The Movement Framework
Two polls can show the same number while describing very different political realities.
A topline measure describes a position. By itself, it does not tell us the path to that position, whether apparent support is durable, or whether changes among different people offset one another. Those are separate questions, and answering them requires more than renaming the headline number.
Position / Direction / Velocity / Volatility is an editorial synthesis of ideas developed across my writing, not a previously published four-variable validated model. The four dimensions organize questions about change. They are not a score, a prediction formula or stages through which an electorate necessarily passes.
Four dimensions of change
Position
Where a measured opinion currently stands. A percentage, distribution or reported preference can describe position, provided we identify the population, question and time of measurement. A national estimate and a measure of conversation on one platform describe different populations. Similar-looking numbers are not automatically comparable.
Direction
Whether the measure is moving positively, negatively or remaining relatively stable over the relevant interval. Direction always concerns a specified measure: support for a proposition, for example, rather than a general judgment that events are becoming better or worse. Different measures can move in different directions without either being incorrect.
Velocity
The rate at which change is occurring. A change over a week is not the same observation as the same change over a year. Comparisons need a consistent interval and sufficiently comparable observations. The speed at which a narrative circulates is also different from the speed at which reported public opinion changes.
Volatility
The degree to which an observed position appears stable or fragile over time. This question concerns durability: does an apparent position persist, reverse or fluctuate? Fragility is often inferred rather than directly observed, so it needs to be distinguished from the measurements used to suggest it.
What volatility does—and does not—mean
Earlier essays also used volatility in a more statistical sense, including variation in sentiment scores over time. That is a concrete description of a measured series. It does not, by itself, establish the fragility of an individual’s preference. The later discussion of durability asks a related but different question.
Volatility is not velocity. A measure can change quickly in a consistent direction without fluctuating markedly. It can also move up and down while showing little net change. Describing both situations as momentum would conceal an important difference.
Volatility is not persuadability, either. Instability in an aggregate measure does not demonstrate that the people represented are responsive to new information. Apparent movement can reflect changing participation, measurement differences or consolidation of existing interpretations. None of those possibilities can be resolved from a topline alone.
Dimensions are different from drivers
In What Moves Voters and The Hidden Metrics Deciding Elections, I discuss four drivers: persuadability, issue salience, message resonance and volatility. That set concerns possible conditions underlying movement. The four dimensions on this page describe different aspects of the movement being examined.
The shared word volatility does not make the two sets interchangeable. Describing how a measure changed is not the same as establishing why it changed. A historical example may suggest an explanation without isolating its cause.
Why this matters
The same reported position can accompany strengthening, weakening or simple fluctuation. Campaigns and other organizations may interpret those situations differently, but the framework does not prescribe electoral tactics. Its purpose here is to make the underlying research questions explicit rather than treating one number as a complete account.
A useful description identifies the measure, population, interval and uncertainty. Repeated surveys can establish some changes; individual trajectories require evidence suited to that question. Two similar endpoints do not prove that nothing happened between them—and they do not prove hidden movement, either. That distinction is central to using this framework with appropriate restraint.
Further reading
What Moves Voters — The separate four drivers and the conditions beneath headline preferences.
The Hidden Metrics Deciding Elections (That Polling Can’t See) — A companion explanation of persuadability, salience, resonance and volatility.
The Volatility Problem — The later emphasis on fragility and durability.
Measuring the Pulse of Change: The Voter Sentiment Volatility Index — An early account of sentiment variation and an emerging index.
Measuring Voter Sentiment Volatility: Staying Ahead in a Rapidly Changing Landscape — The more statistical treatment of variation over time.
Watch
WPA CEO Chris Wilson discusses Presidential polling numbers on Fox News
Earlier discussion of this idea, around 3:11: similar polling comparisons may conceal intervening movement. The explanation is presented as conjecture, not a measured trajectory. Uploaded August 6, 2025; the discussion concerns the 2008 election. It is not the origin of the four-part framework.
Christopher S. Wilson 