Markets, Data, and the Value of High-Tech Patents: A Q&A with Christopher Knittel

Academic affiliate Christopher R. Knittel’s consulting work encompasses a wide variety of practice areas, including energy, intellectual property (IP), technology, antitrust, and damages assessment. What ties this work together, and how does his extensive academic publishing illuminate the topics at issue in his work in high-tech IP cases? Analysis Group Managing Principal Chris Borek and Vice President John Browning spoke to Professor Knittel (who’s on the MIT Sloan School of Management faculty) about the connections informing his academic and consulting work, as well as his role in the landmark Apple v. Qualcomm case.

Your scholarly work covers fields we ordinarily wouldn’t think of as sharing close connections, such as IP and energy and the environment. What, if anything, links them?

Christopher R. Knittel: George P. Shultz Professor, MIT Sloan School of Management; Director, Center for Energy and Environmental Policy Research, Massachusetts Institute of Technology; Co-Faculty Director, The E2e Project

In my academic research on industrial organization, applied econometrics, and energy economics, I focus primarily on analyzing marketplace data to differentiate the impact of pricing from other factors on corporate or consumer behavior. For example, how do gasoline prices affect ethanol production levels upstream, or fuel switching decisions downstream? As it turns out, the same kind of analysis goes into determining the value of patented technologies and scrutinizing claims that a company is overcharging for the use of its patented technologies.

How do those concerns come into play specifically in IP cases?

As you might imagine, high-tech products such as computers, chips, and smartphones present unique challenges for sorting out value contributions for all of the intertwined technologies they employ. A single product or device might incorporate literally hundreds of different patented technologies, making it complex to estimate the value that any one feature or patent contributes to overall value.

Can you give an example of one of these cases?

One of the biggest cases I’ve worked on was Apple v. Qualcomm. I was retained on behalf of Qualcomm in a large-scale IP and antitrust dispute with Apple over IP patented by Qualcomm that was used in Apple’s mobile phones. Qualcomm has an extremely strong patent portfolio, including patents essential to the 2G, 3G, and 4G wireless standards. In the high-tech world, it’s virtually impossible to develop an innovative product in a cost-effective way without employing a wide variety of what have been deemed to be standard essential patents, or SEPs – that is, technologies that are so ubiquitous that they benefit the market as a whole. To encourage adoption of such beneficial technologies, industry participants often agree that the royalty rates negotiated for the use of such technology should be “fair, reasonable, and non-discriminatory,” or FRAND.

At issue in Apple v. Qualcomm, among other things, was whether Qualcomm had been charging Apple “supra-FRAND” royalty rates – that is, rates so elevated that they were no longer FRAND – for the use of its mobile wireless technology. This question was also at issue in the Federal Trade Commission’s antitrust suit against Qualcomm, a case on which I served as a consulting expert and in which the Ninth Circuit Court of Appeals recently held in Qualcomm’s favor.

What was your specific assignment?

A key question in the case was: How much value was Apple deriving from being able to incorporate mobile wireless technology, including Qualcomm’s technology, into its devices? My role was to assess this value, and to do so, I was asked to analyze the price premiums for mobile wireless connectivity that Apple has been able to charge on iPhones since that product line launched in 2007.

In essence, you were placing a value on the ability of an Apple device to connect to the internet while outside the range of Wi-Fi networks.

Exactly.

T. Christopher Borek: Managing Principal, Analysis Group

But mobile connectivity today seems like a ubiquitous entity. How do you begin to measure its value to a consumer, and hence the extent to which Apple profited from incorporating this technology into its products?

Here is where the market data to which we had access is crucial. Apple’s historical pricing practices revealed a clear correlation, both within and among product lines, between the addition of cutting-edge features and higher product prices. This idea makes sense on an intuitive level, but the thorough study of Apple’s pricing that we conducted bore it out. If adding premium features, such as wireless connectivity, to its products leads not only to higher prices but to higher sales, that constitutes strong evidence that a price premium – the levels at which a product’s price exceeds that of a benchmark product – provides an estimate of the lower-bound benefit of utilizing a patented technology – in this case, mobile wireless connectivity.

One of the key analyses we performed was a historical comparison of the prices of Apple’s iPhone models and iPod Touch models.

What did that comparison show, and how did it help you to home in on the value of mobile wireless connectivity to Apple?

Historically, the iPhone and iPod Touch have been quite similar devices along many dimensions. A critical fact is that every iPhone model has incorporated wireless connectivity, whereas the iPod Touch never has. In fact, over the time that Apple has offered the two products, the inclusion of mobile wireless connectivity in the iPhone has always been the key difference between the devices, and at certain times it has been one of just a few important differences between the two products. And Apple has typically generated $300–450 more in revenue on each iPhone sale than they have on each iPod sale.

John M. Browning: Vice President, Analysis Group

But there were additional differences between the two products. So how did you go beyond pairwise comparison to focus on the mobile connectivity price premium?

You’re correct that there was no pair of devices that appeared to differ solely on the dimension of mobile wireless connectivity. That’s where regression analysis was useful: It allows you to isolate a particular feature while holding all other features constant. In this case we used data related to Apple’s pricing, the feature of interest, and control variables to look at the relationship between the two products. Rigorously employing this technique allowed me to precisely estimate the average per-unit price premium in Apple iPhones associated with mobile wireless connectivity.

What was your conclusion?

My analysis showed that the value to Apple of incorporating mobile wireless connectivity into its iPhone model has been substantial, to say the least.

Stepping back from this particular case, how does this kind of calculation relate back to whether particular royalty rates are FRAND?

I’ve been involved in a number of cases where price premiums have been used as an input in a reasonable royalty assessment. In such cases, these price premiums can give a good indication of a hypothetical licensee’s willingness to use the technology at issue, and are also relevant to a number of the Georgia-Pacific factors that the courts use to determine reasonable royalty rates, particularly the factors addressing the benefits and profits of the patented invention. So I see this particular analysis as being an especially powerful element in these often-complex determinations. ■