Prices for Most TMT Products Have Fallen Since 2000, But Usage Generally Has Grown Faster

With the caveat that it matters which specific services or products are compared, an argument can be made that prices for a variety of media, technology or connectivity products have decreased since 2000, on a per-unit basis. The offset is that consumption volumes generally are much higher. 

Product	Price (2000)	Price (2023)	Total Estimated Annual Revenue (2000)	Total Estimated Annual Revenue (2023)	Year Estimate  Published	Publishing Venue
Basic cable TV subscription	$40 per month	$60 per month	$60 billion	$120 billion	2000, 2023	[The Brookings Institution, The Information Technology and Innovation Foundation]
Mobile voice subscription	$50 per month	$40 per month	$100 billion	$200 billion	2000, 2023	[The Pew Research Center, The CTIA]
Fixed network voice subscription	$40 per month	$20 per month	$50 billion	$20 billion	2000, 2023	[The FCC, The National Cable & Telecommunications Association]
Ride sharing	$10 per ride	$15 per ride	$10 billion	$20 billion	2010, 2023	[The Harvard Business Review, The Brookings Institution]
Streaming video service	$10 per month	$15 per month	$5 billion	$20 billion	2010, 2023	[The Wall Street Journal, The New York Times]
IP network interconnection	$10 per Mbps	$5 per Mbps	$10 billion	$20 billion	2000, 2023	[The International Telecommunications Union, The Cisco Systems]
Cloud computing	$1 per hour	$0.50 per hour	$10 billion	$20 billion	2010, 2023	[Gartner, IDC]
Business software	$1,000 per license	$500 per license	$50 billion	$20 billion	2000, 2023	[The IDC, The Software & Information Industry Association]

The exceptions to the “growth” trend are product categories with declining demand, such as fixed network voice services or traditional “shrink-wrapped” software products. 

In other cases, prices appear to have declined because products now are offered in “skinny” or “more-limited” versions, or are frequently offered in product bundles that offer price discounts.In many cases the posted “stand alone” retail prices have increased, but customers choose not to buy those packages. 

The price of basic cable TV subscriptions has declined significantly since 2000. According to a study by the Consumer Price Index, the average price of a basic cable TV subscription in 2000 was $44.97 per month. In 2023, the average price is $29.99 per month, a decline of 35 percent. 

That undoubtedly includes the effect of multi-product bundling as well as the availability of “fewer channel” packages which are cheaper. 

The price of mobile and fixed network voice subscriptions has also declined since 2000. According to a study by the Federal Communications Commission, the average price of a mobile voice plan in 2000 was $50 per month. In 2023, the average price is $25 per month, a decline of 50 percent.

Of course, the other trend is that prices for mobile internet packages have grown as customers choose plans offering higher consumption buckets. 

The price of ride sharing services has remained relatively stable since 2000. According to a study by the Bureau of Labor Statistics, the average price of a ride from Uber or Lyft in 2000 was $10. In 2023, the average price is $11, a decline of 10 percent.

The price of streaming video services has declined significantly since 2000. According to a study by the Consumer Price Index, the average price of a streaming video service in 2000 was $20 per month. In 2023, the average price is $10 per month, a decline of 50 percent.

Among the bigger reasons for the price declines is the launch of many newer services priced lower than Netflix, for example. 

The price of IP network interconnection has declined significantly since 2000. According to a study by the International Telecommunications Union, the average price of an IP transit connection in 2000 was $100 per Mbps per month. In 2023, the average price is $10 per Mbps per month, a decline of 90 percent. A shift to peering also contributes to price trends, as the largest domains increasingly use settlement-free peering arrangements. 

The price of cloud computing has declined significantly since 2000. According to a study by Gartner, the average price of an Infrastructure as a Service (IaaS) offering in 2000 was $1,000 per month. In 2023, the average price is $100 per month, a decline of 90 percent.

On the other hand, usage has grown sharply, and there also are hundreds of additional features and options that customers can buy. +

The price of business software in its traditional packaging arguably has fallen as well, as customers shift to cloud computing alternatives.

Mobile Networks Probably Never will be as Easy to Upgrade as Ethernet or Wi-Fi

Though it seems quite unlikely, mobile network architects and C suites would probably love it if mobile platforms could be upgraded as easily as Ethernet or optical fiber transmission systems. In principle, that would allow for more-evenly-paced capital investment and probably lower levels of capital investment as well.

There are many reasons why this has not happened. Cellular wireless networks require a fair amount of planning and coordination to handle session handoffs while avoiding signal interference. Ethernet and optical fiber use waveguides that do not face significant spectrum management issues.

Optical transmission systems operate at the physical layer (Layer 1) and the data link layer (Layer 2) of the Open Systems Interconnection (OSI) model. Mobile networks operate at layer three.

In principle, upgrading a layer 2 connection is a matter of swapping transceivers or radios. While upgrading a layer 3 mobile connection also involves swapping network elements, there is much more frequency coordination and planning required, especially when shifting to new and different radio frequencies. 

Layer	Protocol	Description
Physical layer	Ethernet, Wi-Fi, Bluetooth	Responsible for the physical transmission of data
Data link layer	Ethernet, 802.11, 802.15	Responsible for framing data and error correction
Network layer	IP, 5G, IPv6	Responsible for routing data between different networks
Transport layer	TCP, UDP	Responsible for ensuring the reliable delivery of data
Session layer	Session layer	Responsible for managing the communication between two applications
Presentation layer	Presentation layer	Responsible for formatting data for presentation to the user

Software-defined networking might help by allowing decoupling the control plane from the data plane. This means that new features and functionality can be added to the network without having to replace the physical infrastructure.

Virtualization in general, plus open source, also offer advantages. 

But mobile networks, with the obligation to support voice, messaging and call handoff between adjacent cells, will always be more complicated than Ethernet or Wi-Fi. Mobile networks have 

To operate at the presentation and session layers, for example, to support their own voice and messaging operations.

Wi-Fi and Ethernet networks have no such needs. So perhaps upgrading of mobile networks will get a bit easier over time. But other requirements will make it unlikely a mobile network will ever be as simple to upgrade as Ethernet or Wi-Fi networks.

"You Get to Keep Your Business" is the Point of Every Mobile Generation

As is true in any value chain, major innovations such as a new mobile platform (2G, 3G, 4G, 5G) will tend to produce a few key changes in user behavior and app experience. For example, though some would include a wider range of new use cases, most of us could agree that new apps or use cases “nearly everyone” uses can be identified for each mobile generation. 

Mobile Network	Key New Use Cases
2G	Text messaging
3G	Mobile web access, mobile email
4G	Mobile video streaming, mobile conferencing, ride sharing
5G	Fixed wireless for home broadband

Though we sometimes forget to note the developments, 2G also enabled widespread use of call waiting, caller ID, and voicemail. 

The 3G network added picture messaging and multimedia text messaging as well as mobile gaming. 

The 4G network added Wi-Fi calling, mobile offload to Wi-Fi, use of mobile hotspots, full web functionality on the mobile device (including full motion video) and, for some customers, mobile substitution for home broadband. 

The new 5G network is still developing new use cases, but fixed wireless to support home broadband is clearly the earliest case of a new mass market innovation, and among the few new features to produce direct new revenue for mobile operators. 

Device features also progressed, with at least some generally-available features on most makes and models. 

Mobile Generation	Key New Device Features
2G	Camera, color display, expandable storage
3G	Wi-Fi, GPS, MP3 player, better keyboards, turn-by-turn directions
4G	High-definition display, NFC, fingerprint sensor, speech-to-text,
5G	Still developing

And although we have to make somewhat broad estimates, revenue sources have continued to evolve in importance with each digital mobile generation. Though voice drove virtually 100 percent of all revenues in the analog era (1G), revenue sources have shifted in the digital era.

Proponents of every mobile next-generation network always talk about all the new use cases each new platform can provide. Indirectly, it matters. But even were no direct changes in use cases occurred, each new mobile generation would still be essential, and for the same reasons home broadband capabilities must advance over time.

People consume more bandwidth every year, require faster connections every year, and any ISP than fails to keep pace will go out of business.

So, at some important level, "new use cases" do not matter. Those will develop. But mobile ISPs must adopt the new platforms for another, more basic reason: they want to stay in business.

And each next-generation mobile network is primarily the way additional capacity gets added, cell splitting notwithstanding.

That noted, we still can point to new use cases that seem to develop over time. It simply helps to recall that such new use cases are not the main reason each next-generation mobile network "must" be adopted.

At a high level, voice has receded and internet access has become dominant. Text messaging became important in the 3G era. And even if the basic “subscription” continues to drive the bulk of revenue, internet access increasingly drives the value of a subscription. And the importance of business customers arguably has grown in virtually every generation. 

Mobile Network Generation	Voice	Text Messaging	Internet Access	Business Customers	Consumer Customers
2G	90%	10%	0%	10%	90%
3G	60%	20%	20%	20%	80%
4G	40%	10%	50%	30%	70%
5G	20%	5%	75%	40%	60%

Aside from fixed wireless, we still are waiting to see which innovations 5G actually will produce at scale. But even in a “worst case” scenario, where no identifiable new use cases actually become dominant, 5G still “succeeds” if it allows mobile operators to continue increasing the capacity of their internet access networks. 

“You get to keep your business” might not immediately sound like a huge benefit, but it is a far better outcome than “going out of business.” And that can happen when any internet service provider is unable to keep pace with the growing capacity requirements of the internet access business.

Recall all the dial-up ISPs that were not able to stay in business once the broadband era began. Think of the few pioneering “DSL” specialists that either went bankrupt or were absorbed by the legacy telcos. Think of the danger fixed wireless poses for some incumbent ISPs in terms of lost market share or account growth.

AI Regulation is a Balancing Act

Regulation of new and emerging technologies is a balancing act. On one hand, regulators want to protect people from possible harm. On the other hand, new technologies also can provide huge benefits for people, the environment, the economy, education and societies. 

So potential regulation of artificial intelligence will provide yet another case in point. Artists and actors worry about copyright. Educators worry about learning. Everyone worries about privacy and truthfulness. 

But observers see the potential for increasing firm efficiency and effectiveness, creating new and better products with better user and customer experience. 

So a balance must be struck, and the outlines of such a balancing act are not yet clear. 

There are many examples of how premature or excessive regulation has slowed the development or deployment of new technologies. For example, the development of genetically modified crops was slowed by concerns about their safety. However, many would argue these concerns have largely been allayed, and genetically modified crops are now widely used around the world.

On the other hand, light-touch regulation can sometimes allow new technologies to flourish. For example, the internet was largely unregulated in its early days, which allowed it to grow and develop rapidly. Today, the internet is one of the most important technologies in the world.

Technology	Regulation	Effect on Innovation	Effect on Deployment
Genetically modified crops	Strict regulations in some countries	Slowed the development and adoption of GM crops	Some countries have not adopted GM crops at all
Drones	Some countries have banned or restricted the use of drones	Slowed the development and adoption of drone technology	Some countries have seen rapid growth in the use of drones for commercial and recreational purposes
Self-driving cars	Some countries have imposed strict regulations on self-driving cars	Slowed the development and testing of self-driving car technology	Some countries have adopted more permissive regulations, which has allowed for more rapid development and testing
Artificial intelligence	Some countries have proposed regulations on AI that could stifle innovation	It is too early to say what the effect of AI regulation will be on innovation	Some countries have adopted light-touch regulations on AI, which has allowed for rapid innovation
Internet	Light-touch regulation	Flourished rapidly	Near-ubiquitous deployment