How the Media Punks Readers with Polls
Did you know that anti-tax advocate Grover Norquist has a 25% higher approval rating than New York Times statistician Nate Silver. More bad news for Silver: he’s trusted by less than half the number of Americans that trust the Gallup Poll.
Is this the sign of a youth rebellion? Young people—a voting block that went in large part for Barack Obama in the 2012 election—support having their state secede from the union by a margin of nearly 2 to 1 over older voters. And did you know that nearly one in twenty Obama voters aren’t convinced that the president won the election legitimately?
Skeptical? Don’t be! These shocking statistics are all from a recent poll by the Democratic polling firm Public Policy Polling.
Actually, be skeptical. Very skeptical.
Citizens need to develop critical thinking skills and learn not to trust the news media when they report on "polls." That’s because while everything I said in that opening paragraph is completely true based on a recent Public Policy Polling poll, it's illustrative of how media spin defines poll reporting.
Of course, the usual spin is anti-conservative. Take this December 4th story from The Hill by writer Justin Sink titled "Poll: Half of Republican voters say ACORN stole election."
Nearly half of Republican voters say that ACORN — the community organizing group that closed in 2010 — aided in stealing the 2012 election for President Obama, according to a new poll released Tuesday.
The survey, conducted by Democratic polling firm Public Policy Polling, found that 49 percent of GOP voters believe that the president did not legitimately win reelection because ACORN interfered with the vote. A full 50 percent of Republicans said Democrats engaged in some sort of voter fraud.
ACORN earned wide ridicule in 2009 after a hidden-camera video from conservative activist James O'Keefe showed ACORN employees advising walk-ins on how they could engage in criminal activity. Questions were also raised about the legitimacy of voter registration forms submitted by the group. The ensuing controversy led Congress to prohibit government funding from going to ACORN. Without federal dollars, the organizing coalition was forced to shut its doors in 2010.
The goal is to make Republicans look conspiratorial and dumb. ACORN was shut down, and half of those kooky GOP voters think that they stole the election! Leave aside the fact that ACORN was only shut down in name only for the time being.
Sink doesn’t provide readers to a link the PPP website. If he did, they might find that a press release from the group titled "A quarter of Republicans want to secede, half think election stolen" that provides the lead to his piece. If you dig a little deeper on the PPP site, you’ll find the actual raw data.
Let’s start with the obvious: 50% of all Republicans didn’t say that. There are about 55 million registered Republicans in the United States.
The raw data on that PPP poll is right here in this press release from the polling company. The poll was conducted on less than 800 Americans. And it was all done via an automated telephone system, which means there was no independent verification of who was being spoken to and no way for the poll takers to ask questions or clarify what a question might mean.
Next, let's look at the question that was actually asked of poll participants to get this result. The Hill article doesn’t tell readers what the question is, but it’s in the linked report: "Do you think that Barack Obama legitimately won the Presidential election this year, or do you think that ACORN stole it for him?"
This is question framed intentionally to get a result. There are two options: Obama won legitimately or it’s ACORN. They don’t ask a broader yes / no question such as “Do you believe there was voter fraud?” Nope—two choices: legit or ACORN.
Picture what is going on here. You get a phone call out of the blue. It’s a pre-recorded message with a poll. There’s nothing riding on it; you know your answers are meaningless and anonymous. You either hang up or go through the dopey process. You know you’re under some pressure to answer or the recording will hang up. There are a series of questions.
You get a question about whether Obama legitimately won. If you think he did, that’s easy—press the button for yes. If you think he didn’t, for any number or reasons, then the answer open to you is ACORN did it. That’s it.
A similar twist occurs on the headline grabber that "a quarter of Republicans want to secede." That’s a great way to target all GOP voters as reactionary kooks, isn’t it? However, I can easily take the same polling data and put a new spin on it.
The “secede number” that jumped out at me was the significantly higher percentage of 18-23 year olds who bought into that position. My hunch: they didn’t actually understand the question. Whatever, it’s a statistic!
I was able to fit this data into an impressive-sounding Tweet, even: "Post-election survey by Democratic pollsters shows key Obama demo of younger voters twice as likely to favor secession (!) as older voters."
And it’s true, or at least technically true. The numbers are right there in the poll. It’s just not relevant to anything in reality. It’s stunt polling, showboating statistics to score partisan points. It’s clever games with words and numbers.
Could PPP really be that cynical? Submitted for your approval; they actually included a fake question with a made-up issue.
Panetta-Burns is a deficit reduction plan. Panetta-Burns is supported by 8 percent of Americans and opposed by 17 percent. But most importantly, Panetta-Burns is not real: It’s an imaginary policy plan created by Public Policy Polling to see how many Americans would profess to have an opinion on a policy that does not exist.
Is this a political poll or a late-night comedy skit?
Playing games with the polls is easy. Here’s a couple of examples of how I manipulated statistics in the opening paragraph:
Norquist’s approval rating is 15%, Silver’s is 12%. Both numbers are small, so rather than say Norquist has a 3% higher rating, I used the more impressive sounding “25% more” since 15 is 25% higher than 12. This is a trick I learned from the Eric Holder Department of Justice.
About 4% of Obama voters told PPP they agreed (2%) or weren’t sure (2%) whether Obama won the election legitimately. I added those numbers together and described them as "aren’t convinced." Four percent sounds small, so I changed it to the more impressive sounding “nearly 1 in 20,” since 1 in 20 would be 5% and 4% is nearly that.
Keep digging in the raw data and you’ll find all sorts of stuff that didn’t make headlines: 33% of Hispanics believe the Democrats engaged in voter fraud. That’s 1 in 3 Hispanics! Nationwide! Liberals viewed Nate Silver more unfavorably than conservatives!
This isn’t news. It’s a prank. It’s a game. And the American people lose when journalists play it.