to ditch diggers) really don't understand why studying the sex lives of crabs can be valuable to everyone. In other words, maybe some open source software should be encouraged. But to take a broad stab at it with business software (especially vertical applications) would be folly (in my opinion). There are numerous examples of this type of subsidy doing harm to an industry. Just look at health care, agriculture, and administration in many enlightened countries. The capitalistic incentive is taken away. Entrepreneurs' efforts aren't rewarded sufficiently to inspire their best efforts. We have a computer industry where it is today because there was the enticement of huge revenues. I'm convinced that without that allure, we wouldn't even have the 8086 processor yet. There's also one other danger in government subsidies: government rarely operates as efficiently as they could. That means it costs $0.43 per user to make the $0.20 software available. That's why subsidies must be very carefully chosen. The other valid issue you bring up is the cost of distribution. The truth is, it costs way more to distribute to 1,000,000 users than 10. Nearly 50% of every dollar in software revenue goes to sales and distribution. Open source tends to totally circumvent this since who really cares if anybody knows about the software? It's not like anybody makes any money when the product is sold (or at least not very much). If it's good enough, the word will eventually get out. Open office has been around since '97 but it has taken nearly a decade for it to really make inroads