Conspiracy theorists hang onto theories despite the indicators you would expect to see if it were true not being there.
The argument
The explanation offered here is epistemological rather than psychological in the usual sense: what goes wrong is the way belief is revised, not the motive for holding it. People often believe in conspiracy theories because they hold on to degenerating research programs. A research programme degenerates when it survives not by successful prediction but by absorbing each disappointment with a further adjustment, so that it explains more and more while forecasting less and less. The test it fails is a straightforward one. If something is true, there are certain outcomes we can reasonably see along the way, because a genuine hypothesis commits whoever holds it in advance to things that ought to show up, and the absence of those things counts as evidence against it. Conspiracy theorists, when these outcomes do not appear, do not let go of the theories. Rather than treating a missing outcome as a reason to re-evaluate, they add whatever is needed to accommodate its absence, which keeps the theory intact at the cost of its ability to be wrong about anything at all. The example given concerns secrecy, and it is a prediction the theory cannot easily avoid making. The sheer amount of people that would be needed to keep a conspiracy secret over time is such that, for many alleged plots, the secret could reasonably be expected to be revealed. Somebody would have talked, or left a record, or been found out. When no revelation comes, theorists ignore this, and the silence is treated as further proof of the conspiracy's reach rather than as evidence that it did not happen, so they continue to cling to the degenerating research program. On this account the belief is sustained by an error in how evidence is handled rather than by paranoia or bad faith.
Premises
Counter-arguments
The Lakatosian language of "degenerating research programmes" describes how conspiracy believers reason but does not fully explain why they adopt or retain the belief in the first place. The sibling positions note that clinging to a failing theory despite disconfirmation is characteristic of motivated reasoning driven by distrust, social identity or psychological need — so the "cognitive error" account may be redescribing a symptom rather than identifying the cause. Moreover, not every unfulfilled prediction marks a degenerating programme: some conspiracies are genuinely hard to disconfirm because absence of evidence is exactly what a believer would expect, so labelling all such persistence a bare cognitive error risks begging the question against them.
Rejecting the premises
[Rejecting P2] That believers "cling" to a failing programme may reflect motivated reasoning rooted in distrust or identity rather than a purely cognitive error, so the cognitive-reaction framing may mistake a mechanism for a cause. [Rejecting P3] The "many people would reveal it" heuristic is itself contestable — believers argue secrets are kept by compartmentalisation — so treating their refusal to update as simple error assumes the conclusion. [Rejecting C] Even if the reasoning pattern is real, it describes how the belief persists, not why it is held; the other positions' social and psychological accounts may do the explanatory work the cognitive story only redescribes.