Imagine two entrances at an event.
Twenty people are lined up on the left. No one is using the right entrance. The signs suggest both are valid, yet stepping toward the empty one feels strangely risky.
Then one person walks through the right entrance without a problem. Immediately, two or three others follow.
Physically, nothing changed. The doorway, rules, and distance stayed the same. Only one new piece of information appeared:
someone else already did it.
That tiny scene connects surprisingly well to multiple-choice quizzes, restaurant lines, office culture, social movements, wartime mobilization, propaganda, elections, and the fear of being the first person in your social circle to choose an unusual life.
Humans do not simply carry an “obey powerful leaders” operating system. The mechanism is subtler and often rational: we use other people’s behavior as information, calculate the cost of standing apart, and rely more heavily on social cues when uncertainty is high.
That mechanism helps civilizations coordinate. Under bad conditions, it can also help groups run in dangerous directions.
1. We watch both the answer and the crowd
Classic social psychology distinguishes two major motives for conformity.[1]
Informational social influence says: “I am not sure. They may know something I do not, so their behavior is evidence.”
Normative social influence says: “I still think I am right, but being the only dissenter may be embarrassing, costly, or socially painful.”
The same public behavior can therefore conceal very different private states.
In a true-or-false classroom quiz, seeing 90% of the room move to the opposite side can reasonably make you wonder whether you missed something. That is information.
Thinking “I still believe my answer, but I do not want to stand alone” is norm pressure.
Modern neuroscience reviews also find that deviation from group judgments recruits systems associated with conflict, aversive experience, valuation, and social evaluation, and that such responses can predict later conforming adjustments.[2]
So “reading the room” is not a uniquely Japanese oddity. Cultures differ in average levels and contexts of conformity, but using others as information and avoiding exclusion are broadly human.
2. The most interesting lesson from Asch is the power of one dissenter
In Solomon Asch’s classic line-judgment studies, even when the visual answer was obvious, participants followed a unanimous majority’s clearly wrong answer on roughly one third of critical trials.[3]
But independence remained common.
The experiment did not show that humans are programmable robots. It showed that many people resist, while unanimity creates real pressure.
When unanimity was broken by even one other person, conformity dropped sharply. In some variants, the crucial feature was not that the dissenter perfectly agreed with the participant; it was that the majority was no longer unanimous.
This is remarkably similar to the empty line.
Nobody uses the right entrance: “Maybe it is not allowed.”
One person uses it: “Oh. It is an available option.”
The first mover does not command anyone. The person makes the option socially visible and legitimate.
Bond and Smith’s 1996 meta-analysis of 133 Asch-type studies from 17 countries also found that conformity varies across cultures and historical periods, but is not confined to one society.[4]
3. A crowd can become self-reinforcing
Mark Granovetter’s threshold model scales this logic from individuals to collective behavior.[5]
Suppose A will act even if nobody else does. B acts if one person has acted. C needs two. D needs three.
Then:
A → A+B → A+B+C → A+B+C+D.
A small difference in the distribution of thresholds can produce dramatically different collective outcomes even when average preferences are similar.
That helps explain why a crowded line can attract more people despite being objectively slower.
“Many people are there” becomes evidence that the line is correct. The empty alternative starts to look suspicious precisely because it is empty.
People join because it looks correct, and it looks correct because people joined.
Centola and colleagues experimentally demonstrated a tipping point in an artificial social-convention setting. In their particular experiment, a committed minority around 25% could overturn an established convention, but 25% is not a universal law. Critical mass depends on network structure, incentives, and the type of norm.[6]
4. The frightening part: everyone does not even have to believe it
Public behavior and private belief can diverge.
Timur Kuran’s analysis of the 1989 East European revolutions emphasized the distinction between private preferences and preferences people publicly express.[7]
A person may privately dislike a regime but fear public opposition, so they behave as if they support it. A neighbor observes that behavior and concludes, “Maybe everyone else supports this.” The neighbor also stays silent.
A society can therefore reach a state in which many people believe that everyone supports something that many people privately dislike.
When public dissent suddenly becomes safer and visible, the process can reverse rapidly:
“Wait, you can say that?” “I think it too.” “So do I.”
A rapid change in visible norms does not necessarily mean millions of private minds changed simultaneously. Hidden preferences may simply have become public.
5. War and propaganda are more complicated than a national brainwashing switch
It is tempting to summarize wartime societies as: leaders demand war, the public believes them, defeat occurs, and everyone changes their mind.
History is messier.
Wartime Japan developed a broad mobilization system involving the National Mobilization Law, the National Spiritual Mobilization Movement, information control, and organized propaganda. Government documents from 1938 record this mobilization apparatus in detail.[8]
Yet identical outward cooperation could arise from different internal states: genuine support, belief shaped by the available information environment, fear of penalties to oneself or family, conformity, resignation, or strategic adaptation.
Research on Nazi-era German radio makes the same point from another direction. Adena and colleagues found that radio effects changed when control of political broadcasting changed. Nazi radio contributed to party recruitment and later anti-Semitic actions and denunciations, but its impact depended strongly on preexisting local anti-Semitism; in historically low-prejudice areas, effects could even move in the opposite direction.[9]
In Rwanda, Yanagizawa-Drott estimated that the violent RTLM radio broadcasts increased participation in the genocide, attributing roughly 51,000 perpetrators—about 10% of overall violence—to the station, while social spillovers into neighboring villages were estimated to be even larger than direct exposure effects.[10]
The scary mechanism is not magical remote control.
It is messages interacting with existing beliefs, changing visible behavior, and that behavior becoming new social information for others.
6. Postwar Japan was not “everyone militarist yesterday, everyone pacifist today”
U.S. observers in 1946 were struck by how rapidly Japanese society accepted surrender and occupation. One official report proposed several explanations: the Emperor’s surrender order, disillusionment with Japanese political and military leadership, and an ability to adapt to a radically changed situation.[11]
But this is an occupation-era American interpretation, not proof of an eternal Japanese personality trait.
Wartime Japan contained support, doubt, fear, opportunism, resignation, and resistance. Postwar Japan contained relief, shock, grief, anger, adaptation, and genuine value change.
Still, when institutions, schools, media, laws, leadership signals, and visible public speech all change together, the answer to “what can be said openly?” can change quickly.
The human brain did not need to be replaced overnight.
The inputs to the same social-learning machinery changed.
7. Elections show social influence—but not a magic control panel
The same logic raises an obvious political question: if people hear that a candidate is popular or likely to win, will they jump on the bandwagon?
Political science has studied this for decades.
In a classic minimal-information election experiment, poll numbers alone were not enough to reliably make voters switch; stronger qualitative cues were needed.[12]
Kalla and Broockman later synthesized 40 field experiments and added nine new ones on campaign contact and advertising in U.S. general elections. Their best average estimate of persuasive effects on candidate choice was approximately zero, with effects appearing in narrower conditions.[13]
So social influence is real, but people are not blank terminals that can be rewritten at will.
Partisanship, prior values, knowledge, social networks, source credibility, message fit, and reactance all matter.
The correct lesson is neither “social pressure is irrelevant” nor “manufacture a majority and control everyone.”
8. Why does an unusual life feel scary?
Bring the model back to personal decisions.
In a social environment where nearly everyone follows one standard career path, alternatives such as traveling while writing for income, combining project work with short-term jobs, moving frequently, or building a portfolio career can feel disproportionately dangerous.
At least four mechanisms contribute.
First, ambiguity aversion. “There is a 30% failure rate” is risk; “nobody knows the failure rate” is ambiguity. People tend to dislike unknown probabilities.[14]
Second, lack of social proof. One person who has lived that way for two years does not prove it will work for you, but it changes “impossible” into “observably possible.”
Third, status quo bias. Existing salary, routine, housing, and status are concrete; the benefits of an alternative future are hypothetical. Samuelson and Zeckhauser documented a systematic preference for preserving the status quo.[15]
Fourth, reputational exposure. Failure may cost more than money. The social cost of being seen to fail after doing something unconventional can affect trust, career expectations, and relationships.
That is why “I can afford the financial loss” is not enough. Money may be reversible while health, safety, reputation, legal status, or shared assets are not.
9. “Nobody does this” is neither a green light nor a red light
The useful question is not “Does anyone do this?”
It is:
Why is this not widely done?
Possible answers are very different:
- regulation makes it difficult;
- the economics are bad;
- safety is poor;
- required skills are rare;
- failure is hard to reverse;
- demand is weak;
- technology only recently made it possible;
- the behavior merely looks socially strange;
- it works, but most people do not want it;
- small groups already do it, but they are invisible to your reference group.
Only some of these are genuine danger signals.
The goal is not to throw away conventional wisdom. It is to stop treating conventional behavior as if it were direct evidence.
10. If you want to be first, build a cheap experiment instead of a heroic leap
The rational first mover is not necessarily fearless.
Often, the rational first mover is simply good at reducing uncertainty cheaply.
Effectuation research uses the idea of affordable loss: under deep uncertainty, decide what you can tolerate losing rather than pretending you can precisely forecast the upside.[16]
Diffusion research likewise finds that trialability—the ability to test an innovation on a small scale—and observability—the ability to see its results—are positively related to adoption.[17]
Applied to life experiments:
- Research the structural facts first: law, fixed costs, taxes, insurance, demand, and failure cases.
- Fix the maximum acceptable loss before estimating the dream upside.
- Test at the smallest meaningful scale.
- Increase commitment only as real information accumulates.
- Set stopping conditions before the experiment becomes emotionally expensive.
- Separate tail risks—debt, health, safety, illegality, shared assets, severe reputational damage—from ordinary expected-value calculations.
- Test reversibility of both the situation and other people’s expectations about you.
- If the experiment affects another person’s money, housing, time, safety, or reputation, treat it as a joint decision rather than a private experiment.
The goal changes from “be brave enough to jump” to:
buy information cheaply.
That is much more reproducible than courage.
Conclusion: society is a distributed computer, and the first mover manufactures precedent
Humans watch one another because no individual can independently calculate everything.
That is adaptive.
But once everyone uses everyone else as an input, feedback loops appear:
“Everyone does it, so it must be right.” → “It looks right, so more people do it.”
Lines, social norms, revolutions, propaganda, elections, and life choices differ enormously in stakes, but they share one structural fact:
other people’s behavior becomes data for the next person.
The same mechanism also works in reverse.
One person uses the empty entrance.
Someone behind them learns that it is allowed.
One person tests an unusual way of working, publishes the actual costs, failure conditions, and results, and makes the experiment observable.
The next person no longer faces pure ambiguity.
The first mover has created more than a success story.
They have created precedent: social information that lowers uncertainty for everyone who comes later.
So there is no need to become a contrarian for its own sake.
Keep “what everyone does” in one column and “what the evidence says” in another.
If the evidence shows genuine danger, stop.
If the only obstacle is lack of precedent, test cheaply.
Once one person walks through safely, the path is no longer an empty line.
References
Sources
- Morton Deutsch & Harold B. Gerard, “A Study of Normative and Informational Social Influences upon Individual Judgment,” Journal of Abnormal and Social Psychology 51(3), 1955, DOI: 10.1037/h0046408
- Haiyan Wu, Yi Luo & Chunliang Feng, “Neural Signatures of Social Conformity,” Neuroscience & Biobehavioral Reviews 71, 2016
- Solomon E. Asch, “Opinions and Social Pressure,” Scientific American, 1955; “Studies of Independence and Conformity,” Psychological Monographs, 1956
- Rod Bond & Peter B. Smith, “Culture and Conformity: A Meta-Analysis of Studies Using Asch’s Line Judgment Task,” Psychological Bulletin 119(1), 1996
- Mark Granovetter, “Threshold Models of Collective Behavior,” American Journal of Sociology 83(6), 1978
- Damon Centola et al., “Experimental Evidence for Tipping Points in Social Convention,” Science 360(6393), 2018
- Timur Kuran, “Now Out of Never,” World Politics 44(1), 1991
- Cabinet Information Department of Japan, Kokumin Seishin Sōdōin Jisshi Gaiyō, Vol. 1, 1938, National Diet Library; related National Mobilization Law materials
- Maja Adena et al., “Radio and the Rise of the Nazis in Prewar Germany,” Quarterly Journal of Economics 130(4), 2015
- David Yanagizawa-Drott, “Propaganda and Conflict,” Quarterly Journal of Economics 129(4), 2014
- U.S. Department of State, Foreign Relations of the United States, 1946, The Far East, Vol. VIII, document 159
- Daniel W. Fleitas, “Bandwagon and Underdog Effects in Minimal-Information Elections,” APSR 65(2), 1971
- Joshua L. Kalla & David E. Broockman, “The Minimal Persuasive Effects of Campaign Contact in General Elections,” APSR 112(1), 2018
- Unraveling Ambiguity Aversion,” Review of Economics and Statistics, 2023
- William Samuelson & Richard Zeckhauser, “Status Quo Bias in Decision Making,” Journal of Risk and Uncertainty 1, 1988
- Saras Sarasvathy, Stuart Read, Nick Dew & Rob Wiltbank, “Affordable Loss,” Strategic Entrepreneurship Journal, 2009
- Fred K. Weigel et al., “Diffusion of Innovations and the Theory of Planned Behavior in Information Systems Research: A Metaanalysis,” CAIS 34, 2014
