Donald Trump said "crime is rising," and PolitiFact rated that statement "Pants on Fire." Eugene Volokh looks into it.
It seems that PolitiFact based its rating on the fact that crime has been on a long downward trend overall for the last 25 years. Volokh writes,
It seems that PolitiFact based its rating on the fact that crime has been on a long downward trend overall for the last 25 years. Volokh writes,
I don't find this a persuasive defense. If the original PolitiFact post had said something like, "The violent crime rate has plummeted in the past 25 years, and while it may have been increasing in the last year and a quarter, that could easily be an anomaly, and our data on that are just preliminary and may not be sound," I would have thought it a sensible criticism of Trump's assertion. We should indeed be cautious about data that are limited to one year, or (as with the 2016 first-quarter data) to a subset of jurisdictions. There is some degree of short-term variation within any long-term trend; data from a year and change aren't really enough to tell whether 1) the long-term violent crime decline has been reversed, or 2) the year was just an anomaly and the decline will continue, or at worst, the violent crime rate will remain flat. For instance, the violent crime rate increased in 2005 and 2006, but those proved to be just small blips in an otherwise substantial decline.
But that's not what the PolitiFact people said; they said "Pants on Fire." They didn't point to the data that suggests that Trump may have been literally accurate as to recent violent crime, even if one can plausibly argue that the 25-year trend is more important than the current 1.25-year possible upswing. Instead, they just categorically asserted that Trump outright lied on this point (unless there's some nuance in "Pants on Fire" that I'm missing). They included a quote from an expert briefly nodding to there being "some spikes in homicide and shootings in certain cities," but the quote was immediately followed by saying that "other cities continue to experience low rates." And though a paraphrase from another expert briefly noted a "possible upward swing in the past year or so," it said that any such upswing "wouldn't show up in the data currently available." There was nothing to acknowledge that some real, albeit preliminary, recent data does indeed suggest that violent crime is rising, though there may be reasons to discount that data.
This is not, I think, how fact-checkers should operate (and I say this as someone who is not a Trump supporter). And though today's update at least mentions the "preliminary figures for 2015 that show crime rising," it doesn't acknowledge what strikes me as the quite misleading analysis in the original post. In any case, review the PolitiFact story and the data I link to above for yourselves and see what you think.
I think that in this case PolitiFact is more "Politi" than Fact, and that does not seem to be unusual for it.
To most people, the statement "crime is rising" would seem to refer to the trend at this time, not the historical trend over a long time. As I have noted many times on this blog, there is a serious lag between crime trends and official data. The data that PolitiFact relies on for its "Pants on Fire" rating end at 2014. PolitiFact says, "Two criminologists we checked with before publication warned us that [preliminary] data [from 2015] may not be indicative of a real trend." (Emphasis added.) You call a person a liar because the data supporting his position may be transitory? They may not.
There are a lot of nuances to this discussion, but overall Trump is more right and less misleading than the people who are calling him a liar. This time.
Just so. I noticed the same thing on PolitiFact a couple of weeks ago.
http://www.crimeandconsequences.com/crimblog/2016/06/looking-for-only-what-you-want.html