How Do You Move Statewide Public Approval 70+ Points in Two Years?

Statewide public approval does not move on argument. It moves when credentialed local sources speak to credentialed local audiences across the full geography of a state, for long enough and at high enough density, that the average citizen's pre-conscious read on the topic changes before the conscious mind ever forms an opinion about it. That is the only thing that has ever produced a 70-point swing in two years, and it is repeatable.

Most people who try to move public opinion at scale start in the wrong place. They start with the message — the talking points, the policy paper, the well-produced ad. The message is the surface artifact of a campaign. The campaign is what happens to the predisposition of the population the message is hitting. If the predisposition has not moved, the message lands on a public that has already privately decided what it thinks. The message gets metabolized into the existing read. The number does not move.

This is the same operator problem that decides a discovery call, scaled up by several orders of magnitude. The forks fire before the words land. The work is to reach the predisposition first.

Why argument alone never moves a stuck number.

A stuck public approval number is stuck for a reason. Some piece of the population's pre-conscious read — fear, stigma, association with the wrong cohort, a half-remembered story from twenty years ago — is doing the holding. That holding pattern is not an opinion. It is a set of conditions that produce an opinion every time the topic comes up.

Argument lands on the conscious narrator. The narrator is the press secretary, not the president. It receives the argument, files it next to the existing read, and produces a sentence that explains why the existing read still holds. I hear you, but I just feel uncomfortable with it. The discomfort is the predisposition. The sentence is the narrator covering for it.

Decision Science is the operator practice of working on the predisposition rather than the narration. Applied at population scale, it becomes Thought Engineering — the discipline of changing what the public encounters by default until the pre-conscious read changes on its own.

Saturation of context, not saturation of argument.

The mechanism that moves a stuck number is not volume of message. It is density of trusted exposure inside the geography where the population already lives.

A citizen in a small Pennsylvania county does not change their mind because they saw a slick ad. They change their mind because over eighteen months they heard a respected local physician speak at the community center, then their pastor reference the same topic in a sermon, then the local newspaper covered an event at the high school, then a friend mentioned that her cousin in another county had been to a similar event. Each exposure is a small pre-conscious commitment. None of them is an argument. The argument is not what moved them. The accumulation of credentialed local sources speaking inside their existing trust network is what moved them.

The 2008 Soon-Brass-Heinze-Haynes work on the pre-conscious window tells us why this is the right unit of intervention. Decisions form across small forks long before the conscious mind narrates a position. Public opinion is the same mechanism running across millions of people at once. The forks are smaller and the timeline is longer, but the architecture is identical.

A statewide number is the sum of millions of pre-conscious commitments. You do not change the sum by shouting louder. You change it by changing the conditions inside which each commitment forms.

How this worked across Pennsylvania.

In 2014 the public approval rating for medical cannabis reform in Pennsylvania sat at 23%. The state's media environment was not moving favorably on its own. Legislators do not move on a reform until the underlying public number moves first, and the public number had been stuck for years. The mandate was to move the number to a point where elected officials could act.

The strategy was not to argue. It was to re-engineer the surface area on which the average Pennsylvanian encountered the topic. A two-year statewide influence campaign was built. Educational materials were designed and distributed at scale. An event model was deployed that covered all 67 counties — credentialed local sources speaking in front of credentialed local audiences, again and again, across the full geography of the Commonwealth. The sponsoring senators of the eventual Medical Marijuana Act adopted the same framework for their public outreach. The number moved from 23% to 96% across the two-year arc. The same architecture was then deployed in West Virginia. Same arc. Same speed. Same result. The full record of the work is on the Medical Marijuana Act Pennsylvania case study.

The campaign was not a campaign in the traditional sense. It was a system that changed what an average citizen of Pennsylvania metabolized by default when the topic came up in conversation.

Where the greats left it.

Edward Bernays — Freud's nephew and the founder of modern public-relations theory — spent his career on this question and stopped at the level of the message. He understood that the public mind ran on associations rather than logic, and he became extraordinarily good at producing associations. He did not have the cognitive-neuroscience evidence base that arrived after him. Solomon Asch ran his conformity studies in the 1950s and showed how powerfully social context overrides private judgment, and stopped at the laboratory. Robert Cialdini named the six levers of influence and stopped at the consumer-products surface. Each of them set the tool down at the edge of their own evidence base. The discipline now in practice picks up where they set the tool down and applies the synthesis at population scale, on policy questions where the cost of a wrong public-opinion read is measured in lives and decades.

Three moves you can run this week.

First, name the predisposition that is holding the number in place. If you cannot say in one sentence what the population's pre-conscious read on your topic actually is — not their stated opinion, the read underneath it — you do not yet have a campaign. You have a message. Sit down with three people who hold the existing position and listen for the predisposition the position is covering for. That is the unit you are working on.

Second, audit your existing campaign material for argument density versus exposure density. Most stuck campaigns are over-indexed on argument and under-indexed on credentialed local exposure. Count the ratio of dollars spent on producing message to dollars spent on getting trusted local sources in front of trusted local audiences. If the ratio is upside down, the budget is going to the wrong layer.

Third, identify the smallest credible unit of geography in which you can run a full-density saturation pilot. A county, a city, a media market. Run a six-month pilot where every credentialed channel inside that footprint is touched at high density. Measure the predisposition shift before and after. If it works at the unit level, you have a model that scales. If it does not, the architecture needs work before any larger spend is justified.

What this costs and what it does not.

A campaign of this kind is not cheap. It is also not as expensive as the paid-media campaigns it replaces. Most political and public-affairs budgets spend a significant fraction of their dollars purchasing attention that lands on a predisposition that is not yet ready to receive the message. The dollars produce impressions. The impressions do not produce shifts. The Behavioral Revenue System framing for public-influence work moves those dollars upstream, into the layer where the predisposition is actually formed, and lets the message at the end of the funnel do less work for more effect.

This is the same architecture the firm uses to install revenue systems for high-trust private practices. The geography is larger and the timeline is longer, but the underlying discipline is the same.

FAQ

Q1: How long does a 70-point statewide opinion swing actually take?

A1: The Pennsylvania reform moved from 23% approval to 96% approval across approximately two years of continuous statewide work. The West Virginia replication ran the same arc on a similar timeline. Faster movement than that is possible on issues where the existing predisposition is shallow, but on entrenched issues where the holding pattern has been in place for decades, eighteen to twenty-four months of saturation is the conservative floor. Anything faster is usually the appearance of a swing rather than the underlying read actually shifting.

Q2: Can you do this with digital media alone, or do you need in-person events?

A2: Digital media is necessary and is not sufficient. The mechanism that produces a real shift is credentialed local sources speaking inside the trust networks the population already operates in. Some of that work happens in digital channels. A great deal of it happens at the community center, the church, the school auditorium, and the local newspaper. The campaigns that have produced real swings have used both, weighted toward in-person credentialed exposure at the densest part of the arc.

Q3: Does this work for opposing a policy as well as supporting one?

A3: The mechanism is symmetric. The same architecture that moves a number from 23% to 96% can hold a number at 23% or move it back down. The discipline is the same. The discipline does not pick the side. The operators do. The firm's standing rule is that the underlying predisposition has to be readable, the population's interests have to be aligned with the campaign's, and the work has to be defensible inside the same Decision Science framework that governs the rest of the practice.

The deeper architecture for population-scale work is described in the Cersosimo method.