Campaigns spend heavily on persuading swing voters, and the term describes at least three distinct groups. Confusing them leads to strategies aimed at people who do not exist.
Switchers are rarer than assumed
Most voters support the same party consistently across elections, and the share who move between parties from one cycle to the next is comparatively small.
Survey measures overstate the group, because people misremember past votes in ways that align with their current preference.
Panel studies that track the same individuals over time find considerably more stability than cross-sectional surveys suggest.
Late deciders are a different population
Some voters know which party they prefer but decide close to the election whether the specific candidate merits their support.
These voters are typically less engaged with political news, which means they receive campaign messages later and from fewer sources.
They are reachable but expensive to reach, since the channels that deliver political information efficiently are precisely the ones they use least.
Turnout decisions are often the real variable
A large share of apparent swing is not people changing sides but supporters of one party voting in greater numbers than supporters of the other.
This produces the same shift in the result while requiring an entirely different campaign response, aimed at mobilisation rather than persuasion.
Distinguishing the two after an election requires data on who voted rather than on how the totals moved, which is available but rarely used in immediate analysis.
Cross-pressure explains much of the behaviour
Voters holding positions associated with different parties have no natural home, and which consideration dominates depends on what the campaign made salient.
Such voters appear inconsistent when they are responding predictably to changes in which issues are prominent at the moment they decide.
This is why campaigns compete so hard over which subject dominates the final weeks, since the argument being had matters more than the argument being won.
The strategic implication
Persuasion and mobilisation compete for the same finite budget, and the correct allocation depends on which group is actually producing the movement in a given contest.
Campaigns that misidentify this spend on advertising aimed at switchers who are not there, or on turnout operations among people who would have voted anyway.
The diagnosis has to be made contest by contest, since the composition of a district determines which of the three mechanisms carries the most weight.