Unexamined or unstated assumptions sometimes are not part of policy debates, but are nevertheless important. Consider, for example, the Telecommunications Act of 1996, the most-significant attempt to update in a major ay U.S. communications policy since 1934, by one measure, and since 1984 by a second measure. On the first score, the basic framework for communications regulation had last been set in 1934, with the passage of the Communiations Act of 1934, which codified most of the rules communications had operated under, for most of the 20th century.
One way of illustrating the challenges of 1934, and the subsequent implicit issues in 1996, is to look at what issues needed to be regulated. In 1934 it was radio licensing and telephone service. In 1996 the issue was really competition in voice communications. One might point out that the historical peak of phone line usage by Americans appears to have occurred sometime between 2000 and 2001, and that usage has been declining, rather steadily, ever since.
On the second score, 1984 represented the voluntary breakup of the old AT&T monopoly, creating separate local telephone and long distance segments of the business. The Bell system breakup also introduced competition in long distance services on a wide basis, for the first time.
The explicit assumption in every policy effort aimed at introding competition into a market is that too much market power exists. The implicit assumption is that newly-competitive policies will have been proven to work when the former incumbents lose market share. Sometimes, the unexamined assumption is that the markets being considered are stable; that end user demand for the product will remain strong, for example.
All of those assumptions ultimately proved incorrect in these cases. The U.S. over-the-air radio market no longer is dominant enough to warrant much concern about market power, within the broader media business. All U.S. radio revenue in the first three quarters of 2009 was a bit shy of $10 billion, for example, and overall industry revenue has been displaced by a shift to television since the 1950s.
The introduction of competition into long distance did provide lower prices and greater options for consumers. But long distance has gone from being the highest profit margin service in communications, able to subsidize telephone service for most consumers, to not being a distinct industry segment, and is among the lowest-margin parts of the entire communications business, if not the worst-margin product in the entire catalog.
The Telecommunications Act of 1996 was designed around the assumption that the voice market was stable. With the rise of all manner of Internet-based applications and businesses, plus mobile services, fixed-line voice has begun a long process of decline as a principal revenue driver for the industry, in the United States and globally.
To use but the most-recent example, voice communications deregulation, there is a difference between applying more competitive measures to a declining market than to a growing or stable market. Adding pressure to a shrinking business likely only enhances its decline.
In the case of the Telecom Act of 1996, regulators essentially proposed "more competition" for a market that was about to begin a decline. People also forget that the dynamics of competition essentially pitted the legacy long distance providers against the local telcos. Between them, AT&T and MCI garnered a majority (60 percent or more) of all customers ever gained by the entire competitive communications industry.
That effort ultimately failed, and most of the U.S. "retail long distance" industry no longer exists.
One can surely debate whether, and when, it ever is useful to prop up a declining industry. The perhaps more-important point is that it doesn't make much sense to apply new constraints to industries or services that are in decline, as "voice" communications surely is.
The unintended danger for virtually all proposals aimed at applying more constraints on large ISPs is precisely that it is the wrong time to do so. The unexamined assumption is that the industry is fundamentally healthy, and can handle a significant new dose of rate, conditions and terms regulation. That might not be the case.
There always is great danger when any single industry faces technology or market displacement. The analogies might be moves to impose more regulation on steel, auto, newspaper or airline industries after it became clear they were in some state of decline. The other temptation is to apply subsidies in an attempt to stave off decline or consolidation, but that typically doesn't work, either.
To be sure, you will not find service provider executives talking about such dangers in public. No executive at a public company can do so. They will talk about how well they are managing the transition to a new business model. They will talk about how well all their new services are doing. They will start reporting revenue in new categories to make the point. They will talk about all the new ways they can make money using new business models. None of that is unexpected, or wrong.
But none of it is certain, either. Communications these days is a fundamentally unstable business facing a wholesale replacement of its legacy revenues. By any measure, that would be destabilizing. So service providers must replace lost revenues, grow new lines of business and also cope with changes to the business ecosystem that put pressure on them in new ways, chiefly by severing the historic relationship between applications and pipes (separation of access from application).
In the past, virtually all networks have been application specific. Networks were built to support a key "killer app," and app provisioning and network services were simply parts of that process.
These days, IP networks are open, intended to support all sorts of applications, irrespective of network ownership and operation. No matter what apps develop in the future, access networks will be needed. The difference is that the ability of an access provider to profit from those apps is unclear. That doesn't mean "access" or "transport" are not essential parts of the ecosystem: they remain essential. What is not clear is the economic value those parts of the ecosystem can drive. And that is the issue.
Prudence might dictate that we not place additional burden on businesses that depend on "voice" revenues, any more than we should unnecessarily burden steel, autos, airlines or newspapers. There are problems aplenty.
Perhaps you cannot imagine a world where telcos and cable companies are smaller, relatively more unimportant parts of the application ecosystem. But it has happened frequently in the past, and has happened in the communications business itself. Retail long distance used to be the cash cow that drove profits and subsidized service for the rest of the business. Voice used to be the killer app whose revenues drove the rest of the business. Fixed-line services used to drive the whole business. Obviously, all those conditions have changed, or are changing.
Under such conditions, it might not be prudent to spend too much time or effort intervening in a business that itself already is changing. Intervention inadvertently could accelerate decline and instability. Right now that's the last thing America needs, especially when it appears another round of capital investment is needed.
Gary Kim has been a digital infra analyst and journalist for more than 30 years, covering the business impact of technology, pre- and post-internet. He sees a similar evolution coming with AI. General-purpose technologies do not come along very often, but when they do, they change life, economies and industries.
