Friday, April 03, 2020

If You're One in a Million...

Many of you are familiar with the saying "If you're one in a million, there are a thousand people just like you in China alone."

It helps illustrate that while one in a million is certainly very rare, on another way of looking it at it's also quite common. A thousand people! You could fill a high school gymnasium with that!

Push the proportion down a bit and things get even more stark. Imagine a political view held by only 1% of the population. That's pretty fringe, right (for reference, 33% of Americans believe that alien UFOs have visited Earth)? But it's also one in a hundred -- in America, that translates to well over three million people. That's a lot! (We explored this dynamic previously in my "how to tokenize with proportions" post.)

One thing I often think about is how modernity and modern technology, in conjunction with our decidedly pre-modern lizard-brains, don't always mesh well. We know, for example, that fat tastes delicious because in the primordial environment it was rare and vital, and thus highly desirable to consume -- unfortunately, this doesn't translate well to a contemporary context where calories and fat are plentiful and we can easily over-saturate ourselves.

I suspect there's something similar going on with political opinions. One of the oft-proclaimed virtues of the internet is it allows you to find communities of like-minded persons no matter how obscure or random the interest. Obsessed with underwater basketweaving? You can find dozens of people who share that passion with minimal effort!

What does it mean when the same is true for political opinions? I suspect our brains have a rough heuristic at the ready that correlates how difficult it is to find holders of a given opinion with how uncommon it is in society. If one struggles to come across individuals who believe ideology X, one assumes that X is rarely believed in a given society. If one comes across X-believers without too much trouble, one infers that X is a common ideology. If 1% of Americans hold a particular political stance, that may be three million people -- but (at least until recently) they're not going to be easy to find via the normal modes of political engagement. If you just read newspaper columns, chatted with your neighbors, watched TV pundits, and so forth, you'd probably come across it rarely, if ever. If one really wanted to find a sizable chunk of Americans who believe this 1% view, one would have to expend considerably more effort.

Now to be clear: what I'm describing is only a heuristic, which means it's imperfect -- there are all sorts of reasons why, for example, a rare opinion might nonetheless be easy to spot "in the wild" (it's favored among extroverts or celebrities, e.g.) or a common one might be rarely seen (it's embarrassing). But it has some logic as a rough-and-ready way of telling us which views are common in our social circle and which aren't. It's not quite the same as the availability heuristic, but it is similar. Call it the search heuristic. Something easy to find upon commencing a search for it is common; something hard to find even when searching for it is rare.

The problem is that if modern technology makes pretty much any opinion with even a speck of public salience "easy to find", that hijacks our heuristic circuitry to make all of these opinions register in our minds as "commonplace". What is the result of that?

One potentially positive result is that it might offset some mechanisms that serve to silence dissident views via the so-called "spiral of silence" -- they learn that they're not alone, and so they're more willing to air their dissident views knowing that there are peers who share their perspective.

But there are also some potential upshots that I'm more ambivalent about. One thing that we might experience is the erosion of perceived consensus -- a sense of widespread opinion balkanization and a corresponding vertiginous inability to tell when there is an opinion that carries significant social agreement. There's a push/pull on this -- sometimes, a feeling of "consensus" is dependent on wrongly not perceiving the existence of dissent, and so the elevation of dissident voices corrects a widespread social misperception. But, assuming "consensus" does not require universal agreement, sometimes, a feeling of dissensus is falsely inspired by the presence of high-profile but ultimately negligible dissenters. To the extent that modern technology makes very small ideological minorities loom larger, we might believe ourselves to be far more disunited than we actually are. And if the search heuristic causes a wide range of opinions (many mutually incompatible with one another) to register as "common", we may have trouble grasping onto distinctions between actually common versus fringe outlooks.

In a similar vein, it is at least plausible that in a democracy there is a prima facie obligation to consider and give airing to certain viewpoints simply by virtue of the fact that they're common. This wouldn't necessarily mean that uncommon views can be automatically rejected, only that they must "earn" their space on the democratic agenda by means other than "because many people believe it". If this is so, then the perception that more views are "common" mean that more views can claim access to this prima facie obligation of consideration. Perhaps that doesn't strike you as a bad thing -- but consider it in the case of, say, openly avowed racism or extremism -- views which might objectively be as rare as ever, but perhaps feel more common than they've been in recent memory.

There are also risks latent even for the holders of the dissident opinions themselves, for they as much as anyone might be mislead into thinking their views are more widely shared than they are. If someone holds a view they know is rare but wish was widely shared, they must endeavor to persuade others to adopt it. If they then, say, run for office on its platform whose tenets are held by only 10% of the population, if (or when) they lose they probably won't be happy but they at least probably won't be confused. Unpopular opinions don't win elections.

But things are different if the search heuristic misfires and makes the dissident believe they are actually expressing a very common view. If they nonetheless persistently lose in the democratic arena, they might suspect bias, corruption, institutional barriers, or other forms of foul play are obstructing them. To be clear: there are many cases where such things are at work; I'm not saying that everyone who believes their views are not carrying the democratic day because of various social biases is simply misleading themselves. But sometimes a democratic spade really is just a spade; and there is at least the potential for this sort of self-deception to accelerate -- the result being greater mistrust and resentment of social institutions.

It's worth noting that there isn't an "objective" way of declaring whether a view is "rare" or not. Much of it already lies in framing: "held by 1% of the population" sounds uncommon, "held by three million Americans" sounds reasonably common. So we can't quite say that, even if the search heuristic is misfiring, it is objectively causing us to label "uncommon opinions" as "common". But I do suspect that our wider net of appraisals around how we relate to an opinion based on its perceived "commonality" are tied to the same set of assumptions under which the search heuristic should function at least roughly well -- meaning that if we no longer exist in that social world, the whole edifice comes under serious strain (if it doesn't collapse outright).

These are preliminary thoughts; they are not wholly hashed out in my mind yet, and I'm curious to hear others' views. Here's the tl;dr

  1. The search heuristic tells us that, roughly speaking, a view that is hard to find upon searching for it is rare, and a view that is easy to find upon searching for it is commonplace.
  2. The social media revolution has drastically reduced the search costs required to find large absolute numbers of persons who hold any particular view, even when they are actually relatively uncommon.
  3. Together, (1) & (2) cause us to mentally code many viewpoints which we'd perceive as uncommon as quite common (since we are able to find examples of them with little effort).
  4. The effects of this are unclear, but may include (a) increased willingness to air dissident views; (b) decreased sense of social consensus; (c) decreased ability to distinguish relatively common versus uncommon views; (d) decreased trust that formal mechanisms for measuring public opinion reliably track actual public viewpoints (even when they are in fact doing so reasonably well).


Jay Reardon said...

This is some brilliant analysis and writing! And this partially explains why minority republicans with extreme views rule the country and have won 3 presidential elections in the last 20 years. Technology easily connects the minority, extremist herd, which tend to be more tribal and disciplined/hardened in their beliefs. (Liberalism and progressivism is inherently flexible and cooperative -- at least it should be -- Bernie Bros are not and that's probably why he ultimately could not win anything nationally.). And also why a consistent majority of Americans agree with taxing the wealthy more, affordable health insurance, free or affordable college education for all, etc. -- yet this and other related issues are never close to being implemented. The US population consistently polls more liberal/progressive than Congress. This is also caused by what I'll call "conclusion by anecdote". Reagan's welfare cheater "driving a caddilac and buying steaks with food stamps" is but one example.

Ed said...

“It is difficult to get a man to understand something, when his salary depends on his not understanding it.” ~ Upton Sinclair