When the networks had to pull back their Florida call for Gore last night and then went for Bush and then went to "too close to call," they were discredited among their viewers. While what the networks did last night won't impact the final outcome of the election, their critics in Washington were given cannon fodder for hearings, and they managed to damage their positions with both candidates in the tight race.
The problem that resulted in last night's bad calls can be traced to decisions made to offset the revenue pressures of the late '80s and early '90s. Top managements pressed individual operating units-in this case, news divisions-to cut costs, and some of the areas cut were those that would have prevented last night's problems.
For current investors and managers of media companies, there is an object lesson: Relatively small savings in areas that aren't used every day can have long-term consequences on the image and positioning of the overall corporation.
Going forward, as managers attempt to match costs to revenue pressures, decisions in areas that could impact overall corporate image can't be left to operating units.
Here's some history that led up to last night's network problems. Before the 1990s, each network ran its own polls during election years. These polls provided interviews with voters and other survey data that let the "decision desks" of each network separately make calls on who won or lost.
While raw-vote collection had been accomplished by the pool called the Network Election Service since the late 1960s, the polling data was collected independently. In 1992, the Voter News Service was created by the networks to pool the collection of polling data, and every network was provided the same data at the same moment. Decision desks still operated independently, but they were using the same data. That's why, last night, most calls were identical and made within a few minutes of each other.
Had each network's independent polling still been in operation, there would have been at least four separate polls of sample sizes similar to that of the one poll on election night. In the very close Florida election, nuances in the data may have caused one network to call Bush, another Gore, and maybe one or two may have held off altogether, saying, accurately, that it was too close to call to the end. Instead, the same data with whatever flaws it may have contained was used by everyone to reach and un-reach and then reach other conclusions.
The one-vs.-multiple poll situation is analogous to the difference in the volatility between owning one stock in a sector and four or five.
The cost saving of shutting down the independent polling operations was seen most in presidential election years but also, to a lesser extent, in off years. We estimate that each network saved between $5 million and $10 million in presidential election years. For Viacom, Disney and Time Warner, this is a fraction of a cent in earnings. Shareholders and managers need to ask whether the damage done to image and future political clout in Washington was worth the price of that cost cut.
Last night's fiasco may be an object lesson for corporate chiefs to pay more attention to longer-term implications of cost cuts as they try to deal with the current, unstable revenue environment. Simply demanding that operating units meet their numbers may result in lower-level decisions that potentially impact the strategic leverage and political clout of the greater corporation, not to mention the longer-term business prospects of the company.
All media conglomerates are at risk: Viacom as it tries to make numbers while most heavily exposed to advertising; Disney as it tries to offset revenue shortfalls from ratings declines; and Time Warner as it works to justify the merger with AOL.
By the way, AOL managers who have not been involved in media will have to learn quickly that the policies, operating practices and political fire that come with running news operations are far beyond those which they have ever experienced. Just ask GE about its early years at NBC.