With the reports of yet another devastating earthquake in
Haiti this morning retaining the broader focus and concern the home front was rocked last evening in a different sort of way.
Scott Brown’s defeat of Martha Coakley in a
Massachusetts special election to fill the vacant seat of Ted Kennedy shows up on the political Richter scale. With polls showing a safe lead in one of the most reliably Democratic states little more than a month ago Coakey’s defeat comes as shocking and sobering news to an admittedly reeling Democratic Party. On the flip side the Scott Brown victory is being touted in Republican circles as conclusive evidence of an electorate rejecting current healthcare reform and many of President Obama’s economic policies.
The finger pointing and debating over the significance of this outcome is already underway. How much of it can be attributed to merely the special circumstances of this particular race, a rejection of a Martha Coakely, the infatuation with Scott Brown as an “independent” Republican voice or long-term political trends remains to be seen. For now lets take a deeper look inside exit are pre-election polling to see how this outcome was achieved and which polling organizations were on top of the rapidly changing race.
While not every vote had been counted as of late last night the Brown victory appears to be in the neighborhood of 52-47%. Rasmussen Reports was one of the few organizations to provide sufficient exit polling on the race. Their data as they suggests depicts a “complex picture” but below are some of the more revealing figures.
· Health care (56%) was biggest single issue with
Massachusetts voters. To the surprise of many onlookers Martha Coakley actually outperformed her opponent in this block 53-46%. Perhaps a surge of Massachusetts Democrats hopeful of retaining the seat for their party and assisting in the passage of the healthcare reform bill produced the result.
· 25% of voters listed the economy as their top concern. Scott Brown took home a narrow 52-47% victory in this category mirroring that of his overall performance.
· Only two other issues described as most important to
Massachusetts voters, national security and taxes, accounted for over 5% of the total. Here is where Scott Brown dominated and ultimately won the race. Brown defeated Coakley by a decisive 67-29% margin on security and thumped the Democrat 87-13% on the issue of taxes.
· President Obama continues to be strongly approved of by majorities in
Massachusetts. The correlation between how voters chose their candidates in response to their level of support for the President points to an election that is perhaps less of a referendum on Obama’s job performance than previously believed. Those in the
BayState who strongly approve of the President voted for Coakley 96-3%. Conversely those who strongly disapprove of his job performance went for Brown almost universally, 97-2%. Brown was similarly supported (95%) by those voters who only “somewhat” disapprove of Obama but Coakley (69%) was nowhere near as popular with the group who somewhat approves of the President.
· Did the Governor hurt Coakley’s chances of victory in
Massachusetts? There was also a gap created by those who both support President Obama but disapprove of Democratic Governor Deval Patrick. Accounting for 15% of the electorate Scott Brown took home a huge 93% of that grouping.
One survey that continues to be referred to comes from
University who as recently as early-November showed Democrat Martha Coakley leading Scott Brown by 31-points. Just over two-months later that same polling firm indicated an incredible reversal of fortune for Republicans and a Brown lead of four-percent. Droves of pollsters chimed in on the race in recent weeks. Below are some of the big the winners and losers from last evening in the business of election surveying.
Gold Medal: Public Policy Polling (51-46%)
PPP again shows us the value of large polling samples. Gathering information from 1,231 likely voters, the largest of any poll, PPP nailed the eventual victory for Scott Brown of five-percentage points and their 51-46% margin was remarkably close to the actual result.
The polling firm who brought us the first information on the Massachusetts Senate race also wound up being one of the most reliable.
Suffolk came within a point of accurately predicting the 52-47% race. Polling likely voters down the stretch instead of registered voters from two months ago Scott Brown’s incredible turnaround from a 31-point underdog to a five-point winner may have been less dramatic than it appears.
Bronze Medal: American Research Group (52-45%)
Although sampling less than half of what went into the Public Policy Polling survey ARG did a similarly excellent job gathering results down the stretch of this abbreviated campaign. One recent survey released on January 14th pitted Scott Brown’s lead at three-points. Three days later another ARG poll had Brown up by seven, meaning the average of the two would have accurately predicted the five-point margin.
Other Notables: Fivethirtyeight.com (50-48%), Politico/Insider Advantage (52-43%), PJM/Cross Target (52-42%) Inside Medford/Merriman River Group (51-41%), Daily KOS/Research 2000 (48-48%), Rasmussen Reports (47-49%)
Nate Silver’s Fivethirtyeight.com website does not conduct polling on their own but rather weighs the results of other polling firms through a combination of timeliness, sample size, previous accuracy and other factors. Three other surveys all showed a safe victory for Republican Scott Brown of nine or ten-percentage points. Daily KOS/Research 2000 meanwhile depicted the race as a dead-heat heading into Election Day. Rasmussen Reports gets left off our “biggest loser” list (see below) for three reasons. First off, as stated above they’ve provided the public with helpful exit polling whereas virtually all other polling firms ignored that particular chore. Secondly they were the first major pollster to show Scott Brown within single-digits of Martha Coakley on Janurary 4th. Lastly they reported on some technical difficulties with their website over the past week which could have kept the January 11th poll showing a narrow 49-47% for Coakley as their most recent.
Biggest Losers: Boston Globe/University of
New Hampshire (36-53%), Blue Mass Group (41-49%)
The Blue Mass Group in affiliation with Research 2000 touted their poll as being more insightful than “robo-pollsters” like Rasmussen and PPP by way of a supposedly more efficient live interviewing process. The survey was paid for by the Political Action Committee (PAC), a reputable organization who none the less receives a majority of its donations from solidly Democratic groups. Their results showing a likely victory for Martha Coakley by 49-41% leaves much to be desired. It was not a good week at the polling end of the Boston Globe meanwhile. Not only was their website apparently hacked, reporting on a Coakley victory plush with statistical voter breakdowns hours before polls even closed on Election Day, their big win for the Democratic candidate never came close to materializing. Working off of a University of New Hampshire survey the Globe seemed fairly confident of a big Coakley victory. Her lead of seventeen-points was a whopping twenty-two off the actual result.
PHOTO CREDIT: ASSOCIATED PRESS / Elise Amendola