Heckuva Job: Computer Model To Predict Organizational Response To Disasters

In the wake of Katrina, a category 5 hurricane that devastated much of the nation’s Gulf Coast region in August 2005, three Rensselaer students traveled to New Orleans to collect paperwork, e-mails, reviews, assessments, and other documents that could provide them with information about how each organization reacted to any given task during the disaster. The recovered paper trail also provided the researchers with insight into a variety of cultural and organizational characteristics that impacted both agencies’ ability to act during the disaster.

Throughout the group’s research, a vast dichotomy between the cultures of FEMA and the Coast Guard became increasingly evident, according to William “Al” Wallace, professor of decision sciences and engineering systems (DSES) at Rensselaer, and principal investigator on the project. The researchers believe these cultural factors ultimately dictated how well each organization was able to carry out its function and responsibilities. 

“FEMA’s fatalist culture, coupled with the loss of its cabinet-level position and budget and rulemaking authority, crippled the agency’s ability to fulfill its normal repertoire of emergency coordination and response during Katrina,” says Wallace. “On the other hand, the Coast Guard had undergone minimal organizational changes and had its pre-existing routines supported, thus it was better equipped to fulfill its duties during the disaster. Additionally, because of the Coast Guard’s hierarchical culture, action orders continually disseminated through the organization’s chain of command to the response team.” 

Today Wallace is leading a team of researchers to construct a computer simulation that models an extreme disaster situation – similar to that of Hurricane Katrina – where decision-makers are forced to shift their attention from one dimension to another, responses often play out over long durations of time, and information demands vary between interacting response organizations.

They’ll then input a series of “what if” scenarios related to organizational structure and culture into the disaster model. Algorithms, or automated reasoning, will predict how each organization’s constraints would affect its ability to effectively react to an emergency. The organizational factors observed by the researchers while studying FEMA and the Coast Guard will be used to test the model and to set the parameters.

“Essentially, the model will be able to determine how well an organization will respond to a disaster based on the rules it is following and its organizational structure,” says Wallace, who warns that the device is not a scenario generator. “It won’t tell you ‘if you have a disaster and you don’t get enough ice to the victims in time, this will happen.’ Instead it will say ‘if you institute these rules and a disaster happens, you will succeed or you will fail.’”

Wallace sees the model as a diagnostic tool that could help local, state, and federal governments shed light on the vulnerability of certain organizational features. It could also aid in the development of more flexible, responsive approaches to risk management, which is key to improving organizational responses to extreme events, according to the researchers.

“When a group of people are ingrained in an organization, it can be difficult to identify the day-to-day operations or procedures that could potentially become roadblocks when responding to certain situations,” Wallace says. “This model will be a tool for organizations to study and reflect on the ways their culture affects their ability to function.”

Two University of Washington researchers, Peter May and Bryan Jones, serve as co-principal investigators on the project, which also will consider how states assign risk priorities, and how the federal government influences those priorities. Other researchers include: Rachel Dowty and Colin Beech, two doctoral candidates in the Department of Science and Technology Studies (STS) at Rensselaer, and Yao Zheng, a senior in mechanical and nuclear engineering at Rensselaer.

Rensselaer’s portion of the collaborative research is funded by a three-year $299,578 National Science Foundation (NSF) Human and Social Dynamics (HSD) grant. An NSF Small Grant for Exploratory Research (SGER) funded the Rensselaer team’s initial travel to New Orleans.