Supplemental: More unworkable work from Pew!

FRIDAY, NOVEMBER 7, 2014

More true belief from Marcotte: It’s painful to watch the way the liberal world now pretends to “reason.”

Consider what happened when Amanda Marcotte engaged in some true belief.

Yesterday, as we quit for the day, Marcotte seemed to have swallowed some rather unlikely statistics.

We were only four paragraphs into a recent piece at Salon. Already, though, Marcotte was serving us this:

MARCOTTE (10/24/14): That this polarization is going on isn’t a myth. Previous Pew research shows the percentage of Americans who are “mostly” or “consistently” conservative has grown from 18% in 2004 to 27% in 2014. During that same period, the percentage of Americans who are “mostly” or “consistently” liberal stayed a little more consistent, growing from 33% to 34% in 10 years. (These statistics don’t measure what you call yourself, but what you rate as on a scale of beliefs about various issues.) While liberals became more liberal, conservatives both became more numerous and more rigidly conservative over time. What gives?

Enter right-wing media, which has a nifty trick of convincing audiences it’s the other guys who are the liars, all while actually being much less trustworthy in reality. From conservative screaming about the “media elite” to Fox News’s old slogan “Fair and Balanced,” conservative media is rife with the message that everyone is out to get you, conservative viewer, and only in the warm blanket of right-wing propaganda will you be safe.

People loved this at Salon. Marcotte was saying The Other Tribe contains the vile propagandists.

Still, there were those peculiar statistics from Pew. Do they seem to make sense?

Tell the truth! Is it your impression that 34 percent of the American people are mostly or consistently liberal? Is it your impression that there are substantially more liberals than conservatives out there at the present time?

Is it your impression that conservatism was on the rocks in 2004—that there were almost twice as many liberals as conservatives in the year when we re-elected Bush?

Those numbers seemed a bit unlikely to us—and we happen to know how bad Pew’s work often is. So we decided to check Marcotte’s work.

This is what we found:

First, Marcotte was possibly conning her readers a bit with her selection of that 2004 starting-point. In fact, Pew has been tabulating the number of liberals and conservatives at intervals since 1994.

These are the numbers they’ve deathlessly found. Click here, scroll down a tad:

Percentages of liberals and conservatives (Pew)
1994: Liberals 21 percent/conservatives 30 percent
1999: Liberals 31 percent/conservatives 20 percent
2004: Liberals 33 percent/conservatives 18 percent
2011: Liberals 31 percent/conservatives 26 percent
2014: Liberals 34 percent/conservatives 27 percent

Those numbers have jumped around a fair bit, especially after 1994, when conservatives were said to be ascendant. But according to Pew, we seem to be living in a plurality-liberal land.

On their face, do those numbers make sense? Do you believe that 30 percent of Americans were conservative in 1994, with the number dropping to 18 percent as of 2004?

That didn’t seem to make sense to us either—until we looked at the way Pew has gathered its numbers.

Gack! In all these surveys, Pew has questioned respondents about the same ten topics. In each case, respondents have been asked to choose between two statements.

One statement represents the “liberal” position; the other statement is “conservative.” These are the choices respondents have always been given for Question 1:

Choices offered for Question 1:
Government is almost always wasteful and inefficient.
Government often does a better job than people give it credit for.

In this case, the second choice is the “liberal” position. The first choice is “conservative.” To review the choices for all ten topics, click here.

In the beginning, this may have seemed like a good way to run a survey like this. But as it turned out, several of Pew’s questions may have been poorly chosen.

Why were there so many fewer conservatives as of 2004? Just a guess: In large part, it seems that respondents answered some of these questions based on their view of the sitting president, rather than based on matters of principal.

Consider Question 1, about wasteful government, with the two choices shown above. In 1994, 74 percent of Republicans and Republican leaners chose the (rather over the top) first position.

(Click here, scroll down to Growing Gaps between Republicans and Democrats.)

People, Bill Clinton was president! In 2004, with Bush in charge, that number had dropped to 46 percent among Republicans and Republican-leaners.

Pew interprets that to mean that we had fewer conservatives in 2004. We’d offer a different, fairly obvious interpretation—it simply means that Republicans thought the government was more frugal under Bush.

In 2014, with Obama in charge, that number was back at 75 percent. In short, that question doesn’t seem to be measuring conservative values. In many cases, it seems to be measuring what respondents think of the sitting president.

Routinely, Pew does bad analytical work. Routinely, true believers use Pew’s data—when their data advance true belief. It seems to us that that’s what was happening here.

Amanda Marcotte is full of fire. That said, she’s often a horrible analyst, assuming she’s actually trying.

This was the latest sorry example. The progressive world will never succeed with people like this in charge.