Wednesday, May 30, 2012

Monday, May 28, 2012

Education opportunities for veterans

The enterprising and ubiquitous June Olsen recently asked for space to present a post about education opportunities for veterans.

Olsen seems on the up and up, but her subject matter is far from the normal fare here, so rather than having a guest post substantially out of character with focus of Sublime Bloviations I proposed directing attention toward Olsen's mission near some appropriate holiday.

It's Memorial Day.

Supreme thanks to all who have fallen in defense of liberty (ours and that of others).

Here's a representative example of Olsen's writing at MilitaryBases.com.

Sunday, May 27, 2012

Nutting doing: PolitiFact's inadequate excuse

This week many liberals jumped on the meme that President Obama has the lowest spending record of any recent president.

Fortunately for all of us, PolitiFact was there to help us find out the truth in politics.

Actually, PolitiFact completely flubbed the related fact check.  And that's not particularly unusual.  Instead, it was the Washington Post's Glenn Kessler and an Associated Press fact check that helped people find the truth in politics.

PolitiFact isn't backing down so far, however.  On Friday PolitiFact offered the following response to the initial wave of criticism (bold emphasis added):
(O)ur item was not actually a fact-check of Nutting's entire column. Instead, we rated two elements of the Facebook post together -- one statement drawn from Nutting’s column, and the quote from Romney.

We haven't seen anything that justifies changing our rating of the Facebook post. But people can have legitimate differences about how to assign the spending, so we wanted to pass along some of the comments.
(Image captured by Jeff Dyberg;
 click image for enlarged view)

PolitiFact also made the distinction on Twitter.

There's a big problem with the attempt to distinguish between checking Nutting's claims and those from the Facebook post:  The Facebook post argues implicitly solely on the basis of Nutting's work.  PolitiFact likewise based its eventual ruling squarely on its rating of the Nutting graphic.

PolitiFact (bold emphasis added): 
The Facebook post says Mitt Romney is wrong to claim that spending under Obama has "accelerated at a pace without precedent in recent history," because it's actually risen "slower than at any time in nearly 60 years."

Obama has indeed presided over the slowest growth in spending of any president using raw dollars, and it was the second-slowest if you adjust for inflation. The math simultaneously backs up Nutting’s calculations and demolishes Romney’s contention.
Credit PolitiFact with accurately representing the logic of the implicit argument.  Without the fact check on Nutting's work there is no fact check of Romney's claim.  Making matters worse, PolitiFact emphasized the claim that Obama "has the lowest spending record" right next to its "Mostly True" Truth-O-Meter graphic.  The excuse that PolitiFact was fact checking the Facebook post completely fails to address that point.  Andrew Stiles is probably still laughing.

Criticisms of Nutting make clear that the accounting of bailout loans substantially skews the numbers in Obama's favor. Using the AP's estimates of 9.7 percent for 2009 (substantially attributable to Obama) and 7.8 percent in 2010, Obama's record while working with a cooperative Democrat-controlled Congress looks like it would challenge the high spending of any of his recent predecessors.  The leader from the Facebook graphic, President Reagan, tops out at 8.7 percent without any adjustment for inflation.  PolitiFact's fact check was utterly superficial and did not properly address the issue.

There is a silver lining.  The Obama administration has so aggressively seized on this issue that PolitiFact will certainly feel pressure to fact check different permutations of Nutting's claims.

I can't wait to see the contortions as PolitiFact tries to reconcile this rating with subsequent attempts.


Saturday, May 26, 2012

Grading PolitiFact: Obama's "Julia" and Medicare

Context matters -- We examine the claim in the full context, the comments made before and after it, the question that prompted it, and the point the person was trying to make.
--Principles of PolitiFact and the Truth-O-Meter
I've written quite a bit about PolitiFact's problem with selection bias.  This fact check serves as an excellent example of a related and perhaps even more pernicious form of bias, that arising from the selection of a story's focus.  Two stories dealing with the same topic, such as the Medicare claim in the Obama campaign's "Julia" ad, can turn out very differently.  And the ideology of the writer or editor can easily alter the story focus.


The issue:

(clipped from PolitiFact.com)


The fact checkers:

Becky Bowers:  writer, researcher
Angie Drobnic Holan:  editor


Analysis:

As I pointed out in an earlier review of a "Julia" ad fact check, the context of the Obama campaign's claims about Mitt Romney has to do with hypothetical comparisons between her life under Obama's policies and Romney's policies.  Here's a closeup of the relevant part of the ad:

(click image for larger view)
For Obama we see the best-case scenario.  For Romney we get the worst-case scenario.  Medicare "could" end as we know it under either president's policies extrapolated into the future.  In Obama's case Medicare could end through a rationing regime.  In Romney's case it would end by adding expenses for recipients of the benefits to help mitigate the third-party payment dynamic that makes Medicare spending unsustainable.  Obama's argument is cherry-picked.

PolitiFact ignores Obama's point of comparison, narrowing the focus specifically to the claim regarding Romney's plan for Medicare:
We wondered, under Romney’s policies, is it true that, "Medicare could end as we know it, leaving Julia with nothing but a voucher to buy insurance, which means $6,350 extra per year for a similar plan"?
PolitiFact, at least for purposes of this fact check, does not wonder whether it is fair and proper for the Obama campaign to compare the cherry-picked rosy picture of Medicare under Obama with the cherry-picked bleak prospects for Medicare under Romney's supposed plan.  Julia's prospects under Romney could have read "Julia enrolls in Medicare, helping her to afford preventive care and the prescription drugs she needs."

Unfortunately for the Obama campaign, PolitiFact did find a few things wrong with the ad.

1)  It bases its statement of Romney's position on the older Paul Ryan budget proposal.  Romney has said he agrees in principle with the bipartisan Wyden/Ryan plan which is similar in some ways to Ryan's old plan and different in others.  Plus Romney has a plan of his own about which we know few specifics.

2)  Because of the outdated analysis, the ad has no foundation for its claim about the $6,350 voucher figure it cites.

3)  Various features in each of the plans undercut the claim that "Julia" would end up with "nothing but a voucher to buy insurance."

But the news for the Obama campaign wasn't all bad.

PolitiFact originally rated the claim "False," but after "readers" pointed out an error in the story, PolitiFact reconsidered and upgraded the rating to "Mostly False."  Woohoo.

What was the error?

PolitiFact: 
(R)eaders pointed out we had incorrectly said that the Wyden-Ryan and Romney plans for Medicare offered traditional fee-for-service Medicare alongside a premium-support or "voucher" system for purchasing private insurance. Instead, both plans offer traditional Medicare inside of a new premium-support system, as another plan that competes with private insurance.
One might well wonder why that fact makes a difference.  Unfortunately, PolitiFact made the original story unavailable.  That makes finding all the specific changes difficult.  But the summary paragraph likely includes most or all of the justification for the higher rating.

PolitiFact:
The Obama campaign ties Romney to an outdated plan with less generous spending growth and no traditional Medicare option. Romney does support a voucher-like system, but the graphic ignores critical facts that would give a different impression about his plan. We rate the claim Mostly False.
See, the Romney plan is "voucher-like" so saying his plan could only give "Julia" a voucher is true enough to qualify as "Mostly False."

By narrowing the focus to a fine point PolitiFact locates a portion with sufficient truth to avoid a "False" rating.  The problem with the fact check isn't the rating on this narrow point.  The problem comes from PolitiFact failing to provide any mention of the way the particle of truth powers a dynamo of deception.  The ad's argument is a stinker.  PolitiFact's ruling gives it more cover than it deserves.


The grades: 

Becky Bowers: D
Angie Drobnic Holan:  D

I've seen so many fact checks much worse than this one from PolitiFact that perhaps I'm being overly generous.  Bottom line:  A serious fact check takes into account the underlying argument and the context.  The overall argument of the Medicare portion of the "Julia" ad was deceptive and designed to scare voters with its manipulated comparison.  PolitiFact did not allow its focus to fall on the main issue.

Thursday, May 24, 2012

PolitiFlub: When words don't really matter that much

The impetus for this PolitiFlub comes from PolitiFact Texas, rating a claim by Democrat Gilberto Hinojosa.

PolitiFact quotes Hinojosa:
"A large majority of the Republican Party believes that this man is a Muslim and was born in a foreign country, was not born in the United States," he said at the Central Texas Democratic Forum on April 24, 2012.
I'm in favor of charitable interpretation for all, but it's appropriate to at least mention each reasonable interpretation.

Hinojosa may have intended to say that a majority of Republicans believe both (and not or) propositions.  In other words, the exact same group in the sample supposedly holds both beliefs.

For example (figures simply for the sake of argument):

52 percent believe Obama is a Muslim and born outside U.S.
8 percent believe Obama is a Muslim, but born inside U.S.
11 percent believe Obama is not a Muslim but was born outside U.S.
(71 percent believe Obama is a Muslim or born outside U.S.)

Consistent with PolitiFact's statement of principles, the fact check should have considered the less charitable interpretation.  It is the best practice to inform the reader when one is using the principle of charitable interpretation in a fact check.

In this case, contrary to PolitiFact's principles, words weren't important.  Hinojosa's specific wording was forgiven without comment in favor of a less literal interpretation.

Those wacky politicians and their fundraising letters

Got a request for campaign dollars today from a Democrat running for Congress.
As the daughter and granddaughter of Holocaust survivors, I have always believed strongly in the tenants of fairness, equality, and justice – in our laws and in our country.
Doubtless this candidate also fully supports the public school system.

PolitiFlub: President Obama's near-sterling record on spending?

MarketWatch's Rex Nutting published a defense of President Obama's spending record on May 22.

Quite a number of conservatives have written to rebut Nutting regarding his point that Obama has not presided over an unusually large increase in federal government spending.  But, strangely, conservative-leaning PolitiFact did not join the chorus of criticism.

Instead, PolitiFact rated a secondhand viral Facebook version of Nutting's argument "Mostly True."

The problem? The flub, that is?

PolitiFact entirely ignores the fundamentally accurate criticism undercutting Nutting's argument in favor of Obama's light-spending ways:  The Bush budget of 2009 provides a very unusual baseline against which to compare Obama's spending.

PolitiFact simply closes its eyes to that context:
Our ruling

The Facebook post says Mitt Romney is wrong to claim that spending under Obama has "accelerated at a pace without precedent in recent history," because it's actually risen "slower than at any time in nearly 60 years."

Obama has indeed presided over the slowest growth in spending of any president using raw dollars, and it was the second-slowest if you adjust for inflation. The math simultaneously backs up Nutting’s calculations and demolishes Romney’s contention. The only significant shortcoming of the graphic is that it fails to note that some of the restraint in spending was fueled by demands from congressional Republicans. On balance, we rate the claim Mostly True.
The rest of the fact check is equally blind to the critical context of the claim. 



Wednesday, May 23, 2012

A perfect time to oppose efforts to protect against voter fraud

What is the extent of voter fraud in the United States?  I sure don't know.  But I find the war between the opposing sides fascinating.

Robyn "Blumñata" Blumner, hard left editorialist for the Tampa Bay Times, enters the fray with "Odd time to purge voter rolls."  Because one wouldn't want to have the purest voter roll just before an important election, right?  Rather, you want to purify the rolls immediately after a big election.  Or at least at some point far enough away from the election so the purge has time to unpurge.

So much for the silly title and on to the silly content.

Our jumping off point is the new Florida effort to purge voter rolls of non-citizens.  Up, up and away:
The process raises uncomfortable comparisons to the Jeb Bush-era error-ridden felon list. That purge list was used to prevent thousands of legitimate voters from casting ballots in the 2000 presidential election — an election decided by 537 votes.
The list prevented thousands of legitimate voters from casting ballots in 2000?   That's news to me.

 Fact check time.

The executive summary of an investigation of the 2000 election by the couU.S. Commission on Civil Rights was apparently not the source of Blumner's claim.  One would think that the direct disenfranchisement of thousands of legitimate voters would serve as exhibit A in such a report, presuming the Commission detected the problem, though an early draft of the report obtained by the Washington Post included the following:
Perhaps the most dramatic undercount in this election was the nonexistent ballots of the countless unknown eligible voters, who were wrongfully purged from the voter registration rolls, turned away from the polls, and by various other means prevented from exercising the franchise. While statistical data, reinforced by credible anecdotal evidence, point to widespread disenfranchisement and denial of voting rights, it is impossible to determine the extent of the disenfranchisement or to provide an adequate remedy to the persons whose voices were silenced in this historic election by a pattern and practice of injustice, ineptitude and inefficiency.
"Countless," so it might have been in the billions.  Though more likely the vague "countless" means they don't have any idea how many.  Several people turned up at the hearings to claim such disenfranchisement.  Assuming their stories are true the number is at least greater than zero.

Hunting in the USCCR data for confirmation of the disenfranchised thousands Blumner alleges turned up nothing.  Agenda journalist Greg Palast named no such number of lost votes.  Lost potential votes, yes--but Blumner uses no such qualifying language.

Don't buy Blumner's number unless it comes with reliable backing.  And good luck ever finding any.

More from Blumñata:
Detzner claims that 180,000 registered Florida voters may not be citizens after their names and other data were "matched" to foreign nationals in an outdated state motorist database. These voters could easily have been naturalized in the years since obtaining or renewing a driver's license or state I.D.
Obviously if the members of the set of 180,000 naturalized at some point then they count as citizens.  However, if they didn't naturalize then they don't count as citizens and it remains true that "180,000 registered voters may not be citizens" until some determination takes place.  Blumner's gratuitously playing up the obvious and admitted lack of certainty regarding the number among the 180,000 who qualify as citizens.


(click image for larger view)

 Blumñata:
Already some of the 2,700 noncitizens on the verified purge list are proving to be citizens. Which leads to the question of whether this is a pure effort to clean up the voter rolls or is there an element of suppressing minority votes?
Fact check.

What is the "verified purge list"?  Supposedly a list from the state that includes only confirmed non-citizens?

Blumner's own paper gives us the answer:
As a result, some citizens could appear to be non-citizens now because the DHSMV computer system doesn't automatically update when someone becomes a citizen, said Chris Cate, a spokesman with the Florida Division of Elections.

"We're not just looking at the matches from highway safety," Cate said. "We're doing a secondary assessment here before we send the names to supervisors of elections. You have to consider that a person's last contact with highway safety can be more than four years ago. Someone could have become a citizen in that time. So you can't presume someone's not an eligible voter."
Is Blumner trying to mislead readers or what?  The state knows that legitimate voters could occur on the list.  That's exactly why they send the list to local election workers. The local workers are charged with trying to verify citizenship.  The best time to do that is in advance of a coming election while allowing enough time to correct as many individual snags as possible.

Blumner:
Monroe County Supervisor of Elections Harry Sawyer, a Republican who is a straight shooter on election matters, says that some of the records Detzner is using are too old to be reliable. Without a confirmation on citizenship one way or another from the voter, Sawyer says, he won't automatically drop anyone off the rolls. But not every local elections supervisor will be this careful.
> Likely de-obfuscation of Blumner's paraphrasing:   Sawyer's stating the obvious again that persons with older driver's licenses are more likely to have achieved citizenship in the interim.   Presenting the same information twice doubles the amount of evidence.  De-obfuscation No. 2:  Sawyer's not acting to remove people from the rolls until he receives confirmation because he's limited from acting because of federal restrictions on his and four other Florida counties.  The information again comes from Blumner's own paper.  And she forgets to mention this?

Blumner:
Florida's noncitizen purge follows those in Colorado and New Mexico. Funny that this is being done by immigrant-heavy swing states with Republican secretaries of state. Both Colorado and New Mexico made a media splash with dire claims that noncitizens were registered and likely voting in large numbers. But a closer look by the nonprofit Brennan Center for Justice found that those states had drawn indefensible conclusions about noncitizen voting and had refused to release evidence that backed up their claims. The allegations of voter fraud, the center suggests, were smoke and mirrors and couldn't be trusted.
Members of the Democratic Party should feel outrage at Blumner's implication that Democrat secretaries of state would resist making it more difficult for non-citizens--who tend to vote Democrat--to vote.

The Brennan Center for Justice evaluation, by the way, has a little hypocrisy problem:
A careful review of the reports and information released (they have not released the complete set of data and methodology used to arrive at their conclusions) shows some serious problems with the methods used based upon the information they did disclose resulting in questionable conclusions.
A definitive finding of "serious problems" depends on a full evaluation of the methodology, at least in a case like this.  None of the supposed problems lacks for a potential solution.  So the evaluation suffers a problem similar to that of which it accuses Colorado and New Mexico.

It should come as no big surprise.  The evaluation was written by Senior Counsel in the Brennan Center's Democracy Program, Keesha Gaskins.  Lawyers too often don't care about the legitimacy of their arguments.  They're often swayed by the desire to convince the audience by any means, including by the use of a host of formal and informal fallacies.  Not that we should dismiss Gaskins' arguments because of that lawyerly tendency.  Each argument stands or falls on its own.

Case in point (Blumner's a former lawyer):
The same can be said for all the dead people voting in South Carolina. That state's attorney general, Republican Alan Wilson, claimed recently that more than 900 votes had been cast by dead people. But after the South Carolina Election Commission looked into the claim, it didn't hold up. There was no evidence that anyone had fraudulently voted in the name of someone dead.
Apparently a confession from the illegal voter is a necessary proof.

Seriously, an investigation of the 207 potential votes by dead people ended up ruling out voter fraud, on reasonable grounds, in all but 10 cases.  In lawyer world that's "no evidence that anyone had fraudulently voted in the name of someone dead."  The investigation did not encompass another 696 suspicious votes from earlier elections.  See what I mean about lawyers?  No evidence isn't the same as no proof.
This is part of a pattern. Republicans actively gin up voter fraud claims to justify turning voting into an obstacle course to dissuade Democratic-leaning constituencies.
Blumner's hilarious.  News reports indicate that the information about potential votes by dead people was uncovered after South Carolina started implementing a voter ID law.  Some pattern.



Tuesday, May 22, 2012

PolitiFlub: PolitiFact Florida and the loss of existing insurance

PolitiFact Florida published a fact check yesterday parallel to one I blasted just last week concerning the loss of existing insurance plans with the implementation of the health care reform bill.

The PolitiFact Florida conclusion:
Our ruling

The U.S. Chamber said, "Obamacare could cause 20 million people to lose their current coverage." It's a claim that is oft-repeated and much exaggerated.

The chamber employs a worst-case projection by a nonpartisan research agency. The agency’s other forecasts are lower, but you wouldn’t know that from watching the ad. And unlike Priebus, the group does not specify that the type of insurance potentially affected is a specific type of insurance -- "employer-based" insurance.

Plus, this 20 million estimate only counts people who receive coverage from their employer -- and not those who might receive better coverage elsewhere.

Most importantly, this figure does not represent uninsured people who will get coverage because of the law.

We rate it Mostly False.
How doth PolitiFact mislead thee?  Let me count the ways ...

1)  "The chamber employs a worst-case projection by a nonpartisan research agency"  Following as it does on the heels of "oft-repeated and much exaggerated," this sounds bad.  But the worst case scenario is exactly how one ought to support a claim of potential damage ("could cause 20 million people to lose their current coverage").  PolitiFact's observation is a positive posing as a negative.

2)  "The agency’s other forecasts are lower, but you wouldn’t know that from watching the ad." Hidden context?  One wouldn't know that "could cause 20 million" calls for a high-end estimate from reading PolitiFact's fact check.  Basic English interpretation tells the viewer that "could cause" implies a high-end estimate and a worst-case scenario.  But the 20 million figure is neither of those in the context of the ad.

3)  "(T)he group does not specify that the type of insurance potentially affected is a specific type of insurance--'employer-based' insurance."  PolitiFact doesn't tell you that by using a number specific to one type of insurance it implies that any loss from other types of insurance gets added to the worst-case scenario number.  Consider all types of insurance and the number can only be equal or higher.  So how is it misleading except by underestimating the worst-case scenario?

4)  "Plus, this 20 million estimate only counts people who receive coverage from their employer -- and not those who might receive better coverage elsewhere."  This isn't even a coherent objection.  PolitiFact is probably trying to say that the 20 million figure includes those who opt for "better" coverage apart from an employer-offered plan.  The report on which PolitiFact relied makes no reference to spurning the employer offer for a "better" plan unless by "better" PolitiFact means cheaper.  The report indicates that customers offered more than one option for health insurance strongly prefer less expensive plans.  The report does not mention any alternative motivation, so PolitiFact apparently dreamed up the one it mentions.  It's possible to argue that those who leave employer insurance because of pricing are not forced into cheaper insurance plans, but PolitiFact's implication that consumers freely choose better plans obscures the factors at work and allows readers to assume that the "better plans" idea comes from the CBO/JCT report.  It doesn't.

5)  "Most importantly, this figure does not represent uninsured people who will get coverage because of the law."  A figure for persons losing current coverage should never represent uninsured people who will get coverage because of the law.  Uninsured people have no existing coverage to lose in the first place.  PolitiFact's most important point is irrelevant to the accuracy of the 20 million figure.  At most, it is relevant as context trying to balance the 20 million figure against a benefit of the health care reform law.

6)  A real fact check would have looked for types of existing insurance other than employer insurance likely lost as a result of the health care law.  The same CBO/JCT report lists such losses on the line below losses to employer insurance.  PolitiFact somehow neglects to notice (the figure is an additional 3 million for the baseline scenario and an additional 1 million for the worst-case scenario):

http://www.cbo.gov/sites/default/files/cbofiles/attachments/03-15-ACA_and_Insurance_2.pdf

Nice job of plowing another train into your existing train-wreck, PolitiFact.
We rate it Mostly False.
Trading on what credibility, pray tell?

How did this fact check pass an editor's examination?

PolitiFlub: Comparing apples and appanges?

PolitiFact determines whether Mitt Romney flip-flopped on President Obama's stimulus plan:
Our ruling
It’s misleading to say that Romney flip-flopped on supporting "the president's Recovery Act." At the time of the CNN interview, no Obama stimulus bill had been introduced, much less enacted. With such complex bills, the devil’s in the details.

But before Obama took office, Romney had spoken in favor of a stimulus -- particularly if the stimulus in question had been passed when a Republican was in the White House. We rate this a Half Flip.
 So he never supported Obama's stimulus but half-flipped on supporting Obama's stimulus.

That's because he supported some type of stimulus according to the description he gave prior to Obama taking office.

This is a little like flip-flopping on Pepsi cola.  You've never liked Pepsi cola, but you do drink Coca-cola.  Since you like Coca-cola you've half-flipped on liking Pepsi cola.

It makes perfect sense once you think about it.

Sunday, May 20, 2012

Grading PolitiFact (Ohio): The Wurzelbacher dilemma: Is an Air Force plumber a plumber?

Context matters -- We examine the claim in the full context, the comments made before and after it, the question that prompted it, and the point the person was trying to make.
--Principles of PolitiFact and the Truth-O-Meter

 Horse feathers.


The issue:

(clipped from PolitiFact.com)


The fact checkers:

Sabrina Eaton:  writer, researcher
Robert Higgs:  editor


Analysis:

As one might anticipate from the epigraph, the context serves as the key to this fact check.

Most (or at least many) people know of "Joe the Plumber," Joe Wurzelbacher, a Toledo man vaulted to modest fame after an unscripted and candid conversation  in 2008 with a campaigning Barack Obama.

Since that initial boost to notoriety, Wurzelbacher entered into a congressional race as the Republican candidate in a redrawn Ohio district against veteran of Congress Marcy Kaptur.  Kaptur squeezed out fellow Democrat and congressional veteran Dennis Kucinich in the Democratic primary.

The fact check concerns Wurzelbacher's status as a plumber, the Kaptur campaign's attempts to exploit that status as a campaign issue and Wurzelbacher's response to Kaptur's tactics.  PolitiFact fact checks the latter, and the former two constitute the critical context.

We'll examine all three issues.

Is Wurzelbacher a plumber?

Friday, May 18, 2012

Another truth-hustler: Eric Alterman

What is a "truth-hustler"?

I've defined it as one who writes on the issue of truth while spinning on behalf of a partisan agenda.

Eric Alterman, formerly part of the left's media influence machine Media Matters for America, comes to our attention as another example of a truth-hustler via his "As a Matter of Fact" entry in his "Think Again" series at the Center for American Progress.

I'm skipping the portion dedicated to Alterman's obsession with Koch Industries.  That leaves us plenty to work on.

Reality-biased?

Alterman presents us with the proposition that "reality has a liberal bias" may serve as a valuable device in explaining (or at least describing) a supposed tendency for conservatives to favor ideology over fact.  The trick, of course, is for Alterman to make his case without favoring his ideology over the facts (hat tip to Dr. Sanity, who is among those to beat me to coining the term "reality-biased" as a play on Left's "reality-based" mantra). 

Does he have good evidence?

Working the refs?

Without seeming self-conscious about either the aim of his column or his history with Media Matters, Alterman tells us that conservatives are effectively working the media referees.  Instead of providing any example in support of his claim, Alterman offers up this odd portrayal of Mark Hemingway's critique of media fact checking:
As The Weekly Standard online editor Mark Hemingway complains, when such fact-checking organizations tend to find many more conservative lies than liberal ones, rather than respond that conservatives tell far more lies, the fact-checkers go looking for liberal fabrications and find them whether they exist or not. How else could the priests of false equivalence maintain their law that "both sides do it" -- what I call "on-the-one-handism" -- which has proven to be a fundamental tenet of Beltway faith?
Somebody please read Hemingway's article and find the complaint described by Alterman, for I can find no evidence of it, unless Alterman only intended the "such fact-checking organizations tend to find more conservative lies than liberal ones" to represent the entirety of Hemingway's relevant complaint.


The refs want to look fair!

Before we know it, Alterman has drifted from conservatives supposedly working the refs to the fact-checkers' attempts to appear fair (Alterman, of course, isn't working the refs!):
The American Prospect's Paul Waldman discerned this dynamic at work in 2011, when he bravely predicted that yet another of these fact-checking organizations, PolitiFact, would choose as its (gimmicky) lie of the year a "'lie' told by Democrats, even if the one it picks is far from the most egregious lie told this year, or even really a lie at all" -- owing to the fact that the previous two years, Republicans had been caught fibbing. Three times in a row is enough to cry partisanship whether it exists or not, so Waldman was on pretty strong ground despite the damage that such a strategy might do to the organization's reputation for intellectual integrity.
I'll partially grant Alterman's point with this example, except it doesn't appear to serve as an example of  "working the refs."  And if predicting the award would go to a Democrat indicates bravery then pin a medal on my co-editor at PolitiFact Bias, Jeff Dyberg, who confidently predicted an Obama win in 2011 for the "Lie of the Year."  On the other hand, the "Lie of the Year" isn't a typical fact check so it doesn't serve Alterman as a particularly strong example, particularly when every past "Lie of the Year" was prone to the same types of criticisms about its legitimacy.

Alterman's subsequent example, Jon Stewart's claim about Fox News viewers being the most misinformed, fares worse if anything.

Alterman's examples actually suggest that some on the Left take poorly-supported propositions as truth and count against others the failure to equally subscribe to those dubious propositions.

The truth, no spin

Alterman's right that fact checkers want to appear fair, and I would grant that PolitiFact's 2011 "Lie of the Year" may reflect an influence from that motivation.  But studies show that referees are biased.  For example, referees working home games tend to favor the home team.  Is that relevant when audience feedback for fact check stories comes primarily from the political left?  And when journalists as a group tend well left of the general population?

Simply in terms of science, we can't accept without good evidence a narrative from liberals that liberals have a better handle on the truth.  We clearly can't just take their word for it.  We shouldn't accept as evidence studies with rampant selection bias problems.  And we should view with suspicion writers such as Alterman who trumpet the high value of dubious evidence.

Thursday, May 17, 2012

The delusions of truth-hustler Chris Mooney

I call the new breed of writers who focus on issues of truth while applying their own spin to truth by the term "truth hustlers."

Chris Mooney is one such.

Mooney writes on a number of subjects concerning science and recently published a book called "The Republican Brain."  I'm focused on Mooney's recent article for the The Nation, since I'm satisfied with the review Mooney's book received from one of his erstwhile colleagues at the Center for Inquiry.

In "Reality Bites Republicans" Mooney continues pushing the meme that liberals are somehow more attuned to the truth than conservatives, with most of his fodder coming from the stats connected to the liberal fact checkers at PolitiFact and the Washington Post.

Mooney:
The Smart Politics blog went on to suggest, based on these numbers, that PolitiFact is biased against the right—precisely the type of knee-jerk centrism that Mann and Ornstein have called into question.
Knee-jerk centrism is hardly worse than knee-jerk liberalism.  So what knee-jerk, if any, are we looking at, here?

I've written a number of times about the Smart Politics study by Eric Ostermeier.  People, apparently including Chris Mooney, don't appear to understand the basis for Ostermeier's criticism.

Ostermeier criticized PolitiFact on the basis of a selection bias problem.  Mooney, cloaked as he is in an ostentatious mantle of science, ought to have thorough familiarity with the concept.

Ostermeier looked into the method PolitiFact uses to choose which claims it would check, and found PolitiFact editor Bill Adair saying "We choose to check things we are curious about. If we look at something and we think that an elected official or talk show host is wrong, then we will fact-check it."  Ostermeier questioned how the above method could explain the numbers:  a roughly even number of ratings from the Republican and Democrat groups with one group faring significantly worse in the ratings.  Ostermeier's making the point that the PolitiFact data do not constitute anything like a fair scientific experiment designed to figure out which party does a better job as a steward of the truth.

Mooney, if he has the type of familiarity with science that he tries to project in his writing, surely knows that the numbers from PolitiFact do not lend themselves to scientific conclusions.  Yet somehow he manages to produce that impression.

Mooney:
After all, there is another possibility: the left just might be right more often (or the right, wrong more often), and the fact-checkers simply too competent not to reflect this—at least over long periods.
Ostermeier allowed for that possibility despite the "knee-jerk" characterization from Mooney.

Mooney:
So which interpretation is correct? While certainly not definitive, a study undertaken for my book The Republican Brain and updated for this article—with dedicated data-gathering and statistical analysis from an assistant, Aviva Meyer—lends additional credence to the latter possibility.
Mooney's study doesn't lend any "additional credence to the latter possibility."  His study of the Washington Post "Fact Checker" column in essence duplicates the Ostermeier study and leaves exactly the same questions about selection bias.  Even if one assumes that the fact checks are free from error and bias, the breakdown of the findings is entirely consistent with the hypothesis that liberal media bias accounts for the result.

Mooney's entire argument in support of liberal truthiness appears to stem from an appeal to the principle of parsimony (Occam's razor):
As Kessler’s words suggest, interpreting these data—or, for that matter, the aforementioned data on PolitiFact—as evidence of a liberal bias among the fact-checkers discussed here would be shortsighted and simplistic. All indications are that both outlets try (or even bend over backward) to be as balanced as possible.

A potentially simpler explanation for these results, then, is that the fact-checkers are simply doing their job—and Republicans today just happen to be more egregiously wrong. Democrats, meanwhile, are certainly not innocent when it comes to making misleading statements, but their pants are not on fire.
Apparently since Kessler professes unconcern about which party he fact checks Mooney believes this indicates something akin to a lack of selection bias.  In truth, it indicates nothing of the kind and we are left to wonder at how Kessler's words are supposed to connect to the conclusion the liberal bias hypothesis is "shortsighted and simplistic."

Mooney says the fact checkers try to be as balanced as possible.  Because they say so?  That's silly.  There is nothing in either Kessler's or PolitiFact's selection process, so far as we know, that would keep the data from manifesting a selection bias problem even if we assume that each individual fact is checked in a neutral manner.

In science, one does not accept a researcher's word that his data is free of selection bias.  One expects a check on selection bias built into the method of selection, often accomplished via a simulation of randomness.

Occam's razor?

Mooney suggests that it is "potentially simpler"  to explain the data by assuming that the fact checkers are neutral while Republicans fib more than Democrats.  Mooney apparently thinks it simpler to suppose that a large group of Republicans lies more compared to Democrats than to suppose bias in a small group of fact checkers.  On the surface, that is the opposite of the way the principle of parsimony is supposed to work.

I challenge Mooney to describe his understanding of Occam's razor as it applies to this issue.

Until he deals with that problem as well as others, he's properly classed as a truth hustler.

PolitiFlub: A static number of federal jobs programs?

What can you say about a fact-checking organization that's often no more accurate than the statements it critiques?

Case in point, PolitiFact did a check on Mitt Romney's claim that the federal government runs 49 jobs programs and oversees those programs using eight different agencies.

PolitiFact used as its key source a Government Accountability Office report that focused on fiscal year 2009.

The current year is 2012, and it is the year in which Romney made his statement.

PolitiFact made a very weak effort to update the GAO numbers with reports of new federal job training programs or changes in the number of agencies overseeing said programs.  Indeed, PolitiFact did not make any mention of the Wall Street Journal column that probably served as Romney's source:
There are no fewer than 49 federal job training programs administered by nine agencies that cost taxpayers some $14.5 billion in 2010. A General Accountability Office performance audit in 2011 looked at fiscal year 2009 and determined that "only 5 of the 47 programs have had impact studies that assess whether the program is responsible for improved employment outcomes."
If the Journal report is correct then Romney is off by one as to the number of agencies overseeing job training programs and did not overestimate the number of federal jobs programs.  PolitiFact, if the Journal report is correct, was accurate regarding the number of agencies but off by at least two regarding the number of programs, with the latter an understatement of the actual number.

It looks very much like PolitiFact simply whiffed on the attempt to check for new information updating the GAO numbers.  Without that information PolitiFact has no business trying to rate the truth of Romney's claim.

Wednesday, May 16, 2012

Grading PolitiFact (Florida): Is U.S. Chamber of Commerce's Bill Nelson ad accurate?

Context matters -- We examine the claim in the full context, the comments made before and after it, the question that prompted it, and the point the person was trying to make.
--Principles of PolitiFact and the Truth-O-Meter
Hogwash.


The issue:

(clipped from PolitiFact.com)

The fact checkers:

Angie Drobnic Holan:  writer, researcher
Louis Jacobson:  researcher
Katie Sanders:  researcher
Aaron Sharockman:  editor


Analysis:

This story by PolitiFact Florida manifests that special type of journalistic incompetence that could easily double as deliberate campaign activity on behalf of the Democratic Party's senatorial candidate, Bill Nelson.

PolitiFact does provide the context of the ad, which gives the lie, directly or indirectly, to pretty much the entire PolitiFact analysis:
U.S. Chamber: "Obamacare will be a nightmare for Florida seniors. Did Bill Nelson consider the consequences when he cast a deciding vote for Obamacare?"

Nelson: "This legislation is gonna let folks that are happy with their insurance keep it …"

U.S. Chamber: "Wrong. 20 million people could lose their current coverage."

Nelson: "... including our senior citizens who are on Medicare."

U.S. Chamber: "But Senator, seniors will see $500 billion in Medicare cuts to fund Obamacare. Call Bill Nelson. Tell him to support the repeal of Obamacare."
It is pretty well established that it was not true of the ACA that folks who are happy with their insurance will necessarily have the opportunity to keep their existing insuranceEven PolitiFact didn't swallow that line from President Obama.  Criticizing that same line from Sen. Nelson constitutes the immediate context of the ad.

PolitiFact focuses on a would-be broader context where the ad supposedly implies that 20 million Medicare beneficiaries will* lose their current insurance:
Here, we’re checking whether "20 million people could lose their current coverage," and whether those people are older Americans on Medicare as the ad strongly suggests.
Don't hold your breath waiting for PolitiFact to substantiate its claim that the ad "strongly suggests" that 20 million Medicare beneficiaries will lose their current coverage.  It never happens.  Instead, we get a series of statements that essentially repeat the charge without any supporting evidence:  "The ad makes it sound like 20 million seniors will be losing coverage," "The ad blurs these distinctions in a way that’s highly misleading," "Yet it seems like that’s the reasonable interpretation of the ad" and "The ad from the U.S. Chamber makes it sound like 20 million older Americans will lose coverage under the health care law."

Apparently since the ad opens with the claim that "Obamacare will be a nightmare for Florida seniors," PolitiFact assumes that some portion of the ad must support that statement.  But that simply isn't the case.  The ad follows up by showing Nelson making false statements about the health plan he supported and the ad questions whether Nelson considered the consequences of his vote.  That's fair game in politics.

Is the jump from the claim that ObamaCare is disastrous for Florida seniors to Nelson's ill-stated advocacy a proper justification for a "Pants on Fire" rating?  PolitiFact wants us to think so, and uses techniques that make the U.S. Chamber of Commerce ad look completely aboveboard by comparison.

Shenanigan A:
We read (CBO) the report, which analyzes the health care law of 2010, also known as the Affordable Care Act. We quickly realized the ad was mixing apples and oranges.
On the contrary, rather than mixing apples and oranges the ad uses the higher boundary of an estimate of those who will lose existing employer-provided coverage to represent the number of citizens overall who will lose their existing coverage.  Logically, the higher boundary on the loss of existing coverage for all Americans is higher than that for just those in the employer-provided market.  The claim that the ad mixes apples and oranges stands if PolitiFact demonstrates that the ad is saying that 20 million Medicare beneficiaries will lose their existing coverage.  PolitiFact never undertakes that demonstration.

Shenanigan B:
Even if the ad had provided proper context of employer-provided insurance, it’s still not exactly the case that 20 million people will lose coverage under the health insurance law.
The ad said "could lose," not "will lose."

Shenanigan C:
Second, some portion of that (20 million) number are people voluntarily switching to other, better coverage -- not being forced out of coverage against their will.
Ah, the old "conjecture as evidence" ploy.  "Are" suggests a fact in evidence.  But the consequences of the law foretold in the CBO report are not yet in evidence.  As chronicled in an earlier "Grading PolitiFact" entry, PolitiFact invented its evidence on this point.  Is it possible that a person will voluntarily leave employer-provided coverage for coverage under an exchange?  Sure, barely.  But subsidized exchange coverage under the health care reform act is not available to those forsaking employer-offered coverage.  Those doing so pay for their insurance out of their own pockets, and in addressing that point the CBO notes that consumers in the insurance market strongly tend to choose the less expensive option.  PolitiFact has an excellent record of ignoring that detail from the report.

Shenanigan D (bold emphasis added):
We contacted the Center for Medicare Advocacy, a nonprofit group that works with Medicare beneficiaries to educate them on the program and help them get access to care. We asked executive director Judith Stein if there was any scenario in which seniors would lose coverage under the new health care law.

"If they’re stating that 20 million people would lose Medicare because of the Affordable Care Act, that is simply not true. In fact, the health care law strengthens Medicare coverage," Stein said.
Note the leading nature of the question posed to Stein.  The question uses as its premise the supposition that the ad says seniors will lose insurance coverage, not that seniors will lose their current plan.  And for many seniors, the current plan is a Medicare Advantage plan.

From the Medicare's chief actuary, Richard Foster:
The new provisions will generally reduce MA rebates to plans and thereby result in less generous benefit packages. We estimate that in 2017, when the MA provisions will be fully phased in, enrollment in MA plans will be lower by about 50 percent (from its projected level of 14.8 million under the prior law to 7.4 million under the new law).
As with the decreased enrollment in employer-provided insurance, the abandonment of existing Medicare Advantage insurance plans will occur primarily because of price rationing.

PolitiFact would leave its readers none the wiser.

Now to evaluate the ad according to an alternative reasonable interpretation.

The ad doesn't support its claim that the health care reform bill will prove disastrous for Florida seniors.  Instead, it makes the case that Bill Nelson's advocacy for the reform bill was misplaced.  That is the meaning of the ad's question "Did Bill Nelson consider the consequences when he cast a deciding vote for Obamacare?"

To support the idea that Nelson's perception of the consequences was awry, the ad shows him making the debunked claim that those who like their current insurance plan will be able to keep that plan.  Nelson expands on that claim by specifically mentioning Medicare.  The response in the ad points to a Medicare "cut" that reallocates the funding toward ObamaCare.  A portion of that reduction to projected Medicare spending comes from the reduced Medicare Advantage rebates mentioned above by Medicare's chief actuary.

PolitiFact crashes into this problem again and again.  Fair interpretation requires charitable interpretation of every statement.  It is flatly improper for fact checkers to arbitrarily prefer one "reasonable" interpretation over another.  The claim that 20 million may lose existing insurance--the claim PolitiFact chose to check--is flatly true by the CBO report because the total number of Americans who may lose existing coverage cannot be less than the number of Americans who may lose employer-based coverage.  PolitiFact's suggestion that the ad implies that the full 20 million come from Medicare carries the appearance of mendacity.


The grades:

Angie Drobnic Holan:  F
Aaron Sharockman:  F

I'm not offering grades for those who merely contributed research.

The two listed above are journalists reporting badly.


*Update/Clarification 5/18/2012:   I wrote early in the post that PolitiFact says "the ad supposedly implies that 20 million Medicare beneficiaries will lose their current insurance."  That statement is accurate in the context of the the entire fact check but is not fully supported by the quotation immediately following.  I considered changing it but decided on an explanatory note instead.

PolitiFlub: The Romney Flip-Flop on man-made global warming?

PolitiFact's rating of a "Half Flip" for Mitt Romney's position on man-made global warming gives me an excuse to dust off the "Flub-O-Meter" and attach the ol' electrodes to PolitiFact's reasoning behind the rating.

The rationale behind the "Half Flip" rating comes from a single statement Romney made on the issue.

PolitiFact (bold emphasis added):
But at a campaign event in Pittsburgh, Pa., a few months later, Romney offered a somewhat different perspective. His comments were videotaped and promoted by the liberal advocacy website Think Progress.

A voter asked Romney, "What is your position on man-made global warming and would you reject legislation, such as cap and trade, which is based on the idea of man-made global warming?"

"My view is that we don’t know what’s causing climate change on this planet. And the idea of spending trillions and trillions of dollars to try to reduce CO2 emissions is not the right course for us. My view with regards to energy policy is pretty straightforward. I want us to become energy secure and independent of the oil cartels. And that means let’s aggressively develop our oil, our gas, our coal, our nuclear power."
PolitiFact detects a contrast between that statement and his other statements, as Romney typically acknowledges a role for humans in causing global warming while stating he doesn't think we know the degree to which humans contribute to the phenomenon.

In the summary paragraph, PolitiFact uses that difference to justify the "Half Flip" rating:
It’s unclear to us whether this was an inadvertent omission or a calculated attempt to say divergent things to different audiences. Because Romney, in our view, is savvy enough to know the difference between suggesting a human role in climate change and leaving it out, we think it’s reasonable to perceive Romney as taking two distinct stances in these two statements. We rate this a Half Flip.
PolitiFact's rating ignores the fact that it is normal to communicate using varying degrees of precision.  There are a good number of reasons for not drawing too much into the one statement from Romney, among them the fact that the question was specifically about man-made global warming.  If Romney didn't think mankind was at all responsible for climate change then he could have answered that he does not believe in man-made climate change.  The most reasonable interpretation takes Romney as referring to the principal cause(s) of climate change when he says the cause is unknown.

It's fairly easy to illustrate this variation in precision using statements by PolitiFact editor Bill Adair regarding PolitiFact's process for choosing which statements to fact check.

Adair last year:
"We are a news organization and we choose which facts to check based on news judgment. We check claims that we believe readers are curious about, claims that would prompt them to wonder, 'Is that true?'"
Adair this year:
"We try to check roughly the same number of claims by Democrats as we do for Republicans."
In both cases Adair was addressing the issue of selection bias.  Like Romney's statements on man-made climate change, Adair's descriptions differ from one another.  PolitiFact would probably conclude that Adair is savvy enough about PolitiFact's operations to know that even back in 2011 that PolitiFact tries to check roughly ht same number of claims from either party.

But is that the correct conclusion?

It's more reasonable to take Adair as simply offering differing accounts of the same process.  It's called "charitable interpretation," and PolitiFact struggles particularly in applying it to the claims of conservatives.

Regarding Romney, his statement was probably neither an inadvertent omission nor a calculated attempt to play to two different audiences.  Romney simply used normal human language, complete with normal ambiguity, to communicate his position on climate change.

Heightening the aggravation of these types of PolitiFact findings, PolitiFact exempts itself from the "Burden of Proof" standard it applies to those it rates.  A politician has the burden of proof for supporting his claims, PolitiFact says.  But PolitiFact can claim a "Half Flip" just because it's supposedly "reasonable" to reach its conclusion.

It's hypocrisy lumped on top of a poor evaluation process.

And it's a "Full Flub" on the Flub-O-Meter.

Three years in with "the Obama Effect," part two

In the first part of "Three years in with 'the Obama Effect" we noted that in Europe the "Obama Effect" was substantial and lasting, though less with Eastern European nations in the small sample than for the nations in Western Europe.

The second chart grouping concerns primarily the Middle East.  The decreases in percentage outnumber the positive ones.  No nation in this group increased over 5 percent.

(click image for larger view)
The chart design may obscure the degree to which perceptions of the United States increased when Obama took office.  Where the bottom bar for each country shows a larger decrease than the top bar it means that only after Obama took office did the perception of the U.S. drop below what it was when President Bush was in office.

With that in mind, the chart shows notable decreases in U.S. popularity in Egypt, Jordan and Lebanon during Obama's presidency.

In the Palestinian Territory, U.S. popularity bumped up only slightly in 2009 and increased a few percentage points since.

In Israel, U.S. popularity dropped a few percentage points when Obama took office and subsequently increased very slightly.

For the four nations represented on the chart only the Palestinian Territory shows any enduring "Obama Effect."

Part 1
Part 2
Part 3



Tuesday, May 15, 2012

Great Ideas Department: Fact checker uses primarily liberal audience to crowd-source the keeping of political promises

What could go wrong?
Our Obameter has been cited frequently in blogs, news stories, television shows and books as the authoritative source about Obama's promises.

We'll be updating the promises in the next few months, particularly the ones rated In the Works or Stalled, and we'd like your help. If you have information about developments on a particular promise, please send it to us. We'd love to get links to news stories, press releases, appropriations documents -- anything that shows the outcome.
This is the type of helping that Media Matters was born to do.

To be sure, journalism finds itself in a tough spot these days.  It's hard not to sympathize with PolitiFact for pawning off some of its research on the crowd.  Can PolitiFact help it if its audience is predominantly liberal?

Though the answer to that question is probably "yes," we can't assume PolitiFact sees it that way.

The crowd-sourcing strategy is understandable, but it will serve as just one more among many factors that skew PolitiFact toward the left.

Elizabeth Warren still pround of Native American heritage?

What "Native American heritage"?
“I am very proud of my Native American heritage, thank you,” said Warren when asked if she disapproved of the school counting her as a minority woman on the faculty. “These are my family stories ... This is our lives and I am very proud of that.”
The Democratic Party's candidate for the Massachusetts senate seat occupied by Scott Brown has failed to provide better proof of her supposed heritage than high cheekbones and family folklore.  Ron Paul might have passed himself off as the offspring of a visiting space alien based on that type of evidence.
The estimable William Jacobson has kept abreast of the story, including the way apparently false media reports (as in the Boston Globe) have helped Warren perpetuate the impression the she has a Native American heritage:
The Globe and the false report of a 1/32 Cherokee connection may have saved Warren’s campaign, as it came at a time when her campaign was in panic and without any evidence to substantiate her claim to Native American ancestry, which she used when a junior faculty member in a law school association directory to obtain “minority law teacher” status.

The false report bought Warren time during which various supportive pundits could opine about what it means to be Cherokee and how dare white people impose their own standards.

This mea culpa should be front page at The Globe.
Maybe The Globe is concerned about hurting Warren's chances in the election.


Monday, May 14, 2012

Illegally registered voters + mainstream journalism = buried lede


Most of the Reuters story deals with a potential 180,000 voters on the rolls who are not legal residents.  The last line of the story is perhaps the most interesting:
Cate said some Florida officials have asked the Obama administration to grant the state access to databases maintained by the U.S. Department of Homeland Security to help determine who is a citizen.

"We've been requesting access, but have so far been denied," he said.
I can imagine legitimate reasons that Homeland Security would not share this information.

If the actual reason is that correlating voter registration rolls with databases showing legal residence is "racist" then I don't count that as a legitimate reason.

Props to the press for even running the story, but we need more information on that caboose of a lede.

(for Glenn Kessler and PolitiFact) How to fact check the job recovery numbers

A valuable media watchdog watchdog post at the new blog "counterirritant" pointed out a problem with Glenn Kessler's fact check of a claim from Republican presidential candidate Mitt Romney.

Kessler writes the fact checker feature for the Washington Post.  Romney claimed that in a normal recovery from recession the country should be adding something like 500,000 jobs per month.

counterirritant:
Kessler decided that the best way to “check” this was determine how frequently 500,000 jobs were created in a month in the last 65 year.
The post goes on to very effectively criticize Kessler's methodology throughout.

By a funny coincidence (general leftward lean of the mainstream media, maybe?), PolitiFact used very similar reasoning on the same claim:
Is 500,000 jobs created per month normal for a recovery?

The short answer is "no."

We arrived at this conclusion by looking at the net monthly change in jobs all the way back to 1970. Since Romney was referring to total jobs, rather than private-sector jobs only, we used total jobs as our measurement. And since Romney was talking about job creation patterns during a recovery, we looked only at job creation figures for non-recessionary periods, as defined by the National Bureau of Economic Research. Finally, we excluded the current recovery.
The Kessler/PolitiFact method is entirely wrongheaded.

Romney didn't claim that 500,000 jobs created per month was a normal figure during a recovery.  I can imagine the furrowed brows of Kessler, Louis Jacobson and other mainstream media fact checkers.  Aren't they the experts?  What am I talking about?

It's actually pretty simple.

The size of the economy changes.  If country A enters a recession losing 1 million jobs  then it takes two months to regain the lost jobs at a rate of 500,000 per month.  If country B experiences a recession losing 10 million jobs then it takes 20 months to regain them at a rate of 500,000 per month.

Not only does the size of the economy vary, but so does the depth of the recession.  The rate of recovery for lost jobs needs  to account for both factors.  Neither Kessler nor PolitiFact gave any apparent consideration to those critical criteria.  It's like comparing prices between now and the 1950s without adjusting for inflation.

The Romney campaign has made a number of statements like the 500,000 jobs claim, and they probably relate to the following chart or one like it from the Bureau of Labor Statistics:

Clipped from nytimes.com

Romney's claim almost certainly derives from the fact that post-war recoveries usually replace lost jobs much faster than the present recovery. 

So how does one check that claim?  It's not that hard.  Take the bottom point of employment, then count the number of months that it takes to get employment back up to the peak level.  Divide the number of jobs lost by the number of months it took to return to return to the employment peak at the start of the recession.  Do the same for each of the post-war recessions, then average the numbers to obtain the average job recovery time.  After all of that, divide the number of jobs lost from the 2007 recession by the averaged job recovery time.

Why didn't the Washington Post or PolitiFact do anything remotely resembling the fact check I just described?  It could be gross incompetence.  It could be ideological bias.  Or it could be both.





Correction 6/7/2012  Removed a redundant division operation from the next-to-last paragraph

Saturday, May 12, 2012

Grading PolitiFact: Marco Rubio and taxing small business

It's the old "true but false" trick.


The issue:

clipped from PolitiFact.com


The fact checkers:

Angie Drobnic Holan:  writer, researcher
Bill Adair:  editor


Analysis:

First things first, the context:




PolitiFact provides a partial transcript with its quotation of Rubio:
To explain his vote, U.S. Sen. Marco Rubio, R-Fla., released a statement that day:

"As someone with a student loan and with a state with so many people with student loans, I support a hundred percent making sure that the interest rates on student loans do not go up," Rubio said.

"I cannot support the way the Democrats want to do it, however, because, they want to do it by raising taxes on small businesses, very small businesses. The kinds of small businesses that give jobs to graduates who not only need low interest rates but need jobs in order to pay their student loans."
Rubio went on:
"We have a plan to keep student interest rates low that doesn't raise taxes.  And I hope the Democrats will give us a vote and a chance to pass that."
Rubio tries to make the point that the Republican plan to keep student loan rates from rising does not raise taxes compared to a Democratic Party plan that does raise taxes, and Rubio mentions the entities affected by the tax hike.

What fact shall we check, PolitiFact?
We decided to check out Rubio's claim that Democrats want to raise taxes on "small businesses, very small businesses … the kinds of small businesses that give jobs to graduates."
It's not often one sees an ellipsis used to take the place of a period.  But at least we have the answer to the question.  Let's see how PolitiFact executes the fact check.

PolitiFact:
To fact-check Rubio's statement, we read the text of the legislation the Democrats were proposing, examined a summary of the bill provided by the nonpartisan Congressional Research Service, and ran the proposal by the experts at the Tax Policy Center, a respected independent think tank that focuses on taxes.
Reading the text of the legislation is a good move.  Going to an expert from the Tax Policy Center isn't a bad move, either, if we overlook the fact that it is associated with the center-left Brookings Institution as well as the fact that the consultation with experts stops there.

It doesn't exactly represent a good faith effort to "interview impartial experts."  But let's move on to PolitiFact's reasoning.

PolitiFact reports that the bill limits the tax increases to S-corporation beneficiaries who a) are married and make over $250,000 adjusted gross income, b) are married filing separately while individually making over $125,000 adjusted gross income or c) are individuals making over $200,000.  Democrats promoting the bill say it is targeted at closing a tax loophole that allows some parties to sidestep payroll taxes.

So let's say all that is true.  Is it therefore false that the Democratic proposal raises taxes on small (and very small) businesses?  If it does not affect small businesses then what types of businesses are affected?

PolitiFact doesn't say, at least not explicitly.  Instead we get an implicit answer:
Rubio said he opposed the Democratic bill on student loan interest because it would raise taxes on "small businesses, very small businesses … the kinds of small businesses that give jobs to graduates."
Actually, the bill changed tax rules only for S-corporations, and only on professionals like lawyers and accountants who could be taking advantage of the tax code to avoid paying payroll taxes ...We rate Rubio’s statement False.
PolitiFact just implicitly informed us that small and very small businesses are not taxed more under the Democratic proposal.  We deduce that therefore if any businesses are affected they are medium, very medium, large and/or very large businesses.

How did we end up with that apparently absurd conclusion?  Easy.  PolitiFact interpreted Rubio to say that the Democratic tax change affected small and very small businesses broadly speaking.  In other words, small businesses in general were the target of the tax change rather than a specific subclass of (small) businesses.

Why use that interpretation instead of taking Rubio to mean that the businesses paying for the tax would fall into the smaller business demographic?

That's the $64,000 question.  PolitiFact offers no reason justifying its interpretation other than to claim that people might well take it that way:
Rubio's statement gives the impression that all kinds of mom-and-pop operations might be subject to new, additional taxes, when actually the bill is aimed squarely at high-income professionals who are taking advantage of a loophole.
Is that good enough?

If it's good enough then why do we not equally pay attention to the risk that voters watching Obama's "Julia" cartoon will think a President Romney threatens their right to fight for equal pay?

Obama "Mostly True" and Rubio "False"?  Seriously?

Rubio's claim is literally true, assuming  some "very small" businesses would end up paying for the closed tax loophole.  One could present all types of caveats about the extent to which Rubio fails to present the full picture, but similar complaints could most certainly apply to the Obama campaign's "Julia" Web ad.

We end up with yet another outstanding illustration of the subjective nature of PolitiFact's grading system.  And once again the conservative receives the harm.


The grades:

Angie Drobnic Holan:  F
Bill Adair:  F


Taking PolitiFact's grading system literally, it should not be possible to grade a literally true statement like Rubio's "False."  Yet somehow you did it, PolitiFact.  Amazing.

PolitiFact has yet to learn how to apply the principle of charitable interpretation with anything remotely resembling fairness, assuming anyone on the staff even knows what it is.

Or maybe they're just in the tank for the Democratic Party.  I can't dismiss that hypothesis when PolitiFact produces stories like this one.


Clarification 5/15/2012: Traded out "In other words, such businesses would feel the effects of the law more commonly than not" in favor of  "In other words, small businesses in general were the target of the tax change rather than a specific subclass of (small) businesses)."  Later added sections b) and c) to the description of parties affected by the tax.  My apologies for the temporary omission.

Three years in with "the Obama Effect," part one

With President Obama's decidedly mixed foreign policy results and an approaching presidential election, I grew curious about the current state of "the Obama effect."  That's the name given to the tendency of foreign nations to view the United States more favorably with the election of President Obama.

I looked at the Pew Research Center data for its surveys on attitudes toward the U.S. of foreign nations.  I was disappointed that Pew Research did not have data posted to allow the comparison of some nations, such as Canada.  In other cases enough data exists for comparisons but not always using the same range of years.  For some nations I deliberately chose a different range to better represent overall trends in U.S. popularity.

I divided the data into three groups to keep the size of the graphic manageable.  The first group primarily represents European nations.


(click image for larger view)

Europe overall makes Obama look pretty good.

Each nation has two bar graphs representing the change in U.S. popularity over time.  The first shows the change from Bush through 2011.  The second reflects the change just since Obama took office.  The effect in France, then, shows an increase in popularity of over 30 percent.  No graph is visible in the space below, which means that the high opinion has held steady.

Western Europe just plain loves Obama, from the look of things.  I don't have any insight into the increase in popularity in Spain.  I don't know what Obama or the U.S. has done to produce it.

Improvements in Russia, I think, result from Obama's weak foreign policy approach. 

Poland surprised me.  Obama improved on Bush over time, which is somewhat perplexing given the Obama administration's soft stance on Russia.  Perhaps Iraq policy accounts for most of the modest change.

Turkey was the only nation from this group where the U.S. was more popular under Bush, but even in this case the numbers show that the drop since Bush was president is small--smaller, in fact, than the drop just since Obama took office.  That means that U.S. popularity initially went up slightly under Obama before sliding down lower than it was in 2008.

For context, it's true that in many cases U.S. popularity in 2008 is somewhat higher than in previous years.  While some of this may stem from Bush's actions in office, such as negotiating the path toward the end of the Iraq War, I wouldn't discount the possibility that the U.S. election had something to do with it.  Depending on when the data were collected, Obama might have something to do with the rise in popularity even in 2008.  The world does watch U.S. elections to some extent.

Parts two and three will show quite a few more negatives.

Part 2
Part 3


Addendum:  The Associated Press was thinking along the same lines and published a story today on international perceptions of President Obama.  Some of the information from that story helps illuminate information on the charts.  The AP story actually paints a less complimentary picture of Obama's image in Europe.

Thursday, May 10, 2012

Good on Brendan Nyhan

I've been critical enough of the research of political scientist Brendan Nyhan that I owe some recognition to a pretty good column he wrote up at Columbia Journalism Review.

The title has a good hook, and it's worth a read:

Obama ‘evolves,’ Romney ‘flip-flops’



Wednesday, May 09, 2012

Grading PolitiFact: Approving a "Julia" misdirection

Context matters -- We examine the claim in the full context, the comments made before and after it, the question that prompted it, and the point the person was trying to make.
--Principles of PolitiFact and the Truth-O-Meter
Sometimes you just have to tip your hat to PolitiFact for its ability to tease a grain of truth out of a highly misleading presentation.

On with the evaluation.


The issue:
(clipped from PolitiFact.com)


The fact checkers:

Becky Bowers: writer, researcher
Bill Adair:  editor


Analysis:

The Obama campaign released a Web-based audiovisual presentation designed to play up the contrast between President Obama and Republican challenger Mitt Romney.


Note that Romney threatens Julia's idyllic life story.  That's the context.

PolitiFact:
The cartoon "The Life of Julia," compares the candidates' impact on Julia at a dozen points in her life.  At age 23, she’s shown reading a newspaper with the headline, "Equal pay 4 equal work." She’s starting her career as a Web designer.

"Because of steps like the Lilly Ledbetter Fair Pay Act, Julia is one of millions of women across the country who knows she’ll always be able to stand up for her right to equal pay," the graphic says.
The "Lilly Ledbetter Fair Pay Act" is very probably widely misunderstood.  It would more accurately go by the name of "The Lilly Ledbetter Exemption From Statute of Limitations Concerns For Fair Compensation Suits."  Admittedly the latter is quite a mouthful.  In a nutshell, the Act doesn't provide for equal pay.  It allows people to sue for existing equal pay rights years after the alleged discrimination took place, shifting a very difficult burden of proof onto an employer.  And in that context, the phrasing from the Obama Campaign is practically Clintonian in its weaselly use of words (take a bow, campaign staff):  Julia "Knows she'll always be able to stand up for her right to equal pay" because the statute of limitations doesn't apply.  Of course the majority of people seeing the ad probably have no idea that's what it means.  And sometimes PolitiFact is concerned about campaign materials that create a distorted view of reality.  But not this day.

PolitiFact notes the intended contrast:
Romney, on the other hand, "has refused to say whether he would have vetoed or signed the Lilly Ledbetter Fair Pay Act," the graphic says.
Horrors.  Seriously, it's a fear-based message.  Women are supposed to fear that Romney would threaten equal work for equal pay.  PolitiFact often brings its most heated criticism to fear-based messaging.  But not this day.

Toward the good, PolitiFact properly starts tracking the key message:
We wondered: Did Romney refuse to say whether he would have vetoed or signed the act? And what would that mean for Julia?
What would it mean for Julia if Romney refused to say whether he would have vetoed or signed the act?  That's the key to the ad.  The ad tries to create worry about rights to equal pay regardless of gender if Romney wins the presidency.  The implication is that Romney represents some type of threat.

For the sake of expediency, let's assume that PolitiFact suitably confirms that Romney "refused" to say whether he would have vetoed or signed the act.  We'll follow the key point of the ad instead of the factoid that supposedly supports the ad's implication.

PolitiFact:
Romney did refuse to say whether he would have signed the bill. But he also said he supported "equal pay for women" and had "no intention of changing that law."
PolitiFact infers, rightly I think, that Romney's reference to "that law" refers specifically to the Lilly Ledbetter Fair Pay Act.  He wouldn't change it, he says.  So what is the consequence for Julia if Romney refuses to say whether he would have signed it?

PolitiFact concludes:
The Obama campaign said Romney "has refused to say whether he would have vetoed or signed the Lilly Ledbetter Fair Pay Act." That’s essentially what happened in Romney’s ABC News interview in April, but the claim leaves out some details that matter to Julia.

The relevant point for Julia isn't whether Romney would sign the bill — it's already the law — but his decision not to change it. Romney also said he  "certainly support(s) equal pay for women," and has "no intention of changing that law."

Still, the Obama campaign is correct that he dodged the question, so we rate the claim Mostly True.
Amazing, no?  The main point of the ad was without merit, but since it is trivially true that Romney refused to say whether he would have signed the bill, PolitiFact rules it "Mostly True."

PolitiFact doesn't always use this standard.  Just ask Mitt Romney.

Romney recently released an ad pointing out that unemployment for American youth is double that for the rest of the population.  PolitiFact found it true but irrelevant since the unemployment rate is nearly always approximately double for that demographic.  Apparently it's not supposed to matter even if the unemployment rate for the general population is high.  Romney ended with a "Half True" rating.

Is the Romney ad more misleading than the Obama presentation?  Seriously?


The grades:

Becky Bowers:  F
Bill Adair:  F

PolitiFact discounted the context of the ad where it counted.

I'd love to see a journalist ask Bill Adair to coherently explain why the Romney rating is lower than the Obama rating.  Though I suppose I can ask him myself, since I'm 99 percent certain he drops by for the occasional visit.

How about it, Mr. Adair?  You have my email address.  Drop me a line and explain why Obama receives the "Mostly True" while Romney gets the "Half True."  What's the objective dividing line between the two cases?  I'll publish it in full if granted permission.