In a recent article, Washington Post technology reporter Timothy Lee suggested that “broadband speeds were stagnating in the United States”, resulting in “slow innovation and poor customer service.” Comcast—the nation’s largest broadband service provider—begged to differ, and provided Mr. Lee with hard evidence indicating that the opposite was true. While Mr. Lee subsequently admitted his error and conceded that “Comcast’s service really has been getting faster”, Mr. Lee attempts to use the same data to argue that Comcast is “acting more and more like a monopolist.”

Specifically, Mr. Lee contends that these data reveal that Comcast is “focus[ing] on maximizing its own profits, without worrying very much about improving the experience of the average customer.”  Mr. Lee reaches this conclusion based on the observation that the speed of Comcast’s entry-level broadband services is not increasing as much as its network’s maximum capability, and that this pattern is “a sign of declining competition at the high-end of the broadband market.”

I don’t buy it, but Mr. Lee’s interpretation of the data brings up some interesting questions.

First, what is the problem with Comcast “maximizing its own profits”? After all, in a capitalist system, firms are expected to “maximize their own profits.”  And, if not “its own profits,” whose profit does Mr. Lee want Comcast to maximize?  As a publicly-held corporation, Comcast has a fiduciary duty to maximize shareholder value (a lesson that the founders of Craig’slist were forced to learn about the hard way in court.)  Also, I am curious—if Comcast were to maximize something other than profits, what would it be, and would that render any better outcomes than profit maximization?   I suppose that Comcast could attempt to maximize the well-being of society.  I hope its management avoids that trap. As Adam Smith observed, “I have never known much good done by those who affected to trade for the public good,” and my experience leads me to believe Mr. Smith was spot on with this idea.  After all, recognized Mr. Smith, “[i]t is not from the benevolence of the butcher, the brewer, or the baker that we expect our dinner, but from their regard to their own interest.”  The FCC concurs: in the National Broadband Plan, the agency concludes that if “service providers … cannot earn enough revenue to cover the costs of deploying and operating broadband networks, including expected returns on capital, there is no business case to offer broadband services …”  (Id. at p. 136) 

Second, Mr. Lee’s statement that Comcast isn’t worried about improving the experience of its average customer is belied by the very same data that undermined his prior argument.  Over the past five years, the speed of Comcast’s most popular and lowest-priced broadband tier increased by 50%, and the speed of its second most popular tier increased by four times.  The average customer has seen significant improvements in connection speeds.   Moreover, Mr. Lee complains that Comcast’s entry-level service is only a “pokey 20 Mbps,” but the fact is that for nearly every Internet user, a speed of 20 Mbps is more than adequate.  This is demonstrated by the fact that for a mere $10 per month, a Comcast subscriber could jump to 50 Mbps, but most don’t.  (And let’s not forget that this “pokey” 20 Mbps is much faster than the average connection speeds in the Asian markets held out as “broadband miracles” including Hong Kong, South Korea, and Japan.)  I recently reduced my speed from about 40 Mbps to 12 Mbps and haven’t noticed much difference (and I stream HD content from Amazon using a Roku box).  Finally, it is also important to recognize that the “experience” of a consumer is not just limited to speed; consumers are actually interested in price-quality combinations, and most are unwilling today to pay for higher speeds.  In the U.S. experience, most speed increases have been free upgrades by the broadband service providers.

Third, Mr. Lee appears most concerned about the “extreme divergence between Comcast’s low-end and high-end speeds.”  I do not know why he thinks this is a problem.  Over time, Comcast, and other Internet providers, will make investments that will dramatically increase the capabilities of the network. But as Comcast’s data show, most consumers aren’t presently interested in these capabilities—the cheapest tier is more than enough.  Thus, Comcast’s investments are mostly forward-looking and presently target the needs of very few customers.  Comcast went big on its upgrades, and even though the widespread need for its highest tier of 505 Mbps isn’t there today, I suspect it will be there before long as the types of content evolve and when it is, Comcast is ready.  And, they’ll upgrade again to those customers that want more. 

Although I do not think Mr. Lee’s arguments make sense, they do offer all of us an opportunity to review the basic economics underlying the broadband issue in this country. The crux of Mr. Lee’s arguments is that the observed spread of speed at the low-end and high-end of the product offering is evidence of “monopoly” behavior or a “sign of declining competition at the high end of the broadband market.” These are strong claims, but they are entirely unsupported.  I was unable to find any economic argument in Mr. Lee’s reporting, much less a formal mathematical proof, that supported the argument that large quality spreads are monopoly behavior.  He needs one.  It is just as legitimate, absent formal analysis, to conclude that the observed quality spreads are evidence of competitive behavior. (Indeed, this could be demonstrated theoretically.)  Ask yourself the question:  do we observe large spreads of quality in competitive markets?  The answer is “Yes.”  For example, using any online musical instrument dealer, I can buy a new Les Paul guitar made by the Gibson Guitar Company for as low as $129 for an “Epiphone” model or pay over $10,000 for a Gibson Custom Shop model.  Now, if Fender Guitars, Gibson’s chief competitor, were to cease operations, would we expect Gibson’s spread of price and quality to increase or decrease?  I have no idea.  That said, we typically view a wide spread on price-quality combinations as a good thing—competitive markets serve the needs of a wide variety of consumers, offering low-price/lower quality products as well as high-price/high quality products. Furthermore, if there is weak competition at the high end of the market, then why does Comcast bother to offer 505 Mbps service at all?  For a competitive advantage—that’s why.  Investments in broadband are driven by the desire to win in the marketplace, and the pursuit of victory is the essence of competition. 

Mr. Lee also concludes that “there’s no technical reason Comcast couldn’t provide 105 Mbps service to everyone currently subscribing to the cheaper 50 Mbps and 20 Mbps tiers.”  Again, Mr. Lee misses the point, ignoring the economics of the industry.  To illustrate, let’s assume, for simplicity, there are only two customers.  One values broadband at $50, the other at $90.  The marginal cost of providing a broadband circuit is $0, but the setup costs of the network are $120.  How can the seller make enough money ($120) to be in business?  The seller could offer the same service to both customers at $60, but only one customer values the service enough to pay that price.  The investment is not profitable.  On the other hand, charging both customers a price of $50 will result in revenues of only $100, and the business plan doesn’t work, the investments are not made, and there’s no broadband.  Thus, at no single price can the network investment be justified.  However, if the seller offers the service to one customer at $40, and the other at $80, then the seller can make enough revenue to justify the costs ($120 total).  But here’s the question—why would one customer pay $80 when the other is paying only $40? Obviously, the seller must offer the customer some advantage to induce her to pay the higher price.  This can be accomplished by differentiating the service, especially if the customer with the higher valuation places a higher value on broadband speed.  Let’s say that differentiation is costless (as does Mr. Lee, though this is not true), and the seller offers a low-end speed of 20 Mbps and a high-end speed of 40 Mbps.  Now, the buyer at the high-end of the market asks whether or not an additional 20 Mbps justifies the $40 increment in spending.  Let’s say it doesn’t, so the higher-valuation buyers defects from the high-end product and buys the 20 Mbps for $40; total revenue is only $80 and the business is a bust.  Instead, let’s assume that the seller offers a higher-end product of 100 Mbps, and now the buyer views the difference as large enough to justify the $40 price spread.  This is a business model which works, and sufficient revenues are obtained to finance the network. This example is simple—in fact, far too simple relative to the complex reality—but it reveals much about the economics of pricing network services. 

In sum, while I disagree with Mr. Lee’s arguments, I do want to thank him for bringing them up.  Debate is healthy, presenting opportunities to clarify arguments.  For various historical and sociological reasons, the Internet, and access to it, engenders strong emotions.  In the passion of these debates, though, the fundamental economic principles are too often under-appreciated.  For the vast majority of people, Internet service is provided by private, for-profit firms.  Lacking a steady flow of public funds, such firms must earn revenues sufficient to satisfy their stockholders and bondholders.  As is often the case in the competitive marketplace, these firms will offer various service qualities at differing prices. Such behavior is part and parcel of the competitive process.  Even in cases where public funding is part of the service offering, firms offer various service qualities at different prices (indeed, Chattanooga’s municipal system, mentioned by Mr. Lee, does so). 

It is not difficult to imagine a world that better suits our individual desires.  I’d personally like every airplane seat to be a first class seat (but sold at coach prices), especially as I think about my upcoming flight to Asia.  It is important, however, to temper our desires with a healthy dose of reality.  Sadly, such temperance is in scarce supply when it comes to broadband policy, which explains, in part, why broadband policy is such a mess.