What’s that, you say? What about the crazy valuations of Internet startups like Amazon.com and Yahoo?

All temporary and all based on, among other assumptions, the false belief that the advantages of the Internet will translate into higher margins. I’m talking about after the smoke has cleared from the current Internet gyrations and discombobulations—what economists refer to as a "long-run equilibrium."

The perception on Wall Street—and just about everywhere else—about the Internet generating more profits per unit of sales is simply wrong. The margins for Internet businesses will be lower, consistently lower, than for brick-and-mortar operations.

The reason, though it seems paradoxical, is that the Internet lowers the cost of doing business. Wall Street analysts like to drool over Internet startups because they see that virtual storefronts are less expensive to create than the real thing. These storefronts do not require real estate, plumbing, showrooms, dressing rooms, heating or air conditioning. Virtual storefronts can be scaled up without having to hire additional employees and build many locations to house them.

Most analysts have assumed that because companies doing business on the Net will have lower costs, this will translate into the ability to generate higher markups and higher margins than their brick-and-mortar counterparts. They often apply these margins to projected sales figures to determine future profits and investment potential. At one level this seems to make sense. After all, companies that have managed to achieve a cost advantage over their competitors, everything else equal, do earn higher profits and margins.

But what is true for individual companies is not true for markets consisting of many companies.

To understand this we have to dig a little deeper. Economists consider markets competitive when easy entry keeps the typical company in the industry earning only normal returns on investment—in other words, returns unimpressive enough to keep potential new market entrants interested.

If all the companies in a competitive industry achieve a reduction in costs, their profits, after an initial and temporary rise, will return to normal when new capacity and new entrants suck up any excess profits. Memory chip manufacturers, for example, constantly have falling costs, but new investment in fabrication plants at every uptick in profits eventually pushes profits back to the low levels that plague very competitive industries.

How do we know this model of competition is not just some aberrant creation from academics who’ve spent too long in their ivory towers? Because there is overwhelming evidence to support many of its implications, including the one that industries with lower costs per dollar of sales also have lower margins.