Friday, September 26, 2003

The Value of Free Software

Microsoft frequently gives away software for education in $20-30 million dollar lots, and sometimes as much as $100 million.

Of course, the price of the giveaway is quoted in if-sold value, which is not what it costs Microsoft to make the donations. A software package that costs $399 at retail, say the standard version of Office XP, costs about $1 to manufacture, including CD-ROM, box, and printed material (bot no manual any more).

Linux vendors also give away software, but their if-sold value is pretty much $0. You can pay for Red Hat or any of the other distributions if you want to, but you can also download them for free. At best, a Linux vendor giving away on operating system and a thousand applications can claim something like $29 a copy as the selling price, or somewhat more for the enterprise version.

This is one of the reasons why Microsoft's donations are treated as big news in the media, while Linux donations are ignored. On the other hand, Microsoft is slashing prices for countries where the governments are thinking seriously about switching to Linux for education and other programs. Thailand has a deal for Microsoft Windows XP and office for $36 a set, combined. So what would happen if Microsoft donations were reported at those prices?

Let's also take a look at the value of a Microsoft or a Linux donation. With Microsoft, if you want to go on using later versions of the software, you have to pay. Maybe you have to pay first-world prices, and maybe you have to pay the new third-world prices, but you have to pay each time, for each program or suite. With Linux, you pay the same price for the OS and the thousand applications each time, whether that is $29 or $0.

What will it take to bring developing countries across the Digital Divide? Well, if you want to go the whole way, you need to provide a computer and software for each of four billion poor people. Why one per person, and not, say, one per family, or even less? First, because we are talking about doing away with the Digital Divide, not building a few narrow bridges here and there across it. So that means that everybody needs easy access. So let's say one in each home, a sufficiont number in every school and library, one for each worker on the job, a bunch of laptops and palmtops--actually we're going to end up at some point with more than one computing device per person.

Now you can spend $500 and up per computer and maybe $36 for Windows and Office. In round numbers, about $2 trillion at one unit per poor person, without talking about all the other software needed, which very quickly exceeds the cost of the hardware, amounting to trillions of dollars more. And that is only the first time. In a few years, you will have to buy the software all over again.

What if you do it with Linux? Well, you can run Linux on a $300 desktop or a $200 handheld, and the software is free. Let us arbitrarily say $250 per unit. Again in round numbers, $1 trillion.

That's the cost. What is the value? The return on investment (ROI) for education is quite astonishing (as Ireland and india have been demonstrating), as is the ROI for health. The ROI for the economic opportunities enabled by computers and communications is inculculable. Once the process gets under way, it creates a wonderful period of exponential growth, where new prosperity leads to further improvements in education, health, opportunity, and so on.

The developing countries will generate wealth enough to pay for the new ICT, whether they do it through commercial software or Free Software. But they will do it much sooner if they can save a few trillion dollars early in the process, when it will make the biggest difference.

Now there is more to the equation than this. I haven't discussed communications equipment and costs, for a start. But that doesn't change the principle.

So forget debt forgiveness for developing countries, which would be worth a few hundred billion dollars. Forget low-cost medicine (only a few billion, so far) and agricultural sibsidies (maybe a hundred billion annually). We're talking about real money here.

Comments: Post a Comment

This page is powered by Blogger. Isn't yours?