Does multidimensional distance matter? Perceptions and acceptance of wind power

ABSTRACT In addressing long-term climate and energy challenges with wind energy, the acceptance of wind power is in a key role. Here, we systematically emphasize the several dimensions of distance, namely spatial, temporal, social, and experiential, and test the difference in attitudes toward near and distant wind turbines in these dimensions. We focus on both attitudes toward wind power and perceptions of its impacts using survey data with spatial information from southwestern Finland. Spatial distance associated significantly and positively with attitudes toward wind power. Regarding the social and temporal dimensions of distance, the direction of the association was against our hypothesis, meaning that the lower the distance was, the more positive were the attitudes. For experiential distance, no association was observed. The results recommend siting wind farms at a distance of over 10 km from population centers. Avoiding conflicts with vacation homes creates a new social challenge for wind power development.


Introduction
Mitigating climate change requires a radical reduction in the use of fossil energy.Wind power is one of the key solutions for achieving the EU's renewable energy targets and shifting the EU to carbon neutrality by 2050 (European Commission, 2019).Under scenarios up to 2050, renewable energy production will more than quadruple, emphasizing wind and solar power generation (e.g.Pleßmann & Blechinger, 2017).Wind power is also a solution for the energy crises caused by the Russian invasion of Ukraine and the increased demand for domestic energy sources in Europe.The corresponding development will be realized in EU member countries such as Finland, where wind energy generation is expected at minimum to triple in the coming decades (Lund et al., 2021).However, while wind power can contribute to solving the long-term challenges of climate change and energy dependence, the short-term consequences in people's everyday environment also affect the acceptance of wind power and may slow down the establishment of new wind farms (Ruddat, 2022).
There is a considerable amount of literature on the acceptance of wind energy and distance from the observer.Most previous studies have focused on the objective spatial dimension of distance (for a review, see Dobers, 2019).While the distance of wind farms from individuals has associated with the acceptance of wind power in some studies (Ladenburg et al., 2013;Ladenburg & Dahlgraad, 2012;Ladenburg & Möller, 2011;Swofford & Slattery, 2010), others have found no clear link between these two variables (Johansson & Laike, 2007;Braunholtz, 2003).When significant, the direction of the association between distance and acceptance has been found to vary from positive (Betakova et al., 2015;Ladenburg & Möller, 2011;Swofford & Slattery, 2010) to negative (Baxter et al., 2013;Meyerhoff, 2013), while situational variation or CONTACT Eija Pouta eija.pouta@luke.fiNatural Resources Institute Finland (Luke), Latokartanonkaari 9 FI-00790 Helsinki, Finland Supplemental data for this article can be accessed online at https://doi.org/10.1080/1523908X.2023.2279053.inconsistencies have also been reported (Baxter et al., 2020).Results concerning the number of turbines have been more consistent, with a general tendency for acceptance being reduced when more turbines can be seen (Ladenburg et al., 2013;Ladenburg & Dahlgraad, 2012), but positive effects of the number of visible turbines have also been reported (Ladenburg & Möller, 2011).
Beyond objective distances, many studies have applied reported experiences of distance (Krause et al., 2016;Langer et al., 2018;Mulvaney et al., 2013).According to Olson-Hazboun et al. (2016), the frequency of seeing turbines was strongly related to support for wind energy, while physical proximity was not.A positive association between visibility and acceptance was also observed by Ladenburg (2010).Firestone and Kempton (2007), as well as Ladenburg (2008), found a negative effect of self-reported visibility on acceptance, but several studies have reported non-significant effects (Ek, 2005;Johansson & Laike, 2007;Ladenburg, 2010).
Some studies have also recognized that different social and temporal settings are important for acceptance.Social distance can be expressed, for example, in relation to permanent or vacation homes (Ladenburg et al., 2013).Studies concerning the temporal dimension and acceptance have demonstrated a tendency for a Ushaped association.Over time, attitudes have ranged from very positive (that is, when people are not confronted by a wind power scheme in their neighborhood) to much more critical (when a project is announced) to positive again (some reasonable time after construction) (Devine-Wright, 2005;Wolsink, 2007).However, according to Windemer (2023), perceptions of a local wind farm did not change following construction, contrasting with common expectations that acceptance will increase over time.
Our contribution is to systematically focus on the different dimensions of distance, namely spatial, temporal, social, and experiential distance, and their effect on the acceptance of wind power.By spatial distance, we refer to (i) the physical distance between an observer and the nearest turbine, or (ii) the density of turbines around the perceiver.Temporal distance describes the perceived proximity of an event in time (e.g.past, present, future).Social distance denotes the closeness between a perceiver and an object in terms of the social setting.Regarding social distance, we assume that the permanent home provides a tighter social distance than other locations of an individual, such as a vacation home or owned land.Experiential distance describes the strength of experiences an individual has with an object (Fiedler, 2007;Föster, 2009).
To gain further insight into the reasons for acceptance, we go beyond general acceptance and examine individual perceptions of the impacts of wind farms.Previous studies have demonstrated that individually perceived impacts of wind power can explain the general acceptance of wind farms.The perceptions especially include negative impacts, such as noise, health risks, and landscape harm, but also positive impacts, such as community benefits, general community enhancement, and a preference for wind-generated electricity (Baxter et al., 2013;Wolsink, 2000).Olson-Hazboun et al. (2016) found that beliefs regarding economic benefits and landscape impacts influenced general acceptance.A distance dependency between perceived impacts and the location of wind turbines has been observed in the spatial and temporal dimensions of distance.For example, Peri and Tal (2020) found that the perceived environmental impacts of turbines were significant whenever sites were located less than 750 m from settlements.According to a study by Betakova et al. (2015), the perceived negative effects increased as a function of the number of turbines in an approximately linear manner.Concerning the temporal association, Eltham et al. (2008) observed that after implementation, opinions regarding the visual attractiveness of wind farms and the importance of the energy security they provide changed.
Our aim is to systematically test the hypothesis of an association of different dimensions of distance (spatial, temporal, social, and experiential) between individuals and wind farms and attitudes toward wind power.We proceed beyond attitudes to individually perceived impacts of wind power and analyze the associations of these perceptions with various dimensions of distance.To attach wind power information to the locations for each individual, we use data collected with a public participation GIS (PPGIS) survey.

Theoretical perspectives and hypothesis
How does the spatial association between wind turbines and individuals associate with individual attitudes and perceptions regarding turbines?A simplified explanation tested in recent decades has been the concept of NIMBY ('not in my back yard').However, in recent years, the NIMBY explanation has more often in the literature been considered as the view of developers and other proponents of wind power regarding local resistance rather than a theory-based description of the association between attitudes and proximity (e.g.Devine Chappell et al., 2021;Wolsink & Breukers, 2010).New theoretical considerations have been proposed in this context.Fujita et al. (2008) suggested construal level theory (CLT) (Liberman et al., 2002;Liberman et al., 2007;Trope & Liberman, 2003) to understand how attitudes toward near and distant objects in the environment are formed.CLT suggests that different dimensions of perceived distance, i.e. spatial, temporal, social, and hypothetical, and the intensity of the experience (Fiedler, 2007;Föster, 2009) can influence the mental representation of an object and its acceptance.CLT suggests that when the distance to an object is perceived as greater, people tend to assign more abstract, global, and general characteristics (e.g.clean energy).If the perceived psychological distance to objects is proximal, the objects are given more concrete characteristics (e.g.noisy) (Hart et al., 2015).
Few empirical studies have focused on different dimensions of distance suggested by CLT and analyzed the role distance plays in public perceptions of wind power.The measures of distance have varied from individual perceptions of distance, suggested by CLT, to geo-based objective measures.Sutter et al. (2019) qualitatively analyzed perceived distance to understand how individuals problematize and make decisions regarding wind energy.Hart et al. (2015) created a 'distance treatment' and measured positive and negative affective perceptions for projects that were proposed to be built at varying distances.Konisky et al. (2021) built on CLT but applied geo-based objective distances in the case of energy infrastructure and found various beliefs of impacts to be an important determinant of project support rather than proximity in general.
The hypotheses we test imply more positive attitudes with larger distances between individuals and wind turbines: H1 Spatial distance associates significantly and positively with the attitude towards wind power.H2 Temporal distance associates significantly and positively with the attitude towards wind power.
H3 Social distance associates significantly and positively with the attitude towards wind power.
H4 Experiential distance associates significantly and positively with the attitude towards wind power.
Furthermore, we test how the dimensions of distance associate with perceptions of the impacts of wind power.In the spirit of CLT, we assume that a smaller distance to turbines raises more local concrete perceptions, but higher distances raise more general principal-type aspects of wind power.We formulate the hypotheses as summary hypotheses over various dimensions of distance: H5 When distance (spatial, temporal, social, experiential) to a wind farm is higher than average, the more general impacts are emphasized by the respondents.
H6 When distance (spatial, temporal, social, experiential) to a wind farm is lower than average, the local impacts are emphasized by the respondents.

Case study area in southwestern Finland
The case study area comprises two regions (provinces), Varsinais-Suomi and Satakunta (see the map in Supplementary material), in moderately densely populated southwestern Finland, where wind energy production will be significantly increased in the near future (Huttunen, 2017).The regional land use plan has mapped 50 areas that are suitable for the expansion of wind farms (Regional Council of Satakunta, 2014;Regional Council of Southwest Finland, 2011).This is a prerequisite for opening the specified planning and development process of a wind power farm (Land Use andBuilding Act 132/1999, 1999).Municipalities have a monopoly over detailed land use planning in their areas.After the completion of a master plan by a municipal administration, private electricity companies may proceed with the license procedures and other operational phases in wind farm planning and construction.

Survey sample
The survey was implemented among two sub-samples in the provinces of Varsinais-Suomi and Satakunta, the first comprising private citizens and the second consisting of private non-industrial forest owners.In the first sub-sample, people were selected from a representative respondent panel by a commercial survey company, Taloustutkimus Oy (Taloustutkimus, 2022).The Internet panel of Taloustutkimus Oy comprised approximately twenty-five thousand respondents recruited to the panel using several contact methods 1 to represent the population.Regarding gender and age, the respondents selected from the panel represented the regional population as well as possible.
The names and e-mail addresses of the second sub-sample were received from the Register of Forest Owners managed by Suomen Metsäkeskus (the Finnish Forest Centre).We sent the survey invitation to all 7,200 private non-industrial forest owners owning forest in the case study area.
After testing the questionnaire in January 2019 with a pilot survey of 200 respondents, i.e. 100 per subsample, we conducted the main survey in February 2019.The practical data collection was organized by Taloustutkimus Oy.The survey was conducted online by sending out an e-mail call and an Internet link to the survey.The invitation e-mail that was sent to respondents of this study did not reveal the topic of the survey.After two reminders, we received 1,276 responses from the first sub-sample, corresponding to a response rate of 25.6%, and 1,176 responses from the second one, resulting in a response rate of 16.6%.The total response rate was 21.1%.
We evaluated the representativeness of the data sets separately (Supplementary material).In the first subsample, the proportions of the genders corresponded with the population.However, our data set including fewer respondents in younger age classes and more in older age classes than in the demographic statistics.The proportion of respondents living in towns and cities was greater in our study sample than in the population.The representativeness of the data set for forest owners was evaluated relative to data from the same provinces in a national study of forest owners carried out by Hänninen et al. (2011).The respondents were somewhat more frequently male in our data, but regarding their age, considerable dissimilarities could only be found in the two oldest age groups.

Questionnaire and measures
The questionnaire in the Internet survey was composed of four sections: (1) the perceptions and attitudes of the respondents regarding the landscape, environmental problems, future generations and energy production, and wind turbines, (2) mapping of the reference locations: the permanent home, a possible vacation home, and a forest lot, (3) attitudes toward compensation for the negative externalities of wind turbines (Mäntymaa et al., 2021;Mäntymaa et al., 2023), and (4) the general socio-demographic characteristics of the respondents.Here, we apply the survey information collected from parts 1 and 2.
The first dependent variable was a general indicator of the attitude towards wind power.The variable was measured on a five-point Likert scale (strongly agreestrongly disagree).The 'do not know' values were handled as missing values.The four items measuring attitudes were 'I support wind power generation', 'I don't like wind turbines', 'I perceive wind turbines as harmful' and 'We need more wind power'.These items were converted so that they were all scored in the same direction, and Cronbach's alpha was 0.935.The overall attitude toward wind power was summarized by taking the mean of these four items and denoting it as 'WP att', for which the higher the value, the more positive the respondent's attitude was toward wind power.
A question set with 21 items (Appendix 1) was used to define the second dependent variables, the perceptions regarding the impacts of wind power.The impacts focused on forestry and agriculture, nature, landscape quality, recreation, and people's well-being, as well as perceptions of wind power as a source of energy.With principal component analysis (PCA), the 21 items were condensed into three uncorrelated principal components (total explained variance 63%; loadings given in Appendix 1).A total of 35% of the respondents left the 21 item set incomplete, probably because they did not have experiences regarding some of the statements, such as statements regarding hunting.As we did not want to leave out responses because of a few missing values from among the 21 items, we used imputation to complete the data. 2  The three principal components were interpreted and named as follows: .

Positive local perceptions (PosLocal)
. Negative local perceptions (NegLocal) . Future energy perceptions (FutureEn) Positive and negative local perceptions describe the locally observed concrete impacts of wind power.Future energy perceptions include more general perceptions about global and national energy policy issues.The signs of the loadings were chosen so that statistical correlations of PosLocal and FutureEn would be related to a positive attitude, and vice versa for NegLocal.
As a part of the same Internet survey, spatial data on the respondents were collected with the map-based tool Harava (Sitowise, 2022).The Harava tool, integrated with the Internet survey, included a zoomable background map presenting the locations of current operating wind farms and proposed wind farm projects.We asked respondents to mark on the map their reference locations, i.e. the locations of their permanent home, possible vacation home, and, for private forest owners, the forest lot.If a respondent provided several locations in each category, for the permanent home we selected the response given first, while for the vacation home and forest lot, the selection was conducted randomly.
The spatial data of Finnish Wind Power Projects by the Finnish Wind Power Association (https:// tuulivoimayhdistys.fi/en/)jointly with Etha Wind, updated in February 2019, contained information on, for example, the name and phase of the projects, the number of turbines, planned megawatts, and the owner.The spatial analyses based on this spatial wind power data set and the survey data were performed with ESRI ArcGIS for Desktop software, version 10.6. 1. (ESRI, 2011).
For spatial distance, we had two measures: the first focused on one wind farm and was the proximity to the wind farm, while the second focused on several farms nearby and was the density of turbines.The proximity to a wind farm was calculated as the Euclidean distance (km) from the respondent's permanent home to the closest wind farm and classified into three categories: within a distance of 0-10 km, 10-30 km, and over 30 km.If a respondent had a vacation home and/or forest lot, the distance was also calculated from these reference locations in a similar manner.Consequently, we obtained from one to three values for each respondent to describe the proximity to the nearest wind farm.The turbine density within 30 km radius, a measure of how many turbines are nearby, was classified into three categories: 0 turbines, 1-20 turbines, and over 20 turbines.The density was also computed separately relative to the permanent home, vacation home, and forest lot, producing from one to three values for density depending on the respondent's ownership of a vacation home and/or forest lot.
Social distance was defined based on the respondent's reference location closest to wind turbines.As we could assume individuals to spend more time at the permanent residence, this location most closely reflected social distance (value 1), with vacation homes next (value 2), and forest lots having the highest distance (value 3).For example, if turbines were closer to the vacation home than to the permanent residence, the social distance was given the value of 2, i.e. the vacation home.
Temporal distance was defined as a dichotomous variable.The value of the variable was dependent on the phase of the closest wind farm project.It obtained values of 1 for operating wind farms or 2 for wind farm projects, meaning those farms that were not in production but in the planning or construction phase.
For experiential distance, we relied on the subjective self-reported variable of visibility.The measure for the visibility of turbines in the respondents' everyday landscape was based on the survey question asking whether operating or planned wind farms were or would be visible from ( 1) their home or (2) along their daily travel route, or (3) neither of these.These three categories were used in the analysis.
The following respondents were excluded from the analysis: respondents with a fully missing question set used in PCA (n = 12), respondents with missing responses in the measures of attitude variable (17), and respondents whose home location was not in the target provinces or was missing (617).Most of the latter owned a forest lot in the study area but did not live there.

Statistical analysis
First, we summarized descriptive information on the key variables, i.e. the dependent variables of attitudes and perceptions and independent variables of distances (Table 1).We also illustrated this information in maps (Supplementary material).Second, the mean of the attitude variable WP att (1-5, 5 more positive) in the classes of various distance variables was compared using ANOVA, having age-gender-group as an additional factor.Post hoc pairwise group-mean comparisons were based on Tukey's 'honestly significant difference' (HSD) test.The analysis was conducted separately for each spatial distance variable.Third, the means of the PCA-based wind power perceptions in the classes of various distance variables were compared (for details, see footnote 2).The R-package (R 4.2.0,R Core Team, 2021) 'mice' was used for the imputations, 'miceadds' for the ANOVAs and their pooling, and 'multcomp' and 'mitml' for the pooled post hoc group comparisons.

Results
Table 1 provides descriptive information on the key variables of the analysis.The attitude toward wind power had values ranging from 1 to 5. The general attitude was rather positive, with a mean of 3.5.The three perception components varied around zero mean, as they were principal components.Of the three perception components, the perception of clean energy had the highest standard deviation.The descriptive statistics for distances reveal that for the majority of the respondents (57%), their reference locations were at a distance of 10-30 km from a wind farm, and the typical density of turbines was over 20 turbines within 30 km (61% of respondents).Regarding temporal distance, the majority (54%) of the spatially closest wind farms were in the operation phase.Concerning social distance, the typical value, i.e. reference location, was the permanent home, as for 58% of the respondents, the spatially closest wind farm was observed from this location.This also reflected the fact that all the respondents had a permanent residence, but only 35% had a vacation home and 60% a forest lot.For experiential distance, 70% of the respondents did not observe wind farms in their daily life.
The analysis of variance presented in Table 2 provides the tests for hypotheses 1, 2, 3, and 4. The table presents the mean attitude in classes of distance variables.The mean attitudes in the classes of spatial distance provide the test for H1.The table presents results for two spatial distance indicators: Euclidian distance in kilometers to the closest turbine and turbine density within a 30-km radius.When the social and temporal dimensions are not taken into account, the attitude differs significantly (p < 0.001) in the classes of Euclidian distance (km), being lowest (3.32) for a distance under 10 km and highest (3.76) for a distance over 30 km.The attitude also differs significantly (p < 0.001) in the classes of turbine density.The attitude is more positive in the lowest turbine density category (3.76) than in the highest turbine density category (3.34).The both indicators suggest that the higher the spatial distance, the more positive was the attitude toward wind power.These results confirm H1, i.e. spatial distance associates significantly and positively with the attitude toward wind power.Furthermore, when the effect of spatial distance is analyzed in different classes of social distance, we observed that the association between spatial distance and attitude was statistically significant only if the class of social distance, i.e. the reference location of the respondents, was the permanent home.This applies to both measures, i.e.Euclidian distance and turbine density.The association of spatial distance and the attitude toward wind power was not sensitive to temporal distance, classified by turbines being either projected or in operation, as we found a statistically significant positive association between the attitude toward wind power and spatial distance in both classes.Second, in Table 2, we provide the means of attitudes in the classes of temporal distance for testing H2.The results demonstrate that temporal distance associates significantly with the attitude toward wind power when the effect of social distance is not considered (p = 0.01) or when the reference location is the permanent home (p = 0.02).However, the direction of the association is against the hypothesis: When the closest turbines were operating, the attitude was more positive than in the case of projected turbines.Thus, the results provide evidence against the hypothesis of a positive association between the attitude toward wind power and temporal distance.
Third, we tested the association of social distance with the attitude toward wind power (H3).The results partly supported the hypothesis, as social distance (permanent home < vacation home < forest lot) associated significantly with the attitude toward wind power (p = 0.002).However, the direction of the association was against the hypothesis.The smaller the social distance was, i.e. permanent home compared to vacation home or forest lot, the more positive was the attitude.A similar effect was observed in both categories of temporal distance.The effect of social distance was, however, sensitive to the fact that only part of the sample had a vacation home or forest lot.If the comparison was conducted for the subsample with all the reference locations available (not reported in Table 2), the difference in attitude was no longer significant (p = 0.3).
Fourth, the experiential distance was not found to associate significantly (p = 0.18) with the attitude toward wind power, providing evidence against H4.The uniformity of experience implies that individuals are somehow affected by wind projects simply by knowing that they exist, regardless of turbines being visible in their daily lives or not.
To test hypotheses H5 and H6, Table 3 presents the analysis of variance in the three wind power perceptions with varying distances, focusing on the main effects.
H5 stated that more general policy-level aspects are emphasized when the distance is higher in its various dimensions.To analyze whether we obtain support for H5, we focus on future energy component scores, which represent the general argumentation for wind power.From the conducted five tests concerning this component, we found a significant association with distance in three of them: with spatial distance in two tests, i.e.Euclidian distance (p < 0.001) and density (p < 0.000), and with experiential distance in one (p < 0.001).The direction of the association followed the hypotheses in all of the three significant comparisons: the higher the distance was, the higher were these general perceptions of wind power as a source of future clean energy.Thus, we can conclude that we obtained support for H5 in the dimensions of spatial and experiential distance, but not in social or temporal dimensions.We tested H6 with the components of positive and negative local perceptions, hypothesizing that the perceptions of local impacts are emphasized when the distance is lower in its various dimensions.In the comparisons of two components in the five measures of distance dimensions, i.e. ten comparisons, we found seven significant differences at the 0.1 significance level.From these seven results, we found significantly higher scores for lower distances in three comparisons, all related to spatial distance.Regarding temporal and social distance, the three significant comparisons were against the hypothesis, showing a lower score for these local perceptional components with lower distances.This result resembles the results from the attitude comparison, where temporal and social dimensions of distance were against the hypothesis.Table 3 reveals that in the case of experiential distance, the results were also against the hypothesis, as negative local perceptions did not differ significantly and positive local perceptions were at a lower level if the experiential distance, i.e. visibility from the home, was lower.

Discussion and conclusions
This study demonstrated that when considering the various dimensions of distance between wind turbines and individuals, only some of these dimensions explain the acceptance of wind power measured as wind power attitudes.Spatial distance associated significantly and positively with the attitude toward wind power, partly confirming our main hypothesis: the higher the spatial distance, the more positive was the attitude toward wind power.If compared with previous studies, our results are similar to those of Swofford and Slattery (2010), Ladenburg and Möller (2011), Ladenburg and Dahlgraad (2012), and Ladenburg et al. (2013), who observed a positive association between spatial distance and wind power acceptance by residents.Our conclusion is opposite to that of Konisky et al. (2021), who found very little evidence that proximity in general is an important determinant of support for a wind power project.In our case, the positive association between spatial distance and attitude was rather consistent with the expectations.
Regarding social and temporal distance, against expectations, the lower the distance was, the more positive was the attitude toward wind power.The difference between the permanent home and vacation home most probably relates to differences in the way of life at these two locations.In the permanent home, people are very dependent on technology and electricity production.In this context, it is logical to accept solutions that sustain a technology-dependent way of life.In the vacation home, the typical expectation in Finland is to lead a simple life in nature and to escape from daily routines that are dependent on technology (Vepsäläinen & Pitkänen, 2010).This raises a new challenge in wind power development, as vacation homes are distributed in rural areas that are also target areas for wind power development.However, from a societal point of view, preventing project development in areas with vacation homes would slow the deployment of wind power.Avoiding wind power development in vacation home areas would also be socially inequitable, as only a minority of the population has access to vacation homes and typically for a short period of time in a year.The challenge for developers and municipalities in vacation regions is how society could overcome the resistance from vacation home owners, given the need to deploy a considerable amount of wind power in an equitable manner.
For the temporal dimension, our results follow previous ones (Devine-Wright, 2005; Wolsink, 2007) demonstrating how attitudes shift in a positive direction as individuals become used to wind farms.The direction of attitudes along temporal distance was in our case not very clear.Although planned wind farms might be more distant in objective time than operating ones, they may be more salient in people's minds, and also closer to the actual moment of building the plants.
The results related to spatial distance contained an element of NIMBYism, namely that people do not want projects close to them.However, for experiential distance based on individual perceptions, no association with acceptance was observed.The results suggest that people do not want projects near them, regardless of whether they are directly affected.This implies that local public opposition need not have a basis in direct experience of the project.It could be, for instance, that people are concerned that projects will affect their private property values or they may be concerned about the effects of turbines on the wider community.The effect of objective spatial distance was especially significant in relation to the permanent home, but not so much in relation to the vacation home or to the forest lot, suggesting that a tighter social community around the permanent home is relevant to acceptability.
Beyond the previous empirical studies of CLT and energy infrastructure (Hart et al., 2015;Konisky et al., 2021;Sutter et al., 2019), we further tested whether the association of distance with perceptions follows the idea of a positive association of distance with more general, politically oriented perceptions and a negative association of distance with more practical, local perceptions.As hypothesized, a high spatial and experiential distance associated with stronger perceptions of clean future energy related to wind power.However, also here, social and temporal distance did not follow the hypothesis.
In more concrete local perceptions, either negative or positive, the hypothesis of stronger perceived impacts associated with a lower distance was confirmed in the case of spatial distance.For temporal and social distance, the results were against the hypothesis, displaying a lower score for these local perceptional components with low distances.
The results of our study were based on the situation before the ongoing strong expansion of wind power generation along the west coast of Finland.Even though this area includes some of the pioneering wind power areas in Finland, the density of turbines is still rather low and distances to the closest wind turbines are relatively high compared to some other countries, such as Germany or Denmark.In the Finnish case, it has still been possible to locate wind turbines in rather remote areas with a low population density.This means that the number of respondents having a very close relationship with intensive wind power farms is still low, and the effect of a very close connection could only be partly captured from our data.
The results are promising for increasing distributed electricity production in the future, and especially if this increase must be achieved with energy from renewable sources such as wind power.Although statistically significant, the practically rather small effect of the proximity of wind turbines provides freedom for decisions on the location of wind farms in land use planning, especially beyond a distance of 10 km from population centers, if the opposition related to siting wind farms in areas with vacation homes can be resolved.

Table 1 .
Descriptive statistics for the key variables.

Table 2 .
The mean of attitude (1 … 5, 5 more positive) in the classes of various distance variables.(ANOVA with post-hoc comparisons using Tukey's 'honest significant difference' method, where different letters indicate significant differences at the 0.1 significance level).

Table 3 .
The mean of wind power perceptions in the classes of various distance variables.(ANOVA with post-hoc comparisons using Tukey's 'honest significant difference' method where different letters indicate significant differences at the 0.1 significance level).(N = 1733).