There is a huge debate underway about whether Democrats should move toward the center of the political spectrum or to the left. As usual, many people who want the party to land at their preferred point on the spectrum also argue that this would be the best electoral strategy, although those are separate issues.

Some observers note that shifting one’s ideological placement is a poor tactic because, as G. Eliot Morris told Paul Krugman, “voters have very poor understanding of what candidates actually stand for at the issue-position level. They also have a very poor understanding of what these ideological labels: moderate, progressive, really even mean.” He also says, “the vast majority of the American public is not consuming the type of information that you would need to know, first off, what issue positions politicians hold and second, what the ideological labeling, the orientation of those, what those issue positions are.”

For some progressive commentators, this kind of evidence counts against moving to the center. A centrist platform won’t help win elections if voters are unaware of candidates’ positions. But the same evidence would also argue against moving to the left. If we assume that people don’t know enough to evaluate policies, and a candidate can equally well propose anything–well, that is a cynical theory and a depressing one if it’s true.

I take a different view. I observe that voters are heterogeneous. They care about various issues, believe various kinds of information that they derive from various sources, identify with various social groups, feel various ways about each major institution (experiencing emotions that range from trust and respect via obliviousness to contempt or fear), vote–or don’t vote–for various reasons, and consider various combinations of policies, personal characteristics, demographic markers, and perceived performance when they assess candidates.

One interpretation is that people are naive or “innocent” about ideology (Kinder & Kalmoe 2017). They have, as Morris says, “a very poor understanding” of the ideological spectrum. I would counter that the ideological spectrum is just one way of organizing beliefs, and probably a poor one. We shouldn’t allow the seating arrangements of the French National Assembly in 1793 (when the Jacobins sat to the left and gave that word its political significance) to mesmerize us. People who organize their political thoughts in other ways may have insights.

Consider John, a major character in Farah Stockman’s nonfiction book American Made: What Happens to People When Work Disappears. John is a union activist who comes from a long family tradition of militant unionism, originally in Kentucky. He knows a great deal about trade policy. Some of his most important sources of news are union publications. There is a Confederate flag in his basement, which has meanings that he may not want to acknowledge but that also stands, in his own mind, for Appalachian workers against elites. Most deeply, he divides the world between workers and managers. He hates talk of “white privilege” because he feels oppressed as a worker. He wants the union to fight the company, and he voted for Trump in 2016. His wife is more favorable to management. On that basis, he describes her as a “liberal.” He is surprised when a Republican politician doesn’t seem to favor US workers over managers, as he would expect.

I disagree with John in many respects, including the way we use words like “liberal.” Still, I could learn a lot from him about trade policy and industrial issues. He is not “innocent” of ideology. (If that flag is anything, it isn’t innocent.) Not is he ignorant or uncaring. He just organizes his beliefs about the world very differently from me because of his accumulated experiences.

If most voters agreed with John, then candidates would be wise to favor both unions and tariffs and to oppose race-conscious policies. But that is not my point. John represents one sliver of a very heterogeneous electorate. A policy recipe that would appeal to him would not work for many others.

I should acknowledge that I know what people mean when they distinguish progressive politicians from centrists. This distinction conveys information to me. If all I know is that politician A is to the left of B, then I will be biased in favor of A. But the information I can glean from these labels is limited, reflecting just one way of organizing the political debate. It is a signal with a whole lot of noise. I would much rather know more than which candidate is considered further left according to a certain elite discourse.

Policy positions do matter, and no one should treat the electorate as ignorant. But it is literally impossible to move to the center–or to the left–if there is no common spectrum.

If you are a candidate, you should generally adopt the policies that you believe are best and advocate for them. If a specific policy is unpopular among swing voters in your constituency, you might need to compromise on it, because you can’t accomplish anything if you lose the election.

You should consider the pros and cons of proposals that elites and specialists would classify as belonging to the left, and the center, and the right. The ideological label of a policy does not tell you whether it is smart.

However, you shouldn’t adopt a miscellaneous list of policies. You should present your ideas coherently. You need a narrative or core theme. But each candidate’s thematic coherence may look distinctive.

You should demonstrate respect for the electorate by endorsing and defending specific positions. But you should also realize that your policy platform is just one factor. At least as important are your biography and record, your rhetorical style, and your modes and methods of campaigning.

Sources: Donald R. Kinder and Nathan P. Kalmoe, Neither Liberal nor Conservative: Ideological Innocence in the American Public (University of Chicago Press, 2017); Farah Stockman, American Made: What Happens to People When Work Disappears (Random House, 2021). The classic source for the idea that American voters do not understand ideology is Phillip E. Converse, “The nature of belief systems in mass publics (1964)” reprinted in Critical Review 18.1-3 (2006): 1-74. Converse does acknowledge that if people’s opinions are “idiosyncratic,” then we will find “little aggregative patterning of belief combinations,” because people may “put belief elements together in a great variety of ways” (p. 44). For him, this would be evidence of ignorance, but I would observe heterogeneity instead.

My own work on this topic includes: “People Are Not Points in Space: Network Models of Beliefs and Discussions.” Critical Review 36.1-2 (2024): 119-145; and “Mapping ideologies as networks of ideas.” Journal of Political Ideologies 29.3 (2024): 464-491.