The Voter Trap
How political competition gets stuck on a question almost nobody
wants, and why nobody can unstick it. This is the static version of
an interactive story.
A pattern across democracies
Across the world’s democracies the same shape keeps appearing: a
voter whose pocketbook points one way and whose cultural loyalties
point the other, and whose vote is decided by the second. She is in
the right-populist coalitions of the United States, Britain, France,
Italy, Hungary, and Poland; the left-populist movements of Chávez,
Correa, and Morales; the nationalist turns of Bolsonaro, Duterte,
Erdoğan, and Modi. In the textbook, competition fixes this. Yet
three things keep happening that the textbook says should not: a
dominant coalition stops competing on the bread-and-butter issues most
of its own voters care about and keeps winning; a rival offers those
voters their ideal on every issue at once and is not believed; and
entry stays legally open, yet for a generation no new party breaks
through.
The shared move
These are the signature of one strategy. As Gidron and Margalit
describe Netanyahu’s coalition, such parties win “not by
moving to the center, but through destroying the social legitimacy of
crossing over to the other side” — anchoring the contest on
an identity question where the rival is fixed in place and cannot
credibly follow. The same move recurs: the 2015 Likud video warning of
Arab voters voting “in droves”; the “non-citizen”
voter claim in the Trump coalition; Orbán’s “Soros
plan”; the BJP’s “infiltrators.” This is a
theory of how that move works. The clearest case on record, where it
played out in print in eleven weeks, is North Carolina in 1898.
The case
On the morning of November 10, 1898, in Wilmington, North Carolina,
armed white men overthrew an elected city government — what
historians call the only successful coup d’état in the
history of the United States. It began at a printing press. Three
months earlier, North Carolina was, by the standards of the American
South, a working multiracial democracy, governed since 1894 by the
Fusion: Black Republicans and white Populist farmers. Picture one
voter, a white tenant farmer in the cotton east. She wants Populist
economics and the racial order she was raised inside. She is
cross-pressured: each party owns half of what she wants.
Two stories, and a drumbeat
A party’s brand is what people will actually believe it when
it speaks, earned slowly by repetition and visible behavior. The
Populists built theirs on economics and clean government
(“Smoke the rascals out!”) and governed across the color
line. The Democratic apparatus, Simmons’s machine and
Daniels’s News & Observer, built its brand on one
axis: “this is a white man’s country and white men must
control and govern it.” Through the fall of 1898 the apparatus
press ran that theme daily. In one documented case, a year-old
Georgia speech defending lynching, untouched by any North Carolina
paper for a year, was revived on August 18, 1898 and kept running:
67 issues with references before the election. Alex Manly, editor of
Wilmington’s Black-owned Daily Record, answered it in an
editorial; the apparatus press attacked his reply for the rest of
the fall. Meanwhile, across the whole window, lynchings in North
Carolina numbered zero, and the pairing the columns obsessed over,
Black men married to white women, amounted to about three couples in
ten thousand. The coverage ran at a hundred to a thousand times the
scale of anything actually occurring. It was not reporting. It was
production.
Raleigh News & Observer, Nov 8 1898 — Library of Congress, sn85042104 (public domain).
The jump, dated by the rival
The dominant coalition does not abandon the economic debate because it
stopped mattering, but because its rival got better at it: four years
of governing made the Fusion credible on bread-and-butter questions,
and the more believable the rival became there, the worse that fight
looked. Past a threshold the best campaign flips all at once. On August
23, 1898, the Democratic Hand-Book launched and the contest
jumped to the racial corner. Nothing outside changed that week; the
trigger was the rival’s own accumulated strength. In the
apparatus’s pages, “ballot security” and the racial
order fused into one subject — the correlation between the two
themes ran near zero before August 23 and about 0.56 by election day.
The bounce
In October the Populists reached for the racial frame themselves,
offering that voter both halves of her ideal on one ticket. It could
not be believed: four years of governing across the color line fixed
where they sat on race, so the claim fell outside their believable
range and was not weighed at all. Some pivots are not discounted; they
are dismissed. In the model’s illustrative units, the pivot cost
0.676 of credibility against a budget of about 0.5.
The trap, and why it holds
The election was held November 8, 1898: the Democratic vote for
governor swung +33.4 points between 1896 and 1900 in the counties
where the Black share of the population was largest, and −1.4
in the mountain counties where it was smallest. Two days later came
the Wilmington massacre and coup, beginning with the burning of
Manly’s press. Two locks held the outcome in place. First, the
winning story is never disproved, because the stories that would
disprove it are the ones no one is left able to tell and be believed;
the voter, honest and attentive, is never surprised, so she never
moves. Second, a newcomer can be believed or be different but not both
at once, so entry stays open in law and closed in practice. No serious
new party broke through for roughly seven decades, until the federal
civil-rights laws of the 1960s rewrote the menu from outside. The voter
got the racial order she wanted; the economics she wanted never came.
She was not deceived, lazy, or bought. The trap is a property of the
system, not a defect in her.
Not only history
Nothing in the mechanism is particular to one election or one dominant
party. Its two-sided form — two coalitions, each deep on its own
cultural anchor, each holding its own cross-pressured voters, no room
between them for a new party — is the natural way to read the
present. You can even keep score: the number of genuinely independent
questions a politics contests at once fell from about 2.5 (1894) to
about 1.6 (1898) in North Carolina’s press, and preliminary
evidence points the same way in modern American survey data. A
measurement, not a prophecy — it says what to watch: the quiet
narrowing of how many questions there are.
Read the full paper at
jccisneros.com/uploads/voter_trap_paper.pdf.