The Americans will always do the right thing. . .after they've exhausted all the alternatives.

--Winston Churchill

One alternative that cannot be exhausted soon enough is banner advertising. I have been eager to see this concept die since it first was introduced in 1995.

Another alternative that I believe should be euthanized is the subscription model for individual periodicals. The marginal cost of distributing the publications online is zero, so subscription models are very difficult to sustain.

Finally, there is the alternative of micropayments, meaning small payments for access to particular slices of information. I am now persuaded by Clay Shirky's argument that the mental transaction costs involved in micropayments are too high to make micropayments workable.

The implausible alternatives all are based on the notion of a content "silo," which is left over from the pre-Internet era. Content is embedded in a silo, such as a traditional magazine or newspaper. This silo is presented to the reader in "take-it-or-leave-it" fashion. The funding comes from specific subscriptions and advertising.

Arriving at a viable economic model for content on the Internet requires changing the "silo" mentality. There are a number of things wrong with it.

First of all, the "silo" model tries to maintain an anachronistic wall between the content in one silo and content in other silos. In the world of physical magazines, it certainly makes sense that a subscription to "Business Week" does not entitle you to read "Forbes." Clearly, they are two separate physical collections of paper.

On the Internet, however, this distinction is not a physical necessity. Most consumers in fact pick and choose articles from a variety of online magazines. In contrast to the physical world, consumers can engage in extensive content aggregation without imposing meaningful costs at the margin.

A second thing wrong with the traditional silo model is that it ignores a critical component of value. Because of the volume of content available on line, the marginal value of additional articles has gone down. The marginal value of annotation, filtering, and recommendations from other consumers has increased. As J.C. Herz put it in an article in "The Industry Standard," the "digital future is probably in the margins, rather than on the pages."

For example, one of Amazon.com's recent innovations consists of book lists created by readers. These lists create connections among various books. The lists are an example of value created in the margins.

Instead of trying to stay within the silo structure, online journalism should evolve toward broad content aggregation with the value of annotation and connections. This will maximize the value of the online component.

For an economic model, I continue to recommend the idea of "clubs." A club would provide content aggregation, recommendation, and annotation services. Journalists would be paid by clubs, rather than by individual publications. For a consumer, joining a club will provide access to value-added services relative to online content.

Most of the articles that you are able to read when you join a club may be freely available without joining the club. What your membership fee would give you is better access to individual authors, as well as to indexing tools and cross-reference tools. Some of these tools would be provided by community members, as in the Amazon book lists.

The raw content is not what you are paying for. The haystack is free. But if you want help finding the needle, you have to join the club.

While I would not pay to subscribe to an individual online journal, I might be willing to pay to join a club that gives me access to a variety of journals as well as to helpful annotations. Annotations might consist of Lawrence Lee's recommendations for articles on Internet marketing or Virginia Postrel's comments on articles about policy issues.

A club could offer several different levels of membership. The most expensive membership could entitle you to personal chat time with famous authors. Or it could entitle you to 24-hour response time to inquiries that you submit to human experts. Or it could entitle you to use the most sophisticated indexing and cross-reference tools.

In a silo model, I only get one periodical's content when I subscribe. With the club concept, my subscription includes access to many periodicals. Some of these are in the haystack (that is, available to people who are not club members), but perhaps not easy to search and cross-reference. Other articles might be licensed only to clubs, so that individuals who are not members would not have access (or have access only with a lag).

How would journalists and other content providers get paid? They will be paid by the clubs. They may be paid for writing articles that the club's managers know will be particularly interesting to club members. They will be paid for permitting their articles to be searched by the club's indexing tools. In addition, writers may be paid for making themselves available to club members for chats or responses to questions.

Online content providers are trapped in the offline metaphor of a magazine, which is a separate silo funded by a combination of subscriptions and advertising. When they have exhausted the alternatives associated with the silo model, perhaps they will come around to the idea of a club.