Effort Justification: Why Hard-Won Things Feel More Valuable

The effect starts from a contradiction a person cannot resolve by changing the facts. You went through something unpleasant to get into a group, and the group turns out to be dull. Those two thoughts do not sit together: nobody suffers for nothing. One of them has to move, and the suffering already happened in front of witnesses, so the thing that moves is the rating of the group. Leon Festinger's dissonance theory predicted exactly this, and Elliot Aronson and Judson Mills built the 1959 experiment to test it.

Three conditions have to hold for the effect to appear. The cost has to be real: Aronson and Mills found that their mild initiation produced ratings almost identical to no initiation at all, because a minute of mild embarrassment is not enough to need justifying. The choice has to feel free: a cost imposed on you can be blamed on whoever imposed it, which removes the pressure to revalue the thing. And the object has to be ambiguous enough to be read generously. A recorded discussion of nothing in particular can be rated 80 or 98; a flat tire cannot.

Effort justification is sometimes described as proof that suffering creates loyalty, and the field evidence does not support that reading. Hein Lodewijkx and Jef Syroit followed newcomers through real initiations in Dutch student groups and reported in 1997 that the members who described their initiation as more severe were not more attracted to the group, and were somewhat less so, with frustration, loneliness and low mood accounting for the link. The laboratory version holds up when the conditions above are met; the practical claim that a harsher initiation builds a stronger member does not.

What the effect does reliably explain is the revaluation of the thing achieved rather than the decision to continue. A member who revalues the group is not deciding whether to spend more; they are reporting a changed opinion about what they already have. That is the dividing line between this effect and the sunk cost fallacy, and it is why the two need separate counter-measures.

Elliot Aronson and Judson Mills published "The effect of severity of initiation on liking for a group" in the Journal of Abnormal and Social Psychology in 1959. Sixty-three college women volunteered for a series of group discussions and were assigned at random to three conditions of 21 each. In the severe condition, joining required an "embarrassment test": reading 12 obscene words and two explicit passages from contemporary novels aloud in front of the experimenter, who said he would judge how embarrassed the reader became. In the mild condition the reading was five words related to sex but not obscene. In the control condition there was no test at all.

Every participant then listened to what she believed was the group's meeting in progress, over headphones, and every participant heard the same tape: three students discussing secondary sex behavior in lower animals as dryly as the authors could arrange, contradicting themselves, mumbling and trailing off. Afterwards each woman rated the discussion and the group members on 14 written scales running from 0 to 15, plus three spoken ratings.

On the nine scales about the discussion, the means were 80.2 for the control condition, 81.8 for the mild condition and 97.6 for the severe condition. Summed across all 17 scales the figures were 166.7, 171.1 and 195.3. The gap between severe and control was reliable, the gap between severe and mild was smaller but still reliable, and the gap between mild and control was not there at all. One participant declined the test and was excluded, and one who guessed the setup was dropped.

Harold Gerard and Grover Mathewson replicated the design in 1966 with a cost that had nothing to do with embarrassment. Instead of reading aloud, participants took electric shocks, described as a test of emotionality, either painful or barely noticeable, and then heard the same dull recorded discussion. The participants who had taken the painful shocks rated the discussion and the group higher, which answered the main objection to the original study: that the 1959 result came from relief at having survived something humiliating, or from the sexual content of the reading, rather than from the effort itself.

The pattern turns up wherever a costly route leads to something whose value is a matter of opinion. The rows below are illustrations, not findings from the studies cited on this page.

The counter is to get a reading of the thing that does not pass through what it cost you.

• Write down what you expect to get, before you pay the cost. A written expectation cannot be revised quietly once the effort has been spent.
• Ask the stranger's question: what would I say about this if it had been handed to me for nothing. If the answer is different from your own rating, the difference is the price of entry.
• Separate the appraisal from the decision. "Was it worth it" is about the past and does not need an answer. "Would I choose this again from here" is the question that affects what you do next.
• Look for a measure outside the experience: what you can do now, what it would cost to replace, what people who got in easily say about it.
• When you are the one designing the door, notice what you are buying. Harder entry does not reliably produce more committed members, and in Lodewijkx and Syroit's field studies it went with frustration and low mood.
• Expect a mild cost to do nothing. The effect needs a real one, which is also why small fees and short forms are not a way to manufacture commitment.

Three nearby effects all involve something you have put in, and they differ in what moves.

• It is not the sunk cost fallacy, which is continuing to spend because of what has already been spent. Effort justification changes your opinion of the thing; the sunk cost fallacy changes your decision about the next payment, and a person can show either one without the other.
• It is not the IKEA effect, which is the specific premium people place on objects they built themselves. Norton, Mochon and Ariely reported in 2012 that builders valued their own assembled items above identical pre-assembled ones, and found the premium only when the assembly was completed successfully, which effort justification alone does not predict.
• It is not post-purchase rationalization, the reasons assembled after a purchase to defend the money spent. That one is about justifying a choice; this one is about justifying a cost, and the cost does not have to be money.