The Left's Big Blunder

The disastrously counter-productive strategy of Obama's supporters

October 15, 2008

[This essay now includes a post-election afterword at the bottom of the page.]

Two campaigns are being waged right now for the presidency of the United States. No, I'm not talking about the Obama campaign and the McCain campaign. I'm talking about the real-world campaign and the meta-campaign.

The real-world campaign involves speeches and proposals and facts and scandals and political positions and news events. These details, however, are becoming increasingly irrelevant, and have become subsumed by the meta-campaign, which consists of perceptions, polls, reactions, analyses and summations. Until very recently, elections were decided by real-world facts -- but not anymore. Facts and events in and of themselves are no longer important; what's important is how everyone reacts to them. And how do we find out the public's mood concerning this or that incident? Why, the media tells us, that's how.

Or so we've been led to believe.

We're all part of the campaign now. Every single one of us. Our opinions, our actions, are bundled together as a group and used as weapons in the race for the White House. When the media reports on what people think, either through public-opinion polling or reportage about anecdotal incidents, it becomes an endless feedback loop, in which the media's representation of most people's purported thoughts is supposed to influence everyone else's thoughts. And then they take another poll to determine how effective the first poll was in influencing public opinion, and the cycle starts all over again. Since everyone now knows that any public expression of their political opinions might be reported by the media, even the most innocent activity becomes a calculated campaign action. Saying how you intend to vote is not simply an expression of how you intend to vote, but rather a component of the public barometer of how the majority intends to vote, which is then used by the media and the blogs to influence everyone else. Nothing is done in all innocence anymore.

It's no longer a matter of dispute that the mainstream media, overall, very strongly leans to the left. Over 90% of journalists classify themselves politically as "liberal" to varying degrees, and innumerable instances of left-wing bias on the part of the media have been pointed out by bloggers over the years. Yes, a small subset of media outlets are identifiably conservative, but they are vastly outnumbered, both in sheer numbers and in influence, by the liberal media. This fact takes on intense importance in an era when the "news" becomes (as it has become) a subjective matter. Nearly any fact or incident can be "spun" to Obama's benefit.

Obama's supporters and his official campaign have taken great advantage of this felicitous informational landscape -- first, that the meta-campaign trumps reality, and second, that the media is cooperative and complicit. For example, after presidential debates, the leading left-wing blogs always coordinate massive online opinion-poll-stuffing campaigns. After the Palin-Biden vice-presidential debate, the overwhelming consensus on conservative and centrist blogs was that Palin had won handily, and that Biden spoke mostly in a soporific monotone while spewing a continuous stream of easily debunked falsehoods. And yet readers of DailyKos, the Huffington Post, Democratic Underground and dozens of other top left-wing blogs swarmed en masse to vote (often repeatedly) in mainstream online polls about the debate, so that afterward, CNN (among many others) could run headlines that said "57% Think Biden Won Debate," basing their conclusion on the results of the online polls. And once enough of these articles get published, then they themselves become "proof" of the debate's supposed outcome, and before long (often just a matter of hours) it becomes a "fact" no longer up for discussion that Biden won the debate. This fact is then referenced by pundits, and slips into supposedly neutral news stories.

The illusory quest for conformist decision-making in the 2008 presidential election

Say, for example, that you were an Obama supporter who watched the Vice Presidential debate and felt that Palin had done well and was a more effective debater than Biden -- though not well enough to change your mind about voting for the Obama-Biden ticket. Immediately afterward you encounter an online poll, asking you to vote on who won the debate. What do you do? I suspect that most, if not nearly all, Obama supporters would lie and still vote in the poll that Biden had won the debate, even though they felt that Palin had in fact defeated him. But why do so? As an Obama supporter, you still want your candidate to win, so that every action you take should revolve around only one question: Will this help Obama win? Probably unconsciously, people assume without really thinking about why, that if enough people say Biden won the debate, then a general consensus will be reached that Biden did win the debate, and as a result some vague category of Americans who up until that point had not been Obama supporters will change their allegiances and in reality vote for Obama-Biden on election day.

The Obama campaign itself also takes advantage of the sympathetic media to construct a facade of inevitability. The campaign will stage-manage crowds and dictate camera angles so that Obama is seen to not only have overwhelming numbers of fans but the correct demographic proportion of fans; the campaign will coordinate Obama appearances to coincide with rock concerts or other festivals so they can point to the huge crowds who showed up to watch Obama; and the media plays right along.

McCain supporters often complain about this strategy by the Left, going to great pains to point out the poll stuffing, the deceptive photos, the crowd overestimation, the slanted media coverage, and so forth. But should conservatives be so concerned? I propose that McCain supporters should be GLAD this is happening -- because the Left is in fact making a disastrous strategic blunder.

A substantial portion of the Left's strategy during this campaign is to create the perception that as many people as possible are supporting Obama. They strive to not simply show that he has a lot of supporters (which, obviously, he has), but to purposely inflate or exaggerate the numbers in order to make his support seem larger than it really is. The drive to do this seems almost automatic; it is assumed by Obama's supporters to be the most effective campaign strategy. It's so automatic that they perhaps are no longer even aware that it is a strategy. But why? What purpose is possibly served by this behavior? Has anyone on the Left ever paused, stepped back, and asked, "Wait a minute -- why are we doing this? Are we sure it's the correct course of action?" Doing everything possible to inflate the perceived support of Democratic candidates has become so de rigueur that the Left has long ago forgotten why they're even doing it.

This essay examines the underlying faulty assumptions of this strategy -- and shows why it's not only counter-productive, but could backfire disastrously.

Who Does the Polling?

A key component of this strategy is an over-reliance on polling, since poll numbers which show Obama apparently in the lead can be used to club undecided voters or McCain supporters into submission. You're all alone. Nobody else thinks like you. Your side is losing. You're out of touch. Change your mind -- join the winning team. But the polls may not reflect what we imagine they reflect.

The actual grunt-work of doing public-opinion polling is a low-paying job that doesn't require much (or any) experience. The people asking the "Who will you vote for?" questions either sit in a room for hours on end making repetitive calls, or walk door-to-door in potentially dodgy neighborhoods -- not particularly high-end work. Who would apply for and accept such a job? Not someone who already has a higher-paying job. Not someone who has experience or skills that would allow them a better career or position. Not someone who is middle-class or middle-aged. Not someone who is wealthy. Generally, I wager, college students, the under-employeed and/or the under-educated make up the majority of people who do the actual rubber-meets-the-road polling of asking the questions and writing down the answers. Though there are no published statistics about the demoographics of pollers themselves, the fact that, for example, The Gallup Poll routinely recruits entry-level call-center employees on college campuses lends support to this supposition. And, anecdotally, when I was a college student I briefly worked for a public-opinion call center doing this exact kind of job, and my fellow pollers were almost all college students or underemployed folks, some of whom only worked a limited number of hours per week so as to still qualify for government benefits.

Now, as we all know, college students (especially these days) are usually at the apex of left-wing sentiment in their lives' political journey. As has been documented extensively, people tend to get more conservative as they get older and acquire more responsibilities. And contemporary college students are almost all being influenced toward leftist ideologies by the liberal atmosphere on most campuses. Some will later grow out of it and slowly shift to the center or right, but while they're in college, they're in the thrall of their doctrinaire professors. The Obama campaign itself brags that they have the under-25 demographic locked up.

Furthermore, chronically underemployed or unemployed people also tend to support left-leaning and populist candidates -- especially candidates like Obama who are promising free health care, more government benefits, job programs in poor areas, and so on. This is not a controversial supposition: it's also been documented extensively that wealthy people tend to support Republicans who promise to lower taxes, and poorer people tend to support Democrats who promise more handouts.

So, I posit that the vast majority of people actually doing the polling are themselves Obama supporters. Not necessarily the CEOs who own the polling companies, but the people actually picking up the phone and making the calls, or walking door-to-door.

This fact has potentially significant implications for the outcomes of polls. Imagine, for a moment, that you were one of the rare McCain supporters in a polling call-center; wouldn't you be a little depressed if person after person you called stated that they were intending to vote for Obama? How could you not be? But what if you were an Obama supporter working in that same call center? Wouldn't you be elated or enthused to hear the votes for Obama piling up? Of course you would. Ah, but pollers are under strict instructions to not reveal their personal opinions to the people they're polling. And I'm sure that most try to follow the rules. But even if you grant that, say, 90% of them manage to maintain complete neutrality, and not let some kind of expectation creep into their voice or attitude, that still leaves 10% who might consciously or unconsciously be slanting the results. And that's all it would take to screw up a poll. Even a 95% honesty rate leaves room for 5% bias in the results, which can be very significant in a close race like this one.

Yet I'm not concerned about the 10% or 5% of pollers who consciously allow their personal opinion to slip into the conversation. I'm concerned about the 90%+ who try to appear impartial. The operative word here is "try" -- because as we learned from an amazing horse (yes, a horse), "trying" isn't good enough.

The Clever Hans Effect

  Clever Hans giving a typical performance to a large crowd, watching carefully for the subtle body-language cues
  which told him when to stop stomping.
What in the world does a performing horse from early 20th-century Germany have to do with the 2008 presidential election? As it turns out: everything.

Clever Hans was a horse in turn-of-the-century Germany who was apparently able to perform mathematical calculations and all sorts of intellectual feats. According to the Wikipedia Clever Hans page,
Hans was a horse owned by a Mr. von Osten, who was a high school math teacher, an amateur horse trainer and phrenologist, and something of a mystic. Hans was taught to add, subtract, multiply, divide, work with fractions, tell time, keep track of the calendar, differentiate musical tones, and read, spell, and understand German. Von Osten would ask Hans, "If the eighth day of the month comes on a Tuesday, what is the date of the following Friday?" Hans would answer by tapping his foot. Questions could be asked both orally, and in written form. Von Osten exhibited Hans throughout Germany, and never charged admission.
After a government investigation failed to discover any fraud in Hans' performances, a psychologist named Oskar Pfungst performed a series of tests on Hans and discovered something astonishing. First off, he proved that Hans was not in fact making the calculations, nor did he understand German, but rather was simply responding to body language and other subtle clues provided by his human trainers. Hans indicated his numerical answers by stomping his foot the appropriate number of times. So for example, if you asked Hans what 2 + 3 was, he'd stomp five times -- and then stop. But it turned out Hans only stopped stomping when he noticed that the humans surrounding him (often there was a crowd at his performances) expected or wanted him to stop stomping. By controlling what Hans could and could not see, and by asking questions to which the humans did themselves not know the answers, Pfungst proved that Hans was literally a one-trick pony: he only knew that he was supposed to start stomping when a human spoke to him, and then stop stomping when the humans exhibited relief or happiness or a relaxation of tension (which they did when he reached the right number of stomps).

But during his investigations, Pfungst discovered something even more significant: The Hans performances were not intentionally fraudulent, because the trainer (and the audience) were exhibiting the body language unconsciously. But that was only the half of it: when Pfungst himself asked Hans to solve problems, even when he intentionally tried to suppress his own subtle visual and physical expressions which might indicate when to stop stomping, Hans still was able to pick up on them somehow and give the right answer. Again, according to Wikipedia,
After formal investigation in 1907, psychologist Oskar Pfungst demonstrated that the horse was not actually performing these mental tasks, but was watching the reaction of his human observers. Pfungst discovered this artifact in the research methodology, wherein the horse was responding directly to involuntary cues in the body language of the human trainer, who had the faculties to solve each problem. The trainer was entirely unaware that he was providing such cues.
Pfungst made an extremely significant observation. After he had become adept at giving Hans performances himself, and fully aware of the subtle cues which made them possible, he discovered that he would produce these cues involuntarily regardless of whether he wished to exhibit or suppress them. This gives the phenomenon an importance which could hardly be exaggerated. Its recognition has had a large effect on experimental design and methodology for all experiments whatsoever involving sentient subjects (including humans).
  Clever Hans correctly performing mathematical calculations
  in a laboratory setting -- still managing to pick up seemingly
  imperceptible cues from the unwitting experimenter.
This phenomenon is now called "The Clever Hans Effect": the attitude of questioners can affect respondents' answers even if the questioner is trying to remain neutral. If a horse can notice subtle verbal or visual cues, then a person -- who is much more attuned to human culture and emotions -- can probably notice them even more so (possibly also without even realizing it).

So: What's the connection to the 2008 presidential election? Well, much of the media analysis, and even the strategies of the campaigns themselves, is based on the ongoing poll results indicating voter preferences state-by-state and nationwide. But I suspect that we are observing the Clever Hans Effect on a massive scale, and that the polls are in fact unreliable. Worse than "unreliable," actually: they are inaccurate because to some degree they reflect not the honest feelings of the respondents but rather what the pollers want to hear. Since, as discussed above, most poll-questioners are likely to be Obama supporters, and since the Clever Hans Effect tells us that they likely slant their questions and/or provide subtle clues as to what the "correct" answer is whether or not they're trying to be neutral and fair, the end result is that the poll results end up being tilted in favor of Obama. Pundits and journalists and campign directors are deriving supposed "information" from the poll results, and basing their actions on them -- even though the polls merely reflect (to a certain degree) what the pollsters wanted to hear. To a certain degree, contemporary polling is one vast demonstration of the Clever Hans Effect. And the end result, once again, is exaggerated support for Obama.

Hang on just a minute, you protest. The hitch is in that phrase "to a certain degree." Even if everything you say is true, I find it hard to believe that anyone would lie to a pollster simply because they detected that the poller preferred a certain answer over another. That's a valid challenge. Yes, it would only take 5% or 10% of respondents being swayed this way to throw a poll off, but seriously, would it ever happen in real life? I mean, c'mon, who would tell a pollster that they intended to vote for Obama when in fact they were undecided, or were actually going to vote for McCain? Is anyone that sensitive to social expectations?

I've got two words for you. And those words are "Solomon Asch."

Conformity and the Asch Experiments

  Solomon Asch (on the far right) and six assistants trick the test subject (third from the right
  with a white shirt, tie, and crew cut) into thinking he's taking a vision test.
Implicit in the Left's continuous attempts to exaggerate Obama's perceived support is the belief that "a crowd draws a crowd" and that undecided voters will be drawn to the Obama camp if they think "everyone else" is supporting him. But is that an accurate assessment? Is there any evidence that it's true?

Well, actually, yes. There is evidence. Or should I use scare quotes: "evidence." I believe that a series of experiments carried out in the 1950s by social psychologist Solomon Asch were in fact the (now long-forgotten) inspiration for and justification for the current strategy, especially the strategy of online opinion-poll-stuffing after debates and other major news events.

Starting in 1951, Asch, a professor at Swarthmore College, ran a series of unusual experiments to generate a quantitative measurement of the subjective term "conformity." The experiments, which many now consider somewhat unethical and a bit sadistic, went like this:

A volunteer was recruited to participate in a vision test. He was brought to a room with seven other volunteers who were also to take the same test, in a group. Little did the volunteer know, however, that his fellow "volunteers" were all confederates of the experimenter, and the test was not a vision test but a psychological torture session designed to elicit conformist behavior. The experimenter would then unveil a pair of displays, one showing a single black line, and the other showing three black lines of varying lengths. The volunteer is told to simply state which of the three lines most closely matches the length of the single line.

  Which line on the right -- A, B or C -- matches the length of the line on the left?
  In the privacy or your own home and your own mind, you'd almost certainly
  answer "A." But if you were in a large group of people who all insisted that "C"
  was the answer, you might very likely answer "C" as well -- either because you
  feared being rejected socially by the group, or because you doubted your own
  perceptions and deferred to the opinions of others.
Sounds simple, right? In every instance, the correct answer was quite self-evident, as only one of the lines was even close to the correct length. The volunteer, who was always placed in the second-to-last position, was only allowed to state his answer after he had heard most of the other faux-volunteers give their answers. For the first two rounds, these confederates were instructed to give the obviously correct answer; in each instance, the test subject would then also give the correct answer. But starting on the third round, the confederates, as instructed by Asch, intentionally gave a consistently wrong answer; the goal of the experiment was to see if the volunteer would "break" and also begin to chime in with the wrong answer as well. Most volunteers would resist for a few rounds, but eventually the majority would cave in at least part of the time and give the wrong answers in complete defiance of their own perceptions. Overall, the test subjects gave the wrong answers 36.8% of the time -- an astonishing result.

What would you do if you had the ability to conduct this "experiment" on a vast scale? And if the results of the experiment were not just of academic interest, but affected the real world? What would you do if you had a monopoly on the media, and could affect each individual's perception of how the general public felt? You could take the Asch experiment nationwide. You could deceive every single individual voter into thinking he was all alone in his opinions. And consequently, due to social pressures to conform, they'd change their allegiances. You could use it to win elections.

That's the position in which the Left -- the Obama campaign, its supporters, and the liberal media -- imagine themselves to be. They're trying to use the principle of behaviorial conformity as a weapon in the campaign. But there is a terrible flaw in their plan. The Asch experiment doesn't work unless the test subject is unaware that he is being duped. And I'm telling the subject right now: you're being duped.

Normative and Informational Conformity

In post-experiment interviews, during which Asch revealed the ruse, the test subjects gave one or the other of two completely different reasons for agreeing with the wrong answers. One personality type said that, although they were fully aware what the correct answer was and that everyone else was giving the wrong answer, they themselves repeated the wrong answer publicly because they "didn't want to go against the grain" or to appear like a freak or an outsider, or be rejected by their peers. This attitude was given the unwieldy moniker "normative conformity", in which "a person publicly accepts the views of a group but privately rejects them." The other personality type doubted their own perceptions, and assumed that if an entire group of people thought Line "C" was the correct answer, then it must in fact be the correct answer, despite the subject's own first impression that it was the wrong answer. This type of person actually changed their opinion based on the group's apparent consensus, and would have answered "C" even if allowed to do so anonymously. This attitude was dubbed "informational conformity," in which "a person accepts the views of the group" as actually valid and adopts them internally.

  Faced with unanimous agreement by the group on the wrong answer, the test subject (#6) peers at the
  cards, deciding whether to trust his own eyes and his own confidence, or whether to instead change his
  mind and trust the group's opinion, or to simply pretend to agree with the group in order to avoid conflict
  or rejection.
The seemingly minor and esoteric difference between normative conformity and informational conformity is actually not esoteric at all -- it is the key factor which defines this campaign and this election. Because it all boils down to this: Obama supporters presume that increasing Obama's perceived support will induce informational conformity in the American public. In other words, Obama supporters operate on the assumption that individual McCain supporters or undecided voters will in actuality change their minds about who to vote for if they perceive that a majority of people are supporting Obama. The imagined line of thinking is, "Gee, if so many people like this Obama guy, then my impression of him must be wrong; I trust the group's wisdom more than my own impressions."

I submit that this assumption is a catastrophic blunder. To the extent that there is any conformist behavior being exhibited by McCain supporters and undecided voters, it is much more likely to be normative conformity. In other words, people who are confronted with apparent overwhelming support for Obama may indeed announce that they too support Obama, but do so only in order to avoid ostracism or accusations of racism. Inside, however, they have not changed their minds. On November 4, they will go into that voting booth, and in total privacy and anonymity, they are free to vote for whomever they want, without fear of social condemnation for doing so. And in such a setting, normative conformity disintegrates, because there is no "norm" to conform to when your vote is anonymous.

If the American voting system was like an Asch experiment, in which your individual vote was displayed on an electronic tote board for all to see above your voting booth when you pressed the button for McCain, then, yes, normative conformity would in fact affect the election; but our anonymous-voting system frees people to vote according to their innermost convictions. The one important fact of the Asch experiment is usually glossed over in the descriptions of it: "most" of the subjects who gave the wrong answers did so out of normative conformity: "When they were interviewed after the experiment, most of them said that they did not really believe their conforming answers, but had gone along with the group for fear of being ridiculed or thought 'peculiar.' A few of them said that they really did believe the group's answers were correct."

  A re-enactment from a documentary showing a realistic simulation of what the
  Asch experiment is like.
Another fundamental flaw in the Left's presumption that informational conformity will apply during this election season is that, as revealed in this good summation of the Asch experiment, the pressure to conform drops precipitously if the subject is aware of even a single fellow dissenter. All it takes is one person to shatter the facade of unanimity, and suddenly the number of conformist answers drop from around 33% to around 8%. With more dissenters, it drops even further. This is crucial, because despite the overwhelming left-wing bias in the media, there isn't an impenetrable wall of pro-Obama propaganda.

The Left's blunder is thinking that they have a monopoly on the entire infosphere. Yes, liberal ideologues dominate the media, but they do not hold a monopoly. Dissenting voices can be heard. But more importantly, the mediasphere is not the entirety of the infosphere. There are many alternate sources of information and opinion -- most significant among them being the political blogosphere, which is about evenly divided between Left and Right, and which has grown to be nearly as important as the media itself. All it takes is five minutes online for anyone to discover that they have millions of cohorts who think exactly like they do.

While conservative-leaning media outlets are in the minority, they certainly exist, and if one is so inclined one has a wide variety of conservative blogs/radio programs/TV shows/newspapers/writers from which to chose. And even though such outlets have less access to the "mainstream" of public discourse -- ridiculous rumors that crop up on left-wing blogs are almost immediately picked up and disseminated by mainstream media sources, whereas conservative blogs have to publish a real scoop with solid factual evidence if they want to get any media attention -- there remains a continuously audible undercurrent of dissent. So, while the Left might (unconsciously) imagine they have the ability to pull an Asch on the whole country, the deception fails to work, because there is always a Rush Limbaugh or a Little Green Footballs or a Fox News or a Michael Savage whispering in the subject's ear, "It's all a trick! Don't believe the lies!" And under those conditions, the level of informational conformity would almost certainly drop to near zero.

And there's another reason why the Asch experiment could never work on a nationwide scale. When these experiments were first conducted, the subjects were all male college students, in the 1950s. The pressures to conform during that era and in that social setting were much greater than now. The whole reason Asch was testing for conformity in the first place was that the 1950s were perceived (and are still perceived) as the era of conformity. Yet that era has long passed. Nowadays, everyone wants to be a nonconformist. And even though there remains an innate human urge to conform, and even though despite our illusions of individuality we sometimes end up conforming anyway, in the 21st century people at least think that nonconformity is hip. So that the group psychological dynamic is now likely to be the opposite of what it was half a century ago.

(For those interested, one of Solomon Asch's original papers can be read here.)

The Race Card and the Revival of the Bradley Effect

The Bradley Effect, a polling phenomenon which has been discussed extensively in this election, is
a discrepancy between voter opinion polls and election outcomes in American political campaigns when a white candidate and a non-white candidate run against each other.... The Bradley effect refers to an alleged tendency on the part of some voters to tell pollsters that they are undecided or likely to vote for a black candidate, and yet, on election day, vote for his/her white opponent.
Despite being a well-known phrase, the Bradley Effect is quite often misreported by the media and misunderstood by the public to mean that whites who are racist will refuse to vote for any black candidate yet will lie to pollsters about their intentions for fear of having their racist attitudes exposed. As a result, polls sometimes over-report support for black candidates in elections when they are running against white candidates. But this is a gross misapprehension of what the term means. First of all, the phrase "Bradley Effect" originally only referred to a bare-bones description of what actually happens in such races: White voters tell pollsters they intend to vote for the black candidate, but on election day they either vote for the white candidate or don't vote at all. Left out of this original definition was any notion of why this happened. But over the intervening decades since the effect was first noticed (when black Los Angeles mayor Tom Bradley lost the 1982 California gubernatorial election to lesser-known white candidate George Deukmajian, despite Bradley apparently having a substantial lead in the polls), additional layers were added to the definition in which the motivation was assumed. The causes are in fact not so clear, and are impossible to study directly. While the media and the general public often assume that the Bradley Effect is caused by actual racism, some astute analysts point out the real cause is more likely to be something much more subtle: That white voters who are in fact not racist will pretend to support a black candidate due to fear of being falsely perceived as racist.

The Bradley Effect seems to have gradually evaporated over the intervening decades since 1982, as black politicians became more commonplace and it was no longer automatically assumed that if you voted against a black candidate you did so only out of racism. That is -- until this election. Until this year, accusing your ideological opponents of racism -- a.k.a. "playing the race card" -- was for a while a taboo strategy, which only served to highlight that your campaign was becoming desperate and had no other valid lines of attack. But as election day 2008 draws near, accusations of racism have escalated exponentially, and now it seems the majority of pro-Obama pundits, journalists and bloggers routinely state as fact that all McCain supporters are racists who refuse to vote for Obama simply because he is black (and not because of his policies). The situation is even more extreme in social interactions in liberal areas, where in casual conversation the race card is played almost continuously. I live in the San Francisco area, in an artsy/intellectual/academic circle, and never once have I heard anyone professing support for McCain. If your boss mocks McCain supporters, if all your co-workers express a desire to for Palin to be raped on national TV, if your family are all Obama volunteers, if the media tries to shame everybody into voting for Obama by stating implicitly and explicitly that only a racist would do otherwise, could you have the nerve to come out of the closet as a McCain voter?

In such an environment, where admitting to disliking Obama in the interpersonal sphere has become the equivalent of social suicide, it seems very likely that the Bradley Effect is not just back, but back with a vengeance. The more that Obama supporters go unchallenged in their blanket accusations of racism against McCain supporters, the less likely anyone will publicly admit to dislike of Obama. Hence, the Bradley Effect is not an artifact of racism, but rather an artifact of false accusations of racism.

So, when the phone rings and the pollster calls -- and your Clever Hans social antennae tell you the pollster is young and liberal and likely an Obama supporter -- would you have the nerve to tell the pollster the truth that you wouldn't vote for Obama in a million years? I mean, they called you; they know your number. They know who you are. Can you be absolutely sure they aren't putting a check mark in the "Racist" box next to your name in some mysterious database?

Polls: Snapshot or self-confirming prophecy?

One odd thing about public-opinion polls is that there's no way to know if they're accurate or not. Except for a poll taken on the very last day of the campaign, when it can be later compared to the actual vote totals, a poll is a self-supporting statement of "fact" that can only be confirmed or disproven by taking yet another poll -- which is just as unreliable as the first one. We do not have access to some secret hyper-accurate invasion of privacy enabling us to peer into voters' hearts to see how they actually intend to vote, and to use that information to assess the accuracy of a poll. So, if a poll is taken a month ahead of time showing a candidate with a five-point lead, and then a month later he in fact wins the election by five points, we have no way of knowing whether or not the poll was simply accurate, or whether it was originally inaccurate, but fed a public perception that the candidate was in the lead, causing many voters to switch allegiances to him out of a desire to "be on the winning team." Do polls reflect reality, or do they create reality?

The entire Democratic strategy in 2008 revolves around the unproven theory that polls do create reality. Otherwise, there would be no point in continuously striving to inflate Obama's perceived public support.

The real question at the end of the day is this: Are people telling pollsters they're supporting Obama due to normative conformity (which is what I suspect) or due to informational conformity (which is what the Left is banking on)? We won't know until November 4. You can lie to a pollster. But you cannot lie to a ballot.

There are other issues which can, and do, seriously affect polling results: what time of day calls are made, how the questions are phrased, how the "internals" (the raw data) are doctored, and so on. (You may be surprised to learn, for example, that the poll numbers you see are rarely if ever the raw totals summarizing the responses of 1,000 people called at random; the percentages of Democrats and Republicans are adjusted to match voter registration rates, totals are fiddled with to include or exclude people who say they are "somewhat likely" or "likely" or "unlikely" to vote, and so on.) But the unreliability of polls overall is a different field of study which others have covered in detail and which is beyond the scope of this essay.

Bluffing When You Have to Show Your Hand: The Worst Possible Strategy

In the majority of strategy games, the most effective plan is to trick your opponent into underestimating you. The goal is to catch them unawares, to spring an attack when they least expect it. But there are exceptions to this rule: in poker, and sometimes in warfare, the most effective strategy can be the reverse -- to bluff, which means to trick your opponent into overestimating your strength.

There's a catch, however: bluffing only works when the rules of the game allow you to avoid a showdown with your opponent. Imagine, for example, that you were playing a hand of high stakes poker against a skilled opponent. You look at your cards and you see that your hand is not particularly good. Under these circumstances, some professional poker players will bet everything they have, and act confident, in hope of bluffing their opponents into folding. When the opponent folds, there is no final showdown; the player who was bluffing collects the winnings without having to reveal his cards, or, even if he does reveal them, doesn't actually need to have the better hand. The opponents gave up because they couldn't take the pressure. But what if the rules of this particular game of poker were changed so that every hand necessarily ended in a showdown: it was not possible to fold, and no matter how much you bluffed, and how much your opponent feared the strength of your hand, you still have to beat his hand at the end of betting. Under these circumstances, bluffing is not only ineffective, it's absolutely foolish, because you're risking everything on a group of cards that are not likely to win.

Similarly, in ancient warfare, if you could convince your opponent that your army was bigger than it really was, you could possibly trick them into surrendering without a fight -- a fight you might have actually lost if you had gone to battle.

Yet presidential elections are not like poker or siege warfare. There is no way to bluff your way to victory. The McCain campaign can't "fold" or surrender, nor can the actual voting on November 4 be cancelled. No matter how much the Obama supporters inflate their apparent strength, at the end of the day they're going to have prove it's true -- or lose the election. Because of this, bluffing in a presidential election is the worst possible strategy, because all you'll end up achieving is to inspire the opposing camp to fight more desperately, since they'll assume their backs are against the wall. Teddy Roosevelt advised that the best political posture is to "Speak softly and carry a big stick" -- but in 2008 the Democrats are yelling at the top of their lungs and carrying only a medium-sized stick.

In 1968, Richard Nixon did the exact reverse of what the Democrats are now doing. Instead of announcing bombastically that he and the Republicans had complete domination of the media and the electorate, his key campaign slogan was "The Silent Majority." Nixon knew full well that the atmosphere of the times made the Republican prospects looks pretty bleak. Youth culture and the social revolution of the 1960s were at a peak; left-wing ideology was soaring in popularity, a huge demographic of young post-war Baby-Boomers had just reached voting age and were thought to loathe the old-fashioned Republicanism; and the media had long had a serious vendetta against Nixon. Everything seemed aligned to ensure a handy victory for Humphrey and the Democrats. But Nixon sensed that behind the media frenzy about hippies and riots and drugs and revolution, there was, literally, a "silent majority" of staid conservative voters whose voices and viewpoints were being ignored. And all he needed to do was to reassure those voters that they existed across the country, even if the media and popular culture ignored them. And, to everyone's astonishment, Richard Nixon, who was considered by many to be a laughingstock, the last holdout of a pathetic dying breed of old fogeys, managed to pull out a razor-thin victory in what seemed like the most hostile possible social environment.

In 2008 there is no silent majority: there is the silenced majority. The unpolled majority. The media is so pro-Obama that the views and the concerns of McCain supporters are for the most part ignored or, at best, mocked. The goal is to foster disillusionment among them, a sense of isolation. To trick the Republicans into all staying home on election day because "there's no hope of winning." Maybe the Democrats can't avoid a showdown on November 4, but if they can convince enough McCain supporters to individually "fold" and not vote at all, then Obama can carry the day.

One of the sources of our current dilemma is that it's no longer possible to tailor different messages to different audiences. In days gone by, leaders could say one thing to one group, and then something else entirely to another group or to the public at large, in order to serve their political purposes. Yasser Arafat, for example, famously would say something conciliatory about the Israel-Palestinian conflict while speaking English to the international press, and after the reporters had all left to file their stories about what a peace-loving moderate he was, would turn around and give a fiery speech in Arabic to his supporters, calling for the extermination of Israelis. Thus, he pleased both sides, well enough to earn him a Nobel Peace Prize from his naive Western audience while simultaneously being hailed as the leader of the jihad against the Jews by his Arab audience. This trick worked quite well for many years until the advent of the Internet, when for the first time the Western audience could finally hear what he was saying when they weren't supposed to be listening.

A similar problem bedevils most politicians in the modern era, especially during any election season that has both a primary and then a general election. Obama for a year clearly spelled out his far-left position when addressing the far-left wing of the Democratic Party, using their support to take the nomination from Hillary Clinton. But now that we're in the general election, Obama is furiously backpedaling at top speed, denying or modifying or attempting to spin not just his primary positions, but a lifetime of far-left activism -- because those positions won't fly with the general public.


Now, it could very well be that, after all is said and done, Obama will indeed win this election -- I can't predict the future any better than can anyone else. The Obama campaign and its supporters are also engaging in many other strategies (unrelated to the exaggeration of his popularity) that have likely been effective -- such as blanketing the airwaves with advertisements, disparaging McCain, insulting Palin, and so on. The unabashed and unapologetic Obama boosterism from the traditional media certainly isn't hurting either. In prior elections, candidates worried about an "October Surprise," some last-minute revelation or scandal that threatens to realign the entire race. But in 2008, two or three October Surprises seem to be cropping up every single day, and there's no reliable way to predict what will happen next (other than that the media will try to emphasize the anti-McCain news and downplay the anti-Obama news). And it may be that less than 50% of the population was ever interested in voting for McCain in the first place, and that an Obama victory was a foregone conclusion long before the campaign even began; I simply don't know. However, if Obama does win, it will be IN SPITE OF the counter-productive antics of his supporters, not because of them. I feel that all the exaggerations and bias polling and online poll-stuffing and comment-spamming have only served to increase a desperate come-from-behind energy in the McCain campaign, and induce a sense of complacency and inevitable victory among rank-and-file Obama voters. However: If McCain wins, then Obama's supporters will only have themselves to blame.

Will the exaggerations become a self-fulfilling prophecy, as assumed, or are Obama supporters spinning further and further away from reality, constructing one unsupportable exaggeration on top of another -- only to be stunned on election day when the actual results, once again, don't match either their pre-vote opinion polling or their post-vote exit polling?

Yet it may very well be that an army of glum, dispirited and pessimistic conservatives will reluctantly trudge to the polls on November 4, each one imagining they are the only remaining person in the entire country voting for McCain, and lo and behold -- they'll turn out to be a silent majority after all.

Post-Election Update

As I stated above, "One odd thing about public-opinion polls is that there's no way to know if they're accurate or not. Except for a poll taken on the very last day of the campaign, when it can be later compared to the actual vote totals...." So, now that we have the results -- how did things turn out after all?

Well, as everybody now knows, the final result of the election was a 6.5-point victory by Obama over McCain. But that doesn't mean this essay was entirely inaccurate.

Because, exactly as I speculated, most of the major polls did in fact overstate Obama's lead in their final election-day predictions. On the last day of polling, the Gallup Poll claimed Obama was going to win by 11 points; Reuters also gave Obama an 11-point lead on November 4; CBS also predicted a 9-point victory for Obama on the last day (and a 13-point lead the day before); the Washington Post's final poll was +9 for Obama as well; The Marist Poll confirmed the +9 results; and the IBD/TIPP poll had Obama at +7.2. Overall, the consensus was very strongly tended toward at least a 9-point or greater Obama victory, since all these polls were taken on the very last day before voting. (Among major pollsters, only Rasmussen (+6 for Obama) and Fox (+7 for Obama) were accurate. A few minor polls had outlying results, but on average they also somewhat overestimated Obama's lead.)

What this means is: The effects described in this essay very likely did happen as I postulated, but not to a large enough extent to overcome Obama's actual strength and McCain's actual weakness. In other words, approximately 3% of people responding to polls did lie and say they supported Obama when in fact they did not (a ~9.5% predicted victory on average vs. a 6.5% actual victory). It's just that McCain was not close enough in real support for the Hans/Asch/Bradley Effect to make the difference.

Bear in mind that nowhere in this essay did I predict a McCain victory; in fact, of all the statements above, the one I think that ended up being the most accurate in describing this election was, "And it may be that less than 50% of the population was ever interested in voting for McCain in the first place, and that an Obama victory was a foregone conclusion long before the campaign even began...". Despite everything else that happened, McCain never came anywhere even close to 50% support in nationwide polls, and with tepid enthusiasm like that, there's simply no way you're going to win an election.

Although the election is now over, and this essay has therefore become a moot point, I'll leave it online for analysts and historians to ponder and discuss when picking through the wreckage of the 2008 campaign.

(If you wish to comment on this essay, you can do so here.)

(Click here to return to the main zombietime page.)