On Sept. 25, 20198 (corrected Dec. 17, 2023) we submitted a complaint to the International Fact-Checking Network regarding PolitiFact’s failure to follow an open and honest corrections policy. A .pdf image of that complaint follows.
Submitted Sept. 25, 2018
On Sept. 12, 2018, PolitiFact published a “PunditFact” fact check of MSNBC host Joe Scarborough. PolitiFact committed a huge mistake in its analysis of Scarborough’s claim and has shown no inclination to fix it after it was pointed out promptly and repeatedly.
Scarborough said the GOP-led Congress is responsible for adding more debt in one year than was added over the first 200 years of American history. PolitiFact, using debt from 1776 through 1976 as its baseline for comparison, found the claim true in terms of nominal dollars.
So far, no real problem.
However, PolitiFact then noted that comparisons done in terms of percentage of GDP serve as a better measure than nominal dollars. PolitiFact confirmed(!) that Scarborough was correct because the debt as a percentage of GDP in 2017 was greater than the debt as a percentage of GDP in 1976.
As PolitiFact/PunditFact put it (bold emphasis added):
As the chart shows, the current level of debt as a percentage of GDP is not unprecedented (it was higher around World War II) and its growth has eased recently since a spike after the Great Recession. The recession both fed higher spending on the national safety net and weakened the denominator, which is gross domestic product.
The chart does show that, when looking at a percentage of GDP, Scarborough is correct in his comparison. Debt as a percentage of GDP in 2017 was far higher (almost 77 percent) than it was in 1976 (about 27 percent).
But the debt as a percentage of GDP in 2017 does serve as any kind of representation of Scarborough’s claim. Scarborough specified added debt, not total (cumulative) debt. The number on the chart PolitiFact used to verify Scarborough’s claim included all debt, including the debt amount from 1976, not merely added debt. To figure the added debt as a percentage of GDP for 2017 involves dividing the 2017 deficit by the 2017 GDP. That number comes out as a fraction of the number PolitiFact used to confirm Scarborough’s comparison.
This is the truth: The chart does not show that, when looking at a percentage of GDP, Scarborough is correct in his comparison. PolitiFact’s claim is flatly false.
The chart contains too little information, by itself, to have any bearing on Scarborough’s comparison. The chart offers no valuable insight into the measurement of added debt in 2017 in terms of percentage of GDP.
We used Twitter on Sept. 12, the day the fact check published, to flag the problem for author Louis Jacobson and PolitiFact.
Later on Sept. 12 we also formally reported the error to PolitiFact using the means it recommends (email to truthometer@politifact.com) :
Dear Truth-O-Meter,
A PunditFact fact check published on Sept. 12, 2018 requires correction.
The fact check examines the claim of one Joe Scarborough that the added debt in 2017 exceeds the cumulative debt over America's first 200 years.
The fact check correctly notes that testing the claim by expressing the comparison as a percentage of GDP serves as a better measure than nominal dollars. But it goes on to describe the wrong comparison:
The chart does show that, when looking at a percentage of GDP, Scarborough is correct in his comparison. Debt as a percentage of GDP in 2017 was far higher (almost 77 percent) than it was in 1976 (about 27 percent).
While it's true that the chart shows the debt as a percentage of GDP as 77 percent in 2017 and only 27 percent in 1976, that is not the comparison Scarborough described. To calculate the added debt for 2017 in this case one needs to calculate the 2017 deficit as a percentage of GDP (the deficit is the same as the added debt for the year). OMB numbers show the 2017 deficit as a percentage of GDP at 3.5 percent for 2017.
We received an automated reply assuring us PolitiFact would read our message and respond appropriately. We have received no other response from PolitiFact.
On Sept. 14, 2018 we used Twitter again to flag PolitiFact’s false reporting. We retweeted an earlier tweet pointing to the problem but adding the #PolitiFactThis hashtag. PolitiFact monitors tweets applying that hashtag to look for fact-checking stories.
We filed this complaint with the IFCN to argue that PolitiFact does not follow its policy on open and honest corrections. But we cannot rule out the possibility that PolitiFact reviewed its work and finds nothing at all wrong with it (after all, we have received no response at all from PolitiFact on this matter). For our part, we cannot image how a fact checker would justify using the total debt in 2017 (as a percentage of GDP) to represent added debt in 2017 (as a percentage of GDP).
We find the error obvious, in other words, and find ourselves at a loss to explain why PolitiFact would resist correcting its work.
PolitiFact’s error remains intact as of Sept. 25, 2018.
Thanks for your attention to this matter.
