On Polling Bias and “The Manufacture of Consent”

The birth of modern propaganda took place in World War I, where an extraordinarily well orchestrated campaign turned America from pacifist to ferociously anti-German in a mere eighteen months. When after the war, the public learned that its beliefs had been turned without their realizing it, some of the key actors. such as Eddie Bernays (often described as the father of the public relations industry) and journalist Walter Lippman defended this type of activity. Lippman, in his 1922 book Public Opinion, contended that most people did not have the time, inclination, or ability to take the time to analyze the relevant information and parse it. Thus, it was important for well informed people to take responsibility for this process, which Lippman called “the manufacture of consent.” This is, of course, an argument for the role of an elite, steering the masses who are less capable of coping with its complexities.

Americans in particular are leery of this sort of thinking; elitism has a very bad name here. Yet there has rarely been a time when the gap between the beliefs of those in the corridors of power and those of ordinary people has been so wide.

One of the reasons may be that the very needs of those at the top of the food chain to sell their messages is leading to distorted feedback. We saw a ham-handed effort earlier this year of a deliberate effort to generate “opinion” data that would support the agenda of a particular group, in this case, the Peterson Institute’s long-running campaign against Social Security and Medicare. But their “America Speaks” campaign backfired rather spectacularly.

An attentive reader pointed out a subtler and more widespread version of this problem, keying off a weekend article in Raw Story, “Poll: Vast majority opposes attack on Iran“:

Fewer than one in five Americans would support a US military strike on Iran if the Middle Eastern country continued to pursue its nuclear program in the face of international sanctions, a new poll indicates.

The poll, carried out in June for the Chicago Council on Global Affairs, finds 18 percent would support a strike on Iran if the country failed to stop its enrichment of uranium. Forty-one percent would urge further economic sanctions against the country, and 33 percent would support further diplomatic engagement.

When asked what the UN should do, the answers were similar: 21 percent support military action, while 45 percent want more sanctions and 26 percent want negotiations.

“Americans are gravely con cerned about Iran’s nuclear program. Yet they are also quite concerned about the possible negative impact of a military strike to try and stop it,” the survey’s authors state. “Only a small minority favors the use of military force now, and if all efforts to stop Iran from develop ing nuclear weapons fail, Americans are essentially evenly divided over whether to conduct a strike.”

The survey (PDF) also finds an electorate that is far less certain of its support of Israel than US political leaders would suggest. By a narrow margin — 50 percent to 47 percent — Americans would oppose the US militarily defending Israel if it were the victim of an unprovoked attack.

If the attack against Israel were retaliation for Israeli military action, even more — 56 percent — would oppose US military intervention, while 38 percent would support it.

Our correspondent’s remarks:

To hear any kind of mainstream discussion, and even most discourse in the blogosphere, it’s taken as a given that Americans: a) by a vast majority will back Israel all the way in anything, b) want Iran bombed if they don’t grovel, and c) despise ‘those Islamic upstarts.’ In fact, this substantive poll with a large sample shows those positions to be false insofar as the public is concerned. Those statements are true only _of the Beltway elite_. In short, those at the top of the pile are running these policies, together with collaborative mass media, in direct opposition to public sentiment; not something one hears reported, unsurprisingly.

There are pols which show more hostile attitudes by the American public yes, but that’s where the plot thickens. It’s not generally well understood the extent to which the polling industry is, largely knowingly, producing ‘results’ which significantly overstate right-wing biases. This is done by the phrasing of the questions and the framing of alternatives. The issue has been discussed among others by George Lakoff, a linguist with considerable expertise in ‘framing’ (who as a dyed-in-the-wool liberal sold himself completely to the Democrats and now, I rather thinks, regrets his credulity in that regard.) The country _sounds_ more right-wing because the right wing media frames the issues and right-leaning language in polls elicits more right-leaning answers than a neutral remark would. I’m not being paranoid on this: it’s a major issue in US political discourse which I’ve only seen a few scattered discussions of in recent years, even on radical websites (whose membership largely prefers conspiracy theories even beyond REAL conspiracies such s the Koch brothers’ dealings).

I haven’t looked at this systematically, but some of my early work was in survey research, and anyone who has had any exposure knows that the results of questionnaires are quite sensitive to the wording of the questions. And I have noticed more than occasionally signs of bias in polling (unnecessarily value-laden phrases, strange either-or pairings that exclude other choices, leading question sequences). I generally take opinion polls when called out of curiosity but I can recall at least a couple of time aborting the call because the questions were very leading. So my hunch is that there is more that a little truth in this argument, even though it would take a fair bit of effort to prove it in a rigorous manner.

Print Friendly, PDF & Email


  1. jbmoore

    I was polled by an automated Democratic polling system last Friday. My answers boiled down to even though I believe that the Democrats are doing a terrible job at governing, I’ll still vote for them rather than let the Republicans finish destroying the system like they almost did in 2008. I don’t know what results the poll will gather, but it’s likely that the Democrats are counting on most voters remembering which party was most responsible for the financial meltdown and the GFC in spite of their own terrible record these past two years.

    1. Doug Terpstra

      The duopoly relies and thrives on such lesser evilism, and such farcical wackiness may well be a product of the consent manufacturing process. It depends how you’re vested in the current system, I suppose, but I’m rooting for creative instability at least, if not creative destruction.

  2. readerOfTeaLeaves

    Years ago, I was involved with some political polling. Different questioners would get such varied responses that I concluded the ability of the polling rep had a significant influence.

    We changed the question order, and often that affected how respondents prioritized issues.

    A sound bite based media has been a disaster for US politics.
    Nevertheless, in 2008, I went to my local precinct caucus and one of the most remarkable things that occurred was the discussion; the number of political issues that were discussed in a very thoughtful fashion was astonishing. Polls miss this kind of information richness IMVHO.

    It’s been my observation that people are far more concerned about college tuition and bailouts than they are about Iran and/or Isreal. None of the 20-and-30-somethings that I know care all that much about ‘age old, ancient grudges’.

  3. bob goodwin

    I think a good populist movement is periodically necessary to purge the rot out of government. But lets please remember that we do want our leaders to lead. Worldwide there was little public support for confronting the Germans at the end of the great depression.

    Containing rougue states is not for the faint-of-heart.

    1. attempter

      I don’t know who your “we” is, but it’s not the citizenry.

      1. The people no longer have “leaders” but alien tyrants. I don’t know who you and your “we” consider to be “our leaders”. I’m not being sarcastic there; I literally have no idea who you’re talking about. Certainly no one in politics, business, the military, the “law”, the media, academia…

      2. We (the real we) do not in fact need “leaders” at all. We can lead ourselves. We can rule and manage ourselves. this has already been historically proven (among especially non-Western peoples, for tens of thousands of years; in Western history, by among other examples the Paris Commune, the Italian syndicalists, the Spanish co-operatives), just as the absolute incompetence and malevolence of all elites has been defintively proven once and for all. Anybody who had any doubts can no longer have them since the crash and the Bailout.

      Those forever define the elites. That’s what “leadership” is, that’s all it is, that’s all it does.

    2. liberal

      “Containing rougue states is not for the faint-of-heart.”

      The idea that Iran is a rogue state on par with Germany on the eve of WWII is laughable.

      1. Chicken Little

        Laughable? I would have gone with spectacularly ignorant or, perhaps, monumentally moronic and myopic, but I guess I can live with laughable.

  4. GaryD

    Americans are anti-war? What a crock. Every hamfisted invasion of the past 40 years as been backed by a huge majority, and Iran won’t be any different. We just haven’t got the memo to support it yet.

    1. DownSouth

      Mass mobilizations went hand-in-hand with the advent of democracy. Much more on this can be found in Jonathan Schell’s The Unconquerable World.

      I just visited Washington DC for the first time, and in the Smithsonian Museum of American History saw that Sherman’s scorched earth march through Georgia was informed by his at-the-time radical belief that wars are conducted between peoples, and not between armies, as was the traditional view.

      It was Clauzwitz who spoke of the concept of total war that Sherman emabraced, which requires complete mass mobilization. Mass mobilzation is possible only with highly perfected propaganda machines, which are a by-product of democractic endeavors to sway and influence large masses of voters.

      The people of the United States, however, are hardly unique in being vulnerable to these sorts of propaganda campaigns.

      1. emca

        Clausewitz was a first-hand student of the Napoleonic Wars of the early 19th C. It was the armies of France under the direction of Napoleon Bonaparte which first resurfaced many of the concepts of modern warfare an initial served so nobly in thawing and upending ‘Coalitions’ organized by other European nations; coalitions designed to play upon the fear of a spreading revolutionary menace, but in reality were yet another European ploy to meddle in and advantage another country’s internal woes (aside, the search for a more conciliatory government did not begin U.S. adventures in Vietnam or Iraq).

        Among the tactics the Revolutionary forces championed: total warfare, the use of mass conscription to raise an army, an reform of the French military apparatus.

        The above mentioned Sherman is the best known example of the first, mass conscription (as you imply) go hand in hand with democracy, and the third is a little mentioned (unfortunately) tactic of any organization implied in democratization of government entities, i.e. which would include the removal of vested interest, cronyism, merit based on inherited, internal privilege, etc. by which all human enterprise tends to descend.

        Clausewitz by the way is a difficult read in his native German language let alone translation or interpretation (one can readily appreciate the enormity of the problem by attempting to read Nietzsche). On cursory examination, he appears primarily as advocate of dynamic, fluid, ever-evolving, quick-adapting methods for pursuing warfare. Probably a must read by military strategists, if you can figure out what he’s talking about.

        Jonathan Schell sounds to be a more approachable writer.

      2. Jack Parsons

        To discuss propaganda campaigns, it is easier to discuss grand public service efforts like littering (Keep America Beatiful), flu shots, etc. Easier because they step outside of politics and it is easier to see them clearly.

    2. LeeAnne

      “a huge majority” were shocked by 9/11 and hadn’t a clue other than the usual suspects.

      Most people understandably do not have the time or inclination to search for political news. TV controlled by the wrong people treats them like the sitting ducks they are for emotional messages whatever audacious idea comes to the minds of the psychopaths of which there are many in power positions like Dick Cheney and Rumsfeld, the king of flu vaccine investments and distribution control.

      These men are BIG war criminals; guilty of crimes against humanity; particularly against US, Iraq and Afghanistan humanity.

      Even after the propaganda that Saddam was THE responsible party for 9/11 had lightened up, friends of mine in academe were disheartened to find that relatives in the mid-west, by the time of the 2004 elections, still believed that Saddam was responsible.

    3. liberal

      “We just haven’t got the memo to support it yet.”

      Of course.

      The polling _is_ useful, because, as I believe Meersheimer or Walt said, support for Israel in the US is a mile wide and an inch deep.

      1. Tao Jonesing

        Clearly, I could have been more careful and precise in what I said.

        To the extent that a blog fact-checks and analyzes the news, it can, indeed, be closer to the truth than what we get from the new.

        On the other hand, to the extent that a blog uncritically accepts a news story as a starting point of analysis (which happens all the time regardless of ideological bent), then it is just the second derivative of the truth.

  5. GaryD

    Bob you moron. Do you seriously think that Iran (or Iraq or freaking Granada or bloody Afghanistan) is the same threat as the Nazis who had the entire Germany industrial base at their disposal? You’re a bloody idiot. Actually, you’re probably yet another corporate war whore who makes his living selling death and misery. Nice work if you can stand it, jerk.

    1. liberal

      “Actually, you’re probably yet another corporate war whore who makes his living selling death and misery.”

      No, he’s probably a neoconservative whore who places the putative (not actual, of course) interests of Israel above those of the US.

  6. Tao Jonesing

    Ah, Bernays and Lippmann, two of my favorite examples of the cynical aftermath of classical liberalism.

    If people want to get a real taste of what these men had to say, check out Bernays’ 1929 classic “Propaganda” (available free on the intertubes) and Lippmann’s 1937 “The Good Society” (the inspiration for neoliberalism). Lippmann’s “Public Opinion” (also available for free online and on the Kindle) is a masterful work that anticipates modern cognitive science, but the fact is that it remains hopeful compared to “The Good Society” (I don’t view “Public Opinion” in quite the same way as Yves does). Also check out Adam Curtis’s “Century of the Self” and “Pandora’s Box” documentaries, available on YouTube and Amazon.com.

    At this point, pretty much any “news” item that appears at any major media outlet should be viewed as an opinion piece. This means all polls and all “on this hand — on the other hand” news items. News is the first derivative of the truth. Blogs are the second derivative of the truth. Except in the few cases in which objective facts are actually provided, the only way to discern what is true is by comparing the magnitude and direction (the vector) of change relative to the asserted start point and stated desired end point. Simplifying, listen to what they say they are going to do and compare it to what they actually do, and don’t let your biases distract you from reality.

    1. attempter

      News is the first derivative of the truth. Blogs are the second derivative of the truth.

      By that you mean you think the MSM is likely to be closer to the truth than blogs? I’d say the opposite.

      While blogs have plenty of biases, they’re far more likely to be biased in favor of the truth than the corporate media, which we know is, as Jay Rosen put it, “viciously biased” against truth.

    2. Francois T

      “News is the first derivative of the truth. Blogs are the second derivative of the truth.”

      hmmm hmmm! Must be why a blog like fivethirtyeight.com nailed the 2008 elections so well compared to the “predictions” of the obsessive-compulsive mainstream media who wouldn’t let the facts interfere with their pathological need for horse races and keeping scores.

  7. Avg John

    ?? So which poll is the most reliable? I guess we will find out whats on people’s minds come November. For me, I get a basic idea of the political leanings or bias of a particular public news outlet, be it t.v., radio, or internet, and expect that any polls( or articles about polls) will support that bias, or attack it accordingly.

    Vote the rascals out. As many times as is necessary.

  8. Kevin de Bruxelles

    On building consent most people will support a government’s “information” campaign if they support the underlying policy being pushed. For example different people will complain about say a powerful propaganda campaign by the government in favour of the public option (which of course would never happen in real life) as opposed to those who complain about a publicity campaign to bomb Iran, for example.

    But for a potential attack on Iran I wouldn’t expect much a of an preparatory propaganda barrage (that is over and above the normal everyday level that we experience anyway) since surprise would be desired in any attack that supposedly will attempt to cripple a nuclear program. Besides, with the tiny exception of Ron Paul, the elites of both parties are overwhelmingly behind an attack on Iran; there is no need to provide anyone any political cover. Elites would expect the natural patriotism of the masses to bubble to the surface after a few days of video of the attacks, (along with some engaging story lines of women, minorities and maybe even some handicap people shown in an heroic light, no doubt) and these elites would expect the initial polls to be supportive.

    Also the intensity of the propaganda barrage surely has to have a relationship to dependence the elite have on the masses to execute the war. Obviously in WW1 and WW2 the elites needed huge numbers of their people to field their armies and to man the factories. Coercion is more expensive than persuasion so propaganda was used. In the run-up to Iraq there was a huge international propaganda campaign (which largely failed outside of the US) but since smaller armies are needed now the failure was not catastrophic. In the US once could imagine the motivation of the propaganda was to lessen any political uncertainty that may have arisen from the attacks (in other words Bush wanted a second term). Most of the international damage was fixed by first channelling the anger resulting from US actions in Iraq towards the singular person of George W. Bush. Then this “bad” person was replaced by the “good” Obama and everything was forgotten. On Iran there will be British and French participation from the get-go and the idea would be to get it over with before any widespread protests could take place.

    So far from there being any debate in the aftermath of an attack on Iran by the US on the wisdom of such actions; there will more likely be a discussion on whether “enough” was done. Liberals on D-Kos, etc. will strongly back Obama on the attack and highlight the international participation, while the Right will grumble about him being a second Jimmy Carter.

    1. doom

      Exactly. Debate will be confined to, “Is the president strong enough?” You won’t hear a peep about, “Would this war be a crime of aggression?” Now that aggression has been defined in law, manufacture of consent will have be accompanied by lots more arm-twisting in the UN, because some country needs to volunteer to veto a Security Council referral of the crime to the ICC. It will be interesting to see how the propaganda explains this process. The US press has blacked out the story of progressively stronger supranational restrictions on US action.

  9. Ex-Con John

    It’s not just the polls, it’s the biased reporting on the polls. The question will be something like ‘Would you support an attack on Iran if Iran attacked the U.S first?’ Most people would answer yes to this and it will be reported as ‘In latest poll, majority of Americans support war against Iran.’

    I find the reporting on the polls to be much worse than the biased questions. If the media reported the entire question as opposed to an implied, sensationalized version of the question, most people could figure out for themselves if the question was valid.

  10. attempter

    And I have noticed more than occasionally signs of bias in polling (unnecessarily value-laden phrases, strange either-or pairings that exclude other choices, leading question sequences). I generally take opinion polls when called out of curiosity but I can recall at least a couple of time aborting the call because the questions were very leading.

    That’s a pet peeve of mine, since out of curiosity I too sometimes think it might be fun to take political quizzes and such. But I always have to give up on them because the questions are so wrongly framed.

    By deceptive design no poll or quiz ever stipulates from what point of view it wants the respondent to treat the questions. Thus it seeks to further the fraud of an absolute, i.e. elite-enabling status quo, point of view.

    So I start looking at the questions and usually right from the start my answer in each case is, “it depends…”. Usually for most questions I have co-existing, often opposite answers depending on what point of view is the wellspring of the answer. Often I have to say to myself, “my real answer is this, but from the questionnaire’s point of view the “right” answer is probably the opposite..”

    It’s frustrating, and I largely gave up on even starting to take quizzes. (The last time I looked at the political compass quiz I got about five questions into it before I gave up. And that’s one of the better-conceived quizzes, having two political axes rather than just the one “left-right” line.)

    And that’s a quiz drawn up without any ulterior motive I can think of. Imagine these polls which are trying to generate an ideologically desired outcome.

    So if you’re going to answer one, you can’t give your answer in a vacuum, but you have to think of what your “correct” answer would be given the poll’s bias.

  11. Ignim Brites

    Polling is typically little more than bourgeois bolshevism. This is especially true of issues polling in contrast to to election horse race polling. Especially nefarious are those polls that break out results by “educational” level. As though the major problem with America is not the extraordinarily poor education received by our best and brightest.

  12. Digoenes

    The country sounds more right-wing because the right wing media frames the issues and right-leaning language in polls elicits more right-leaning answers than a neutral remark would.

    I agree. Indeed your correspondent provides a good example in support of this hypothesis in his quotation of this survey question:

    Americans would oppose the US militarily defending Israel if it were the victim of an unprovoked attack.

    Given the history of Israel’s interactions with the Palestinians over the past 30 or more years (to say nothing of their periodic incursions into their nominally sovereign neighbor to the north), I think the phrase “unprovoked attack” is virtually lacking in semantic content, unless of course you anticipate a rogue division of the Swiss Guard marching across the Golan into Israeli territory. The use of the word “unprovoked” shows very clearly the pre-existing bias of those who created the question.

    Finally, I wonder whether Gary D. thinks his fellow citizens would enthusiastically support a prolonged full bore war against the Islamic Republic of Iran if it required (as it quite probably would) the reinstatement of the military draft, heavy casualties, emergency war taxes, substantially higher oil prices and likely rationing.

  13. john

    How about the manufacture of dissent? Emails flying around, screaming about everything from imams to (ha!) “most hostile, anti-business administration ever” to hatred of white people to whatever else can be scared up. This country is an absolute fucking disaster.

  14. Mark

    I’ve never understood why anyone answers pollsters questions and yet many seem to regard it as their civic duty. Your information has value, your privacy has value, why would you give it away for free?

    Anyway, pollsters business people trying to remain in business, i.e., make a profit. They sell their services to willing buyers. If the buyer wants to use the poll for the purpose of publicizing the results to further the buyer’s agenda then the pollster will accommodate the buyer or risk going out of business.

    The notion of pollsters being some form of “truth-seeker” is as quaint as the notion of news organizations being objective purveyors of information rather than a part of the entertainment industry seeking to earn a profit.

    1. aet

      My polling firm would pay people to take part in some of our polls: and that was thirty years ago.
      perhaps you are not familiar with how very many kinds of research into public opinions, about all kinds of things, not necessarily political, goes on.

        1. Mark

          It seems to me that my clear meaning was that pollsters must oblige their clients. Note the “if” in following quote: “If the buyer wants to use the poll for the purpose of publicizing the results to further the buyer’s agenda then the pollster will accommodate the buyer or risk going out of business.”

          Was it really necessary to say the obvious: If a client wants an accurate picture of the public, be they consumers or voters, rather than a “talking point”, then a pollster must be able to provide as well.

          1. aet

            True, the use to which the sponsor of the work will put that work to effects the work.
            But polling is useful, far beyond the use of their results as a way to manipulate public opinion.
            After all, election day is a poll, too.

  15. Siggy

    “. . . anyone who has had any exposure knows that the results of questionnaires are quite sensitive to the wording of the questions”.

    You were expecting maybe people would answer something other than what you asked? Well, probably not, since the objective of the inquiry was a gaussian chart that demonstrated a central tendency toward a point of view that you were being asked to champion. The contemplated polemic would require some statistical support. And the polemic may well have had no relationship to reality.

    Now September is hurrying into October and it will soon be time for the World Serious and nothing is more serious than the World Serious. It is that grand event when the Boys of October, many of whom are closer to forty than to twenty, cavort and the spectating wannabees gorge on beer and chips and nachos and deep fried pigs ears and other such gastric abuses.

    There was a time when I was immortal. You are immortal until you turn thirty. Then you become very mortal and all manner of debilities begin to arise. Mostly you get fatter than you were. Now that extra weight is an extra load all over your body. It even preoccupies a great many of your brain cells. It’s a Gresham’s dynamic, overeating engenders too much brain activity to the business of keeping track of the needs of all those recently acquired extra fat cells. But then, we are, indeed, fast approaching the time of the World Serious.

    So, let us dispose of the this business of the seriousness of polls and like instruments of propaganda.

    Siggy’s first dictum: Don’t respond!

    Second dictum: Pay very little attention to them.

    Third dictum: Such attention as you give them, direct your focus to just what the poll appears to be supporting and does that point of view set right with the world as you know it.

    Fourth dictum, Understand that more often than not, the polster is the lackey of some charlatan who is envious and wanting of the contents of your purse, pocket or pocketbook.

    As to the manufacturing of consent, I thought that everyone understood that that was and is theater in progress. It’s a drama with an unending parade of players and scenes, all to very little purpose. After all we will soon have before us in that vast wasteland of televison, the World Serious. And a hat tip to you Newton N. Minow.

    And oh yes, Dear, Dear, Red Smith, would that you could rejoin us and tell us in your Pulitzer Prize winning prose, inning by inning just what transpired. But then the Herald Tribune is gone and the New York Times gives every appearance of limping off the stage as well. So, if you returned, where would you ply your wordsmith’s trade? Would it be here in cyberland, blogsville?

    Lets take a poll: Should Red Smith return to us via the blogoshere?

    What say you, one and all?

  16. JTM

    Health care reform is a good and rather frustrating example of selective and distorted reading of polls. By the Republicans for obvious reasons, but also by the press for less obvious reasons. But the result has been a manufacturing of opposition when in fact a majority favor the legislation, or want more.
    Consistently reform has been labeled “unpopular” because a majority or plurality have usually “disapproved” of the specific legislation as proposed and passed. Republicans tout this as supporting their position, i.e. no reform. But when a full breakdown has been reported, a majority has consistently either approved or wanted stronger reform. It is really the Republican position that is unpopular.
    The press has fallen into this I think largely in order to present a simple, bipolar (as it were) picture, when in fact it is a three-dimensional issue. The press is often accused of turning everything into a horse race. In fact they much prefer a prize fight. Just two sides, with maximum drama.
    So reporting of the true picture has been distorted through Republican tendentiousness and media laziness. And the more people hear that something is “unpopular”, the less popular it is likely to become.

  17. MarkJ

    I suggested (on Nate Silvers website fivethirtyeight.com) that a polling certification organization be formed that pollsters could subscribe to and be required to meet at least a minimum level of objectivity in the makeup of the poll and the interpretation of the poll’s results.

Comments are closed.