Managing Flash Floods: A Cultural Perspective of
Risk Perception
Ashley R. Coles
University of Arizona
Abstract
This research investigates how culture affects risk perception and behavior in the
context of flash floods. Approximately 1,000 citizens of Tucson, Arizona will fill out
surveys assessing cultural characteristics as well as historical and hypothetical behavior
during flash flood scenarios. Neighborhoods will be selected from those with regular
street flooding problems and the highest amount of diversity regarding race and
socioeconomic status. The cultural factors explored here include trust, social
incorporation, self- and community-efficacy, social autonomy, and time orientation.
Multiple linear regression analysis will reveal which of these factors predict whether
people say they would cross a flooded roadway or wash. If cultural variation appears to
predict stated behavior, this study confirms that risk managers must consider cultural
variation in order to effectively execute hazard mitigation.
Introduction
Risk perception research has proven useful in natural hazards mitigation by
investigating how people might respond to hazard information, and how risk managers
might alter warnings or dissemination methods to stimulate a wider practice of what they
consider the “appropriate” response. These studies often assume that actions result from
either universal human nature or from psychological traits unique to each individual. In
most cases, risk managers attribute failure to respond appropriately to irrationality or lack
of information (Douglas, 1992; Jasanoff, 1998; Slovic, 1999; Frewer, 2004). Thus, risk
communication research and practice have focused on tailoring the message until people
understand and cease engagement in risk-taking behaviors (Douglas, 1992; Kasperson &
Kasperson, 2005). However, risk perception studies often fail to include the assessment
of how factors such as cultural context and social networks may affect how people
perceive threats and respond to warnings, not to mention how cultural norms may
influence the distribution of vulnerability. The studies that do consider the effects of
culture tend to focus on a singular aspect, and many do not contain information that risk
managers may utilize in practical application.
This study aims to fill this research gap by working directly with risk managers to
create a comprehensive understanding of how several cultural factors affect propensity to
engage in risk-taking behavior, using the example of crossing a flooded roadway or wash.
Investigators will examine five key cultural factors known to affect risk perception and
how these factors predict participants’ historical or hypothetical behavior in flash flood
scenarios. These factors include (a) trust in the hazard science as well as the government;
(b) social incorporation, or the extent of social networks; (c) self- and communityefficacy, which describes how confident people feel in their ability to handle both
extreme events and ordinary life; (d) social autonomy, the degree of freedom to fill any
social role; and (e) time orientation, whether people focus on the past, present, or future.
Survey questions will elicit responses that reveal each participant’s cultural
characteristics, the predicting power of which will be determined through multiple
regression analysis. The survey will also present past and hypothetical situations as both
a series of questions and a few short vignettes for which participants must decide whether
or not they would cross the flooded street or wash. Investigators hypothesize that
propensity to claim tendencies to engage in risk-taking behaviors will vary among
cultures, and that the multiple regression analysis will reveal which cultural factors have
the strongest predictive powers. The information gathered in this project may help flood
control managers understand risk-taking behavior as it varies with culture, rather than
attributing such behaviors to irrationality or lack of information.
Literature Review
Douglas (1992) suggests that the lens through which to view risk perception and
response is culture. Culture refers to a way of life learned from and shared by a social
unit, including – but not limited to – attitudes, beliefs, values, and habits. Values and
social norms shape an individual’s worldview, and therefore affect how the individual
will understand threats and determine appropriate methods of threat avoidance or
amelioration (Douglas, 1992). Tobin and Montz (1997) also recognize the need to
consider non-geophysical factors to understand all aspects of natural hazards. They
define natural hazards as the “potential interaction between humans and extreme natural
events (p. 5),” and natural disasters as the actual interactions that have impacts on society
economically or physically. They claim that humans are not only necessarily part of this
system, but that everyday social factors such as goals, concerns, values, and institutions
play a large role in how people perceive, mitigate, and respond to hazard events. This is
not to say that other factors, such as previous experience with a particular hazard, do not
influence behavior before, during, and after an event, but that in fact the enormous impact
of culture on people’s values and beliefs may affect just how these other factors play a
role in an individual’s actions.
Cultural psychologists explain that culture has such a strong effect on an
individual’s actions because culture and psyche mutually and constantly constitute one
another. Norms, values, and practices shape how individuals process information and
make decisions, and the resulting behaviors either maintain or reshape those norms,
values, and practices (Fiske et al., 1998; Kitayama & Markus, 1995; Markus et al., 1997;
Shweder, 1995). While many regard culture as a national or multi-national variable, it
also includes subgroups based on “gender, ethnicity, religion, cohort or generation,
historical period, profession, social class, and country of origin” (Kitayama & Markus,
1995, p. 368). To obtain a comprehensive understanding of how culture and behavior
construct one another, researchers must consider multiple levels as well as multiple
factors of culture.
Theoretical and applied hazards research tend to focus on trust, social
incorporation, self- and community-efficacy, social autonomy, and time orientation as the
key cultural factors that affect risk perception and response (e.g., Douglas, 1992; McGee
& Russell, 2003; Slovic, 1999). Trust holds such strong influence over an individual’s
risk perception because it serves as an adaptation to an increasingly complex
environment. To reduce the necessity for learning every detail about every possible risk,
individuals place trust in those whom they believe to both have the knowledge and
willingness to share accurate information about particular risks (Slovic, 1999; Siegrist
and Cvetkovich, 2000; Savadori et al., 2004; Lang & Hallman, 2005). In the case of
flash floods, the public must place trust in both the science behind risk assessment as well
as the government that sponsors such research and holds responsibility for flood
mitigation. The fragility of trust makes the issue much more salient, particularly in
matters of risk management. As Slovic (1999) notes, “distrust, once initiated, tends to
reinforce and perpetuate distrust” (p. 698). Once an institution has earned a reputation
for distrust, individuals tend to avoid contact or automatically reject even trust-building
information. The resilience of opinions formed based on distrust thus often creates a
barrier that prevents experts from communicating new or amended hazard information.
The search for appropriate media and messages for risk communication becomes futile
when the public lacks trust in the messenger, because “if trust is lacking, no form or
process of communication will be satisfactory” (Slovic, 1999, p. 697).
Social incorporation as defined by Douglas (1992) refers to the extent of an
individual’s social network as well as the degree of connection between its members.
Extensive social networks facilitate appropriate precautions and actions for hazard events
by increasing the likelihood that information will be heard and believed by individuals.
Many studies have shown that people seek confirmation of the severity of a hazard and
the need to prepare or respond appropriately from a variety of sources before they will
take action (Mileti, 1995; Brilly & Polic, 2005). These sources may be different
agencies, government or emergency management officials, or the media, but often people
consult with friends, family, or neighbors before making a decision. Scherer and Cho
(2003) have developed the social contagion theory, which states that individuals model
their attitudes and behaviors after those of the members of their social networks. This
helps to explain studies that have shown that people evacuate more readily if they know
that neighbors, friends, or family members are evacuating (Mileti 1995; Dow & Cutter,
2000). The social amplification of risk framework (SARF) provides a theoretical model
for how information about hazards and risk perceptions may be amplified or attenuated as
people consult with their social networks (Kasperson, 1992; Masuda & Garvin, 2006).
Rumors or information will be propagated as it fits preconceived mental models, beliefs,
or values (Kasperson, 1992). This means that if the hazard message conforms to these
predetermined notions of reality, the receiver will attempt to spread information and
amplify the apparent risk. Otherwise, the threat will be downplayed by either circulating
contrary information or failure to propagate any information at all. Kasperson (1992)
describes culture as a “super-variable” that affects all stages within SARF, including
amplification and attenuation processes. Since culture determines an individual’s
worldview, it is likely to affect whether individuals amplify or attenuate hazard
information (Kasperson, 1992; Masuda & Garvin, 2006). Sáenz (2003) would agree, and
states that just educating the public is insufficient, because “individuals and groups
possess a series of subjective or affective components that impede or allow for the
incorporation of new knowledge” (p. 124).
Efficacy describes whether individuals or groups determine that a risk may
actually be reduced as well as whether they are capable of taking the proper precautions
or reactions during a hazard event (Tobin & Montz, 1997; Paton & Johnston, 2001).
Both self- and community-efficacy are primarily related to an individual’s apparent locus
of control, the latter referring to an aggregate of individuals. An internal locus of control
puts the individual in charge of his or her own destiny, and the resulting higher sense of
efficacy increases the likelihood that an individual will react to warnings and take
measures to secure their life and property. However, should the individual assume they
have complete control over their destiny, regardless of their awareness of the hazard in
question, they may feel invincible and take no precautionary or reactionary measures. On
the community level, a high sense of efficacy will encourage and foster collective action
and cooperation. Individuals with an external locus of control tend to feel helpless in
hazard situations and depend on others for protection, rather than attempt to protect
themselves. Certain religious beliefs contribute to fatalistic tendencies because people
may assume that they cannot alter their pre-determined destiny by taking precautionary or
reactionary measures (Douglas, 1992; Inelman et al., 2004).
Social marginalization also produces this sense of fatalism by reducing resources
and options available to people during exposure to a hazard. Social autonomy measures
variation in both real and perceived access to information and materials as a function of
the degree to which individuals have the freedom to move between various social roles
(Douglas, 1992). Many factors determine social status including age, gender, race,
ethnicity, education, socioeconomic status, or any other standards by which societies
produce social hierarchies. Even in the United States, widely considered an egalitarian
society, certain groups must cope with downward social constitution. Markus, Steele,
and Steel (2000) define downward social constitution as
the experience of being in a setting where, based on a given group identity, one is
exposed to a potentially limiting and devaluing concert of representations,
historical narratives, possible judgments, treatments, interactions, expectations,
and affective reactions. (p. 235)
Thus, not only will certain groups find access to resources limited, they will come to
expect such limitations. This may drastically affect decision-making not only by
restricting which options an individual has available, but also which options an individual
even considers.
As mentioned before, time orientation refers to whether individuals focus on the
past, present, or future. Cultures that focus on the past tend to have difficulty
conceptualizing unforeseen or unpredictable events (Douglas, 1992). This produces
beliefs that because an event has failed to occur in the past, it probably will not occur in
the future. When a major event does occur, however, many see it as anomalous and
believe that it will not happen again within their lifetime (Tobin & Montz, 1997). This
also fails to account for natural and anthropogenic changes in the environment, which
will affect the magnitude and frequency of future hazards. On the other hand, little
emphasis on the past causes people to disregard the frequency of hazard events despite
experience with them, and little emphasis on the future leads to the tendency to take
action (or fail to take action) in order to improve current conditions at the expense of
those to come. In the context of natural hazards, present-focused individuals may opt out
of insurance and invest the money they would have spent in other ventures or luxuries.
This occurs despite knowledge of past events and probability of future events to come.
This does not necessarily reflect lack of reasoning, as Douglas (1992) is quick to point
out, but rather indicates a preference. Wisner et al. (2004) would agree that these
individuals have merely decided that their current standard of living is more important
than their future one, which may or may not include a hazard event. Those with a
particular concern for the future will likely exercise more caution for fear of the
unknown, and will therefore engage in risk-taking behavior less often.
Theoretical Framework
This research falls within the psychometric paradigm commonly used in hazards
and risk perception research (Slovic, 2000). However, previous studies conducted under
this paradigm mostly focus on psychological functioning at an individual level. This
project integrates cultural psychology theory by attempting to discern how culture affects
psychological processes, which then affect behavior. This process will be examined by
assessing the five aforementioned cultural characteristics, and then determining how
these characteristics differ among the groups, as well as how they influence risk
perception and behavior. The following factors are expected to increase the likelihood
that an individual will report that they have historically or would hypothetically cross a
flooded roadway or wash:
Hypothesis 1: Low trust in the science behind risk assessments, the government,
or both will increase reports of risk-taking behavior because individuals will not accept
information from a distrusted source.
Hypothesis 2: Low social incorporation will increase reports of risk-taking
behavior because individuals have less opportunity to receive new information.
Hypothesis 3: High self-efficacy will increase reports of risk-taking behavior
because individuals will believe themselves capable of surviving the risk unscathed.
Hypothesis 4: Either high or low social autonomy may increase reports of risktaking behavior. Those with low social autonomy may feel they have no choice but to
take the risk. However, those with low social autonomy have more to lose by taking the
risk, so those with high social autonomy will probably report more risk-taking behavior
than those with low social autonomy.
Hypothesis 5: Past or present time orientation will increase reports of risk taking
behavior because those individuals have less concern for the future and the consequences
of their actions.
Method
Tucson, Arizona provides an optimal site for this case study for both physical and
political reasons. Hundreds of streets and intersections flood when it rains due to a
combination of poor drainage systems and numerous wash crossings without bridges.
Flood control managers have created several sign and road barrier projects in an attempt
to deter motorists from driving into flooded roadways. An Arizona state law now
obligates individuals to pay fines in order to offset costs incurred by the city if they drive
around a barricade and require rescue (Arizona Revised Statutes, 2006). Despite these
and other programs designed to inform the public about the dangers of driving through
flooded roadways and otherwise discourage such behavior, losses to life and property
continue to occur. This, coupled with Tucson’s rich variety of cultural and
socioeconomic demographics provides a perfect context within which to test the theory
that cultural factors affect risk perception and behavior during a flash flood. In addition,
interactions with flood risk managers in Tucson have revealed the tendency to assume
irrationality or lack of education as the cause for this particular risk-taking behavior. To
ensure that the information gathered through this research provides a substantial degree
of utility for the risk managers, they will be included throughout most stages of the
project. The Climate Assessment for the Southwest research institution will host and
provide support for initial, intermediary, and final meetings between the flood risk
managers and the survey research team. Background information, research plans, and
results will be discussed so that the final results may be used in future implementations of
risk management policy.
Ideally this research would target members of the population who have crossed
flooded roadways or washes and compare them to those who have not. However, the
ephemeral nature of the former group inhibits efforts to locate them. Many have crossed
without any trouble, and many have perished in the process. This project therefore
requires widespread dispersal of surveys in order to capture members of many different
cultural groups who have, have not, probably will, and probably will not cross flooded
roadways and washes. Locations of flood barriers and problem areas have been mapped
out in relation to neighborhood association boundaries. Surveys will be mailed to
approximately 1,000 Tucson residents in neighborhoods with at least one flood barrier or
sign, giving preference to neighborhoods that the U.S. Census 2000 tract data show as the
most culturally and socioeconomically diverse. In order to potentially increase response
rate, local neighborhood associations will be contacted in advance and asked for support.
Each household will receive a copy of the survey in both English and Spanish, along with
a brief description of the research and a reply-paid envelope. To further maximize
response rates, a postcard reminder will arrive a week later, and a second copy of the
questionnaire will arrive after another two weeks (McGuirk & O’Neill, 2005). To
compensate respondents for their participation and to encourage response, all respondents
that return a survey will be entered in a random drawing for a prize.
The social stigma attached to crossing a flooded roadway or wash may pressure
participants to answer strategically about past and possible future experiences in order to
appear more rational. The instructions provided with the survey will attempt to
counteract this behavior by carefully explaining that the investigators seek information
about the logical thought processes that lead to the decision to engage in this behavior,
and do not assume that such a decision results from irrationality (Iarossi, 2006). This
should help to reduce some of the disparity between reported and actual behavior. The
use of a self-completed survey rather than an interview may also encourage honest
responses that participants might be reluctant to provide in face-to-face contact
(Buckingham and Saunders, 2004).
Questions will be directed so as to elicit information regarding the five cultural
factors outlined above through everyday practices and beliefs, as well as through explicit
situations related to flash floods. This information will be compared to participants’
stated typical or hypothetical responses to situations in which they encounter a flooded
street crossing. Types of possible answers to survey questions will include discrete
categories, Likert scale ratings, and rankings. Since multiple choice answers tend to limit
responses to those preconceived by the investigator, the survey will also include space for
optional short answer responses that will allow the participant to elaborate or explain
their selection (McGuirk and O’Neill, 2005). This survey design will allow investigators
to perform quantitative analyses on the data without sacrificing the opportunity for
participants to provide deep, rich, and varied information that may open discussions for
later studies.
Data will be stored in a Microsoft Excel spreadsheet, which will allow for the
performance of statistical calculations as well as simultaneous storage of the open-ended
responses with their corresponding multiple-choice answers. Quantitative analysis of the
results will include a multiple linear regression of the five cultural factors to see how well
each one predicts historical and hypothetical risk-taking behavior. This project requires
this type of analysis because the cultural groups and their effects overlap and confound
analysis of each mediating variable alone. Qualitative analysis will include the openended explicated responses, which will be coded by the investigators and then used to
theorize about the results, particularly where the hypotheses fail.
Results
The multiple linear regression analysis will reveal which of the five factors have
the strongest predicting power of participants’ stated historical or hypothetical behavior.
Groups of participants based on race or ethnicity, gender, socioeconomic status, level of
education, and age should vary among their responses to whether they have, have not,
would, or would not cross a flooded roadway or wash.
Hypothesis 1: Low trust in the science behind risk assessments, the government,
or both will increase reports of risk-taking behavior because individuals will not accept
information from a distrusted source. Investigators expect level of education to
positively correlate with trust in science. Groups that may display low trust in the
government because of historical experiences include Mexican Americans, American
Indians, and individualistic European Americans (for the latter, see Bellah et al., 1985).
Hypothesis 2: Low social incorporation will increase reports of risk-taking
behavior because individuals have less opportunity to receive new information.
Independent cultures, such as European Americans, tend to have smaller social networks
and less connection to those networks than those in interdependent cultures, such as
Asian Americans, Mexican Americans, and American Indians (Fiske et al., 1998;
Kitayama & Markus, 1995; Markus et al., 1997).
Hypothesis 3: High self-efficacy will increase reports of risk-taking behavior
because individuals will believe themselves capable of surviving the risk unscathed.
Investigators predict that European Americans, those with high socioeconomic status,
those with high education, young adults, and males will have higher self-efficacy than
other groups.
Hypothesis 4: Either high or low social autonomy may increase reports of risktaking behavior. Those with low social autonomy may feel they have no choice but to
take the risk. However, those with low social autonomy have more to lose by taking the
risk, so those with high social autonomy will probably report more risk-taking behavior
than those with low social autonomy. Groups with high social autonomy include
European Americans, those with high socioeconomic status, those with high education,
young adults, and males.
Hypothesis 5: Past or present time orientation will increase reports of risk taking
behavior because those individuals have less concern for the future and the consequences
of their actions. Investigators expect weaker correlations between particular groups and
time orientation, because factors such as low socioeconomic status may force individuals
to focus on meeting current needs instead of allowing for future preparation, while other
groups may have the option to prepare for the future but choose to focus on the present or
past.
Discussion
The analysis will show which factors have the strongest effect on participants’
self-reported risk-taking behavior, but will also show how these factors might work in
conjunction with or against each other. Hypotheses 1, 2, and 3 appear slightly more
robust because they lack the internally confounding factors found in Hypotheses 4 and 5,
but combinations with other factors may affect the magnitude and direction of influence.
Should the reported risk-taking behaviors tend to vary predictably by cultural
characteristics, flood risk managers will have to consider the ways culture may increase
risk-taking behavior rather than attribute this tendency to irrationality or lack of
information. The difficulty of assessing actual behavior requires the use of reported
behavior as a proxy, which does not always match, but nevertheless should still provide
useful findings for risk managers, hazards research, and cultural psychology research.
References
Arizona Revised Statutes (2006). Liability for emergency responses in flood areas. Title
28, Article 10, Section 910. Retrieved December 9, 2006 from
http://www.azleg.state.az.us/ArizonaRevisedStatutes.asp
Bellah, R. N., Madsen, R., Sullivan, W. M., Swidler, A., & Tipton, S. M. (1985). Habits
of the heart: Individualism and commitment in American life (pp. 142-153). New
York: Harper Row.
Brilly, M. & Polic, M. (2005). Public perception of flood risks, flood forecasting and
mitigation. Natural Hazards and Earth System Sciences, 5, 345-355.
Buckingham, A. & Saunders, P. (2004). The Survey Methods Workbook: From Design to
Analysis. Cambridge: Polity Press.
Douglas, M. (1992). Risk and Blame: Essays in Cultural Theory. New York: Routledge.
Dow, K. & Cutter, S.L. (2000). Public orders and personal opinions: Household strategies
for hurricane risk assessment. Environmental Hazards, 2, 143-155.
Fiske, A. P., Kitayama, S., Markus, H. R., & Nisbett, R. E. (1998). The cultural matrix of
social psychology. In D.T. Gilbert, S.T. Fiske, & G. Lindzey (Eds.), Handbook of
Social Psychology (pp. 915-930). New York: McGraw-Hill.
Frewer, L. (2004). The public and effective risk communication. Toxicology Letters,
149, 391-397.
Iarossi, G. (2006). The Power of Survey Design. Washington, D.C.: The World Bank.
Inelmen, K., Say, A.I., & Kabasakal, H. (2004). Participation lethargy in disaster
preparedness organization within the framework of a Turkish CBO. International
Journal of Sociology and Social Policy, 24(10), 130-158.
Jasanoff, S. (1998). The political science of risk perception. Reliability Engineering
and System Safety, 59, 91-99.
Kasperson, R.E. (1992). The social amplification of risk: Progress in developing an
integrative framework. In S. Krimsky & D. Golding (Eds.), Social Theories of
Risk. Westport: Praeger Publishers.
Kitayama, S. & Markus, H.R. (1995). Culture and the self: Implications for
internationalizing psychology. In N.R. Goldberger & J.B. Veroff (Eds.), The
culture and psychology reader (pp. 366-385). New York: New York University
Press.
Lang, J.T. & Hallman, W.K. (2005). Who does the public trust? The case of
genetically modified food in the United States. Risk Analysis, 25(5), 1241-1252.
Markus, H.R., Mullally, P.R., & Kitayama, S. (1997). Selfways: Diversity in modes of cultural
participation. In U. Neisser & D. Jopling (Eds.), The conceptual self in context: Culture,
experience, self-understanding (pp. 13-26). New York: Cambridge University Press.
Markus, H.R., Steele, C.M., & Steele, D.M. (2000). Colorblindness as a barrier to
inclusion: Assimilation and nonimmigrant minorities. Daedalus, 129(4), 233-257.
Masuda, J. R. & Garvin, T. (2006). Place, Culture, and the Social Amplification of Risk.
Risk Analysis, 26(2), 437-454.
McGuirk, P.M., & O'Neill, P. (2005). Using questionnaires in qualitative human
geography. In I. Hay (Ed.), Qualitative Research Methods in Human Geography
(2nd ed., pp. 147-162). Oxford: Oxford University Press.
Mileti, D. S. (1995). Factors related to flood warning response. US-Italy Research
Workshop on the Hydrometeorology, Impacts, and Management of Extreme
Floods. Perugia, Italy.
McGee, T. K. & Russell, S. (2003). "It's just a natural way of life..." an investigation of
wildfire preparedness in rural Australia. Environmental Hazards, 5, 1-12.
Paton, D. & Johnston, D. (2001). Disasters and communities: vulnerability, resilience and
preparedness. Disaster Prevention and Management, 10(4), 270-277.
Tobin, G.A., & Montz, B.E. (1997). Natural Hazards: Explanation and Integration. New
York: The Guilford Press.
Sáenz Segreda, L. (2003). Psychological interventions in disaster situations. In J. Zschau
& A. Küppers (Eds.), Early Warning Systems for Natural Disaster Reduction (pp.
119-124). Berlin: Springer-Verlag.
Savadori, L., Savio, S., Nicotra, E., Rumiati, R., Finucane, M., & Slovic, P. (2004).
Expert and public perception of risk from biotechnology. Risk Analysis, 24(5),
1289-1299.
Scherer, C. W. & Cho, H. (2003). A social network contagion theory of risk perception.
Risk Analysis, 23(2), 261-267.
Shweder, R. (1995). Cultural psychology: What is it? In N.R. Goldberger and J.B.
Veroff (Eds.), The Culture and Psychology Reader (pp. 41-86). New York: New
York University Press.
Siegrist, M. & Cvetkovich, G. (2000). Perception of hazards: The role of social trust
and knowledge. Risk Analysis, 20(5), 713-719.
Slovic, P. (1999). Trust, emotion, sex, politics, and science: Surveying the riskassessment battlefield. Risk Analysis, 19(4), 689-701.
Slovic, P. (2000). The Perception of Risk. London: Earthscan Publications, Ltd.
Wisner, B., Blaikie, P., Cannon, T., & Davis, I. (2004). At Risk: Natural Hazards,
People’s Vulnerability and Disasters. New York: Routledge.