Risk Perception and Communication in the Porto Region (PR), Portugal

Modern cities developed in the wake of the upheaval of the industrial revolution and the economic, social, technological and political changes that took place from the mid-eighteenth century until the middle of the following century [1]. Currently, cities are seen as centres par excellence for incubating progressive ideas, designs and new formulations for products and services. They are also connecting points for global networks, where the economy can find resources and partners that link them to new markets. Associated with the growth of the urban population over recent decades, a series of sprawling urban households and conurbations have appeared. Sometimes much larger and with a larger population than the original cities, they are recognised in the term "metropolitan area", which very often exceeds their administrative limits.

Cities are generally regarded as areas of greater risk, basically because they have the potential for suffering greater damage. This includes potential victims (including deaths and injuries, with different levels of severity), displaced people and people suffering psychological harm. Economic, material, environmental, and functional damage are likely to be more severe since they are estimated in monetary sums or strategic values, according to the various elements exposed to risk. Such sums and values will correspond to the market cost of their recovery/restoration which must take into account not only the nature and features of the damaged element, but other factors, too, that could influence this cost, such as other direct and indirect economic losses that result from the cessation of operability, activity or work [2].

Interest in the study of risks has mainly grown since the second half of the 1950s, partly because of the increase in disasters but also thanks to higher economic and social costs generated not only by human and economic losses, but also by the recovery and reconstruction of the affected areas [3]. At the same time, the growing importance of analysing the behaviour of communities in relation to risk is acknowledged. In this regard, in addition to the elements present in the risk areas (people and their assets and possessions) that are subject to possible loss, it is also important to consider the level and extent of the damage that these exposed elements could suffer. Such damage would be associated with their intrinsic characteristics and their degree of protection, as well as the ability to both anticipate and respond to risk [4]. The idea of population involvement in risk management is consequently reinforced [5], going beyond a view that sees crises as merely technical issues and ignoring the importance of community involvement, in particular how the public perceives them [6]. Perception thus plays a major part in motivating individuals to take action to avoid, mitigate, adapt to, or even ignore risks [7]. According to Wachinger et al. [7] risk perception is the process of collecting, selecting, and interpreting signals about the uncertain impacts of events and it involves multiple influencing factors in a very complex framework [8]. But risk perception depends on an individual’s subjective judgement and evaluation of a specific risk [9] which can be perceived as potentially dangerous by one person, whilst it may be considered safe by someone else. Risk management, then, is the modulated mental models and the psychological mechanisms that people use when they judge, evaluate, tolerate, and react to risks [10]. Furthermore, it is how individuals and communities perceive the complex and varied factors which interfere in risk perception. These certainly include social media and capital, media influence, personal experience, values, worldviews, and the influence of individual adaptation strategy through learning from past events and education [11-13]. We can thus say that risk communication is a social process by which people become informed about hazards, are influenced towards behavioural change and can participate in decision making about risk issues in an informed manner [14, 15]. According to Covello [16], risk communication is a two-way process between the communicator(s) and the recipients of the messages in a process of exchanging information among interested parties about the nature, magnitude, significance, or control of a risk. Other definitions emphasize the importance of risk management [17], the need for dialogue between communicators and stakeholders [18], and the necessity of ongoing risk monitoring [19]. Other authors have emphasized the emergency management efforts [20] and the crisis and emergency risk communication [21], where supports pre-crisis communication is classified as a method to increase the effectiveness of the response stages and reduce harm in the resolution stages [22]. In this context, pre-crisis communication is directed to the public and response community to provide risk messages, warning, and guidance regarding preparation. Strategies include building alliances, developing consensus recommendations, and testing messages with specific publics [23]. Effective risk communication requires understanding where the public is coming from in order to convince them to prepare better for risks. Mental models provide a framework to understand pre-existing public perceptions of less-familiar and higher-dread risks and what communication messages, and the way of disseminations, can be developed and tested to improve awareness, understanding, and preparedness [24, 25].

This work sets out to identify the main risks, perceived as most important because they are present in the Porto region (PR), and to identify the relationships between the factors that generally influence this perception, especially the socio-economic and demographic characteristics of the population (sex, age, education, income). It is also intended to understand how people living in the PR perceive the support of public entities in crisis situations, which bodies are responsible for pre-crisis communication, and the most effective way to disseminate risk messages in crisis situation. It is assumed that knowledge and understanding of these factors will help to inform decision makers better with respect to the type of variables they should consider when conceptualising risk communication strategies, to make them more efficient by including them in plans and preventive measures to combat and mitigate natural risks.

3.1 Questionnaire design

The questionnaire entitled Perception of risks and communication in the Porto Region (PR) was randomly applied to 248 residents in the PR in April and May 2020. Online surveys are an instrumental methodology widely used in the study of risk perception [29]. From the total, 244 questionnaires were validated. According to the Survey system sample size calculator (https://www.surveysystem.com/sscalc.htm), for a resident population of about 700 000 inhabitants, the confidence interval obtained was 6.27 (a confidence level higher than 90%). Several authors [30] also consider the need for quantitative structuring of the results observed on the ground, to define and systematise response structures and patterns. Keeping the proposed objectives in mind, the questionnaire is disimproved into five parts (Figure 2).

2.png

Figure 2. Components of the perception of risks and communication in the PR questionnaire

The PR covers several Seveso industries, mainly in the municipality of Matosinhos, along with several industries that handle or store materials of this nature. Finally, the risk of urban and industrial fire is also very significant in all of the municipalities under analysis, mainly because of the high population density in these locations. In the first, “Subjects’ Characterisation Data”, information about the sex, age, academic education, and income" of the respondents is considered. In the second part respondent’s population was asked to evaluate the risk, for 13 different typologies, considering the harmful consequences probability, or expected losses (deaths, injuries, property, livelihoods, economic activity disrupted or environment damaged) resulting from interactions between natural or human induced hazards and vulnerable conditions”.

They were classified using a qualitative scale, ranging from zero/minimum to maximum on a scale of 1 to 5 where: 1 - zero/minimum; 2 - low; 3 - moderate; 4 - high; 5 – maximum. The third part, called “Risk communication”, was intended to identify which bodies are responsible for risk communication. Four bodies were considered (central government, municipal councils, Portuguese Environment Agency, and civil defence authority). In keeping with the second part of the questionnaire, these had to be classified using a qualitative scale, ranging from “not responsible” to “very responsible” on a scale of 1 to 5 where: 1 - not responsible; 2 - not very responsible; 3 - responsible; 4 - quite responsible; 5 - very responsible.

The fourth part entitled “Forms of warning in crisis situation” was intended to analyse what respondents considered to be the most effective way of warning people in crisis situations. They were classified using a qualitative scale, ranging from "not at all effective" to "very effective" on a scale of 1 to 5 where: 1 - not at all effective; 2 - not very effective; 3 - effective; 4 - quite effective; 5 - very effective, the forms being short message service (SMS), megaphone around the streets, radios, and sirens.

Finally, the fifth part, entitled “Support from government entities in the event of crisis”, considered the support from municipal councils, central government, civil defence authority, and insurance companies, which the respondents had to classify using a qualitative scale, ranging from “no support at all” to “a great deal of support” on a scale of 1 to 5 in which: 1 - no support at al.; 2 - not much support; 3 - supported; 4 - quite a lot of support; 5 - a great deal of support.

The questionnaires were applied online with the collaboration of some Universities, City Councils and Schools, through the respective social networks (Facebook and WhatsApp), since they were applied during the month of March 2020, in a pandemic situation and with a declared state of emergency in Portugal. Due to these constraints, it is expected that certain population groups were not covered in a representative way.

Cronbach's α test was applied to measure the internal reliability (consistency) of the multiple Likert scale for the different items used in the survey [31]. Internal consistency describes the extent to which all the items in a test measure the same concept or construct, and so it is connected to the inter‐relatedness of the items within the test [32]. Cronbach's α reliability coefficient ranges from 0 to 1. The closer Cronbach's α coefficient is to 1.0, the greater the internal consistency of the items on the scale. The α coefficient obtained for the 21 analysed items is 0.886, suggesting that the data have a high internal consistency.

3.2 Statistical analysis

Next, to detect homogenous groups in the collected data, the cluster analysis approach was used, in other words, a multivariate approach. Indeed, hierarchical cluster analysis is one of the methods used to measure the hierarchy of the proximity between objects, which in this case are risks. The average linkage technique was used to achieve this. Some of the characteristics of this method are lower sensitivity to noise than shown by nearest-neighbour chain methods, and complete-linkage clusterings, presentation of good results for Euclidean distances and other distances, and a tendency to form groups with a number of similar elements. The dendrogram is thus the diagrammatic representation of the proximity matrix among samples, given that it organises the risks according to their similarity or dissimilarity, as perceived by the respondents.

Various statistical methods were used to estimate the impact of particular factors on risk perception. In most cases they included correlation analysis [33], independent samples t-tests [34], one-way analysis of variance ANOVA [35]. Quantitative statistics were primarily used to describe and summarise the features of the socio-demographic characteristics of the respondents (age, gender, education, and income). Analysis of variance (ANOVA) was then performed to examine the mean ranks of two or more independent variables with the null hypothesis of equality. Analysis of variance (ANOVA) was then used to identify whether or not there were any statistically significant differences between the impact factors on risk perception.

Anthropogenic and mixed risks are the most perceived risks for the PR, especially the risk of pollution, industrial accidents, and transport accidents in general. The risk of heat and cold waves is the most perceived of the natural risks. The fact that it is an urban area may justify this perception, given that respondents have greater contact with the manifestation of this type of risks. In fact, the rapid expansion of cities to accommodate the population growth, combined with inappropriate land use planning, increased the exposure of people and economic assets to hazards. Disaster risks in urban areas resulting in a process of “risk accumulation” where risk is amplified by human activities. Nevertheless, urban areas can also provide opportunities for reducing risks. Cities are usually the economic drivers within their countries and the centres of intellectual, political, business and financial activities. Urban centers typically have people with comparatively higher levels of education in better informed communities and more powerful local response capacity.

In this sense, a number of authors consider socio-demographic characteristics as important factors in perception in studies on risk perception, including age, gender, educational level, and income [9, 36]. Some studies suggest, especially in urban areas, that the most disadvantaged population, both from an economic and housing point of view, tends to increase the perception of risk and take fewer preventive actions [37]. In our study, regarding the level of education, respondents with higher education show a better perception of risk than respondents with a lower educational level.

Some authors suggest that people who have been educated to a higher level tend to acquire and understand new information more easily, which can somehow influence knowledge about potentially dangerous processes and thus increase sensitivity to issues related to safety and risk perception [13, 38]. They also tend to be more aware of the mitigation actions taken by governments and local experts, and thus enhance the internalisation of their actions, increasing the degree of control over the different risks [39]. Although significant statistical correlations between the level of education and the perception of risk were found in this work, other studies did not find any correlation, not even a negative one, between the perception of risk and the level of education [40]. This somehow confirms the idea that there is a significant complexity associated with perception, and a need to develop more work in different socio-economic and cultural contexts, with the inclusion of more analysis variables. On the other hand, emotions play a very important role in how the population perceives risks, and naturally how they evaluate their safety. Such situation is clearly observable in risk perception studies in urban areas where technological risks are perceived as high conditioned by fear or alarm [41]. The conceptualization of the perception and acceptance of risk are a complex process influenced not only by the objective nature of a risky technology but also by various underlying social and psychological factors of individuals in interaction with communication processes [42]. The presence of industrial complexes considered potentially dangerous by the media and local public opinion, not only as pollutant emitters, but also associated with the risk of explosion in the study area, such as the Matosinhos refinery, for example, may influence the perception of risks, advocating industrial risks as higher than the others.

Gender has been suggested as a factor strongly related to perception and attitudes towards risk, with several studies suggesting that, on average, men have lower levels of perception than women, especially in works that focused on the risk of flooding [43]. However, in relation to this factor, other studies report opposite results [44], which may be related to the different roles of women in society, which vary with the socio-cultural context, and which may influence perception. In certain societies, women are perceived as being physically more vulnerable [45], which may justify responses that are more targeted to a greater concern for health and the family [46]. Regarding technological risks, several studies point out that women perceive more risk than men, especially younger and more educated ones [47].

Our study found few statistically significant correlations between gender and the perception of risk, even though they suggest a greater perception by women. These results may be related to the urban context to which the sample belongs, which is characterised by a socio-cultural context that tends to mitigate gender differences as opposed to what happens in more traditional socio-cultural contexts [48]. The results of this work also suggest that people (regardless of age, gender and education) feel more supported by the municipal councils and civil defence authorities in a crisis situation than by the central government, and even less by insurance companies. These results are in agreement of several works showing that population have greater trust in regional/municipal institutions compared to national bodies [13].

In fact, the need to establish trust with the public is fundamental to the effectiveness of pre-crisis communication messages and strategies [49]. Moreover, it may possibly be that more effective risk communication, based on an appropriate socio-cultural context could alleviate this distrust. In fact, these are still the institutions, together with the APA, to which the respondents ascribe greater responsibility in pre-crisis communicating, reinforcing the idea of a devaluation in relation to the role of the central government in risk management. This conclusion is even more clearly discernible in the younger age group. If we also consider that this is the age group with the lowest perception of risk, the results can be seen to justify the importance of the role of school and education as a risk mitigation factor, which reinforces the relevance, for example, of including this topic in the teaching curriculums [13, 39].

Warning the local community via SMS, in a crisis situation, is the option preferred by the surveyed sample. However, the results suggest that the majority of the older population would rather be warned by radio station announcements. The effectiveness of risk communication must thus consider, in addition to the socio-cultural context, the specific characteristics of the population, particularly certain factors such as age, gender and education These results are in agreement with the findings of Ellen et al. [50], in which plurality in communication, which means communicating using the language of the different perspectives, is therefore very important. A range of communication tools that use different approaches, SMS, Facebook, twitter, as well as traditional media (radio, TV) will help to reach the different stakeholders and to integrate them into the decision-making process, thereby increasing support for decisions taken. On the other hand, the risk communication over time leads to improved risk perception. In this way, provision of risk information can significantly benefit the change of risk perceptions, with subsequent changes of the impacts of risk perceptions on risk behaviours [50]. Any population exposed to hazards natural ones such as earthquakes, storms, wildfires or floods; or technological ones, such as explosions, chemical spills, train crashes and so on wants and needs to be optimally informed about risk characteristics, preventative measures, and appropriate behaviours during emergencies. Authorities have to draft sensible planning, prepare coping strategies and communicate the relevant information effectively to residents, people in their workplaces, and communities as a whole. The more disaster management requires active the involvement of residents, the more vital risk information/ communication/education become. Furthermore, in the case of controversial risk sources (e.g., the positioning of an airport or a waste incineration facility), public discussion, participation of stakeholders and possibly joint conflict resolution are required. All these situations involve social processes which are usually subsumed under the (umbrella) term "risk communication", and the exchange of risk information between interested parties (individuals, groups, institutions) is at its very core [15, 25]. Evidently risk communication is the indispensable link between risk perception and risk management. Given the high relevance of effective disaster preparedness, risk communication programmes need to be based on a sound understanding of the underlying sociopsychological processes and preconditions for successful communication.