Home / Blogs

Moore's Law and the Economics of Abundance

Tom Evslin

Moore's Law explains why the price of everything electronic keeps going down; but now Moore's Law is starting to have an effect on much more than technology prices. The costs of energy, medicine, law, education, financial transactions, and government itself are falling because of Moore's Law's relentless progress. But these cost decreases are not being fully reflected in the prices we pay for all these things. Both government and private sector monopolies work hard (and sometimes together) to prevent plunging technology prices from cutting into the rents they've become accustomed to collecting.

Quick primer on Moore's law; feel free to skip this paragraph if you don't need it.

Gordon Moore, co-founder of Intel, observed that process improvements resulted the number of transistors which could be placed on a chip doubled roughly every eighteen months while the cost of making these chips stayed the same. In the real world this means that cost of computer memory and computing power comes down 50% every 18 months. For example, in 1968, I was responsible for the purchase of 256,000 bytes of memory for a mainframe; it cost $100,000 or $382 for each thousand bytes. Today on Amazon you can buy an 8 Gigabyte (8 billion bytes) memory card for $11.93. Ignoring the free reader which Amazon throws in, this comes to $0.0000015 for each thousand bytes. If you do the math, you'll find that, in this case, prices went down even faster than Moore predicted. And I've ignored inflation.

Data communications prices plummet

How much data you can transmit for a given price in a given period of time depends on how much of it can be squeezed into a copper, fiber, radio wave or other kind of pipe. The cost of data squeezing depends on the cost of the computer power which does the squeezing. So it shouldn't be any surprise that data communication costs follow Moore's law down further and further. In 1996 you paid $19.95 for the ability to send 32 kilobits (roughly 32 thousand bits) of data per second over a dial up phone line continuously, about $.61 per thousand bits per second. Today I have a $39.55 connection which lets me transmit 10 megabits (a little more than 10 million bits) per second, $.004 per thousand bits per second. This improvement, astonishing as it may be, is actually slower than Moore's law would predict.

Prices for talking also affected

Plain old phone calls within the United Sates used to be distance and time sensitive and cost up to $1.00/minute. You can pay those rates today if you make the mistake of using a telephone in an expensive hotel. Now unlimited domestic calling is part of most landline plans and, even on cell phones, doesn't cost much. Similarly international calling rates to places like China and India, which used to be priced in dollars, per minute now are priced in pennies. In a strange reversal of value, video calling on Skype from anywhere there's a broadband connection to anywhere else is free (once you pay for the computers and Internet connections). On the other hand, if you take your US cell phone abroad and call or get called from home, you're back to over a dollar per minute if you're not careful.

What's going on here? You can transmit and receive huge quantities of data on the Internet without worrying about how far it's going or whether it crosses national borders. The amount of data necessary to represent a voice call is laughably small compared to a graphic, a video, or even high definition sound. Why should this small amount of data carrying capacity be priced at such high rates?

The answer is toll booths! The old voice telephony structure consisted of national monopolies — some owned by government, some government sanctioned. These monopolies became enormously profitable for their owners; the value of voice communication is high so high tolls could be charged. At first high tolls were fair compensation for the high cost of laying undersea cable. But the costs stayed high even as technology running at Moore's Law pace kept doubling and redoubling the amount of data which could be stuffed into a cable. Replacing cable with fiber obviously required more capital; but the capacity of the fiber was so high that capacity was no longer constrained. Volume didn't grow to fill the capacity because the national monopolies conspired with the help of their governments and a UN body called the ITU (International Telecommunication Union) to keep international calling rates high.

The Internet happened. And those of us in the nascent Voice over IP industry realized we could move data as voice over these new channels. We were more efficient than the traditional carriers at packing voice data into the fiber which made up the Internet; but, more important, we could bypass the toll booths which had been set up at national borders and the ITU rules for keeping prices high. International calling rates collapsed like a pin-pricked balloon. India and China fought at first to keep the high revenues their national monopolies earned (and the huge bribes paid for the right to interconnect). But both those countries realized that they were sacrificing their whole economies to protect telecom revenue and loosened up. Cheap communication has been a huge enabler of their economic booms in a global economy.

Skype video calls can be free because they don't use any of the voice infrastructure which is still under the control of the descendants of the old monopolies. Most of the landline infrastructure in the US belongs to the reconstituted heirs of Ma Bell. Governments initially allocated radio spectrum for cell phone service in a way which assured that duopolies would control most markets for cellular service — hence relatively high cell phone rates. As lobbyists know, government can be a great help in maintaining artificial scarcity and protecting monopoly rents.

How about financial transactions?

Well, the cost of trading stock has gone from dollars per hundreds of shares to fixed fees of almost nothing for trading unlimited quantities. The government did a good job of breaking up the cozy old boy networks and fixed transaction costs. Technology reduced the real cost of trading. The market worked like it should have and prices for trading plummeted. Unfortunately bankers found some other ways to collect rents, but that's another story for another day.

Back in the 1970s I was hired to write some software to move transactions electronically between banks in order to reduce the volume of checks being flown to clearing centers every night on little jets (how FedEx got its start and its name). Part of the specification was that it took one day to clear a local transaction; two days for a regional one; and three days if the transaction was inter-regional. "It's gonna cost more in programming to hold all these transactions than just clear them instantly," I said naively. "I mean electronic travel doesn't take days."

The answer was — and still is — that the banks get interest on the float — the uncleared checks. The Federal Reserve decreed that the banks keep this benefit even though the "checks" no longer existed. In doing this they also passed up the very real value of knowing quickly that there are insufficient funds to cover a transaction. Couldn't be a better example of government action to preserve rents for favored constituents.

Medicine, energy, government, education and more

I'll be writing more about costs which technology is driving down — and attempts, often successful, to keep prices high regardless. Just think, if the price of medical technology went down as fast as the price of consumer technology, a cat scan would probably cost a nickel by now!

I'll also be speaking about Moore's Law and the Economics of Abundance at the Ethan Allen Institute's Sheraton Economic Series at 7:00PM on March 30 (this Wednesday). It's free and no reservation required. Details here.

Follow CircleID on

If you are pressed for time ...

... this is for you. More and more professionals are choosing to publish critical posts on CircleID from all corners of the Internet industry. If you find it hard to keep up daily, consider subscribing to our weekly digest. We will provide you a convenient summary report once a week sent directly to your inbox. It's a quick and easy read.

I make a point of reading CircleID. There is no getting around the utility of knowing what thoughtful people are thinking and saying about our industry.

Vinton Cerf, Co-designer of the TCP/IP Protocols & the Architecture of the Internet

Share your comments

The exception Alan Wilensky  –  Mar 29, 2011 1:54 PM PDT

Ah! You missed one anachronism: EDI is still priced in several pennies per KC - ostensibly because there is a higher standard of bailment (responsibility to deliver) and higher 'touch' support services for the networks of trading partners on a VAN or B2B Hub.

I suggest you discuss opportunity cost Mike O'Donnell  –  Mar 30, 2011 9:00 AM PDT

Thanks for the summary of exponentially decaying prices, and the market frictions that can delay our enjoyment of the decay.

I encourage you to go on and discuss the economic implications of the price decay itself.

A particularly important consequence of Moore's law for computing power: in many cases opportunity cost dominates rational planning. Whatever money I spend for computing equipment this year prevents me from getting much more with the same money next year.

In some cases, the quickest way to solve a particular computing problem posed in year nnnn, with particular financial constraints, is to do nothing for several years, then solve the problem on the faster equipment available later.

Even when a problem can be solved by computation on current equipment, the value of a quick solution must be very high to justify a large immediate purchase. I note that two key applications of "supercomputers" have been movie animation and weather prediction, where the value of a solution decays very rapidly with time, sometimes faster than the Moore's law decay of the computing cost.

It would be very interesting to learn more about the particular ways in which the dominance of opportunity cost, and/or the application to tasks whose value delays faster than the exponential decay of the cost, affect these other commodities subject to Moore's lawish behavior.

Mike O'Donnell

Pretty thoroughly wrong Richard Bennett  –  Apr 04, 2011 9:32 PM PDT

Moore's Law is a wonderful thing, but it doesn't apply to specialized analog chips or to trenches and towers. I've been forced to take this article apart: http://www.hightechforum.org/misunderstanding-moores-law/

To post comments, please login or create an account.



IP Addressing

Sponsored byAvenue4 LLC

DNS Security

Sponsored byAfilias

New TLDs

Sponsored byAfilias


Sponsored byVerisign

Domain Names

Sponsored byVerisign