See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/286168094

The Theories of Accident Causation

Chapter · December 2015

DOI: 10.1016/B978-0-12-800113-4.00005-5

CITATIONS READS

0 2,990

2 authors, including:

Whitney DeCamp
Western Michigan University
40 PUBLICATIONS 316 CITATIONS

SEE PROFILE

All content following this page was uploaded by Whitney DeCamp on 17 February 2019.

The user has requested enhancement of the downloaded file.

 The Theories of Accident Causation

Whitney DeCamp, Ph.D.

Kevin Herskovitz, M.S.

“They’re funny things, Accidents. You never have them till you’re having them.”
- Eeyore (from A. A. Milne’s Winnie The Pooh)

Accidents occur everyday and, one way or another, will impact virtually everyone.
During the year 2012, there were more than 2.8 million on-the-job nonfatal injuries in the United
States (Bureau of Labor Statistics, 2013). That same year, there were also 4,628 on-the-job
fatalities (Bureau of Labor Statistics, 2014). Note that other incidents, such as workplace
violence, add to this data as well, though a vast majority are related to accidents. Further
highlighting the impact of accidents is the costs absorbed by organizations. Cost considerations
include workers’ compensation case management, the use of paid-time-off/sick time, short- and/
or long-term disability, worker replacement costs (i.e. training of an employee to replace the
injured worker), and time and money spent investigating the accident with follow up corrective
actions which can include policy and/or equipment change or upgrades.
Even further expanding upon the impact of accidents is the great number of accidents that
do not result in injuries. In an early study of accidents, H. W. Heinrich (1950) found that for
every serious injury, there were 29 minor injuries and 300 accidents resulting in close calls. At
that rate, even assuming that all injuries (major or otherwise) are included in the official
statistics, there would be an additional 42 million accidents that go unreported. Figure #1
demonstrates Heinrich’s (1950, p. 24) “Foundation of a Major Injury,” sometimes also referred
to as the Injury Pyramid.

Figure #1: Foundation of a Major Injury

 It’s important to note that OSHA uses the term “incident” to refer to these events, while
The National Safety Council – and typically the general public on the whole – uses the term
“accident.” These terms are essentially interchangeable, but for the purposes of this text, we will
use “accident” as defined by The National Safety Council: An accident is an undesired event that
results in personal injury or property damage. This definition implies two important points.
First, accidents are unavoidable; the chance of one occurring will virtually always be present.
Second, the chance of an accident occurring is a variable that can be changed. While it is
impossible to prevent all accidents, it is possible to decrease their rate of occurrence.
Understanding the cause of such phenomenon is key to decreasing the rate at which
accidents occur. Determining the true root cause of each accident is the only way to formulate
effective prevention strategies. Presented below are a few of the most common theories used to
explain accidents. As with theories discussed in other chapters, these are not perfect and will not
explain every accident in full detail. Rather, they provide a nomothetic explanation that seeks to
explain what usually happens and attempts to address the most common underlying causes.

Heinrich’s Domino Theory

Heinrich’s (1950) theory explains accidents using the analogy of dominos falling over
one another and creating a chain of events. While this theory is not the most advanced or
complex theory, it is especially noteworthy as one of the first scientific theories used to explain
accidents. It is often still referenced today, seven decades later.
When dominos fall over, each tips the next enough to push it over and continue the
process until all the connected dominos have fallen. However, if just a single domino is removed,
the entire process ceases. Heinrich explains accident causation in the same way.

Figure #2: Heinrich’s Domino Theory

 As you can see from the figure, Heinrich identified five stages of accident causation. The
first stage, the social environment and ancestry, encompasses anything that may lead to
producing undesirable traits in people. It is worth noting that Heinrich’s inclusion of genetics and
ancestry is very much a product of the time it was written. A modernized version of this theory
would likely use the term “inherited behavior,” similar to how alcoholism and temperaments can
be inherited. This stage of accident causation is quite similar to the social learning theories
discussed in the criminological theories chapter of this textbook.
The second stage, faults of a person, refers to personal characteristics that are conducive
to accidents. For example, having a bad temper may lead to spontaneous outbursts and disregard
for safety. Similarly, general recklessness can also be one of the manifestations of poor
character. Ignorance, such as not knowing safety regulations or standard operating procedures, is
also an example of this stage.
The third stage, an unsafe act or condition, is often the identifiable beginning of a specific
incident. Unlike the first two stages, which affect the probability of accidents occurring, this
stage is closer to the accident in terms of temporal proximity. This can include a specific act that
is unsafe, such as starting a machine without proper warning, or failing to perform appropriate
preventative actions, such as using guardrails or other safety measures. In essence, this stage
entails acts (or failures to act) that occasionally cause accidents.
The next stage, logically, is the accident itself. This, in and of itself, needs little
explanation. It is, simply, when something occurs that is undesirable and not intended. The final
stage, injury, is the unfortunate outcome of some accidents. Whether an injury occurs during an
accident is often a matter of chance and not always the outcome. This relationship highlights the
relationships between stages in terms of causality. An accident occurring is not a sufficient cause
for an injury, but it is a necessary one. Similarly, the undesirable characteristics in stage two do
not always occur in poor environments, but could not occur without such environments.
Given this necessary causality, the most important policy implication is to remove at least
one of the dominos, which can in turn lead to a healthy subculture through positive accident
prevention training and seminars. An organization may not be able to weed out all of the people
with undesirable characteristics, but it can have a procedure in place for dealing with accidents to
minimize injury and loss.

Ferrell’s Human Factor Model

Unlike Heinrich, who explained accidents with a single chain reaction in vague terms,
Ferrell’s model incorporates multiple causes and is very specific about these causes (Heinrich,
Petersen, & Roos, 1980). Additionally, Ferrell defines accidents in terms of being the result of an
error by an individual. As such, he explains his theory using the assumption that accidents are
caused by one person.
Ferrell identifies three general causes of accidents: overload, incompatibility and
improper activities. Each of these are actually broad categories that contain several more specific
causes. Improper activities is perhaps the simplest of the concepts, as it encompasses two
straightforward sources of accidents. First, it is possible that the responsible person simply didn’t
know any better. Alternatively, he or she may have known that an accident may result from an
action, but deliberately chose to take that risk. The incompatibility cause is slightly more
complex than improper activities. It encompasses both an incorrect response to a situation by an
individual, as well as subtle environmental characteristics, such as a work station that is
incorrectly sized.
The remaining cause, overload, is the most complex of Ferrell’s causes. It can further be
broken down into three subcategories. First, the emotional state of the individual accounts for
part of an overload. These states include conditions such as unmotivated and agitated. Second,
the capacity refers to the individual’s physical and educational background. Physical fitness,
training, and even genetics play a part of this. Situational factors, such as exposure to drugs and
pollutants, as well as job related stressors and pressures, also affect one’s capacity. Finally, the
load of the individual can also contribute to an overload. This includes the difficulty of the task,
the negative or positive effects of the environment (noise, distractions, etc.), and even the danger
level of the task. Separate from each other, overload, incompatibility, and improper activities can
all cause a human error to occur, which can lead to an accident.

Petersen’s Accident/Incident Model

Petersen’s model is largely an expansion upon Ferrell’s Human Factor Model (Heinrich,
Petersen, & Roos, 1980). The notion of an overload, caused by capacity, state, or load, is very
similar to Ferrell’s work. However, a few changes and refinements do exist. First, Petersen
conceptualized the environmental aspect of incompatibility (work station design and
displays/controls) as a different part of the model, calling them ergonomic traps. Additionally,
Petersen also separated a decision to err from the overload cause. Further, Petersen also specified
separate reasons to choose to err. These reasons include: a logical decision due to the situation
(primarily for financial cost and temporal deadlines), an unconscious desire to err (psychological
failings), and perceived low probability of an accident occurring. The latter of those reasons, the
perception of low accident probability, can include both actual instances of an accident being
extremely unlikely, as well as the natural inclination of a human to disregard his or her own
mortality. This aspect of Petersen’s model is akin to criminology’s rational choice perspective
(see the criminological theories chapter), as it makes the same assumptions of human rationality
and hedonistic calculus.
Another noteworthy contribution is Petersen’s recognition that human error is only part
of a larger model. A system failure, the inability of the organization to correct errors, was added
as a possible mediator between errors and accidents. These failures have a range of possible
occurrences. The failure of management to detect mistakes and a lack of training are but two
examples of systems failures. Even poor policy itself can lead to a systems failure that does not
prevent an accident from occurring following a human error.

Systems Models

Most of the theories thus far discussed focus on human errors and environmental flaws. A
systems model theory approaches the relationship between persons and their environments
differently. Rather than the environment being full of hazards and a person being error prone, a
system model view sees a harmony between man, machine, and environment. Under normal
circumstances, the chances of an accident are very low. Once someone or something disrupts this
harmony by changing one of the components or the relationships between the three, the
probability of an accident occurring increases substantially.
 Another aspect of the systems model is what is referred to as risk-taking. Whenever
someone chooses to do something, there is an associated risk (Firenze, 1978). Smaller tasks and
risks are often calculated on an unconscious level. For example, when one chooses to drive to
work each morning, that person weighs the risks (slight chance of being in a car accident) and
the benefits (making a living) and decides the benefits outweigh the risks. This hedonistic
calculus, as with Petersen’s model, is quite similar to the rational choice perspective. Just as
potential criminals may weigh the risks of being caught, managers, safety specialists, and
supervisors consider the chances of injury or financial loss. The decision to move forward with
the task is only taken when it is decided the potential benefits outweigh the potential loss. In a
real life example of this type of risk taking behavior, Ford was once accused of deciding that the
risks of releasing a defective vehicle (several fatalities that would result in wrongful deaths) were
not enough to outweigh the benefit (not having to pay to fix all the defective vehicles). While
subsequent reports have shown that this accusation is false to a large extent (Schwartz, 1991),
this particular case has often been cited as an example of the ethical and financial calculations of
risk-taking.
Firenze (1978) suggests considering five calculated risks and benefits:

1) Job requirements
2) The capabilities and limitations of the worker in relation to her or her job
3) The potential gain upon succeeding
4) The potential consequences upon failure
5) The potential loss of not attempting the task

Additional information about these five factors becomes available through feedback after an
initial attempt. In other words, a common task previously taken has well known risks and
benefits, while a new task often has more unknown factors.

Reason’s Swiss Cheese Model

Perhaps the most popular systems theory on accident causation is James Reason’s Swiss
Cheese Model, originally proposed in 1990. According to this theory, every step in a process has
the potential for failure (Reason, 2003). Each layer of defense is represented by a slice of Swiss
cheese, and the possible problems or failures in that defense are represented by the holes in the
cheese. There are two types of failures that can occur: active and latent. Active failures are
unsafe acts that directly contribute to an accident. Latent failure are conditions that exist that
may lay dormant for a period of time until they lead to an accident. An example of a latent
failure could be the lack of a policy describing how a given work task should be completed
safely.
For an accident to happen, the holes have to line up – no layer of defense caught the
problem. If the holes do not line up, then the problem was caught, and no accident occurs.
For example, the first layer could be the policy that addresses work needing completed.
The next layer could be the personal protective equipment (PPE) required to be worn to complete
the task. If the policy does not address the PPE, or requires the wrong PPE, the problem passes
that line of defense and through the first hole. Choosing to not wear the PPE, or wearing the
wrong PPE as described in the policy allows passage through the next hole. Assuming there are
no additional layers of defense, these conditions will allow an accident to happen. Figures 3
illustrates this model. Despite successive layers of defense, small holes in each layer allows for
some possibility that the defense will be ineffective. If the circumstances are right, these holes
can align and allow an accident to occur.

Figure 3: How an accident occurs:

(Image source: Davidmack, Wikimedia Commons, CC BY-SA 3.0)

The Integration of Theories and General Safety Program Implications

It is important to understand that each theory of accident causation does not explain every
accident. Each theory explains only a portion of accidents, and all of these theories are
incomplete as evidenced by the number of published works countering each theory. It is
therefore important to recognize that true accident prevention - the reduction of the
probability of accidents - can only occur when all possible causes are addressed. Focusing
on only one or two theories is simply not enough. Further, there are numerous theories not even
briefly discussed in this chapter. Safety specialists and individuals with related duties are highly
encouraged to consult additional information about accident causation.
There are numerous program implications that can be derived even from the few theories
discussed in this chapter. Many of these are common sense, as they are often used practices.
First, most theories and models agree that human error is always a possible cause of accidents.
An effective strategy is to train employees carefully and continually. Better safety training and
increased knowledge and awareness of possible dangers can only decrease the chance of an
accident occurring.
Second, socialization and subculture are also a common thread in accident causation.
This further underscores the need for regular training and safety programs. An unsafe employee
not only increases the risk of causing an accident, he or she can also corrupt future staff and
make the problem grow exponentially. A safety awareness program is a good example of how to
approach this problem. Regular meetings and positive safety posters are some of the tactics an
awareness program can utilize. Keeping employees motivated to stay safe is another contributing
factor to a successful safety program. The two-factor theory of motivation, also called the
Motivator-Hygiene theory (Herzberg, Mausner & Snyderman, 1959), suggests that employees
should be exposed to motivators (positive rewards) and hygiene factors (routine parts of a job,
such as a good working environment, that prevent dissatisfaction). Management should
understand the importance of maintaining a positive subculture and be trained with intervention
strategies for problem employees. Management buy-in into the safety program is also extremely
important.
Third, the physical environment is also an important aspect of accident causation that
must be addressed. In addition to obvious implications (guard rails, safety warnings, hardhats,
etc.), the subtle relationships between man and the environment must also be considered.
Ergonomic designs, often used to increase productivity, can also increase a worker’s comfort.
Stress and boredom can play a role in human error, so keeping agitators to a minimum through
ergonomic designs may also be helpful.
Finally, don’t rely solely on conventional thinking. Safety, like Security, is often reactive
in nature; being proactive and using outside-the-box thinking can further a safety program
substantially. Offering incentives and rewards to safety-oriented workers is a relatively new
approach that, at minimum, gets attention. The status quo can also be challenged by simply
asking if more can be done to keep employees safe. An important part of any program is a
regular evaluation to make sure it is working. Statistical analyses of accident rates, surveys of
individuals’ perceptions of safety and inspections by safety specialists are all examples of
potential indicators of program effectiveness. Confirmation of the findings by using multiple
indicators is important to validate findings. If possible, different programs should be
implemented within different environments so that effectiveness (or lack thereof) can be
compared. If a program is not working, ask how it could be better. If it is working, ask the same.

Emerging Trends

The main goal of any good safety program is to meet legal requirements and prevent
accidents, but since accidents can never be completely prevented, a secondary goal is to be
prepared for the inevitable. In the aftermath of the terrorist attacks of September 11, 2001,
interest in emergency management has heightened. However, such interest has largely
overlooked individual accidents, especially on the small scale, in favor of Terrorism and other
acts of malicious or intentional harm (Haddow & Bullock, 2006). In the wake of Hurricane
Katrina, the focus of emergency management has at least partially included non-malicious
events. Where does this leave the safety specialist concerned with individual accidents and
workplace safety?
Despite the lack of interest by the public and the media, accident prevention continues to
be an important topic. Fiems and Hertig (2001) noted that fines by the Occupational Safety and
Heath Administration (OSHA) have been increased and more being imposed more liberally than
in years past for violations of unsafe working conditions. Additionally, more states are
legislating safety standards and security organizations are placing more emphasis on providing
both security and safety. Accident prevention increasingly rests with the individual organization.
Organizational economic situations may play a role in accident prevention as well. A
fiscally healthy organization may not see the Workers’ Compensation “bottom line” as a
significant number, or they may see it as something out of their control. Struggling
organizations do right by critically looking at every dollar spent, and focusing energy
appropriately in controlling costs, especially considering the $58,000 cost of the average
disabling workplace injury (National Safety Council, 2012). Not only can reducing accidents be
 a huge money saver, it can easily be argued it’s the right thing to do. A workforce that feels safe
is better equipped – literally and figuratively – to be more productive as well.
Since Heinrich’s Domino theory in 1936, knowledge about accident causation and its
counterpart, accident prevention, has grown remarkably. What once was the only theory
explaining accidents has served as the foundation of a discipline home to many theories,
perspectives, and implications. This increase in knowledge, both among safety professionals and
other individuals, has a substantial impact on safety in the modern world. Together with
technological advances in safety and communication, accident causation theory and accident
prevention are more advanced than ever, as at least partially evidenced by the 32% reduction in
non-fatal injury and illness rates over the last 10 years (Bureau of Labor Statistics, 2013).
Understanding and quantifying causation will lead us to a more scientific approach and greater
cost-effective intervention strategies.

References

Bureau of Labor Statistics. (2014). Revisions to the 2012 Census of Fatal Occupational Injuries
(CFOI) counts, April 2014. Retrieved from:
http://stats.bls.gov/iif/oshwc/cfoi/cfoi_revised12.pdf
Bureau of Labor Statistics. (2013). Employer-Reported Workplace Injuries and Illnesses – 2012.
Retrieved from: http://www.bls.gov/iif/oshwc/osh/os/ostb3569.pdf
Fiems, R. A., and Hertig, C. A. (2001). Protection Office Guidebook. Naples, FL: International
Foundation for Protection Officers.
Firenze, R. J. (1978). The Process of Hazard Control. New York: Kendall/Hunt.
Haddow, G. D. and Bullock, J. A. (2006). Introduction to Emergency Management (2nd ed.).
Oxford: Butterworth-Heinemann.
Heinrich, H. W. (1950). Industrial Accident Prevention (3rd ed.). New York: McGraw Hill.
Heinrich, H. W., Petersen, D., and Roos, N. (1980). Industrial Accident Prevention. New York:
McGraw-Hill.
Herzberg, F., Mausner, B., and Snyderman, B. B. (1959). The Motivation to Work. New York:
John Wiley.
National Safety Council. (2012). Estimating the Costs of Unintentional Injuries. Retrieved from:
http://www.nsc.org/news_resources/injury_and_death_statistics/Pages/
EstimatingtheCostsofUnintentionalInjuries.aspx.
Occupational Safety & Health Administration (OSHA). (n.d.) Accident/Incident Investigation.
Retrieved from: https://www.osha.gov/SLTC/etools/safetyhealth/mod4_factsheets_
accinvest.html
Reason, J. (1990). Human Error. Cambridge: Cambridge University Press.
Schwartz, G. T. (1991). The myth of the Ford Pinto case. Rutgers Law Review, 43, 1013-1068.

View publication stats