Tuesday, January 29, 2013

PolitiFact 2012 Truth Index Outbreak Confirmed

Following the election, it was apparent from Facebook posts that PolitiFact had gone into "clean up" mode and had taken a hiatus from fact-checking.  They needed the rest after kicking it up to about 238 rulings in October, about 40 percent more than their usual output.  It updated its Obamameter of promises, and  only two, yes, one plus one, rulings were published between the election and the new year.
 
Because of those two rulings and the PolitiFact states continuing to churn out  rulings at a lower pace, PolitiFact's 1,961 rulings for 2012 exceeded their production for all other years (2011 was 1,919, second place).  If it continued at same rate as in the first ten months of 2012, there would have been over 2,100 rulings for 2012.  Through the end of 2012, since PolitiFact began in 2007, it has done a total of 6,658 rulings.
 

PolitiFact Truth Index Weekly
PolitiFact's Truth Index App has
rather limited info.
As usual, Republicans comprised 55 percent of the rulings in 2012; the Democrats only 37.5 percent.  The congressional majority and the predominance of Republican legislatures and governors in PolitiFact states Florida, Georgia, Virginia, Ohio, New Jersey, Texas and Wisconsin contributed to the Republican numbers.  PolitiFacts New Hampshire, Oregon and Rhode Island, while having more Democrats in power, do not contribute as many rulings overall as the other seven.   Those three states combined contributed 299 rulings for 2012, compared to 1,170 for the other seven heavily Republican states.
 
Recently PolitiFact noted in an article that since the election, the Truth Index had an "outbreak" up the "truthiness scale" in its week-to-week graphic that I'm assuming is available on its android application (I don't have a smart phone because in my circumstances I really don't need one).  Using PolitiFact's Truth Index methodology, the increase in "truthiness" is pretty striking as shown here, smoothed out month to month for the entire year of 2012.  The only other difference to PolitiFact is, it's being shown for Democrats (blue line) versus Republicans (red line), with the grand average as the dotted purple line (which includes those with another or no political affililiation).   The number of rulings published  is shown in parentheses next to the month.
PolitiFact Monthly Truth Index 2012 D vs R
Click to enlarge:  The Truth Index went into a narrow range between parties in the 3
months prior to the election.  Note this is not Truth Index+, but PF's methodology.

The chart has what one might call a "bowl" pattern bottoming out in July; after bounding back in August, the Democrats and Republican resumed a rather closely corresponding line with both pivoting up sharply following the election.  So, yes, this confirms PolitiFact's breakout.
 
The question about this chart that most may wonder about:  what caused the Republicans to "bomb" in July?  The percent of False and Pants on Fires of the total rulings for that month was the highest for the year, at 38.4 percent.  A quick check on Pants on Fire revealed the most likely answer.  July was the "month" of the ruling that received the most votes from PolitiFact readers as the 2012 Lie of the Year, although it was not selected for that "prize" by its editors.  It was the statement about Obamacare being the biggest tax increase in history.  It was repeated that month four times, along with another Obamacare myth about a 3.8 percent real estate transaction tax (ruled on twice).
Finally, what happened to PolitiFact Tennessee (as of January 29, 2013)? They haven't done a ruling since late November, and they are not on the state drop-down at the PolitiFact website.  Searches don't seem to be turning anything up with any sort of "announcement."  Well I hope PolitiFact Tennessee doesn't just waltz away.

Wednesday, January 23, 2013

Confounded by Contextual Complications

It is our policy to make on-the-record email correspondence public where the content affects the context of a story.
Zebra Fact-check, January 7, 2013
 
Fact checkers should emphasize such contextual elements, not bury them.
Sublime Bloviations "Grading PolitiFact", August, 2012
 
(For the second damn time, practice what you preach!)
 
 
Right wingers feel vindicated that Fiat/Chrysler has announced it will be building Jeeps in China, especially at those websites critical of PolitiFact. Instead of beating the same dead horse again and again, I'm going to take another look at this in the terms of the right-wing argument: that if any (even a very small) percentage of production at the Jeep plants in the United States is replaced by production in China, then Mr. Romney's statement (see above graphic) is true in all respects, and needs no further explanation.
 
The right-wing website Zebra Factcheck (ZFC) uses the word "context" in almost every one of its "fact-checks." It seems to feel that it's so important that  they even have a tile symbol for it. For example, in one of its more recent posts on Marco Rubio, it claims that " In effect, PolitiFact takes Rubio’s tweet out of context."
"Context" has its own symbol at Zebra Factcheck (circled)
Zebra Factcheck's littered graphics has a white tile symbol for "Out of Context"
 On his review of Romney's "Jeeps to China" comment, anti-PolitiFact website(s) PolitiFact Bias and Zebra Factcheck writer Bryan White duly notes the "context" he feels is crucially important to Romney's claims.
That’s right. PolitiFact takes a preferred Obama campaign claim about its rescue of the auto industry and states it matter-of-factly in a fact check of Romney. No additional context needed, apparently.

The media treatment of the dispute between the campaigns misses the context while focusing on the particulars. Romney is not simply making statements about Obama’s role in the auto bailout. His campaign is answering the Obama campaign’s messaging and pointing out missing parts of its story.
The only context I can see that was missed (according to Zebra Factcheck) appears to be what was presented by one of Romney's spokespeople to the New York Times in response to Romney's ad:
The facts are clear: despite his false and misleading attacks, President Obama took the auto companies into bankruptcy,” said a spokeswoman, Amanda Henneberg. “His mismanagement of the process has exposed taxpayers to a $25 billion loss.”
Romney had proposed "letting Detroit go bankrupt" in opposition to what President George Bush decided to initiate with a bridge loan to GM and Chrysler in 2008. He said the government "should provide guarantees for post-bankruptcy financing." But even Michigan's future (Republican) Governor Rick Snyder and (Republican) Congressman Fred Upton, acknowledged that "there was no one that could have picked up those pieces other than the federal government." The respected, not really liberal Economist had this to say: