Here’s the real reason why most of the press’s political analysis is useless

Virtually all political analysis shares a common premise that voters are divided into various ideological camps, and that while they might be coaxed to support a politician from a different spot on the spectrum they will naturally gravitate to candidates who share their label. Every day, you will find opinion columns and news reports that assume, for example, that Joe Biden’s supporters would search for another moderate Democrat if he fades, or that warn that Elizabeth Warren or Bernie Sanders would struggle to win the votes of mainstream liberals in Wisconsin, or otherwise begin with similar “political lane”-based premises.

But in the cumulative exit polls for the 2016 Democratic primary, Bernie Sanders beat Hillary Clinton among voters who wanted the next president to be more liberal than Obama, just as you would expect, but he also won over Democratic voters who wanted a more conservative president than Obama. Clinton won handily by dominating among those who told pollsters that they thought Obama was about where they wanted the next president to be ideologically.

Currently, and very much contrary to the conventional wisdom, progressives Elizabeth Warren and Bernie Sanders are tied as the most frequently cited second choice by those who prefer the more moderate Joe Biden, according to Morning Consult’s polling. Warren supporters are only slightly more likely to name Sanders as their second choice (28 percent) than to pick Biden (24 percent), and the same is true of Sanders voters–32 percent say Warren is their number 2, and 28 percent pick Biden. The top second choice among those backing Pete Buttigieg, who is often described as competing in Biden’s “moderate” lane, is Warren.

The reality is that most voters aren’t nearly as ideological as the people who write about politics (and political junkies on Twitter) believe. There is some evidence that the electorate has become more ideological with our rising polarization, but even as more identify as “very liberal” or “very conservative,” when you drill down a bit, most  people still aren’t consistently so in their actual beliefs.

One shouldn’t overstate the case. It’s not that ideological affinity plays no role whatsoever in voters’ candidate selection, or that voters who place a great deal of emphasis on ideology don’t exist. Rather, perceived ideological alignment is only one among a  mishmash of factors, often irrational, that go into voters’ choice of candidates. The media tend to report as if it’s the primary factor, but it’s probably pretty low on the list after personal affinity–the one you’d want to have a proverbial beer with–cues from your in-group (or in-groups), the view that a primary candidate can win in the general election, etc.

When we get to a general election, the storyline tends to shift to independents, who are often portrayed as moderate “swing-voters” that might be scared away from a party that moves too far toward the left or right. Such voters do exist–they’re not entirely mythical creatures like leprechauns–but not in significant numbers. Most independents are “closet partisans” who vote consistently for the same party (when they vote at all), and studies suggest that, on average, their politics don’t differ significantly from those of partisans.

There’s a correlate when the media cover various policy proposals. Many pundits and reporters assume that various proposals will appeal to or alienate this or that group of voters. This is based on what political scientists Christopher Achen and Larry Bartels call “the folk theory” of democracy. The idea is that people identify with a party or a candidate because they perceive that doing so will advance their own rational interests.

But in their book, Democracy for Realists, the scholars make a pretty strong case that voters tend to work backwards, first developing an attachment to a candidate (or party), and then working backwards to support the policies for which the candidate advocates. They wrote in The Washington Post in 2016 that “decades of social-scientific evidence show that voting behavior is primarily a product of inherited partisan loyalties, social identities and symbolic attachments. Over time, engaged citizens may construct policy preferences and ideologies that rationalize their choices, but those issues are seldom fundamental.”(As someone who writes frequently about healthcare policy, it’s quite clear that most of the people yelling at me on Twitter aren’t well versed in the subject but nonetheless have strong beliefs that correlate perfectly with whether or not my writing supports their favorite candidate’s position.) There are exceptions–Achen and Bartels point to single-issue voters who care deeply about issues like guns or abortion–but for the most part, affinity for a candidate comes first.

There’s an important caveat here. It’s very likely that lazy political analysis can create a feedback loop that effectively makes ideology more important to voters than it would otherwise be. A broad body of research has shown that most people don’t follow politics very closely and aren’t well-versed in policy details, so they tend to look for cues among their social affinity groups and other influencers. The political press influences perceptions and the constant drumbeat of reporting and analysis about how a given candidate may or may not align with this or that group likely creates a self-fulfilling prophesy by telling voters who is and who isn’t in their “lane,” or which candidates are “one of them.”

There is evidence that a candidate perceived as being outside the mainstream will face a small penalty with voters for that perception–much smaller than most pundits would tell you, but a penalty nonetheless. Most political journalists have a fairly strong bias toward centrism, and it’s quite likely that hearing over and over again that a given candidate’s policies are too far to the left or right to attract broad support can bring that penalty down on a campaign even if voters wouldn’t otherwise feel that those policies are problematic.

With some exceptions, the political press in this country is abysmal at analyzing politics consistent with empirical research into how it all works. Political journalists not only have a tendency to prize moderation, but most remain deeply wedded to the “folk theory” of democracy, and that results in coverage that skews public opinion in all sorts of ways that are ultimately deleterious to healthy democratic debates.