The presented survey was conducted in six European countries as an online study. A total of 2454 subjects participated. Two main research questions were investigated: firstly, how does the cognitive, moral, and affective framing of radio frequency electromagnetic field (RF EMF) exposure perception influence RF EMF risk perception? Secondly, can the deployment of mobile phone base stations have greater acceptance with RF EMF exposure reduction? The findings with respect to the first question clearly indicated that the cognitive framed exposure perception is the main determinant of RF EMF risk perception. The concomitant sensitivity to exposure strength offers an opportunity to improve the acceptance of base stations by exposure reduction. A linear regression analysis supported this assumption: in a fictional test situation, exposure reduction improved the acceptance of base stations, operationalized as the requested distance of the base station from one’s own home. Furthermore, subjects with high RF EMF risk perception were most sensitive to exposure reduction. On average, a 70% exposure reduction reduced the requested distance from about 2000 meters to 1000 meters. The consequences for risk communication are discussed.
The fast-growing use of wireless communication technologies has stimulated concerns about the rise of public exposure to radio frequency electromagnetic fields (RF EMF) and fostered the prevailing societal discussion about potential human health risks. It is unsurprising that the public risk perception of RF EMF is at a constantly high level across Europe, especially for mobile phone base stations [
Several approaches in dealing with this resistance against the deployment of mobile phone base stations are proposed, for example, early public participation, improved risk communication, and measurement campaigns, which became an important topic for policy makers [
The low EMF exposure future networks (LExNet) project follows a different approach. It assumes that more acceptance of mobile phone base station deployments from the public could be achieved by reducing the exposure to RF EMF. LExNet expects that this reduction will result in a lower risk perception and therefore higher acceptance of EMF technologies [
In a previous study (see [
It seems reasonable to assume that people select one of these frames when they assess the risk of an exposure scenario. In other words, the cognitive frame may not necessarily be the dominant frame. People could focus on moral or affective aspects of exposure. This view is in line with the concept of intuitive toxicology developed by Slovic et al. [
The study presented in the following examines fours questions in detail: What do people believe about the level of RF EMF they are exposed to? How is RF EMF risk perception influenced by cognitive, moral, and affective frames of RF EMF exposure perception? Can the acceptance of base stations in one’s own neighborhood be improved by RF EMF exposure reduction? Does RF EMF risk perception mitigate the effects of exposure reduction on the acceptance of base stations?
While the first question is self-explanatory, the remaining three questions require elaboration. Whether exposure reduction results in more acceptance of telecommunication networks, especially base station deployments, depends in our view on the frame in which an RF EMF exposure situation is viewed. People’s acceptance of base stations will be only a matter of RF EMF exposure reduction if a cognitive frame is applied. If RF EMF exposure is perceived solely within a moral or affective frame then information about any reduction of exposure will have little or no impact on the acceptance of base stations.
Furthermore, the dominant exposure perception frame might vary for different RF EMF exposure sources. Therefore, the impact of exposure frames has to be investigated across various sources of exposure.
However, even when a cognitive frame is applied in exposure perception, that is, when risk perception is based on people’s belief about exposure levels, it is still open how much RF EMF exposure reduction is required in order to accept the deployment of a source of exposure. Therefore, we are interested in the effects of different exposure reductions. Furthermore, the acceptance of a base station deployment can be operationalized by the required distance from the base station to one’s own home: the further the distance the higher the acceptance. This approach seems to be highly promising because people intuitively understand distance as “measure” for safety [
The survey was conducted in August 2014 in six European countries as an online study by a professional survey company. A total of 2454 interviewees participated. After quality control 1809 respondents remained for analysis (German sample
The questionnaire consisted of 33 questions. All questions were translated into the languages of the participating countries and double-checked with retranslation back into English. An introduction to the survey informed the participants about the main research aims and what participation in the survey involves, including how anonymity of the survey is ensured. Some questions were introduced with additional information, for example, a technical background. In addition, we provided some background information about the LExNet project.
For investigating the exposure perception frames, we used questions guided by pictures, describing various exposure situations in a vivid way. We selected five scenarios: (1) exposure through using a mobile phone for calls, (2) exposure through laptop use, (3) exposure through using a WLAN (WiFi) router in a close position, (4) exposure through having a mobile communication mast (base station) on a roof close to one’s home, and (5) exposure through the use of cell phones by others (the picture displayed a person using a mobile phone in public transport, sitting close to another passenger). The pictures were randomized in order of presentation.
We asked the subjects for how strong they consider the RF EMF exposure situation as well as for their affective and moral evaluation of the situation, all on a 5-point Likert scale. Furthermore, we asked the respondents to assess the level of danger of each of the five exposure situations (see Table
Questions on affective, moral, subjective exposure perception and risk perception of various exposure situations shown in pictures.
Question | Answer option |
---|---|
Affective evaluation: |
5-point Likert scale from 1 = “very positive” to 5 = “very negative” |
|
|
Moral evaluation: |
5-point Likert scale from 1 = “not at all” to 5 = “yes absolutely” |
|
|
Subjective exposure perception: |
5-point Likert scale from 1 = “low” to 5 = “high” |
|
|
Risk perception: |
5-point Likert scale from 1 = “not dangerous” to 5 = “very dangerous” |
Following a study from Wiedemann and Claus [
Some of our demographic, political, and belief-related questions were derived from the “European Social Survey” [
The mean age of the participants was about 40 years, with 49.1% male and 50.9% female. The mean of respondents’ education years of 15.2 indicates a sample of well-educated people, with a mode value (
The general perceived level of RF EMF exposure is indicated by Figure
Perceived level of exposure (on a 5-point Likert scale from 1 “not at all” to 5 “to a very high degree”; question: “thinking about your daily life, to which degree do you think you are exposed to electromagnetic fields from electronic devices (like mobile phones, WiFi router) and base stations?”).
These results show that many respondents believe that they are exposed to RF EMF to a high degree.
The question is, do people—when they assess the riskiness of an exposure situation—take exposure levels into account or are their risk perceptions rather influenced by moral or affective evaluation of RF EMF exposure situation? This question was researched for five exposure situations: using a mobile phone (MP) for calls, exposure through the use of cell phones by others, laptop use, using WLAN router in a close position, and exposure from a base station; see Table
Means and variance of affective and moral evaluation, subjective exposure perception, and risk perception of various exposure situations, on 5-point Likert scale from 1 = “very positive” to 5 = “very negative” for affective evaluation; from 1 = “not at all” to 5 = “yes absolutely” for moral evaluation; from 1 = “low” to 5 = “high” for exposure evaluation; and from 1 = “not dangerous” to 5 = “very dangerous” for risk perception.
Evaluation of various sources of EMF exposure |
|
Mean | Variance |
---|---|---|---|
Mobile phone (MP) calls | |||
Affective evaluation | 1536 | 3.03 | .862 |
Moral evaluation | 1648 | 2.81 | 1.648 |
Subjective exposure perception | 1643 | 3.34 | 1.472 |
Risk perception | 1654 | 3.01 | 1.268 |
WLAN close position | |||
Affective evaluation | 1546 | 2.90 | .836 |
Moral evaluation | 1630 | 2.67 | 1.411 |
Subjective exposure perception | 1639 | 2.90 | 1.359 |
Risk perception | 1627 | 2.76 | 1.296 |
MP use by others | |||
Affective evaluation | 1547 | 3.05 | .893 |
Moral evaluation | 1659 | 2.59 | 1.383 |
Subjective exposure perception | 1632 | 2.55 | 1.350 |
Risk perception | 1640 | 2.44 | 1.231 |
Laptop use on the lap | |||
Affective evaluation | 1572 | 2.91 | 1.025 |
Moral evaluation | 1655 | 2.69 | 1.535 |
Subjective exposure perception | 1637 | 2.91 | 1.482 |
Risk perception | 1642 | 2.81 | 1.448 |
Base stations | |||
Affective evaluation | 1629 | 3.59 | 1.423 |
Moral evaluation | 1672 | 3.64 | 1.593 |
Subjective exposure perception | 1657 | 3.86 | 1.389 |
Risk perception | 1667 | 3.76 | 1.393 |
As demonstrated by Table
To analyze whether people’s risk perceptions of various sources of EMF exposure are based on affective and moral frames or on a cognitive (taking exposure into account) frame, linear regressions were computed for all five exposure situations using risk perception as the dependent variable and the affective and moral evaluation as well as the subjective exposure perception as independent variables. Table
Linear regression of affective, moral, and exposure evaluation on concerns about various sources of EMF exposure (risk perception), beta values indicated,
Dependent variable risk perception of |
|
| ||
---|---|---|---|---|
Affective | Moral | Exposure | ||
Mobile phone (MP) calls | .092 |
.302 |
.584 |
.672 |
WLAN close position | .051 |
.292 |
.629 |
.756 |
MP use by others | .004 | .222 |
.718 |
.790 |
Laptop use on the lap | .072 |
.269 |
.670 |
.822 |
Base station | .061 |
.208 |
.711 |
.811 |
Furthermore, as indicated by the beta values in the regressions, the influence of the affective frame on RF EMF risk perceptions is more or less negligible while the moral frame plays a role (
Regarding the acceptance of RF EMF technologies we focused on the question of whether reductions of RF EMF exposure influence the acceptance of base station in one’s own neighborhood. Specifically, we asked the respondents for the minimal distance (in meters) in which they would accept a base station close to their home for four different exposure conditions: (1) current exposure level without any reduction, (2) exposure level reduced by 30%, (3) exposure level reduced by 50%, and (4) exposure level reduced by 70%. For the analysis we excluded subjects answering a distance higher than 10 000 meters, that is, people who are in fundamental opposition to base stations (
Firstly, we have a look at the distribution of required distances for the exposure scenario (1), that is, the current exposure level. More than 25% (
Accepted distances of base stations close to one’s home in meter with 0% exposure reduction, indicated by numbers of respondents. For respondents with distance <10 000 meters (
In addition, we asked the subjects to consider the effects of fictional exposure reductions on their willingness to accept a base station deployment in their own neighborhood. The comparison between the four fictional exposure situations indicates a consistent picture: The higher the exposure reduction, the lower the distance in which a base station in the vicinity of one’s home is accepted. While the median of the distance for the baseline exposure situation (0 = % reduction) is at 1000 meters, the median of the distance decreases to 700 meters for 30% exposure reduction, decreases to 500 meters for 50% exposure reduction, and finally remains at 500 meters for the highest exposure reduction of 70% (see Figure
Median of distance in meters in which a base station close to one’s home is accepted for 0%, 30%, 50%, and 70% exposure reduction. For respondents with distance <10 000 meters (
It seems reasonable to assume that RF EMF risk perception will influence the required distances across to the four exposure reduction scenarios. In order to test this hypothesis five risk perception groups are distinguished based on the scores for the perceived risk of base stations. The scores refer to one of above-mentioned pictured-guided scenarios that focused on the exposure by having a mobile communication mast (base stations) on a roof close to one’s home.
The frequency distribution of these risk perception scores is depicted by Figure
Frequency distribution of the scores for risk perception of base stations on a 5-point Likert scale (from 1 = “not dangerous” to 5 = very dangerous).
The different colored lines presented in Figure
Effects of RF EMF exposure reduction on the acceptance of base stations in dependency of risk perception groups. Acceptance measured by the question “roughly at what distance (meters) would you accept a base station close to your home?,” “
A general linear model with repeated measures was calculated using the different risk perceptions of base stations as “between subject factor” and the four exposure reduction scenarios as “within subject factor.” The required distance to one’s own home was used as the dependent variable. The results show a significant main effect for the repeated factor exposure scenario (
Using the results from Figure
Table
Bonferroni post hoc test for in between group differences among different risk perception groups and accepted distances to a base station for various exposure reduction scenarios. Question exposure reduction: “roughly at what distance (meters) would you accept a base station close to your home?,” “…if the exposure was reduced by 30%?,” “…if the exposure was reduced by 50%?,” and “…if the exposure was reduced by 70%?” Risk perception measured by the question “how dangerous do you consider living close to the building with the antennas?” on 5-point Likert scale from “1 = not dangerous” to “5 = very dangerous.”
Exposure reduction | RP groups | Mean distance (meters) | RP groups | ||||
---|---|---|---|---|---|---|---|
(1) low | (2) | (3) | (4) | (5) high | |||
0% | (1) low | 536,55 | .869 | .103 | .000 |
.000 | |
(2) | 1130,37 | .869 | 1 | .000 |
.000 | ||
(3) | 1349,87 | .103 | 1 | .000 |
.000 | ||
(4) | 2161,26 | .000 |
.000 |
.000 |
.140 | ||
(5) high | 2569,40 | .000 |
.000 |
.000 |
.140 | ||
|
|||||||
30% | (1) | 465,94 | 1 | .157 | .000 |
.000 | |
(2) | 929,71 | 1 | 1 | .001 |
.000 | ||
(3) | 1122,58 | .157 | 1 | .003 |
.000 | ||
(4) | 1702,04 | .000 |
.001 |
.003 |
.049 | ||
(5) | 2102,26 | .000 |
.000 |
.000 |
.049 |
||
|
|||||||
50% | (1) | 369,45 | 1 | .140 | .000 |
.000 | |
(2) | 789,11 | 1 | 1 | .010 |
.000 | ||
(3) | 963,82 | .140 | 1 | .038 |
.000 | ||
(4) | 1378,79 | .000 |
.010 |
.038 |
.022 | ||
(5) | 1767,50 | .000 |
.000 |
.000 |
.022 |
||
|
|||||||
70% | (1) | 329,30 | 1 | .250 | .003 |
.000 | |
(2) | 685,19 | 1 | 1 | .108 | .000 | ||
(3) | 814,74 | .250 | 1 | .295 | .000 | ||
(4) | 1093,99 | .003 |
.108 | .295 | .017 | ||
(5) | 1449,78 | .000 |
.000 |
.000 |
.017 |
The results indicate constant significant differences in the between group comparison among respondents with lower ((1), (2)) and higher risk perception ((4), (5)) (range from
Due to the fact that one of the requirements of the general linear model with repeated measurements is not fulfilled (Levene’s test of equality of error variances) nonparametric test was calculated. The Friedman analysis of variances of ranks and the pairwise comparisons indicate significant differences between the four exposure reduction scenarios (
Finally, for exploratory reasons, we conducted several regression analyses with gender, age, and education as predictor variables and risk perception of various sources of EMF exposure as dependent variables (see Table
Linear regression of age, gender, and education on concerns about various sources of EMF exposure (risk perception), beta values indicated,
Dependent variable risk perception of |
|
| ||
---|---|---|---|---|
Gender | Age | Education | ||
Mobile phone (MP) calls | .077 |
−.030 | .029 | .008 |
WLAN close position | .105 |
−.001 | .038 | .012 |
MP use by others | .043 | −.029 | .000 | .003 |
Laptop use on the lap | .085 |
−.071 |
.025 | .015 |
Base station | .047 | −.018 | .016 |
.007 |
The findings indicate various significant findings, especially for gender (mobile phone calls:
The chosen approach, based on a community sample, allows for testing of relationships between exposure perception, risk perception, and the effects of exposure reduction on acceptance of base station deployments. However, there are some limitations. The present study is based on a cross-sectional research design. This design has restrictions concerning the interpretation of statistical associations. Any causal interpretation of the established statistical associations needs further support from a randomized controlled study. Furthermore, one must be cautious in extrapolating these results to the general population. Firstly, the present study is based on an online survey that limits the scope for the generalization of the findings, as people without Internet access are not taken into account. Secondly, the country samples are not drawn randomly from the populations. Therefore, an extrapolation from our sample to the general population is restricted although the chosen sample is community based and represents a diversity of educational backgrounds. For the same reasons, conducting a cross-cultural analysis would not be practical.
The presented research provides new insights into how people evaluate the risk of various RF EMF exposure situations. Firstly, our respondents believe that all considered exposure situations expose people to at least a medium level of RF EMF. The highest exposure is attributed to base stations and the lowest to the exposure caused by other persons’ cell phone use. Our data show also that base stations tend to be associated with negative feelings. All other exposure situations do not elicit negative feelings. For moral concerns, the findings point at a more complex picture. All RF EMF exposure situations are associated with moral concerns, at least to a certain degree. Only base stations stimulate reasonably strong moral concerns. It must be noted that the type of moral concern was not specified in our questions. Therefore, we can only speculate about the moral dimension that the respondents refer to. According to Haidt’s Moral Foundation Theory [
The determinants of intuitive EMF risk perception were analyzed by linear regressions across the four exposure situations. A consistent picture emerged: EMF risk perception is mainly affected by exposure perception and also to a certain degree by moral concerns. The affective evaluation of the exposure situations plays only a minor role. Both the high proportion of explained variances and high beta values for exposure perception in our regressions models clearly support this interpretation. In terms of risk communication, these findings provide a positive message. They suggest that, in principle, people should be sensitive to exposure reductions, simply because the level of exposure is an important factor when they evaluate EMF risks. Otherwise, any information about exposure reduction would have little or no impact.
This leads to the conclusion that a significant first step of risk communication is the framing of the EMF controversy. Communication should underline the importance of exposure issues and reduce the impact of moral frames. But even when the public acknowledges that exposure is the crucial point it does not necessarily mean that any reduction of RF EMF exposure—regardless of its amount—leads to reduced risk perceptions and to more acceptance of telecommunication technologies. The question “how safe is safe enough?” posed by Fischhoff et al. [
For a communication strategy that builds upon exposure reduction, however, several points have to be taken into account that can only be briefly outlined here. Firstly, our previous study [
What seems to be lacking is the comparative view, that is, comparing the impact of these exposure characteristics for making trade-offs. Secondly, further research is needed in order to analyze the weight that lay people assign to the various RF EMF exposure conditions. Thirdly, built on this insight, one could explore which communication strategies are appropriate in order to strengthen informed judgments about the impact of various exposure sources. The dissemination of knowledge that explains that the assessment of the impact of the exposure of a base station requires trade-offs is especially important, for example, between exposure strength of the base station and the distance of the base station to the exposed people.
The authors declare that there is no conflict of interests regarding the publication of this paper.
This paper reports work undertaken in the context of the project LExNet. LExNet is a project supported by the European Commission in the 7th Framework Programme (GA no. 318273). For further information, please visit