Polls and the line between truth and conspiracy theory

In the poll on conspiracy theories I mentioned a few days ago, I mostly focused on the item about vaccines, mentioning in passing the fact that Democrats (and liberals) bought into far fewer conspiracy theories than Republicans (or conservatives). I didn't point out that, of the "conspiracy theories" Democrats were more likely to accept, several require a rather fine parsing to register as conspiracy theories (rather than simply an over-broad but accurate account of history).

For instance, the PPP poll asked whether "the CIA was instrumental in distributing crack cocaine into America’s inner cities in the 1980s." There's no doubt at this point, as even the CIA acknowledges, that contractors the CIA sponsored to bring guns to the Contras were hauling drugs back. The CIA claims to have been shocked (shocked!) that the smugglers they employing could possibly be smuggling drugs, and denies helping these smugglers evade the DEA. Whether you believe that or not, there's no doubt that the CIA was helping the folks who were moving drugs, especially cocaine, into the country. Does that make the CIA "instrumental"? Probably not, but is that a distinction you make when a pollster's recording is on the phone with you?  (For what it's worth, political Independents were the most likely to agree with that poll item.)

These subtle shades of meaning are even more important on the question about whether "President Bush intentionally misled the public about the possibility of weapons of mass destruction in Iraq to promote the Iraq war." As a simple historical matter, President Bush, his administration, and other advocates for war definitely made false claims about Iraq's WMDs; the main justification for war was the claim that those weapons – and programs to create more – existed, but they didn't. After the invasion, extensive and unhindered investigations found no evidence of active nuclear, biological, or chemical weapons programs, nor of usable stockpiles.

To turn this factual assessment into a conspiracy theory liberals might endorse, PPP adds a layer to the question, asking whether President Bush "intentionally misled." It's a question of mental state: did he know his claims were false, or did he craft a biased process for gathering and vetting intelligence that generated biased evidence and analysis, thus allowing him and his staff to overestimate and overstate the threat posed by Saddam Hussein? Posed with that question, 14% of Democrats felt that President Bush didn't lie (15% were unsure), while 73% of Republicans felt he didn't lie.  Quite a gap.

A big part of that gap comes because of a counterfactual belief that's widespread in Republican circles: the belief that Saddam Hussein really did possess weapons of mass destruction.  A survey from Dartmouth, conducted in late April, 2012, found that 63% of Republicans agreed that: "Iraq had weapons of mass destruction when the United States invaded in 2003."  By contrast, 63% of Democrats correctly recalled that there were, in fact, no WMDs in Iraq when we invaded (15% thought there were WMDs, 22% were unsure). That partisan difference in factual knowledge shows up consistently in public opinion polls ever since the invasion.

So, combining those polls, we learn that 63% of Republicans think there were WMDs in Iraq, and an additional 10% (give or take the margin of error) know there weren't, but want to give George W. Bush credit for having fooled himself into believing the BS he was selling.

Combining those polls gets tricky though, since more Democrats seem to think Bush lied (per the PPP poll) than think there weren't WMDs in Iraq (in the earlier poll). I'd guess that the strong wording of the question triggered a stronger partisan, even tribal response. Republicans were more likely to try to defend Bush, while Democrats were more likely to vilify him.

You can see this effect in the question about climate change, which 58% of Republicans in the PPP survey said was "a hoax." But that same week, a new survey from the researchers who produce the "Six Americas" assessment of public opinion on climate change showed that only a quarter of Republicans or Republican-leaning Independents think climate change is not happening. Half know that it is happening, and more – 6 in 10 – think we should do something about it.

Again, phrasing the question in a more politicized way and in stronger terms made people more likely to reject the science. The Six Americas survey defined climate change carefully, as "the idea that the world’s average temperature has been increasing over the past 150 years, may be increasing more in the future, and that other aspects of the world’s climate may change as a result." That careful, scientific framing strips the issue of its political context: it isn't a question about economic policy or allegiance to a partisan agenda. Asking whether "global warming" (the more politicized term) is a "hoax" taps into the partisan framing. Senator James Inhofe has used that phrase as a book title and in Senate hearings, to attack both the science of climate change and the policies proposed to address it.

While you might expect that stronger charges would attract less support – logically, the subset of people who think global warming is a hoax ought to be nested within, and therefore smaller than the subset who think it isn't happening – we see that they don't. Placing charges into a political context draws in people who know that the underlying factual claim is wrong, but still want to endorse their tribe's position. As I've said before, it's often worth being wrong in order to maintain group membership.

More like this