On December 24, 1913, a Christmas party was being held on the second floor of a hall in Calumet, Michigan. The party had been organized by the Western Federation of Miners for the families of mine workers who were in facing what would prove to be a lengthy strike. Despite the tense mood, the party was meant to lift the spirits of everyone present. At least, it was until someone (nobody knows who), suddenly yelled "Fire!" and triggered a panic. The four hundred people present all rushed down the stairs and began searching for exits. Though there are differing accounts of what happened and why people panicked, seventy-three men, women, and children died, mostly by being crushed to death during the mad rush for escape. There was no fire.
On June 6, 1944, a fire broke out at a performance of the Ringling Brothers and Barnum and Bailey Circus in Hartford, Connecticut. About 168 people died, many of whom were trampled by audience members trying to flee the scene. Others were found to have asphyxiated due to being buried under dozens of bodies. Only a small percentage of the casualties actually died from fire or smoke inhalation.
In January, 2005, 291 Hindu pilgrims attending the annual Kalubai Jatra pilgrimage at the Mandhardevi Kalubai temple in India's Maharashta district died when a stampede broke out. Witnesses later stated that the stampede began with a fire caused by exploding gas canisters in nearby shops. There were far more deaths from the stampede that followed than the actual fire.
In virtually every natural disaster, terrorist attack, or fire, there is always the risk of mass panic which can often cause more casualties than the disaster itself. While "fight or flight" is a natural human response when people think they are danger, this instinctual reaction may lead us to make very unwise decisions, especially if we are part of a large group of people trying to do the same thing. Even when fire escapes and exits are clearly marked, the danger associated with panic can't be underestimated.
Unfortunately, researchers trying to study panic in emergency situations often have difficulty studying how people react in real-life situations. Along with the problems associated in questioning survivors after an actual disaster, there are also the ethical issues involved in studying panic behaviour in laboratory experiments. To get around these difficulties, researchers have been working with computer simulations intended to mimic actual emergencies. An intriguing new example of this was recently published in the Journal of the Royal Society Interface. This new study examined crowd behaviour in a virtual environment and made some surprising findings about how people can react to danger. Conducted by a team of researchers at the Max Planck Institute for Human Development, Disney Research Zurich, ETH Zurich, and Rutgers University, the study consisted of different experiments using 36 participants interacting in a 3-D virtual setting which could be manipulated in different ways. This allowed to researchers to simulate crowd behaviour in high and low-stress conditions..
Each participant was placed in front of a computer screen that provided a first-person view of the environment, including the avatars of the other participants. Using a computer mouse and keyboard, they could navigate freely in the virtual environment. By offering financial compensation and setting time limits, the researchers were able to create high-stress conditions in which each participant was expected to carry out specific tasks. What they found was that the way people behaved in the virtual environment greatly resembled how people often behave in real-life.
To simulate a real-world emergency, the researchers had the participants evacuate from a large virtual room through exits that acted as bottlenecks. As the researchers expected, the density of participants leaving the room grew as the bottleneck grew smaller. Still, under non-stress conditions, participants showed little difficulty exiting in an orderly fashion with little real conflict.
To simulate an emergency evacuation under high stress, the researchers set up four exits from the virtual room but, unknown to the participants, three of those exits were blocked. For some of the participants however, arrows were provided showing the correct exit and all other participants were aware that some would be given this information. In this experiment, participants were paid in points that could be converted into money afterward. The more points you had, the more money you could collect later. To add to the stress, the participants evacuated the room under different conditions. In the first condition, participants only had fifty seconds to leave the room and would receive 50 points if they succeeded. They were also penalized points if they ran into anyone else while evacuating. In the second condition however, participants were penalized 100 points for failure (with no reward for succeeding). There were also simulated fires at several of the blocked exits along with flashing lights and lower lighting overall to boost the tension participants experienced.
Results showed that participants were far more likely to collide with one another under high-stress conditions, even if it meant losing a considerable amount of points, in order to escape the room in time. The level of crowding was also much greater under high-stress conditions, often to the point of violating safety standards. The greatest amount of crowding occurred in (i) areas where a decision needed to be made, (ii) areas surrounding the exit where bottlenecks occurred and caused congestion, and (iii) dead ends where the flow of people returning after exploring a wrong option encountered the flow of those moving in the opposite direction. Since everyone knew that some participants would have correct information about which exit to use, there was also considerable "herding" behaviour with most people deciding to go in the same direction, even if the direction happened to be the wrong one.
Though there is no way to be sure whether the behaviour seen in simulated experiments like these exactly match how people really behave in life-threatening situations, it seems clear that bottlenecks and lack of information can often lead to dangerous outcomes in emergencies. More research is definitely needed but these results highlight the critical importance of well-marked exits and good emergency planning to avoid the kind of serious injury and loss of life that scan often occur in situations where people need to evacuate in a hurry.
So pay more attention to your surroundings when you go out, especially if you are in a crowded theatre or stadium.. Know where the exits are, be alert to fire alarms and emergency broadcasts, and, most of all, don't panic. The life you save may be your own.