Why Do Anti-Dumping Measures Not Apply to Software?
Flexible Costing Lets Coders Charge What They Please
By Roger L. Kay
If an American chip company were doing with microprocessors what Google is doing with
software, it would be hauled in front of the World Trade Organization (WTO) in short order.  Why is
it okay to give software away for free — intentionally losing money in the process to gain market
share — when it's not okay to do so with other, more tangible goods?  Answer: people just don't
see software as a tangible thing.

Last week, a friend drew my attention to this vexing little problem.  The WTO concerns itself with,
among other things, dumping.  On its Website, the WTO defines dumping as “ … the introduction
of a product into the commerce of another country at less than its normal value”

Now, Google could drive a truck through the word “normal.”  In Google’s world, normal is giving
software away: Google Apps, Google Earth, Google Toolbar, Chrome, Android, and other
programs are all conferred on the end user free of charge.  So, from a WTO perspective, Google is
cool.  It gives software away in its “home market.” So, in theory, it can give its code away
anywhere.  True enough.

But this perspective doesn’t quite capture what is going on in the software world, and Google
relies on the fact that people don’t really understand how the company’s business practices have
changed the game.  Google can continue to pose as the great diplomat, making the world safe for
information flow, while in reality it is pillaging industry after industry as it “dumps” software on the
market.

I got curious enough to go back to Bill Gates’s Open Letter to Hobbyists, that canonical document,
vintage 1976, the first major attestation to the value of software and the harbinger of the huge flood
of money that would wash into Microsoft’s coffers over the next few years.  

The Open Letter is one of the most misquoted documents of our time.  It is often summarized as
being the venue where Gates said, “Software is not free.”  He said no such thing, at least not in the
letter.  Never much of a sound bite kind of guy, Gates laid out in forceful, if uninteresting, terms an
argument for why hobbyists should be paying for their software, rather than just copying it and
passing it around to their buddies.  At first, this line of reasoning may sound draconian and
curmudgeonly, but, like many Gates pronouncements, it is much subtler than that.  While he does
berate hobbyists, he saves special ire for those who obtain software for free and then go on to sell
it as part of a hardware offering.  Here, he signals for the first time what would become his
trademark willingness to sue over intellectual property.

What I found interesting in the detail of the letter, though, was Gates’s justification for charging for
software.  You may or may not remember that IBM, the Goliath of the day, had a habit of NOT
charging for software.  It sold mainframes for $10 million and filled them up with this inexplicable
stuff, software, and did lots of hand tooling in order to make the large systems do what customers
wanted them to, charging for services, which also turned out to be a great business, but not for
software.  Gates whined that it had cost his company $40,000 in computer time to develop Altair
Basic, the language used to program the first personal computer.  He also pointed out that given
the many months it took to develop the software, he and his pals made only $2 per hour in
royalties because only 10% of the hobbyists had actually paid for their licenses.  Now, this was
1976, and even then $2 per hour wasn’t much.  I made $2 per hour, doing much humbler work,
while employed at my dad’s computer company in 1970.

So, in 1976 Gates framed the value of software in terms of computer time on large mainframes
and programming labor costs.  Of course, he would later abandon this view and begin to charge
for end-user value rather than cost.  Some of this software-value morph can be attributed to the
great plasticity of software costing.  Particularly since the advent of distribution by download, the
marginal cost of an additional copy of a given program is zero.  This principle is what lay beneath
the hobbyists’ glib redistribution of software.  Instinctively, they knew that copying a string of bits,
once it existed, was a free operation.  This principle also underlies Google’s model.  

However, Gates reminded the hobbyists that, although software’s marginal cost was zero, its
capital investment could be substantial.  But how to allocate the fixed costs?  Clearly, at some
point the investment in any given generation of Windows is fully amortized and then some, but
Microsoft doesn’t just start giving it away.  It continues to charge full boat.  Even loaded up with the
unprofitable Windows Live effort and all sorts of R&D, employee severance, and the like, the
Windows Division at Microsoft made a 78% operating profit in 2FQ10, which ended Dec. 31,
2009.  The gross margin on Windows has been estimated by some to be as high as 90% in
some quarters, a figure Microsoft is not too anxious to confirm, given that the number tends to
demonstrate just how little software actually does cost.

You can see the pricing flexibility phenomenon in the way Microsoft prices Windows and Office.  
For Office, the company gets as much as $499 for the full Monty in retail.  In China, to keep Linux
out, the company distributed a bundle of Windows and Office for three bucks in 2007.  Pricing is
totally arbitrary and based on competitive considerations more than anything else.

And then along comes Google, making lots of software but charging not a dime for it.  And as
Google “dumps” software onto the market, it undermines companies like Microsoft that charge for
it.  

Of course, lest we forget, Google gives away its software because it can charge advertisers
handsomely on the back end.  But still, software is clearly a tangible good, a legal fact to which
longtime paying Windows customers and losers of intellectual property court cases can attest.

© 2010 Endpoint Technologies Associates, Inc.  All rights reserved.
Debate Over the Tangibility of Diaphanous Goods