Why do revolutions happen? Why are they so difficult to anticipate? Some of the most instructive answers point to three factors: (1) preference falsification on the part of rebels or revolutionaries, (2) diverse thresholds for revolutionary activity, and (3) social interactions that either do or do not trigger the relevant thresholds. Under conditions of actual or perceived injustice or oppression, true preferences and thresholds are probably impossible to observe; social interactions are impossible to anticipate. Even if we could observe factors (1) and (2), the challenge of anticipating factor (3) would make it essentially impossible to foresee revolutions. For all their differences, and with appropriate qualifications, the French Revolution, the Russian Revolution, the fall of Communism, and the Arab Spring were unanticipated largely for these reasons. And in light of factors (1), (2), and (3), it is hazardous to think that the success of successful revolutions is essentially inevitable. (The same is true for the failure of unsuccessful revolutions.) History is only run once, so we will never know, but small or serendipitous factors might have initiated (or stopped) a revolutionary cascade. The #MeToo movement can be seen as such a cascade, marked by factors (1), (2), and (3). For that movement, as for successful revolutions, we might be able to point to some factors as necessary conditions, but hindsight is hazardous. It is also important to note that in revolutions, as in #MeToo, preferences and beliefs are not merely revealed; they are also transformed. Revolutionary activity, large or small, puts issues about preference falsification, experience falsification, and adaptive preferences in a new light.


I. Unpredictable Revolutions

Why do revolutions happen? Why are they so hard to anticipate? Why do they seem to come out of nowhere? My aim here is to cast some light on these questions and, in the process, to help explain #MeToo. I shall begin with some general remarks on revolutions and their genesis and then turn to #MeToo—which is not quite a revolution, of course, but which has something in common with one.

To vindicate the premise of my opening questions: Lenin was stunned by the success and speed of the Russian Revolution.1 Tocqueville reported that no one foresaw the French Revolution.2 The Iranian Revolution of 1789 was unanticipated.3 More recently, the Arab Spring was unanticipated by many of the best analysts in the United States, the United Kingdom, and elsewhere.4 Puzzlingly, revolutions seem to come in waves; they spread rapidly within countries and across countries, for reasons that remain unclear.5 It is tempting, and not unhelpful, to speak of demonstration and contagion effects. But what exactly do those terms mean?6 In what sense is revolution, or some kind of revolt, “contagious”?

A. Three Factors

Some of the most illuminating explanatory work on this subject points to three factors: (a) preference falsification, (b) diverse thresholds, and (c) interdependencies.7 When the three are taken together, the difficulty of anticipating such movements, or revolutions in particular, becomes less puzzling. I will introduce complications in due course, but these three factors tell us much that we need to know.

1. Preference falsification

Preference falsification exists when people conceal, or do not reveal, what they actually prefer.8 They might say they like the existing regime when they despise it. They might silence themselves. Their friends and neighbors might have no idea what they actually think. To that extent, people live in a world of pluralistic ignorance, in which they do not know about the preferences of others.9 Under regimes that are oppressive (in one or another respect), preference falsification is common. Because of oppression, it is difficult to learn what people actually think.10

For those who want to predict revolution or revolt, the problem is that the law, or social norms, can draw a wedge between private preferences and public preferences.11 The law matters if citizens lack freedom of speech and if dissent is punished. Social norms matter if people will be ostracized, in some sense, if they reveal their distress, anger, indignation, or dissatisfaction. Perhaps they will be shunned; perhaps powerful people will punish them in one or another way; perhaps their employment prospects will be compromised. In any of these cases, people might not merely silence themselves; they might say that they are happy with the status quo when they are not. Consider some chilling words from a computer programmer from Syria:

When you meet somebody coming out of Syria for the first time, you start to hear the same sentences. That everything is okay inside Syria, Syria is a great country, the economy is doing great. It’ll take him like six months, up to one year, to become a normal human being, to say what he thinks, what he feels. Then they might start whispering. They won’t speak loudly. That is too scary. After all that time, even outside Syria you feel that someone is listening, someone is recording.12

2. Diverse thresholds

Different people will require different levels of social support before they will rebel or say what they actually think.13 Some people might require no support at all; they are rebels by nature. They might be courageous, committed, or foolhardy. Call them the “zeroes.” They might well turn out to be isolated; no one may join them, in which case they might look radical, foolhardy, or even crazy. Other people might require a little support; they will not move unless someone else does, but if someone does, they are prepared to rebel as well. Call them the “ones.” Others might require more than a little; they are the “twos.” The twos will do nothing unless they see the zeroes and the ones, but if they do, they will rebel as well. The twos are followed by the threes, and the fours, and the tens, and the hundreds, and the thousands, all the way up to the infinites (defined as people who will not oppose the regime, no matter what).14

Outside of science fiction, it is not possible to see people’s thresholds. People may not quite know whether they themselves are threes, fours, or tens. They might turn out to be surprised. Consider the relevant words of John Adams, writing with evident amazement about the American Revolution: “Idolatry to Monarchs, and servility to Aristocratical Pride was never so totally eradicated from so many Minds in so short a Time.”15

3. Interdependencies

Interdependencies point to the fact that the behavior of the ones, the twos, the threes, and so forth will depend crucially on who, if anyone, is seen to have done what. Suppose that the various citizens are in a kind of temporal queue. The zeroes go first, then the ones, then the twos, then the threes, and so forth. (Or perhaps vice-versa. Or perhaps it is all random.) Under imaginable assumptions, a rebellion will occur, but only given the right distribution of thresholds and the right kind of visibility.16 If the ones see the zeroes, they will rebel, and if the twos see the ones, they, too, will rebel, and if the threes see the twos, they will join them.17 If the conditions are just right, almost everyone will rebel.18

But it is important to see that the conditions have to be just right. Suppose that there are no zeroes, or that no one sees any zeroes. If so, no rebellion will occur. If there are few ones, the regime is likely to be safe. If most people are tens or hundreds or thousands, the same is true, even if there are some ones, twos, three, fours, and so forth.

4. Unpredictability

We should now be able to see three reasons why revolutions may be impossible to predict. First, we do not know what people’s preferences are. By hypothesis, they cannot be observed. Second, we do not know what people’s thresholds are. They too are unobservable. Third, we cannot anticipate social interactions—who will say or do what and exactly when. It is important to emphasize the third point.19 Even if we could identify people’s preferences and specify their thresholds, we would not be able to know, in advance, the nature of social interactions. The point bears on revolutions in general and on #MeToo in particular. In the case of oppressive societies, it may be possible to know that people are widely miserable or dissatisfied. In the context of sexual assault and sexual harassment, it is reasonable to assume that dissatisfaction is widespread. But that is not enough.

These points suggest that even if new technologies make it increasingly possible to identify private preferences—for example, by exploring people’s online behavior—we will still not be able to predict revolutions.20 To be sure, we would know something important: a revolution is more likely if people secretly hate the regime. We could certainly learn from that fact. Secret opposition may be necessary for revolution, but it is not sufficient. To know what will happen, we would need to know about people’s thresholds as well. As I have noted, obtaining that knowledge will inevitably be difficult; it might be impossible. And even if we overcome that challenge, we would need to know who interacts with whom, and who sees whom, and when. No one has that kind of prescience. But the answers to those questions may well determine outcomes.21

These points help explain not only why revolutions are unpredictable but also why they are often a product of seemingly small, random, or serendipitous factors—of who did what when, or who heard what when, or whether some kind of butterfly flapped its wings at the right moment.22 We might think that Regime “A” was bound to fall, but it really was not. It happened to fall. The same is true if it does not fall. It happened not to fall.23 Counterfactual histories can be illuminating insofar as they illustrate this point.24

B. Complications

This is a very simple account, of course, and it needs to be complicated in multiple ways. For present purposes, consider these points.

First, people’s preferences may be adaptive to the status quo.25 People might not have to work hard to shut themselves up. They might not even think that the status quo is bad. Consider these words from a woman in North Korea: “It never occurred to me that I could or would want to do anything about it. It was just how things are.”26 The most important word here is “want.” To be sure, fully adaptive preferences are an extreme case, even under conditions of real fear.27 It might be better to speak of partially adaptive preferences, in which people are aware that something is wrong, or bad, or horrific, but the awareness takes the form of a small voice in the head, to which people do not pay a great deal of attention. But the idea of preference falsification is too simple when people’s preferences are an artifact of the status quo. Whether we are dealing with preference falsification, adaptive preferences, or partially adaptive preferences cannot be answered in the abstract.

Second, the very word “preferences” is under-descriptive or perhaps misleading. It might be better to speak of people’s beliefs, experiences, or values. Under an oppressive regime, people might believe that terrible injustices are committed or that their values are being violated. To be sure, they are also concealing or falsifying what they prefer, but that is hardly an adequate account of what is happening. They are concealing or falsifying their deepest convictions. They are concealing or falsifying what actually happened to them. (Talk about fake news).

Third, and crucially, rebels are not doing a full analysis of the costs and benefits of rebellion. They rely on mental shortcuts, or heuristics, in deciding what to do and when.28 For that reason, available incidents or outcomes might affect probability judgments.29 If a town suddenly falls to rebels, or if a government collapses, other rebels might believe that the probability of success is high.30 The availability heuristic, as it is called, works with emphatically social forces, producing availability cascades, as specific incidents or results move rapidly from one person to another, altering judgments about what is likely to happen.31 A revolutionary movement might be fueled or halted by an availability cascade.

Fourth, fate is not only in the hands of revolutionaries. There is also the regime, and there are also counterrevolutionaries, and there may well be counterrevolution. As a revolutionary cascade starts to develop, the regime is likely to do something. For example, it might try to entrench pluralistic ignorance by hiding or preventing visible rebellion or mass demonstrations.32 It might allow dissent and disagreement—until they become too visible.33 It might make concessions, hoping to retain power. It might try to dissuade the hundreds and the thousands. It might bring out its guns. It might kill people.34 If the goal of the regime is to maintain power, the choice among these options can be very difficult. For example, violence might be effective in quelling revolution, but it might also foment more of it.35

II. #MeToo

Turn to #MeToo in this light. All three conditions are met. The qualifications are relevant as well.

First, with respect to sexual assault and sexual harassment, preference falsification has run rampant.36 Victims have silenced themselves.37 In some cases, they have said that all is or was well, when it is or was anything but that.38 These points are true and important, but they are inadequate and under-descriptive. What many women (and many, but fewer, men) did not reveal—what they kept private—was a set of experiences, alongside evaluative judgments about those experiences. We might want to speak, in the case of #MeToo, of experience falsification. Self-silencing has been important, of course, but actual falsification of experience—with an employer, for example—might be more searing.39

Experience falsification or self-silencing can be a product of many different factors. With respect to sexual violence or sexual harassment, it may be a product of a rational calculation of likely costs and benefits, given the risks of disclosure. Some women who did come forward with accusations of assault and harassment pre-#MeToo have been ridiculed or disparaged, or worse, providing a signal to other victims about what might happen if they spoke out and thus tilting the cost-benefit analysis in favor of staying silent.40 If cases of this kind were highly visible and thus cognitively “available,” the availability heuristic would lead people to think that probability of damage or harm from disclosure could be quite high. But we need not invoke the availability heuristic. A 2003 study, cited by the EEOC in 2016, indicated that 75% of employees who spoke out against workplace mistreatment faced some form of retaliation.41 To the extent that victims of sexual harassment were aware of the risk of retaliation, that awareness provided a reason to falsify their experiences or at least not to speak about them.

Second, different women had and have different thresholds for disclosing their experiences and their judgments. Some women are ones, others are twos, others are tens, and others are hundreds or thousands. For one reason or another, some may be infinites.42 (They might be frightened; they might have some kind of loyalty to the perpetrator; they might not want their lives to be disrupted; they might cherish their privacy.) Some might not have clarity on what their thresholds are. They, and we, learn about that only ex post. Consider the following words from Beverly Young Nelson, who accused Republican Senate candidate Roy Moore of having sexually assaulted her in 1977:

I thought that I was Mr. Moore’s only victim. I would probably have taken what Mr. Moore did to me to my grave, had it not been for the courage of four other women that were willing to speak out about their experiences with Mr. Moore. Their courage has inspired me to overcome my fear.43

Third, social interactions are, and continue to be, crucial to #MeToo. Under certain conditions, the threes and the fours would silence themselves, because the ones and the twos were silent too. But #MeToo has benefited from the visibility of those who spoke out and the multiple interactions made possible by social media. Within 24 hours of Alyssa Milano’s initial tweet, 45% of all U.S. Facebook users had friends in their networks who had posted with #MeToo.44 Once the ones and the twos spoke out, the threes and the fours felt safer or emboldened.

It is important to say that this account is barebones and highly stylized, and that it misses a great deal. I emphasize five points here. First, the #MeToo movement is not opposing a regime, at least not in the usual sense.45 Rather than rebelling against a government, the women (and men) of #MeToo are uniting around a similar or common experience and rebelling against a practice and also against institutions (some of which may be in government). While cascading accusations against individual perpetrators have been crucial—for example, more than two dozen women spoke out against Roger Ailes before he was ousted at Fox News—the larger movement might be understood as a challenge to a system of sex discrimination and to institutions that engage in or perpetuate it.46 To the extent that we are speaking of institutions, it is not so much of a stretch to say that regime change, at least of a sort, is involved.

Second, there is the question of granularity—of exactly what happened, and when, and why. Answering that question would reveal not only informative detail but also conceptual surprises.

Third, there is the crucial role of salience in the #MeToo movement.47 Some twos are different from other twos, and the same is true for threes and fours, for one reason: their own statements and actions are especially salient. In the context of #MeToo, Ashley Judd might have made all the difference.48 Catharine MacKinnon has suggested that Judd’s celebrity and salience were not the only thing that made her an ideal first-mover; she was also, importantly, “somebody whose credibility is not readily attackable and who wasn’t suing at the time.”49 In revolutionary movements in general, what is salient, and what is cognitively available, greatly matters. As I have suggested, rebels do not make elaborate cost-benefit analyses. They use mental short-cuts, and availability is especially important.50

Fourth, descriptive social norms, which capture what people actually do, greatly matter. Other things being equal, people are more likely to change their behavior to comply with a norm if they believe that most other people are compliant, and less likely to do so if they believe that most other people are noncompliant. A prominent study found that visitors to a national park who saw signs informing them that many past visitors had stolen petrified wood from the park became more likely to steal petrified wood—and that visitors who saw signs informing them that the vast majority of visitors had left the wood in the park became less likely to steal petrified wood.51 The #MeToo movement appears to have benefited from a shift in descriptive norms, suggesting that speaking out or objecting is not inconsistent with usual behavior.

While it seems highly unlikely, we cannot rule out the possibility that attention to widespread harassment will serve to inform (some) male perpetrators that they are simply behaving like many other men, thus reducing their incentive to behave differently.52 But it is also true that, to the extent #MeToo succeeds in changing norms, a new (and beneficial) wave of preference falsification may lead potential harassers to condemn the behavior rather than to support it.53 There is much to be learned about this topic.54

Finally, #MeToo is not simply about the revelation of preferences, experiences, beliefs, and values. It is also about the transformation of preferences, beliefs, and values—most obviously on the part of perpetrators, but equally relevantly on the part of victims. Any social movement helps to alter preferences, beliefs, and values. It casts a new light on past experiences. It does not merely elicit preexisting judgments. It produces fresh ones.55 Part of the point of #MeToo, and one of its achievements, is to turn a sense of embarrassment and shame into a sense of dignity.

Recall the statement from a computer programmer from Syria:

When you meet somebody coming out of Syria for the first time, you start to hear the same sentences. That everything is okay. It’ll take like six months, up to one year, to become a normal human being, to say what he thinks, what he feels. Then they might start whispering. They won’t speak loudly.56

But eventually they might.

  • 1Asef Bayat, The Arab Spring and its Surprises, 44 Dev. & Change 587, 587–88 (2013).
  • 2Id. at 587.
  • 3Id. at 588.
  • 4Id. at 587.
  • 5Kurt Weyland, Making Waves: Democratic Contention in Europe and Latin America Since the Revolutions of 1848, 1–7 (2014).
  • 6See id. at 7–11, for valuable discussion, emphasizing the availability and representativeness heuristics. Weyland’s exploration of availability and representativeness has implications for rebellions of many kinds, and not merely revolutions; #MeToo could easily be studied with reference to those heuristics. I offer some brief remarks at various points below.
  • 7See generally Timur Kuran, The Inevitability of Future Revolutionary Surprises, 100 Am. J. Soc. 1528 (1995); Timur Kuran & Diego Romero, The Logic of Revolutions, in 2 The Oxford Handbook of Public Choice (Roger D. Congleton et al. eds., 2018); Merouan Mekouar, Protect and Mass Mobilization: Authoritarian Collapse and Political Change in North Africa (1st ed. 2016); Muzammil Hussain & Philip Howard, What Best Explains Successful Protest Cascades? ICTs and the Fuzzy Causes of the Arab Spring, 15 Int’l Stud. Rev. 48 (2013) (emphasizing the importance of communications technologies).
  • 8Timur Kuran, Private Truths, Public Lies: the Social Consequences of Preference Falsification 4–5 (1995). The literature on “informational cascades” is also relevant, but revolutions go well beyond those. See generally Sushil Bikhchandani et al., A Theory of Fads, Fashion, Custom, and Cultural Change in Informational Cascades, 100 J. Pol. Econ. 992 (1992); Susanne Lohmann, I Know You Know He or She Knows We Know You Know They Know: Common Knowledge and the Unpredictability of Informational Cascades, in Political Complexity: Nonlinear Models of Politics (Diana Richards ed., 2000).
  • 9See, e.g., Kuran, supra note 8, for an especially good account see also D. Garth Taylor, Pluralistic Ignorance and the Spiral of Silence: A Formal Analysis, 46 Pub. Opinion Q. 311 (1982); For a valuable account with special reference to law, see Richard McAdams, The Expressive Powers of Law: Theories and Limits, 136–62 (2015).
  • 10See generally Wendy Pearlman, We Crossed a Bridge and It Trembled: Voices from Syria (2017), for first-hand reports.
  • 11See Kuran, supra note 8, at 84–102.
  • 12Pearlman, supra note 10, at 4.
  • 13See, e.g., Mark Granovetter, Threshold Models of Collective Behavior, 83 Am. J. Soc. 1420 (1978), for the classic account; Kuran, supra note 8, at 60–83.
  • 14The infinites deserve some attention. Their motivations are undoubtedly varied; they may involve identity, habit, fear, loyalty, or something else.
  • 15Gordon Wood, The Radicalism of the American Revolution 169 (1993).
  • 16See generally Granovetter, supra note 14, for the clearest explanation.
  • 17Id. at 1424–25.
  • 18See id. at 1431. See also Heng Chen & Wing Suen, Falling Dominoes: A Theory of Rare Events and Crisis Contagion, 8 Am. Econ. J.: Microecon. 228, 239 (2016), for an emphasis on the importance of beliefs and on their fragility.
  • 19See generally Matthew Salganik, Peter Sheridan Dodds & Duncan J. Watts, Experimental Study of Inequality and Unpredictability in an Artificial Cultural Market, 311 Sci. 854, 854–56 (2006), emphasizing the unpredictability of social exchange.
  • 20See generally Seth Stephens-Davidowitz, Everybody Lies: Big Data, New Data, and What the Internet Can Tell Us About Who We Really Are (2017).
  • 21See Salganik, supra note 20, at 854–56, for relevant findings. See also Matthew Salganik & Duncan Watts, Leading the Herd Astray: An Experimental Study of Self-fulfilling Prophecies in an Artificial Cultural Market, 71 Soc. Psychol. Q. 338, 338–55 (2008); Matthew Salganik & Duncan Watts, Web-Based Experiments for the Study of Collective Social Dynamics in Cultural Markets, 1 Topics in Cognitive Sci. 439, 439–68 (2009).
  • 22See generally Catharine MacKinnon, Butterfly Politics (2017); Paul Ormerod, Butterfly Economics: A New General Theory of Social and Economic Behavior (1998).
  • 23To be sure, the factors that underlie any fall, or failure to fall, deserve close attention.
  • 24See generally Cass R. Sunstein, Historical Explanations Always Involve Counterfactual History, 10 J. Phil. Hist. 433 (Issue 3, 2016).
  • 25See Jon Elster, Sour Grapes: Studies in Subversion of Rationality, 25 (1983).
  • 26Choe Sang-Hun, North Korea #MeToo Voices: ‘They Consider Us Toys,’ N.Y. Times (Oct. 31, 2018), https://www.nytimes.com/2018/10/31/world/asia/north-korea-women-metoo.html [https://perma.cc/VLP5-CES3].
  • 27Cf. Serene Khader, Adaptive Preferences and Women’s Empowerment (Studies in Feminist Philosophy) (2011) (discussing the complexity of the idea of adaptive preferences in the face of personal agency).
  • 28See Weyland, supra note 5, at 35–38.
  • 29Kurt Weyland, The Arab Spring: Why the Surprising Similarities with the Revolutionary Wave of 1848?, 10 Persp. on Pol. 917, 921 (2012).
  • 30See id.
  • 31See Timur Kuran & Cass R. Sunstein, Availability Cascades and Risk Regulation, 51 Stan. L. Rev. 683, 685 (1999).
  • 32Gary King, Jennifer Pan & Margaret E. Roberts, How the Chinese Government Fabricates Social Media Posts for Strategic Distraction, Not Engaged Argument, 11 Am. Pol. Sci. Rev. 484, 496 (2017).
  • 33Id.
  • 34See generally Pearlman, supra note 10.
  • 35See Edward L. Glaeser & Cass R. Sunstein, A Theory of Civil Disobedience, Nat’l Bureau of Econ., 2–3, 13–17 (2015), https://www.nber.org/papers/w21338 [https://perma.cc/8N4R-339M].
  • 36But see Catharine MacKinnon, #MeToo Has Done What the Law Could Not, N.Y. Times (Feb. 4, 2018), https://www.nytimes.com/2018/02/04/opinion/metoo-law-legal-system.html [https://perma.cc/NQ78-LWD7] (“Women have been saying these things forever. It is the response to them that has changed.”). MacKinnon is surely right on this point. It is also true that some women said these things privately rather than publicly—and some spoke to no one at all.
  • 37Timur Kuran, who introduced the concept of preference falsification, has used the concept to explain the pre-#MeToo silence around sexual harassment and assault, drawing comparisons to the fear-induced preference falsification that preserved communist rule for decades. Shankar Vedantam, The Psychological Forces Behind a Cultural Reckoning: Understanding #MeToo, (NPR: Hidden Brain Radio Broadcast Feb. 5, 2018) (in conversation with Princeton psychologist Betsy Paluck), https://www.npr.org/templates/transcript/transcript.php?storyId=582698111 [https://perma.cc/933R-WW23] (“KURAN: For decades, communism survived by making the populations it ruled afraid to express opposition to the principles of communism and express opposition to the dictatorships that were running the Soviet bloc countries. VEDANTAM: The silence around Harvey Weinstein’s sexual misconduct operated in the same way. KURAN: It was an open secret for decades, we’ve learned, in Hollywood and in circles that Harvey Weinstein traveled in, that he was a predator of young women, but also, that if anybody called him out on this, he would ruin their careers.”)
  • 38See, e.g., Cass R. Sunstein, Unleashed, 85 Soc. Res.: An Int’l Q. 73, 73–92 (2018).
  • 39A vivid treatment of experience falsification is The Tale (Gamechanger Films et al. 2018), https://www.amazon.com/gp/product/B07D8MCN2B/ref=atv_feed_catalog [https://perma.cc/MPR5-C8DZ].
  • 40Vedantam, supra note 39.
  • 41U.S. Equal Emp. Opportunity Comm’n, Select Task Force on the Study of Harassment in the Workplace: Report of Co-chairs Chai R. Feldblum & Victoria A. Lipnic, 16 (2016), https://www.eeoc.gov/eeoc/task_force/harassment/upload/report.pdf [https://perma.cc/CJ8B-TE3G].
  • 42See Lisa Bonos, Not Everyone with a #MeToo is Posting Their Story. Here’s Why Some Are Refraining, Wash. Post (Oct. 19, 2017), https://www.washingtonpost.com/news/soloish/wp/2017/10/19/not-everyone-with-a-metoo-is-posting-their-story-heres-why-some-are-refraining/?noredirect=on&utm_term=.502bb45e04b7 [https://perma.cc/23KN-RBSA], for insights from the “infinites.”
  • 43See Text of Beverly Young Nelson’s Accusation against Roy Moore, N.Y Times (Nov. 13, 20-17), https://www.nytimes.com/2017/11/13/us/politics/text-beverly-young-nelson-statement.html [https://perma.cc/6X4F-FZAD] (last visited Jan. 25, 2019).
  • 44See More than 12M “Me Too” Facebook Posts, Comments, Reactions in 24 Hours, CBS News (Oct. 17, 2017), https://www.cbsnews.com/news/metoo-more-than-12-million-facebook-posts-comments-reactions-24-hours [https://perma.cc/P33G-HKUZ]. Note also that it might be easier to use Twitter, to reveal an experience or to state agreement, than to speak offline, or to attempt to show support or to attract attention that way.
  • 45But see Ella Nilsen, More than 100 Members of Congress Want the Oversight Committee to Investigate Trump, Vox (Dec. 12, 2017), https://www.vox.com/policy-and-politics/2017/12/12/16766800/democratic-congress-members-trump-investigation-women [https://perma.cc/3VYD-GCSR] (suggesting that, for some, #MeToo may be viewed as a tool to oust President Trump, who has been accused of sexual assault by multiple women).
  • 46See Gabriel Sherman, The Revenge of Roger’s Angels, N.Y. Mag. (Sept. 5, 2016), http://nymag.com/intelligencer/2016/09/how-fox-news-women-took-down-roger-ailes.html [https://perma.cc/4FP4-RLLS].
  • 47See generally Weyland, supra note 29 (emphasizing availability); see also Margaret E. Tankard & Elizabeth Levy Paluck, Norm Perception as a Vehicle for Social Change, 10 Soc. Issues & Pol.’y Rev. 181, 185 (2016).
  • 48See, e.g., Stephanie Zacharek, Eliana Dockterman & Haley Sweetland Edwards, Time Person of the Year 2017: The Silence Breakers, Time (Dec. 6, 2017), http://time.com/time-person-of-the-year-2017-silence-breakers/ [https://perma.cc/7PWN-6AY4] (noting Judd’s importance).
  • 49Brock Colyar, The Ms. Q&A: Catharine MacKinnon Weighs in on the #MeToo Movement, Ms. Mag. Blog (July 30, 2018), http://msmagazine.com/blog/2018/07/30/ms-qa-catharine-mackinnon-weighs-metoo-movement/ [https://perma.cc/53QS-PSKN].
  • 50See Weyland, supra note 29, at 35–38.
  • 51See Robert B. Cialdini et al., Managing Social Norms for Persuasive Impact, 1 Soc. Influence 3, 7–8 (2006).
  • 52See Siri Uotila, Could the #MeToo Movement Inadvertently Lead to More Sexual Harassment?, Behav. Scientist (Apr. 16, 2018), http://behavioralscientist.org/ask-a-behavioral-scientist-could-the-metoo-movement-inadvertently-lead-to-more-sexual-harassment/ [https://perma.cc/GW6A-C59N].
  • 53See id.
  • 54It is important to note that populations are heterogeneous. Some people will be very glad to deviate from a new or emerging norm; they are defiant. Robert Kagan & Jerome Skolnick, Banning Smoking: Compliance Without Enforcement, in Smoking Pol’y: Law, Pol., & Culture 78, 78–81 (Robert Rabin & Stephen Sugarman eds., 1993), for relevant discussion.
  • 55On the difference between the two, see Cass R. Sunstein, How Change Happens (2019).
  • 56Pearlman, supra note 10, at 4.