There has long been reason to suspect the data which the cable industry provides to various reporting services like Warren Communications News, Kagan Research, and Nielsen Media Research for U.S. cable coverage and subscribers precisely because the cable industry has considerable incentive to lie about it. Specifically they have incentive to under-report both coverage and subscribers so as to avoid a finding that the 70/70 limit – that seventy percent of American homes are passed by cable and that seventy percent of homes subscribe to cable – has been reached, thus triggering additional FCC regulation of the industry. The numbers have danced around the mid- to upper-60% range reported in these sources since 2004, only tipping over in Warren Communications News’ Television and Cable Factbook, which recently reported a 71.4% take rate to the FCC.1 When it became clear that the FCC was prepared to take action to invoke the 70/70 rule on the basis of the Warren data, the managing editor of Warren Communications News’ Television and Cable Factbook immediately called its own data into question in an interview in Communications Daily:
The figures from the Television and Cable Factbook aren’t well suited to determining whether the threshold has been met, said Managing Editor Michael Taliaferro. Taliaferro said Factbook figures understate the number of homes passed by cable systems — and the number of subscribers — because not all operators participate in its survey. “More and operators are just not giving up” those numbers, he said. “We could go with two dozen footnotes when we start to report this data.” Cable operators participating in the Factbook survey said they passed 94.2 million homes and had 67.2 million subscribers.
The FCC official who asked him for the cumulative figure didn’t say how it would be used, Taliaferro said. If he had known, he would have provided a list of caveats, he said. “It would have been a very lengthy email,” he said. Taliaferro said he did point out the shortcomings in a phone conversation with the FCC official but didn’t put it in writing because he wasn’t asked to. “I had no idea what they were doing with it.”2
More below…
One can only imagine the sort of pressure which was brought to bear by an irate cable industry against the reporting service which had revealed — on the basis of industry data – that the 70/70 test had been passed. However, Taliaferro’s explanation made little sense. If the problem is industry understating number of households passed and number of subscribers because cable operators refuse to provide the data, as Taliaferro suggests, then Warren’s Television and Cable Factbook must understate the number of households passed and subscribers. This means that the real numbers — the numbers we would have if all the cable providers coughed up the data — have as a matter of mathematical certainty to be greater than 70% coverage and 70% subscription. Taliaferro, attempting to please the cable companies, in fact gave evidence that the Warren figure of 71.4% of homes having subscribed to cable (the take rate) as of October 10, 2007 has to be an understatement of the reality.
In fact, there are several problems with the data on offer from Warren, including the fact that it relies only on information provided by NCTA members, ignoring overbuilders and much smaller systems which still meet the 36-channel requirement, as well as the incentive of NCTA members to understate their coverage and take rates to avoid further regulatory scrutiny. These problems also impact the data reported by Nielsen Media Research and Kagan Research, as well as the fact that they involve more estimation than the Warren numbers, since Warren Communications News tends to be treated by the cable industry as a very industry-friendly, tame trade press venue and cable operators therefore provide more proprietary data to it than to the other reporting services. It was a cable industry blunder of the first order that it provided to Warren in the first place the data which it then forced Warren to try to discredit.
The question remains however: is there a reliable source of data on cable coverage and take rates in the U.S.? Answering this required, first, determining a source to which the cable industry would have little incentive to lie about either matter. ABI Research collects data from U.S., Canadian, and Mexican cable providers, including overbuilders and smaller system providers, on coverage and subscription in North America. The U.S. providers have hugely less incentive to lie about these figures because they are merged with the Canadian and Mexican data and, thus, do not on first look provide a ready gauge of the U.S. situation. If it were possible to determine the values of the Canadian and Mexican coverage and subscription data and subtract it from the ABI North American data, the remainder would be the numbers provided by the U.S. cable industry in a venue where they had hugely less incentive to lie about them and in which under-reported overbuilders and smaller providers would be represented. The Australian international cable and broadband reporting firm, Paul Budde Communications Pty., Ltd., proved to be precisely such a source of the Canadian and Mexican data. Being a reporter which does not fixate on the U.S. market, Budde is less likely to be susceptible to the kinds to pressures to which U.S. reporters are subject from the cable industry, and Budde’s international reporting is highly regarded among industry analysts. The ABI Research North American coverage and subscription data and the Paul Budde Communications Pty., Ltd, data for Canada and Mexico were obtained.
ABI Research reports 133.71 million households passed by CATV in North America in 2006.3 Subtracting the 12.6 million Canadian households4 and 10.2 Mexican households5 passed, we are left with a total of 110.91 million households passed in the U.S. ABI Research reports 99.61 million CATV subscribers in North America in 2006.6 Subtracting the 8.65 million Canadian CATV7 and 4.97 million Mexican subscribers,8 we calculate 85.99 million CATV subscribers in the U.S. This gives us a take rate of 77.53%, which is comparable to the Warren Communications News take rate of 71.4%, but calculated from completely independent data, and which is a figure which appears to address the underreporting problem which Warren Communications News cited in calling its own data into question.
Warren reports 94.2 million homes passed and 67.2 million subscribers, which is 16.71 million homes and 18.79 million subscribers fewer than arrived at by our alternative methodology. We attribute this discrepancy to the following factors: major cable providers have incentives systematically to understate their coverage and take rates to avoid regulatory scrutiny even to relatively trusted reporting firms like Warren (and it is difficult to plausibly dance in the upper-60% range for a take rate without fumbling and reporting figures which place the industry as a whole over 70%), a factor which does not come into play so intensely since the U.S. are melded into North American figures, and the ABI data take account of overbuilders and smaller providers, many of which are privately owned and do not report to state and federal capital market regulatory agencies. Furthermore, the 77.53% take rate we calculate is remarkably close to that estimated by SBC in 2005:
In the Eleventh Report, the Commission concluded that the first 70 percent statutory threshold has been met: 79.8 percent of all occupied households are now passed by cable systems with 36 or more channels. However, it also found that the second 70 percent statutory threshold had not been met: of the 84,415,707 occupied households passed by systems with 36 or more channels, only 58,177,885 (or 68.9 percent) subscribed to cable.
NCTA’s own updated data now suggest that this 68.9 percent figure may have increased to well beyond 70%. NCTA’s website states that, as of February 2005, there were 73,219,360 total subscribers to cable service. Last year, the Commission found that there were 8,063,920 subscribers to cable systems with less than 36 channels. It is reasonable to conclude that this number has not increased over the past year, because it is unlikely that anyone would construct a new system with such limited capacity. It is far more likely that some such small systems have been upgraded in the last year. But using this same number of 8,063,920 subscribers to smaller systems, that would leave at least 65,155,440 subscribers to cable systems with 36 or more channels, or 77.2 percent of all households the Commission found were passed by systems with 36 or more channels (at least as of its last report).9
SBC used the NCTA’s own data to arrive at an estimate of a 77.2% take rate.
Another indirect method of estimation is to take the FCC’s 2006 estimate of 84% of U.S. television households having MPVD access and subtract out the 16.2% DBS penetration rate of MPVD households established by the GAO in a 2005 study,10 resulting in an estimated CATV take rate of 70.392%, which is consistent with the Warren figure, if somewhat lower than the figures produced by our method and that of SBC.
It should be noted that all three independent methodologies, as well as the Warren data, reveal a take rate in excess of 70%.
For more than a decade now the public interest community has complained that the cable industry systematically understates its coverage and take rates. The ABI and Budde data merely reinforce this contention. If the cable industry has evidence to refute these findings, then let them present that coverage and take rate data to the FCC under oath on penalty of perjury, rather than by pressure on reporting services and innuendo.
1 “November FCC Meeting to Focus on Cable Industry,” Communications Daily, November 14, 2007, 2-4.
2 Ibid., 3-4.
3 ABI Research, “Cable TV Market Forecasts,” 2007, Table 1-1.
4 Paul Budde Communications Pty. Ltd., “Canada – Convergence -Triple Play & Digital TV,” October 28, 2007, 8.
5 Paul Budde Communications Pty., Ltd., “Mexico – Convergence -Triple Play & Digital TV,“ October 28, 2007, 13-18.
6 ABI Research, op. cit., Table 1-2.
7 Paul Budde Communications Pty. Ltd., ”Canada – Convergence -Triple Play & Digital TV,“ October 28, 2007, 8.
8 Paul Budde Communications Pty., Ltd., “Mexico – Convergence -Triple Play & Digital TV,” October 28, 2007, Table 9.
9 Reply Comments of SBC, Inc., MB Docket No. 05-255, October 11, 2005, 14-15.
10 U.S. Government Accountability Office, “Direct Broadcast Satellite Subscribership Has Grown Rapidly, but Varies across Different Types of Markets,” GAO-05-257, April 2005, 29.