CBS'and NBC's news divisions took some blame for bad calls during the election-night debacle, passed some of it to the Voter News Service consortium, recommended changes, and, in separate reports, provided-particularly in the case of CBS-a heretofore unseen level of detail regarding election-night coverage.
CBS' 87-page report, released last week, provides a nearly minute-by-minute account of its experience, highlighting the events that led to the mistaken Florida calls.
Kathleen Hall Jamieson, dean of the Annenberg School for Communication at the University of Pennsylvania, who produced the report with CBS executives Linda Mason and Kathleen Frankovic, wrote, "For me, the underlying problem with coverage by the broadcast networks of the election returns is overconfidence in the ability of the system in place at each network to protect viewers from misinformation.
"Confidence in the statistical models programmed by Voter News Service (VNS) was misplaced," she added. "That problem was compounded by a flaw in the Decision Desk setup at CBS News, which relied solely upon the VNS numbers without taking into account the breaking news from Florida. This situation left Dan Rather, the anchor of CBS News and the embodiment of the network's Election Night coverage, at a disadvantage, since he received his information from the producers of the Election Night broadcast as they received the calls from the CBS News Decision Desk."
Among the major recommendations by the authors of the CBS study are these:
Assign a member of CBS News senior management to head the so-called Decision Desk who would be able to "withstand the competitive pressure" to call a race because other networks may be doing so. Also, assign a correspondent to dissect tight races and conflicting data, and use more sources to determine a race's outcome.
Move the Decision Desk into the election-night studio, to improve communications, instead of working "in a vacuum as it did this time in an office three floors from the studio." (NBC-in its own, far shorter report also released last week-suggested just the opposite. It proposed isolating its decision-makers, so that they could make their calls without being influenced by the pace and competition on the set.)
Develop a new terminology to describe races, such as reporting a state is "leaning" toward a candidate, without actually awarding the state to a candidate.
Tell viewers more often and with detail how and why calls are-and aren't-made.
CBS News chief Andrew Heyward said the "candid report stands as a testament to the seriousness and persistence of our effort to maintain the highest credibility with the American public we serve. Furthermore, we hope this document will serve as a valuable blueprint for CBS News election coverage for many years to come."
VNS-which has made a preliminary report to members NBC, ABC, CBS, CNN, Fox and the Associated Press-had no comment on the studies and is continuing toward its own internal report. An independent review from the Research Triangle Institute is expected later this month.
"While VNS' fundamental methodology is sound, its system needs improvement, including a computer upgrade," NBC's report said, "VNS' projection models, data feeds to members, and internal quality-control methods appear to have been inadequate in a contest as close as the Florida presidential race. Additional exit polling may be called for in races expected to be close."
NBC News Vice President Bill Wheatley conceded, "There was some bad data in the VNS system." Further compounding the errors was a system malfunction that reduced the ability of VNS to cross-check with Associated Press numbers.
Both CBS and NBC say they will invest additional resources in their own reporting and in VNS-or a new consortium-and refrain from most projections until a state's polls are closed. The networks also plan to do better cross-checking. Neither network appeared ready to abandon the concepts of projecting winners or of a consortium for election-data gathering.
That brought disagreement from Carl Gottlieb, a veteran broadcast journalist and deputy director of the Project for Excellence in Journalism. "There are ways to economize," he said, noting that the move to consolidate election polling a decade ago was borne out of economy, not to advance journalism. "This is not one of them. But these reports indicate a serious attempt to overhaul the system. To that, I say great! But we'll see. The networks say it's more important to get it right than to get it first. But the temptation not to be last is always going to be very great. That competitive nature is what makes journalism in this country so great."
Both networks disputed the contention-made most prominently by Rep. Billy Tauzin (R-La.)-that there was a bias in election calls that favored Democrat candidate Vice President Al Gore over Republican Texas Gov. George W. Bush.
The CBS report said its study "concluded that there is no evidence of either intentional or unintentional bias in the timing of the calls" and called Tauzin's charges "serious but not accurate."
Wheatley said NBC's investigation "found no evidence of that in any possible way. We believe that there's not a trace of it in our own decision-making." Tom Goldstein, dean of the Graduate School of Journalism at Columbia University, helped produce the NBC report with NBC execs David McCormick and Maya Windholz.
A Tauzin spokesman said that, while the reports "may have satisfied some of the network brass, in our minds, it fell short of expectations. There are still a lot of unanswered questions."
Spokesman Ken Johnson said Tauzin, chairman of the House Telecommunications Subcommittee, will continue with plans for a hearing on what went wrong on election night, probably sometime next month. "Given the resources and technology available to all the networks, this simply should not have happened."