Skip to content
The OCR Glossary

Reputation Cascades

Jean-Philippe Bonardi

The theory of reputation cascades explains why rational individuals who care about their own reputation might adopt the opinion of others, sometimes contrary to their own existing belief, and why this process—the cascade—leads to the formation of general opinions that are sometimes very stable over time. Reputation cascades apply to what individuals say about firms and can therefore be a strong driver of corporate reputation.

The theory rests on three fundamental assumptions. The first is that most people are rationally ignorant when it comes to firms: Because the costs of getting informed about these firms are high and the benefits relatively low, people often decide to remain ignorant and to rely on the opinion of others about these firms. It is thus the opinions of a few that will be at the origin of companies’ reputation, rather than the individual opinions of most people. Following this, the second assumption is that these few individuals driving corporate reputation are those who can be seen as experts, for example, researchers, reporters covering firms, and financial analysts, whose role is to investigate and give their opinions about firms. The third assumption, however, is that these experts also care about their own individual reputation and face a cost if their opinion is against the trend set by others. As a result, some of these experts might decide to falsify their opinion, not by ignorance but to enhance their own reputation, leading to the formation of a common public opinion about a firm even though their true opinion is in fact a lot less homogeneous. This last assumption is the one that creates the possibility of individual experts’ reputations leading to choices cascading into a homogeneous public opinion about a firm—that is, a corporate reputation. This entry covers the drivers of experts’ opinions about firms, the dynamics of reputation cascades among firms, and the implications for corporate reputation.

The Drivers of Experts’ Opinions About a Firm

Analytically, the starting point is the choice faced by an expert who must convey a preference regarding a firm or what this firm does. If experts do not feel any social pressure related to their evaluation of a firm, they will give their private opinion when asked about this firm. In some cases, however, experts are concerned about the social pressure created by other experts and may falsify their private opinion—for instance, knowingly falsifying their analysis of a firm, concealing disturbing information, looking at certain research areas without considering others, or self-silencing. Experts, however, vary in their responses to prevailing social pressures, and one individual may resist the pressure that another chooses to accommodate through preference falsification.

When an expert is asked to publicly express an opinion, he receives benefits or incurs costs as a result. Three factors enter experts’ calculations: (1) the satisfaction that one is likely to obtain from the opinion about a firm that emerges, (2) the reward or punishment associated with a chosen preference, and (3) the benefits one derives from truthful self-expression. Of these three considerations, the one that is the least subjective and the most interesting from a managerial point of view is the second: rewards or punishments for a certain opinion about a firm. Sanctions imposed on experts include being ostracized in conferences or public debates, being hindered in the development of their career, or finding it difficult to acquire research support. Rewards include attracting a wide audience to escalate their career and pushing an attractive story in the media.

The Dynamics of Reputation Cascades Among Experts

When an expert expresses his opinion about a firm, his choice can be modeled as being between two alternatives: (1) the position 0, or a low opinion about the firm, and (2) the position 100, indicating a high opinion. Each expert has a threshold, between 0 and 100 percent of the mean estimated collective opinion, at which he decides to falsify his private preference. The mean collective opinion characterizes the firm’s reputation. For instance, if an expert has a private preference x = 30, then given a choice of 0 or 100, he would rather support the former. But there is a threshold at which social pressure will push him to express 100. Assume that this expert’s threshold is 60. If the collective opinion is below 60, he will express 0, whereas if the collective opinion among experts is above 60, he will pick 100.

Figure 1 shows a distribution of thresholds among experts who have to make a statement about a company. On the graph, the vertical axis represents the cumulative distribution of thresholds, found by plotting, for each value of the expected mean collective opinion—the percentage of the expert community with a threshold at or below that level. Here, 17 percent of the experts have a threshold equal to 0: They already believe that the company is a great one (that 100 is the right opinion) and do not have to falsify their private opinion. On the other hand, 80 percent of experts have a threshold below 100: Their preference can be falsified if they expect the collective opinion to be high enough. Similarly, at 30 percent of the estimated mean collective opinion among those experts, 58 percent of them will have reached their threshold and will then falsify their private opinion to follow the trend.

Figure 1 Experts’ Thresholds as Drivers of a Firm’s Reputation

Source: Adapted from Bonardi and Keim (2005, p. 563).

Looking now at the dynamics, if the expected public opinion starts at 10, it appears that 32 percent of the population has a threshold at or below 10. So this share of the experts will give its support to 100, and the remaining 68 percent will support 0, suggesting that the firm has a relatively poor reputation. However, beyond going from 10 to 32 percent, further reestimation will take place, driving experts’ public opinion higher until no more upward estimation occurs as all experts have reached their thresholds. Here, it happens at 78 percent of the collective opinion, which is a stable equilibrium. A large majority of experts now support publicly the position 100 (indicating a great corporate reputation), even though only 10 percent of them were really convinced when the process started. The reputation cascade has worked to transform the collective opinion of the experts, who will then search for information to support their viewpoint and provide this information to rationally ignorant voters.


Several implications for corporate reputation can be derived from this analysis. The first is that a good reputation often generates an even better one by making it costly for many experts to express their true opinion. When a stable equilibrium is reached, a firm’s reputation is likely to remain strong over time as many experts will hesitate to make contrary statements. This mechanism, however, also works the other way round: The public opinion of experts might cascade down—which happens when the curve characterizing the distribution of thresholds goes under the median—and some firms might find it difficult to improve their reputation as they cannot push many experts to publicly express their positive private impression.

In many cases, in fact, several stable equilibria exist, and a small change in a few experts’ opinions might make a firm’s reputation cascade down from a strong reputation to a bad one. Managers’ task is then to find a way to influence experts’ thresholds—that is, through the sanctions and rewards related to the public opinion these experts take.

Bonardi, J.-P., & Keim, G. D. (2005). Corporate political strategies for widely salient issues. Academy of Management Review, 30(3), 555–576. doi:

Kuran, T. (1995). Private truths, public lies. Cambridge, MA: Harvard University Press.

Kuran, T. (1998). Ethnic norms and their transformation through reputational cascades. Journal of Legal Studies, 27, 623–659.

See Also

Information Intermediaries; Media Reputation; News Media; Public Opinion; Stakeholders

See Also

Please select listing to show.