The Joint Update Rule and Five Exploratory Tests
Authors/Creators
- 1. Phenikaa University
- 2. Faculty of Nursing, Widya Mandala Surabaya Catholic University
Description
The Correction Freeze Problem (CFP) proves four routes by which a corrective mechanism stops registering change. Each route has relatives in earlier work, so we ask what CFP adds. We argue that its distinctive content lies in the update rule that joins the routes: a belief is corrected only when the mechanism is active and the change is large enough to notice. The rule predicts that large changes are registered only when stakes look high, and that small changes are missed even then. We report five exploratory tests on U.S. data, each with its rules fixed in advance. In individual inflation forecasts, households responded strongly to large gasoline price moves when prices were high (about 0.76 percentage points for a 10% rise) and not detectably when prices were ordinary, meeting the strict form of the prediction for large changes, although its significance rests mainly on 2022. Splitting the forecast result into its two margins pointed the same way: when the rule says correction is on, fewer households kept their forecast unchanged and more revised in gasoline's direction, although only the first difference was significant. National searches for gasoline prices showed the same pattern in a graded form that the fixed rules judged inconclusive, echoing earlier support for the cost route in inflation searches and attention to Lake Mead. Two readings of Route 3 and one of hysteresis failed. The evidence is consistent with the joint rule but does not yet separate it from attention models with a fixed cost of attending. The joint rule is CFP's most promising prediction of its own and merits a registered test on new data.