Fortunately for my adopted state of Florida, Irma weakened considerably as it moved northward. When it reached my adopted city of Tallahassee, it was barely a tropical storm. While it did some damage, it was nothing compared to last year’s storm. While this was a good thing, there can be a very minor downside when dire predictions turn out to be not so dire.
The problem is, of course, that people might take such dire predictions less seriously in the future. There is even a term for this: hurricane fatigue. When people are warned numerous times about storms and they do not prove as bad as predicted, people tend to get tired of going through the process of preparation. Hence, they tend to slack off in their preparations—especially if they took the last prediction very seriously and engaged in extensive preparations. Such as buying absurd amounts of bottled water. The problem is, of course, that the storm a person does not prepare for properly might turn out to be as bad or worse than predicted. Interestingly enough, inductive reasoning is the heart of this matter in two ways.
Inductive reasoning is, of course, logic in which the premises provide some degree of support (but always less than complete) for the conclusion. Inductive arguments deal in probability and this places them in contrast with deductive arguments—they are supposed to deal in certainty. That is, having all true premises in a deductive argument is supposed to guarantee a true conclusion. While there are philosophers who believe that predictions about such things as the weather can be made deductively, the best current reasoning only allows inductive reasoning regarding weather prediction. To use a simple illustration, when a forecast says there is a 50% chance of rain, what is meant is that on 50% of the days like this one it rained. This is, in fact, an argument by analogy. With such a prediction, it should be no more surprising that it rains than it does not.
While the computer modeling of hurricanes is rather complex, the predictions are still inductive in nature: all the evidence used in the reasoning can be true while the conclusion can still be false. This is because of the famous problem of induction—the gap between the premises and the conclusion means that no matter how strong the reasoning of an inductive argument, the conclusion can still be false. As such, any weather prediction can turn out to be false—even if the prediction is 99.99% likely to be accurate. As such, it should be expected that weather predictions will often be wrong—especially since the models do not have complete information and are limited by the available processing power. That is, there is also a gap between reality and the models. There is also the philosophical question of whether the world is deterministic or not—in a deterministic world, weather would be fully predictable if there was enough information and processing power available to create a perfect model of reality. In a non-deterministic world, even a perfect model could still fail to predict what will happen in the real world. As such, there is both a problem in epistemology (what do we know) and metaphysics (what is the nature of reality).
Interestingly enough, when people start to distrust predictions after past predictions turn out to be wrong, they are also engaging in inductive reasoning. To be specific, if many predictions have turned out to be wrong, then it can be reasonable to infer that the next prediction could be wrong. That is certainly reasonable and thinking that an inductive argument could have a false conclusion is no error.
Where people go wrong is when they place to much confidence in the conclusion that the prediction will be wrong. One way this can happen is through a variation in the gambler’s fallacy. In the classic gambler’s fallacy, a person assumes that a departure from what occurs on average or in the long term will be corrected in the short term. For example, if a person concludes that tails is due because they have gotten heads six times in a row, then they have committed this fallacy. In the case of the “hurricane fallacy” a person overconfidently infers that the streak of failed predictions must continue. The person could, of course, turn out to be right. The error lies in the overconfidence in the conclusion that the prediction will be wrong. Sorting out the confidence one should have in their doubt is a rather challenging matter because it requires understanding the accuracy of the predictions.
As a practical matter, one way to address hurricane fatigue is to follow some excellent advice: rather than going through mad bursts of last second preparation, always be prepared at the recommended minimum level. That is, have enough food and water on hand for three days and make basic preparations for being without power or evacuating. Much of this can easily be integrated into one’s normal life. For example, consuming and replacing canned and dried goods throughout the year means that one will have suitable food on hand. There are also one-time preparations, such as acquiring some crank-powered lights, a small solar panel for charging smart phones, and getting a basic camp stove and a few propane canisters to store.
This does lead to a final closing point, namely the cost of preparation. Since I have a decent income, I can afford to take the extra steps of being always ready for a disaster. That is, I can buy the lights, stove, propane, and such and store them. However, this is not true of everyone. When I was at Publix before the storm, I spoke to some people who said that it was hard for them to get ready for storms—they needed their money for other things and could not afford to have a stockpile of unused supplies let alone things like solar panels or generators. The upfront cost of stockpiling in preparation for the storm was also a challenge—there are, as far as I know, no emergency “storm loans” or rapid aid to help people gear up for impending storms. No doubt some folks would be terrified that storm moochers would be living fat on the public’s money during storms. However, storm aid does sound like decent idea and could even be cost saver for the state. After all, the better prepared people are before the storm, the less the state and others must do during and after the storm.