Last week the European Commission competition authorities charged Intel with anticompetitive behavior in the market for microprocessor chips, and levied a €1.06 billion ($1.45 billion) fine on the company. Some commentators attacked the ruling as ridiculous on its face. I disagree. Let me explain why the European action, though not conclusively justified at this point, is at least plausible.
The starting point of any competition analysis is to recall the purpose of competition law: not to protect rival firms (such as AMD in this case), but to protect competition for the benefit of consumers. The key is to understand what is fair competition and what is not. If a firm dominates a market, and even drives other firms out, but does so by producing better products at better prices, they deserve applause. If a dominant firm takes steps that are aimed more at undermining competition than at serving customers, then they may be crossing the line into anticompetitive behavior.
To do even a superficial analysis in a single blog post, we’re going to have to make some assumptions. First, for the sake of this post let’s accept as true the EC’s claims about Intel’s specific actions. Second, let’s set aside the details of European law and instead ask whether Intel’s actions were fair and justified. Third, let’s assume that there is a single market for processor chips, in the sense that any processor chip can be used in any system. A serious analysis would have to consider carefully all of these factors, but these assumptions will help us get started.
With all that in mind, does the EC have a plausible case against Intel?
First we have to ask whether Intel has monopoly power. Economists define monopoly power as the ability to raise prices above the competitive level without losing money as a result. We know that Intel has high market share, but that by itself does not imply monopoly power. Presumably the EC will argue that there is a significant barrier to entry which keeps new firms out of the microprocessor market, and that this barrier to entry plus Intel’s high market share adds up to monopoly power. This is at least plausible, and there isn’t space here to dissect that argument in detail, so let’s accept it for the sake of our analysis.
Now: having monopoly power, did Intel abuse that power by acting anticompetitively?
The EC accused Intel of two anticompetitive strategies. First, the EC says that Intel gave PC makers discounts if they agreed to ship Intel chips in 100% of their systems, or 80% of their systems. Is this anticompetitive? It’s hard to say. Volume discounts are common in many industries, but this is not a typical volume discount. The price goes down when the customer buys more Intel chips — that’s a typical volume discount — but the price of Intel chips also goes up when the customer buys more competing chips — which is unusual and might have anticompetitive effects. Whether Intel has a competitive justification for this remains to be seen.
Second, and more troubling, the EC says that “Intel awarded computer manufacturers payments – unrelated to any particular purchases from Intel – on condition that these computer manufacturers postponed or cancelled the launch of specific AMD-based products and/or put restrictions on the distribution of specific AMD-based products.” This one seems hard for Intel to justify. A firm with monopoly power, spending money to block competitor’s distribution channels, is a classic anticompetitive strategy.
None of this establishes conclusively that Intel broke the law, or that the EC’s fine is justified. We made a lot of assumptions along the way, and we would have to reconsider each of them carefully, before we could conclude that the EC’s argument is correct. We would also need to give Intel a chance to offer pro-competitive justifications for their behavior. But despite all of these caveats, I think we can conclude that although it is far from proven at this point, the EC’s case should be taken seriously.
“If a firm dominates a market, and even drives other firms out, but does so by producing better products at better prices, they deserve applause.”
Maybe in the “real” world, but in the legal world, see the ruling in the U.S. case against ALCOA in the late 40s or early 50s. “Alcoa said that if it was in fact deemed a monopoly, it acquired that position honestly, through outcompeting other companies through greater efficiencies. Hand applied a rule concerning practices that are illegal per se here, saying that it doesn’t matter how Alcoa became a monopoly, since its offense was simply to become one.” from the article http://en.wikipedia.org/wiki/Alcoa .
It is refreshing to read an evaluation of whether Intel actually did or did not do something wrong. Other forums that I have read simply lambasted the EUs decision as being another example of oppressive government regulation without delving into the legitimacy of that decision.
And there is the additional issue that Intel paid a large retailer chain for exclusivity. The European Commission has been fighting “exclusive dealership contracts” for some years and it is no surprise that it is heaped on with the other issues.
Argument: A consumer can make a much better choice when he’s able to make side-by-side comparisons between two machines. If he has to drive to different dealerships it takes more time to do a proper comparison.
Oh, for crying out loud, what is it that confuses you so much about this? Offering a volume discount is okay, what Intel offered is not a volume discount but a market share discount, which is not okay for anybody, monopoly or not. A volume discount is when you get better prices if you buy 2000 chips instead of just 1000. A marketshare discount is when you get better prices if you buy 80% of your chips from one manufacturer, rather than just 50%. This is clearly exclusionary and only designed to drive business away from your competitors. An example, if company A buys 2000 chips from Intel but this is only 50% of its products using Intel chips, might get a lower discount than a company B which buys only 1000 chips from Intel but as a percentage equips 80% of its products from Intel. If it was simply a volume discount, then the company that sells 2000 chips rather than 1000 chips should get the better discount without question, regardless of percentage. But here the company that buys 80% from Intel rather than 50% gets the bigger discount.
I will assume (possibly incorrectly) that you are not merely trolling. A volume discount has nothing to do with anti-competitive practices, and isn’t what Intel did in any case. The reason a company gives a volume discount is that it accurately reflects its lower cost – if I can make one big sale rather that 10 little sales I save money on salespeople, setup, warehousing, shipping, all down the line. If all Intel had done was give volume discounts there would have been no complaints.
As Ed said, Intel made payments to companies *unrelated* to any sale it was making, simply to cancel projects with AMD, its competitor. In a properly functioning market, such behavior would have no effect (because you have no market power, and the bribee would just move to a third provider). This is the behavior of a monopolist.
As an aside, this is part of what happened to torpedo the One Laptop Per Child project, which drew Intel’s ire when its CPU wasn’t Intel’s.