top of page

The Voter Named John

12h
4 min read

Meet John

John is not a real person, but you have met him. He is a hypothetical voter introduced by Verlan Lewis and Hyrum Lewis in The Myth of Left and Right (2023) to expose a flaw in the way we measure political opinion. John holds two strong views. He favors strict restrictions on firearms. He also favors strict restrictions on abortion. On each question he is about as far from the middle as a person can be.

Now ask a pollster to place John on the left-right spectrum. The standard method scores each position and averages. John's gun position pulls him toward one end; his abortion position pulls him toward the other. The arithmetic lands him squarely in the center. John, the man with two of the most uncompromising positions in American politics, is recorded as a moderate.

The Lewises use John to make a point that is almost embarrassingly simple once stated: a one-dimensional scale cannot distinguish between a person who holds mild views on everything and a person who holds extreme views that happen to point in opposite directions on the conventional map. Both average to the middle. The line reports them as the same voter.

This is not a technicality of survey design. It shapes how campaigns, journalists, and parties think about the public. When analysts say that "moderates" decide elections, they imagine a bloc of cautious, split-the-difference citizens. Some such people exist. But mixed in with them, and statistically indistinguishable from them under the usual method, are the Johns: voters who are intense rather than lukewarm, and intense about issues the two bundles have divided between them.

The Lewises offer a companion example that makes the same point from the other direction. The same ruler that calls John a moderate places Adolf Hitler and Milton Friedman at the same "far right" end. In both cases the instrument is reporting team association, or the lack of it, and dressing the result up as a reading of conviction.

How many Johns are there?

More than one might think. John's particular combination, restrictive on guns and restrictive on abortion, is familiar among many religious voters, including in many Black and Hispanic congregations whose members have long been counted as reliable members of one team on the strength of one issue while holding the other team's position on another. The mirror-image voter, permissive on both, exists in large numbers too, and is equally invisible on the line. Then there are the cross-pressures that involve neither of those issues: the union member who wants trade protection and lower immigration; the small-business owner who wants deregulation and a public health system; the veteran who wants a strong military and a government that stops sending it abroad. Every one of these people is a John. Every one of them averages out.

Earlier posts in this series described Philip Converse's finding that most citizens do not reason along the spectrum, and Ellis and Stimson's finding that label and preference diverge for millions. John is what those findings look like one person at a time. He is not confused. He has simply declined to accept a bundle that was assembled, as the Lewises argue, by history and group loyalty rather than by any principle that would require his two positions to agree.

Under plurality voting with two dominant parties, John has three options. He can vote for the team that shares his gun position and swallow its abortion position. He can do the reverse. Or he can stay home. What he cannot do, in most states and most years, is vote for someone who holds both of his views, because the ballot rarely offers such a person and the debate stage almost never does.

Over time this has a corrosive effect that goes beyond John's own frustration. The parties learn that John is "theirs" on one issue and stop competing for him on the other. His second issue becomes, in effect, free: the team that has him need not deliver on it, and the team that lacks him need not try. The average that erased John from the survey also erases him from the agenda.

Where the evidence pushes back

The strongest counterpoint is that John's type, while real, may be rarer than the argument implies among people who actually vote. Matthew Levendusky's work on partisan sorting shows that engaged partisans have increasingly brought their issue positions into line with their party's bundle, so fewer active voters hold cross-cutting combinations than did fifty years ago. Scholars in the essentialist tradition would add that John's two positions may not be as unrelated as they seem; some would argue both express a single disposition toward protecting the vulnerable, which would make John coherent on a different axis rather than off the map entirely. Finally, two-dimensional surveys, which the next post discusses, can locate John quite well; the problem is specifically the one-dimensional summary, not measurement as such.

Why this matters for the ballot

A voter who is intense on two issues and averaged into the middle is not represented by a choice between two bundles that each contain half of what he believes. Fair ballot and debate access for independents and third parties is how a John gets to vote for both of his convictions at once.

— Michael W. Cunningham, Ph.D.

Cover image: "Suburban Americana Panorama" via Wikimedia Commons. See the file page for author and license.

 
 
 

Recent Posts

See All
A Veterans Day Promise:

Dr. Mike Cunningham's Veterans Day message on honoring those who served and keeping our promises to them.

 
 
 

Comments


bottom of page