PolitiFact Detractors: Virginia Whine Drunk Deeply

Recently I did a post on what I believed to be right-wing bias by PolitiFact (PF) Wisconsin by looking at the timing—PolitiFact Wisconsin published very negative rulings on Democrats, and positive rulings (or nothing) on Republicans, in the period just before the June 5 recall election.

Well, PolitiFact (PF) Virginia’s detractors, in this case, the Republican Party of Virginia, published a massive tome in what they called an “Open letter to the Commonwealth of Virginia.” It can only be described as an 86-page collection of “sound bites/talking points” claiming something similar against PolitiFact Virginia as published by the Richmond Times Dispatch. I should say, it claimed A LOT of things.

The timing they claim is that PolitiFact Virginia buried Republican positive fact-checks in the weekend edition of the Richmond Times-Dispatch when fewer people viewed the news and saved negative Republican fact-checks for during the week when more people would see them, and vice versa for Democrats.

Let me address each of the claims from the Virginia Republican Party before responding to that timing claim, the third item on this list.

1. PolitiFact Virginia “disproportionately” targets more Republicans. This was also addressed by the response to the Republican Party by PolitiFact Virginia writer Warren Fiske, but let’s review it with more specifics: 

According to Wikipedia page for Virginia “Of the state's 11 seats in the U.S. House of Representatives, Republicans hold eight and Democrats hold three.” There are two Democratic U.S. Senators, but a Republican Governor (Bob McDonnell) and a Republican Lieutenant Governor (and Attorney General). And of course, Eric Cantor (R) is the House Majority Leader. The lower house of the Virginia General Assembly is 67% Republican. The upper house of the Virginia General Assembly is split 50/50 for forty Senators.

Counting all seats held, there are currently 98 positions held by Republicans and 57 positions held by Democrats in the state of Virginia. That means the Republicans hold 70 percent more positions.

Based on that percentage, you’d think PF Virginia would likely do more fact-checks on Republicans in a similar proportion. They’ve actually done less, about 50 percent more instead of 70 percent. So it can be argued that maybe PF Virginia doesn’t do enough on Republicans.

2. PolitiFact Virginia gives more negative rulings to Republicans.  In 2010, I posted a finding that PF Virginia actually favored the Republicans in what was then my version of the Truth Index, and tied that into the fact that the Richmond Times-Dispatch endorsed McCain in 2008 and Bush in 2004 (and that Eric Cantor’s wife was on the board of the company that owns it).

But let’s get more current on the actual Truth Index. The report by the Republican Party of Virginia says there’s been 208 rulings through July 10, 2012. I think they also include the Flip-o-Meter. My stats on PF Virginia are only through the end of June and don’t include the Flip-o-Meter, so I have 199 rulings.  It looks like (from their chart below) there's at least seven total Half- and Full-Flips. That means we have a difference of one or two rulings. Close enough.

Click to enlarge:  Virginia Republicans want to convey that PolitiFact publishes more "False" type rulings.

The report by the Republican Party of Virginia published a chart (see above) of the rulings based on the raw number, an absolute measure which obscures the real picture. The chart appears to be slightly off, to perhaps make it look worse for Republicans, at least compared to my raw numbers. This is particularly obvious in the “False” category, where they show what appears to be 13 Falses for Democrats and 23 Falses for Republicans. The total number of Falses for Republicans is accurate, but there should be 17 Falses for Democrats. Apparently they want to make it look like they’re the victim of “False" rulings by PolitiFact.

But it’s clear this approach was used to show it more in a negative light to favor the Republicans. A more equitable approach would be to show it more relatively by percentages, displayed similarly to the PolitiFact app, shown here. It shows a different picture:

Click to enlarge:  Virginia Republicans and Democrats are in a dead heat with regard to the PF Truth Index.

So on a relative basis, what we have is a higher percentage of Republican statements rated True, and a lower percentage rated False, than the Democrats. The Virginia Republicans probably don’t want you to know that. In the Pants on Fire category they are exactly even. In terms of the Truth Index, there is ONE percentage point difference, which means the Republicans and Democrats are in a statistical tie.

3. “The Weekend Dump”. Here’s how the Virginia Republican Party describes it:

PolitiFact Is More Likely To Dump Positive Republican Rulings Late On A Friday, Or Over The Course Of The Weekend, As Well As Save Negative Republican Rulings For Publication During The Week (Monday Through Thursday.

I started recording the date the ruling was published by PolitiFact in 2012, so I can only go by data I have for the first six months of 2012. In addition, the date that the Richmond Times-Dispatch published the ruling could be different from the date PolitiFact published it. For example, so far in 2012 it appears PolitiFact never publishes anything on Saturday, although the Republican Party’s study (Page 11) says that four (4) False rulings on Democrats were “buried” in a Saturday edition of the Richmond Times-Dispatch. (These could be rulings from PolitiFact National, which is a different matter.)

So, using the data for the first six months of 2012, and assuming that the Richmond Times-Dispatch published the ruling the same day as PolitiFact, I found that there is a bit of truth to their claim, although they leave a few things out. Two thirds of the PolitiFact Virginia rulings were published Monday through Thursday. The Republicans have a lower Truth Index than Democrats during this time period, a negative 22 versus a negative 4 for the Democrats. Friday through Sunday, however, both the Democrats and Republicans have a positive Truth Index, 50 for the Democrats and 30 for the Republicans.

The proportion of Republican to Democrat rulings was similiar for both periods: Republican rulings outnumbered those done on Democrats by a ratio of 3 to 2 during the week compared to 2 to 1 over the weekend (Monday through Thursday, 12 Democrat  and 18 Republican; Friday through Sunday, 5 Democrat  and 10 Republican).

In other words, more negative rulings seem to appear for both parties Monday through Thursday, while the weekend rulings appear to be more positive, for both parties. It should be noted that so far in 2012, the previous Republican advantage on the Truth Index at PF Virginia has reversed slightly (Republicans are at a negative 4 and Democrats at a positive 12), so these figures could be different if we took them back to 2011 and prior, and they might very likely favor the Republicans.

I won’t get into their last objection, where they parse individual rulings with Republican talking points, as it appears to be the same contention I’ve been hearing from other mostly conservative PolitiFact detractors: that many PF rulings are just opinion masquerading as fact. That would take me a while, because the Republicans have a lot more resources than me, as evidenced by this “screed.” But their take on the numbers shows that figures can lie and liars can figure.

Postscript (7/17/2012):  It appears that PolitiFact National is more implicated in a "weekend dump" with a bias than PolitiFact Virginia.  But that's only from the view of publishing on the internet; I can't speak for the Richmond Times-Dispatch or the Tampa Bay Times in terms of what or when they publish.