Do We Want Our Predictions to Fail?

A version of this article first appeared on Quartz.

In her 1951 book The Origins of Totalitarianism, in the section on propaganda, Hannah Arendt discusses a concept she calls “infallible prediction”:

The chief qualification of a mass leader has become unending infallibility; he can never admit an error. The assumption of infallibility, moreover, is based not so much on superior intelligence as on the correct interpretation of the essentially reliable forces in history or nature, forces which neither defeat nor ruin can prove wrong because they are bound to assert themselves in the long run. Mass leaders in power have one concern which overrules all utilitarian considerations: to make their predictions come true.

Arendt’s work is in a phase of popularity. I, for instance, discuss her concept of the ‘banality of evil’ here and here. But little attention has been paid to infallible prediction. It has a lot to teach us.

1. Seeking confirmation

We as voters value prescience in our leaders. Those who predict events in politics (like, say, Brexit or a rigged election) are attempting to signal political expertise. And surely we want leaders who have political expertise—though we might disagree about what constitutes it. Hence, predictions serve as a means of acquiring and consolidating power. We clamor for them.

However, to make a prediction is to open oneself up to error. It puts one’s worldview and expertise on the line. Just as a confirmed prediction can be politically useful, a failed prediction can be politically embarrassing. Unfortunately, failed predictions rarely prompt self-reflection. More often, politicians get away with amending, qualifying, and hedging their predictions into something that becomes vacuous or trivially true. To admit error is to surrender political authority. So we see an urge never to admit error.

It is a common tendency, found in each of us. Yet there is something deeply worrying about it. Worldviews are difficult to dislodge. And in politics, careers depend on the validity of certain worldviews. Given how fundamental they are, the worldviews themselves are at play in the assessment of predictions. In other words, predictions originate from a given set of beliefs about the world. And the confirmation or disconfirmation of the prediction is filtered through the same set of beliefs. The beliefs that comprise our political worldviews are closely tied to our identities, and so it would take a lot for us to question them, even if the world has given us good reason to. Reality becomes whatever squares with the prior beliefs. Confirmation pops out vividly, whereas the problematic cases fade into the background and disappear. Complicated issues become simple. Simple issues become complicated.

II. Creating Confirmation

What happens when you add power to this tendency? Arendt argues that the predictions of mass leaders are different. The concept of “infallible prediction” is, one would think, an outright contradiction. A prediction must be fallible if it is to be a prediction at all. If I predict that I will eat dinner, because I have the power to make it true, I am not in any meaningful sense making a prediction. Predictions are about an outside world that I don’t control. Fallibility assumes that we do not have the power to manipulate and mold the world into what we want. Powerful political leaders have that ability. If someone were to predict, say, a war with China in 5-10 years, given enough political power, this prediction becomes something else.

Arendt argues that it is in fact a statement of intent. Leaders announce intentions in the form of prophecy. Given how predictions function in politics, it is a highly effective way to consolidate power. When the war comes, the leader looks wise and prescient and reaps the rewards. They can proclaim that their expertise is the true expertise. The defenders of the leader will point to the confirmation as validation of their support. Any critique of the leader will seem misplaced.

Plus, by phrasing intentions as predictions, leaders can hide from blame. Recall that predictions are about parts of the world that I don’t control (like the outcome of a sports game). A prediction is daring because we are powerless to bring about what we want. Hence, the actions of the leader appear as inevitable forces of nature, not the free choices of an agent. When the war comes, as Arendt says, “the ‘prophecy’ becomes a retrospective alibi: nothing happened but what had already been predicted.” In hindsight, it was fated all along.

Simply put, mass leaders use their power to make their predictions come true.

III. US

What does all this mean for us? At the core of infallible prediction is the urge to confirm one’s worldview and be unresponsive to evidence. It comes from a desire to pick out the convenient pieces of reality or, with the addition of power, to create a reality in the image of a worldview. It comes from the urge to hold up one’s confirmed predictions for all to see.

How many of us fit that description right now? We find the desire in ourselves when we read old philosophy books in search of confirmation of our views—when we find quotes to serve as convenient condemnations of our enemies. We want our predictions to be true. So we pick out the pieces of reality that enable us to portray prescience. There is pride in it. But when our predictions are dire, we face a dilemma. Do we want to be right about something dreadful?

The critical next question is about power. Do we have the ability to mold reality into the confirmation of our predictions? Perhaps political power is no longer grounded in the people. Perhaps it never was. We face the same dilemma regardless: do we want to pick out and create pieces of reality in order to sit in pride with our confirmed predictions, or are we willing to work hard to ensure that our predictions turn out wrong?