So, we hear about a spat between Donald Trump and researcher Frank Luntz, head of Luntz Global, which claims to specialize in message creation and image management for commercial and political clients. Luntz is often seen on FOX, conducting live focus groups in which a carefully-chosen audience is given electronic dials which register the degree of its reactions to tv commercials, sound bites from political speeches and such. The wiz-bang dials and attendant graphs are intended to lend an air of science to his show.
I know a little about message creation and research. What Mr. Luntz does is less science, more puppet show.
I was for many years Creative Director of a major international ad agency, headquartered in New York City. You could say I was a “Mad Man.” One fine day, the Director of Research at our agency came by my office for what had become a regular, friendly visit. She was well known in the field, and I enjoyed her company. She spoke with a charming European accent, but unlike myself, was inclined to treat advertising less as art, more as business.
At the time, I was working on a major advertising campaign for a well known client – an American institution. This client was about to invest tens of millions of dollars in a national ad campaign. My creative team had spear-headed the ad ideas which got client approval provided we could show positive testing (research results) of the idea (or as we called it; the concept).
Having got past our small talk, my colleague flat out asked me how I wanted the test results to come out. She realized how important this project was to our ad agency, to our client, and to me and my staff. She was prepared to help.
On the subject of audience research, polling, focus group testing and such, her attitude was, it’s pseudo-science. She could manipulate. She could guarantee the results.
This particular testing would be done via a “focus group” similar to Mr. Luntz’s. Our agency had a special conference room dedicated to this purpose. Since we were located in midtown Manhattan near Grand Central Terminal, our independent test firm often gathered volunteers from crowds of commuters coming and going through Grand Central. They would gather in our conference room; about 10 or 15 subjects and a test firm moderator – someone like Mr. Luntz. The session would be taped. The moderator began chatting and asking questions from a questionnaire generated by the moderator and our Research Director. The subjects might be shown a film or tv storyboard, or music, or whatever was being researched. The questions generally involved the subjects’ reactions to what they were being shown; likes, dislikes, why, et cetera.
This conference room was nicknamed the “fish bowl” because there was a second, smaller room, separated from the conference room by a “two-way” mirrored wall. The smaller room was darkened so our ad agency people were able watch the proceedings, unseen. We could also send spontaneous questions to the moderator. Sitting in the dark, listening to the spontaneous thoughts of the subjects, it was often difficult to not burst out laughing. These were obviously intelligent people, but when verbalizing their reactions to ideas, celebrities or music, they often sounded like idiots.
The reactions were charted, data tabulated, tapes edited and presented to our clients a few days later. There were certain “norms” against which the results were compared, and adjustments were made to the bottom lines if warranted.
In the 1930s, George Horace Gallup himself was VP in charge of copy
research at another major ad agency; it was there that Mr. Gallup first formed his own polling firm, the
American Institute of Public Opinion. There has always been a close relationship between advertising, audience research and politics.
In the 1960s, journalist and Washington insider Joseph Alsop argued that many people believe polling is
an exact science. To the contrary, Gallup’s polling record up until then
had demonstrated that if it was science, just how imprecise science could be (see Alsop’s New Yorker article).
Gallup, I am told, was fond of quoting French bishop, politician and Napoleonic diplomat Charles Maurice de Talleyrand’s 18th Century maxim: “The only thing wiser than anybody is everybody.” Hence Gallup expanded his sampling of opinions to as broad a cross-section of America’s electorate as his new organization could muster. To be fair, if success in this field is measured by how many election results match the polling stats, Gallup had some bull’s eyes along with the failures. Even a blindfolded man throwing enough darts is bound to hit a target. Occasionally. But would you bet money on his next throw?
I’m not saying polling isn’t of value. I’m saying it has been corrupted. A cozenage in the trick bag of politicians and the media. It doesn’t simply measure; it attempts to influence. Which is precisely what Frank Luntz attempted to do.
Now, I don’t know about you, but I cringe when I hear a politician, pundit or adviser, using a polling statistic, declares “...The American people believe such and such...” Hogwash! No one, not even Gallup, knows what “the American people” believe. There are well over 300 million Americans. Opinions are infinitely nuanced and quixotic.
The worst part is, when a presidential candidate wins by, say, 51% of the vote, it means the other 49% are treated like irrelevant losers for at least the following four years! Presidents often can’t shake their party associations, so 49% of the voters go under-represented. This is why polling should not be used as a partisan tool. But no thinking person believes it isn’t. And Luntz proved it.
A well respected exec of another major ad agency also wrote on the subject. He opined that the phraseology of polling is “loaded.” Since questions are usually written or formulated by researchers and pollsters, the result, by intention or not – consciously or not – is that it steers the subject to give responses the pollster wants.
Ask a young “likely voter,” self-identified as “liberal,” if she would vote for Candidate A who promises to lower her taxes and balance the budget; or for Candidate B who also promises to lower her taxes, balance the budget and limit access to birth control. If that’s all the subject hears about the two candidates, the pollster knows he can put a check-mark in column A.
Subjects are human; they come into the process with their own biases or preconceived notions of political parties or candidates they have heard about. Few voters who claim to be neutral, or “undecided,” are actually neutral – whether or not they are aware of their own biases.
Some subjects are reluctant or embarrassed to express what they really think, so may give answers which are politically correct. Once in the voting booth, it may be a different story.
People may not actually know the answers to questions such as “how do you feel about...” or “what’s your impression of so and so...” Most of us have seen interviews with “the man on the street” who can’t even name our current Vice-President. Some can tell you trivial details about the Kardashians or Katie Perry, but can’t reliably name the country to the north or south of the United States. Many younger people have little interest in the machinations of governance. But they vote. And they can be counted on to vote for narrow self-interests.
In recent times, because our voting population seems so evenly divided, results of presidential elections have been uncomfortably, even suspiciously close. One percentage point can make all the difference. Which is why opinion polling is so controversial.
Daily polling, rolling averages and the like, whether intended or not, influence potential candidates and voters alike. These “snapshots in time” are least controversial – least critical – when the spread between major candidates, in the last critical weeks of campaigning, approaches double digits. I know this is a shocking revelation, but politicians will say whatever they believe will get them more votes. If candidate #1 who says “A-B-C” sees his polling numbers go down, looks at candidate #2 who gets a bump in his polling by saying “J-K-L,” candidate #1 will find a way to turn “A-B-C” into “J-K-L.” After awhile, nobody knows who believes what. Opinions? Garbage in – garbage out.
Maybe it doesn’t matter in the long run. People will vote for the candidate they think they know and “like” – especially if that candidate promises things the voter wants or at least agrees with – whether or not that candidate is up to the job. So if polling is nothing more than a measure of popularity, at least it serves some purpose, shallow as it may be.
Oh, if you’re wondering how our ad campaign tested. Our research department announced to our client that it scored higher than any other ad campaign they had hitherto tested for this or any other client. The client was extremely pleased and funded our ad campaign 100%. Quite an accomplishment, eh? Until I remembered my research director’s lesson. The results certainly were influenced by her questionnaire and, I suspect, by our friendship.
Which brings us back to Mr. Luntz. It was clear, watching his focus group opining on Mr. Trump’s campaigning, that Frank had carefully loaded his group, knowing by his pre-interviews that they would not produce fair or balanced responses to Mr. Trump, but would produce the responses Frank wanted. This was far from science. This was Frank’s way of “getting even” with Trump, who had apparently recently declined to participate in some business dealing with Luntz.
Well, polling in politics is not going away. It is my own notion that despite the claim that there are more “conservative” voters than “liberal” voters in the U.S., the conservative candidate needs a minimum lead of 8 honest points in order to beat his opponent. He needs that much to overcome media bias, polling effects, college professors indoctrinating young voters, and shenanigans with the actual votes. Horrors! you say? But think of it. Conservatives tend to be conservative with the vote counting, whereas liberals tend to be, well... liberal with the counting, accidentally overlooking or excluding cast votes of certain groups, et cetera and ad nauseam.
As Joseph Stalin said, “Those who vote decide nothing. Those who count the vote decide everything.”
