From April 5-20 more than 30 volunteers, mainly college students in the San Francisco area, polled a random sample of the U.S. population on their knowledge and views concerning Constitutional rights and the War on Terrorism. From over 1000 people contacted, 215 from 46 states agreed to participate. The results are discussed below. Retro Poll uses a unique methodology, which investigates peoples' background knowledge on the subject matter before asking their opinions. This allows an assessment of the extent to which background knowledge or its absence contributes to particular political views.
After an opening opinion question: describe the impact of the War on Terrorism on our civil rights in the U.S. (33.6% said it's strengthening them, 22.9% said no impact, and 19.2% said it's removing important rights and guarantees) the next 12 questions measured basic knowledge in several areas.
Disturbing for our democracy, only 40% could name the three branches of the U.S. government and 37% could not even name one branch. People fared better on whether 4 different statements contained elements of the Bill of Rights. Sixty six percent knew of the right to a speedy and public trial. Fifty seven percent were aware of the protections of the Fifth Amendment against coerced self-incrimination. Seventy four percent knew that indefinite punishment and detention was not a right. And seventy nine percent knew that we are (or used to be) protected from unreasonable searches and seizures in our homes and private lives.
Because of concerns that the public may believe that only non-citizens can be subject to detention without trials and other new executive measures, Retro Poll assessed whether people knew that over 2/3 of the people of Japanese ancestry put in concentration camps in 1941 were actually U.S. citizens. Fifty nine percent of respondents knew it. Fifty four percent knew that the U.S. funded Osama Bin Laden in the 1980s, 70% knew that Iraq invaded Kuwait in 1990, and 55% knew that Senator Joe McCarthy used perjured witnesses. We consider this a good degree of historical knowledge by many in these specific areas.
On the other hand, 42% continue to believe there is evidence (only 36% say there is not) that Saddam Hussein worked with the 9-11 terrorists, a key message of the Bush team that has never been bolstered by fact. And when assessing who supports the war against Iraq we have now shown in two consecutive polls (total sample size 365) that it is mainly people who accepted the government's hoopla about an Iraqi threat on face value (they remain 2:1 for war). Indeed the association between being fooled by this government disinformation and supporting the war is statistically significant at the .005 level in the current poll and at the .001 level in the earlier poll. People who are not fooled by the government propaganda are still 60% against this and future unprovoked wars.
When provided information from a Federal Law suit against Florida which was settled with the admission that Governor Jeb Bush's people had removed more than 90,000 legal registered voters, mostly African American, from the voter rolls before the 2000 election, 43% of poll respondents deduced correctly that this assured the election of George W. Bush.
Exploring the issue of a "climate of fear" Retro Poll found that 61% believed that there is now a "general climate of fear in the U.S." That group was then given 6 items to rate as contributing "much, some, little or none" to the climate of fear. When those who said "much" or "some" for each question were combined, the respondents ranking for causes of fear were: Terrorist actions (85%) media hype (83%); failure to catch Bin Laden (77%); homeland security alerts (73%); poor relations with other nations (71%); Saddam (71%); and the anthrax attacker still on the loose (65%). With "media hype" getting the most votes for causing "much" fear (57%), the poll provides evidence that the public is very suspicious of the way that TV (where 2/3 get their news) and other media process and manipulate information.
People were asked whether they support or reject three specific elements of the USA Patriot act. First was that Government officers may enter homes without our knowledge, copy what they want and not even tell us they were there. An amazing 89% rejected this principle. The next question was about librarians being required to divulge peoples reading habits and forbidden from telling you they've given information. Sixty five percent (65%) rejected that. Lastly people were asked about the monitoring of phones, e-mails and Internet activity without probable cause. Seventy seven percent (77%) rejected that provision of the law.
Our analysis then looked to see if people who believe that the War on Terrorism is protecting their right to dissent and strengthens civil rights answered differently on the Patriot Act questions. As expected they were not as strongly opposed to the intrusions into their privacy and their basic liberties. Surprisingly however, this group was also overwhelmingly opposed to each Patriot provision (80%, 55%, 70%--although the differences were statistically significant in two of the three question: X squared <0.02). One possible interpretation of this contradiction is that people who believe that the War on Terrorism is protecting their rights might have been unaware that the Patriot Act powers attacked their rights. Would it be unreasonable to hold the media in general, and TV in particular (where two out of three respondents get much of their information on current events) responsible for the lack of exposure of the Patriot Act's anti-constitutional provisions and the attendant general public confusion?
Another series of questions asked whether people support or reject several other elements of the War on Terrorism. These questions were repeated from an earlier poll: should the U.S. support international attempts to prosecute war crimes and crimes against humanity (83% supported in this poll vs. 94% last September)? Lengthy detentions for citizens and non-citizens that the government arrests without charges, proofs or trials (83% rejected vs. 78% in September). The use of outlawed interrogation techniques such as torture (88% rejected vs. 85% in September). A requirement that the U.S. must prove accusations against other countries before attacking them (69% supported this vs. 89% in September, a significant change). And war against Iraq or other countries the US labels as supporting terrorism when they are not attacking anyone (44% support, 36% reject and 20% don't know). Excepting the last question, the public puts itself in dramatic and overwhelming opposition to the policies, positions and behaviors of the U.S. government, though many could be unaware of those actual policies.
The question on war against various nations contains compound elements. By including other unnamed nations as well as Iraq, the poll intentionally incorporates the vagaries that are part of the double speak issued by Government officials and advisors about attacking other nations. However, these vagaries have unclear impacts upon peoples' understanding of the question, so that this element should be reworded for future polls.
The demographics of the poll sample revealed problems that exist for all major polls-problems that are not being debated and addressed in the media. Although the sample frame is constructed to include proportional representation for various ethnicities and geographical locations, Retro Poll finds that African Americans and Latin Americans are significantly underrepresented in the response group. This is the result of a higher proportion of refusals to participate among these groups. Gallop and other polls claim that they adjust their results to assure that underrepresented groups are weighted to the appropriate level. Doing this does not, however, solve the sampling bias problem.
Poll organizations know that usually in the general population over 70% of people contacted will refuse to be polled. This introduces an unmeasured bias into the sampling frame. The major poll samples (and ours as well) cannot be said to be random samples even if the phone lists are generated randomly. It would take extensive and ongoing evaluation of refusers to assess whether and when there is a self-selection bias introduced by their non-participation. The fact that higher non-participation rates are seen in the African American and Latin American communities in our polls adds fuel to the argument that there must be socio-political and economic reasons why many people do not opt to participate in polls. This flaw in the public opinion methodology is mainly ignored because an open discussion of this poll weakness might weaken the heavy reliance of politicians and media corporations upon them for marketing and manipulation purposes. Unlike Retro Poll, polling organizations get paid big bucks to carry out these polls.
The under representation
of Blacks and Latinos in the current Retro Poll probably reflects greater
bias than does the problem of small sample size. Critics may focus on
Retro Poll's small sample size, but avoid the more serious overall polling
problem of lack of randomness due to self-selection bias. The Retro Poll
Committee believes that random telephone poll results (including our own)
do not have a high degree of reliability. However, looking at within poll
comparisons, correlations and cross tabulations as we have done in this
analysis can provide very powerful and reliable information on the strength
of association of peoples' responses to different questions. Retro Poll
will continue to look at correlations between knowledge and opinion in
its future polls. The Retro Poll Committee believes that the mass media
and TV in particular are a key contributor to poor public knowledge on
critical current events and that public views largely reflect such misinformation,
rather than particular values or ideologies.
| POLLS | RESULTS OF POLLS
| CORRECT ANSWERS AND CITATIONS RELATED TO EACH POLL
ABOUT OUR POLLING STATISTICS | MISSION STATEMENT | HOW YOU CAN HELP | ADVISORY BOARD |
SUBSCRIPTIONS | NEWS | FREQUENTLY ASKED QUESTIONS | OTHER | CONTACT