Biggest Senate upsets since 1978

[Mr.Phips]

In 1978, there were several upsets that almost nobody saw coming.  One was Gordon Humphrey's defeat Tom McIntyre in New Hampshire.  McIntyre was above 50% and leading by 20+ points right up until the polls closed.  However, Humphrey ended up winning by two points.

Another big upset that year was Roger Jepsen's defeat of incumbent Dick Clark in Iowa.  Clark was ahead by double digits right up until the end.  However, Jepsen won narrowly.

The final big upset that year was incumbent Charles Percy's win over challenger Alex Seith in Illinois.  Not only did Seith lead Percy in final polls, but he was often above 50%.  Somehow, the polls were totally wrong and Percy won narrowly.

In October 1984, Mitch McConnell was trailing incumbent Walter Huddleston in Kentucky by almost 30 points, with Huddleston close to 60% in the polls. However, a late shift and Reagan's coattails allowed McConnell to pull off a huge upset that even Republicans were shocked by.

In the fall of 1988, Senator John Melcher of Montana was ahead of Conrad Burns by about 20 points, hovering around the 50% mark or a little above.  However, a late surge and a campaign appearance by George Bush allowed him to unseat the incumbent by a 52%-48% margin in a race that really shocked a lot of election watchers.  I remember the news outlets actually calling this race for Burns with about 20% reporting based on exit pols with Melcher actually in the lead until about 50% reported.

The last big upset I can think of is Saxby Chambliss' 53%-46% win over Max Cleland when no poll showed him ahead.  Most polls showed Cleland ahead by high single digits and in the high 40's.  Polling appeared to overestimate black turnout and underestimate Republican turnout in the suburbs.   

[Mr.Phips]

John Edwards defeating Lauch Faircloth in 1998 was considered a mild upset, IIRC.

Most observers actually expected Faircloth to lose by October.  He was starting to trail in polls and even fired his pollster. 

[Mr.Phips]

Harry Reid's 2010 victory was pretty unexpected, as was Bill Frist's defeat of Jim Sasser in 1994.

Frist clearly had the momentum in the final week and even started leading Sasser in many polls.  It was the margin that shocked everybody. 

[Mr.Phips]

2004 Maj L Daschle, surprising that a sitting opposition leading is defeating for reelection and IL 2010, Kirk winning Obama's ol' senate seat. Shoe-in for Democrats.

Daschle was behind in pretty much every poll after September and was considered more likely than not to lose by election day.  Kirk was ahead in most polls, although Obama could have done a lot more to help Giannaoulias.  No excuse for that loss. 

[Mr.Phips]

1) The conservative Peter Fitzgerald defeating Moseley-Braun in IL in 1998? Actually, you could say he almost blew that one. One more week and Moseley-Braun, who had all the momentum in the last two weeks probably overtakes him. Still though, a conservative republican winning a senate seat in IL? Unlikely.

2) Saxby Chambliss defeating Max Cleland in GA in 2002 was not expected.

3) Maria Cantwell knocking out Slade Gorton in WA state in 2000. At the beginning of that race (despite Gorton's relatively weak win compared to the rest of the GOP in '94), most thought Gorton would win.

4) Stabenow's victory over Spence Abraham in 2000.

The only thing surprising about the Mosely-Braun loss was that she almost won.  She was behind by double digits for most of September and October, although the final Zogby poll put her ahead 45%-42%.

Gordon's loss wasnt that surprising. Most fall polls showed the race pretty tight.

The 2000 Michigan race was one that closed fast in the final week and Stabanow was almost certainly pulled accross the finish line by Gore's larger than expected victory in the state.