What Evidence Would Ever Change Their Minds?

One of the simplest ways to test whether a policy position is evidence-based or ideological is to ask a single, unfashionable question: what evidence would change your mind? Not what evidence would reinforce your position, not what data you already trust, but what outcomes would count against the policy you support. In science, this is called falsifiability. A claim that cannot, even in principle, be contradicted by evidence is not empirical at all. It is a belief.

This distinction matters because evidence-based policy is supposed to work the same way science does: hypotheses are proposed, interventions are implemented, and outcomes are evaluated against expectations. If the outcomes diverge, the hypothesis is revised or abandoned. When this process breaks down, when outcomes are endlessly reinterpreted to preserve the original claim, the language of evidence remains, but its substance disappears.

Applied to nicotine policy, this test is devastatingly simple. If vaping prohibition is justified on public health grounds, then there must exist observable outcomes that would indicate it is failing. Rising smoking rates should count. The growth of illicit markets should count. Enforcement collapse should count. Displacement from regulated products to unregulated ones should count. Increased dual use, reduced cessation success, or worsening health inequities should count. If none of these outcomes qualify as disconfirming evidence, then prohibition is not being defended as a public health intervention but as an article of faith.

What makes this so stark is that these outcomes are not obscure or hypothetical. They are precisely the risks critics of prohibition warned about in advance. When predicted harms materialise and are still dismissed, the problem is no longer uncertainty or imperfect data. It is the refusal to allow evidence to function as evidence.

Start with smoking itself. The core justification for banning or severely restricting vaping is that it allegedly threatens the long-term decline in smoking, particularly among young people. Vaping is framed as a “gateway”, a “renormalisation” risk, or a derailment of decades of tobacco control progress. On this logic, smoking outcomes are central to the policy’s moral and empirical legitimacy.

Yet when smoking rises after restrictions are imposed, the response is not reconsideration but reinterpretation. The increase is described as “temporary”, “noise”, “lagged effects”, or blamed on unrelated factors such as stress, cost-of-living pressures, population changes, or the pandemic. The causal arrow is never allowed to point back at policy. Smoking outcomes are only admissible as evidence when they align with the desired narrative.

This creates an asymmetry that would be unacceptable in any genuinely scientific framework. If smoking declines, the policy is credited. If smoking rises, the policy is insulated. Success is causal; failure is incidental. At that point, the metric is no longer being used to test the policy but to protect it.

This is not how evidence-based policymaking works. If a policy is introduced explicitly to reduce smoking, then a sustained increase in smoking following its implementation is not an inconvenience to be explained away, it is a direct challenge to the policy’s rationale. The refusal to treat it as such reveals that the outcome being optimised is not smoking reduction but something else entirely, often moral conformity to abstinence, symbolic toughness, or institutional consistency.

The same pattern appears even more clearly in discussions of black markets. Prohibition advocates routinely acknowledge the existence of illicit vaping markets, but only to weaponise them rhetorically. The black market is cited as evidence that vaping is dangerous, uncontrolled, and therefore must be banned even harder. Its existence is treated as proof of the product’s inherent threat rather than a predictable consequence of regulatory design.

The possibility that prohibition caused the black market, or at least dramatically expanded it, is excluded from consideration. This exclusion is not accidental. If illicit markets were recognised as policy-generated harms, they would constitute powerful falsifying evidence. Instead, the logic is inverted: the failure becomes the justification.

In any other domain of public policy, this reasoning would be recognised as circular. Alcohol prohibition, drug prohibition, and even housing or labour market interventions are routinely judged in part by whether they create unmanageable black markets. In nicotine policy, however, the same outcome is treated as confirmation that the policy has identified a uniquely evil substance rather than a uniquely flawed approach.

This inversion is philosophically revealing. The policy is insulated from blame by redefining all negative outcomes as intrinsic to the product rather than contingent on the regulatory framework. The damage caused by the policy is treated not as a warning but as justification.

Enforcement failure follows the same pattern. When laws are unenforceable at scale, when retailers openly flout regulations, consumers easily access banned products, and authorities lack the resources or capacity to control supply, this is normally considered a serious policy flaw. Enforcement feasibility is a core component of policy evaluation, not an afterthought.

But in the prohibitionist framework, enforcement failure is never evidence against the policy’s design. It is always framed as insufficient enforcement, insufficient funding, insufficient penalties, or insufficient political will. The conclusion is never “this policy does not work as intended”, but “we have not yet applied it forcefully enough”.

This is another classic sign of non-falsifiability. If a policy’s failure always implies the need for more of the same policy, then there is no imaginable outcome that would count as evidence against it. The policy becomes self-sealing.

At this point, the structure becomes clear. No empirical outcome is permitted to count against prohibition. Positive outcomes validate it. Negative outcomes are reclassified as reasons for intensification. Ambiguous outcomes are selectively interpreted. Conflicting data are sidelined or dismissed as methodologically flawed, biased, or morally suspect.

This is the hallmark of a closed epistemology: a system of belief that absorbs all evidence without risk of refutation. It is not that evidence is absent, but that its role has changed. Evidence is no longer used to test assumptions. It is used to rationalise conclusions that have already been reached.

The contrast with harm reduction approaches is instructive. Harm reduction makes explicit, testable claims. If providing lower-risk alternatives does not reduce smoking prevalence, does not reduce disease burden, does not improve cessation outcomes, or leads to worse population-level health, then the approach should be revised or abandoned. Those are falsifiable conditions.

They are uncomfortable precisely because they expose the policy to the possibility of failure. Harm reduction accepts that good intentions are irrelevant if outcomes are bad. That is why it attracts such hostility from ideologically committed prohibitionists: it competes on results rather than moral posture.

Prohibitionist rhetoric avoids this exposure entirely. Ask advocates what data would convince them that banning vaping was a mistake, and the question is either deflected or answered in ways that make falsification impossible. Smoking increases? That doesn’t count. Illicit markets? Expected side effect. Enforcement collapse? A reason for harsher laws. Youth experimentation declines, but adult smoking rises? Only youth outcomes matter. Adult cessation improves, but youth vaping exists. Still unacceptable.

Notice how the goalposts move depending on which outcome preserves the policy’s legitimacy. The standard is never applied consistently across populations, timeframes, or harms. What matters is not coherence but insulation.

When every conceivable outcome is compatible with the policy’s righteousness, the policy is no longer empirical. It is ideological. It is grounded not in outcomes but in a prior moral judgement: that non-medical nicotine use is inherently wrong, and that its elimination is desirable regardless of consequences.

This matters because public health claims carry authority. They justify coercion, restriction, surveillance, and punishment. They override individual autonomy in the name of the collective good. If those claims are shielded from falsification, then public health ceases to be a scientific enterprise and becomes a moral enforcement mechanism dressed in epidemiological language.

At that point, evidence no longer disciplines power. It serves it.

The question, then, is not whether prohibitionists can cite studies or statistics. They can, and they do. The question is whether there exists any real-world evidence that would cause them to reconsider their position. If the answer is no, then the debate is not about data at all. It is about values, institutional incentives, and the refusal to admit that a policy meant to protect health might be actively undermining it.

Once you see this, the contradictions make sense. Rising smoking does not trouble the framework because smoking reduction is not the true falsifier. Black markets do not trouble it because market realism is not the goal. Enforcement failure does not trouble it because coercion, not effectiveness, is the measure of seriousness.

What would trouble it, what must be avoided at all costs, is the admission that controlled, consumer-driven harm reduction might work better than prohibition. That possibility is excluded not by evidence but by ideology.

An ideology that cannot be falsified should never be allowed to govern public health.