ILLUSTRATION BY GORDON STUDER

By Frances Milliken, Joe C. Magee, Nancy Lam, and Daniel Menezes

An analysis of this real-world situation shows how the response was shaped by the different levels of power of those involved.

ne of the most troubling aspects of Hurricane Katrina and its aftermath was the seemingly confused and slow response of government officials and of large nonprofit relief organizations to the crisis. It often seemed as if there was a “disconnect” between what was being conveyed by some of the most powerful people in the country and what we were seeing on television. Those who watched the coverage may remember the desperate calls to a television reporter from doctors and nurses stranded in a hospital in New Orleans with no electricity. Millions of people were aware of the worsening plight of these healthcare workers and their patients, yet government aid did not arrive for days after the August 29, 2005, disaster.

Why was the government’s response to Hurricane Katrina so slow and seemingly uncoordinated? One possible explanation is that the officials in charge thought about the situation in a systematically different way than the first responders and victims did. These differences in the interpretations of the situation may have stemmed from actual differences in the amount of knowledge individuals had or from differences in the perspectives they had about the situation. In particular, the level of power they possessed may have affected how they thought about the situation.

Previous laboratory research has suggested that a person’s level of power can affect how he or she interprets events. Research suggests, for example, that individuals with power may tend to perceive problems or crises as less threatening than those with less power. If this were found to be true in the context of a real situation, this could cause high-level officials to respond to an emerging problem in a way that people with less power perceive to be inadequate.

In our analysis, we sought to build a bridge between research on organizational decision making and research on the psychology of power by showing empirically how power-holders may have differed from those with less power in their interpretations of a high-consequence, real-world event. The aftermath of Hurricane Katrina, during which multiple interdependent individuals and organizations attempted to work together to respond to the storm, was an ideal setting for us to explore.

In the case of Hurricane Katrina, the loosely coordinated organization that was responsible for preserving lives and infrastructure shared many of the features of the traditional, pyramid-like structure of hierarchical organizations: a small number of federal government officials with a great deal of power at the top, state and local officials and military personnel in the middle, followed by an array of volunteers with relatively little power, and, finally, thousands of powerless victims.

Talk to Me

Research on communication in hierarchies suggests that the transfer of information up and down the line is rarely smooth. Those at the top not only have the power to affect the outcomes of those at the bottom – thus creating incentives for those lower down to communicate selectively based on what they think power-holders want to hear – but they also may have a different perspective on the nature of the issues being confronted. Interviews and testimony suggest that communication up and down the hierarchy in the networked response to Hurricane Katrina was deficient, and it is widely believed that the suffering was significantly worse as a result. Pinning down the exact causes of miscommunication, however, has been a topic of intense debate.

We present a unique reason why the response to Hurricane Katrina was deficient. We argue that differences in power among the people responsible for coordinating and delivering relief, and between those people and the hurricane’s victims contributed to differences in their interpretations of the event as it was unfolding, which, in turn, impeded effective communications and a coordinated response.

In a crisis, those who anticipate, notice, or recognize problems must communicate with the people who have the power to make decisions about how to solve them, so how those problems are interpreted by the individuals who are in charge is critical. Further, those with the resources to solve the problems need to develop solutions that are then implemented by those in the field. Problem solving in organizations thus almost always involves information to be relayed up and down a hierarchy. If people at the top of a hierarchy tend to think in reliably different ways about issues, they may generate different solutions than what those lower down anticipated. Sometimes problem solving is further complicated by involving multiple organizations that are themselves nested in a hierarchy of power, as was the case after Hurricane Katrina. To develop our hypotheses about how individuals at different levels of this disaster relief hierarchy interpreted the storm and its aftermath in systematically different ways, we built on psychological perspectives of the effects of power on cognition and behavior.

Transformative Power

Power is an intriguing lens through which to examine how people view the world. Our hypotheses were based on two well-researched theoretical frameworks of how power transforms individuals psychologically. The first proposes that individuals with greater power tend to focus more on the positive, rewarding aspects of their environment rather than on negative, threatening aspects. Studies have found, for example, that power is positively associated with the anticipation of positive outcomes in risky situations and with the experience of positive emotion. We hypothesized, therefore, that there would be a positive relationship between individuals’ level of power and the extent to which their communication about Hurricane Katrina tended to be marked by the expression of positive thoughts. Similarly, high-power individuals faced with interpreting environmental threats or crises may tend to be more focused on the future in an effort to “find the positive” in the situation. Thus, we predicted that there would be a significant relationship between individuals’ level of power and the likelihood of being future-oriented in their communication.

A second framework that we used to guide our research posits that any form of psychological distance – such as differences in power – affects how individuals interpret their environment. Hurricane Katrina could be described abstractly as “a vicious storm wreaking havoc on anything in its path” or more concretely as “a level 4 hurricane that is forcing people to climb up to their roofs and wave white towels for help.” People who are very powerful may tend to construe their world in more abstract terms than others. Thus, we hypothesized that there was a positive relationship between individuals’ level of power and how abstractly they communicated about Katrina.

Power, also is associated with a greater sense of perceived control and certainty about the future. Not surprisingly, high-power individuals perceive themselves as having more control than people with lower levels of power, which translates into their having greater confidence in predicting the outcomes of events. In addition, psychological distance tends to increase this confidence, because specific details that might present roadblocks are not a focus. This suggests power-holders had increased confidence during the Katrina crisis, and we investigated this hypothesis as well.

The Method

To test our hypotheses, we analyzed the content of individuals’ descriptions of the events during the 10 days following the hurricane, examining direct quotations of people involved in relief and recovery. These included victims as well as people who were organizing and informing the response.

To capture a range of media (e.g., local versus national; print versus radio and television), we chose The New York Times, The Times-Picayune of New Orleans, CNN, and National Public Radio. Our final sample consisted of 888 unique quotations by 272 unique speakers. Federal officials and governors had the most power, followed by lesser state and local officials. Police, fire, and military personnel had more power than disaster responders and experts. Victims had the least power.

Our Findings

We found that the language that our speakers used to communicate about the hurricane varied as a function of the amount of power the speakers held. As expected, people with more power tended to use more abstract language. The data also suggested that people who had power were more likely to focus on the future and to use positive language when they spoke about the situation. Finally, those with the most power in this situation (the federal officials) showed a tendency to use language that reflected a higher degree of certainty and confidence than that used by individuals with less power. Our findings, thus, provided strong support for our hypotheses.

ur study contributes to the literature on power and on sense-making and communication in organizations by showing that power appears to play a central role in the sense-making process, potentially producing a “sense-making gap” between high- and low-power individuals’ interpretations of events. Given that most, if not all, organizations are characterized by a hierarchy of power, understanding this gap is important to constructing a more accurate model of organizations as information-processing and sense-making systems.

One could argue that high-level government officials, like other top managers, are supposed to “see the forest and not the trees.” They should be thinking somewhat abstractly and searching for common features across situations so as to apply standardized solutions. One could also argue that it is useful for high-level managers to be positive and future-oriented in their thinking. But what if these ways of perceiving and interpreting events are systematically different from those of their subordinates? If leaders respond to a situation from a viewpoint that is misaligned with others’ viewpoints, they could respond in ways that subordinates find surprising. For example, because high-power individuals are more likely to see the positive in a negative situation, they may be slower to respond to a crisis than their subordinates think they should be. They may also respond in ways that others lower down in the hierarchy perceive as insufficient given the magnitude of the problem. Leaders could also lose credibility with subordinates if they are perceived to not have a good understanding of events on the ground, which could damage communications within the hierarchy. Lower-level employees could then elect to withhold critical information about problems from their bosses, believing that their managers only want to hear positive news and, perhaps, are not really interested in solving problems. Over time, this could cause leaders to become out of touch with “the reality on the ground.” If leaders are perceived to be out of touch, their effectiveness would eventually erode because they would not, in fact, have a good understanding of the complex issues and problems facing the organization. Elected officials could be voted out of office, and managers who have lost support – or subordinates who do not feel supported by management – could quit.

Conclusion

The finding that powerful individuals interpret the world in very different terms than those with little power might not surprise anyone familiar with organizations. But that psychological theories of power can explain and predict this phenomenon in the context of a large-scale disaster is remarkable. Power colors how people perceive, interpret, and communicate about the situations they encounter in ways that are often not consciously recognized or understood. Understanding more about these effects will be critical to helping ensure effective communication and coordination up and down organizational hierarchies.

FRANCES J. MILLIKEN is professor of management and organizations and Peter Drucker Faculty Fellow at NYU Stern. JOE C. MAGEE is an affiliated assistant professor of management and organizations at NYU Stern and assistant professor of management at the NYU Robert F. Wagner Graduate School of Public Service. NANCY LAM is a doctoral student in the management department at NYU Stern. DANIEL MENEZES graduated in May 2008 with a BA in Metropolitan Studies from the NYU Department of Social and Cultural Analysis.

The authors thank the NYU Center for Catastrophe Preparedness and Response for the grant that supported this research.