Broadband in America

Eli M. Noam

Montpellier, France

November 2001

 

 

For electronic media, transmission technology is destiny: it affects format, content and economics. It used to be expensive to move information, now it is cheap. We can do old things in new ways, new things in old ways, and new things in new ways.

 

Past technology enabled us to create transmission networks of two types: those that moved a lot of bits and waves, shared by many. Think of it as a fat party line. This we called broadcasting and cable TV, both “synchronous” forms of communication. The second type of network moved a relatively small number of bits, but it did so on an individualized, non-shared, “asynchronous,” basis, giving everyone a skinny but individual line. This we call telephony. The two different applications were based on the cost of delivery.

 

This is in the process of changing. So now we are in the midst of a historical move: from the kilobit stages of  individualized communications to that of megabit stage , and within the reasonable future, to the gigabit stage . 

 

How is broadband progressing in America? In many ways, very well.  In others, only moderately. Let me explain.

 

Broadband is a big success in terms of supply of availability.  It is now available to 70% of all US households via cable TV, to 45% of the US population via telephone DSL, and to virtually everybody via satellite with return phone channel, and soon via return satellite uplink. It’s available across the country, including in most rural counties.

 

On the whole, the upgrade to a higher speed availability has been phenomenal in historic terms, considering that it’s been an infrastructure upgrade, not like the narrowband internet.  Narrowband, which required mostly a modem, the equipment cost was mostly a matter of consumer investment, and network upgrades were in the core, which is relatively much cheaper. In contrast, DSL and cable modem require network providers to do things to their local network, which is the more expensive upgrade.

 

This success in rollout means that several arguments for regulatory change – some embodied in the so-called Dingell-Tauzin bill in particular that is before Congress, which attempts to deregulate, among other things, high speed local access—that is, DSL and fiber-- by the Bell companies, with the argument that it retards the rollout of broadband, became outpaced by the developments in the market. The Bell companies had no choice but to meet competition from the cable industry, Dingell-Tauzin or not, asymmetric regulation or not.

 

Where BB exhibits a less dramatic development is on the user side. So far, BB has been a supply push story.  The demand side of the equation has been more mixed. 

 

On the one hand, growth has been decent.  Looking at the date, cable was there first and forced phone companies to follow. That’s competition.  You can also see that the ratio of cable and DSL has stabilized at 60:40.  There are some other minor players, too, such as fixed wireless, satellite (which may be 200,000 and growing once Hughes rolls out two-way satellite service), and fiber (about half a million now).

 

The big providers of broadband are, not surprisingly, the two largest cable companies and the two largest phone companies.  There was a total of nine million subscribers of broadband, almost 10% of households, and about 20% of households with residential Internet access.  6.4 million were cable modem users in the case of 2001, or about 9.1% of the 70 million households able to receive the service.  Of these, AOL Time Warner had 1.7 million subscribers and ATT Broadband had 1.4 million.  DSL had about three million subscribers.  Of these, SBC had 1.2 million, and Verizon had one million.

 

Cable Growth rate was 15 % in the 3rd quarter of 2001, faster than earlier in 2001. Considering that the economy is weak and that other electronic services are slowing, this is remarkable.

 

If one compares the adoption rate of broadband with other consumer technologies, it also looks pretty good.  Growth rates have tracked those of other consumer information devices, they were twice as fast as cell phones, vastly faster than color TV, about the same as for PCs, and significantly slower than for black and white TV.  And recall that it required extensive infrastructure construction or upgrade, which makes this growth rate quite impressive.

 

This looks good.  But, in international terms, the US take rate is far behind that of Korea or Canada, though ahead of the major European countries.  Korean broadband has both cable and DSL, and three major competitors.

 

So the question is, as much as the growth has been heart warming, as nice as it is to have the glass of subscribership be 1/10 full, why are so many people NOT choosing broadband when it is available?

 

 What are some of the reasons for the demand not being stronger? 

 

  1. The price of broadband is not low. They have first come down, but now they have recently gone up again by 15%, which you cannot see on this chart. Consumers are price sensitive. In surveys, consumers indicate their price resistance. Also, take rates are definitely related to income, of course, and more strongly than narrowband internet service.

 

Price is based partly on cost. DSL is not cheap to install. It is maintenance intensive. Each truck roll is about 200 dollars. And often, you need several service units.  So if you have to go back to fix an installation, you burn up a lot of money.  Phone companies have pushed self-installation, but that’s still not easy to install for mass market consumers once you get beyond the net-heads.

 

  1. Competition is the reason for rising DSL prices and service fulfillment being slow.  Intra-modal competition options has declined, at least for DSL.  It has never existed for cable modem service.

 

 

Remember the days when everything involving the internet was supposed to be wide open, when you couldn’t tell a dog on the internet?  Well, those days seem over.  Economies of scale are back.  And broadband seems to have increased these economies.

 

For a while, we could ignore these economies, because the inefficiency of the incumbents masked them, and provided an umbrella. But the inefficiency has declined with threats of competition, and now economy of scale and scope are back, and now the small entrants are on the ropes.

 

Local DSL competitors are on the ropes. CLECs, DLECS, and BLECS are bleeding. Resellers can’t make it. DSL companies are bankrupt.  Wall Street, always quick to spot a trend after it has happened, has shut down DSL financing. IPOs are gone.  Junk bond funding is gone.  Equipment vendor financing is comatose, together with the financial health of the equipment vendors themselves. And for existing equities, entrant share prices have dropped by a depressing 96%.

 

Here are the DSL industry statistics from last year.  On the one hand, this is positive. Isn’t it great not to have to listen anymore to those purveyors of hype, about how bits play by different business rules than atoms, how the silicon economy is different from the carbon one, and how a P/E ratio need not have any E that stands for earnings, as long as that e-stands instead for electronics?

 

On the other hand, it reduces competitive pressures for higher innovation and lower prices.  We are therefore now at a cross road about telecom regulation. We now have in America an FCC chairman who has signaled to stay out of merger reviews and not to impose any conditions. We have a Justice Department that is likely to be friendly to mergers, witness its political treaty with Microsoft. And we have a Wall Street community that is shifting its financial bets and political weight from the entrants to the incumbents, now that it has discovered the probability of market power. All this creates a window of opportunity to major mergers.

 

Thus, on the supply side, there has been no real development of intra-platform competition. There is little telecom vs telecom, or cable vs cable. And now, even the top two satellite TV providers want to merge. No more star wars for American DBS.  On the other hand, there has been a trend towards inter-platform competition.  Telecom vs cable. And, in the future, US fixed wireless, 3G, or more likely, W-LANs which already reaches 11 MBps, in contrast to the slower, ISDN-like speed of 3G.  Soon, there will also be competition from satellite.

 

  1. Content.  The major problem of broadband is not supply push but demand pull. People need a reason to upgrade, and to pay maybe $50 per months for the connectivity.

 

What would broadband be used for?

Even the industry has no idea. The president of AT&T, the company with the second most broadband customers, recently opined that he has no idea what consumers could use his own service for.  This must lead to a truly inspired marketing campaign by AT&T.  A slogan like: “Buy it, maybe you can use it for something”.

 

Also, there is only so much free time and attention to go around in a day. Especially now that TV viewing this year in the US has gone up again, even before September 11.  And web browsing generally has declined.

 

And there is no indication that broadband users use the internet much more than narrowband users.  Broadband users internet usage, in minutes per hour during a 24 hour period, are very similar to narrowband users – just a bit more.  The main difference is around three in the morning, which probably means just that always on broadband is on while the user is sleeping.

 

I suppose broadband usage includes some high speed web surfing, and always-on is the main driver of usage in entertainment, mostly for media use, audio and especially video.

This is the weak element of broadband today – Content.

 

According to my calculations, the distribution costs of TV over the internet are 1.85 m¢ per second of content and per user.[1]  This is 40 times higher than the distribution cost per cable channel.  The reason is that asynchronous individualization requires significantly larger transmission resources than synchronous broadcasting.  You saw that quite clearly on September 11, when TV was great, email was great, but internet news-sites were mediocre. Internet-TV can function economically, in competition, only as a premium medium or a specialized medium.

 

When it comes to content itself, the whole point of internet TV is to be more than regular standard, linear TV.  The interactivity and multimedia aspects of the medium require additional features beyond straight video. And after some initial amateur period, it cannot possibly be produced cheaply.

 

Partly as a result of these economic realities, the content providers are not doing well.  Last year, in the months known as Black September, many of them got wiped out.  They have not found a business model to support their content. Or, put differently, they have not found a content model that would support their business. Video over the internet is not likely to be a huge driver of demand, or driver of bits, or driver of dollars.  We must wait for a decent system of micropayments.

 

How does this all add up?  It’s a classic chicken and egg situation. Content is the chicken, demand is the egg.  In this situation, what has government policy been up to?

 

There are several policy recommendations.  First, there is the need to prioritize widespread deployment and defer new regulation in the early stages while avoiding present-day policy making that is based on presumptions about the final form of broadband markets and by favoring policies facilitating rapid deployment over intervention.  Second, the structure of regulation needs to emphasize facilities-based competition by favoring alternatives to physical unbundling and anticipating that facilities-based competition will not occur in all places.  Appropriate policies need to be fashioned to address these gaps.  Third, there needs to be a move toward a more coherent, consistent policy framework for broadband.  Fourth, we need to take active steps to promote increased or accelerated deployment, including at the local level by exploring public-sector initiatives that foster market entry.  Fifth, there needs to be an increase in local capacity to promote broadband deployment by supporting planning grants for localities to explore options and by providing cost-sharing for field trials, including local-government-sponsored initiatives.  Sixth, we must

 defer development of a universal service policy for broadband until the nature of broadband services, pace of deployment, distribution of access, and social significance become clearer.  And finally, we must support research and experimentation, on access technologies, especially targeting the needs of non-incumbent players and other areas that are not targets of stable private-sector funding, and on economic, social, and regulatory factors.

 

 

 

Conclusion: 

We are now at the verge of the individualized broadband pipe.  It’s proceeding pretty well in America.  Whether it will take 5 or 15 years, in historic terms, is quite secondary.  Primary is the trend, which is positive.  And primary is the hope, the hope for getting it right this time, the hope for a new medium of individuality and community, of liberty and equality , of creativity and diversity, of commerce and of leisure. The reality will always fall short, but the hope should not be curtailed in advance.

 

Yes, speeding up the process by government is helpful.  And avoiding an early over-regulation is likely to be costly. But at the same time, just as important as when we reach the individualized BB pipe is what we get when we reach it.  And it therefore it is time for us to consider what we want to reach -- in terms of media structure, media content, e-commerce, globalism. Those are questions that should be discussed. We should not regulate in advance. But we should think in advance.  And that’s what we should do, on both sides of the Atlantic. Because if we don’t know where we want to go, we might actually end up there.