Monday, September 29, 2008

Rhetoric: Opinion Polls

The use of polls in politics is often a subject of discussion.

People frequently refer to opinion polls in political debate, without giving much care to whether the polls report accurate data. Unless a poll is conducted among a large, representative sample of the group whose views or opinions it is attempting to represent, it risks being inaccurate. The poll's questions must also be crafted carefully, or they will lead respondents to give certain answers.

But apart from issues of accuracy, there are some other ways that polls are misused in political debate.

The "Basing Their Views on Polls" Accusation

Politicians and pundits frequently accuse their opponents of being influenced by polling data. They claim, for instance, that their opponents are using polls to determine their political positions: if a poll says that the voters want X, then their opponent (so the claim goes) decides to support X. Their political views are based on what the polls say voters want.

There are a couple of points on which to challenge this sort of claim.

In the first place, these accusations are seldom backed up with conclusive evidence. How would you prove this sort of claim, after all? You'd need to do more than just show that someone supported X and that a poll showed that voters also supported X. You'd need to show that the former led to the latter. That is, you'd need to show more than just correlation, but causation as well.

In the second place, it's arguable that reflecting the will of the voters is what politicians are supposed to do. As I say, this is arguable. It is often argued that it demonstrates a lack of character for politicians to simply reflect the will of the voters. And others argue that the job of our elected representatives is to exercise their best judgment, even if that judgment is contrary to what the voters think is best.

But this is all debatable. There certainly ought to be some connection between the positions of politicians and will of the voters. And it's not obviously wrong for politicians to be abiding by what they want (though it's not obviously always right, either).

And, even if it were demonstrably true that someone had adopted a political position simply in order to abide by what the voters want, that tells us nothing about whether that position is wrong. It is ad hominem reasoning to argue that a position should be rejected because someone else adopted it on the basis of a poll.

At any rate, there is nothing intrinsically wrong with politicians using polling. In fact, it can be extraordinarily helpful when it comes to knowing where to direct their efforts. Suppose that polling reveals to a politician that voters overwhelmingly agree with his position on abortion, but overwhelmingly disagree with his position on taxes: this information will indicate that his time would be better spent trying to convince people that his views on taxes are correct, rather than his abortion policies, which his constituents already agree with.

Appealing to the Majority

Polls are often used in another way, to try to convince people out of their positions.

For instance, it is often pointed out to a politician that their position on a particular issue runs contrary to what polls say the people want. It usually comes in the form of a statement along the lines of:
"Polls show that a majority of people in the country or in the world oppose you on X, or have a negative opinion of you: how do you respond to this?"
The idea is to put the politician on the spot. How can he oppose the will of the majority?

But the majority isn't always right. Sometimes they are wrong. To argue that "a majority believes X, therefore X is true" is to make a fallacious argument known as argumentum ad populum (or an "appeal to poularity").

So, if someone points out that a poll says that a majority of people hold a particular belief, it is important not to infer -- or to allow anyone to imply -- that that belief is therefore correct.

Polls Seldom Ask About Justification

Related to the "appeal to the majority" matter, it is important to note that polls -- though they often ask people their opinions about what is good and bad, right and wrong, etc. -- they seldom ask voters why they hold those opinions. That is, they seldom ask voters to defend their opinions, or to explain how they justify holding those opinions.

Polls often don't check to see if an opinion is an informed opinion.

As such, merely looking at polling data, you get little or no idea about whether peoples' opinions are based on actual facts or sound reasoning. Often times, if you ask people to defend their opinions, you discover that their opinions are based on false beliefs or flawed reasoning.

As such, we should be unimpressed by reports that "a majority of people believe such-and-such" if we are given no indication that that belief is based on true premises and valid reasoning.


Polls are not necessarily a bad thing. But they don't necessarily carry all the weight that people imagine that they do, either. If someone brings up a poll in a political discussion, be careful about what conclusions are being offered on the basis of the poll.

"According to a June Gallup report, most Republicans (58 percent) believed that God created humans in their present form within the last 10,000 years. Most Democrats and independents did not agree. This anti-intellectualism is antediluvian. No wonder a 2009 Pew Research Center report found that only 6 percent of scientists identified as Republican and 9 percent identified as conservative. Furthermore, a 2005 study found that just 11 percent of college professors identified as Republican and 15 percent identified as conservative. Some argue that this simply represents a liberal bias in academia. But just as strong a case could be made that people who absorb facts easily don’t suffer fools gladly."
-- Columnist Charles Blow, December 7, 2012.

Comment: Blow is appealing to polling data to argue that Republicans are "anti-intellectual", which amounts to caricaturing them as "stupid" or perhaps as not caring about facts. Blow also considers a causal connection between political affiliation and scientific occupation, though isn't this a case of false causation?

CARNEY: It's not good government for one party in Congress to refuse to acknowledge what a compromise has to include, a compromised position that is not just the President's position, is not just the Democratic Party's position, but it's the position of the majority of the American people. I mean, I think we've seen data again today that reinforced that fundamental fact. And it's certainly not good government -- the reference you made to our debt ceiling debacle -- to even hint at the possibility of holding the American economy hostage again to the ideological whims of one wing of one party in Congress. That’s unacceptable.

REPORTER [unidentified]: Jay, speaking of the debt ceiling, does an agreement to raise the debt ceiling have to be part of an agreement to avert the fiscal cliff?
CARNEY: We're not going to negotiate over what is a fundamental responsibility of Congress, which is to pay the bills that Congress incurred. It should be part of the deal. It should be done and it should be done without drama. We cannot allow our economy to be held hostage again to the whims of an ideological agenda.
-- White House briefing with Press Secretary Jay Carney, December 4, 2012.

Comment: Carney is making a claim about what Americans want (he seems to cite polling data to back up his assertion). Also, Carney is indulging in "hostage-taking" rhetoric. Related to that, why can't Republicans in Congress bargain in exchange for agreeing to raise the debt ceiling? If they asked for spending reductions in return for raising the debt ceiling, why couldn't that be cast as "comprehensive" legislation? Why must the two be unrelated? Hasn't unrelated legislation been attached to defense spending bills in the past? Was that "hostage-taking"?

"But you know who doesn't want entitlement reform? Voters. Democratic voters, independent voters, and, yes, Republican voters. The Washington Post / ABC News poll asked voters about raising the Medicare eligibility age from 65 to 67 proposal that has been floated by Republicans in Congress. 67% of people oppose raising the Medicare eligibility age, including 71% of Democrats, 62% of independents, and 68% of Republicans. 68% of Republican voters oppose a Medicare reform proposal being floated by congressional Republicans."
-- TV pundit Lawrence O'Donnell, November 29, 2012.

Comment: First, this is a hasty generalization. Just because a majority of voters oppose raising the age of Medicare eligibility doesn't mean they oppose other Medicare reforms. Second, O'Donnell's use of opinion polls seems to be making an appeal to popularity. Also, by pointing out that Republicans (who O'Donnell frequently opposes), agree with him, O'Donnell seems to be making an "even my opponents agree" argument.

"Sen. Bernie Sanders from Vermont has spoken on this program about the need to protect programs for the middle class in debt negotiations. Sen. Sanders [I-VT] released this statement to the Ed Show tonight: "What [presidential advisor] David Plouffe has stated deeply concerns me. Despite Mr. Plouffe's assertions, the American people have been clear, both through their votes in the election and in poll after poll after poll. At a time when the middle class is disappearing and the number of people living in poverty is at an all-time high, the American people have demanded that there be no benefit cuts to Social Security, Medicare, and Medicaid, and that the wealthiest people and largest corporations in this country, who are doing phenomenally well, must be asked to play a significant role in reducing the deficit." Here, here, Bernie! I'm on board with that! The public agrees with Sen. Sanders, I'm not the only one. In the latest CNN poll, 56% of Americans believe that taxes for the wealthy should be raised to help pay for programs such as Medicare and Medicaid."
-- TV pundit Ed Schultz, November 26, 2012.

Comment: Citing an opinion poll that claims 56% support for a positions indicates a majority, but does it indicate what Americans want as a whole? Plus, is Schultz making an appeal to popularity?

"You see, Mr. President, real leaders don't follow polls; real leaders change polls."
-- Gov. Chris Christie (R-NJ), August 28, 2012, giving the keynote address at the GOP National Convention.

Comment: This is another derisive caricature, accusing President Barack Obama of simply adopting positions according to polls.

(The list above is not intended to be a comprehensive record of all relevant examples.)

No comments: