Are Polls Polluting Politics?

Howard Kurtz
8:37 a.m.  Tuesday  9/24/96 

       Last night, ABC’s World News Tonight led with its tracking poll, showing Dole trailing by 12 points. I open Newsweek and (after skimming–okay, reading–the cover piece on adultery) see its poll has Dole down by 16. Others surveys still put the race at 20. Who’s right? Who the hell knows? It’s only a blurry snapshot. Some polls are done over weekends when many people aren’t home. If polls are indeed “white noise,” as Eleanor Clift says, they’re sure getting white-hot treatment from news organizations (which pour big bucks into them and then trumpet the results).
       Is the problem over-interpretation by TV & newspaper “dopes,” to use Fred Barnes’ word? Sometimes. Polls can actually be quite useful in telling us about the attitudes that underlie political preferences. It’s fascinating that more people now think Clinton better able than Dole to handle the crime issue, a Democratic weakness for 30 years. Or that those who think the country’s on the “right track” is at the highest level in several years. That may tell us more about 1996 than all the horse-race polls put together.
       As for Arianna Huffington’s contention that polls will turn pols into “slavish followers,” I ‘m afraid we’re already in the era of government by focus group, and not just at campaign time. I’m sure the V-chip and school uniforms tested off the charts for the Clintonites.
       The moderator suggests a lack of evidence that poor poll numbers depress turnout (although common sense suggests less enthusiasm for a candidate 20 points behind). But that’s too narrow a focus. Presidential elections ought to be ABOUT something. Granted, the candidates have made this one about a succession of small issues (Dole blaming Clinton for soaring teen-age drug use isn’t going to wash, I suspect). To the extent that poll-driven journalism has turned this campaign into a snooze, we may be missing a golden opportunity for a national debate. The ‘88 campaign was largely about flags and furloughs, but the biggest issue the next year turned out to be the massive savings and loan bailout. Guess it didn’t show up in the polling data. Eleanor Clift
9:41 a.m.  Tuesday  9/24/96 

       Polls were the decisive factor in keeping Ross Perot out of the debates. He met the other criteria of being on the ballot in every state and therefore having at least a theoretical chance of winning the presidency. But because he registered only 4 to 6 percent of the popular vote in current polling, the debate commission decided he could in fact not win.
       I hold no particular brief for Perot. I think he abused the democratic process by refusing to debate former Colorado Governor Richard Lamm, his only rival for the Reform Party nomination. But taking today’s polling numbers, which are after all only a snapshot, and holding them up as evidence that Perot cannot win an election six weeks from now is misusing polls as a predictive tool. Perot was only at 7 percent in the polls when he gained entry into the debates four years ago, and most analysts failed to grasp the impact he would eventually have on the election, and on the policy agenda pursued by President Clinton and the Democratic Congress. This time the race between the two major-party candidates is so lackluster, Perot could at least offer some comic relief. Arianna Huffington
11:59 a.m.  Tuesday  9/24/96 

       I will follow our moderator’s categorization of the three aspects of the potential significance of polls–incidentally, Herb, what’s wrong with alliteration?!
       1) Influence on election outcomes: Non-existent. It has been clear for months to all but the GOP establishment that Dole would lose to Clinton. A good counterfeit will win against a composite anytime. Clinton could have only lost against the real thing–someone clear and unequivocal about what he believed.
       2) Predictors of election outcomes: I agree with Fred that their reliability is often overestimated, but when we get into double-digit numbers–and all the latest polls that Howard mentioned give Clinton a double-digit lead–it’s hard to doubt their significance as predictors.
       3) Their influence on politicians either as candidates or policy-makers: This is still for me the most fascinating aspect of polling. If we accept, as Howard seems to, that there is nothing to be done about the current reality of “government by focus group” we might as well wave good-bye to the idea that what can be achieved through Democratic process will lead to significant reform–whether in entitlements or in the area of the role of government versus the role of citizens and communities.
       I agree with Eleanor that polls are “political white noise” and that most voters discount them. The problem is that politicians quote them with the reverence traditionally reserved for the Holy Gospel. In the heyday of the Contract with America, Newt Gingrich was proudly declaring at every opportunity that everything in the contract was a 70 percent item–a lot of good that did him and the GOP!
       When will politicians wake up to the fact that campaigning and governing by permanent polling is not only bad for the country, but bad for them? Herb Stein
5:31 p.m.  Tuesday  9/24/96 

       I would like to make one more try at the question of whether polls affect the election outcomes. There seems to be a common phenomenon of polls tightening as the election draws nearer, and at the actual election being closer than they had previously suggested. One might expect that if the polls show a big gap in, say, September, the gap would become larger thereafter, as the contributors and supporters of the trailer became discouraged. But that does not seem to be the case.
       Clift gives the most dramatic example of the influence of polls on the election. That is the case of Perot, who is excluded from the debates because his poll numbers are low. Probably no one thinks that Perot would be elected if he were included in the debates. Clift thinks he might bring comic relief. But he could bring more than that. He could confront the leading candidates with questions that Barbara Walters doesn’t ask them, influence the issues discussed in the campaign and thus possibly influence the policies followed by the victor after the election, and change the historical background from which the 2000 election will occur. Is the Perot case a misuse of polls?
       Kurtz says: “To the extent that poll-driven journalism has turned the campaign into a snooze, we may be missing a golden opportunity for a national debate.” But maybe there is no golden opportunity for a national debate. Maybe we are all centrists now, and there is nothing to debate about. In that case the poll-driven journalism has not turned the campaign into a snooze, but the snooze has driven the journalists to the polls to fill the vacuum left by the absence of a national debate.
       During the Republican and Democratic conventions SLATE held panel discussions in which all the panelists agreed that the conventions were pretty boring and irrelevant. I suggested that might be because the federal government had become irrelevant. This may also explain why the campaign is a snooze.
       Another possibility is that there are real issues but that the candidates have failed to make them seem important, and so the press has failed to give them the attention they deserve. I understand that was what Barnes was saying about the treatment of taxes in the Dole campaign. Can you visualize action by the candidates that would drive the daily poll numbers out of the headlines and, in consequence, also change the poll numbers?
       I would like to return later to Huffington’s point about the influence of polls on policy, but meanwhile invite the comments of the other panelists on her argument. Incidentally, I have no objection to alliteration. On the contrary, I am amused by it and pleased to think that I encouraged it by the title I gave for the subject of this panel. Fred Barnes
5:44 p.m.  Tuesday  9/24/96 

       1) Here’s where Arianna goes wrong. She says Republicans were too poll-conscious in trying to balance the budget (a “poll-driven” goal) partially by cutting Medicare, absent a “public consensus” to attack Medicare. Wrong on both counts. Polls didn’t drive this. Republicans didn’t need a poll to tell them they should balance the budget; that’s GOP dogma. And they ignored polls telling them to shy away from Medicare and went ahead, against Haley Barbour’s advice. This was leadership, bad politically for them now, good for the country when Medicare is reformed in 1997 along GOP lines with Clinton’s assent.
       2) Howie’s complaint, both Monday and Tuesday, isn’t really about polls. What he really dislikes is racehorse journalism, a phenomenon that preceded heavy use of polls but has been worsened by overindulgence in polls.
       3) Contrary to what Eleanor said, Perot’s position now isn’t comparable to his position in 1992. Now, he’s a diminished figure, having never reached into the 30s, even the 20s, in polls this year, as he did in 1992. Also, Reform Party members were polled in the nominating process, and only about 30,000 voted for Perot, roughly the number who voted for Lamar Alexander in the New Hampshire primary. As a result, his candidacy is illegitimate and he shouldn’t be included in the debates.