Ue

Ja

b

c

d

a

ARRA

KPUFPLP

1

re2iw2tladBs

h0

Land Use Policy 58 (2016) 491–501

Contents lists available at ScienceDirect

Land Use Policy

j o ur na l ho me page: www.elsev ier .com/ locate / landusepol

tility-scale solar and public attitudes toward siting: A criticalxamination of proximity

uliet E. Carlisle a,∗, David Solan b, Stephanie L. Kane c, Jeffrey Joe d

Dept. of Political Science and Philosophy, University of Idaho, United StatesEnergy Policy Institute, Boise State University, United StatesWashington State University, United StatesIdaho National Laboratory, United States

r t i c l e i n f o

rticle history:eceived 22 April 2016eceived in revised form 2 August 2016ccepted 4 August 2016

eywords:ublic opiniontility scale solaracility sitingroximityand typeslace attachment

a b s t r a c t

Public opinion polls show that the American public strongly supports the development of large-scalesolar power facilities. Yet, often with renewable energy development, when specific developments areproposed, they are met with local opposition. In the past, many scholars relied upon explaining suchopposition in terms of a NIMBY (Not In My Backyard). However, NIMBYism is criticized as an overlysimple, incorrect, and pejorative characterization of opposition. Yet, while some criticize NIMBY expla-nations, other research demonstrates that distance indeed matters. Research also demonstrates that placeattachment, socio-demographic characteristics, and project-related characteristics also matter. Our studyintegrates these different factors to better understand the nature of support for large-scale solar devel-opments. Specifically, we consider visual impact of large-scale solar facilities and what effects distancebetween different types of land and the proposed solar facility might have on public support. Therefore,we examine proximity but not just proximity to one’s residence but rather to different types of land.

Our data are from a 2013 telephone survey (N = 695) from six Southern Californian counties (Inyo, Kern,Riverside, San Bernardino, San Luis Obispo, and Ventura), selected based on existing and proposed solardevelopments in those areas and available suitable land. Findings suggest that the visual impact of large-scale solar facilities does matter for support and that preference for buffer sizes, and thus proximity ofproposed large-scale solar facilities, do change depending on the type of land being considered.

© 2016 Elsevier Ltd. All rights reserved.

. Introduction

In March 2015, the U.S. Energy Information Administration (EIA)eleased a report announcing California as the first state to gen-rate more than 5% of its electricity from utility-scale solar (EIA,015). Solar power is growing and especially so in California. In fact,

n 2014 California generated 9.9 million megawatt-hours (MWh),hich was a 6.1 million MWh increase over 2013 figures (EIA,

015). This growth in solar energy production in California is largelyhe result of new utility scale facilities including Topaz, Desert Sun-ight, Ivanpah, and Genesis. The sheer growth in solar is greatlyttributed to state policies including state renewable portfolio stan-

ards (RPS) and incentives (rebates and net-metering policies).oth policies and geo-physical conditions have led to the growth inolar production in states other than California. Arizona and Nevada

∗ Corresponding author.E-mail address: carlisle@uidaho.edu (J.E. Carlisle).

ttp://dx.doi.org/10.1016/j.landusepol.2016.08.006264-8377/© 2016 Elsevier Ltd. All rights reserved.

have experienced an increase in solar production due to their obvi-ous solar resources and New Jersey and Massachusetts as a resultof their state RPS. More importantly, growth in solar productionis not contained to the U.S. With the more recent decrease in thecost of solar system manufacturing, Photovoltaic (PV) systems haveexperienced considerable growth since 2003, not only in the U.S.but also in China, Japan and Germany, especially.

In terms of renewable energy (RE) resources, while utility-scalesolar electricity production trails behind wind, it is still a promis-ing source of energy to help alleviate the growing dependence onfossil fuel-based energy. In fact, the EIA forecasts solar electric-ity generation to increase by almost ten percent annually through2035 (EIA, 2012, p. 90). Moreover, the data on public opinion sug-gests that an overwhelming proportion of Americans support solarenergy development (Carlisle et al., 2014, 2015; Farhar, 2003) andthe public is even willing to pay more for clean energy production

(Farhar, 2003). As well, the 17 solar energy zones in six South-western states–California, Nevada, New Mexico, Arizona, Utah, andColorado (Cart, 2012) demonstrates the Obama administration’s

4 Use Po

ep

gaoaaitarpoat

Piafaslteesffca

2

dA22sK2OpawaeRse(iat2e

rwtso

92 J.E. Carlisle et al. / Land

fforts to deliver on its promise to make renewable energy a largerortion of the nation’s energy portfolio.

President Obama has made policy advances regarding thereater use of RE. However, these policies are often met, evenmong pro-RE groups, with criticism due to the expedited naturef the permitting process. Environmental and conservation groupsre concerned about the impact of solar facilities on the ecosystem,nd so despite widespread support for RE development in general,ncluding solar, specific projects are often met with strong opposi-ion (Klick and Smith, 2009). While large-scale solar developmentsre likely to occur away from people’s neighborhoods to moreemote areas where geographic and solar attributes are enhanced,ublic opposition still exists. Thus, a fundamental aspect of devel-ping and expanding solar is to understand factors affecting publicttitudes toward the resource in general, as well as those specifico place and geography.

This research considers public attitudes toward utility-scaleV solar development.1 While much of the research on prox-

mity examines distance in terms of distance between a facilitynd one’s home, our research also considers respondents’ pre-erred distance between a proposed large-scale PV solar facilitynd different land-types. In particular, we consider the effect ofocio-demographic and place attachment predictors on support forarge-scale solar developments in terms of land-type (e.g. agricul-ural land, recreation areas, wetlands, wildlife migratory routes,tc.). Our findings demonstrate that support is dependent on sev-ral socio-demographic and place-related predictors. Additionally,upport is also related to how respondents perceive the solaracility will visually impact the landscape. Finally, we find that pre-erred buffer distance between a proposed utility scale solar facilityhanges depending on the proximity between a proposed facilitynd the type of land. Therefore, [buffer] size does matter.

. Previous research

Many scholars have measured public attitudes towards energyevelopment in the U.S. and Western Europe (Ansolabehere 2007;nsolabehere and Konisky 2009; Sovacool 2009; Van der Horst007; Walker 1995; Wüstenhagen et al., 2007; Wolsink and Bürer,007), much of it with regard wind energy projects and controver-ies (see Bell et al., 2005; Ladenburg, 2008; Klick and Smith, 2010;rohn and Damborg, 1999; Swofford and Slattery, 2010; Wolsink000, 2007; Warren et al., 2005; Warren and McFadyen, 2010).verall, research demonstrates that respondents generally sup-ort RE development (Bell et al., 2005; Devine-Wright, 2005; Klicknd Smith, 2010; Warren et al., 2005; Wolsink, 2000), especiallyhen compared to other energy sources such as nuclear (McGowan

nd Sauter, 2005). While support for renewable energy tends tobb and flow with fluctuations in gas prices (increased support forE increase with upticks in gas prices), it has remained relativelytable (Gallup, 2013; McGowan and Sauter, 2005; Smith, 2002),xcept for a recent dip in overall support between 2011 and 2013Gallup, 2013). In addition, public support for government fund-ng of alternative energy projects has recently declined, especially

mong Republicans (Pew, 2012). Among different RE types, solarends to be the most positively regarded (Gallup, 2013; Greenberg,009); and wind to be the most polarizing (DTI Scottish Executivet al., 2003). However, few studies in any countries examine public

1 Large-scale solar facilities or utility-scale solar facilities are different fromesidential rooftop solar, solar panels on commercial or public buildings, andidespread installation of panels on public infrastructure such as utility poles. For

he purposes of this study, each large-scale solar facility is intended to power thou-ands of homes and businesses, requiring significant land-coverage in the hundredsr thousands of acres per project, depending on specific installation size.

licy 58 (2016) 491–501

attitudes towards utility scale solar energy development by itself,although work by Carlisle et al. (2014, 2015) has certainly made acontribution in this area.

In terms of local planning and development, research demon-strates that while general support for RE is often widespread,opposition to specific facility proposals exists. Resistance to theCape Wind Project, which was proposed for construction on Horse-shoe Shoals in Nantucket Sound near Cape Cod, Massachusetts, isa notable example. Notwithstanding, scholars have begun to moveaway from the NIMBY explanation, due to criticisms that NIMBY ispejorative, oversimplified, and tends to consider opposition in self-ish or irrational terms or the result of ignorance. Rather, scholarshave found that opposition can be both very informed (Michaudet al., 2008; Petts, 1997) and rational (Gross, 2007). In particular,more recent literature looks beyond NIMBY and considers a vari-ety of other explanations built upon a psychological environmentaltheoretical framework, finding that variation in support and oppo-sition for specific facility proposals is quite nuanced. Thus, suchresearch considers the relationship between support and oppo-sition to RE and demographic factors, socio-psychological factors(knowledge, direct experience, environmental and political beliefs,place attachment); and contextual factors (technology type andscale, institutional structure, and incentives).

Recent research indicates that support and opposition towardRE varies according to demographic variables including age,income, education, and gender (Firestone and Kempton, 2007;Ladenburg, 2009). Devine-Wright (2008) cites several studies con-ducted in the UK that demonstrate the significant impact of age onsupport for RE, although there are contradictory findings regardingthe nature of the relationship. For example, older individuals aremore opposed to or less willing to pay for RE than younger indi-viduals (MORI Social Research Institute for Regen SW, 2003; seealso Ottman and Herbert, 1993; Vorkinn and Riese, 2001; Zarnikau,2003) while other studies find a U-shaped relationship where bothyounger and older respondents are less opposed to RE than aremiddle-aged cohorts. Still other studies find older respondentsare less opposed to nuclear energy than are younger respondents(Populus, 2005). Similarly, research considering the impact of sexon support for RE projects is mixed. While some research findswomen to be more environmentally concerned (Mohai, 1992) andsupportive of renewables than men, men tend to demonstrategreater awareness and greater support for solar, nuclear, and wind(Brody, 1984; Corner et al., 2011; Department of Trade and Industry,2003DTI Scottish Executive et al., 2003; Klick and Smith, 2010).Income and class have both been found to be positively corre-lated with support for renewable, nuclear, and wind energy (Corneret al., 2011; Firestone and Kempton, 2007; MORI Social Research forRegen SW, 2004).

Place attachment is a collective orientation that describes theprocess of becoming attached to an environmental setting (Vorkinnand Riese, 2001). Moreover, place attachment allows that this ori-entation need not be exclusively positive. Manzo (2003, 2005)characterizes place attachment as a positive connection with whatis familiar, such as home or neighborhood, and others link placeattachment to length of residence (Ahlbrandt, 1984; Taylor, 1996).For environmental psychologists, place-identity relates to thedimensions of self that develop through interaction with the envi-ronment via beliefs, preferences, feelings, values, etc. (Proshanskyet al., 1983). When change is proposed to a place, it can be perceivedas a “disruption” or “threat” and can be met with action in orderto preserve the community or neighborhood to which individualsare likely closely attached. Threats or disruptions to place attach-

ment can result from development, crime, neighborhood decline,and even natural disasters (Brown and Perkins, 1992). Individu-als develop a sense of place from the environmental experiences

se Po

tc

iwaaetctwtd“Tl

tcopi2podtpTvea

tdoenmVan(pmhtaitLnlsfSaStsiitl

another objective of this study.The literature thus provides a solid framework for understand-

ing factors that might play a role in influencing public perceptions

2 Up until recently there has been little effort to examine public opinion on solarenergy independent of other renewable energy sources. Also, solar use as a viablerenewable energy source is more recent in large part due to the reduction in costof solar technology. With the cost of solar technology dropping, solar’s popularityhas increased. There is much research on public attitudes toward wind no doubtto its role as a major source of renewable energy production, ranking second tohydropower in terms of proportion of total U.S. energy generated it (4.4%). On theother hand, solar, which makes up only 0.4% of the total U.S. energy generated (www.eia.gov). Globally, growth in solar PV module production grew by an average of 78%

J.E. Carlisle et al. / Land U

hat they accumulate over time and such experiences are usuallyharacterized as positive, but not necessarily so.

Place attachment often implies a localness due to the fact thatndividuals are most likely to use local areas more than those

ho live far away. Thus, locals are likely to have greater placettachment to local areas than are those who live elsewhere. As

result, scholars have found that the neighborhood, residentialnvironment, or community is to what individuals tend to attachhemselves. However, areas that possess value as a national symbolan also garner such place attachment by individuals who live far-her away. Moreover, places, and thus place attachment, can varyith regard to scale (e.g. a house, playground, or forest, etc.) and

angibility of an area. Finally, Riger and Lavrakas (1981) identify twoistinct dimensions of neighborhood attachment: “rootedness” andbonding,” the latter of which also identified as “local bonds” byaylor et al. (1985) and centers on social bonds and also includes

ength of residence.In terms of the site screening and selection process, we believe

he proximity literature to be promising. Again, proximity studiesonsider distance from respondents’ locations to assess whetherpposition is a function of location in relation to the owners’roperties. A less explored extension of this literature—which has

nconsistent findings (Van der Horst, 2007; Swofford and Slattery,010 Warren et al., 2005; Devine-Wright, 2005)—is to examineublic attitudes toward the development of infrastructure nearbyr on specific types of land. This is a new function because respon-ents are not only queried for preferences based on distance fromheir property or personal location, but location and distance ofroposed project sites from specific parcels based on the land use.his is more in line with what Devine-Wright has noted as socialalue resulting in “place protective” actions, in turn extending it tonvironmental attitudes (Devine-Wright, 2009), although a moreccurate statement would be to call it “use protective.”

Blaming local residents’ opposition to proposed RE infras-ructure has long been a favored explanation for delays in itsevelopment. Even so, researchers have found opposition the resultf a variety of factors. Surprisingly, studies indicate that evennvironmentalists have opposed projects because of the potentialegative impact of solar facilities on rare desert plants and ani-als (Cart, 2012). Another solar project, proposed for the San Luis

alley of Colorado, garnered opposition when both local residentsnd environmental groups expressed concern about the potentialegative impact the project would have on the local ecosystemFarhar et al., 2010). Even beyond the protective opposition that theublic demonstrates toward RE developments (and other develop-ents as well) due to the negative impact of the ecosystem and

abitat, public opposition to energy development has also beenhe result of people’s identification with the land (Devine-Wrightnd Howes, 2010). Additionally, many scholars have investigatedndividual preferences of landscapes and landscape attributes. Inruth, the natural and physical environment is not homogenous.ikewise, perceptions and perhaps protective attitudes toward theatural and physical environment vary depending on land type or

and use. More generally, research has found a strong relation-hip between socio-demographic characteristics and preferenceor natural landscapes over developed ones (Herzog et al., 2000;cott, 2002). Additionally, agrarian, water, and open landscapesre also viewed positively (Herzog, 1985; Kaplan and Kaplan, 1989,trumse, 1994). In accord with this research, the question we seeko answer is whether support or opposition to proposed large-scaleolar energy development differs based on the land type near whicht is proposed to be built. Similarly, we also seek to understand

ndividual level and project related characteristics explain the rela-ionship between support for solar, desired buffer distances, andand types.

licy 58 (2016) 491–501 493

Certainly, NIMBYism and proximity are tightly connected. Theidea of “backyard” implies close proximity and thus opposition asthe result of how proximate a proposed “something” is to one’sbackyard. Interestingly, research to date has not found consistentresults examining the effect of proximity in terms of support for oropposition to RE facilities. Swofford and Slattery (2010) find thatwith regard to wind farms, those living closest to the wind farmpossess the most negative attitudes towards it. However, otherstudies (Braunholtz, 2003; Krohn and Damborg, 1999; Warrenet al., 2005) have found the exact opposite, with those living closestto wind farms having the most favorable opinions towards them.Breukers and Wolsink (2007) investigation of the acceptability ofwind turbines finds that different types of landscapes determinessupport or opposition to them. For example, Wolsink demonstratesthat some people only accepted wind turbines offshore and pub-lic acceptance was lowest near residential areas, scenic places,or places representing nature. Molnarova et al. (2012) find thatphysical attributes of both the landscape and the wind turbinesinfluenced attitudes more than did demographic and attitudinalfactors. While very little research exists on the nature of public sup-port for large-scale solar facilities,2 Carlisle et al. (2014, 2015) doconsider support/opposition to large-scale solar in terms of placeattachment and predictors of support among Americans, respec-tively. However, their research does not address the relationshipbetween proximity and support for these facilities and thus is anobjective of the research herein. Additionally, we consider the rela-tionship between support for solar facilities and the proximity ofproposed solar facilities to different types of land (e.g. agricultural,residential, breeding grounds, recreational areas, etc.).

Finally, contextual factors have proven to be particularly rel-evant to explaining support and opposition to RE. In particular,project visibility and its visual impact have been shown to be sig-nificant predictors of public support, especially with regard to bothon- and off-shore wind developments; projects that are out of viewtend to garner greater public acceptance (Brittan, 2002; Jones andEiser, 2010; Ladenberg, 2009; Phadke, 2010; Warren et al., 2010;Breukers and Wolsink, 2007). Studies have found that the effectof visual impact of wind turbines can be mixed where some indi-viduals consider the presence of turbines on the land or sea-scapeto be positive whereas for others the visual presence is negative(Breukers and Wolsink, 2007). Ladenberg (2009) considers prox-imity (a proxy for prior experience) and visual impact, finding thatpeople with prior experience from offshore wind farms locatedfarther from the coast have a more positive attitude of the visualimpacts from offshore wind farms than those who have lived nearwind farms nearer to the coast. Certainly, the infrastructure size,potential landscape disturbance, and geographical requirementsdiffer for wind turbines and PV solar facilities. Yet, the effect ofvisual impact on public attitudes for PV solar is unknown and is

from 2006 to 2011. In the U.S. In the U.S. we see a similar trend with a significantgrowth in consumption of solar PV since 2006, but still it lags behind the growth inconsumption of wind energy. (www.eia.gov). Finally, surveys do not usually querythe public on non-salient issues and due to the low saliency of solar historically, wehave seen few if any surveys solely dedicated to public attitudes of solar.

4 Use Po

asoviA“sSunmlpf

1

2

3

3

hwssOpdocTphccmcSr

wivGpodws

iJvc

bolic measures were initially designed and used by Cotton andDevine-Wright (2011) to reflect the natural beauty of the local-ity, the presence of a strong community bond, and socio-economic

5 American Association of Public Opinion Research. 2011. Standard Definitions:Final Disposition of Case Codes and Outcome Rates for Surveys. Available at: http://

94 J.E. Carlisle et al. / Land

bout and support for solar energy, in particular utility-scale PVolar in Southern California. In this study, we consider a varietyf the environmental psychology explanations (i.e. place relatedariables), in combination with demographic factors while focus-ng more specifically on visibility and distance of proposed facilities.

unique aspect of our research design is that we test the effect ofproximity” in terms of buffer distance between a proposed large-cale solar facility and different land-types. The U.S. Southwest, orouthern California, more specifically, is the most likely locale fortility-scale solar developments and understanding the level andature of support of Southern Californians is relevant for policy-akers and stakeholders. Further, understanding how distance and

and-types relate to support or opposition to PV solar is useful forlanners and developers worldwide. Our research questions are as

ollows:

Does support for large-scale solar facilities differ depending onwhether the facility will be visible or not?

Do respondents prefer different sized buffer zones between var-ious land-types proposed for large-scale solar construction?

Do place related variables (e.g. sense of place, place attachment,etc.) predict public attitudes regarding buffer distances regard-less of land-type development?

. Data

We used a dual-frame telephone survey methodology, withousehold Random Digit Dialing (RDD) landline (n = 5,442) andireless RDD telephone numbers (n = 5,000). Both frames were

imple random samples of numbers from within six counties inouthern California–Inyo, Kern, Riverside, San Bernardino, San Luisbispo, and Ventura, which were chosen because of their closeroximity to many utility-scale solar facilities in various stages ofevelopment (proposed, under construction, or operating). Basedn data from the Solar Energy Industries Association,3 there isonsiderable solar development occurring in Southern California.herefore, the goal behind this sampling decision was to collectublic opinion data from people who are more likely to have first-and experience with utility-scale solar facilities sited in theirounties. Moreover, Southern California makes for an interestingase study to study solar development. It is viewed as one of theost desirable places to live in the U.S., generally with higher per

apita income and higher property values than much of the Unitedtates, and it is a region with people who are very proactive in bothegional and local politics.

While the primary purpose of our project and data collectionas to support the development of a GIS based siting tool that

ncludes sociopolitical constraints and public preferences, our sur-ey is also designed to test other research questions as well. TheIS-based siting tool assesses alternatives to identify feasible andotentially optimal sites for utility-scale solar development. Inrder to design this tool, our survey assesses the value that respon-ents place on a variety of geographic features and land-types andhich of these they would like to see protected from the siting of

olar or those that they consider preferable for siting solar.

Survey data were collected on Wincati telephone interview-

ng software. Calls began 21 March 2013 and continued until 13une 2013.4 The SSRU employed a Spanish-language speaking inter-iewer. In total, 695 interviews were completed, with 82 interviewsompleted in Spanish. The final response rate (AAPOR2) for the two

3 http://www.seia.org/research-resources/major-solar-projects-list.4 Wincati, v 5.0. 2012. Sawtooth Technologies, Inc. Northbrook, IL.

licy 58 (2016) 491–501

frames combined is 9.7 percent,5 the final cooperation rate is 27.7percent, and the final refusal rate is 27.1 percent. Heavy fires withevacuations were reported in some of the counties in early Maywhile this study was underway, however we expect little impacton response rate.6 All statistical analyses were conducted using IBMSPSS Statistics, version 22.0.

3.1. Measures

To gauge support for utility-scale solar projects we use severalvariations of our “support for solar” question. Specifically, we askeach respondent about support for construction of large-scale solarin general (in the U.S.) and near to where they live. In addition,we ask about support for large-scale solar nearby but not visible.An example of our “support for solar” items is as follows7: “Sup-pose the construction of a large solar facility was planned near towhere you live. How strongly would you support or oppose its con-struction?” Answer categories for all “support for solar” items arebased on a five-category Likert scale from 1 (strongly oppose) to5 (strongly support), which have been collapsed and recoded intoa 3-category variable, 1 (oppose); 2 (neither oppose nor support);and 3 (support).

To gauge preferred buffer sizes between different land-typesand proposed large-scale solar facilities we ask respondents, “Howmuch buffer distance is acceptable between a large solar facilityand [a residential area; existing agricultural land; an area used asnesting sites or breeding grounds by wildlife; an area of culturalor historical importance; recreation areas (e.g. hunting, fishing, orhiking locations); wetlands; an area used as a migration route bywildlife; existing solar]?” Answer categories for all buffer distanceitems are self-reported preferred distance and range from adjacentto over 5000 miles. While most of the self-reported distances wereconveyed in miles by the respondents, some were in non-mile dis-tances (e.g. yards, block, kilometers, acres, etc.) were converted tomiles and recoded as a four-category Likers scale as follows: 1) Lessthan a mile; 2) 1–5 miles; 3) 6–10 miles; and 4) More than 10 miles.“Don’t know” and “No preference” were coded as system-missingand thus omitted from the analyses.

Our predictor variables include those that typically relate sig-nificantly to support and opposition to energy facilities and alsoare often standard predictors of environmental and political atti-tudes. These include demographic factors such as sex, age, race,education, and urbanicity; socio-psychological factors such as placeattachment, party identification, and belief in the seriousness of cli-mate change; and project-related factors including the perceivedimpacts—both positive and negative—of a proposed solar facilityand its construction.

Place attachment is measured by proxy with length of residencyin the area. In addition, we also capture place-related symbolicmeanings with five Likert-type items. The five place-related sym-

www.aapor.org/AM/Template.cfm?Section=Standard Definitions2&Template=/CM/ContentDisplay.cfm&ContentID=3156.

6 The bulk of the 2013 fire season fell after mid-June. There were three fires thatdestroyed structures damage during the call period − the Springs fire in Venturacounty which destroyed 10 structures (5/2–5/9), the Powerhouse fire in Los Angelescounty which destroyed 58 structures (5/30–6/10), and the Summit fire in Riversidecounty which destroyed 2 structures (5/1–5/4). Given the long calling period, mostof which occurred prior to any major fires, and the fact that the largest and mostdisruptive fires occurred after the calling period concluded, we expect little impacton response rate.

7 Wording of all items available upon request.

se Policy 58 (2016) 491–501 495

vtatriotso

ipfjL(wasW

mipanytplIvasctpftacloo(

4

stdtbihamio

ol

J.E. Carlisle et al. / Land U

itality and are as follows: I don’t want the landscape of this placeo change; this place has a great community of people; this place is

beautiful area; nature is unspoiled in this place; and this place isoo quiet—it needs to liven up a bit. All items, excluding length ofesidency in the area, included a 5-category response format rang-ng from 1 (strongly agree) to 5 (strongly disagree) with a midpointf neither agree nor disagree. While we are not certain the direc-ion of the relationship that place attachment will have on degree ofupport, we expect that it will have a strong and significant impactn support.

Project-related items capture the perceived impacts, both pos-tive and negative, such as jobs, too much traffic, and changes inroperty values that might result from building a large-scale solar

acility in the vicinity, as well as attitudes about the proceduralustice of the decision-making process. These items also includeikert-type answer categories ranging from 1 (strongly agree) to 5strongly disagree), except for perceived impact on property value,hich ranges from 1 (increase greatly) to 5 (decrease greatly) and

re included based on previous research that demonstrates theirignificance in terms of wind siting (Devine-Wright, 2013; Devine-

right and Howes, 2010).We used a factor analysis using the principal components

ethod with a varimax factor rotation to assess the underly-ng dimensions of our place-related and interpreting change viaroject-related measures. Our analyses of the twelve items reveals

four factor structure (Table 1). The first factor, “Good Commu-ity,” consists of four items explaining 60% of the variance andields an eigenvalue of 7.0. The four items loading on the first fac-or capture an underlying positive regard for the community, itseople, and an unwillingness for it to change and are highly corre-

ated with a Cronbach’s alpha of 0.65. The second factor, “Positivempacts—Community,” consists of two items explaining 13% of theariance and yields an eigenvalue of 1.0.8 These two items capturen underlying positive economic benefit from building large-scaleolar in terms of industrialization and jobs and are also highlyorrelated with a Cronbach’s alpha of 0.76. Factor three, “Posi-ive Impacts—Personal,” also consists of two items capturing theositive impact that would result from building a large-scale solar

acility nearby but do so more in terms of the personal impact. Thesewo items explain 12% of the variance and yield an eigenvalue of 1.5nd a Cronbach’s alpha of 0.5. The fourth factor, “Negative Impacts,”aptures negative aspects of solar development with three itemsoading and explaining 11% of the variance with an eigenvaluef 1.0 and low reliability with a Cronbach’s alpha of 0.27. Lengthf residency in the area failed to load on any of the four factorsTable 2).

. Results

To begin our analyses we first consider support for large-scaleolar development. More specifically, Fig. 1 illustrates the propor-ion of Southern Californians who (1) support large-scale solarevelopment in the U.S.; (2) near to where they live; and (3) nearo their home, but where developers have guaranteed it will note visible from their home. With (3), we predict that support will

ncrease relative to baseline support for a solar facility near theirome. Overall, and in all three circumstances, Southern Californiansre very supportive of large-scale solar development. Approxi-

ately 80% of respondents support large-scale solar development

n the U.S., with only a very small proportion (5%) of respondentspposing construction. Similarly, respondents are overwhelmingly

8 Several place-related items included in Devine-Wrights (2011) analyses weremitted from ours due to survey constraints. Two of the omitted items would have

ikely loaded on the second factor and one likely on the first.

Fig. 1. Public Support/Opposition for Solar Development.

supportive of large-scale solar development where they live, withapproximately 75% of respondents supporting large-scale solardevelopment and 10% opposing. Additionally, we see that in termsof proposing a large-scale solar nearby to where the respondentlives, when the question includes the wording that developers haveassured that the solar facility will not be visible to the respondent,support increases to 80% (opposition remains the same at 10%).Moreover, a paired samples t-test reveals a statistically reliable dif-ference between the mean level of support for building large-scalesolar in the nearby but not visible (M = 2.66, SD = 0.66) versus themean level of support for building large-scale solar developmentsnearby (M = 2.71, SD = 0.62); t(666) = −2.37, p < 0.02). We also detecta statistically significant difference between support for buildinglarge-scale solar “nearby” (M = 2.76, SD = 0.53) versus the meanlevel of support for building large-scale solar developments in theU.S. (M = 2.84, SD = 0.442); t(628) = 4.54, p < 0.000).

4.1. Buffer distance & land types

We now turn our attention to public preferences for buffer dis-tances between large-scale solar and a variety of land-types (Fig. 2).Typically, proximity is considered in terms of proximity of a pro-posed facility to one’s residence, but we suspect that not all landis created equal. That is, people value to a greater or lesser degreecertain types of land and thus prefer different sized buffers, accord-ingly. As results demonstrate, respondents indeed prefer differentsized buffers based on the particular type of land. For example, theonly type of land where respondents prefer a buffer of less than1 mile is that which is currently used for an existing solar facility.Specifically, land that is currently used for existing solar, 45% ofrespondents prefer a buffer of less than 1 mile, 15% prefer a bufferbetween 1 and 5 miles, 10% prefer a buffer of 6–10 miles and 25%prefer a buffer of 11 miles or more (X̄ = 31.77 miles; M = 1.0 mile).For all other types of land, except breeding grounds and wildlifemigration routes, the preferred buffer distance is 1–5 miles. Infact, for both breeding ground and wildlife migration route, themodal distance category is 11+ miles (X̄ = 37.15 miles; M = 10.0miles; X̄ = 35.38 miles; M = 6.08 miles, respectively). Another strik-ing finding is that for recreation areas, nobody selected less than1 mile as the preferred distance. The modal category is 1–5 milesbut a large proportion of respondents (40%) preferred more than 11miles (X̄ = 33.82 miles; M = 5.0 miles). These findings present clearsupport for H2. Moreover, the findings suggest that NIMBY, in itsmost basic form, has been interpreted quite narrowly to mean only‘Not In “My” Backyard’—one’s own residence—without any regardfor other land types that people might care about (e.g., national

parks, etc.). Our findings show that local concern (or NIMBY) is notjust based on this narrow interpretation or operational definition ofNIMBY. Therefore, we believe that the definition or interpretationof NIMBY needs to be changed and/or expanded to include more

496 J.E. Carlisle et al. / Land Use Policy 58 (2016) 491–501

Table 1Rotated Component Matrixa.

Extraction Method: Principal Component Analysis.Rotation Method: Varimax with Kaiser Normalization.a

a. Rotation converged in 5 iterations.

Table 2Mean and Median Preferred Buffer Distance between Large-Scale Solar & Land-type (in Miles).

Residential Areas Agricultural Land Breeding Ground Cultural Area Recreation Areas Wetlands Existing Solar

Mean (X̄) Median(M) Mean(X̄) Median(M) Mean(X̄) Median(M) Mean(X̄) Median(M) Mean(X̄) Median(M) Mean(X̄) Median(M) Mean(X̄) Median(M)21.72 5.00 19.85 5.00 37.15 10.00 27.58 5.00 33.82 5.00 35.38 6.08 31.77 1.00

istanc

rpam

4

fb

Fig. 2. Preferred Buffered D

easons for opposition and that the use of the term be used lessejoratively. Opposition to particular proposed facilities is not justbout selfish motivations (e.g., my property values), but somethingore, perhaps attachment to different land types or land uses.

.2. Probit models

The purpose of our study is to understand whether supportor solar depends on a variety of contexts, including visibility anduffer distance between proposed large-scale solar and land-type

e from Various Land Types.

in Southern California. We now turn our analysis to models thathelp us explain support and preferred buffer distance. First, we con-sider support for a facility proposed generally, proposed nearby towhere the respondent lives, and also proposed nearby but withassurance it will not be visible. Using PLUM (ordinal probit), weregress many standard predictors of environmental and political

attitudes including demographic variables such as sex, age, race,education, and urbanicity, and socio-psychological measures suchas party identification and belief in the seriousness of climate

se Po

csaa

Td(ltfc(pwn

01milipTt(dlpasd

sawanpPttaaptva

ndcpdpttnlatplls

J.E. Carlisle et al. / Land U

hange (our proxy for environmentalism), on our measures of solarupport. We also include four factors capturing place attachmentnd project-related aspects such as the perceived positive and neg-tive effects of large-scale solar development.

The results from our first set of probit estimations are found inable 3. To remind readers, our dependent variables ask respon-ents about support for construction of large-scale solar in generalin the U.S.), near to where they live, as well as support forarge-scale solar nearby but not visible. Answer categories forhe dependent variables are based on a five-category Likert scalerom 1 (strongly oppose) to 5 (strongly support), which have beenollapsed and recoded into a 3-category variable, 1 (oppose); 2neither oppose nor support); and 3 (support). Thresholds (or cut-oints) demonstrate where on the dependent variable an individualill transition from one category of the variable to the next (oppose,

either oppose nor support, and support).Several variables demonstrate statistical significance at the

.05 � level. In terms of support for solar in the U.S. (Model), our results show that education, belief in seriousness of cli-ate change, rural, positive impacts—community, and positive

mpacts—personal (these latter two measures capturing the under-ying positive impacts from building large-scale solar in terms ofndustrialization, jobs, bringing life to the place, and increasedroperty values), are all significant and in the positive direction.hus, those who are more educated and more strongly believe inhe seriousness of climate change are more likely to support solarB = 0.13 and p ≤ 0.05; B = 0.17 and p ≤ 0.05, respectively). Respon-ents who are more likely to see the positive impacts resulting from

arge-scale solar development in the U.S. are more likely to sup-ort large-scale solar development (B = 0.37 and p ≤ 0.001; B = 0.44nd p ≤ 0.001, respectively). Rural dwellers are more likely thanuburbanites (the omitted category) to support large-scale solarevelopment (B = 0.43 and p ≤ 0.05)

In looking at support for large-scale solar nearby (Model 2),everal demographic variables including race and location (ruralnd urban) demonstrate strong and positive relationships. Thoseho are white/Caucasian, rural dwellers and urban dwellers

re more likely to support large-scale solar development thanot than are non-whites and suburban dwellers (B = 0.69 and

≤ 0.001; B = 0.49 and p ≤ 0.01; B = 0.04 p ≤ 0.05, respectively).arty identification, also demonstrates an expected significant rela-ionship insofar as Republicans are less likely than Democratso support the construction of large-scale solar nearby (B = 0.43nd p ≤ 0.001). Similar to Model 1, positive impacts–community,nd positive impacts—personal (again, capturing the underlyingositive impacts from building large-scale solar in terms of indus-rialization, jobs, bringing life to the place, and increased propertyalues), are both significant and in the positive direction (B = 0.40nd p ≤ 0.001; B = 0.28 and p ≤ 0.01, respectively).

We next consider support for a large solar facility being builtearby but with a guarantee it would not be visible to the respon-ent (Model 3). Our results for this model are similar to the resultsonsidering support for solar in the U.S. (Table 3, Model 1) and sup-ort for solar nearby (Table 3, Model 2). In particular, the resultsemonstrate a significant and positive relationship with race andositive impacts (community). Thus, whites and those who believehat positive impacts would result from building large-scale solar inerms of industrialization and jobs are more likely to support solarearby and solar that is not visible than are non-whites and those

ess likely to perceive positive impacts to the community (B = 0.61nd p ≤ 0.01; B = 0.35 and p ≤ 0.001, respectively). Significant rela-ionships exist for party identification and length of residency, our

roxy for place attachment. More specifically, when a proposed

arge-scale solar facility that is guaranteed to not be visible, Repub-icans and those who are more attached to place are less likely toupport its construction than are Democrats and those less attached

licy 58 (2016) 491–501 497

to place (B = −0.30 and p ≤ 0.05; B = −0.22 and p ≤ 0.01, respec-tively).

Our second set of probit models consider the impact of predic-tors on buffer distances for a variety of different land types. Again,our dependent variable is preferred buffer sizes between differ-ent land-types and proposed large-scale solar facilities, which wasmeasured by asking respondents, “How much buffer distance isacceptable between a large solar facility and [a residential area;existing agricultural land; an area used as nesting sites or breedinggrounds by wildlife; an area of cultural or historical importance;recreation areas (e.g. hunting, fishing, or hiking locations); wet-lands; an area used as a migration route by wildlife; existingsolar]?” All respondent answers were converted (when necessary)to miles and recoded as a four-category Likers scale as follows:1) Less than a mile; 2) 1–5 miles; 3) 6–10 miles; and 4) Morethan 10 miles. “Don’t know” and “No preference” were coded assystem-missing and thus omitted from the analyses. Thresholds(or cut-points) demonstrate where on the dependent variable anindividual will transition from one category of the variable to thenext.

We find that overall, many of the same predictors demonstratesignificance across several dependent variables (or land-types). Wewill highlight a few of the significant effects, the results of whichare found in Table 4. To begin, the predictor with the strongesteffect across the most models is race. Specifically, for seven of theeight models where race proves to be significant, the nature of theeffect is negative such that whites are more likely than non-whitesto prefer a smaller buffer distance from land characterized as resi-dential, agricultural, breeding grounds, cultural areas, recreationalareas, wildlife areas, and containing existing solar. Also, our mea-sures capturing place related symbolic meaning perform very wellin many of our models. Those who perceive solar construction tocreate negative impacts are more likely to prefer larger buffer dis-tances from land characterized as breeding grounds, cultural areas,recreation areas, wetlands, and existing solar. Additionally thoserespondents who perceive the construction of large scale solar toyield positive impacts for the community are more likely to prefersmaller buffer distances between the proposed solar facility andresidential areas, agricultural land, cultural areas, and recreationareas. Place attachment and the factor described as “good commu-nity” (capturing place-related symbolic meanings); however, onlydemonstrate significance in two of the models. The factor describedas “good community” has a significant and positive impact (respon-dents prefer a larger buffer distance) in terms of land characterizedas residential and recreation area. Those who are more attached toplace (e.g. lived there longer) are more likely to support a smallerbuffer distance between land characterized as a breeding groundor wildlife migration route. As well, the factor described as positiveimpacts (personal) only demonstrates a significant and negativeeffect in one model—residential areas—so that those who perceivelarge-scale solar to yield a personal positive impact are more likelyto prefer a small buffer distance. Another variable that yields astrong and significant impact across several models is sex. Thatis, males are more likely than females to prefer larger buffer dis-tances for four of the land types, including residential, agricultural,breeding grounds, and wetlands.

5. Discussion and conclusion

The purpose of this study was to test the nature of therelationship between a set of predictors including demographic,

socio-psychological and those relating to place attachment andplace-related symbolic meaning on public support and opposi-tion for large-scale solar development. In particular, we soughtto test these predictors across different scenarios including vis-

498 J.E. Carlisle et al. / Land Use Policy 58 (2016) 491–501

Table 3Probit Model of Support for Solar.

Model 1: U.S. Model 2: Nearby Model 3: Not Visible

Variable Estimate Std. Error Estimate Std. Error Estimate Std. Error

Threshold 1 −1.19 0.66 −1.67** 0.61 −2.28*** 0.66Threshold 2 −0.19 0.48 −1.00 0.61 −1.77** 0.65Sex (male vs. female) −0.04 0.18 −0.03 0.17 −0.07 0.17Age 0.09 0.07 0.10 0.06 −0.03 0.07Education 0.13* 0.07 0.06 0.06 0.12* 0.06Race (white vs. non-white) −0.03 0.22 0.69*** 0.20 0.61** 0.21Party Identification (Republican high) −0.07 0.13 −0.43*** 0.13 −0.30* 0.13Seriousness of Climate Change 0.17* 0.09 0.06 0.08 0.05 0.08Rural 0.43* 0.22 0.49** 0.20 −0.13 0.20Urban −0.22 0.24 0.04* 0.22 −0.12 0.23Length of Residency in the Area −0.05 0.08 −0.11 0.07 −0.22** 0.08Good Community −0.07 0.10 −0.06 0.09 −0.08 0.09Positive Impacts (Community) 0.37*** 0.08 0.40*** 0.08 0.35*** 0.08Positive Impacts (Personal) 0.44*** 0.12 0.28** 0.10 0.17 0.11Negative Impacts 0.07 0.09 0.11 0.08 0.05 0.09

Cox and Snell 0.15 0.18 0.14

* p ≤ 0.05.** p ≤ 0.01.

*** p ≤ 0.001.

Table 4Preferred Buffer Size between Large-scale Solar and Land-type.

Residential Areas Agricultural Land Breeding Ground Cultural Area Recreation Areas

Variable Estimate Std. Error Estimate Std. Error Estimate Std. Error Estimate Std. Error Estimate Std. Error

Threshold 1 −1.64*** 0.51 −1.39** 0.51 −1.87*** 0.51 −1.34** 0.50 −4.72*** 0.75Threshold 2 −0.29 0.51 −0.25** 0.51 −0.79 0.50 0.01 0.50 −1.72** 0.61Threshold 3 0.03 0.51 0.13 0.51 −0.27 0.50 0.50 0.50 −1.31 .61*

Sex (male vs. female) 0.31* 0.14 0.49*** 0.15 0.29* 0.14 0.24 0.14 0.22 0.17Age −0.08 0.05 −0.02 0.05 −0.02 0.05 −0.05 0.05 −0.14* 0.06Education −0.11* 0.05 −0.17*** 0.05 0.00 0.05 −0.16*** 0.05 −0.05 0.06Race (white vs. non-white) −0.52** 0.17 −0.72*** 0.17 −0.58*** 0.17 −0.37* 0.17 −0.62** 0.21Party Identification (Republican high) −0.10 0.11 −0.01 0.11 −0.06 0.10 0.01 0.10 −0.22 0.13Seriousness of Climate Change 0.09 0.07 0.02 0.07 −0.01 0.07 0.18** 0.07 −0.09 0.08Rural 0.02 0.16 0.00 0.16 −0.01 0.15 0.08 0.16 0.07 0.19Urban 0.17 0.21 −0.04 0.24 0.30 0.21 0.24 0.20 0.12 0.25Length of Residency in the Area −0.04 0.06 −0.04 0.06 −0.13* 0.06 0.00 0.06 −0.04 0.07Good Community 0.14* 0.07 0.01 0.07 0.04 0.07 0.11 0.07 0.21* 0.09Positive Impacts (Community) −0.17** 0.07 −0.14* 0.07 −0.03 0.07 −0.19** 0.07 −0.21* 0.07Positive Impacts (Personal) −0.12** 0.08 0.01 0.09 0.03 0.08 −0.12 0.08 0.12 0.11Negative Impacts 0.08 0.07 0.11 0.08 0.15* 0.07 0.14* 0.07 0.25** 0.09

Cox and Snell 0.20 0.24 0.16 0.20 0.26

Wetlands Wildlife Migration Route Existing Solar

Variable Estimate Std. Error Estimate Std. Error Estimate Std. Error

Threshold 1 −1.68*** 0.51 −1.81*** 0.51 −0.15 0.11Threshold 2 −0.47 0.51 −0.82 0.50 0.40 0.51Threshold 3 −0.07 0.51 −0.28 0.50 0.68 0.51Sex (male vs. female) 0.36** 0.14 0.23 0.14 0.28 0.14Age −0.06 0.05 −0.01 0.05 −0.04 0.05Education −0.08 0.05 −0.03 0.05 −0.13** 0.05Race (white vs. non-white) −0.24 0.17 −0.57*** 0.17 −0.69*** 0.17Party Identification (Republican high) −0.07 0.10 −0.05 0.10 0.08 0.11Seriousness of Climate Change 0.06 0.07 0.05 0.07 0.12 0.07Rural 0.04 0.16 −0.02 0.15 −0.01 0.16Urban −0.01 0.20 0.36 0.21 0.22 0.21Length of Residency in the Area −0.09 0.06 −0.14* 0.06 0.05 0.06Good Community 0.06 0.07 0.09 0.07 0.08 0.07Positive Impacts (Community) −0.09 0.07 −0.01 0.07 −0.16 0.07Positive Impacts (Personal) −0.03 0.08 0.09 0.08 0.09 0.08Negative Impacts 0.15* 0.07 0.06 0.07 0.16** 0.07

Cox and Snell 0.18 0.18 0.24

* p ≤ 0.05.** p ≤ 0.01.

*** p ≤ 0.001.

se Po

iWtimwNfaida

pitvvdltadtms1iaoWUmomedevaiehh

satjmaprtesttfpi

sbnb

J.E. Carlisle et al. / Land U

bility of proposed facility and proximity to various land-types.e utilized variables that help understand support and opposition

o proposed energy facilities in a manner that moves beyond thencreasingly criticized NIMBY explanation. As many scholars have

ore recently found, support and opposition to proposed facilities,hether energy facilities or other types, is more nuanced than theIMBY explanation allows. In this spirit, scholars have found that

acility characteristics, e.g. size and scope and visibility, can have strong effect on public support or opposition. And while NIMBYs criticized as over-simplistic, it is not to suggest that proximityoes not matter. Indeed our research demonstrates that proximitys measured by buffer distances certainly matter!

Overall, Southern California residents are overwhelming sup-ortive of large-scale solar under a variety of circumstances,

ncluding generally (in the U.S.) and locally. Moreover, a guaran-eed lack visibility of a proposed nearby large-scale solar facilityersus one where respondents were not guaranteed of lack ofisibility does alter public support. Also, we find that respon-ents preferred different sized buffer distances between a proposed

arge-scale solar facility and various types of land. For example,he preferred buffer distance between a proposed solar facility andn existing one is less than 1 mile whereas, the preferred bufferistance between a proposed large-scale solar facility and residen-ial areas, cultural areas, and recreation areas is between 1 and 5

iles. And, the preferred buffer distance between a proposed large-cale facility and wildlife migration route and breed grounds is1 miles or more. Clearly, respondents are concerned about the

mpact of solar on wildlife and this finding should serve as reassur-nce to wildlife experts who have sought to understand the impactf solar development to the natural habitat, especially wildlife.hile the potential for solar development in the Southwestern

nited States is great, the potential harm to the natural environ-ent is also great given its large biodiversity and precariousness

f the desert ecosystem due to impacts already resulting from cli-ate and human causes (Lovich and Bainbridge, 1999; Mittermeier

t al. 2002; Randall et al., 2010; CBI, 2010). Additionally, in theesert Southwest there exist many “hotspots” for threatened andndangered species (Flather et al., 1998). Understanding how indi-iduals who live in these areas where solar is currently being builtnd/or proposed think about land, buffers, and about the perceivedmpacts of these solar facilities on natural resources is certainly rel-vant to stakeholders. It now appears that wildlife experts at leastave the support of the public in protecting the wildlife areas andabitats.

In terms of support for solar and preferred buffer distances,ome of the strongest predictors in our research are those thatre project-related. For example, we explored the factors that cap-ure the perceived impacts, both positive and negative, such asobs, too much traffic, or increased/decreased property values that

ight result from building a large-scale solar facility nearby, as wells attitudes about the procedural justice of the decision-makingrocess. Our findings demonstrate the perceived positive impactsesulting from large-scale solar development are positively relatedo support for building a large-scale solar facility nearby. Also, sev-ral demographic variables (sex and race) consistently demonstrateignificance across a number of our models, while age and educa-ion demonstrate significance less frequently. These results seemo reaffirm the positive impacts some individuals believe will resultrom the development of such an infrastructure and how perceivedositive benefits stemming from a proposed large-scale solar facil-

ty seem to lead people to find smaller buffer distances acceptable.In terms of opposition to large-scale solar development, the

trongest predictor is party identification. However, in terms ofuffer distances, party identification failed to reach statistical sig-ificance. What we find is that buffer distance is driven mostlyy race and sex and our factors capturing place related symbolic

licy 58 (2016) 491–501 499

meaning. Other variables also demonstrate significance, but notconsistently across the various land-types. Scholars of public opin-ion have long been aware of the strong and significant role thatpredispositions such as values, worldviews, political ideology, andpartisan identification have on opinions. More specifically, parti-san identification in one of the key predictors of whether or notone believes in the seriousness of climate change (McCright andDunlap, 2011). Here, our findings do not entirely support this logic.In terms of solar development in Southern California, it appears tobe a partisan issue but in terms of buffer distance, both race and sexas well as perceptions of change really seem to be driving supportor opposition.

The relationship between race and support for solar could possi-bly be the result of two similar and well-documented relationshipsbetween race and environmental concern. One perspective, draw-ing on Maslows (1970) hierarchy of needs psychological model,hypothesizes that minorities (Because they are disproportion-ately poor), like the lower classes, are more focused on strugglingto address their more basic physiological and economic needsrather than spending time being concerned about higher order,socio-psychological needs such as a clean environment. There-fore, whites, more so than blacks, will exhibit pro-environmentalattitudes and behaviors. In the context of our study we see thisrelationship illustrated insofar as whites are more likely to supportsolar development than are non-whites.

Another perspective considers how the 1970’s energy crisis andthe growing threat of smog, nuclear risks, industrial chemical pollu-tion and hazardous waters gave rise to the modern environmentalmovement (Martinez-Alier, 2002). It is argued that since theseenvironmental threats are more prevalent in poor and minoritycommunities one can observe what has been termed “environ-mentalism of the poor.” So, rather than whites being the onlyenvironmentally interested and concerned citizens, blacks andother minorities too have demonstrated, as a result of the envi-ronmental threats in their own backyards, environmental attitudesand behaviors on par with their white counterparts and their con-cern has been witnessed by the growing strength of the of thegrassroots “environmental justice movement” (Austin and Schill,1994; Bullard, 1994; Taylor, 1992). Mohai and Bryant (1998) foundthat blacks are as concerned about the same kinds of environ-mental issues as are whites and that blacks were also concernedabout preservation issues about which it has long been thoughtthat blacks were unconcerned. Moreover, their study revealed thatblacks are actually more concerned than whites on local environ-mental issues, which supports the relative deprivation explanation,thereby discounting Maslow’s hierarchy of needs and the culturalbackground arguments. In the context of our study, our resultssupport this perspective in that non-whites are more likely thanwhites to act in a place protective manner and thereby supportgreater buffer distance between large-scale solar construction anddifferent types of land (recreation areas, breeding grounds, and thelike). While our results on race and support for solar and supportfor buffer distance are seemingly contradictory this is hopefullysomething that additional research can sort out.

One limitation of this study and the literature in general isthat there is a lack of understanding in regard to the accuracy ofpeople’s distance perceptions in the abstract, versus objectivelygauging visual or topographic distances beyond limited spaces suchas neighborhoods. Existing research on distances and perceptiontends to focus on personal orientation and neuropsychology andneuroimagery within space (Lee and Tversky, 2005). These includestudies on nearness and environmental space between landmarks

in small areas such as campuses (Worboys et al., 2004), and trans-portation and examining linear distance through walking, driving,or commuting from place to place, generally finding that the moreturns and transfers of modes elongates people’s perception of dis-

5 Use Po

tfthpeto

sdtp(fasofarioUsslospsewiaui

R

A

A

A

B

B

B

B

B

B

(

C

C

00 J.E. Carlisle et al. / Land

ance versus actual distance traveled (Montello, 1997). An area forurther study might include experiments with subjects to assesshe ability to estimate distances visually and to understand betterow people translate distance to a personal belief about buffers andarticular technologies. Based on our present study, policy mak-rs, solar developers, and the public should pay close attentiono preferred buffer distances as proxies for intensity of potentialpposition or preference in regard to proposed solar sites.

Another limitation to our study is that we consider large-scaleolar in a general sense rather than focus on any specific solarevelopment project, proposed, under construction, or in opera-ion. Research finds that public opposition tends to be highest whenrojects are proposed and then ebbs once construction is completedWarren et al., 2005). However, we believe this limitation to beairly minor because we are trying to understand public attitudesbout solar development generally rather than any relation to anypecific site in an area where the probability of some solar devel-pment is high and not hypothetical. Moreover, many of the solaracilities that are proposed, under development, or already oper-ting in this region are small-scale (e.g. not utility-scale) and, as aesult, many people will not have direct knowledge of these specificnstallations by name. However, what they should have knowledgef is that solar development in Southern California (and perhaps the.S. Southwest) is imminent. Perhaps by considering and discussing

pecific solar developments with respondents in a more local casetudy, we might see similar effects. A final limitation might be theack of control for solar penetration. That is, certainly some areasf California and the Southwest, in general, have experienced moreolar development than others. By controlling for degree of solarenetration we may find that support and the predictors of thatupport vary as well. We hope to explore this in a future paper, how-ver. Nevertheless, we believe our most noteworthy contributionith this research is that we consider large-scale solar development

n terms of different geo-spatial aspects including buffer distancesnd visibility. It is our hope that our findings are helpful to others innderstanding and rethinking the effect and importance of prox-

mity and will useful to policymakers, developers, and planners.

eferences

hlbrandt Jr., R.S., 1984. Neighborhoods, People and Community. Plenum, NewYork.

nsolabehere, S., Konisky, D.M., 2009. Public attitudes toward construction of newpower plants. Public Opin. Q. 73 (3), 566–577.

nsolabehere, S., 2007. Public Attitudes Toward America’s Energy Options:Insights for Nuclear Energy. MIT Center for Advanced Nuclear Energy Systems,MIT-NES-TR-008 (accessed 17.08.16.) http://ceepr.mit.edu/files/papers/2007-002.pdf.

ell, D., Gray, T., Haggett, C., 2005. The ‘social gap’ in wind farm siting decisions:explanations and policy responses. Environ. Politics 14 (4), 460–477.

raunholtz, Simon, 2003. Public Attitudes to Windfarms: A Survey of LocalResidents in Scotland. http://www.scotland.gov.uk/Resource/Doc/47133/0014639.pdf.

reukers, S., Wolsink, M., 2007. Wind power implementation in changinginstitutional landscapes: an international comparison. Energy Policy 35 (5),2737–2750.

rittan, G.G., 2002. A view from lake como. In: Pasqualetti, M., Gipe, P., Righter, R.(Eds.), Wind Power in Vie. Academic Press, San Diego, CA, pp. 215–221.

rody, C.J., 1984. Differences by sex in support for nuclear power. Soc. Forces 63(1), 209–228.

rown, Barbara B., Perkins, Douglas, 1992. Disruptions in place attachment. In:Altman, I., Low, S. (Eds.), Place Attachment. Plenum Press, New York, pp.279–304.

CBI) Conservation Biology Institute, 2010. Recommendations of IndependentScience Advisors for the California Desert Renewable Energy Conservation Plan(DRECP), CBI. (6 July 2011; www.energy.ca.gov/2010publications/DRECP-1000-2010-008/DRECP-1000-2010-008-F.PDF).

arlisle, J.E., Kane, S.L., Solan, D., Joe, J.C., 2014. Support for solar energy: examining

sense of place and utility-scale development in California. Energy Res. Soc. Sci.3, 124–130.

arlisle, J.E., Kane, S.L., Solan, D., Bowman, M., Joe, J.C., 2015. Public attitudesregarding large-scale solar energy development in the US. Renew. Sustain.Energy Rev. 48, 835–847.

licy 58 (2016) 491–501

Cart, J., 2012. U.S. Solar Plan Creates Energy Zones, Excludes Sensitive Lands. LosAngeles Times. July 25 (accessed 01.04.13.). http://articles.latimes.com/2012/jul/25/local/la-me-0725-solar-zones-20120725.

Corner, A., Venables, D., Spence, A., Poortinga, W., Demski, C., Pidgeon, N., 2011.Nuclear power, climate change and energy security: exploring British publicattitudes. Energy Policy 39 (9), 4823–4833.

Cotton, M., Devine-Wright, P., 2011. NIMBYism and community consultation inelectricity transmission network planning. In: Devine-Wright, P. (Ed.),Renewable Energy and the Public: From NIMBY to Participation. Earthscan,London, pp. 115–130.

Department of Trade and Industry, Scottish Executive et al. (2003). Attitudes andKnowledge of Renewable Energy amongst the General Public: Report Findings.

Devine-Wright, P., 2005. Beyond NIMBYism: towards an integrated framework forunderstanding public perceptions of wind energy. Wind Energy 8 (2), 125–139.

Devine-Wright, P., 2008. Reconsidering Public Acceptance of Renewable EnergyTechnologies: A Critical Review. Delivering a Low Carbon Electricity System:Technologies, Economics and Policy. Cambridge University Press, Cambridge,pp. 443–461.

Energy Information Administration, 2012. Annual Energy Outlook 2012 withProjections to 2035 (accessed 04.19.13.). http://www.eia.gov/forecasts/archive/aeo12/pdf/0383(2012).pdf.

Energy Information Administration, 2015. California First State to Generate Morethan 5% of Electricity from Utility-scale Solar (accessed 04.09.15.). http://www.eia.gov/todayinenergy/detail.cfm?id=20492.

Farhar, B.C., 2003. Trends in Public Perceptions and Preferences on Energy andEnvironmental Policy: Executive Summary. National Renewable EnergyLaboratory (accessed 01.08.13.). http://www.nrel.gov/docs/legosti/old/4857a.pdf.

Firestone, J., Kempton, W., 2007. Public opinion about large offshore wind power:underlying factors. Energy Policy 35 (3), 1584–1598.

Flather, C.H., Knowles, M.S., Kendall, I.A., 1998. Threatened and endangered speciesgeography. Bioscience 48 (5), 365–376.

Gallup, 2013. Americans Want More Emphasis on Solar, Wind, Natural Gas(accessed 10.01.13.). http://www.gallup.com/poll/161519/americans-emphasis-solar-wind-natural-gas.aspx.

Greenberg, M., 2009. Energy sources, public policy, and public preferences:analysis of US National and site-specific data. Energy Policy 37 (8), 3242–3249.

Gross, C., 2007. Community perspectives of wind energy in Australia: theapplication of a justice and community fairness framework to increase socialacceptance. Energy Policy 35 (5), 2727–2736.

Jones, C.R., Eiser, R.J., 2010. Understanding ‘local’ opposition to wind developmentin the UK: how big is a backyard? Energy Policy 38 (6), 3106–3117.

Klick, H., Smith, E.R.A.N., 2010. Public understanding of and support for windpower in the United States. Renew. Energy 35 (7), 1585–1591.

Krohn, S., Damborg, S., 1999. On public attitudes towards wind power. Renew.Energy 16 (1), 954–960.

Ladenburg, J., 2008. Attitudes towards on-land and offshore wind powerdevelopment in Denmark; choice of development strategy. Renew. Energy 33(1), 111–118.

Ladenburg, J., 2009. Visual impact assessment of offshore wind farms and priorexperience. Appl. Energy 86 (3), 380–387.

Lee, P., Tversky, B., 2005. Interplay between visual and spatial: the effect oflandmark descriptions on comprehension of route/survey spatial descriptions.Spat. Cognit. Comput. 5 (2–3), 163–185.

Lovich, J.E., Bainbridge, D., 1999. Anthropogenic degradation of the southernCalifornia desert ecosystem and prospects for natural recovery and restoration.Environ. Manag. 24 (3), 309–326.

MORI Social Research Institute for Regen South West, 2003. Public AttitudesTowards Renewable Energy in the South West (accessed 01.04.13.). http://www.regensw.co.uk/content-download/REPORTMORIPublicAttitudes.pdf.

Manzo, L.C., 2003. Beyond home and haven: toward a revisioning of emotionalrelationships with places. J. Environ. Psychol. 23 (1), 47–61.

Manzo, L.C., 2005. For better or for worse: exploring multiple dimensions of placemeaning? J. Environ. Psychol. 25 (1), 67–86.

McCright, A.M., Dunlap, R.E., 2011. The politicization of climate change andpolarization in the American public’s views of global warming. 2001–2010.Soc. Q. 52 (2), 155–194.

McGowan, F., Sauter, R., 2005. Public Opinion on Energy Research: A Desk Study forthe Research Councils. Sussex Energy Group (accessed 01.18.3.). http://www.epsrc.ac.uk/SiteCollectionDocuments/Publications/reports/EnergyAttitudesDeskStudySussex2005.pdf.

Michaud, K.E.H., Carlisle, J.E., Smith, E.R.A.N., 2008. Nimbyism vs.environmentalism in energy development attitudes. Environ. Politics 17,20–39.

Mittermeier, R.A.M., Pilgrim, C.G., Fonseca, J., Konstant, G., William, R., 2002.Wilderness: Earth’s Last Wild Places (No. 333.782 W673w). CEMEX, México,MX.

Mohai, P., 1992. Men, women, and the environment: an examination of the gendergap in environmental concern and activism. Soc. Nat. Resour. 5 (1), 1–19.

Molnarova, K., Sklenicka, P., Stiborek, J., Svobodova, K., Salek, M., Brabec, E., 2012.Visual preferences for wind turbines: location: numbers and respondent

characteristics. Appl. Energy 92, 269–278.

Montello, D.R., (1997). The perception and cognition of environmental distance:direct sources of information. In S.C. Hirtle & A.U. Frank (eds.), Spatialinformation theory: a theoretical basis for GIS (pp. 297–311). Proceedings ofCOSIT ’97. Berlin: Springer-Verlag.

se Po

O

P

P

P

R

R

S

S

S

T

J.E. Carlisle et al. / Land U

ttman, J.A., Herbert, H., 1993. Green Marketing. NTC Business Books,Lincolnwood, IL.

hadke, R., 2010. Steel forests or smoke stacks: the politics of visualisation in theCape Wind controversy. Environ. Politics 19 (1), 1–20.

opulus, 2005. Energy Balance of Power Poll (accessed 01.08.3.).www.populuslimited.com/pdf/2005 07 01 times nuclear.pdf.

roshansky, H.M., Fabian, A., Kaminoff, R., 1983. Place-identity: physical worldsocialization of the self? J. Environ. Psychol. 3 (1), 57–83.

andall, J.M., Parker, S.S., Moore, J., Cohen, B., Crane, L., Christian, B., Cameron, D.,MacKenzie, J., Klausmeyer, K., Morrison, S., 2010. Mojave Desert EcoregionalAssessment. Unpublished Report. The Nature Conservancy, San Francisco,California. 106 pages + appendices. Available at: http://conserveonline.org/workspaces/mojave/documents/mojave-desertecoregional-2010/@@view.html (accessed August).

iger, S., Lavrakas, P.J., 1981. Community ties: patterns of attachment and socialinter-actions in urban neighborhoods. Am. J. Community Psychol. 9 (1), 55–66.

mith, E.R.A.N., 2002. Energy, the Environment, and Public Opinion. Rowman &Littlefield Publishers Incorporated, Lanham, MD.

ovacool, B.K., 2009. Exploring and contextualizing public opposition to renewableelectricity in the United States. Sustainability 1, 702–721.

wofford, J., Slattery, M., 2010. Public attitudes of wind energy in Texas: local

communities in close proximity to wind farms and their effect ondecision-making. Energy Policy 38 (5), 2508–2519.

aylor, R.B., Shumaker, S.A., Gottfredson, S.D., 1985. Neighborhood-level linksbetween physical features and local sentiments: deterioration, fear of crime,and confidence. J. Archit. Plann. Res. 2 (4), 261–275.

licy 58 (2016) 491–501 501

Taylor, R.B., 1996. Neighborhood responses to disorder and local attachments: thesystemic model of attachment social disorganization, and neighborhood usevalue. Sociol. Forum 11 (1), 41–74.

Van der Horst, D., 2007. NIMBY or not? Exploring the relevance of location and thepolitics of voiced opinions in renewable energy siting controversies. EnergyPolicy 35 (5), 2705–2714.

Vorkinn, M., Riese, H., 2001. Environmental concern in a local context: thesignificance of place attachment. Environ. Behav. 33 (2), 249–263.

Wüstenhagen, R., Wolsink, M., Bürer, M.J., 2007. Social acceptance of renewableenergy innovation: an introduction to the concept. Energy Policy 35 (5),2683–2691.

Walker, G., 1995. Renewable energy and the public. Land Use Policy 12 (1), 49–59.Warren, C.R., McFadyen, M., 2010. Does community ownership affect public

attitudes to wind energy? A case study from south-west Scotland. Land UsePolicy 27 (2), 204–213.

Warren, C.R., Lumsden, C., O’Dowd, S., Birnie, R.V., 2005. ‘Green on green’: publicperceptions of wind power in scotland and Ireland. J. Environ. Plann. Manag.48 (6), 853–875.

Wolsink, M., 2000. Wind power and the NIMBY-myth: institutional capacity andthe limited significance of public support. Renew. Energy 21 (1), 49–64.

Wolsink, M., 2007. Wind power implementation: the nature of public attitudes:equity and fairness instead of ‘backyard motives’. Renew. Sustain. Energy Rev.

11 (6), 1188–1207.

Worboys, M., Duckham, M., Kulik, L., 2004. Commonsense notions of proximityand direction in environmental space. Spat. Cognit. Comput. 4 (4), 285–312.

Zarnikau, J., 2003. Consumer demand for ‘green power’ and energy efficiency.Energy Policy 31 (15), 1661–1672.