As they pass through those public schools I spent most of my childhood and adolescence avoiding, normies are taught that polling by outfits like Gallup and Pew Research Center can tell us what the public thinks. But what if there is no "public opinion" to measure in the first place?
This sounds crazy at first. We're bombarded with poll results, especially during election seasons. The news is full of headlines like "68% of voters disapprove of the president's handling of X" or "Candidate Y leads by 3 points in key swing state." These numbers seem solid, scientific even. But they're built on a shaky foundation.
Consider the state of the presidential race: Joe Biden is out, Kamala Harris is in, and the pollsters are working overtime. Every day brings new numbers. "Harris leads by 5 points nationally." "64% of Democrats approve of the switch." "Harris trails in key swing states." These polls are treated like scientific measurements of public sentiment. But what if they're really just elaborate illusions?
Pierre Bourdieu, a notoriously abstruse French sociologist,1 argued that "public opinion" as measured by polls doesn't actually exist.2 This might sound absurd at first. After all, we can see the numbers. But Bourdieu's point is deeper: the entire enterprise of polling is built on faulty assumptions that distort our understanding of political reality.
Let's look at how this plays out with Harris's recently-coronated candidacy. Bourdieu writes: "The opinion poll is, at the present time, an instrument of political action; its most important function is perhaps to impose the illusion that a public opinion exists, and that it is simply the sum of a number of individual opinions." In other words, polls don't just measure opinion — they create it.
Consider a typical poll question: "Do you approve or disapprove of Kamala Harris as the Democratic nominee?" Sounds straightforward, right? But Bourdieu would point out several problems:
It assumes everyone has an opinion. But many people might not have thought much about Harris or the unusual circumstances of her nomination. As Bourdieu puts it, "the first condition for the production of opinions is to be able to perceive a question as being political." Not everyone sees the world through this lens.
It treats all opinions equally. But should the view of someone who has closely followed Harris's career carry the same weight as someone who barely knows who she is? Bourdieu argues that "by gathering a plurality of opinions which do not have the same real importance, the results are very severely distorted."3
It imposes a framework on the issue. Bourdieu notes that "the simple fact of asking everyone the same question implies the hypothesis that there is a consensus about the problem, that is, an agreement about which questions are worth asking." But who decided this was the right question? Why not ask about specific policies, or the process of replacing Biden?
These aren't just academic quibbles. They fundamentally shape our political discourse. When we see a poll showing "53% support Harris," it creates the impression of a clear public mandate. But this "public opinion" is largely a mirage.
Bourdieu goes further, arguing that polls serve a specific political function: "The 'public opinion' which is stated on the front page of the newspapers in terms of percentages... is a pure and simple artefact whose function is to conceal the fact that the state of opinion at any given moment is a system of forces, of tensions."
In other words, reducing complex political realities to simple percentages hides the real power dynamics at play. The replacement of Biden with Harris wasn't driven by polls, though such polls were supposedly shown to the ailing president — it was a decision made by party insiders based on various political calculations. But polls are now being used to justify that decision after the fact.
We might see headlines like "Majority of Democrats say Harris was right choice, poll shows." But this doesn't mean public opinion drove the decision. Instead, the decision is driving the creation of a public opinion to support it.
Bourdieu explains: "The prohibition will be strongest when the group is nervous." Democratic leaders were clearly nervous about Biden's chances, so they made a dramatic change. Now, pollsters (consciously or not) are crafting questions to demonstrate public support for that change.
This dynamic becomes even clearer when we look at how polls about Harris are being used. Every slight percentage-point shift within the "margin of error" is breathlessly reported as if it represents a real change in public sentiment. But Bourdieu would argue that what we're really seeing is the "mobilization of opinion."
He writes: "The opinion poll treats public opinion like the simple sum of individual opinions, gathered in an isolated situation where the individual furtively expresses an isolated opinion. In real situations, opinions are forces and relations of opinions are conflicts of forces."
Think about how different someone's response might be in a quick phone survey versus a long conversation with friends about Harris's candidacy. The poll captures a snapshot, but misses the complex, evolving nature of how people actually form political views — and the odd ways they often articulate them.4
This matters because polls don't just measure opinion — they shape it. When we constantly see poll results about Harris, it influences how we think about her candidacy. If we're told that "most Americans think X about Harris," we're more likely to consider X ourselves, or at least view it as a legitimate position.
Bourdieu argues that this creates a false consensus: "It imposes the idea for instance that in any given assembly of people there can be found a public opinion, which would be something like the average of all the opinions or the average opinion. The 'public opinion' which is stated on the front page of the newspapers in terms of percentages (60% of the French are in favor of .... ) is a pure and simple artefact."
So how should we approach the polling around Harris's candidacy? First, we need to be much more skeptical of poll results, especially when they're used to justify political narratives. We should ask: Who commissioned this poll? How were the questions worded? What options were respondents given?
Second, we should recognize that our own opinions are being shaped by the polls we see. It's a feedback loop: polls influence opinion, which influences future polls. Being aware of this can help us think more independently about Harris and her policies.
Finally, we should look for better ways to understand political realities than simplistic poll numbers. This might mean seeking out diverse perspectives, diving deep into policy details, or getting involved in local politics to see how things really work.
The replacement of Biden with Harris is a seismic political event, driven by complex factors that no poll can fully capture. As Bourdieu puts it: "The state of opinion at any given moment is a system of forces, of tensions." Understanding those forces and tensions requires more than just looking at poll numbers about Harris's favorability or her chances against Republican opponents.
This isn't to say the public's views on Harris don't matter. Of course they do. But "public opinion" as presented in polls is a crude approximation at best, and a manipulative fiction at worst. Real democratic engagement requires citizens who can see past these illusions and grapple with the messy, complex reality of politics.
So the next time you see a poll about Harris or any other aspect of the 2024 election, remember Bourdieu's warning: "Public opinion does not exist" — at least, not in the simple form polls pretend to measure. The real story is always more complicated, more dynamic, and more interesting than any poll can show.
Translation is partly to blame, but Outline of a Theory of Practice — what a mouthful! — is some tough sledding. Sociological Questions is far more accessible.
Follow that link to download the PDF of “Public Opinion Does Not Exist,” which is about as good as this stuff gets.
I argue something similar in my essay about the “median voter.”
“I’m a little bit religious and a little bit superstitious” or “I love going out, but I also like to stay in and curl up on the couch.”