When the Issue is Acceptable Deaths,
'Taboo or Not Taboo – That is the Question' ~ #36

By Alan F. Kay, PhD
© 2004, (fair use with attribution and copy to authors)
April 14, 2004

My colleagues and I came upon an unusual opportunity to ask a couple of taboo questions, Q71 and Q72, in ATI# 16 (June 1991) that came in part from the question-set dubbed "the People's Military Policy,” described in "The Polling Critic" Column # 22, "How Bush Lost Non-U.S. Public Opinion," June 5, 2003.  Questions on how many deaths are acceptable in a military intervention were asked only once, 13 years ago.  It would be difficult, if not impossible, to get reliable, meaningful data on such questions now.  Nevertheless, the findings are extremely critical to decisions that are going to be made on terrorism in Iraq, Afghanistan, and the United States, but affecting the whole world, particularly the monotheistic world of Christians, Muslims and Jews.

Thirteen years ago in ATI#16, a battery of six items, Q63-68, had this frame:

From time to time the behavior of dictators, like Manuel Noriega in Panama and Saddam Hussein in Iraq, threatens U.S. and international security.  I am going to read you some things that dictators may do.  For each one, please tell me whether you agree or disagree that the United States and the United Nations should take clear action to stop it.

First/(How about) [asked in random order]              

Over 98%, an amazingly large percentage, of the sample (986 out of 1002 respondents) agreed, somewhat or strongly, that the United States and the United Nations should take clear action to stop at least one of the six dictator actions.  I have never seen such a high percentage of agreement with any proposal before or since.

 The 986 were then asked in two half samples, A (Q69) and B (Q70), nearly identical questions,

In some cases, you were willing to support clear action to stop what a dictator was doing.  Please tell me which of the following two actions is closer to what you would favor:

The closeness in results from the two half samples tells us that the different wordings did not make any significant difference in findings, which seems sensible.  It is reasonable and not a surprise that a respondent favoring military intervention and combat accepts that some loss of life would be inevitable. These responses practically begged ATI to let the next questions go after what people consider an acceptable loss of life.  Still, how to do that successfully was not easy to figure out.  The ATI survey design team chose for all 986 respondents to ask first question Q71 and then Q72, as follows:

I would like to get some idea of what you think "too much loss of life" is in a military intervention [Rotate items in Q71 and in Q72]

First think about___/ Next think about____

What would be the rough figure you would use as a limit to an acceptable number of___ [Q71] U.S. deaths / [Q72] deaths of citizens of the dictator's country?

Just stop me when I reach the limit you would use. [read list and stop when respondent makes choice]

Here are the responses

	Acceptable deaths:	U.S.	non-U.S.
1.	Zero/None [not read]	12%	9%
2.	One	22%	18%
3.	10	9%	6%
4.	100	16%	14%
5.	1,000	16%	16%
6.	10,000	10%	13%
7.	100,000	3%	6%
8.	1 million	1%	2%
9.	No limit	4%	7%
	Refused/Don't Know	7%	10%

Item 1 was not read, but volunteered by respondents. The options offered, in the order of asking, were items 2 through 9, the size of which increased by a factor of 10 from each item to the next.  Remarkably, every statement in the following paragraph is true for both the U.S. and non-U.S. data. 

"Not more than one death" had the most support, and combined with volunteered item 1, essentially meant that no deaths were acceptable to about one-third of the 79% of the public willing to use military force (46% of the total public).  Roughly two-thirds of those willing to use military force (a majority of about 53% of the total public) had no such squeamishness.  They were ready to give numerical estimates.  Of the percentages who voiced an acceptance of the various choices offered regarding death tolls, the maximum was 16% for both U.S. and non-U.S. deaths (at 1,000), and this 16% was a larger aggregation than those who refused or gave no answer. For both U.S. and non-U.S. deaths above "1,000," support fell off from the 16% maximum reaching a minimum of 1% or 2% at  1,000,000, but rose again with the No limit option to a support level that was less than the support for 10,000 and more than for 100,000.  Both for U.S. and non-U.S. deaths in the order that items were asked, following an initial increase from zero/none to one acceptance switches exactly four times: decrease, increase, decrease, increase.  This is a most complex behavior of findings on the relationship between one variable (here the number of deaths) and another (percent accepting the deaths), and as far from the "normal" bell-shaped curve as I have ever seen.

Beyond all that, there is a perhaps surprisingly small difference between U.S. and non-U.S. deaths found acceptable.  The data tells us that, on a net basis, a large fraction, 87% or 88%, of respondents did not make a distinction between U.S. and non-U.S. death acceptability, while 12% or 13% did, including about a 3% increase in those saying "Don't Know," or not giving a substantive answer, who did deal with the acceptability of U.S. deaths, but could not do so for non-U.S. deaths.

Although there are many other ways one might try to phrase questions on this subject, to my knowledge no other pollster has ever probed "acceptability" of deaths.  Should it be a taboo?  I do not know.  As a result of a personal commitment to work only for clearly beneficial social and environmental change, a commitment that was not so important to me in 1991 but is now, I will certainly not participate again in any survey that asks questions on the acceptability of deaths.

It is nonetheless true that one could hardly imagine a question that would be more on the mind of most people than how long the Middle-East coalition forces, the public, and all others involved will continue to accept the kind of hard-line behavior that is the hallmark of this administration as it urges all to stay the course.  Though it is unpopular to say so, acceptance of casualties will not continue forever.

No polling can anticipate whether and how the public will turn on this issue during the coming months.  We have never been quite this way before and there are a number of choices the administration and the election campaigns may make that are somewhat unknowable, yet bare on the outcome.

 

>>> 2.5  The Polling Critic

Table of Contents

Return to Top