Fact check in order for President Obama, Martha Coakley campaign Web site (Updated)

In the Massachusetts special election for the Senate, Democratic candidate Martha Coakley has portrayed Scott Brown as a Republican "in lock step" with the Republican Party.

President Obama, during his remarks supporting Coakley at a campaign appearance yesterday, repeated the Coakley charge:

So I hear her opponent is calling himself an independent. Well, you've got to look under the hood -- (laughter) -- because what you learn makes you wonder. Now, as a legislator, he voted with the Republicans 96 percent of the time -- 96 percent of the time. It's hard to suggest that he's going to be significantly independent from the Republican agenda. When you listen closely to what he's been saying, it's very clear that he's going to do exactly the same thing in Washington.
(Chicago Sun-Times, yellow highlights added)

Scott Brown allegedly votes with "Republicans" 96 percent of the time.  Is it true?

Martha Coakley's Web site provided additional detail on a page called "Rhetoric vs. Reality":

Brown On lockstep Republican Voting Record

Rhetoric:

Brown:  The allegation that I vote 96 percent of the time with Republicans is inaccurate." (Boston Globe, 1/14/10)

Reality:

Brown has voted with the Minority Leader of the Massachusetts Senate, Richard Tisei, 546 times from 2007-present.  During that same period Brown voted against the Minority Leader only 25 times.  In other words, Brown has voted with the current Republican Minority Leader 96% of the time from 2007-present [Insta-Trac/The Advance Research Group Database]
(Web site style faithfully reproduced)

Two questions ought to immediately occur to the would-be fact checker.  First, why or how does Richard Tisei represent "Republicans"?  Second, given the potential for votes that are neither "yea" nor "nay," does the math add up?


Was Tisei properly representative of "Republicans"?

Insta-Trac's description of its service offers some insight into the choice of Tisei:

• See a legislator’s votes on all law enforcement votes
• Or a sub-set of the law enforcement cluster, such as a member’s votes just on guns and weapons
• How one legislator voted with or against one or more other legislators
• The missed votes of a legislator
• How a legislator voted with and/or against the leadership of his/her party

The Coakley campaign, then, had a number of options open when it came to comparing Scott Brown's voting record.  The best way to determine how Brown voted compared to "Republicans" would have involved using the third option listed above.  Take all the Republican votes on a given issue, and determine whether Brown voted with or against the majority of Republicans.

The Coakley campaign conspicuously avoided using that method.  They did produce a comparison of Brown's votes compared with "the leadership of his/her party," however, finding that Brown voted with the leadership 90 percent of the time over the course of his career in state government.

These numbers produce many questions about the methods used by the Coakley campaign.  Why use the Tisei comparison if one to Republican leaders as a group was available?  And why not do the comparison as it should have been done?


Do the numbers add up?

Another page at Coakley's Web site offers an entirely different set of figures for the same comparison even though the two sets were published less than week apart.  The numbers quoted above result in a figure of 95.6 percent, which is fairly rounded up to 96 percent.  From the earlier (Jan. 10) page, however, the figure comes to 94.7 percent.  The latter is not fairly represented as 96 percent.  One set of figures is wrong, at minimum.

I have a request for a free trial of Insta-Trac's service pending.

Afters:

Will PolitiFact check President Obama's accuracy?  Don't hold your breath.


Update:

This story apparently has some legs.

I have tentative evidence that PolitiFact is interested in this story, but from the perspective of fact checking Scott Brown's denial that he voted with Republicans 96 percent of the time.

A note for my readers:  I haven't shown my entire hand on this item.  I have avoided mention of the aspect of the story that I would most expect reporters with a liberal bias to flub.  Don't let me down, PolitiFact!