Moore's Law and Diminishing Economic Returns

I learned a great deal over the weekend about how to get a great deal on a new computer. I’m going to pass this information along to you. It has to do with Moore’s law.

Gordon Moore is a co-founder of Intel, the microcomputer chip manufacturing company. Intel has dominated chip production for a generation. It invented the first silicon chip with embedded transistors. In 1965, Moore observed that every 18 months, chip capacity doubles. Nothing else in nature or society doubles this rapidly.

Raymond Kurzweil invented speech recognition software. He has looked at the evidence and has concluded that this process has been going on since 1910. Calculating power per dollar doubled every three years from 1910 to 1950. It doubled every two years from 1950 to 1965. It doubled every 18 months from 1966 to 2000. It is now doubling every year. In other words, the rate of change is accelerating.

If this rate of increase continues — and Moore himself is highly skeptical about this — then by 2023, the computer chip’s capacity per $1,000 will equal the human brain. By 2036, one penny will buy this much capacity. In 2049, we will be able to buy a chip with the capacity of the entire human race’s brains for $1,000. In 2059, this will cost one penny.

The growth of computer capacity is rising so fast today that computer manufacturers are now finding it difficult to deliver new machines at a profit. The machine becomes technically obsolete too fast. The buyer knows that the latest and greatest machine will be unsalable within months. On the used computer market, it will sell for a fraction of its price today. It pays the consumer to wait to buy.

Yet, worldwide, the computer market continues to grow. Price competition is bringing the price of computer power into the budget range of millions of Asians. Americans may be jaded, but Asians aren’t. For them, the computer revolution is barely beginning. The Internet beckons.

Asian governments, especially China’s, see what is coming: unofficial information and opportunities that they would rather not share with the public. The days of the information gatekeepers are numbered. This has never happened before in human history. “Every man a publisher” is becoming a possibility.


One of Keynesianism’s arguments against price deflation is that it is bad for business. Consumers know that lower prices are coming. It pays them to wait to purchase anything. It makes them smarter shoppers. So, the Keynesian says, consumer demand falls despite falling prices. The recession gets worse. So, as soon as it looks as though monetary policy is moving toward stability — a rare event — mainstream economists and central bankers start worrying.

What we have seen with computational power since 1910 is a standing testimony against any argument against deflation. Price competition broadens the market, attracting new buyers. Example: it looked in 1960 as though American phone companies would soon saturate the market: one phone line per home. Now look at the phone market: an Internet line, a regular phone line, a kids’ phone line, and maybe a small business line. Then there is the cell phone. This doesn’t count cable TV. As for businesses, they buy lots of phone lines.

Falling prices don’t scare off buyers; they attract buyers. When one group of buyers delays buying, content with their existing units, another group rushes in to get what its members could not afford before. They don’t want to wait. They have waited forever. They want in on the deal. From Model T Fords to the latest desktop computer, falling prices increase demand.

That’s why all the hand-wringing over the threat of deflation — for which there is very little statistical evidence — is so much wasted worry. We should be so fortunate!

To understand the benefits of falling prices, let me tell you about my experience in buying a new Dell computer.


I have problems with my old computer. This will sound crazy. It won’t recognize a modem. Anyway, it won’t recognize it after I bring the machine home from the repair shop. In town, it hooks up to the Internet with no problem. Then I bring it home. I turn it on. It works. But if I turn it off and then turn it on again, it loses all trace of any modem. I have to take it to town again. This has been going on for two weeks. Nobody can fix it.

Because of the falling price of computers, a computer is now a throw-away item. My unit is a 1998, obsolete, 500 megahertz machine. The cheapest new desktop machine is more than twice as fast. Most are four times as fast.

The computer companies are facing intense competition today. Small companies can go on Ebay and sell new machines without monitors for about $500. The units come with warranties. The growth of on-line purchasing is unstoppable. The information cost of finding a new machine keeps falling.

Falling information costs are now a major factor in pressuring retail computer sellers to sell lower. What I found last weekend about how Dell sells was a revelation to me.

On Saturday, I read through a Dell Small Business catalogue dated September 2002. On page 5, there is an ad for a nice little desktop computer: the Dimension 4500S. It has a small vertical case no taller than the monitor, and half as wide as a conventional case. It’s a 2 Ghz machine that uses a Pentium 4 processor, a 20 GB hard drive, a 15″ monitor (13.8″ viewable). Price: $599. Not bad. Even better: there is a $100 rebate.

For an extra $70, you can buy a CD/RW (read/write) drive. For $50, you can upgrade to 256 MB of SDRAM. That’s a total of $619. If you order on-line before Sept. 25, shipping is free.

Now comes the fun part.

I called. I was told that this machine is no longer being manufactured. But I could order a Dimension 2300.

I had been on the Dell site earlier last week. The 2300 is a more expensive machine.

I asked what it would cost with the features I wanted, but without a monitor. The salesman refused to give me a price until I supplied my name and address. He said this was because of varying shipping costs. He had to know where I live. So, he planned to charge me for shipping. But the catalogue announces: “FREE Ground Shipping.” His sales pitch assumed that the caller wasn’t responding to the catalogue, even though the phone number I used is provided at the bottom of each page in the catalogue. After I gave him this information, I was told that I could buy the entire package for $798. This included a $100 rebate.

If I figure that my $120 in upgrades would have been offset by the absence of a monitor, as well as cheaper shipping, Dell was trying to get me to shell out an extra $299 over the $499 rebate catalogue price. That’s a 60% price increase. I passed.

Then I went to Dell’s Web page listed in the catalogue:

I found that I could still order the model I wanted. It’s not out of production after all. In fact, if I ordered it on-line, I could skip the $99 shipping fee. I bought the machine with the monitor and the upgraded SDRAM and the read/write CD drive for $650, delivered. I really don’t need the monitor, but I can use it.

This model doesn’t have a modem, but I prefer an external modem anyway. With an external modem, I can see when it’s connected to the Internet. It has lights.

Dell is hurting. To pay for its telephone sales force, Dell appears to have adopted a familiar sales technique. It used to be called “bait and switch.” I was told that the model I wanted was no longer available, despite the fact that I was ordering from a September catalogue. “The catalogue was printed three months ago,” the salesman told me.

As I learned ten minutes later, the model was still available on-line at a special price: a $100 rebate and no shipping.

The salesman was obviously using a different sales pitch for old-fashioned buyers who still prefer to buy over the phone. Those who buy on-line are more price sensitive. They must be offered a better deal. They know that Ebay or other sources are out there.

Dell is taking advantage of information differences among different classes of buyers. The company is selling higher-priced machines to people who have poorer information skills. The buyer over the phone is more easily upsold by a salesman with a story: “I’m sorry, that model is out of production.”

When I decided to pass on the higher priced model, I made a mental decision not to buy from Dell again. I decided not to trust Dell’s catalogues. That meant that I would no longer trust their phone salesmen. But then I thought, “I wonder if I can beat the system. Can I use on-line ordering to get what I want at a price the salesman says is unavailable?” I found that I could.

Dell got my money. I ordered the machine that I wanted. But I learned a lesson: price competition is cutting profit margins at Dell to such an extent that different classes of customers are treated differently in terms of information.

Now you know. Don’t order from Dell by phone. Order on-line. But before you do, check here to see the “steal of the day.”

The same thing is happening to the airlines. On-line ticket ordering is making highly price competitive deals available. The deals offered on the phone are not as good.

The large carriers have always made their profits off their business travellers. Their profit margins depend on three or four high-priced business travellers per flight. Now the discount carriers with low overhead are making low fares available to business flyers. The major airlines must match the offers to remain competitive. The majors are getting killed. One by one, the airline transport unions are biting the dust. They are having to back off. Price competition is fierce. The Web is the great equalizer. Yet it’s the Sabre reservation system that the Web sellers use. American Airlines developed the Sabre system. American Airlines has laid off 20,000 workers.

This is good for consumers. But the majors are howling. “If we go out of business, consumers will face higher prices!” There are still congressmen who are calling for the re-regulation of the airline industry for the sake of “maintaining competition,” and the majors are in agreement. The producers’ hatred of price competition is always with us.

Price competition works for the consumer. If the old-line, high-overhead sellers get hurt, too bad. The free market is structured in terms of consumers’ sovereignty, not producers’ sovereignty, a fact that outrages older producers with established markets. Price competition erodes established markets. This is why they hate price competition. It lets newcomers beat the Establishment to death. This is why price competition gets a bad press — but not on-line. The bad press is the paper-based press that survives on advertising paid for by established firms.

Fortunately, the home computer industry was never regulated by the Federal government. It came out of nowhere before Congress ever got its hooks into it.

Dell faces competition from no-name producers on Ebay. Now that computers are throw-away items, buyers can risk buying on Ebay. This is great for buyers and tough for producers with high overhead. That’s the way it should be. Michael Dell became a billionaire by offering a price-competitive product line to IBM’s PC’s, beginning in 1983. Now it’s his turn to scramble.

When sellers scramble, buyers win.


My need to buy a new computer is not high. Had it not been for that “no-modem” glitch, I would have stayed with my 1998 Dell model. For accessing the Internet, I don’t need anything faster than a 1998 computer. I won’t in two or three years, either.

Most people use five or six pieces of software: a word processor, an e-mail/browser program, and a utility program or two. Most of these programs are cheap: under $50. Some are free. The most important utility programs that I use are Norton AntiVirus and a $25 program called TextPad, which strips out HTML code and creates clean ASCII files out of Web pages. I also use ZoneAlarm, a firewall program. (If you use a cable modem, download it.) I use Cookie Cutter. I don’t add programs often. I’m still looking for a good free-form data base for Web pages, e-mails, and notes. With these programs, I don’t need any more computer speed. I don’t use CAD/CAM programs or digital movie editors. I don’t use a spread sheet.

Yet the chips keep getting faster. The chip producers are not being driven by consumer demand to make these improvements. They are being driven by the fear of falling behind their competitors. Consumers are not pressuring computer companies to produce ever-faster machines. This is why advertising revenues have collapsed. Computer magazines are thin today. Ziff Davis Publishing is no longer a cash cow. Computer buffs are not reading every page to find out what the latest and greatest is. The increase in computer speed is not matched by an increase in productivity. The return on investment for buyers is low. The law of diminishing returns has set in.

The result is a reduction of the rate of economic obsolescence for old machines. Rapid technical obsolescence does not produce rapid economic obsolescence. Old machines can do the grunt work of new machines with no noticeable reduction in efficiency. Updated software doesn’t run that much worse on a 1998 machine. In 2005, that will still be true for most machines and most operations. Your old machine can’t be sold for much money, but it need not be replaced. Since you can’t get much for it, you might as well keep it.

When a computer breaks, it should be replaced. Fixing it is a waste of time and money. If it can’t be tweaked and put back into production in a couple of hours, replace it. But computers rarely break down. I’m typing this report on a 166-Mhz Dell that has to be seven years old. I keep it separate from my Internet machine, for virus reasons. I use a 1983 PC AT keyboard. My keyboard is worth five times to me what my machine is. It’s not digital. It holds its value. For WordPerfect for DOS users, it actually goes up in value. Newer keyboard models are less productive because of the function key placement.


Some of you are aware that the government uses what is called a “hedonic index” to measure increases in output. Quality changes that are said to be the result of increasing computer speed are used to judge the productivity of the economy. But as computer speeds accelerate, and as output that results from increased computer speed ceases to increase, the hedonic index becomes increasingly a statistical game that makes the economy look better. What we want is cheaper food, cheaper car repairs, and cheaper housing. What we get are faster computers, which we no longer need. Yes, Asians benefit, but the economic index is for Americans.

Moore’s law is still operational. Computer chip density doubles every 12 months. If this were matched by an increase in general productivity, we would all be as rich as kings in two decades. But the law of diminishing economic returns has set in. Output is no longer rising rapidly as a result of faster chips.

The government’s statisticians are busy making the economy look as though it’s growing faster than it really is. Imperceptible increases in output are being counted as significant because of the measurable increase in chip speed. Meanwhile, consumer debt rises and corporate debt rises even faster. What is accelerating is the debt/output ratio. It takes more and more debt to achieve one percent economic growth.

The U.S. economy is climbing a wall of debt. The law of diminishing returns is making itself felt in the debt/output ratio. It takes ever-larger infusions of credit to keep the system running.

The statisticians can play all the games they want. The fact is, the stock market is heading down. So are corporate profits. So is corporate investment. Businesses profitability is kept low by fast-depreciating computer technology. It no longer pays to keep upgrading. By heralding high tech investing as the wave of the future and the key to future growth, the touts — Alan Greenspan is chief — persuaded investors to buy shares, or at least not sell them. But the investment in high tech has not produced comparable productivity. Businesses have stopped making these huge commitments that depreciate rapidly and do not lead to greater profit.

This is why the NASDAQ really is dead. It isn’t coming back. The mania of high tech investing is over. All that remains are the depreciation schedules. The best you can say is that there is no compelling need for businesses to upgrade. That spells years of low growth for technology companies.


I have been working since July, 2001 on a manual. It’s for an industry that is highly profitable. It is essentially a mom & pop industry. It is immune to would-be Wal-Marts. A friend of mine got into it in 1991 with a $35,000 investment. He took in $1.4 million in fiscal 2002. The real estate side of the operation has increased his net worth by $500,000. It will be $1 million in ten years. Yet he can run his entire operation with $1,000 worth of used computers. All he needs is one copy of QuickBooks Pro per location, a $200 program. He could run this on a $100 used computer.

Eventually, I will go public with the full story. My point here is that this industry is pure service. It is low tech. Yet it is highly profitable. I look at the inventory expenses of a business like auto parts or toys or retail book sales, and I see huge outlays. Yet there are lots of little service businesses that can do quite well without inventory or computerized delivery systems.

I know very few people whose livelihood is dependent on Moore’s law. At the same time, I see industries that are high tech with razor-thin profit margins. The airlines are the best example, but so are steel and autos. Price competition is ruthless. It spares no producer’s feelings. It respects no one’s monopoly. The Web is the most price-competitive institution in history. You had better have a niche market if you want to survive. I can do no better than to quote Norm Brodsky: “Barriers to Entry,” Inc. Magazine (Oct. 1, 2001).

Given a choice, moreover, I’d always go for a hard-access business over an easy-access one.

Why? Because in an easy-access business, the product or service you’re selling will eventually become a commodity, if it isn’t one already. Not that there’s anything wrong with commodities. You’re just severely restricted in what you can charge for them, so you’re forced to operate with much thinner gross margins than those you’d enjoy in a business that is more difficult for competitors to enter.

That’s why I pay through the nose to get an electrician, plumber, or common skilled worker to make house calls. Skilled workers with low-tech tools can name their price. We have sent our children to college. The best and the brightest are earning sociology degrees at good universities and getting entry-level $25,000 a year jobs. But journeymen plumbers are living well. There is a lesson here. There is no prestige in cleaning out septic tanks — just a lot of money.


The biggest firms are struggling. They have adopted high technology to secure a cutting edge. But he who lives by the cutting edge dies by the cutting edge. These huge firms have used high technology to give them an advantage, but, year by year, Moore’s law is lowering the barriers to entry. The high tech advantage must be paid for. Front-runners must keep shelling out big money to maintain their advantage. Intel is the model. As a new technology spreads, more newcomers can get into an industry. The cost of maintaining a significant lead is reaching the law of diminishing returns. This is great for consumers and bad for “buy and hold” investors.

Digital technology is a democratizing force. It keeps getting cheaper: Moore’s law. As digits get cheaper, more of them are demanded. More users of digital technologies become skilled. This is why the return on invested capital keeps falling. American companies are seeing their market share taken away by Asian firms. This will not change. Digital technology spreads. It doesn’t respect borders.

It goes to the highest bidder.

Trailing edge technology is becoming so powerful technologically that its ever-lower price overcomes the advantage that cutting edge technology once provided. What is happening in desktop computers is happening in every area of the economy. High technology no longer secures a long-term stream of income for producers.

Half of the Fortune 500 firms in 1980 were not on the list in 1990. If you bought and held, you survived only because the stock market rose in response to retirement money pouring into a narrow market. Those days are over for our lifetime. Now, you must buy the right stocks. Buy and hold won’t do it for you any longer.

Respect Moore’s law. It takes away advantages that were once thought to be secure. To the extent that any industry’s barrier to entry is digital, that industry is now on the defensive. The days of digital wine and roses are over. It’s hangover time.

September 27, 2002

Gary North is the author of Mises on Money. Visit For a free subscription to Gary North’s twice-weekly economics newsletter, click here.

Copyright © 2002