ECONOMIC MODELS

The economic issues facing digital libraries are huge, but fall into a small number of categories:

1. how to pay for the cost of creating and operating them
2. how to pay for the necessary infrastructure (e.g., networks, Internet backbone, multimedia displays)
3. how to pay for content

Charging Structures

The first problem is how to assemble money to finance digital libraries. The second is how a portion of the surplus is to be distributed to content owners. The following are possible models for the financing of digital libraries:

• Microcharging. Because digital libraries are mediated by computer, it is possible to obtain precise data on each type of usage made of material. A charge can then be imposed on the user and paid directly to the content owner. While such a mechanism may have to deal with huge number of transactions each of very small size, it is not unprecedented. After all, telephone companies routinely measure usage, although they have the advantage of having billions of dollars in capital tied up in the measurement equipment and software. The real problem with microcharging is that it dampens consumer interest in material. If one has to pay for each screen one views, one is motivated to browse less. Instead of learning, which is the primary reason for using a DL, one must spend time worrying about how much the material will cost. An effective plan should eliminate this effect.
• The utility model. A first step away from microcharging is the electrical utility model. Each use is measured in "clickls" (to use a phrase coined by financial writer Andrew Tobias), and the user pays a monthly bill based on (but not necessarily proportional to) the number of clickls. As with an electric bill, discounts can be offered for higher usage and payments spread out over the year.
• Subscriptions: the HBO model. A problem with the utility model is that there is no cap on the potential cost of information. More use means higher fees. The solution used by Home Box Office is to charge a fixed rate per month for unlimited usage. Heavy users are subsidized by the occasional users, all of whom pay the same amount. Such a structure allows users to access content freely without worrying about costs, but generates very large revenues to be divided among content owners. How the division should be accomplished is a matter for separate debate.
• Public subsidy. A government may decide that access to information is so essential a component of ordinary life that it makes such access "free" to all citizens. The word "free" is in quotation marks because it of course is not free, but is paid for invisibly through taxes. An example is the U.S. Interstate Highway System, financed by the federal government through tax revenue. No charge is imposed (directly) on individual users because the system is thought to be so essential to U.S. daily life and commerce that tolls are regarded as detrimental.

Creation and Operation

It is difficult to assess the relative merits of the charging schemes discussed above because of a severe lack of real data. Japan is now engaged in a large-scale experiment to understand the economics of DLs. It's going to take time, however. When asked how much it costs to digitize a scientific journal, officials at the Nara Institute replied that they had no idea. This suggests that the Nara Institute of Science and Technology (NAIST) was given a mandate to implement a DL, largely ignoring the cost, with the idea that its economics would be studied later, once the technology has matured to a point that meaningful measurements can be taken. It is difficult to imagine a U.S. funding agency taking such an approach; more likely it would insist on an advance study to predict the savings to be realized.

Priorities

Which works are to be digitized first? If everything will ultimately be digitized, it is tempting to say that order does not matter, but digital libraries must achieve a critical mass in order to be useful, attract additional funds, and grow. The matter of prioritizing is difficult because of competing objectives. These objectives include:

• Offering essential materials. Users expect a DL to have what they want, or they will turn to other sources for it. This suggests that the best use of resources is to input the most commonly referenced books, such as dictionaries, almanacs and encyclopedias.
• Gathering the classics. The idea of a large library without Moby Dick is laughable. However, that novel can be found in every library and bookstore in the United States. It is unlikely that anyone would read it sitting at a CRT screen, so is it important to digitize it? The answer is yes. People expect libraries to offer certain items, and the decision to omit classics because they are available elsewhere is harmful because it represents the erroneous value judgment that introducing the essential works of a culture can be postponed.
• Preserving fragile materials. We cannot afford to wait until manuscripts decay beyond recognition before scanning them.
• Archiving one-of-a-kind items. Of certain critical materials, such as the Dead Sea Scrolls, only one copy exists, which means that only a selected few scholars are even able to access them, and then only under restricted conditions. At what rate should we be digitizing such items so they may be studied by the entire world?
• Avoiding duplication. With so much digitizing going on, it is important to eliminate wasteful duplication occasioned by the failure of institutions to exchange materials.

Paying For Content

It is axiomatic, and required by the legal system of all developed countries, that authors be paid for their work. With few exceptions, it is a pipe dream to imagine that even the inexpensive publishing outlet of the Internet will motivate authors to create and disseminate their works for free.

Professor Makoto Nagao's proposal is to allow the author of each work to set his own price for a menu of uses and allow market forces to operate as a natural regulator (Nagao n.d). He also asks the author to renounce any fee where only a small amount would be charged, thus avoiding the inconvenience and overhead of dealing with small numerical quantities. The notion of creating a free market is sound, since any attempt to subvert supply and demand is not likely to survive for long. However, there is no requirement that an owner allow his work into the marketplace at all, which is a significant obstacle. The recalcitrant owner can keep his work bottled up by setting an exorbitant price. It seems better to combine free negotiation with an appeal to an arbitration body if the owner and user cannot agree. Overall, it seems better to have a universal compulsory licensing scheme with fixed costs similar to the U.S. phonorecord provision.