Ignore The Horse Race, Pay Attention To The Trend

Now that the presidential campaign is in full swing for the media and political communities (as well as for the 12% of Americans who are paying attention), here are some observations about what the polls are showing.

Ignore the Horse Race, Pay Attention to the Trend

Poll standings at this point in the cycle are most often misleading as to the outcome of the presidential election. However, the trends in public attitudes, especially in the President’s approval ratings, are more prophetic.

October 1991– Bush approval ratings are still in the mid-sixty range. Remember: “Who can beat him” and the “the stature gap.” Trend – Bush’s ratings are on one of the steepest downward slopes in polling history.
October 1987– By a margin of 4-to-3 Times Mirror surveys found voters are inclined to cast a ballot for a Democratic candidate over a Republican. Trend – By October, Reagan’s Gallup approval ratings at 50% are in a partial recovery from Iran Contra doldrums (as low as 40% in February ’87).
October 1983– Mondale leads Reagan by a 50% to 44% margin in the Gallup Poll. Trend Reagan’s approval ratings up from 35% in January ’83 to 45% by October. Grenada, a month later, propels Reagan over 50% mark.
October 1979– Carter leads Reagan 48% to 42%, but the President trails Kennedy by 2-to-1 for nomination among Democrats. Trend – The Iranian hostage crisis subsequently boosts Carter’s ratings long enough for him to dispatch Kennedy in the early primaries.
October 1975– Quite a muddled picture as Ford leads Humphrey, Scoop Jackson and Sargent Shriver, but trails Kennedy in the Gallup Polls of October (Gallup does not test unknown Jimmy Carter against Ford). Trend – No consistent pattern is evident, ups and downs in Ford’s ratings over the course of the year. Public divided 44% approve, 44% disapprove by the end of October.
Trends in ’95 – Steadily rising opposition to GOP policies and increasing generic support for an independent candidate are evident. Despite flat approval ratings, in two-way match ups Clinton is now ahead of Dole, who he had trailed earlier in the year.

Polls Apart on Powell

The polls all agree that Colin Powell would run a very strong race against Bill Clinton as the Republican nominee. But they disagree as to how strong a race he would run as an independent candidate. Some show him leading Dole and Clinton as an Independent, while others show him either trailing Clinton, or in a statistical tie with him. The sharpest differences were obtained by Newsweek and Time magazine when each commissioned polls on the weekend of September 14 to test Powell’s strength as an Independent. Time found a statistical tie – Powell at 34%, Clinton 32%, and Dole 25%. Newsweek found Clinton clearly ahead (35%) of both Dole (29%) and Powell (27%).

Timing, design of sample and question wordings were comparable for both polls. However, the Time survey questioned its sample about the three-way race after having first asked it about a two-way race between Clinton and Dole. Newsweek’s three-way ballot test did not follow a two- way preference question.
The very same pattern was observed in 1992. Perot polled five percentage points better against Bush and Clinton after a two-way test, than in a “cold” three-way measurement in a Times Mirror split ballot experiment in June 1992. [ “The Year of the Outsiders,” Times Mirror Center for the People & the Press, June 16, 1992, p.8.]

The Times Mirror experiment and the more recent divergent Powell results suggest that independent candidates generally seem more appealing to voters after considering the prospects of a choice of only major party candidates. However, the fact that Powell’s lead over Clinton is so consistent as a Republican, but so method sensitive when he is cast as an Independent, is also an indication of the “softness” and potential volatility in polling about third party candidates next year.

John Paul Who?

The New York Times recently reported the results of a CBS/NYT poll that found that as many as 28% of Catholics could not say whether they had a favorable or unfavorable opinion of the Pope. Other media polls such as the ABC Washington Post poll showed only 7% of Catholics unable to respond to this most basic of all polling evaluation questions.

The reason for the difference is that the CBS/NYT poll adds to their standard question formulation a phrase “or haven’t you heard enough about him yet to have an opinion”. The intent of the phrase is to discourage people who are ill informed from giving answers. It is not certain that this phrase only achieves this purpose. It may also make it easy for people not to express an opinion they hold, but are reluctant to give. The CBS/NYT poll, perhaps in a bow to the second point of view, now often uses a follow up question which asks the undecideds – “if you had to choose what is your opinion …” (in response, only 4% of Catholics did not pass judgement on the Pontiff).

How hard a polling organization pushes for answers is one of the major differences between otherwise seemingly comparable surveys. The most notable example of this is in the Presidential approval question. The wording, which originates with the Gallup Poll, is the same for all the major media polls, but results vary because polling organizations differ in how they instruct their interviewers to probe for an answer from respondents who initially say “I don’t know, or I approve of some things and disapprove of others.”

The confusing consequence of this is that ABC Washington Post and the Los Angeles Times show over 50% approving of Bill Clinton (52% and 51%, respectively), while the other national polls have his approval rating in the 40’s – NBC/WSJ (46%), CNN/USA Today/Gallup (46%), CBS/NYT (45%), and Times Mirror (45%).