Progress Snapshot
Release 1.3 June 2005

by Adam Peters *

Last week, the FCC released its latest report on telephone subscribership in the United States. [1]The subscribership data contained in the report is gleaned from a Current Population Survey conducted by the Census Bureau on a monthly basis, though the Bureau only collects phone subscribership data every four months. At first glance, the bottom line looks relatively benign. In March 2005, the number of households subscribing to telephone service was 92.4%. However, a look at the percentage of households without telephones over the past two years reveals a noteworthy trend:

To put the most recent figure in perspective, the last time that telephone penetration reached this low level (at least on paper) was nearly eighteen years ago, in November of 1987, when the FCC estimated that 7.7% of households did not have telephones.

What's behind this trend? From a review of the report, it is difficult to discern whether, or which, demographical "groups" of customers are being dropped from the network, at least with the information that is traditionally collected and reported by the FCC. But a couple of conclusions can be reached at the outset. First, the statistical decrease does not appear to be the result of a price increase in local phone service. On the same day that it released its subscribership statistics, the FCC also released its annual Reference Book of Rates, Price Indices, and Household Expenditures on Telephone Service. In 2004, rates for local phone service declined a modest $0.21 from the previous year, for a national average rate (including taxes and fees) of $24.31.

Second, the FCC is trying to more accurately account for wireless substitution in the data set. To be sure, the FCC's recent approach to wireless substitution can be described as guarded, discounting it entirely in its ill-fated Triennial Review Order and calling it "premature" to consider a separate relevant market for wireline and wireless service in its analysis of the Cingular/AT&T Wireless merger. [2]In previous surveys, the Census Bureau asked respondents: "Is there a telephone in this house/apartment?" In all likelihood, the notion of a "telephone" as being a traditional landline phone would probably lead some respondents to answer in the negative. In its March 2005 survey, however, the Bureau asked the following question for the first time: "Does this house, apartment, or mobile home have telephone service from which you can both make and receive phone calls? Please include cell phones, regular phones, and any other type of telephone."

There are two new aspects to this survey question. As the FCC itself noted in its report, the inclusion of the word "service" in the query might lead customers who had a phone but nevertheless had their service disconnected to answer "no," which would cause part of the reported decrease in subscribers between November 2004 and March 2005. On the other hand, by making substitution by wireless or another device more explicit in the survey question, it is somewhat surprising that capturing data on more customers who have "cut the cord" would not offset the number of consumers who may have had their service disconnected. An increase in disconnected customers may be due, in part, to substitution itself, as many wireless carriers are not subject to provider of last resort obligations and it is easier to disconnect customers for non-payment in certain states.

The FCC report states only that the decrease in subscribership is "statistically significant." The significance could be statistical or one of survey quality. This raises the question of whether the new or the old survey is more accurate. Does the current survey bias toward over- or under-reporting subscribership? And what of the old survey question? It was thought to have caused an under-reporting error, hence the change in the survey question. To be sure, whether this downward trend is merely a statistical problem or an indication of real, decreased subscribership requires further study. Data collection and reporting may be more accurate now, and false positives may have skewed the penetration estimate upwards for many years. Or, it may just be more difficult to capture user data in an increasingly intermodal world.

On the other hand, if the downward trend is indeed real, it would be astonishing considering the runaway costs of the USF. The universal service contribution factor has ballooned to 11.1% and total high-cost support payments have increased nearly a billion dollars between 2000 and 2004. [3] To its credit, the FCC has worked to identify and reach out to underserved communities in recent years, including tribal areas. However, following on the heels of exposed fraud in the E-Rate program, a truly significant loss of network users would constitute another black eye for the administration of the USF. If the federal government does not have the resources to do so, perhaps Congress should consider giving the states a more active role in the distribution, oversight and enforcement of the USF.

Footnotes:

* Adam Peters is research fellow and regulatory counsel with The Progress & Freedom Foundation. The views expressed are his own.

1. Alexander Belinfante, Telephone Subscribership in the United States (Data through March 2005), Industry Analysis and Technology Division, Wireline Competition Bureau (Rel. May 25, 2005).

2. Notably, the FCC's rhetorical posture towards substitution of wireline service by VoIP has been far more positive. In his statement supporting the recent order placing a 911 mandate on VoIP services that terminate on the PSTN, Chairman Martin emphasized that the issue was one "of life or death for the millions of customers that subscribe to VoIP service as a substitute for voice service." Commissioner Copps, who opined in the Cingular/AT&T Wireless Order that "we won't be hearing so much rhetoric in the future about the power of wireless as an intermodal competitor," emphasized that the VoIP 911 mandate was aimed "squarely at substitutes for basic telephony."

3. Available Here