In the mid-1970s, U.S. telephone companies began adopting metered service pricing for telephone directory assistance. Previously, unlimited free directory assistance calls were allowed under the flat rate for local telephone service. The new pricing structure had a fixed allowance of free calls (typically three to five) and a per-call charge for calls above that allowance.
Concern about efficiency and excessive use motivated the pricing change. Economists reified economic theory: “A principle of welfare economics is that each commodity should have associated with it a price or charge reflective of its marginal social cost.” “What commodities?” and “How many commodities?” from this perspective are merely issues of number theory. In practice, that means that a string of words (for example, “directory assistance”) defines a commodity. If a market doesn’t exist for it, then reason is needed to explain the missing market. Such an approach utterly lacks appreciation for actual economic ecology.
Uncharged directory assistance service created business problems. Directory assistance calls had significant marginal cost. With uncharged directory assistance, high-volume directory-assistance callers created costs that greatly outbalanced the revenue that they provided. For example, in the early 1970s a major New York City newspaper used directory assistance as a costless credit check:
our ad takers would call directory assistance to confirm the telephone number provided by the individual placing the ad, and we assumed that if the person had a telephone listing they were paying their monthly bills. It was a quick and easy way to get credit approval on a relatively low-dollar sale amount at no cost to the newspaper.
In fact, residential customers had a directory-assistance call distribution more skewed to high call volumes than did business customers. Residential customers also were more likely to call directory assistance for numbers that could be found in their local telephone book. Loneliness, not shrewd business practices, was probably the most important driver of high directory-assistance use. In any case, as an AT&T manager explained in 1976, “directory assistance cost had become a significant and escalating component of telephone service charges — approximately 50 cents per customer per month in 1974.”
Metered pricing for directory assistance greatly cut service use. Metered pricing reduced the total volume of directory service calls by about two-thirds in the mid-1970s. Much of this reduction came from among high-volume users. Before metered pricing, about 89% of directory service calls were calls made above the threshold of three calls per customer per month. With metered pricing in Virginia in 2008, that share was only an estimated 33%.
A more subtle effect of metered pricing was to greatly increase the share of customers who did not use directory service. In the early 1970s under flat-rate pricing, the share of customers who made no directory assistance calls was 30%. With metered pricing in Virginia in 2008, the share of customers who made no directory assistance calls was 82%. Customers in Virginia at that time had a monthly allowance of three free directory assistance calls. Creating a priced service apparently discouraged a large share of customers from using it, even though they could freely do so within their allowance of free calls.
More imaginative consideration of the possible product space might have created a better information business. Suppose two different directory assistance services were offered. One would be unlimited directory assistance on a best-effort basis. It would be included as a free service within flat-rated local telephone service. By having a fixed number of salaried telephone operators employed for best-effort, free service, the marginal cost of additional free directory assistance calls would be zero to the service provider. The other directory assistance service would be a premium, directory assistance service with a guaranteed response window (service level agreement). It would have a per-call price corresponding to the expected marginal cost of additional operators needed to meet the service level agreement. Both the free service and the premium service would be available to all customers at all times.
The free service/premium service combination might easily have created a better business than the single-service metered pricing. With single-service metered pricing, only about 5% of customers actually incurred directory assistance charges. The free service might have retained as directory assistance users the large share of directory assistance users that metered pricing alienated. The free-service users might in some circumstances convert to premium service and create on average a higher share of charged customers. Moreover, the 5% of customers who actually paid for directory assistance did so only after exhausting their allowance of free calls. Premium service allows charging a segment of customers for every directory assistance call. The small share of customers willing to pay for directory assistance might also have been willing to pay for every directory assistance call.
Most importantly, a separate, free, directory assistance service could have served as a testbed for new technologies and new services. It would have provided a natural space for early offering of non-live-operator services. Similar, low-cost, free information services could have been offered. Experiments with advertising could have been tried. In short, much of what is now happening on the Internet could have grown out of a innovative restructuring of the directory assistance business.
Metered pricing to constrain high-volume users of a flat-rated service is textbook economics. Directory assistance has missed huge business opportunities. Innovative business development requires thinking beyond what’s in textbooks.
* * * * *
Data (for U.S.): effects of metered pricing on directory assistance use (Excel version); directory assistance business growth (Excel version).
 See Daly, George, and Thomas Mayor, “Estimating the Value of a Missing Market: The Economics of Directory Assistance,” Journal of Law and Economics, v. 23 (1980). pp. 147-166. The first quote in the above paragraph is the first sentence in id. This line of thinking, which is particularly associated with Harold Demsetz’s work on property rights, has been highly influential among economists.
 Kuehn, Richard A., “A $1.5 Billion Niche” Business Communications Review, July 2003, p. 65. Other evidence indicates that the value of the U.S. directory assistance business in 2003 was about $6 billion.
 For a graph of percent of directory assistance calls versus percent of lines (customers) for residences and businesses about 1973, see McDonald, J., “The Use of Management Science in Making a Corporate Policy Decision — Charging for Directory Assistance Service,” Interfaces, v. 7, n. 1, pt. 2, Nov. 1976, p. 9. Residences had a higher percent of calls than businesses for every level of lines. Among residential customers, 64% of directory assistance calls were to get telephone numbers listed in the customer’s telephone directory. The corresponding figure for business customers was 52%. Calculated from chart in id., p. 8
 Id. p. 6.
 The data and sources for the statistics in this and the previous paragraph are in the spreadsheet on metered pricing of directory assistance. The spreadsheet includes additional data consistent with the points made above.
 See spreadsheet on metered pricing.