Incumbent races closer than they appear

By NICK PANAGAKIS

How will poll undecideds decide? The most common answers to this question are that undecideds either break equally or break in proportion to those stating a preference for a candidate. But the answer should be: It depends on who the candidates are.

There is increasing agreement that undecideds behave differently when races include incumbents: These undecideds appear to end up voting for the challenger most of the time. They fit the Incumbent Rule.

In November 1987 in this column I documented undecided behavior with data from 71 final incumbent polls (our own polls plus others from CBS, Gallup, Gordon Black Research and Market Opinion Research). Since then 84 more polls were added. Almost all were final polls within four weeks of election day, most within two weeks. These polls covered both general and primary elections and Democratic and Republican incumbents.

Of 155 incumbent polls, most or all undecideds went for the challenger 127 times.

The equal split rule would be true if cases deciding for incumbents about equaled those for challengers, and the proportional rule would be true if cases deciding for incumbents were greater (since most incumbents lead in polls and win most elections).

For poll users and news reporters this means:

• Incumbent races should not be characterized in terms of point spread. If a poll shows 50 percent/40 percent and 10 percent undecided, a 10-point spread will occur only if undecideds split equally: that is, a 55 percent/45 percent outcome on election day. Since most of the 10 points will go to the challenger, incumbent polls are a lot closer than they look: 50 percent/40 percent is likely to become 52 percent/48 percent on election day. If a poll is a mirror of public opinion, think of an incumbent poll as one where objects are closer than they appear.
• An incumbent leading with less than 50 percent (against one challenger) is frequently in trouble; how much trouble depends on how much less than 50 percent. A frequent pattern is for incumbents ahead in the polls with 50 percent or less to end up losing. The important question is whether poll results are reported with an understanding of how undecideds decide.
• Many polls may have been improperly analyzed and reported. Many polls remembered as wrong were, in fact, right. It's only natural to interpret the term "undecided" literally.

Why do undecideds end up voting for challengers? It seems that the undecided are not straddling the fence unable to make a choice — the traditional interpretation made for decades. It's best to think of undecided voters as undecided about the the incumbent, as voters who question the incumbent's performance in office. Those having trouble with this decision end up voting against the incumbent.

The 28 exceptions to the Incumbent Rule help support the theory on why this happens. Many challengers who did not get a majority of undecideds in the election were recent or current holders of an office equal to the one they were seeking. Voters were equally or more familiar with the challenger's past performance in a similar office, so the challenger assumes incumbent

April 1989 | Illinois Issues | 28

characteristics. Other exceptions include well-known challengers or short-term incumbents.

Some examples of these exceptions include: Carolina Lt. Gov. Jordon v. incumbent Gov. Martin in 1988, Nebraska ex-Gov. Kerrey v. one-year incumbent Sen. Karnes in 1988, Florida's former Gov. Graham v. incumbent Sen. Hawkins in 1986, Missouri ex-Gov. Bond v. incumbent Gov. Teasdale in 1980, and Hubert Humphrey III as a well-known challenger in the Minnesota Senate race against incumbent Durenberger in 1988. Illinois examples include former Cook County State Atty. Carey v. incumbent Cook County Board President Dunne in 1982 and two-year incumbent Mayor Bilandic, who split undecided voters with Jane Byrne in 1979.

Cases like these cover 17 of the 28 exceptions to the Incumbent Rule. Of course, there are incumbents who turn their campaigns around in the last week after final polls are taken: for example, incumbent Missouri Sen. Danforth in 1982.

There are interesting patterns in the 127 polls where most undecideds voted for challengers:

• 98 were plus or minus 4 percentage points from the incumbent's actual election result. The most frequent result was a two-point gain.
• In 41 cases, the incumbent got less votes than his poll percentage. This means about one in four of all 155 polls exceeded the incumbent's election result.
• In 78 cases, challengers gained 10 or more points in the election over their poll percentage.

In short, most polls appear to merely estimate support for the incumbent, and all or most undecideds end up voting for the challenger.

Most troublesome to explain are polls showing an incumbent ahead but who ends up losing the election. Some examples: in 1986, both Wisconsin incumbent Gov. Earl and incumbent Attorney General La-Follette were ahead in three late polls with less than 50 percent but lost by 5 and 7 points on election day. In 1986, showed Georgia incumbent Sen. Mattingly ahead by 10 points, but he gained only one more point to lose with 49 percent. In 1984, incumbent Illinois Sen. Percy led with 45 percent and 49 percent in final polls and wound up losing the election 48 percent to 50 percent. In 1983, Chicago incumbent Mayor Jane Byrne was leading with 35 percent in both final media polls but got no more when all 16 percent who were undecided voted for her challengers. In 1982, a final weekend poll had incumbent Gov. Thompson down to 50 percent; he got none of the 6 percent undecided and barely won by a few thousand votes. As long ago as 1976 Illinois incumbent governor Dan Walker was leading in the Democratic primary by 3 points at 46 percent points three weeks before election day and wound up losing with 46 percent.

There is one important caveat: The Incumbent Rule has mixed (or no) application in presidential races, apparently because after several months of primary campaigning, the national conventions and media attention, incumbent and challenger records are equally known.□

Nick Panagakis is president of Market Shares Corporation, a marketing and public opinion research firm headquartered in Mount Prospect. The firm conducts both private and public polls and has news media clients in Chicago, Kansas City and Milwaukee.

April 1989 | Illinois Issues | 29