Are Crowds Wise or Mad?

Are Crowds Wise—or Mad?

By Alan Hall, originally published in the June 2011 Socionomist Download the Complete Issue (924 KB)

Study Confirms That Herding Undermines the Wisdom of Crowd Effect

Scottish journalist Charles MacKay1 got it right in 1841; crowds are mad. And now we have evidence from the laboratory.

A new study, “How social influence can undermine the wisdom of crowd effect,” by Jan Lorenz, Heiko Rauhut, Frank Schweitzer and Dirk Helbing, provides important support for socionomic theory. It confirms that social interaction and resulting herding severely undermine the utility of the wisdom of crowd effect (WCE). The experiments show that WCE is reliable only when it derives from the aggregated knowledge of independent individuals.

What is the Wisdom of Crowds Effect?

In other words, it is useful only when crowds are not acting like crowds at all.

The study shows that “even mild social influence” is sufficient to undermine the utility of WCE. It also provides the important new finding that social interaction produces convergent opinions and undue increases in confidence in those opinions. The results support the socionomic case that herding is a major factor behind asset pricing.

When Does Wisdom Become Madness? Here’s a summary of the findings:

  • Groups in which individuals have no knowledge of others’ estimates have more diverse opinions and produce a wider range of responses that is more likely to include the correct answer.
  • Groups in which individuals have knowledge of others’ estimates have less diverse opinions and produce a narrower range of responses that is less likely to include the correct answer.
  • Most important, individuals in the second kind of group develop greater confidence in their estimates. They herd as a result.

Interacting crowds exhibit convergent opinions even when there are factual answers to questions. Socionomic theory posits that when there is no objective answer to a question, such as, “What is this stock really worth,” social interaction further amplifies the convergence of opinions. In financial markets, socionomics proposes that the result is non-rational valuation, Elliott wave patterns, and—at certain points in those patterns—manias and crashes.

Lorenz et al’s study asked participants various questions, each with a factual answer. For example, one question was, “How many murders were registered in Switzerland in 2006?” The researchers asked the participants the same question five times via a computer system that recorded all the answers.

The participants were divided into three groups: those who were given no information about past answers, those who were informed of the answers of participants during the preceding round, and those who were informed of the trajectory of participants’ answers from all prior rounds.

The study uncovered several socionomically important items:

  1. The crowd becomes no better at determining the correct answer when it has information about others’ answers. Sometimes it becomes worse. The study’s lead author, Jan Lorenz, elaborated, “On average there is no [net] effect.”3
  2. When participants are informed of other participants’ estimates, the range of revised estimates narrows dramatically. The authors write, “In the no information condition, there is no convergence of estimates, whereas estimates converge in the aggregated and full information conditions.”3
  3. Finally, as participants gain information about others’ estimates, “individuals’ confidence is substantially and significantly boosted in the aggregated and full information conditions in comparison with the control condition without social influence.”4 Lorenz went on to explain: “Large variance in individual estimates serves as a warning sign for low certainty. …This signal gets washed away by social influence, but we may not recognize [it] because some variance remains.”3

Social Influence Leads to Opinion Convergence


Social Influence Leads to Opinion Convergence: Sample results from the Lorenz et al study. Individuals in the group on the left were given no information about other participants’ answers; the group on the right was told the range of answers of the group from the preceding round. The five color blocks display the “core ranges” of the answers in each round, with light yellow the largest and red the narrowest range. Note how the diversity of opinions narrowed in the group on the right but not in the group on the left.


The authors noted, “the wisdom-of-crowd indicator tends to decline over time under conditions of social influence. This effect is substantial and statistically significant for all questions.”4

In other words, when social influence was introduced, formerly independent individuals herded more tightly each time they reconsidered a question, and the benefit of aggregating multiple estimates was lost. The degradation of WCE occurred even though the survey questions had a factually correct answer.

The authors wrote, “Presumably, herding is even more pronounced for opinions or attitudes for which no predefined correct answers exist.” We wholeheartedly agree with the authors. Prechter and Parker argued in 2007 that herding becomes far stronger with respect to financial-valuation questions—which have only subjective answers. In “The Financial/Economic Dichotomy in Social Behavioral Dynamics: The Socionomic Perspective” (2007), authors Robert Prechter and Wayne D. Parker posited that humans naturally tend toward consensus—especially under conditions of uncertainty—because they have evolved to herd as a primal survival tactic. The Law of Patterned Herding describes crowd behavior in situations where social influence supersedes individual reason:

Social systems comprising homogeneous agents uncertain about other agents’ valuations that are critical to survival and success provide a context in which an endogenously regulated aggregation of unconscious herding impulses constitutes a pattern of social mood, which in turn motivates social actions.5

Herding may facilitate individual survival in the evolutionary sense, but it can wreak havoc in the social arena. William H. Whyte in 1952 coined the term “groupthink,” which he described as “rationalized conformity—an open, articulate philosophy which holds that group values are not only expedient but right and good as well.”6 Research psychologist Irving Janis tied groupthink to a number of American foreign policy disasters and other faulty decisions in history. Groupthink suppresses both the expression and subsequent evaluation of non-core views.

When the Wisdom of Crowd Effect Fails The WCE becomes less useful or disappears when estimates converge due to herding. Groups tend to coalesce around select but uninformed opinions and develop increased confidence in those opinions.

Why does this happen? One reason is that WCE is a statistical phenomenon, not a social/psychological effect. The authors wrote, “As social influence among human group members may trigger individuals to revise their estimates, it can have a substantial impact on the statistical wisdom-of-crowd effect in societies.”

Lorenz told us:

It is not rational for an external observer to rely too much on the wisdom of crowd effect when the group is under social influence, especially with respect to certainty about the fact that the aggregate outcome is central in the cloud of estimates.3

When the Wisdom of Crowd Effect Works WCE is useful when individuals form estimations independently. Heterogeneity of opinion, according to the authors, “generates a more accurate aggregate estimate than the estimates of single lay or expert decision makers.” Even this benefit, however, is limited to questions that have factual answers, as questions with subjective answers have no objective facts against which to measure accuracy. The study’s authors agree:

[C]onsensus formation has often been investigated for questions for which there are no well defined correct answers … . The herding effects created in this way prevent an objective measurement of quality. Therefore, such settings do not reveal whether social influence works in favor or to the disadvantage of the wisdom of the group.

How Herding Undermines WCE The study’s authors describe three ways in which knowledge about others’ estimates narrows diversity of opinions and undermines WCE.

  1. Social Influence Effect—The authors write, “Social influence promotes a convergence of estimates [and] strongly reduces the group’s diversity without significantly reducing its collective errors.” They cite research on hit songs as an example: It has been demonstrated that almost any song of average quality may become a hit if social influence is introduced by publishing the number of downloads.7 In this case, the popularity of a song and its perceived quality emerge through the process of interactive downloading and rating. The herding effects created in this way prevent an objective measurement of quality.
  2. Range Reduction Effect—Under social influence, estimates tend to be more narrowly distributed. Peers and the broader social climate influence even experts, and thus expert opinion is susceptible to convergence and narrow distribution as well. Decision makers who seek counsel from such experts will select from a narrower range of opinions. None of the expert opinions may be correct.
  3. Confidence Effect—Increased confidence is a psychological consequence of the previous two statistical effects. Social influence produces opinion convergence, which bolsters confidence and, according to the authors, “undermines the wisdom of crowds by boosting the subjective and decreasing the objective reliability of the crowd.”

WCE in Finance? Our socionomic model implies that social influence, opinion convergence and increased confidence about the direction of prices all intensify near significant turning points. This is why money flows, sentiment indicators and studies such as Frazzini and Lamont (2006) show that investors tend to buy late into rallies and sell near market lows.8

While extremely one-sided market opinion is slightly more prevalent during market advances, it is more focused at market lows, when fear and pessimism are also highly concentrated. Strong opinion convergence seems to be a better timing tool at lows than at highs. For example, Figure 2 shows that since 2000, the 30-period moving average of the S&P DSI (, a poll of short-term S&P Futures traders, has dipped below 22 percent bulls eight times, each time an important low. This suggests that the strength of opinion convergence fluctuates with social mood and at multiple degrees of scale.

Opinion Convergence and Stock Prices

Figure 2

What This Means for Socionomics This research demonstrates that interacting crowds produce and embrace erroneous conclusions. The undermining of WCE under social influence supports the socionomic premise that unconscious herding drives crowd behavior in financial markets, a context where both interaction and uncertainty are among the defining features of the landscape. Investors herd all the time, relying on rationalization over reasoning, the latter being the province of independent minds. Herding produces subjective valuation, which is subject to swings in social mood.

In the past decade, researchers have more frequently questioned and tested traditional financial and economic ideas. Their results have produced a succession of new findings consistent with socionomic theory. The study by Lorenz et al is an important contribution to this growing body of knowledge.

Alan Hall writes   for The Socionomist.


1MacKay, C. (1841). Extraordinary popular delusions and the madness of crowds. 2Galton, F. (1907). Vox populi. Nature 75:7. 3Lorenz, J.; email exchange with The Socionomist (June 8, 2011). 4Lorenz, J., Rauhut, H., Schweitzer, F. & Helbing, D. (2011). How social influence can undermine the wisdom of crowd effect. Proceedings of the National Academy of Sciences. 5Prechter, R. & Parker, W. (2007). The financial/economic dichotomy in social behavioral dynamics: the socionomic perspective. The Journal of Behavioral Finance, 8 (2), 84-108. 6Safire, W. (2004, August 8). The way we live now: 8-8-04: on language; groupthink. The New York Times, Retrieved from 7Salganik MJ, Dodds PS, Watts DJ (2006). Experimental study of inequality and unpredictability in an artificial cultural market. Science 311:854-856. 8Frazzini, A., & Lamont, O. (2008). Dumb money: mutual fund flows and the cross-section of stock returns. Journal of Financial Economics, 88. Retrieved from

Socionomics InstituteThe   Socionomist is designed to help readers understand and anticipate   waves of social mood. We also present the latest essays in the field of socionomics,   the study of social mood; we anticipate that many of the hypotheses will be   subjected to scientific testing in future scholarly studies.

The Socionomist is published by the Socionomics   Institute, Robert R. Prechter, Jr., president. Alan Hall, Ben Hall, Matt   Lampert and Euan Wilson contribute to The Socionomist. Mark Almand,   executive editor. Chuck Thompson, editor.

We are always interested in guest submissions. Please email    manuscripts and proposals to Ben Hall via   Mailing address: P.O. Box 1618, Gainesville, Georgia, 30503, U.S.A. Phone:  770-536-0309.

All contents copyright © 2011 Socionomics Institute.    All rights reserved. Feel free to quote, cite or review, giving full credit.  Typos and other such errors may be corrected after initial posting.

For  subscription matters, contact Customer Service: Call 770-536-0309  (internationally) or 800-336-1618 (within the U.S.). Or email

For  our latest offerings: Visit our website,,  listing BOOKS, DVDs and more.

Correspondence is welcome, but volume of mail often precludes    a reply. Whether it is a general inquiry, socionomics commentary or a research  idea, you can email us at

Most economists, historians and sociologists presume that    events determine society’s   mood. But socionomics hypothesizes the opposite: that social mood determines   the character of social events. The events of history—such as investment   booms and busts, political events, macroeconomic trends and even peace and war—are   the products of a naturally occurring pattern of social-mood fluctuation. Such   events, therefore, are not randomly distributed, as is commonly believed, but   are in fact probabilistically predictable. Socionomics also posits that the   stock market is the best available meter of a society’s aggregate mood,  that news is irrelevant to social mood, and that financial and economic decision-making  are fundamentally different in that financial decisions are motivated by the  herding impulse while economic choices are guided by supply and demand. For  more information about socionomic theory, see (1) the text, The     Wave Principle of Human Social Behavior © 2011, by Robert Prechter;  (2) the introductory documentary History’s     Hidden Engine; (3) the video Toward       a New Science of Social Prediction, Prechter’s 2004 speech before   the London School of Economics in which he presents evidence to support his   socionomic hypothesis; and (4) the Socionomics Institute’s website,  At no time will the Socionomics Institute make specific recommendations about  a course of action for any specific person, and at no time may a reader, caller  or viewer be justified in inferring that any such advice is intended.

Leave a comment

Comment Here:

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: