My point is that even if the level of competition within Internet industries is equal to that of their brick-and-mortar counterparts, the typical Internet company will earn lower margins on sales because it will have lower costs.

Grocery chains, for example, are famous for their low margins, only pennies on the dollar. They often trumpet this fact to demonstrate how competitive their market is. But these low margins have nothing to do with any extra competitiveness of the grocery industry. They are due instead to the low value added per dollar of sales. (Restaurants have margins 10 times those of grocers, even though the restaurant business is just as competitive, because chefs add so much value to the offerings.) The grocery business is the final distribution stage of the food industry, after most of the value has already been created. Low-wage clerks scan the price into a computer after consumers carry the product to the checkout. Since the value added by the supermarket is very small relative to the price of most of the items, the theory correctly predicts that the markup will be very small.

More generally, in competitive markets, if companies in industry A have a smaller investment relative to sales than those in industry B, then the margin on sales will be smaller in industry A.

Which brings us to the denouement of this little story. If Internet retailers have lower costs than brick-and-mortar retailers, there are two possible long-term outcomes, but both of them portend that stores on the Net will have lower margins than brick-and-mortar companies.

First, let’s assume that consumers find brick-and-mortar companies and Net companies to be perfect substitutes for one another. So, for example, consumers might be completely indifferent about whether to buy a book online or in a physical store. As a result, all the brick-and-mortar booksellers will be driven out of business by the more efficient Internet sellers, just as horses and buggies were driven out of business by automobiles. This may take a while, but eventually the industry would consist only of Internet booksellers, and their margins would be lower than was the case for the previous generation of booksellers since their costs, investments and value added per dollar of sales are lower.

Far more likely, in my opinion, is a scenario where some consumers prefer to buy online and others prefer to buy in brick-and-mortar stores, just as some consumers prefer shopping by mail order and others by traversing shopping malls. Competition in each segment will occur largely between companies in that segment, just as convenience stores largely compete with other convenience stores and supermarkets with other supermarkets. Assuming that each market is competitive, every drop of excess profit will be squeezed out of the market for the typical company, but since the investment per dollar of sales is lower for Internet companies, so too is their average margin on sales.