Deciphering poll results for lesser-known candidates
By NICK PANAGAKIS
Polls conducted early in a political campaign are frequently described as measuring little more than name recognition because people tend to vote for the name they know. If an early poll shows a great difference between the recognition of candidates, the better known candidate will have the advantage in the voter preference question. However, analysis can allow for differences in name recognition to provide more useful data on voter preferences. An example is the primary challenge to U.S. Rep. Gus Savage, a Democrat from the 2nd Congressional District in Chicago.
We conducted a poll in September in the district, interviewing 400 voters likely to vote in the 1990 Democratic primary. The poll showed that 35 percent would vote for five-term incumbent U.S. Rep. Savage and 40 percent would vote for Mel Reynolds a five-point lead for the challenger. But the poll also showed that Reynolds was known to only 45 percent of the district voters.
What is the outlook in this primary contest? Looking at voters who know both Reynolds and Savage, the poll showed 49 percent of them voting for Reynolds and 27 percent voting for Savage. That indicates bright prospects for the challenger as he becomes better known during the campaign.
Savage, now under investigation by a congressional ethics committee, can be compared with Wisconsin's 1986 incumbent attorney general, Bronson LaFollette, who was also under investigation for ethics at the time of his reelection campaign. Several weeks before election day, LaFollette led challenger Donald Hanaway 40 percent to 34 percent. Hanaway was known to only 43 percent of the state's voters, but he led LaFollette among these voters 46 percent to 31 percent. On election day, Hanaway won with 54 percent of the vote.
There are other examples of lesser known challengers who ended up winning elections. They showed early in the campaign that they were not unpopular with the voters who knew them. And they maintained their positive images during the later stages of the campaign. In 1976 at the time of the Illinois primary, when Republican James R. Thompson was first nominated for governor and Michael J. Howlett had won the Democratic nomination, both candidates were even at 38 percent in voter preference. At that time Thompson was known by 62 percent of all voters (he was least known downstate). Among voters who knew him, Thompson led Howlett with 64 percent, about what he received in the November general election.
There are cases, too, where lesser known candidates are unpopular. When the preferences of voters who know them are analyzed, these candidates show no improvement in their chances.
Results of our September poll in the Illinois 2nd Congressional District are unusual because the lesser known challenger is out in front. The poll indicates the problems voters have with longtime incumbent Savage. Forty percent said he was not the kind of congressman they could be
November 1989 | Illinois Issues | 30
proud of. More voters disagreed than agreed (43 percent to 35 percent) that Savage had a high standard of personal conduct and ethics. By an even higher margin (50 percent to 22 percent), voters said Savage was absent too often for key votes in Congress.
Savage has usually won the primary election with 50 percent or less of the vote, defeating two or more opponents. In 1988 he won with 52 percent of the vote. He was able to do this because 8,705 10th Ward voters voted on the Republican side in that primary compared to the 360 to 430 GOP ballots in earlier primaries. Presumably, these new 10th Ward Republicans would have voted for Savage's opponents had they remained in the Democratic camp.
When the September poll was taken, state Sen. Howard Brookins said he might enter the race against Savage for the Democratic nomination. In a three-way contest, the poll showed him trailing: Savage had 37 percent; Reynolds 25 percent, and Brookins only 14 percent. Brookins was known by only a few more voters than Reynolds, but Brookins is not helped by looking only at the preferences of those voters who know all three candidates. In that case, the race between Savage and Reynolds becomes closer, but Brookins still trails badly: Savage has 35 percent, Reynolds 33 percent and Brookins 15 percent.
Another way to gauge the ultimate outcome of a Savage-Reynolds primary from the poll is to look at the undecideds. In this column last April, I showed how undecided voters usually favor challengers over incumbents. If more undecideds decide in favor of Reynolds than Savage, his lead widens to more than 5 points.
Research for the April column, based on 155 incumbent outcomes, also showed that incumbents who received less than 50 percent in the polls against a single opponent were in serious trouble and almost always lost the election. Savage's percentage of the vote misses the 50 percent mark by 15 percent. □
Nick Panagakis is president of Market Shares Corporation, a marketing and public opinion research Firm headquartered in Mount Prospect. Panagakis, a member of the National Council on Public Polls, is best known for pre-election and exit polls conducted for the news media in Illinois, Missouri and Wisconsin.
November 1989 | Illinois Issues | 31