Why We Are Poor at Predicting How We Will Feel

Introduction

Almost every decision rests on a prediction about feeling.

Take the job or stay put. Have the difficult conversation or let it go. Move, leave, apply, decline. In each case the reasoning runs through an estimate of how the outcome will feel, and for how long.

People make these estimates constantly and with considerable confidence. Researchers began testing them against what actually happened, and found a specific and consistent pattern of error, one that has direct bearing on why difficult things get avoided.

What People Get Right, and What They Get Wrong

The forecasts are not wrong in every respect.

People are reasonably good at predicting the direction of a feeling. Asked whether losing a job would feel bad, they say yes, and it does. Asked whether a promotion would feel good, they say yes, and it does.

Where the predictions break down is on intensity and duration.

Across studies covering academic rejection, romantic breakups, election results and other outcomes, people consistently overestimated how long the emotional effect would last. The distress was real. It was also shorter-lived than the person had expected, often considerably.

The same applies to positive events. Anticipated good outcomes tend to produce less enduring improvement in mood than forecast, which is part of why achievements can feel oddly flat.

Researchers named this the impact bias: a systematic tendency to overestimate the emotional impact of future events in both directions.

The Coping You Do Not Know You Will Do

The explanation offered for the durability part of the error is the more interesting piece of the theory.

When people imagine a bad outcome, they imagine the event. What they do not imagine is themselves responding to it: reinterpreting, finding explanations, adjusting expectations, discovering compensations, and generally getting on with it.

Daniel Gilbert and colleagues described this as a psychological immune system: a set of largely automatic processes that reduce the impact of adverse events, operating without deliberate effort and, crucially, without being visible to the person in advance.

The result is that forecasts systematically omit the forecaster. The event is modelled in detail; the person's response to it is not modelled at all.

The forecast leaves out the forecaster: specifically, everything they will do to make the outcome bearable.

Why the Event Looms So Large

A second contributor concerns attention rather than coping.

When asked to predict how an event will affect them, people focus on that event. This is reasonable; it is the question. But it produces a distorted estimate, because in the actual future the event will be one item among many.

Researchers termed this focalism. Someone imagining life after a relationship ends pictures the absence, in detail. They do not picture the commute, the work deadline, the boiler that needs fixing, the friend's birthday. All of these will compete for attention and dilute the emotional weight of the focal event, and none of them appear in the forecast.

Studies that prompted people to consider what else would occupy their time produced more moderate and more accurate predictions, which supports the account and suggests the error is not fixed.

Where the Story Has Been Overstated

This research has entered popular circulation in a considerably stronger form than the evidence supports, and the corrections matter.

The most-cited early study reported that lottery winners were not much happier than controls and that people with serious spinal injuries were happier than expected. It is a genuinely important study and it involved very small samples: 22 lottery winners, 29 people with paralysis and 22 controls, 73 in total, with a response rate as low as 41 per cent in one group. It should never have carried the weight placed on it.

More importantly, subsequent research using large longitudinal panels has found that adaptation is real but frequently incomplete. Following people through the onset of severe disability, or through unemployment, has shown lasting reductions in life satisfaction that do not return to baseline. People adapt more than they expect. They do not always adapt fully, and for some events they adapt very little.

There is also a methodological challenge to the forecasting research itself. Some researchers have argued that part of the apparent bias reflects participants misunderstanding what they were being asked: predicting their overall mood rather than their feelings specifically about the event, and that when the question is posed more precisely, forecasting accuracy improves substantially.

The impact bias has not been overturned by this. But the honest version is narrower than the popular one: people underestimate their adaptation, the effect is more reliable for duration than intensity, and it varies a great deal by event.

What This Does Not Mean

One reading of this research would be genuinely harmful, and it needs ruling out explicitly.

It does not mean that present distress is exaggerated, or that anyone should be told their suffering will pass. The findings describe averages across groups. They predict nothing about any individual, they say nothing about how long a particular person's grief or difficulty will last, and the evidence on incomplete adaptation shows that for some events and some people it does not resolve.

If you are in the middle of something painful, this article is not an argument that you are overestimating it.

It is an account of how predictions about future feeling go wrong, which is a different thing from an account of how feelings actually go, and applies to the forecast rather than to the experience.

Why It Matters for Avoidance

The practical significance lies in what predictions are used for.

Avoidance is almost always built on an affective forecast. The reason a person does not make the call, attend the event or raise the subject is rarely a belief that something terrible will happen externally. It is a belief about how it will feel, that the discomfort will be unbearable, or will not stop.

That belief is a prediction of exactly the kind this research examines. And the research suggests it errs in a specific direction: overestimating intensity and, particularly, duration.

This connects to what is known about approaching feared situations. The value of doing the difficult thing lies in generating information that contradicts the prediction, and a substantial part of that information concerns feeling. Not only the disaster did not occur, but the discomfort was survivable and it ended.

Avoidance prevents both discoveries, which is why the forecast never gets corrected and why it tends to become more extreme rather than less over time.

Making a Better Forecast

Two adjustments follow from the research, and both are modest.

The first is to include yourself in the prediction. When imagining a difficult outcome, the picture almost always omits your own response to it: the adjusting, reframing and problem-solving that will begin more or less immediately and that has begun after every previous difficult outcome in your life.

The second is to widen the frame. The question is not how this event will feel, but how a Tuesday six months from now will feel, containing this event alongside everything else a Tuesday contains. Those are different questions with different answers, and the first is the one people habitually ask.

None of this makes dread irrational. Some things are genuinely as bad as anticipated, and the research is about a tendency rather than a rule.

But when a prediction about future feeling is the main reason something is being avoided, it is worth noticing that the prediction comes from a system with a known and directional flaw, and that the only way to find out is the thing being avoided.

If avoidance is significantly restricting your life, that is worth raising with a GP or a qualified professional. It responds well to structured support, and it tends to expand rather than resolve when left alone.