The Commoditization Of Software – Or Why Economics Doesn't Apply To Software
from the wait-a-second... dept
John Carroll has written his latest open source trolling piece arguing that open source programmers are doing a disservice to their industry by making people think software should be free. He tries to explain the basic economics of software and gets it all wrong. While initially he says that prices are set by supply and demand, he then ignores that in saying that prices are really set by what people expect to pay for the product. He also claims (and makes any economist gasp as he does) that price is not an inherent attribute of any product. I’d direct Mr. Carroll to an introductory economics textbook, where very early on they would show a nice little equation (and all the reasons backing it up) why price equals marginal cost of a product. The main condition being, of course, that the market be truly competitive. In such a market, the price will always be driven down to the marginal cost, regardless of what people expect to pay or what the producer wants them to pay. In the case of software, the marginal cost is zero. Certainly, it’s not a truly competitive market, so open sourcers are jumping the gun a bit in their pricing – but that doesn’t mean they’re devaluing their work at all. They just realize (perhaps inherently) that over time, it’s going to be impossible to justify charging high prices for software. That doesn’t mean, however that software has no value or that open source, free software is bad. Suddenly, free software becomes a free input into a larger business proposition, and opens up plenty of opportunities for businesses built on that software. People like Carroll need to get past the idea of charging for every individual thing they produce, and learn how to take advantage of the (dare I say?) inherent nature of goods. Just because he misunderstands the economics of free, doesn’t mean that everyone else should stop working on open source software – though, he seems to think that’s the case.
Comments on “The Commoditization Of Software – Or Why Economics Doesn't Apply To Software”
New Economics
Mike, I agree with you, however I point out that the basic economic model today is far different than common sense. In today’s society the rules of supply and demand are more the rules of supply and cost, demand is almost assumed and discounted.
I have a supply, and even though my supply is artificially limited and the demand is really small, I should still make my software available for $2000 and you better buy it because we want you to. It is our right and privilege to charge as much as we want, and if you find a better and cheaper product, you are a thief and a criminal for not paying for our rediculously overpriced product. After all it is your responsibility as a good capitalist to assure that we stay in business.
In a way, the open source model is returning to the days of old, where price was based on supply and demand, only not only money, but time and effort. If you like our product, you pay us for what you think our product is worth…believe it or not, I pay Redhat ($180+/year), Mandrake ($60-80/year), and for OpenBSD CDs ($60/year or more), etc. regularly for their software, even though I could download it for free, merely because I think their product is worthwhile and wish to support their efforts, not because they think their product is worth far more than it really is.
For the most part, the big software companies are anti-capitalistic, much more fascist in their approaches.
Economics and software development
Actually, I think you are both right.
The price paid for a product is a price agreed upon by both the seller and the buyer. If they cannot agree on a price, then said transaction will not happen. It matters little on whether said item is made from solid gold or generated for free with a ‘cp’ command, until the buyer and the seller can agree on a price, there’s no deal. If said price does not cover the costs of creating the object (ie for software, developing the object, or creating the first copy), the sellers will go out of business. And then in the future such objects will not be created, even if the cost of creating copies is zero. The costs of creating the first object can never be recovered.
Now in the environment of software, there’s a move afoot to make the price zero. As long as people are willing to provide software in exchange for no money, then such transactions will continue. However, those developers who want money for software will find themselves out of business, and if customers are not willing to pay, that’s the way it happens.
The difference comes when people decide that all software should be free. They take copies without the consent of the manufacturer. Since said manufacturer wants some kind of compensation, they are not in agreement to this transaction, and yet it happens anyway. This is copyright infringement, popularly characterized in the media as ‘theft’ or ‘piracy’. This action is acellerating the demise of a business model. If the model is doomed, it will disappear, but it still does not give people the right to take something without the concent of the supplier.
To wit: just because I can sell you something does not obligate me to sell it to you. It is still mine, and I can dictate the terms under which I will permit you to have it. If you disagree, you can go elsewhere. If my terms are too restrictive, many people will go elsewhere, and I will go out of business — but none of this gives you the right to just take it.
Re: Economics and software development
I would like to add look at software from a economic model its a product with a large fix cost (developing the first copy) and zero variable cost (the ‘cp’ command). The only other goods that had simular model to this “before the internet” that I can think of are the radio and tv. Both have large fixed cost, transmission, production etc… but the cost of making each copy, effectively each additional listener is free.
So what model work in radio and tv… the comercial model. TOday the music and soon the movie business will feel this same large fix cost and zero variable cost model hit them… its already hitting the music biz
The quuestion, which I don’t have the answer too, is what is the right economic solution for a large fixed cost zero variable cost product…
anyone?
Re: Re: Economics and software development
what is the right economic solution for a large fixed cost zero variable cost product…
You could spread the fixed cost over a number of the zero variable cost products, then at some point you will cover the fixed cost and then begin to make a profit after that. A software product can only be driven to zero variable cost if there is an equal zero cost product in the market place, equal being judged by each customer. It is hard to see how software is a zero variable cost item, even bandwidth or time to create a copy has a cost.
Re: Re: Economics and software development
The quuestion, which I don’t have the answer too, is what is the right economic solution for a large fixed cost zero variable cost product…
Oh yeah. That’s not so hard, actually. The trick is just turning around the way you look at the products. When you look at it as a “product”, then you have a problem, since the competitive forces will drive the price to zero (no matter what the fixed costs are – the fixed costs are “sunk costs” and don’t factor into the pricing).
Instead, look at the the zero marginal cost item as a “free input” and tons of business models start to show up in the creative mind. Something that’s free isn’t a product, it’s a zero-cost input into a larger product… or, a “promotion” for something else.
Most specifically, any zero marginal cost item becomes a promotional item for a service. You shouldn’t expect anyone to pay for a zero-marginal cost item that’s already been created, but you absolutely can (and perhaps should) expect them to pay for your ability to recreate a different, but related product in the future. It’s the consultant business model. You’ve shown some expertise in the ability to do a service, and you can even demonstrate past successes, but you get people to pay you to create the next success in a way that specifically benefits them.
Re: Re: Re: Economics and software development
I wonder whether the airline industry could do something like this. They have a high fixed cost and almost zero variable cost structure too.
Incidentally they have also been capital destroying almost since they were set up. Of course, there’s a difference since there isn’t anyone selling airline flights for free.
it seems economics does apply to software
interesting that an article claiming that economics doesn’t apply to software goes to such lengths showing that it does.
i think we can all agree that Excel is not a Chevrolet and it’s silly to think the two economically commensurate by tweaking some parameters. Same for tunes versus bagels. If I eat your only bagel, you go hungry, but in the economics of Software Theft, all that is lost is an increment of scarcity.
Others have argued better than I that in Free Software you have an inverse-Tragedy of the Commons.
In the Glorious Future, we’ll probably see a lot of software that solves individuals’ problems get repurposed increasing its constituency beyond the parochial interests of lone-wolf developers who are out to solve Their Own Problems. The Garage Tinkerer paradigm.
I wonder if we’ll ever see patronage-funded software? IBM seems to be doing something like that these days. Mitch Kapor’s Chandler project may be analogous to what the local billionaires did for the local museum. And we’ve seen the Mormon Church put together a lot of geneological software.
I guess what I’m rambling about is that commercial enterprises are not the only way to produce software.
No Subject Given
I didn’t read the article..so please forgive me if I missed a point inside…
I’m just thinking here that one point that these articles and discussions sometimes fail to mention is that by and large (at least in my experience) open-source programmers don’t see their software as a “product” in and of itself…but as a service…ie: you make the software…give it away, let people use it and it becomes a calling card and you sell either support, customization or consulting.
The only people who think there’s no money to be made in open-source only see software as a product rather than an opportunity.
Misunderstanding marginal cost
I’d direct Mr. Carroll to an introductory economics textbook, where very early on they would show a nice little equation (and all the reasons backing it up) why price equals marginal cost of a product.
Price equals marginal cost in the theoretical realm of perfect competition. Lacking that, marginal cost is usually lower than price, simply because the model doesn’t reflect the fast changing, dynamic environment that exists in the real world.
What, however, is marginal cost? In a world where the price of software is high, there is lots of investment flowing into the industry, and that combined with expectations of high price make the value of software developers go up. In other words, your marginal costs go up because of higher price expectations.
Marginal cost isn’t some inherent price point you can stick a pin in. Marx thought that with his labor theory of value, but modern economics views price as a very intangible creature. Marginal cost depends on demand, and demand depends on intangibles related to human motivations. No one knows why cabbage patch dolls were “in-demand” in the 80s. Whatever the reason, that made Coleco able to charge $60-80 bucks for those ugly little things.
Misunderstanding marginal cost
I’d direct Mr. Carroll to an introductory economics textbook, where very early on they would show a nice little equation (and all the reasons backing it up) why price equals marginal cost of a product.
Price equals marginal cost in the theoretical realm of perfect competition. Lacking that, marginal cost is usually lower than price, simply because the model doesn’t reflect the fast changing, dynamic environment that exists in the real world.
What, however, is marginal cost? In a world where the price of software is high, there is lots of investment flowing into the industry, and that combined with expectations of high price make the value of software developers go up. In other words, your marginal costs go up because of higher price expectations.
Marginal cost isn’t some inherent price point you can stick a pin in. Marx thought that with his labor theory of value, but modern economics views price as a very intangible creature. Marginal cost depends on demand, and demand depends on intangibles related to human motivations. No one knows why cabbage patch dolls were “in-demand” in the 80s. Whatever the reason, that made Coleco able to charge $60-80 bucks for those ugly little things.