Tuesday, December 08, 2015

First Ban All Muslims? Donald Trump, Polls And Echidne Researches


I have so hoped that Trump would just go away.  He's not planning to go away, however, because he has far too much fun by showing how silly a sizable fraction of American voters are, how they find in Trump something that they can relate to, how he is this man who speaks plainly and who tells what's on the minds of many of his supporters, how he appeals to that (bad) idea in many of us that politics is a lot easier than bankrupting firms and Donald has shown us that he can do the latter so why not the former.

Today he went all fascist, so sadly I had to go and do research.  The reason:  Trump argues that all Muslim immigrants and even Muslim tourists should be banned from entering the United States, because so many Muslims hate Americans and the American way of life:

"Donald J. Trump is calling for a total and complete shutdown of Muslims entering the United States until our country's representatives can figure out what is going on," a campaign press release said.*
What motivated Trump to make this announcement?  Partly his permanent-gold-foot-in-the-mouth disease, but partly a survey from last June.

This survey, by the Center for Security Policy**, purportedly interviewed six hundred American Muslims about their beliefs, and, according to both the website of the center and Trump, found that 

...a quarter of Muslims living in the U.S. believe violence against Americans is justified as part of a global jihadist campaign.
Here is the summary of the survey, and here are the cross-tabulations with some abbreviated forms of the questions it used.  I went through everything so that you don't have to, but, really, always verify.

When someone presents a poll or a survey which is supposed to reflect the average opinions of some wider group the first thing you should ask if it actually represents those opinions.  This particular survey fails to pass that test.  The reason is simple:  It was an Internet convenience survey, not a survey which made an effort to get a representative sample of all American Muslims into the study.

We don't know if the people who self-selected into the survey are roughly the same in their opinions as American Muslims, on average.  We don't even know if the respondents are Muslims!  I am not saying that they aren't.  Simply that we have no way of knowing, and neither do we have any way to tell how representative the views in the survey are of all Muslims in this country.  As one critic noted, strictly all we know is this:

This survey does not represent the views of American Muslims. It only represents the views of the 600 Muslims that it polled.
Assuming the respondents were mostly Muslims...

There are ways of judging the likelihood that the survey was not representative of the American Muslim populations.  For example, a representative survey would have roughly the same percentages of Sunnis, Shias etc. as their percentages in the general Muslim population.

But that is not the case in the Center for Security Policy survey.  Based on Wikipedia (yes, I know), roughly 50% of American Muslims are Sunnis, 16% Shias and 22% non-affiliated.  Yet in the linked survey Sunnis are only 40%, the Shias 13% and the non-affiliated 39%.  So the non-affiliated look to be strongly over-sampled.

This matters on two levels.  First, it's evidence that the survey probably isn't representative, even as a fluke.

Second, the findings in the survey are driven by the high percentages given as a yes-answer to the question*** by the Shias (40%) and the non-affiliated (34%) in the survey.  Only 14% of Sunnis in the survey responded affirmatively to that question, and they are the largest Muslim group in the US.

As an aside, something smells off in those survey answers, because most recent terrorist acts against the West have been committed by Sunni Muslims, not by Shias or the non-affiliated.****   One would expect the Sunni percentages to be higher.  Or at least I did.

---------------

*I know what's going on.  As Kamel Daoud, a French writer, wrote, the current situation has both a mother and a father.  The mother is the Iraq invasion (and the Israel-Palestine conflict, I 'd add).  The father is Saudi-supported extreme Wahhabism, the so-called petro-Islam, disseminated with money (from the sale of oil, ultimately).  For various inane-but-power-related reasons those parents are invisible in most of the writings about Islamic terrorism.  We go after the existing children while more monster children are being birthed by that unholy marriage.

** The man who runs it, Frank Gaffney, has a reputation as someone who fans the flames of anti-Muslim bigotry.  (I'm no longer calling that Islamophobia, by the way, because I criticize religions on this blog, including Islam, and that kind of criticism is important.  Besides, it's not irrational for women to be afraid of extremist Abrahamic religious interpretations.  It's rational.  But bigotry against Muslims is horrible.)

The site is biased, obviously.  Note, also, this juxtaposition:

According to the just-released survey of Muslims, a majority (51%) agreed that “Muslims in America should have the choice of being governed according to shariah.”  When that question was put to the broader U.S. population, the overwhelming majority held that shariah should not displace the U.S. Constitution (86% to 2%).
The two questions are not the same.  At the same time, I'm not a fan of shariah, and don't wish to see it offered as a choice in the US.  This is because women get a very rough deal in all currently used interpretations of shariah, and women in fundamentalist families might not be able to escape using the shariah courts in the case of family disputes.  Those cover divorces (much easier for men than women), child custody (the basic rule is that custody goes to the father at some point) and inheritance (daughters inherit half of what sons do).   In general, ancient religious laws treat women abominably and Islam is no exception to that.


*** The statement the respondents were asked to agree or disagree with is this:

Violence against Americans here in the United States can be justified as part of the global jihad.
You may have spotted that there's a way of understanding that question to be about someone else's opinions, not about the opinions of the respondent.  That shouldn't happen in a well-crafted survey.

**** Wahhabists and Salafists, the most likely denominations from which terrorism comes, are both Sunnis.