User Tools

Site Tools


research:rural_areas:common_results:results2

Basic Statistical Analysis

Summary

In this section, the interviewees' opinion on wind energy is further analysed and tested for potential factors of influence, as described in detail in the Method of Analysis. This assessment uses subsampling, correlation analysis via the chi-square test and correlation coefficient analysis and GIS based analysis for looking at the influence of the NIMBY attitude, landscape perception, information management and economic benefits.

The first step of analysis showed that the NIMBY phenomenon is not of great importance in the analysed region. This strengthened even more the need for the subsequent further analysis: The subsample analysis of conditions that would make wind energy reasonable for opponents did not show one overall condition that would satisfy all opponents but showed a relatively equal distribution of preferences. Nevertheless, it was possible to confirm three of the hypotheses about the influences on social acceptance with statistical significance. It was shown that the social acceptance is higher:

  • when people disagreed that wind turbine destroys the landscape
  • when people are informed about wind energy projects with good timing and good quality,
  • when people agreed that there is a direct economic benefits from wind energy projects,

The hypothesis on the distance factor could however not be verified as no significant correlation between distance and acceptance level could be found.

In this section, there are most of the times two questions combined in one graph in order to visualise their relation. For making the graphs understandable, it had to be deviated slightly from the general project's layout. The dark gray indicates that people have agreed to a question/ checked it as valid. The green to red scale is reserved for indicating the acceptance level, which is visualized by the arrow below the graphs.

The methods' description for each step of analysis is always directly accessible via the links following each paragraph. The references of section are listed within the references section of the methods page to where they are also linked.

Subsampling

1) NIMBY

When trying to explain the lack of acceptance of wind energy, NIMBY is often amongst the first cited arguments (See literature review). Assessing the opinion of the interviewed people under the perspective of NIMBY however did not produce results which would align with the respective expectation: First of all, among our data there were only two respondents who showed strong NIMBY attitude according to the subsampling method adopted for this case. Therefore, it was not possible to draw any interpretation on what are the disadvantages they see in wind energy as well on the question of under which conditions they would favour or accept wind energy facilities within their neighbourhood.

However, taking one step back some analyses were possible. For instance, there are 65 people who have marked the option “3000m or more” as the minimum distance of wind turbines to villages, even though they previously stated a neutral or good opinion about wind energy. Furthermore, after checking who also indicated that they have either tried to stop the project or have spoken against it, only 10 people remain in the selected subset.

Subset Explanation Number of people
Q.7 (1, 2 and 3) people in favor of wind energy 118
Q.7 (1, 2 and 3) and Q.18 (4 and 5) people in favour but opposing local project 65
Q.7 (1, 2 and 3) and Q.11 (5 or 6) people in favour but wanting >3000m distance 11
Q.7 (1, 2 and 3) and Q.11 (5 or 6) and Q.18 (4 and 5) people in favour but opposing local project& wanting >3000m distance 10

(click to go back to respective methods to see the detailed explanation again)

It appears that the general categorisation as NIMBY is not a very illustrative approach for the analysed case study. Moreover, it is consequently hard to make reliable statements about what disadvantages of wind energy they see, as each sample is very small. The graphical representation of the shares thus has to be seen under consideration of the total number for that subset, n=10.

{{:research:rural_areas:common_results:nimby1.1.png?300| Fig. 1: NIMBY Statements on wind energy, n=10

Note: by clicking on the image, the image will open in a new window with a better resolution

These results show that the analysis should be extended beyond NIMBY in order to gain relevant insights on the reasons and factors behind people's opinions and positions. It thereby confirms what has been found in the literature review that indicates the necessity of going further (see paragraph on NIMBY). This shall be attempted in the following section.

(for more methodological details click to go back to respective methods)

2) Conditions under which opponents would support wind energy

Extending the subset to either all supporters or all opponents of wind energy (Q.7), produces sample sizes that allow to look further into the conditions under which wind energy would be reasonable for the interviewees. The rate of agreement for each aspect in question 12 asking under which conditions one would agree to the use of wind energy is higher among the supporters than among the opponents (figure 2 and figure 3). There no aspect displays more than 27%, which is represented by the option “none”, meaning that no condition would persuade them. The supporters on the other hand have agreed to at least 40% of the conditions, with the exception of the options “environmental friendly”(38%), “other” (23%) and “none” (6%). It seems that supporters are generally more open for many different conditions than opponents. Moreover, among the supporters of wind energy, the greatest share has stated that if wind energy is an “income for the municipality” (63%) or an “alternative to nuclear and fossil fuels” (61%) they would further support the development of wind energy. Concerning the opponents however, it appears that there is not one single condition that would be able to persuade all or even a great share of them to support wind energy; instead all conditions show similarly low shares. This confirms the idea that acceptance cannot be increased by simply pressing one 'magic button'.

Fig. 2: Opponents' conditions for wind energy Fig. 3: Supporters' conditions for wind energy

(for more methodological details click to go back to respective methods)

Correlation Analysis

3) Impact on Landscape

The strong variance between opponents' and supporters' views does seemingly not hold true when asking whether the wind energy development destroys the landscape. Most people, including many that have a high general acceptance, do nevertheless perceive a negative impact on the landscape caused by a wind park (see frequency analysis' results for Q.6 and qualitative analysis' results for Q.15.4). However, when looking at the graph that compares in detail the level of acceptance, indicated at the y-axis, between those that do perceive a negative visual impact and those that do not, indicated at the x-axis, a trend becomes visible: people that are rather against wind energy (right side of the scale), perceive the landscape more often as disturbed. The lines in the graph serve to visualise this trend, that is proven by the Chi-Square test that shows a significant correlation between that perception of landscape destruction and a low acceptance of wind energy (figure 4, p=0,004652). This finding is closely related to the factors “subjective value of the landscape” and “visibility” described in the synopsis. Nevertheless, the identified correlation can merely generally support the idea that the impact on the landscape influences the acceptance, but cannot provide an insight in how strong this factor is compared to other aspects.

Fig. 4: Correlation between landscape destruction and social acceptance

(for more methodological details click to go back to respective methods)

4) Distance

Visibility of wind turbines and the resulting perceived degradation of the landscape can be influenced by topographic aspects, vegetation cover and also the distance to settlements. Moreover, the interviewees frequently complained about too many wind turbines close to their neighbourhood (see qualitative analysis' results for Q.15.4). It thus seemed reasonable to have a closer look whether the general link between acceptance and distance could be proven by itself and outside of the narrow definition of NIMBY.

The GIS analysis showed that Niederer-Fläming is the municipality with wind turbines installed the closest to settlements among the three analysed municipalities. For Niederer-Fläming both, the minimum distance between a village's centre point and turbines (938 meters in relation to the village Werbig) as well as the median distance resulting from all distances to turbines (3291 meters) were the shortest. However, the difference in distance to the closest turbines for Dahme/Mark and Uebigau-Wahrenbrück account to only a couple of hundred meters (figure 5).

Fig. 5: Correlation between distance to wind turbines and social acceptance

(for more methodological details click to go back to respective methods)

Concerning this analysis on the scale of municipalities, we cannot see any trend in the relation between the acceptance level and distance. Otherwise, for instance, Niederer-Fläming would display the lowest level of acceptance as it is in any relation the closest to the wind turbines. Dahme/Mark instead shows the lowest level (median 3, neutral).

However, when considering the density of wind turbines (WT) we can see the expected trend. Dahme/Mark has the highest density (0,38 WT/km²) and the lowest acceptance level. Niederer-Fläming is in the middle with 0,34 WT/km² and an acceptance level between neutral (median 3) and rather in favour (median 4). Uebigau-Wahrenbrück with 0,28 WT/km² has the lowest density and the highest acceptance level. In the (figure 6) each black dot represents one wind turbine and each column is represented by the wind turbines in each village.

Fig. 6: Correlation between density of wind turbines and social acceptance

(for more methodological details click to go back to respective methods)

When looking at the scale of villages instead of municipalities though, the aforementioned relation cannot be detected. The figure 7 does not show the expected trend, which would be: short distance and high density (upper left of the scatterplot) resulting in a low acceptance level (red) and long distance and low density (bottom right of the scatterplot) resulting in a high acceptance level (green). Instead no correlation can be distinguished.

Fig. 7: Correlation between distance/density of wind turbines and social acceptance

(for more methodological details click to go back to respective methods)

Furthermore, still at the scale of villages, the perceived distance also does not seem to influence the acceptance level. In figure 8 the blue buffer zones around the villages represent the median perceived distance in each village (captured from the question Q.17, how far the interviewees think that the closest wind turbines are installed) and the blue dots indicate the existing turbines (data from the Landwirtschafts- und Umweltinformationssystem des Landes Brandenburg (LUIS-BB) concerning immission control (Landesamt für Umwelt Brandenburg, 2016). Thus, if one assumed a correlation with the acceptance level, one would expect to see low levels of acceptance when the buffer circle is small (people think that the turbines are very close) and high acceptance when the buffer circles are large (people think that the turbines are further away). However, for example the habitants of Rosenthal and Waltersdorf think that the turbines are relatively close and still have on average a high acceptance, while in Ihlow they rather oppose wind energy even though they think the next turbine is rather far away. Moreover, people actually most of the time estimate the distance correctly, as there are neither many turbines within the buffer circle nor much further away. An exception to this are the two villages Hermsdorf and Gebersdorf, whose habitants have stated considerably shorter distances than what they actually are. However, there are planned priority areas close by so that people might already anticipate the installation of wind turbines.

Fig. 8: Perceived and real distance to wind turbines

(for more methodological details click to go back to respective methods)

This lack of correlation, however, should be addressed with care. It is important to mention that when the analysis goes to the scale of villages the results are sometimes impaired. The small sample size for each village give results not as reliable as we would like they to be as well they even limit the analysis in several villages, represented in grey because of n<5. These villages with less than 5 questionnaires could not be considered because of the statistical insignificance. However, the findings are coherent with the results of several previous studies on the link between distance and social acceptance (Petrova 2016, Hübner & Pohl 2015). Thus, the general statement that there is no correlation might be considered reliable. Nevertheless, the villages within the sample were all located relatively close to wind parks, so that the analysis was limited to a rather small scale. Consequently, it cannot be said with certainty, that there would be no correlation at all to distance, if the scale of analysis was extended, as for example what Guo et al. 2015 discovered.1)

Generally, it should moreover be kept in mind, that in this analysis by looking at the villages only the static distances to the wind turbines were measured. This builds on the underlying assumption that people perceive the wind turbines only and mainly from their place of living. However, this view appears greatly limited when referring to the results of question 16, which indicate that most people feel strongly disturbed by wind turbines when they are going for a walk outside of town, not when they are at home (see frequency analysis' results for Q.16). Moreover, even if it was not asked for in the questionnaire, it is very probable that people leave their villages for their daily travel to work. Consequently, people are exposed to and also very likely influenced in their opinion by many more wind turbines than only those in the direct vicinity of their villages. However, assessing the resulting potential impact does exceed the limits of feasibility of this project, not least due to the limited enquiry scope of the provided questionnaire.

5) Information Management

As mentioned before, the questionnaire in Havelland-Fläming also did not detail on participation issues. 2) Therefore, the possibility should be considered that the insignificance of participation possibilities that was indicated in the analysis of the open questions (see qualitative analysis' results) might not reflect the entire reality. Following the theory of consistency effects, respondents are more likely to pick up atopic again if it has already been mentioned before in the questionnaire. Consequently, not asking about participation aspects also reduces the likelyhood of respondents to detail about them later on. Consequently, it cannot be deducted with certainty that people are not concerned about participation issues and would thus behave significantly divergently from literature's results. This consideration is supported by the statistical tests conducted on the link between the information management, which can be seen as an aspect of participation, and the acceptance level, as this does show significant dependencies.

Time of information

For one, the detailed analysis showed that behind that majority of people stating that they have been informed too late (n=110, see frequency analysis' results for Q.9), there is a considerable link to a low level of acceptance of wind energy (figure 9). Within the group of people that felt informed early, the connection to a high level of acceptance seems to be even more marked. However, the comparably small sample size of that subgroup needs to be considered (n=29). Consequently the size of the pies was adjusted to represent their sample sizes. Nevertheless, the relation between acceptance and time of information in the overall sample proofs statistically significant with p=0,00004.

Fig. 9: Time of information and acceptance

Scale of information

Besides this correlation, another one is appearing when looking at the scope of information received. It has already been described that the majority of people felt insufficiently informed (n=126, see frequency analysis' results for Q.10). Moreover, the less the people felt informed, the less they are likely to support the use of wind energy (figure 10). Once again, this relation appears even stronger when looking at the group of people that think they have been informed completely, but in this case the sample size is particularly small (n=5). Overall, the influence of the scale of information on social acceptance proved to be significant (r=0.45 - correlation coefficient).

Fig. 10: Scale of information and acceptance

The above described relations can particularly be regarded as issues of public participation, as the respondents received their information on wind energy projects to a large extent from the municipality (see frequency analysis' results for Q.8), representing in this - once again limited - questionnaire the public decision-making authorities. Moreover, this hints to the great responsibility but also potential of the municipal communication - even though the influence of the reporting by the media cannot be neglected (see media analysis) as it is the 2nd biggest source of information. Generally, these findings on the information managment do clearly support the synopsis' findings on the importance of the timing and the quality of information on wind energy projects for their public acceptance.

(for more methodological details click to go back to respective methods)

6) Direct Benefits

Frequently cited in literature, considerably often chosen in the questionnaire (see results for Q.12 , Q.5 and Q.6) and addressed in the comments (see results) is the aspect of economic benefits and their influence on the acceptance of wind energy. Indeed, also the statistical analysis revealed several significant relations between the level of acceptance and economic issues.

(for more methodological details click to go back to respective methods)

Profitability

Already the idea of wind energy's general profitability correlates with the use of wind energy. The more economically profitable people perceive wind energy, the more likely they are to support it (p=0,000000000000009, figure 11).

Fig. 11: Profitability and acceptance, n=168

Job creation

Visually perceivable and statistically proven, such a positively correlated relation does also exist between the perception of wind energy as a job creator and its acceptance (p=0,0000001, figure 12). Not surprisingly, this relation is also confirmed, if job creation can be chosen as a condition that would render wind energy reasonable (p=0,002, figure 13). This observation actually aligns with and concretises the above described results for the conditions under which opponents would support wind energy (jump back). None of the aspects would truly persuade the opponents to change their opinion, while supporters are rather optimistic.

Fig. 12: Job creation and acceptance 1, n=171 Fig. 13: Job creation and acceptance 2, n=184

Income for municipality

Another significant relation can be shown with regard to the answers to the question whether wind energy generates income for the municipalities (p=0,02, figure 14): Those who expect an additional income for the municipality, more frequently support wind energy. In this context it ought to be noted that the majority of people estimates the local public income to be relatively moderate (see results for Q.20) and also does not think that it could lead to higher revenues (see results for Q.6.6), but states that they would support wind energy if it did so (see results for Q.12.5). Between this position and acceptance there is also a significant correlation (p=0,000146465, figure 15), which might be explainable just as described with respect to jobs above.

Fig. 14: Income for the municipality and acceptance 1, n=162 Fig. 15: Q7-Q12.5 Income for the municipality and acceptance 2, n=183 .
To sum up this section on economic benefits, a combined graph of the three assessed elements has been created, making the statistically proven influence also visually comprehensible (figure 16).

Fig. 16: Direct Benefits and acceptance

These findings do generally align with literature that assigns a great importance to benefits sharing and also the financial involvement of the community. The positive impact of job creation on the acceptance is however seen critically, as the general local job creation potential of wind energy deployment is doubted (see Benefits sharing and jobs).

Concluding remarks

The confirmation of the links between acceptance and the impacts on landscape, the information management and direct benefits as well as the falsification of a connection with the actual or perceived distance are results that are consistent with the findings within the literature. This may give rise to the hypothesis that there are overall conditions valid in all cases where the development of wind energy meets locally affected people. However, as the analysis of the NIMBY phenomenon showed, it can be too short sighted to use general concepts for explaining local conditions. Nevertheless, these concepts may serve as a valuable starting point for analysis and also further on for developing answers to the concerns of the public, as there already exists a broad ' fundus' of recommendations.

To sum up the findings of this section and put them into practice, one might recall the so called “ENUF” framework: Engage local people early in discussions, Never use NIMBY to prematurely dimiss opposition, Understand the specific perceptions of the area and Facilitate follow-up discussion on local creation of benefits (Petrova 2013).

1)
Guo discovered a inverse U-shaped relation, describing that acceptance increases when theoretical wind energy development is moved from local level to a more distant county level and decreases when it is thought on a national level, far away from the interviewees location. However, the author highlights that the compensation power of economic benefits might be linked to the special importance that economic benefits have within this studied region, as it is relatively undeveloped compared to the high-income countries in Europe. Unaffected by this however, is the finding that environmental concerns increase the closer the development is located and thus decrease the acceptance.
2)
Only the Uebigau-Wahrenbrück survey had explicitly addressed the participation in the decision-making process (Question 8.1 and 8.2). However, due to the statistically unreliable return rate, the mainly negative answers cannot be analysed.
research/rural_areas/common_results/results2.txt · Last modified: 2017/02/02 16:50 by admin