Lingle defeated Hirono on every island except Kaua'i, and won by a margin of 52%. (1) A white, Jewish woman winning the seat in Hawaii is being called "revolutionary". (2) This failed forecast can largely be blamed on the deviation from the model. Because of Patsy Mink's (D) unfortunate death, the seat in the House was too be filled by an appointment of the newly elected Governor. Hawaii newspapers, as well as my website, were predicting that voters would elect a Democratic Governor in response to Mink's passing.

This was not the case. Because of the extreme circumstances, a portion of the model was not taken into as much consideration as it should have been: name recognition. Four years ago, Lingle came close to defeating the then incumbent for the Gubernatorial seat. When the seat was open and Lingle ran again, voters know her as not only a name, but a popular one. Lingle had such an extreme case for name recognition that it can be deemed a huge factor in her victory. Although Hawaii is largely Democratic and votes Democratic, (the state House and Senate both just received Democratic majorities),  they elected a Republican Governor to the seat because of name recognition- even in light of the death of the beloved Patsy Mink.
 

Additionally, Lingle's campaign strategy was remarkably different from that of her last race. In her previous race against the incumbent, Lingle spent most of her money, time, and energy winning the primary elections alone. While her television campaign was strong, she did "fizzle" out at the end- when she needed it the most. Her competitor ran the majority of his television ads in the two weeks before election day, and in those two weeks, voters heard very little from Lingle.(3) In this election, 30 second commercials were hitting Hawaiian televisions less than 24 hours before the vote.(4) Several Hawaiian newspapers, including the Honolulu Advertiser and the Honolulu Star Bulletin, mention this as a key factor in her campaign.

While it is difficult to analyze this forecast in terms of the model because I deviated so much from the model because of the extenuating circumstances, some comments on Lingle in relation to the model are warranted. Lingle may have ran a tighter campaign with slightly more money. She may have had name recognition. Voter turn out may have been in her favor. But the remaining factor, incumbency, favored Hirono. The previous Governor was Democratic, representing the largely Democratic Hawaii. However, this one factor against Lingle may have been tainted: there were allegations of Gov. Ben
Cayetano inappropriately using his gubernatorial powers to sabatoge her campaign. While Cayetano did use his tax-payer supplied staff to "look into" Lingle's proposed Gubernatorial plan for faults that could be publicized, many Hawaiians feel he acted inappropriately. (5) This could have worked against Hirono's advantage as the candidate for the party of the retiring incumbent. Thus, with name recognition, a more effective television ad campaign, and despite incumbency disadvantages and Democratic support in light of the passing of Patsy Mink, Lingle's victory is not an entire deviation from the model.

However, it does teach a lesson about the way in which we use the model. While Hirono had an incumbency advantage as a Democrat, that advantage could very well be tainted by the actions of the current Democrat in office. If Cayetano had been overwhelmingly popular at the time of the election, Hirono might have received more votes as a democratic candidate. Moreover, while Cayetano is a Democrat, he only won against Lingle in the last election by a small margin. Therefore, while the democratic party held an advantage due to incumbency, which is used in the model as a major indication of projected election victory, was tainted because a) Cayetano was not entirely popular at the time of the election and b) he won a narrow victory previously. Thus, incumbency may be everything in an election, but it is dependent on the actions and performances of the incumbent. This was an open seat because Cayetano had reached his term limit- but future forecasters should beware: it is possible to have both an open seat AND a vulnerable incumbent in the same election.

Lingle is the first woman governor for the state of Hawaii.

Click HERE to see the original forecast of the election.

(1) http://www.swiminfo.com/lane9/news/4367.asp
(2)http://www.swiminfo.com/lane9/news/4367.asp                                                                                 (3) http://the.honoluluadvertiser.com/article/2002/Sep/24/ln/ln03a.html   (4)http://the.honoluluadvertiser.com/article/2002/Sep/24/ln/ln03a.html   (5)http://starbulletin.com/2002/08/16/news/story3.html