Lies, Damn Lies, Statistics, and a Media Happy to Abuse Them

The Wall Street Journal reports ($) today that support for free trade is fading among Americans who are likely to vote Republican.  Perhaps that’s true.  It certainly wouldn’t be surprising given the way most Americans are misled by their political representatives and the mainstream media about how to measure trade’s impact on the economy.

But something really smells about today’s lead article in the WSJ.  The WSJ/NBC News poll upon which the article is based simply doesn’t support the author’s conclusions.  In fact, the article is misleading in ways I find inexcusable for a newspaper of that caliber.   If you weren’t already, you should be highly skeptical of polling results (at least as reported second hand).

The third paragraph in the article reads: “Six in 10 Republicans in the poll agreed with a statement that free trade has been bad for the U.S. and said they would agree with a Republican candidate who favored tougher regulations to limit foreign imports.”  Next to that paragraph is a graphic box with a bar chart showing responses to the question: “Is foreign trade good or bad for the U.S. economy?”  The “Good” bar showed 32%; the “Bad” bar showed 59%.

Here’s the first problem.  That question (“Is foreign trade good or bad for the U.S. economy?”) was not asked in the poll.  The second problem: no questions were asked about whether the respondents would agree with a Republican candidate who favors tougher regulations to limit foreign imports.  But that didn’t stop the author from reporting that phantom result in paragraph three.

Here is a link to the subject WSJ/NBC poll.  Question 10 is the only question about trade, which gives two statements and asks the respondent to reveal which statement comes closer to his/her point of view.

Statement A: “Foreign trade has been good for the U.S. economy, because demand for U.S. products abroad has resulted in economic growth and jobs for Americans here at home and provided more choices for consumers.” (32% of Republicans agree)

Statement B: “Foreign trade has been bad for the U.S. economy, because imports from abroad have reduced U.S. demand for American-made goods, cost jobs here at home, and produced potentially unsafe products.” (59% of Republicans agree)

From these results, John Harwood concludes that “six in 10 Republicans in the poll agreed with a statement that free trade has been bad for the U.S. and said they would agree with a Republican candidate who favored tougher regulations to limit foreign imports.”

But as you can see, there is a clear bias in the manner of phrasing the questions.  You’re not agreeing that foreign trade is good or bad, but that it’s good or bad because… And respondents are more likely to be familiar with one of the offered consequences of trade.  Certainly, the issue of “potentially unsafe products” is fresh on our minds, thus respondents are basically escorted to that answer.

What bugs me most about this is that the competing statements: foreign trade has been good for the U.S. economy vs. foreign trade has been bad for the U.S. economy would have been perfectly objective phraseology.  Why introduce subjective perspectives?

That a professional polling agency would introduce such obvious bias into its polls and a major newspaper would ignore the obvious problems with the results is troubling.  For all we know, Ron Paul and Mike Gravel are the leading candidates for their respective parties’ nominations.

CORRECTION: The poll did ask about a Republican candidate who favored tougher regulation. See here.