Richard Stallman's new article: Overcoming Social Inertia

Alex Hudson home at
Wed Nov 7 14:51:15 UTC 2007

On Wed, 2007-11-07 at 14:29 +0100, David Picón Álvarez wrote:
> From: "Alex Hudson" <home at>
> > That's not really true. There are many examples of low marginal cost
> > items - software isn't particularly special in that regard.
> All goods of zero or next to zero marginal cost I can think of are either 
> very, very cheap, or there is some kind of artificial regime that restricts 
> their distribution in order to make them artificially expensive. 
> Counterexamples are welcome, though, I might be wrong about this.

Well, you seem to have an unwritten implication of "unless it's cheap
it's artificially expensive", which would be difficult to argue
against :) Cups of coffee are very, very cheap to make - but you don't
see the likes of Starbucks selling them for pennies/Euro cents, in fact,
the price is actually going up not down. 

Many consumer goods are now extremely cheap to make. Obviously, there
are other costs involved - particularly fixed costs of staffing and rent
if you have a shop or something - but the actual cost per unit is pretty
much unimportant for large numbers of products. 

And of course there are other types of product, like VOIP service, where
the cost of having a customer is extremely low (sometimes nothing,
depending on how your call routing works). Yet many people pay
handsomely for the service.

> It is true that software has relatively high fixed costs, which (absent 
> altruism) have to be paid by users somehow. However, I think it is reasonbly 
> clear that a substantial amount of proprietary software is being sold at 
> what can only be considered obscene per-copy profits. Even once a normal 
> rate of profit is recovered from a given piece of software, per-copy pricing 
> usually does not change, thus generating super-profits that are only 
> expected in monopoly conditions.

I think you're conflating two principles there. Obviously, having a
monopoly gives you the ability to make mega-profits. The fact that some
leading proprietary software makers are in that "happy" position makes
it difficult to talk about, because I would argue that it's the monopoly
which creates the massive profits, not the proprietary model.

> > The higher the fixed cost, the more risk a business takes trying to sell
> > that product - you can make a big loss if the product doesn't sell.
> You could finance the fixed costs by getting prospective users to pay for 
> the software in advance, or through milestones. Of course, many software 
> houses choose not to go this way, because they hope for the super-profits 
> that come after the fixed costs are amortized and every copy comes basically 
> for free.

That's really an argument about business model. Getting users to sign up
for software and amortise the costs that way is a great idea; it's
thoroughly impractical though, because you'd need many users up-front
willing to pay the price. The lifespan of a piece of software is usually
a few years, which is much easier to write costs off over. It's the same
for free software, too.

> There is no guarantee that a proprietary software house will be profitable, 
> but I would argue that, if the threshold of amortizing fixed costs is 
> passed, very very high rates of profit are assured. Also, I doubt that 
> Microsoft is as much of an oddity. Sure, it is the clearest and biggest 
> instance, but there are plenty of smaller once. Oracle, Adobe, SAP, come to 
> mind. Surely there are a lot more which are doing very well by 
> regional/national standards, just not spectacularly well like the previous 
> ones.

I think discussing this in terms of Microsoft, Adobe, Oracle, SAP, and
Apple, IBM, etc. is really missing the point. These guys are global
multi-national corporations. They're not representative of proprietary
software houses, and they're pretty much bound to make huge profits no
matter what they do, because that's the nature of the beast: they serve
a huge market and have fearsome economies of scale.

And it's not like software licensing is the highest margin product
necessarily: in Oracle's last 10K filing, they say that their support
business has margins of 84%. Practically every one of their software
customers buys support as well, which is an ongoing revenue. 

At the end of the day, prices are set pretty much what a market will
bear, give or take a business's goals wrt. volume. I don't think
software, free or not, would work much differently: although I do think
there are well-known proprietary vendors who have a very comfortable
monopoly in the market, I don't think that it's actually that common.

I also don't think free software is any kind of economic tool: while the
ability to share notionally does change things, in practise people copy
proprietary software so much I don't think it changes much at all.



More information about the Discussion mailing list