Biased Sources

I recently wrote an article about child gun accidents and the NRA, arguing that these accidents pose a serious challenge to the organization’s ideology that it is unprepared to meet. A gentleman on Twitter took me to task for a source I cited on child fatalities. I linked to a report produced by Everytown for Gun Safety/Moms Demand Action, which counted child gun deaths to show that they were higher than official numbers had indicated. The Twitter-gentleman chastised me for relying on “biased” and “agenda-driven” sources.

It’s tempting to simply write off his complaint as an ad hominem; unable to deal with the facts, he attacks the source. Ultimately, I do think this is the case. But there’s an important question here about how casual consumers of news should treat reports by political groups. Because many of us don’t have the time to wade into the methodology of every 50-page report, we need shorthand methods for determining which facts to trust. Isn’t a principled policy of discounting anything put out by a political group an effective way of decreasing one’s chances of being hoodwinked? Now, I have some cause to doubt that my noble twitter-interlocutor actually does apply the same principle to statistics produced by those on the other side of the gun control debate. But let’s say he did. His rationale was as follows: “I am an ordinary person trying to get to the truth. But I’m not a statistician and I have no time to parse every source. So I just write off anything that comes from a lobby group, because they have a vested interest in lying to me or fudging. I know that even though they could be right, if they had found evidence suggesting they weren’t right, they wouldn’t tell me. So I only pay attention to relatively neutral sources. (Of course, I know that there’s no such thing as complete neutrality. But there are degrees, and some sources have vested interests that makes the likelihood of them lying extremely plausible.)”

This is actually quite reasonable. And if one adopted it, one would discount numbers put out by Moms Demand Action. After all, they have an interest in inflating the numbers of gun accidents! But the gun accidents report actually provides a useful illustration of why biased sources are extremely important, and unfortunately we can’t use the corner-cutting technique of only taking in information from neutral sources.

The problem here is obvious: an organization’s interest causes it to do the research that others without that interest won’t do. Moms Demand Action cares strongly about gun deaths. They want to institute gun controls. The NRA wants to stop them. Now, if both organizations equally suspect that the CDC numbers under-report child gun deaths, MDA is likely to spend the resources to try to prove this. The same holds true with opposite politics. If the “Send ‘Em Home PAC” (a fictional anti-immigration organization) suspects that crimes by unauthorized aliens are higher than the numbers suggest, and “Mi País, Su País” (a pro-immigration group) suspects the same, I can bet which one is going to produce a report…

Okay, but there’s a problem. Both of these organizations want to drag the numbers in their favored direction. So as we said before, even though they could be correct, they’re still untrustworthy. Right, this is a problem. But it can’t lead us to simply discounting the source, because we’d miss all sorts of useful data that these sources’ biases lead them to dredge up. Glenn Greenwald is not objective on the national security state; one can suspect that if he found something out that made the NSA look reasonable and responsible, he’d be much less likely to write about it than something that made them look fiendish and totalitarian. That means we probably can’t trust Glenn Greenwald for an accurate total impression of the NSA. But if we just write him off, we miss the fact that his anti-NSA views lead him to spend a lot of effort uncovering facts on the NSA, because he has a strong interest in finding those facts. Nobody else is as dedicating to finding a certain class of facts than he is. So if CDC numbers on gun deaths are under-reported, one can expect Moms Demand Action to be the ones with the strongest possible interest in discovering that information. If they ignore all sources with strong interests, they miss important facts that these sources’ interests lead them to want to present.

This does, unfortunately, leave the news consumer in a bind. They can’t take the easy way out, of simply trying to find relatively objective sources of information and discounting anything that comes from, say, The Heritage Foundation or MoveOn. The consumer has hard work to do; they need to engage with actual arguments and facts. If the Heritage Foundation is right about something, then it’s right. The Heritage Foundation would say that a market solution had worked best even when it hadn’t (or simply wouldn’t discuss times when it hadn’t), but that doesn’t tell us anything about whether the Heritage Foundation’s conclusion that a market solution worked in a circumstance is correct. Sadly, finding the truth is hard. Political bias is extremely frustrating; it upsets me constantly that both the NRA and the gun control groups are unwilling to acknowledge the points that hurt their case. Neither cares about truth, so we the consumers have to swim through an ocean of bullshit in order to reach the Isle of Veracity. But until people start basing their politics on reality, rather than their realities on politics, it’s all we can do.