Does thinking about death make people harsher? What the evidence now says

The idea: we build meaning to hold back fear

In 1973 the anthropologist Ernest Becker argued in The Denial of Death that much of what people do, from religion to national pride to career ambition, works as a shield against knowing we will die. The book won the Pulitzer Prize in 1974.

Three psychologists, Jeff Greenberg, Sheldon Solomon and Tom Pyszczynski, turned Becker's argument into a testable theory called terror management theory. Its central prediction: if you remind people of their own death, they defend their worldview harder. They reward people who uphold their values and punish people who break them.

The judges and the bond

The best-known test ran in 1989. Twenty-two municipal court judges in Tucson, Arizona filled in a questionnaire. For half of them, the questionnaire asked what they thought would happen to them when they died and what emotions that thought stirred. The other half answered neutral questions. Then every judge read the same hypothetical case, a woman arrested for prostitution, and set a bond.

Judges who wrote about death set an average bond of $455. The others set $50. Same case, same law, and a nine-fold difference. The same paper ran five more experiments with students and found the pattern again: harsher punishment for people who broke cultural rules, bigger rewards for a hero who upheld them.

It is a striking result, and it spread fast. Textbooks, jury consultants and popular books still repeat it.

Two decades of support

By 2010 the idea rested on a large body of work. A meta-analysis that year pooled 277 experiments from 164 articles and found a moderate average effect of death reminders on how strongly people defended their views and their self-esteem. The effect looked bigger in American samples, in college students, and when researchers waited a few minutes between the death reminder and the measurement.

Those same patterns carry a warning. An effect that shows up mostly among American undergraduates, and only under certain timing, may depend on details that small studies cannot pin down.

Then the big replications

Psychology's replication crisis hit this theory directly.

Study	What it did	What it found
Many Labs 4 (Klein et al., 2022)	17 labs, 1,550 participants. Half the labs worked from the published method; half got step-by-step advice from the original authors.	No mortality salience effect under any condition, with or without the original authors' help.
Schindler, Reinhardt and Reinhard (2021)	A pre-registered study with 1,356 participants, plus an analysis of their earlier experiments.	No significant effect; the statistics favoured no effect at all.
Schindler et al. (2023)	A meta-analysis of one popular strand of the research, corrected for publication bias.	The raw effect looked moderate. After correction it shrank, in several analyses to nothing.

Publication bias means studies that find an effect get published and studies that find nothing sit in drawers. Pool only the published ones and an effect can look solid when it is small or absent. The 2010 meta-analysis could not fully correct for this.

So what is true?

Three things hold up.

• The judges study happened. Its numbers are real. But it used 22 judges and one hypothetical case. A study that small can produce a large difference by chance.
• The broad claim is not settled. The largest, most careful tests, including ones designed with the original authors, did not find the effect. That does not prove death never shapes judgement. It means a quick questionnaire about death does not reliably make people harsher in a lab.
• Becker's idea stands as philosophy. Whether fear of death drives culture is a question about meaning. A failed lab effect neither proves nor disproves it.

Why this matters beyond psychology

The judges study gets quoted in arguments about courts, policing, elections and war. It is tempting to use: it seems to show that judgement bends to something irrelevant. Sturdier research makes that point better. (How hindsight and emotion colour blame.)

But a claim should carry the weight its evidence can bear. A single 1989 study of 22 judges, in a field whose central effect has since failed large replications, cannot carry an argument about how every justice system works. Anyone building a case on it hands the other side an easy rebuttal.

How to check a famous psychology finding yourself

1. Find the original paper. Note how many people took part. Under 50 is small.
2. Search the finding's name with "replication" or "Many Labs". Large multi-lab projects since 2015 have tested many classic effects.
3. Look for a meta-analysis, then check whether it corrects for publication bias. Newer ones usually do.
4. Ask whether the result came from students in one country. If so, it may not travel.
5. Quote the finding with its size and its status: "a 1989 study of 22 judges found…; large replications since have not found the effect."