|
|
|||||||
|
|
  |
 |
 |
||||
2.3a Earlier Campaigns
ATI polling had been organized in 1987 as a cooperative project to help all candidates running for president in '88 find out what the public wanted the new President to do on national security matters. At the start of the project an ATI bipartisan team headed by Maddy Hochstein, of the Daniel Yankelovich Group, visited all 13 campaign headquarters (7 Democrat and 6 Republican) and offered to test national security questions that the candidates might wish to have included in our surveys. Twelve candidates submitted questions. The lone exception, Mike Dukakis, was a little surprising since Mike's pollster, Marttila and Kiley, was ATI's pollster too. ATI promised to keep confidential the candidate author of a specific question. Only Maddy and I, as the project organizer, received them. I kept my copy in a secure place to stop leaks. ATI briefed the campaigns on our findings after each survey. Within months into the campaign, most of the candidates' questions had already been covered in the briefings; I took out my confidential copy of the candidates' original questions and studied them.
They covered many topics from all different points of view, not unlike the ATI surveys themselves. But something more important became very clear. The candidates, clearly without any contact between themselves, had asked ATI to find out in effect if the public believed them and/or their competitors on just about every policy position they had taken on national security. Their basic question was, "Did the public believe (agree with) the national security statements and positions that were then coming out publicly in their own speeches and releases?" This was not what the ATI surveys were supposed to be for. The ATI surveys were designed and the question wording developed to ask the public what they wanted the next president to do; they were "what-to-do" questions, which by 1992 ATI had dubbed, "public-interest poling." The candidates' questions were all bent on asking how should the candidates best express publicly their own convictions on issues. They were "what-to-say" questions, not "what-to-do" questions. The different purposes of the two kinds of questions lead to very different survey designs and findings. It was understandable that candidates would do this, but it made me sad to think that the candidates wanted to get confirmation that their positions were acceptable to the voters rather than what the voters themselves wanted. The same pattern was pretty much repeated in the '92 election.
By the time of the '96 election ATI had discovered something new: campaign pollsters familiar with the art and science of survey and question design, if they polled enough, covered an issue from various points of view and studied the effect of question wording variations, would eventually find from "what-to-say" surveys pretty much what you could find more accurately, completely, and quickly from "what-to-do" i.e., "public-interest polling".
Beginning after the ‘94 election, these effects were starting to be found by campaign pollsters working for real candidates. Here is the story. In his book, “Behind the Oval Office”, pollster Dick Morris, explained how he designed and conducted a program that seemed to play a significant role in re-electing President Clinton in 1996. The program required testing public support for hundreds of new policy proposals that Clinton himself liked. Morris gathered these proposals from Clinton by questioning one at a time his cabinet level advisors for a year. Morris' program required sequential media blitzing of one major metropolitan area after another. Prior to each blitz, Morris by quiet polling found a proposal that was strongly supported (usually 75+%) by the local public. The heart of the blitz was a TV spot of Clinton promoting the chosen proposal that he liked and he already knew the public by 75+% really liked.
The ultimate cost of the program was $85 million, largely spent on the TV spots and other media promotion. After each blitz, Morris again polled to show the President’s popularity generally rising slightly – Morris took note of increases as small as 1%. As the program continued, the cumulative effect boosted Clinton’s nationwide approval rating into the 50-60% range, quite high for a President beleaguered by growing scandals. Clinton had decided that selling White House access was the only way he could raise the enormous amount of money the program demanded. Presidents had been selling access for years but never on such a large scale.
Republican accusations, amplified by the mainstream news media, started an avalanche of scandals that dogged Clinton's second term -- initially selling Lincoln bedroom overnights and White House breakfasts/coffees to $100,000 plus campaign donors and later with an even greater scandal -- the Monica Lewinsky, Paula Jones, affairs, et al. The scandals led to legal challenges beginning with investigations of Clinton's involvement in Arkansas corruption, dominated by the impeachment hearings and still dogging him as he left office in 2001.
Morris’ polls, which cost only a few of the 85 million, obtained findings that accurately showed what the public wanted in the many issue areas they touched on. They fall under the heading of Governance Advice discussed on pp.58-61 in "Locating Consensus for Democracy" (LCD) where it explains why it is fair to label such polls as “public-interest polls”. One caveat – I have not seen the precise wording of the questionnaires and so cannot make as definitive judgment as I otherwise would. Morris, like many in Washington, has low ethical standards. Evidence that this is so: (1) His foolish, adolescent affair with a prostitute that forced Clinton to drop him as pollster for a time and brought Morris lots of personal trouble, (2) His total acceptance of polling on the Internet, described in his book “VOTE.COM”, without once mentioning the main problem present on the Internet for many years after Morris published his book, how to obtain a credible unbiased sample. As he was positioning himself to enter a huge potential market for polling on the Internet, Morris simply failed to mention this problem in VOTE.COM. A demonstration of the dominating effect of this problem is given in the section Prodigy Example of p.341 (LCD). After the election, Andrew Kohut, a media-prominent pollster with the Pew Foundation, showed that Clinton's bump-up from the Morris’ program had faded by election day. Of course, the program might still have had a residual longer-term significant but unknown effect. There would be no way to measure that.
| >>> 2.4 Supporters of Public-Interest Polling |